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.
158 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
159 the right size to place the terminating semicolon on its own folded
160 line, the header hash was calculated to an incorrect value thanks to
161 the (relaxed) space the fold became.
167 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
168 than -2003 did; needs libidn2 in addition to libidn.
170 JH/02 The path option on a pipe transport is now expanded before use.
172 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
173 Patch provided by "Björn", documentation fix added too.
175 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
176 missing a wire-to-host endian conversion.
178 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
179 close after a BDAT command line could be taken as a following command,
180 giving a synch failure. Fix by only checking for synch immediately
181 before acknowledging the chunk.
183 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
184 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
185 macro. Patches provided by Josh Soref.
187 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
188 Previously we did not; the RFC seems ambiguous and VRFY is not listed
189 by IANA as a service extension. However, John Klensin suggests that we
192 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
193 the dkim code may be unix-mode line endings rather than smtp wire-format
194 CRLF, so prepend a CR to any bare LF.
196 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
197 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
199 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
200 the first were themselves being wrongly included in the feed into dkim
201 processing; with most chunk sizes in use this resulted in an incorrect
202 body hash calculated value.
204 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
205 DKIM signature block, for verification. Although advised against by
206 standards it is specifically not ruled illegal.
208 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
210 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
211 missing a body hash (the bh= tag).
213 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
214 It seems that HAProxy sends the Proxy Protocol information in clear and
215 only then does a TLS startup, so do the same.
217 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
218 TCP connections (such as for Spamd) unless the daemon successfully set
219 Fast Open mode on its listening sockets. This fixes breakage seen on
220 too-old kernels or those not configured for Fast Open, at the cost of
221 requiring both directions being enabled for TFO, and TFO never being used
222 by non-daemon-related Exim processes.
224 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
225 endings, at least on the first header line. Try to canonify any that get
226 past that check, despite the cost.
228 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
229 now limited to an arbitrary five deep, while parsing addresses with the
230 strip_excess_angle_brackets option enabled.
232 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
233 instead leave the unprompted TLS handshake in socket buffer for the
234 TLS library to consume.
236 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
238 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
240 JH/16 Drop variables when they go out of scope. Memory management drops a whole
241 region in one operation, for speed, and this leaves assigned pointers
242 dangling. Add checks run only under the testsuite which checks all
243 variables at a store-reset and panics on a dangling pointer; add code
244 explicitly nulling out all the variables discovered. Fixes one known
245 bug: a transport crash, where a dangling pointer for $sending_ip_address
246 originally assigned in a verify callout, is re-used.
248 PP/06 Drop '.' from @INC in various Perl scripts.
250 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
252 PP/08 Reduce a number of compilation warnings under clang; building with
253 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
254 should be warning-free.
256 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
258 HS/01 Fix portability problems introduced by PP/08 for platforms where
259 realloc(NULL) is not equivalent to malloc() [SunOS et al].
261 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
262 chunk. This allows us to accept broken chunked messages. We need a more
263 general solution here.
265 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
266 already-broken messages in the queue.
268 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
270 JH/19 Fix reference counting bug in routing-generated-address tracking.
276 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
277 supports it and a size is available (ie. the sending peer gave us one).
279 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
280 years of being deprecated). The replacements are the ACLs
281 acl_smtp_mime and acl_not_smtp_mime.
283 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
284 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
285 or even an in-clear connection were permitted. Now, if the host lookup
286 was dnssec and dane was requested then the host is only used if the
287 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
288 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
289 if one fails this test.
290 This means that a poorly-configured remote DNS will make it incommunicado;
291 but it protects against a DNS-interception attack on it.
293 JH/04 Bug 1810: make continued-use of an open smtp transport connection
294 non-noisy when a race steals the message being considered.
296 JH/05 If main configuration option tls_certificate is unset, generate a
297 self-signed certificate for inbound TLS connections.
299 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
300 in rewrites and routers.
302 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
303 and logged a warning sing 4.83; now they are a configuration file error.
305 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
306 (lacking @domain). Apply the same qualification processing as RCPT.
308 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
310 JH/10 Support ${sha256:} applied to a string (as well as the previous
313 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
314 a cutthrough deliver is pending, as we always want to make a connection.
315 This also avoids re-routing the message when later placing the cutthrough
316 connection after a verify cache hit.
317 Do not update it with the verify result either.
319 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
320 when routing results in more than one destination address.
322 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
323 signing (which inhibits the cutthrough capability). Previously only
324 the presence of an option was tested; now an expansion evaluating as
325 empty is permissible (obviously it should depend only on data available
326 when the cutthrough connection is made).
328 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
329 the relevant preceding SMTP command did not note the pipelining mode.
331 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
332 Previously they were not counted.
334 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
335 as one having no matching records. Previously we deferred the message
336 that needed the lookup.
338 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
339 distinguished as "(=".
341 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
342 for missing MX records. Previously it only worked for missing A records.
344 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
346 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
347 after the data-go-ahead and data-ack. Patch from Jason Betts.
349 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
350 even for a "none" policy. Patch from Tony Meyer.
352 JH/22 Fix continued use of a connection for further deliveries. If a port was
353 specified by a router, it must also match for the delivery to be
356 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
357 When one of the recipients of a message was unsuitable for the connection
358 (has no matching addresses), we lost track of needing to mark it
359 deferred. As a result mail would be lost.
361 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
363 JH/25 Decoding ACL controls is now done using a binary search; the source code
364 takes up less space and should be simpler to maintain. Merge the ACL
365 condition decode tables also, with similar effect.
367 JH/26 Fix problem with one_time used on a redirect router which returned the
368 parent address unchanged. A retry would see the parent address marked as
369 delivered, so not attempt the (identical) child. As a result mail would
372 JH/27 Fix a possible security hole, wherein a process operating with the Exim
373 UID can gain a root shell. Credit to http://www.halfdog.net/ for
374 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
377 JH/28 Enable {spool,log} filesystem space and inode checks as default.
378 Main config options check_{log,spool}_{inodes,space} are now
379 100 inodes, 10MB unless set otherwise in the configuration.
381 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
382 Previously it only applied to the main-section connection policy
385 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
387 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
388 by me. Added RFC7919 DH primes as an alternative.
390 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
391 are not in the system include path.
393 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
394 GnuTLS, when a session startup failed (eg because the client disconnected)
395 Exim did stdio operations after fclose. This was exposed by a recent
396 change which nulled out the file handle after the fclose.
398 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
399 signed directly by the cert-signing cert, rather than an intermediate
400 OCSP-signing cert. This is the model used by LetsEncrypt.
402 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
404 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
405 an incoming connection.
407 HS/02 Bug 1802: Do not half-close the connection after sending a request
410 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
411 fallback to "prime256v1".
413 JH/34 SECURITY: Use proper copy of DATA command in error message.
414 Could leak key material. Remotely exploitable. CVE-2016-9963.
420 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
421 and 3.4.4 - once the server is enabled to respond to an OCSP request
422 it does even when not requested, resulting in a stapling non-aware
423 client dropping the TLS connection.
425 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
426 support variable-length bit vectors. No functional change.
428 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
429 The I= interface field on outgoing lines is now after the H= remote
430 host field, same as incoming lines. There is a separate
431 outgoing_interface log selector which allows you to disable the
434 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
435 If not running log_selector +smtp_connection the mainlog would be held
436 open indefinitely after a "too many connections" event, including to a
437 deleted file after a log rotate. Leave the per net connection logging
438 leaving it open for efficiency as that will be quickly detected by the
439 check on the next write.
441 HS/01 Bug 1671: Fix post transport crash.
442 Processing the wait-<transport> messages could crash the delivery
443 process if the message IDs didn't exist for some reason. When
444 using 'split_spool_directory=yes' the construction of the spool
445 file name failed already, exposing the same netto behaviour.
447 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
448 mime_regex ACL conditions.
450 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
451 to DSN fail messages (bounces): remote IP, remote greeting, remote response
452 to HELO, local diagnostic string.
454 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
455 log line to debug. Even when configured with a tls authenticator many
456 client connections are expected to not authenticate in this way, so
457 an authenticate fail is not an error.
459 HS/02 Add the Exim version string to the process info. This way exiwhat
460 gives some more detail about the running daemon.
462 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
463 matter for fast-change records such as DNSBLs.
465 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
466 retry record, even if constant. There may be multiple transports with
467 different interface settings and the retry behaviour needs to be kept
470 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
472 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
474 JH/10 Bug 840: fix log_defer_output option of pipe transport
476 JH/11 Bug 830: use same host for all RCPTS of a message, even under
477 hosts_randomize. This matters a lot when combined with mua_wrapper.
479 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
480 ${quote_pgsql:<string>} operator.
482 JH/13 Bug 1708: avoid misaligned access in cached lookup.
484 JH/14 Change header file name for freeradius-client. Relevant if compiling
485 with Radius support; from the Gentoo tree and checked under Fedora.
487 JH/15 Bug 1712: Introduce $prdr_requested flag variable
489 JH/16 Bug 1714: Permit an empty string as expansion result for transport
490 option transport_filter, meaning no filtering.
492 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
494 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
495 defaults to "*" (all hosts). The variable is now available when not built
496 with TLS, default unset, mainly to enable keeping the testsuite sane.
497 If a server certificate is not supplied (via tls_certificate) an error is
498 logged, and clients will find TLS connections fail on startup. Presumably
499 they will retry in-clear.
500 Packagers of Exim are strongly encouraged to create a server certificate
501 at installation time.
503 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
504 with the $config_file variable.
506 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
507 in transport context, after the attempt, and per-recipient. The latter type
508 is per host attempted. The event data is the error message, and the errno
509 information encodes the lookup type (A vs. MX) used for the (first) host,
510 and the trailing two digits of the smtp 4xx response.
512 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
513 to write to mainlog (or rejectlog, paniclog) in the window between file
514 creation and permissions/ownership being changed. Particularly affects
515 installations where exicyclog is run as root, rather than exim user;
516 result is that the running daemon panics and dies.
518 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
520 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
521 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
522 "pri" and "weight". Note that the previous implicit priority given by the
523 list order is no longer honoured.
525 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
528 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
529 by defining SUPPORT_SOCKS.
531 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
532 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
533 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
534 variables are renamed to proxy_{local,external}_{address,port}.
536 JH/27 Move Internationalisation support from Experimental to mainline, enabled
537 for a build by defining SUPPORT_I18N
539 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
540 of the query string, and make ${quote_redis:} do that quoting.
542 JH/29 Move Events support from Experimental to mainline, enabled by default
543 and removable for a build by defining DISABLE_EVENT.
545 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
547 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
548 cached by the daemon.
550 JH/32 Move Redis support from Experimental to mainline, enabled for a build
551 by defining LOOKUP_REDIS. The libhiredis library is required.
553 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
554 keys are given for lookup.
556 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
557 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
558 only supported when built with TLS support. The PolarSSL SHA routines
559 are still used when the TLS library is too old for convenient support.
561 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
562 openssl_options), for security. OpenSSL forces this from version 1.1.0
563 server-side so match that on older versions.
565 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
566 allocation for $value could be released as the expansion processing
567 concluded, but leaving the global pointer active for it.
569 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
570 and to use the domains and local_parts ACL conditions.
572 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
573 incorrectly not doubled on cutthrough transmission, hence seen as a
574 body-termination at the receiving system - resulting in truncated mails.
575 Commonly the sender saw a TCP-level error, and retransmitted the message
576 via the normal store-and-forward channel. This could result in duplicates
577 received - but deduplicating mailstores were liable to retain only the
578 initial truncated version.
580 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
582 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
584 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
585 we're in there, support oversigning also; bug 1309.
587 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
589 HS/04 Add support for keep_environment and add_environment options.
591 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
592 either intentional arithmetic overflow during PRNG, or testing config-
595 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
596 delivery resulted in actual delivery. Cancel cutthrough before DATA
599 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
600 rejects a recipient: pass the reject to the originator.
602 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
603 Many were false-positives and ignorable, but it's worth fixing the
606 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
607 for the new environment-manipulation done at startup. Move the routines
608 from being local to tls.c to being global via the os.c file.
610 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
611 an extract embedded as result-arg for a map, the first arg for extract
612 is unavailable so we cannot tell if this is a numbered or keyed
613 extraction. Accept either.
619 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
622 JH/02 The smtp transport option "multi_domain" is now expanded.
624 JH/03 The smtp transport now requests PRDR by default, if the server offers
627 JH/04 Certificate name checking on server certificates, when exim is a client,
628 is now done by default. The transport option tls_verify_cert_hostnames
629 can be used to disable this per-host. The build option
630 EXPERIMENTAL_CERTNAMES is withdrawn.
632 JH/05 The value of the tls_verify_certificates smtp transport and main options
633 default to the word "system" to access the system default CA bundle.
634 For GnuTLS, only version 3.0.20 or later.
636 JH/06 Verification of the server certificate for a TLS connection is now tried
637 (but not required) by default. The verification status is now logged by
638 default, for both outbound TLS and client-certificate supplying inbound
641 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
644 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
645 Status Notification (bounce) messages are now MIME format per RFC 3464.
646 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
647 under the control of the dsn_advertise_hosts option, and routers may
648 have a dsn_lasthop option.
650 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
651 default, modifiable by a malware= option. The list separator for
652 the options can now be changed in the usual way. Bug 68.
654 JH/10 The smtp_receive_timeout main option is now expanded before use.
656 JH/11 The incoming_interface log option now also enables logging of the
657 local interface on delivery outgoing connections.
659 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
660 if the interface and destination host and port all match.
662 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
665 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
666 Patch from Andrew Lewis.
668 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
669 now supports optional time-restrictions, weighting, and priority
670 modifiers per server. Patch originally by <rommer@active.by>.
672 JH/16 The spamd_address main option now supports a mixed list of local
673 and remote servers. Remote servers can be IPv6 addresses, and
674 specify a port-range.
676 JH/17 Bug 68: The spamd_address main option now supports an optional
677 timeout value per server.
679 JH/18 Bug 1581: Router and transport options headers_add/remove can
680 now have the list separator specified.
682 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
685 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
688 JH/21 Support for the A6 type of dns record is withdrawn.
690 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
691 rather than the verbs used.
693 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
694 from 255 to 1024 chars.
696 JH/24 Verification callouts now attempt to use TLS by default.
698 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
699 are generic router options now. The defaults didn't change.
701 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
702 Original patch from Alexander Shikoff, worked over by JH.
704 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
705 files if ZCAT_COMMAND is not executable.
707 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
709 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
711 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
712 Normally benign, it bites when the pair was led to by a CNAME;
713 modern usage is to not canonicalize the domain to a CNAME target
714 (and we were inconsistent anyway for A-only vs AAAA+A).
716 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
718 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
719 when evaluating $sender_host_dnssec.
721 JH/31 Check the HELO verification lookup for DNSSEC, adding new
722 $sender_helo_dnssec variable.
724 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
726 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
728 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
730 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
731 documented as working, but never had. Support all but $spam_report.
733 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
734 added for tls authenticator.
736 HS/03 Add perl_taintmode main config option
742 TL/01 When running the test suite, the README says that variables such as
743 no_msglog_check are global and can be placed anywhere in a specific
744 test's script, however it was observed that placement needed to be near
745 the beginning for it to behave that way. Changed the runtest perl
746 script to read through the entire script once to detect and set these
747 variables, reset to the beginning of the script, and then run through
748 the script parsing/test process like normal.
750 TL/02 The BSD's have an arc4random API. One of the functions to induce
751 adding randomness was arc4random_stir(), but it has been removed in
752 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
753 function when detected.
755 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
756 cause callback expansion.
758 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
759 syntax errors in an expansion can be treated as a string instead of
760 logging or causing an error, due to the internal use of bool_lax
761 instead of bool when processing it.
763 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
764 server certificates when making smtp deliveries.
766 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
768 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
770 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
772 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
773 Merged patch from Sebastian Wiedenroth.
775 JH/05 Fix results-pipe from transport process. Several recipients, combined
776 with certificate use, exposed issues where response data items split
777 over buffer boundaries were not parsed properly. This eventually
778 resulted in duplicates being sent. This issue only became common enough
779 to notice due to the introduction of connection certificate information,
780 the item size being so much larger. Found and fixed by Wolfgang Breyha.
782 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
783 size buffer was used, resulting in syntax errors when an expansion
786 JH/07 Add support for directories of certificates when compiled with a GnuTLS
787 version 3.3.6 or later.
789 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
790 is EXPERIMENTAL_EVENT, the main-configuration and transport options
791 both become "event_action", the variables become $event_name, $event_data
792 and $event_defer_errno. There is a new variable $verify_mode, usable in
793 routers, transports and related events. The tls:cert event is now also
794 raised for inbound connections, if the main configuration event_action
797 TL/06 In test suite, disable OCSP for old versions of openssl which contained
798 early OCSP support, but no stapling (appears to be less than 1.0.0).
800 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
801 server certificate names available under the smtp transport option
802 "tls_verify_cert_hostname" now do not permit multi-component wildcard
805 JH/10 Time-related extraction expansions from certificates now use the main
806 option "timezone" setting for output formatting, and are consistent
807 between OpenSSL and GnuTLS compilations. Bug 1541.
809 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
810 encoded parameter in the incoming message. Bug 1558.
812 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
813 include certificate info, eximon was claiming there were spoolfile
816 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
818 JH/14 Log delivery-related information more consistently, using the sequence
819 "H=<name> [<ip>]" wherever possible.
821 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
822 are problematic for Debian distribution, omit them from the release
825 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
827 JH/16 Fix string representation of time values on 64bit time_t architectures.
830 JH/17 Fix a null-indirection in certextract expansions when a nondefault
831 output list separator was used.
836 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
837 checkers that were complaining about end of non-void function with no
840 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
841 This was a regression introduced in 4.83 by another bugfix.
843 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
845 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
846 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
852 TF/01 Correctly close the server side of TLS when forking for delivery.
854 When a message was received over SMTP with TLS, Exim failed to clear up
855 the incoming connection properly after forking off the child process to
856 deliver the message. In some situations the subsequent outgoing
857 delivery connection happened to have the same fd number as the incoming
858 connection previously had. Exim would try to use TLS and fail, logging
859 a "Bad file descriptor" error.
861 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
862 utilities have not been installed.
864 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
865 temporary space as the ACL may create new global variables.
867 TL/01 LDAP support uses per connection or global context settings, depending
868 upon the detected version of the libraries at build time.
870 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
871 to extract and use the src ip:port in logging and expansions as if it
872 were a direct connection from the outside internet. PPv2 support was
873 updated based on HAProxy spec change in May 2014.
875 JH/02 Add ${listextract {number}{list}{success}{fail}}.
877 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
878 Properly escape header and check for NULL return.
880 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
883 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
885 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
886 characters in header names, implemented as a verify condition.
887 Contributed by Michael Fischer v. Mollard.
889 TL/05 Rename SPF condition results err_perm and err_temp to standardized
890 results permerror and temperror. Previous values are deprecated but
891 still accepted. In a future release, err_perm and err_temp will be
892 completely removed, which will be a backward incompatibility if the
893 ACL tests for either of these two old results. Patch contributed by
894 user bes-internal on the mailing list.
896 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
898 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
899 selectors, in both main and reject logs.
901 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
904 JH/07 Add malware type "sock" for talking to simple daemon.
906 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
908 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
909 routers/transports under cutthrough routing.
911 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
912 numbers. Touch up "bool" conditional to keep the same definition.
914 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
916 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
918 JH/12 Expand items in router/transport headers_add or headers_remove lists
919 individually rather than the list as a whole. Bug 1452.
921 Required for reasonable handling of multiple headers_ options when
922 they may be empty; requires that headers_remove items with embedded
923 colons must have them doubled (or the list-separator changed).
925 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
926 view the policy declared in the DMARC record. Currently, $dmarc_status
927 is a combined value of both the record presence and the result of the
930 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
932 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
933 dnslookup router and the smtp transport (applying to the forward
936 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
937 of ldap servers used for a specific lookup. Patch provided by Heiko
940 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
941 New variable $lookup_dnssec_authenticated for observability.
943 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
944 Patch submitted by Lars Timman.
946 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
948 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
949 Requires trusted mode and valid format message id, aborts otherwise.
950 Patch contributed by Heiko Schlichting.
952 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
953 certextract with support for various fields. Bug 1358.
955 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
956 is requested by default, modifiable by smtp transport option
959 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
960 operate on certificate variables to give certificate fingerprints
961 Also new ${sha256:cert_variable}.
963 JH/23 The PRDR feature is moved from being Experimental into the mainline.
965 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
966 Christian Aistleitner.
968 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
970 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
971 file. Patch from Wolfgang Breyha.
973 JH/25 Expand the coverage of the delivery $host and $host_address to
974 client authenticators run in verify callout. Bug 1476.
976 JH/26 Port service names are now accepted for tls_on_connect_ports, to
977 align with daemon_smtp_ports. Bug 72.
979 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
980 support and error reporting did not work properly.
982 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
983 and is readable. Patch from Andrew Colin Kissa.
985 TL/14 Enhance documentation of ${run expansion and how it parses the
986 commandline after expansion, particularly in the case when an
987 unquoted variable expansion results in an empty value.
989 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
991 PP/02 Fix internal collision of T_APL on systems which support RFC3123
992 by renaming away from it. Addresses GH issue 15, reported by
995 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
997 TL/15 SECURITY: prevent double expansion in math comparison functions
998 (can expand unsanitized data). Not remotely exploitable.
1005 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1007 PP/02 Make -n do something, by making it not do something.
1008 When combined with -bP, the name of an option is not output.
1010 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1013 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1014 $sender_host_name and config options to manage this, and basic check
1017 PP/05 DSCP support for outbound connections and control modifier for inbound.
1019 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1020 (Only plugin which currently uses this is kerberos4, which nobody should
1021 be using, but we should make it available and other future plugins might
1022 conceivably use it, even though it would break NAT; stuff *should* be
1023 using channel bindings instead).
1025 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1026 name; added for Sendmail compatibility; requires admin caller.
1027 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1028 just ignore it); requires trusted caller.
1029 Also parse but ignore: -Ac -Am -X<logfile>
1032 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1034 TL/02 Add +smtp_confirmation as a default logging option.
1036 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1037 Patch by Magnus Holmgren from 2007-02-20.
1039 TL/04 Bugzilla 1281 - Spec typo.
1040 Bugzilla 1283 - Spec typo.
1041 Bugzilla 1290 - Spec grammar fixes.
1043 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1045 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1047 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1048 mailing list by by Dmitry Isaikin.
1050 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1052 JH/02 Support "G" suffix to numbers in ${if comparisons.
1054 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1056 NM/01 Bugzilla 1197 - Spec typo
1057 Bugzilla 1196 - Spec examples corrections
1059 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1061 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1062 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1065 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1066 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1068 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1069 "acl {{name}{arg}...}", and optional args on acl condition
1072 JH/05 Permit multiple router/transport headers_add/remove lines.
1074 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1076 JH/07 Avoid using a waiting database for a single-message-only transport.
1077 Performance patch from Paul Fisher. Bugzilla 1262.
1079 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1082 JH/09 Add $headers_added variable, with content from use of ACL modifier
1083 add_header (but not yet added to the message). Bugzilla 199.
1085 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1086 Pulled from Bugzilla 817 by Wolfgang Breyha.
1088 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1090 (nb: this is the same fix as in Exim 4.80.1)
1092 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1095 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1098 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1100 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1101 advertise SMTP AUTH mechanism to us, instead of a generic
1102 protocol violation error. Also, make Exim more robust to bad
1103 data from the Dovecot auth socket.
1105 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1107 When a queue runner is handling a message, Exim first routes the
1108 recipient addresses, during which it prunes them based on the retry
1109 hints database. After that it attempts to deliver the message to
1110 any remaining recipients. It then updates the hints database using
1113 So if a recipient address works intermittently, it can get repeatedly
1114 deferred at routing time. The retry hints record remains fresh so the
1115 address never reaches the final cutoff time.
1117 This is a fairly common occurrence when a user is bumping up against
1118 their storage quota. Exim had some logic in its local delivery code
1119 to deal with this. However it did not apply to per-recipient defers
1120 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1122 This change adds a proper retry rule check during routing so that the
1123 final cutoff time is checked against the message's age. We only do
1124 this check if there is an address retry record and there is not a
1125 domain retry record; this implies that previous attempts to handle
1126 the address had the retry_use_local_parts option turned on. We use
1127 this as an approximation for the destination being like a local
1128 delivery, as in LMTP.
1130 I suspect this new check makes the old local delivery cutoff check
1131 redundant, but I have not verified this so I left the code in place.
1133 TF/02 Correct gecos expansion when From: is a prefix of the username.
1135 Test 0254 submits a message to Exim with the header
1139 When I ran the test suite under the user fanf2, Exim expanded
1140 the header to contain my full name, whereas it should have added
1141 a Resent-Sender: header. It erroneously treats any prefix of the
1142 username as equal to the username.
1144 This change corrects that bug.
1146 GF/01 DCC debug and logging tidyup
1147 Error conditions log to paniclog rather than rejectlog.
1148 Debug lines prefixed by "DCC: " to remove any ambiguity.
1150 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1152 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1153 Bug spotted by Jeremy Harris; was flawed since initial commit.
1154 Would have resulted in OCSP responses post-SNI triggering an Exim
1155 NULL dereference and crash.
1157 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1159 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1160 Bug detection, analysis and fix by Samuel Thibault.
1161 Bugzilla 1331, Debian bug #698092.
1163 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1165 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1166 Server implementation by Todd Lyons, client by JH.
1167 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1168 config variable "prdr_enable" controls whether the server
1169 advertises the facility. If the client requests PRDR a new
1170 acl_data_smtp_prdr ACL is called once for each recipient, after
1171 the body content is received and before the acl_smtp_data ACL.
1172 The client is controlled by both of: a hosts_try_prdr option
1173 on the smtp transport, and the server advertisement.
1174 Default client logging of deliveries and rejections involving
1175 PRDR are flagged with the string "PRDR".
1177 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1178 fclose(). Diagnosis by Todd Lyons.
1180 PP/17 Update configure.default to handle IPv6 localhost better.
1181 Patch by Alain Williams (plus minor tweaks).
1184 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1185 This is now consistent with GnuTLS, and is now documented: the
1186 previous undocumented portable approach to treating the option as
1187 unset was to force an expansion failure. That still works, and
1188 an empty string is now equivalent.
1190 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1191 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1192 not performing validation itself.
1194 PP/20 Added force_command boolean option to pipe transport.
1195 Patch from Nick Koston, of cPanel Inc.
1197 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1200 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1202 PP/21 Fix eximon continuous updating with timestamped log-files.
1203 Broken in a format-string cleanup in 4.80, missed when I repaired the
1204 other false fix of the same issue.
1205 Report and fix from Heiko Schlichting.
1208 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1209 Report from Prashanth Katuri.
1211 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1212 It's SecureTransport, so affects any MacOS clients which use the
1213 system-integrated TLS libraries, including email clients.
1215 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1216 using a MIME ACL for non-SMTP local injection.
1217 Report and assistance in diagnosis by Warren Baker.
1219 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1221 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1223 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1224 last id that failed so it may be referenced in subsequent ACL's.
1226 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1229 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1230 ldap library initialization, allowing self-signed CA's to be
1231 used. Also properly sets require_cert option later in code by
1232 using NULL (global ldap config) instead of ldap handle (per
1233 session). Bug diagnosis and testing by alxgomz.
1235 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1236 the src/util/ subdirectory.
1238 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1239 renamed to Transport Post Delivery Action by Jeremy Harris, as
1242 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1243 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1244 redis_servers = needs to be configured which will be used by the redis
1245 lookup. Patch from Warren Baker, of The Packet Hub.
1247 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1249 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1250 hostname or reverse DNS when processing a host list. Used suggestions
1251 from multiple comments on this bug.
1253 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1255 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1256 Missed a few lines, added it to make the runtest require no keyboard
1259 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1260 contains upper case chars. Make router use caseful_local_part.
1262 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1263 support when GnuTLS has been built with p11-kit.
1269 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1271 This, or similar/improved, will also be change PP/11 of 4.82.
1277 PP/01 Handle short writes when writing local log-files.
1278 In practice, only affects FreeBSD (8 onwards).
1279 Bugzilla 1053, with thanks to Dmitry Isaikin.
1281 NM/01 Bugzilla 949 - Documentation tweak
1283 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1286 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1288 PP/02 Implemented gsasl authenticator.
1290 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1292 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1293 `pkg-config foo` for cflags/libs.
1295 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1296 with rest of GSASL and with heimdal_gssapi.
1298 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1299 `pkg-config foo` for cflags/libs for the TLS implementation.
1301 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1302 properties get this fed in as external SSF. A number of robustness
1303 and debugging improvements to the cyrus_sasl authenticator.
1305 PP/08 cyrus_sasl server now expands the server_realm option.
1307 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1308 Patch by Jeremy Harris.
1310 PP/10 Added dbmjz lookup type.
1312 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1314 PP/12 MAIL args handles TAB as well as SP, for better interop with
1315 non-compliant senders.
1316 Analysis and variant patch by Todd Lyons.
1318 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1319 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1320 Patch from Dirk Mueller <dmueller@suse.com>
1322 PP/13 tls_peerdn now print-escaped for spool files.
1323 Observed some $tls_peerdn in wild which contained \n, which resulted
1324 in spool file corruption.
1326 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1327 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1328 or write after TLS renegotiation, which otherwise led to messages
1331 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1332 as a tracking header (ie: a signed header comes before the signature).
1333 Patch from Wolfgang Breyha.
1335 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1336 comma-sep list; embedded commas doubled.
1338 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1340 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1342 Report and patch from Dmitry Banschikov.
1344 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1345 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1346 failures. We appear to now support TLS1.1+ with Exim.
1348 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1349 lets Exim select keys and certificates based upon TLS SNI from client.
1350 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1351 before an outbound SMTP session. New log_selector, +tls_sni.
1353 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1354 NULL dereference. Report and patch from Alun Jones.
1356 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1357 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1358 Not seeing resolver debug output on NetBSD, but suspect this is a
1359 resolver implementation change.
1361 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1362 Left warnings. Added "eximon gdb" invocation mode.
1364 PP/21 Defaulting "accept_8bitmime" to true, not false.
1366 PP/22 Added -bw for inetd wait mode support.
1368 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1369 locate the relevant includes and libraries. Made this the default.
1371 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1372 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1374 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1375 This may cause build issues on older platforms.
1377 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1378 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1379 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1380 Added SNI support via GnuTLS too.
1381 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1383 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1385 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1386 Applied second patch from Janne, implementing suggestion to default
1387 multiple-strings-in-record handling to match SPF spec.
1389 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1391 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1392 read-only, out of scope).
1393 Patch from Wolfgang Breyha, report from Stuart Northfield.
1395 PP/29 Fix three issues highlighted by clang analyser static analysis.
1396 Only crash-plausible issue would require the Cambridge-specific
1397 iplookup router and a misconfiguration.
1398 Report from Marcin Mirosław.
1400 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1402 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1403 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1404 As part of this, removing so much warning spew let me fix some minor
1405 real issues in debug logging.
1407 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1408 assignment on my part. Fixed.
1410 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1411 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1412 Janne Snabb (who went above and beyond: thank you).
1414 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1415 string otherwise requires a connection and a bunch more work and it's
1416 relatively easy to get wrong. Should also expose TLS library linkage
1419 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1420 64-bit ${eval} (JH/03).
1422 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1423 GNU libc to support some of the 64-bit stuff, should not lead to
1424 conflicts. Defined before os.h is pulled in, so if a given platform
1425 needs to override this, it can.
1427 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1428 protection layer was required, which is not implemented.
1429 Bugzilla 1254, patch from Wolfgang Breyha.
1431 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1432 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1433 tls_dhparam take prime identifiers. Also unbreak combination of
1434 OpenSSL+DH_params+TLSSNI.
1436 PP/39 Disable SSLv2 by default in OpenSSL support.
1442 PP/01 Solaris build fix for Oracle's LDAP libraries.
1443 Bugzilla 1109, patch from Stephen Usher.
1445 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1447 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1450 TF/02 Fix a couple more cases where we did not log the error message
1451 when unlink() failed. See also change 4.74-TF/03.
1453 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1454 lock up or crash if it happened to be inside a call to libc when it
1455 got a SIGUSR1 from exiwhat.
1457 The SIGUSR1 handler appends the current process status to the process
1458 log which is later printed by exiwhat. It used to use the general
1459 purpose logging code to do this, but several functions it calls are
1460 not safe for signals.
1462 The new output code in the SIGUSR1 handler is specific to the process
1463 log, and simple enough that it's easy to inspect for signal safety.
1464 Removing some special cases also simplifies the general logging code.
1465 Removing the spurious timestamps from the process log simplifies
1468 TF/04 Improved ratelimit ACL condition.
1470 The /noupdate option has been deprecated in favour of /readonly which
1471 has clearer semantics. The /leaky, /strict, and /readonly update modes
1472 are mutually exclusive. The update mode is no longer included in the
1473 database key; it just determines when the database is updated. (This
1474 means that when you upgrade Exim will forget old rate measurements.)
1476 Exim now checks that the per_* options are used with an update mode that
1477 makes sense for the current ACL. For example, when Exim is processing a
1478 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1479 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1480 must specify per_mail/readonly. If you omit the update mode it defaults to
1481 /leaky where that makes sense (as before) or /readonly where required.
1483 The /noupdate option is now undocumented but still supported for
1484 backwards compatibility. It is equivalent to /readonly except that in
1485 ACLs where /readonly is required you may specify /leaky/noupdate or
1486 /strict/noupdate which are treated the same as /readonly.
1488 A useful new feature is the /count= option. This is a generalization
1489 of the per_byte option, so that you can measure the throughput of other
1490 aggregate values. For example, the per_byte option is now equivalent
1491 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1493 The per_rcpt option has been generalized using the /count= mechanism
1494 (though it's more complicated than the per_byte equivalence). When it is
1495 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1496 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1497 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1498 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1499 Note that using per_rcpt with a non-readonly update mode in more than
1500 one ACL will cause the recipients to be double-counted. (The per_mail
1501 and per_byte options don't have this problem.)
1503 The handling of very low rates has changed slightly. If the computed rate
1504 is less than the event's count (usually one) then this event is the first
1505 after a long gap. In this case the rate is set to the same as this event's
1506 count, so that the first message of a spam run is counted properly.
1508 The major new feature is a mechanism for counting the rate of unique
1509 events. The new per_addr option counts the number of different
1510 recipients that someone has sent messages to in the last time period. It
1511 behaves like per_rcpt if all the recipient addresses are different, but
1512 duplicate recipient addresses do not increase the measured rate. Like
1513 the /count= option this is a general mechanism, so the per_addr option
1514 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1515 example, measure the rate that a client uses different sender addresses
1516 with the options per_mail/unique=$sender_address. There are further
1517 details in the main documentation.
1519 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1521 TF/06 Removed a few PCRE remnants.
1523 TF/07 Automatically extract Exim's version number from tags in the git
1524 repository when doing development or release builds.
1526 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1527 Bugzilla 879. Patch from Paul Fisher.
1529 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1530 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1533 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1535 PP/05 Variable $av_failed, true if the AV scanner deferred.
1536 Bugzilla 1078. Patch from John Horne.
1538 PP/06 Stop make process more reliably on build failure.
1539 Bugzilla 1087. Patch from Heiko Schlittermann.
1541 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1542 Bugzilla 1089. Patch from Heiko Schlittermann.
1544 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1545 Bugzilla 1131. Patch from Holger Weiß.
1547 PP/09 Handle IPv6 addresses with SPF.
1548 Bugzilla 860. Patch from Wolfgang Breyha.
1550 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1552 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1555 PP/11 match_* no longer expand right-hand-side by default.
1556 New compile-time build option, EXPAND_LISTMATCH_RHS.
1557 New expansion conditions, "inlist", "inlisti".
1559 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1561 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1563 PP/14 fix log_write() format string regression from TF/03.
1564 Bugzilla 1152. Patch from Dmitry Isaikin.
1570 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1572 PP/02 Harmonised TLS library version reporting; only show if debugging.
1573 Layout now matches that introduced for other libraries in 4.74 PP/03.
1575 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1577 PP/04 New "dns_use_edns0" global option.
1579 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1582 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1583 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1585 TK/01 Updated PolarSSL code to 0.14.2.
1586 Bugzilla 1097. Patch from Andreas Metzler.
1588 PP/07 Catch divide-by-zero in ${eval:...}.
1589 Fixes bugzilla 1102.
1591 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1594 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1595 format-string attack -- SECURITY: remote arbitrary code execution.
1597 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1598 time unintentionally subject to list matching rules, letting the header
1599 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1600 arbitrary string expansion). This allowed for information disclosure.
1602 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1603 INT_MIN/-1 -- value coerced to INT_MAX.
1609 NM/01 Workaround for PCRE version dependency in version reporting
1612 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1613 This fixes portability to compilers other than gcc, notably
1614 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1616 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1617 makefiles for portability to HP-UX and POSIX correctness.
1619 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1620 Also via indented variable definition in the Makefile.
1621 (Debugging by Oliver Heesakkers).
1623 PP/02 Restore caching of spamd results with expanded spamd_address.
1624 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1626 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1627 Improves build reliability. Fix from: Frank Elsner
1629 NM/02 Fix wide character breakage in the rfc2047 coding
1630 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1632 NM/03 Allow underscore in dnslist lookups
1633 Fixes bug 1026. Patch from Graeme Fowler
1635 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1636 Code patches from Adam Ciarcinski of NetBSD.
1638 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1641 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1642 is logged, to avoid truncation. Patch from John Horne.
1644 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1645 Patch from Jakob Hirsch.
1647 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1648 SQL string expansion failure details.
1649 Patch from Andrey Oktyabrski.
1651 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1652 Patch from Simon Arlott.
1654 PP/09 New lookups functionality failed to compile on old gcc which rejects
1655 extern declarations in function scope.
1656 Patch from Oliver Fleischmann
1658 PP/10 Use sig_atomic_t for flags set from signal handlers.
1659 Check getgroups() return and improve debugging.
1660 Fixed developed for diagnosis in bug 927 (which turned out to be
1663 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1664 Patch from Mark Zealey.
1666 PP/12 Bugzilla 1056: Improved spamd server selection.
1667 Patch from Mark Zealey.
1669 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1670 Based on patch from Heiko Schlittermann.
1672 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1673 Patch from Uwe Doering, sign-off by Michael Haardt.
1675 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1676 variable declaration deep within a block. Bug and patch from
1679 PP/15 lookups-Makefile IRIX compatibility coercion.
1681 PP/16 Make DISABLE_DKIM build knob functional.
1683 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1684 Patch by Simon Arlott
1686 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1687 variable argument macros. Our copy now differs from upstream.
1693 TF/01 Failure to get a lock on a hints database can have serious
1694 consequences so log it to the panic log.
1696 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1697 controlled using the smtp_confirmation log selector.
1699 TF/03 Include the error message when we fail to unlink a spool file.
1701 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1702 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1703 for maintaining out-of-tree patches for some time.
1705 PP/01 Bugzilla 139: Documentation and portability issues.
1706 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1707 Handle per-OS dynamic-module compilation flags.
1709 PP/02 Let /dev/null have normal permissions.
1710 The 4.73 fixes were a little too stringent and complained about the
1711 permissions on /dev/null. Exempt it from some checks.
1712 Reported by Andreas M. Kirchwitz.
1714 PP/03 Report version information for many libraries, including
1715 Exim version information for dynamically loaded libraries. Created
1716 version.h, now support a version extension string for distributors
1717 who patch heavily. Dynamic module ABI change.
1719 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1720 privilege escalation vulnerability whereby the Exim run-time user
1721 can cause root to append content of the attacker's choosing to
1724 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1727 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1728 If dropping privileges for untrusted macros, we disabled normal logging
1729 on the basis that it would fail; for the Exim run-time user, this is not
1730 the case, and it resulted in successful deliveries going unlogged.
1731 Fixed. Reported by Andreas Metzler.
1737 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1738 only prepend for the Resent-* case. Fixes regression introduced in
1739 Exim 4.70 by NM/22 for Bugzilla 607.
1741 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1742 increasing numbers of administrators be bitten by this.
1744 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1746 PP/03 Bugzilla 994: added openssl_options main configuration option.
1748 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1750 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1752 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1754 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1755 without a peer certificate, leading to a segfault because of an
1756 assumption that peers always have certificates. Be a little more
1757 paranoid. Problem reported by Martin Tscholak.
1759 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1760 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1761 NB: ClamAV planning to remove STREAM in "middle of 2010".
1762 CL also introduces -bmalware, various -d+acl logging additions and
1763 more caution in buffer sizes.
1765 PP/09 Implemented reverse_ip expansion operator.
1767 PP/10 Bugzilla 937: provide a "debug" ACL control.
1769 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1771 PP/12 Bugzilla 973: Implement --version.
1773 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1775 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1777 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1779 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1780 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1781 ignore trailing whitespace.
1783 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1785 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1788 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1789 Notification from Dr Andrew Aitchison.
1791 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1792 ExtendedDetectionInfo response format.
1793 Notification from John Horne.
1795 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1798 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1799 XSL and documented dependency on system catalogs, with examples of how
1802 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1805 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1806 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1807 configuration file which is writeable by the Exim user or group.
1809 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1810 of configuration files to cover files specified with the -C option if
1811 they are going to be used with root privileges, not just the default
1814 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1815 option (effectively making it always true).
1817 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1818 files to be used while preserving root privileges.
1820 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1821 that rogue child processes cannot use them.
1823 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1824 run-time user, instead of root.
1826 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1827 Exim run-time user without dropping privileges.
1829 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1830 result string, instead of calling string_vformat() twice with the same
1833 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1834 for other users. Others should always drop root privileges if they use
1835 -C on the command line, even for a whitelisted configure file.
1837 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1839 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1845 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1846 $data_path, and $header_path variables; fixed documentation bugs and
1849 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1850 exipick to access non-standard spools, including the "frozen" queue
1853 NM/01 Bugzilla 965: Support mysql stored procedures.
1854 Patch from Alain Williams
1856 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1858 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1859 Patch from Andreas Metzler
1861 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1862 Patch from Kirill Miazine
1864 NM/05 Bugzilla 671: Added umask to procmail example.
1866 JJ/03 installed exipick 20100323.0, fixing doc bug
1868 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1869 directory. Notification and patch from Dan Rosenberg.
1871 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1873 TK/02 Improve log output when DKIM signing operation fails.
1875 MH/01 Treat the transport option dkim_domain as a colon separated
1876 list, not as a single string, and sign the message with each element,
1877 omitting multiple occurences of the same signer.
1879 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1880 Bugzilla 985, 986. Patch by Simon Arlott
1882 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1883 Patch by Simon Arlott
1885 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1886 MBX locking. Notification from Dan Rosenberg.
1892 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1894 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1896 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1898 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1900 NM/04 Bugzilla 915: Flag broken perl installation during build.
1906 TK/01 Added patch by Johannes Berg that expands the main option
1907 "spamd_address" if it starts with a dollar sign.
1909 TK/02 Write list of recipients to X-Envelope-Sender header when building
1910 the mbox-format spool file for content scanning (suggested by Jakob
1913 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1914 (http://www.dcc-servers.net/) support via dccifd. Activated by
1915 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1917 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1918 by Mark Daniel Reidel <mr@df.eu>.
1920 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1921 When building exim an external PCRE library is now needed -
1922 PCRE is a system library on the majority of modern systems.
1923 See entry on PCRE_LIBS in EDITME file.
1925 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1926 conversation. Added nologin parameter to request.
1927 Patch contributed by Kirill Miazine.
1929 TF/01 Do not log submission mode rewrites if they do not change the address.
1931 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1933 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1934 log files in place. Contributed by Roberto Lima.
1936 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1938 TF/03 Bugzilla 615: When checking the local_parts router precondition
1939 after a local_part_suffix or local_part_prefix option, Exim now
1940 does not use the address's named list lookup cache, since this
1941 contains cached lookups for the whole local part.
1943 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1944 Robert Millan. Documentation is in experimental-spec.txt.
1946 TF/04 Bugzilla 668: Fix parallel build (make -j).
1948 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1950 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1951 Patch provided by Jan Srzednicki.
1953 TF/05 Leading white space used to be stripped from $spam_report which
1954 wrecked the formatting. Now it is preserved.
1956 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1957 that they are available at delivery time.
1959 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1961 TF/08 TLS error reporting now respects the incoming_interface and
1962 incoming_port log selectors.
1964 TF/09 Produce a more useful error message if an SMTP transport's hosts
1965 setting expands to an empty string.
1967 NM/06 Bugzilla 744: EXPN did not work under TLS.
1968 Patch provided by Phil Pennock.
1970 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1971 Patch provided by Richard Godbee.
1973 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1974 acl_smtp_notquit, added index entry.
1976 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1977 Patch provided by Eugene Bujak.
1979 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1980 accept(). Patch provided by Maxim Dounin.
1982 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1983 Patch provided by Phil Pennock.
1985 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1987 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1988 Patch provided by Brad "anomie" Jorsch.
1990 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1991 Patch provided by Dean Brooks.
1993 TK/05 Add native DKIM support (does not depend on external libraries).
1995 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1996 Patch provided by Graeme Fowler.
1998 NM/16 Bugzilla 851: Documentation example syntax fix.
2000 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2002 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2005 NM/19 Bugzilla 745: TLS version reporting.
2006 Patch provided by Phil Pennock.
2008 NM/20 Bugzilla 167: bool: condition support.
2009 Patch provided by Phil Pennock.
2011 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2012 clients. Patch provided by Phil Pennock.
2014 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2015 Patch provided by Brad "anomie" Jorsch.
2017 NM/23 Bugzilla 687: Fix misparses in eximstats.
2018 Patch provided by Heiko Schlittermann.
2020 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2021 Patch provided by Heiko Schlittermann.
2023 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2024 plus update to original patch.
2026 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2028 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2029 Patch provided by David Brownlee.
2031 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2033 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2035 NM/30 Bugzilla 888: TLS documentation bugfixes.
2037 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2039 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2040 Unlike the original bugzilla I have changed all shell scripts in src tree.
2042 NM/33 Bugzilla 898: Transport filter timeout fix.
2043 Patch by Todd Rinaldo.
2045 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2046 Patch by Serge Demonchaux.
2048 NM/35 Bugzilla 39: Base64 decode bug fixes.
2049 Patch by Jakob Hirsch.
2051 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2053 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2055 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2057 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2063 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2064 ALT-N's libdkim that I have put here:
2065 http://duncanthrax.net/exim-experimental/
2067 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2068 were called 'true' and it seems that C99 defines that as a reserved
2069 keyword to be used with 'bool' variable types. That means you could
2070 not include C99-style headers which use bools without triggering
2071 build errors in sieve.c.
2073 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2074 as mailq or other aliases. Changed the --help handling significantly
2075 to do whats expected. exim_usage() emits usage/help information.
2077 SC/01 Added the -bylocaldomain option to eximstats.
2079 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2081 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2083 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2089 PH/01 Another patch from the Sieve maintainer.
2091 PH/02 When an IPv6 address is converted to a string for single-key lookup
2092 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2093 dots are used instead of colons so that keys in lsearch files need not
2094 contain colons. This was done some time before quoting was made available
2095 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2096 (notated using the quote facility) so as to distinguish them from IPv4
2097 keys. This meant that lookups for IP addresses in host lists did not work
2098 for iplsearch lookups.
2100 This has been fixed by arranging for IPv6 addresses to be expressed with
2101 colons if the lookup type is iplsearch. This is not incompatible, because
2102 previously such lookups could never work.
2104 The situation is now rather anomalous, since one *can* have colons in
2105 ordinary lsearch keys. However, making the change in all cases is
2106 incompatible and would probably break a number of configurations.
2108 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2111 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2112 conversion specification without a maximum field width, thereby enabling
2113 a rogue spamd server to cause a buffer overflow. While nobody in their
2114 right mind would setup Exim to query an untrusted spamd server, an
2115 attacker that gains access to a server running spamd could potentially
2116 exploit this vulnerability to run arbitrary code as the Exim user.
2118 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2119 $primary_hostname instead of what libspf2 thinks the hosts name is.
2121 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2122 a directory entry by the name of the lookup key. Previously, if a
2123 symlink pointed to a non-existing file or a file in a directory that
2124 Exim lacked permissions to read, a lookup for a key matching that
2125 symlink would fail. Now it is enough that a matching directory entry
2126 exists, symlink or not. (Bugzilla 503.)
2128 PH/03 The body_linecount and body_zerocount variables are now exported in the
2131 PH/04 Added the $dnslist_matched variable.
2133 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2134 This means they are set thereafter only if the connection becomes
2137 PH/06 Added the client_condition to authenticators so that some can be skipped
2138 by clients under certain conditions.
2140 PH/07 The error message for a badly-placed control=no_multiline_responses left
2141 "_responses" off the end of the name.
2143 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2145 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2146 (without spaces) instead of just copying the configuration text.
2148 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2150 PH/11 Added $max_received_linelength.
2152 PH/12 Added +ignore_defer and +include_defer to host lists.
2154 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2155 way in which PCRE > 7.0 is built.
2157 PH/14 Implemented queue_only_load_latch.
2159 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2160 MAIL command. The effect was to mangle the value on 64-bit systems.
2162 PH/16 Another patch from the Sieve maintainer.
2164 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2166 PH/18 If a system quota error occurred while trying to create the file for
2167 a maildir delivery, the message "Mailbox is full" was not appended to the
2168 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2169 only to a quota excession during the actual writing of the file.
2171 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2172 characters?) which causes problems in log lines. The DN values are now
2173 passed through string_printing() before being added to log lines.
2175 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2176 and InterBase are left for another time.)
2178 PH/21 Added message_body_newlines option.
2180 PH/22 Guard against possible overflow in moan_check_errorcopy().
2182 PH/23 POSIX allows open() to be a macro; guard against that.
2184 PH/24 If the recipient of an error message contained an @ in the local part
2185 (suitably quoted, of course), incorrect values were put in $domain and
2186 $local_part during the evaluation of errors_copy.
2192 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2193 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2196 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2197 issue a MAIL command.
2199 PH/02 In an ACL statement such as
2201 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2203 if a client was not listed at all, or was listed with a value other than
2204 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2205 the condition was not true (as it should be), so access was not denied.
2206 The bug was that the ! inversion was incorrectly passed on to the second
2207 item. This has been fixed.
2209 PH/03 Added additional dnslists conditions == and =& which are different from
2210 = and & when the dns lookup returns more than one IP address.
2212 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2213 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2215 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2216 FSYNC, which compiles an option called disable_fsync that allows for
2217 bypassing fsync(). The documentation is heavily laced with warnings.
2219 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2221 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2222 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2223 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2224 including adding "make clean"; (3) Added -fPIC when compiling the test
2225 dynamically loaded module, to get rid of a warning.
2227 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2228 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2229 = 0s. The bug is otherwise harmless.
2231 PH/07 There was a bug in the dovecot authenticator such that the value of
2232 $auth1 could be overwritten, and so not correctly preserved, after a
2233 successful authentication. This usually meant that the value preserved by
2234 the server_setid option was incorrect.
2236 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2238 PH/09 Installed PCRE release 7.0.
2240 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2241 run for batched SMTP input. It is now run at the start of every message
2242 in the batch. While fixing this I discovered that the process information
2243 (output by running exiwhat) was not always getting set for -bs and -bS
2244 input. This is fixed, and it now also says "batched" for BSMTP.
2246 PH/11 Added control=no_pipelining.
2248 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2249 patch, slightly modified), and move the expansion of helo_data till after
2250 the connection is made in the smtp transport (so it can use these
2253 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2255 PH/14 Added log_selector = +pid.
2257 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2259 PH/16 Add ${if forany and ${if forall.
2261 PH/17 Added dsn_from option to vary the From: line in DSNs.
2263 PH/18 Flush SMTP output before performing a callout, unless control =
2264 no_callout_flush is set.
2266 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2267 was true (the default) a successful delivery failed to delete the retry
2268 item, thus causing premature timeout of the address. The bug is now
2271 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2273 PH/21 Long custom messages for fakedefer and fakereject are now split up
2274 into multiline responses in the same way that messages for "deny" and
2275 other ACL rejections are.
2277 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2278 with slight modification.
2280 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2281 draft, changing the syntax and factoring some duplicate code".
2283 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2284 for deliveries of the second and subsequent messages over the same SMTP
2287 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2288 ${reduce, with only minor "tidies".
2290 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2292 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2293 expansion side effects.
2295 PH/27 When a message times out after an over-quota error from an Exim-imposed
2296 quota, the bounce message says "mailbox is full". This message was not
2297 being given when it was a system quota that was exceeded. It now should
2300 MH/03 Made $recipients available in local_scan(). local_scan() already has
2301 better access to the recipient list through recipients_list[], but
2302 $recipients can be useful in postmaster-provided expansion strings.
2304 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2305 in the case of a MAIL command with additional options following the
2306 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2307 were accidentally chopped off.
2309 PH/29 SMTP synchronization checks are implemented when a command is read -
2310 there is a check that no more input is waiting when there shouldn't be
2311 any. However, for some commands, a delay in an ACL can mean that it is
2312 some time before the response is written. In this time, more input might
2313 arrive, invalidly. So now there are extra checks after an ACL has run for
2314 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2315 pipelining has not been advertised.
2317 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2319 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2320 returned in the SMTP response when smtp_return_error_details was false.
2321 This has been fixed.
2323 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2324 instead of the C I/O that was originally supplied, because problems were
2325 reported on Solaris.
2327 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2328 Exim which did not show up earlier: it was assuming that a call to
2329 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2330 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2331 was a macro that became an assignment, so it seemed to work. This has
2332 changed to a proper function call with a void return, hence the compile
2333 error. Exim's code has been fixed.
2335 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2338 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2340 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2342 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2343 because it turns out to be too big or there is a timeout) there is no
2344 "Completed" line in the log. When some messages of this type were
2345 selected by exigrep, they were listed as "not completed". Others were
2346 picked up by some special patterns. I have improved the selection
2347 criteria to be more general.
2349 PH/38 The host_find_failed option in the manualroute router can now be set
2350 to "ignore", to completely ignore a host whose IP address cannot be
2351 found. If all hosts are ignored, the behaviour is controlled by the new
2352 host_all_ignored option.
2354 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2355 homing or because of multiple MX records with /mx) generated more than
2356 one IP address, and the following item turned out to be the local host,
2357 all the secondary addresses of the first item were incorrectly removed
2358 from the list, along with the local host and any following hosts (which
2359 is what is supposed to happen).
2361 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2362 whoever called Exim into the -H file. In the case of the daemon it was
2363 behaving confusingly. When first started, it used values for whoever
2364 started the daemon, but after a SIGHUP it used the Exim user (because it
2365 calls itself on a restart). I have changed the code so that it now always
2368 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2369 message are rejected with the same error (e.g. no authentication or bad
2370 sender address), and a DATA command is nevertheless sent (as can happen
2371 with PIPELINING or a stupid MUA), the error message that was given to the
2372 RCPT commands is included in the rejection of the DATA command. This is
2373 intended to be helpful for MUAs that show only the final error to their
2376 PH/42 Another patch from the Sieve maintainer.
2378 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2379 Eximstats - Fixed some broken HTML links and added missing column headers
2381 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2382 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2384 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2386 PH/43 Yet another patch from the Sieve maintainer.
2388 PH/44 I found a way to check for a TCP/IP connection going away before sending
2389 the response to the final '.' that terminates a message, but only in the
2390 case where the client has not sent further data following the '.'
2391 (unfortunately, this is allowed). However, in many cases there won't be
2392 any further data because there won't be any more messages to send. A call
2393 to select() can be used: if it shows that the input is "ready", there is
2394 either input waiting, or the socket has been closed. An attempt to read
2395 the next input character can distinguish the two cases. Previously, Exim
2396 would have sent an OK response which the client would never have see.
2397 This could lead to message repetition. This fix should cure that, at
2398 least in a lot of common cases.
2400 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2401 advertised in response to EHLO.
2407 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2408 fixed by 4.65/MH/01 (is this a record?) are fixed:
2410 (i) An empty string was always treated as zero by the numeric comparison
2411 operators. This behaviour has been restored.
2413 (ii) It is documented that the numeric comparison operators always treat
2414 their arguments as decimal numbers. This was broken in that numbers
2415 starting with 0 were being interpreted as octal.
2417 While fixing these problems I realized that there was another issue that
2418 hadn't been noticed. Values of message_size_limit (both the global option
2419 and the transport option) were treated as octal if they started with 0.
2420 The documentation was vague. These values are now always treated as
2421 decimal, and I will make that clear in the documentation.
2427 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2428 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2431 MH/01 Don't check that the operands of numeric comparison operators are
2432 integers when their expansion is in "skipping" mode (fixes bug
2433 introduced by 4.64-PH/07).
2435 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2436 child addresses, Exim now panics and dies. Previously, because the count
2437 is held in a short int, deliveries were likely to be lost. As such a
2438 large number of recipients for a single message is ridiculous
2439 (performance will be very, very poor), I have chosen to impose a limit
2440 rather than extend the field.
2446 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2447 leftover -K file (the existence of which was triggered by #402).
2448 While we were at it, introduced process PID as part of the -K
2449 filename. This should rule out race conditions when creating
2452 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2453 processing considerably. Previous code took too long for large mails,
2454 triggering a timeout which in turn triggers #401.
2456 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2457 in the DK code in transports.c. sendfile() is not really portable,
2458 hence the _LINUX specificness.
2460 TF/01 In the add_headers option to the mail command in an Exim filter,
2461 there was a bug that Exim would claim a syntax error in any
2462 header after the first one which had an odd number of characters
2465 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2466 callout verification, Exim cached a "reject" for the entire domain. This
2467 is correct for most verifications, but it is not correct for a recipient
2468 verification with use_sender or use_postmaster set, because in that case
2469 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2470 case of MAIL FROM:<> rejection from other early rejections (e.g.
2471 rejection of HELO). When verifying a recipient using a non-null MAIL
2472 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2473 Whatever the result of the callout, the value of the domain cache is
2474 left unchanged (for any other kind of callout, getting as far as trying
2475 RCPT means that the domain itself is ok).
2477 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2480 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2481 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2482 session, the daemon ignores ECONNECT errors and logs others; it now
2483 ignores EPIPE as well.
2485 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2486 (quoted-printable decoding).
2488 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2489 later the small subsequent patch to fix an introduced bug.
2491 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2493 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2495 PH/08 An error is now given if message_size_limit is specified negative.
2497 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2498 to be given (somewhat) arbitrary names.
2500 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2503 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2504 miscellaneous code fixes
2506 PH/10 Added the log_reject_target ACL modifier to specify where to log
2509 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2510 hostname. This is wrong, because it relates to the incoming message (and
2511 probably the interface on which it is arriving) and not to the outgoing
2512 callout (which could be using a different interface). This has been
2513 changed to use the value of the helo_data option from the smtp transport
2514 instead - this is what is used when a message is actually being sent. If
2515 there is no remote transport (possible with a router that sets up host
2516 addresses), $smtp_active_hostname is used.
2518 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2519 tweaks were necessary in order to get it to work (see also 21 below):
2520 (a) The code assumed that strncpy() returns a negative number on buffer
2521 overflow, which isn't the case. Replaced with Exim's string_format()
2523 (b) There were several signed/unsigned issues. I just did the minimum
2524 hacking in of casts. There is scope for a larger refactoring.
2525 (c) The code used strcasecmp() which is not a standard C function.
2526 Replaced with Exim's strcmpic() function.
2527 (d) The code set only $1; it now sets $auth1 as well.
2528 (e) A simple test gave the error "authentication client didn't specify
2529 service in request". It would seem that Dovecot has changed its
2530 interface. Fortunately there's a specification; I followed it and
2531 changed what the client sends and it appears to be working now.
2533 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2536 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2537 address A is aliased to B and C, where B exists and C does not. Without
2538 -v the output is "A verified" because verification stops after a
2539 successful redirection if more than one address is generated. However,
2540 with -v the child addresses are also verified. Exim was outputting "A
2541 failed to verify" and then showing the successful verification for C,
2542 with its parentage. It now outputs "B failed to verify", showing B's
2543 parentage before showing the successful verification of C.
2545 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2546 look up a TXT record in a specific list after matching in a combined
2549 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2550 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2551 they consult the DNS. I had assumed they would set it the way they
2552 wanted; and indeed my experiments on Linux seem to show that in some
2553 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2554 To be on the safe side, however, I have now made the interface to
2555 host_find_byname() similar to host_find_bydns(), with an argument
2556 containing the DNS resolver options. The host_find_byname() function now
2557 sets these options at its start, just as host_find_bydns() does. The smtp
2558 transport options dns_qualify_single and dns_search_parents are passed to
2559 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2560 of host_find_byname() use the default settings of RES_DEFNAMES
2561 (qualify_single) but not RES_DNSRCH (search_parents).
2563 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2564 spool_read_header() do less string testing, by means of a preliminary
2565 switch on the second character of optional "-foo" lines. (This is
2566 overdue, caused by the large number of possibilities that now exist.
2567 Originally there were few.) While I was there, I also converted the
2568 str(n)cmp tests so they don't re-test the leading "-" and the first
2569 character, in the hope this might squeeze out yet more improvement.
2571 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2572 flag allowing group syntax was set by the header_syntax check but not
2573 turned off, possible causing trouble later; (2) The flag was not being
2574 set at all for the header_verify test, causing "group"-style headers to
2575 be rejected. I have now set it in this case, and also caused header_
2576 verify to ignore an empty address taken from a group. While doing this, I
2577 came across some other cases where the code for allowing group syntax
2578 while scanning a header line wasn't quite right (mostly, not resetting
2579 the flag correctly in the right place). These bugs could have caused
2580 trouble for malformed header lines. I hope it is now all correct.
2582 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2583 with the "reply" argument non-NULL. The code, however (which originally
2584 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2585 but it didn't always do it. This confused somebody who was copying the
2586 code for some other use. I have removed all the tests.
2588 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2589 feature that was used to support insecure browsers during the U.S. crypto
2590 embargo. It requires special client support, and Exim is probably the
2591 only MTA that supported it -- and would never use it because real RSA is
2592 always available. This code has been removed, because it had the bad
2593 effect of slowing Exim down by computing (never used) parameters for the
2594 RSA_EXPORT functionality.
2596 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2597 authenticator to fail if there's a tab character in the incoming data
2598 (there should never be unless someone is messing about, as it's supposed
2599 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2600 if the connection is using TLS or if the remote IP is the same as the
2601 local IP, and the "valid-client-cert option" if a client certificate has
2604 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2605 authenticators. This can be used for authorization after authentication
2606 succeeds. (In the case of plaintext, it servers for both authentication
2609 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2610 if any retry times were supplied.
2612 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2613 connection, where there is no client IP address to check. In this
2614 situation, the verify now always succeeds.
2616 PH/25 Applied John Jetmore's -Mset patch.
2618 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2620 PH/27 In a string expansion for a processed (not raw) header when multiple
2621 headers of the same name were present, leading whitespace was being
2622 removed from all of them, but trailing whitespace was being removed only
2623 from the last one. Now trailing whitespace is removed from each header
2624 before concatenation. Completely empty headers in a concatenation (as
2625 before) are ignored.
2627 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2628 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2630 PH/29 [Removed. This was a change that I later backed out, and forgot to
2631 correct the ChangeLog entry (that I had efficiently created) before
2632 committing the later change.]
2634 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2635 address errors (4xx response to RCPT) over the same connection as other
2636 messages routed to the same hosts. Such deliveries are always "forced",
2637 so retry times are not inspected. This resulted in far too many retries
2638 for the affected addresses. The effect occurred only when there were more
2639 hosts than the hosts_max_try setting in the smtp transport when it had
2640 the 4xx errors. Those hosts that it had tried were not added to the list
2641 of hosts for which the message was waiting, so if all were tried, there
2642 was no problem. Two fixes have been applied:
2644 (i) If there are any address or message errors in an SMTP delivery, none
2645 of the hosts (tried or untried) are now added to the list of hosts
2646 for which the message is waiting, so the message should not be a
2647 candidate for sending over the same connection that was used for a
2648 successful delivery of some other message. This seems entirely
2649 reasonable: after all the message is NOT "waiting for some host".
2650 This is so "obvious" that I'm not sure why it wasn't done
2651 previously. Hope I haven't missed anything, but it can't do any
2652 harm, as the worst effect is to miss an optimization.
2654 (ii) If, despite (i), such a delivery is accidentally attempted, the
2655 routing retry time is respected, so at least it doesn't keep
2656 hammering the server.
2658 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2659 in ${readsocket because some servers need this prod.
2661 PH/32 Added some extra debug output when updating a wait-xxx database.
2663 PH/33 The hint "could be header name not terminated by colon", which has been
2664 given for certain expansion errors for a long time, was not being given
2665 for the ${if def:h_colon_omitted{... case.
2667 PH/34 The spec says: "With one important exception, whenever a domain list is
2668 being scanned, $domain contains the subject domain." There was at least
2669 one case where this was not true.
2671 PH/35 The error "getsockname() failed: connection reset by peer" was being
2672 written to the panic log as well as the main log, but it isn't really
2673 panic-worthy as it just means the connection died rather early on. I have
2674 removed the panic log writing for the ECONNRESET error when getsockname()
2677 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2678 runs only) independently of the message's sender address. This meant
2679 that, if the 4xx error was in fact related to the sender, a different
2680 message to the same recipient with a different sender could confuse
2681 things. In particular, this can happen when sending to a greylisting
2682 server, but other circumstances could also provoke similar problems.
2683 I have changed the default so that the retry time for these errors is now
2684 based a combination of the sender and recipient addresses. This change
2685 can be overridden by setting address_retry_include_sender=false in the
2688 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2689 remote server are returned as part of bounce messages. This was not
2690 happening for LMTP over a pipe (the lmtp transport), but now it is the
2691 same for both kinds of LMTP.
2693 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2694 in header lines that were in fact CNAMEs. This is no longer the case.
2696 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2697 and queue runs started by the daemon processed all messages. This has
2698 been fixed so that -R and -S can now usefully be given with -q<time>.
2700 PH/40 Import PCRE release 6.7 (fixes some bugs).
2702 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2704 PH/42 Give an error if -q is specified more than once.
2706 PH/43 Renamed the variables $interface_address and $interface_port as
2707 $received_ip_address and $received_port, to make it clear that these
2708 values apply to message reception, and not to the outgoing interface when
2709 a message is delivered. (The old names remain recognized, of course.)
2711 PH/44 There was no timeout on the connect() call when using a Unix domain
2712 socket in the ${readsocket expansion. There now is.
2714 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2715 be meaningful with "accept".
2717 SC/01 Eximstats V1.43
2718 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2720 SC/02 Eximstats V1.44
2721 Use a glob alias rather than an array ref in the generated
2722 parser. This improves both readability and performance.
2724 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2725 Collect SpamAssassin and rejection statistics.
2726 Don't display local sender or destination tables unless
2727 there is data to show.
2728 Added average volumes into the top table text output.
2730 SC/04 Eximstats V1.46
2731 Collect data on the number of addresses (recipients)
2732 as well as the number of messages.
2734 SC/05 Eximstats V1.47
2735 Added 'Message too big' to the list of mail rejection
2736 reasons (thanks to Marco Gaiarin).
2738 SC/06 Eximstats V1.48
2739 Mainlog lines which have GMT offsets and are too short to
2740 have a flag are now skipped.
2742 SC/07 Eximstats V1.49 (Alain Williams)
2743 Added the -emptyok flag.
2745 SC/08 Eximstats V1.50
2746 Fixes for obtaining the IP address from reject messages.
2748 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2749 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2750 whitespace changes from 4.64-PH/27
2752 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2755 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2756 are found, allow negative numbers in numeric criteria)
2758 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2760 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2761 to match changes made in 4.64-PH/43
2763 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2765 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2766 "domain" is actually an IP address, and force a failure. This locks out
2767 those revolvers/nameservers that support "A-for-A" lookups, in
2768 contravention of the specifications.
2770 PH/48 When a host name was looked up from an IP address, and the subsequent
2771 forward lookup of the name timed out, the host name was left in
2772 $sender_host_name, contrary to the specification.
2774 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2775 restricted to single-key lookups, Exim was not diagnosing an error if
2776 * or *@ was used with a query-style lookup.
2778 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2780 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2781 long ago, but noone interested enough thought of it. Let's just say that
2782 the "1.1" means that there are some new functions that weren't there at
2783 some point in the past.
2785 PH/51 Error processing for expansion failure of helo_data from an smtp
2786 transport during callout processing was broken.
2788 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2789 tested/used via the -bh/-bhc/-bs options.
2791 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2792 bug, fixed in subsequent PCRE releases).
2794 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2795 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2797 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2803 SC/01 Use a glob alias rather than an array ref in eximstats generated
2804 parser. This improves both readability and performance.
2806 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2807 Don't display local sender or destination tables in eximstats unless
2808 there is data to show.
2809 Added average volumes into the eximstats top table text output.
2811 SC/03 Collect data on the number of addresses (recipients) as well
2812 as the number of messages in eximstats.
2814 TF/01 Correct an error in the documentation for the redirect router. Exim
2815 does (usually) call initgroups() when daemonizing.
2817 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2818 with consistent privilege compared to when running as a daemon.
2820 TF/03 Note in the spec that $authenticated_id is not set for local
2821 submissions from trusted users.
2823 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2824 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2826 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2827 by adding some example configuration directives to the default
2828 configuration file. A little bit of work is required to uncomment the
2829 directives and define how usernames and passwords are checked, but
2830 there is now a framework to start from.
2832 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2833 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2834 without this. I don't know how relevant this is to other LDAP libraries.
2836 PH/02 Add the verb name to the "unknown ACL verb" error.
2838 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2840 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2842 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2843 directory not expanded when it should be if an expanded home directory
2844 was set for the address (which is overridden by the transport).
2846 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2849 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2850 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2851 because it is too late at that time, and has no effect.
2853 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2854 security issue with \' (bugzilla #107). I could not use the
2855 PQescapeStringConn() function, because it needs a PGconn value as one of
2858 PH/08 When testing addresses using -bt, indicate those final addresses that
2859 are duplicates that would not cause an additional delivery. At least one
2860 person was confused, thinking that -bt output corresponded to deliveries.
2861 (Suppressing duplicates isn't a good idea as you lose the information
2862 about possibly different redirections that led to the duplicates.)
2864 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2865 systems where poll() doesn't work, in particular OS X.
2867 PH/10 Added more information to debugging output for retry time not reached.
2869 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2870 operations in malware.c.
2872 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2875 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2876 syslog_duplication set false, log lines that would normally be written
2877 both the the main log and to the reject log were not written to syslog at
2880 PH/14 In the default configuration, change the use of "message" in ACL warn
2881 statements to "add_header".
2883 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2884 not followed by a command (e.g. "seen endif").
2886 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2887 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2890 PH/17 Added extra conditions to the default value of delay_warning_condition
2894 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2895 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2896 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2899 The Auto-Submitted: and various List- headers are standardised, whereas I
2900 don't think Precedence: ever was.
2902 PH/18 Refactored debugging code in route_finduser() to show more information,
2903 in particular, the error code if getpwnam() issues one.
2905 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2906 This is apparently needed in addition to the PH/07 change above to avoid
2907 any possible encoding problems.
2909 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2910 but not after initializing Perl.
2912 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2913 output them only if debugging. By default they are written stderr,
2914 apparently, which is not desirable.
2916 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2919 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2922 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2924 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2925 authenticated or an ident call has been made. Suppress the default
2926 values for $authenticated_id and $authenticated_sender (but permit -oMai
2927 and -oMas) when testing with -bh.
2929 PH/24 Re-jigged the order of the tests in the default configuration so that the
2930 tests for valid domains and recipients precede the DNS black list and CSA
2931 tests, on the grounds that those ones are more expensive.
2933 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2934 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2935 This bug exists in every version of Exim that I still have, right back to
2938 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2939 However, an attempt to turn on case-sensitivity in a regex key by
2940 including (?-i) didn't work because the subject string was already
2941 lowercased, and the effects were non-intuitive. It turns out that a
2942 one-line patch can be used to allow (?-i) to work as expected.
2948 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2949 other effects) broke the use of negated acl sub-conditions.
2951 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2954 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2955 "Deny" causes Exim to reject the incoming connection with a 554 error.
2956 Unfortunately, if there is a major crisis, such as a disk failure,
2957 tcp-wrappers gives "deny", whereas what one would like would be some
2958 kind of temporary error. A kludge has been added to help with this.
2959 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2960 554 error is used if errno is still zero or contains ENOENT (which occurs
2961 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2964 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2966 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2967 errors. Otherwise a message that provokes a temporary error (when other
2968 messages do not) can cause a whole host to time out.
2970 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2971 addresses were routed directly to files or pipes from a redirect router.
2972 File deliveries just didn't batch; pipe deliveries might have suffered
2975 PH/06 A failure to get a lock for a hints database would erroneously always say
2976 "Failed to get write lock", even when it was really a read lock.
2978 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2979 of 0600. This has been changed to use the value of the lockfile_mode
2980 option (which defaults to 0600).
2982 PH/08 Applied small patch from the Sieve maintainer.
2984 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2985 folder from quota calculations, a direct delivery into this folder messed
2986 up the contents of the maildirsize file. This was because the regex was
2987 used only to exclude .Trash (or whatever) when the size of the mailbox
2988 was calculated. There was no check that a delivery was happening into an
2989 excluded directory. This bug has been fixed by ignoring all quota
2990 processing for deliveries into excluded directories.
2992 PH/10 Added the maildirfolder_create_regex option to appendfile.
2998 PH/01 The code for finding all the local interface addresses on a FreeBSD
2999 system running IPv6 was broken. This may well have applied to all BSD
3000 systems, as well as to others that have similar system calls. The broken
3001 code found IPv4 interfaces correctly, but gave incorrect values for the
3002 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3003 that it would not match correctly against @[] and not recognize the IPv6
3006 PH/02 The ipliteral router was not recognizing addresses of the form user@
3007 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3009 PH/03 Added disable_ipv6.
3011 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3012 decoded form, because it is most often used to construct To: headers
3013 lines in autoreplies, and the decoded form may well be syntactically
3014 invalid. However, $reply_address has leading white space removed, and all
3015 newlines turned into spaces so that the autoreply transport does not
3018 PH/05 If group was specified without a user on a router, and no group or user
3019 was specified on a transport, the group from the router was ignored.
3021 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3022 for visible compile-time settings that can be used to change these
3023 numbers, for those that want even more. Backwards compatibility with old
3024 spool files has been maintained. However, going back to a previous Exim
3025 release will lost any variables that are in spool files.
3027 PH/07 Two small changes when running in the test harness: increase delay when
3028 passing a TCP/IP connection to a new process, in case the original
3029 process has to generate a bounce, and remove special handling of
3030 127.0.0.2 (sic), which is no longer necessary.
3032 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3033 be the same on different OS.
3035 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3038 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3039 whether --show-vars was specified or not
3041 JJ/02 exipick: Added support for new ACL variable spool format introduced
3044 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3045 syntactically invalid From: or Reply-to: line, and a filter used this to
3046 generate an autoreply, and therefore failed to obtain an address for the
3047 autoreply, Exim could try to deliver to a non-existent relative file
3048 name, causing unrelated and misleading errors. What now happens is that
3049 it logs this as a hard delivery error, but does not attempt to create a
3052 PH/11 The exinext utility has a -C option for testing purposes, but although
3053 the given file was scanned by exinext itself; it wasn't being passed on
3054 when Exim was called.
3056 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3057 an end-of-file indication when reading a command response.
3059 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3060 compiled. In many other places in Exim, IPv6 addresses are always
3061 recognized, so I have changed this. It also means that IPv4 domain
3062 literals of the form [IPV4:n.n.n.n] are now always recognized.
3064 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3065 used if the router is not running as root, for example, when verifying at
3066 ACL time, or when using -bh. The debugging output from this situation was
3067 non-existent - all you got was a failure to exec. I have made two
3070 (a) Failures to set uid/gid, the current directory, or a process leader
3071 in a subprocess such as that created by queryprogram now generate
3072 suitable debugging output when -d is set.
3074 (b) The queryprogram router detects when it is not running as root,
3075 outputs suitable debugging information if -d is set, and then runs
3076 the subprocess without attempting to change uid/gid.
3078 PH/15 Minor change to Makefile for building test_host (undocumented testing
3081 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3082 additional section of a DNS packet that returns MX or SRV records.
3083 Instead, it always explicitly searches for A/AAAA records. This avoids
3084 major problems that occur when a DNS server includes only records of one
3085 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3086 fixed another bug: if SRV records were looked up and the corresponding
3087 address records were *not* found in the additional section, the port
3088 values from the SRV records were lost.
3090 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3091 using the correct key (the original address) when searching the retry
3092 rules in order to find which one to use for generating the retry hint.
3094 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3095 adding the default one. Similarly, if it contains a Reply-To: header, the
3096 errors_reply_to option, if set, is not used.
3098 PH/19 When calculating a retry time, Exim used to measure the "time since
3099 failure" by looking at the "first failed" field in the retry record. Now
3100 it does not use this if it is later than than the arrival time of the
3101 message. Instead it uses the arrival time. This makes for better
3102 behaviour in cases where some deliveries succeed, thus re-setting the
3103 "first failed" field. An example is a quota failure for a huge message
3104 when small messages continue to be delivered. Without this change, the
3105 "time since failure" will always be short, possible causing more frequent
3106 delivery attempts for the huge message than are intended.
3107 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3109 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3110 $1, $2, $3) because the numerical variables can be reset during some
3111 expansion items (e.g. "match"), thereby losing the authentication data.
3113 PH/21 Make -bV show the size of off_t variables so that the test suite can
3114 decide whether to run tests for quotas > 2G.
3116 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3117 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3118 filecount value is greater than 2G or if a quota value is greater than 2G
3119 on a system where the size of off_t is not greater than 4, a panic error
3122 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3123 never match. The debug and -bh output now contains an explicit error
3124 message indicating a malformed IPv4 address or mask.
3126 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3127 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3128 PH/23 above applies.
3130 PH/25 Do not write to syslog when running in the test harness. The only
3131 occasion when this arises is a failure to open the main or panic logs
3132 (for which there is an explicit test).
3134 PH/26 Added the /no_tell option to "control=freeze".
3136 PH/27 If a host name lookup failed very early in a connection, for example, if
3137 the IP address matched host_lookup and the reverse lookup yielded a name
3138 that did not have a forward lookup, an error message of the form "no IP
3139 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3140 could be logged. Now it outputs the IP address instead of "NULL".
3142 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3143 allows the sender and the authenticated sender to be set when
3144 submitting a message from within Exim. Since child_open_exim() is
3145 documented for local_scan(), the new function should be too.
3147 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3148 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3149 results in an empty string is now treated as unset.
3151 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3153 PH/31 Added sender_verify_fail logging option.
3155 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3156 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3157 tidied the source and removed it altogether.
3159 PH/33 When a queue run was abandoned because the load average was too high, a
3160 log line was always written; now it is written only if the queue_run log
3161 selector is set. In addition, the log line for abandonment now contains
3162 information about the queue run such as the pid. This is always present
3163 in "start" and "stop" lines but was omitted from the "abandon" line.
3165 PH/34 Omit spaces between a header name and the colon in the error message that
3166 is given when verify = headers_syntax fails (if there are lots of them,
3167 the message gets confusing).
3169 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3170 names, as there are now some PTR records that contain slashes. This check
3171 is only to protect against broken name servers that fall over on strange
3172 characters, so the fact that it applies to all lookups doesn't matter.
3174 PH/36 Now that the new test suite is complete, we can remove some of the
3175 special code in Exim that was needed for the old test suite. For example,
3176 sorting DNS records because real resolvers return them in an arbitrary
3177 order. The new test suite's fake resolver always returns records in the
3180 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3181 bounces) except when queue_only is set, to avoid logging races between
3182 the different processes.
3184 PH/38 Panic-die if .include specifies a non-absolute path.
3186 PH/39 A tweak to the "H" retry rule from its user.
3188 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3189 a label. They prevented compilation on older perls.
3191 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3192 a warning to be raised on newish perls.
3194 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3195 on queue. Changes to match documented behaviour of showing count of
3196 messages matching specified criteria.
3198 PH/40 Changed the default ident timeout from 30s to 5s.
3200 PH/41 Added support for the use of login_cap features, on those BSD systems
3201 that have them, for controlling the resources used by pipe deliveries.
3203 PH/42 The content-scanning code uses fopen() to create files in which to put
3204 message data. Previously it was not paying any attention to the mode of
3205 the files. Exim runs with umask(0) because the rest of the code creates
3206 files with open(), and sets the required mode explicitly. Thus, these
3207 files were ending up world-writeable. This was not a big issue, because,
3208 being within the spool directory, they were not world-accessible. I have
3209 created a function called modefopen, which takes an additional mode
3210 argument. It sets umask(777), creates the file, chmods it to the required
3211 mode, then resets the umask. All the relevant calls to fopen() in the
3212 content scanning code have been changed to use this function.
3214 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3215 to 24 hours. This avoids potential overflow problems when processing G
3216 and H retry rules. I suspect nobody ever tinkers with this value.
3218 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3220 PH/45 When the plaintext authenticator is running as a client, the server's
3221 challenges are checked to ensure they are valid base64 strings. By
3222 default, the authentication attempt is cancelled if an invalid string is
3223 received. Setting client_ignore_invalid_base64 true ignores these errors.
3224 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3225 they are received. Thus, the responses can be made to depend on the
3226 challenges. If an invalid string is ignored, an empty string is placed in
3229 PH/46 Messages that are created by the autoreply transport now contains a
3230 References: header, in accordance with RFCs 2822 and 3834.
3232 PH/47 Added authenticated_sender_force to the smtp transport.
3234 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3236 PH/49 Installed latest patch from the Sieve maintainer.
3238 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3239 was also set, the appendfile transport was unnecessarily scanning a
3240 directory of message files (e.g. for maildir delivery) to find the count
3241 of files (along with the size), even though it did not need this
3242 information. It now does the scan only if it needs to find either the
3243 size of the count of files.
3245 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3247 PH/52 Two bugs concerned with error handling when the smtp transport is
3250 (i) Exim was not creating retry information for temporary errors given
3251 for individual recipients after the DATA command when the smtp transport
3252 was used in LMTP mode. This meant that they could be retried too
3253 frequently, and not timed out correctly.
3255 (ii) Exim was setting the flag that allows error details to be returned
3256 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3257 recipients that were returned after the DATA command.
3259 PH/53 This is related to PH/52, but is more general: for any failing address,
3260 when detailed error information was permitted to be returned to the
3261 sender, but the error was temporary, then after the final timeout, only
3262 "retry timeout exceeded" was returned. Now it returns the full error as
3263 well as "retry timeout exceeded".
3265 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3266 do this, and (what is worse) MTAs that accept it.
3268 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3269 will now be deprecated.
3271 PH/56 New os.c-cygwin from the Cygwin maintainer.
3273 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3274 formats (previously only available in exim formats via -bpr, -bpru,
3275 and -bpra. Now also available in native and exiqgrep formats)
3277 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3278 with very large, slow to parse queues
3280 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3282 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3284 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3285 responses to authentication challenges, though it was showing the
3286 challenges; (ii) I've removed the CR characters from the debug output for
3289 PH/58 Allow for the insertion of a newline as well as a space when a string
3290 is turned into more than one encoded-word during RFC 2047 encoding. The
3291 Sieve code now uses this.
3293 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3294 data_4xx, lost_connection, tls_required.
3296 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3297 message was being sent as an SMTP response.
3299 PH/61 Add -l and -k options to exicyclog.
3301 PH/62 When verifying, if an address was redirected to one new address, so that
3302 verification continued, and the new address failed or deferred after
3303 having set something in $address_data, the value of $address_data was not
3304 passed back to the ACL. This was different to the case when no
3305 redirection occurred. The value is now passed back in both cases.
3307 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3308 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3309 use login_cap.h, so on its own it isn't the distinguishing feature. The
3310 new name refers directly to the setclassresources() function.
3312 PH/65 Added configuration files for NetBSD3.
3314 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3316 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3317 is preferred over IPv4.
3319 PH/68 The bounce_return_message and bounce_return_body options were not being
3320 honoured for bounces generated during the reception of non-SMTP messages.
3321 In particular, this applied to messages rejected by the ACL. This bug has
3322 been fixed. However, if bounce_return_message is true and bounce_return_
3323 body is false, the headers that are returned for a non-SMTP message
3324 include only those that have been read before the error was detected.
3325 (In the case of an ACL rejection, they have all been read.)
3327 PH/69 The HTML version of the specification is now built in a directory called
3328 spec_html instead of spec.html, because the latter looks like a path with
3329 a MIME-type, and this confuses some software.
3331 PH/70 Catch two compiler warnings in sieve.c.
3333 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3334 function verify_get_ident() calls ip_connect() to connect a socket, but
3335 if the "connect()" function timed out, ip_connect() used to close the
3336 socket. However, verify_get_ident() also closes the socket later, and in
3337 between Exim writes to the log, which may get opened at this point. When
3338 the socket was closed in ip_connect(), the log could get the same file
3339 descriptor number as the socket. This naturally causes chaos. The fix is
3340 not to close the socket in ip_connect(); the socket should be closed by
3341 the function that creates it. There was only one place in the code where
3342 this was missing, in the iplookup router, which I don't think anybody now
3343 uses, but I've fixed it anyway.
3345 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3346 well as to direct DNS lookups. Otherwise the handling of names in host
3347 lists is inconsistent and therefore confusing.
3353 PH/01 Two changes to the default runtime configuration:
3355 (1) Move the checks for relay_from_hosts and authenticated clients from
3356 after to before the (commented out) DNS black list checks.
3358 (2) Add control=submission to the relay_from_hosts and authenticated
3359 clients checks, on the grounds that messages accepted by these
3360 statements are most likely to be submissions.
3362 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3364 (1) Generate an error if the third argument for the ${prvs expansion is
3367 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3370 (3) Reset the variables that are obtained from the first argument of
3371 ${prvscheck and used in the second argument before leaving the code,
3372 because their memory is reclaimed, so using them afterwards may do
3375 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3376 one (it's much easier than Tom thought :-).
3378 (5) Because of (4), we can now allow for the use of $prvscheck_result
3379 inside the third argument.
3381 PH/03 For some reason, the default setting of PATH when running a command from
3382 a pipe transport was just "/usr/bin". I have changed it to
3385 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3386 anything to be listed in the output from -bV.
3388 PH/05 When a filter generated an autoreply, the entire To: header line was
3389 quoted in the delivery log line, like this:
3391 => >A.N.Other <ano@some.domain> <original@ddress> ...
3393 This has been changed so that it extracts the operative address. There
3394 may be more than one such address. If so, they are comma-separated, like
3397 => >ano@some.domain,ona@other.domain <original@ddress> ...
3399 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3400 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3401 not being looked up in the rDNS to get a host name, Exim was showing the
3402 IP address twice in Received: lines, even though the IP addresses were
3403 identical. For example:
3405 Received: from [1.2.3.4] (helo=[1.2.3.4])
3407 However, if the real host name was known, it was omitting the HELO data
3408 if it matched the actual IP address. This has been tidied up so that it
3409 doesn't show the same IP address twice.
3411 PH/07 When both +timestamp and +memory debugging was on, the value given by
3412 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3413 called by the debug printing, thereby overwriting the timestamp buffer.
3414 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3416 PH/08 When the original message was included in an autoreply transport, it
3417 always said "this is a copy of the message, including all the headers",
3418 even if body_only or headers_only was set. It now gives an appropriate
3421 PH/09 Applied a patch from the Sieve maintainer which:
3423 o fixes some comments
3424 o adds the (disabled) notify extension core
3425 o adds some debug output for the result of if/elsif tests
3426 o points to the current vacation draft in the documentation
3427 and documents the missing references header update
3431 o fixes a bug in processing the envelope test (when testing
3432 multiple envelope elements, the last element determined the
3435 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3436 Electronic Mail") by including:
3438 Auto-submitted: auto-generated
3440 in the messages that it generates (bounce messages and others, such as
3441 warnings). In the case of bounce messages for non-SMTP messages, there was
3442 also a typo: it was using "Auto_submitted" (underscore instead of
3443 hyphen). Since every message generated by Exim is necessarily in response
3444 to another message, thes have all been changed to:
3446 Auto-Submitted: auto-replied
3448 in accordance with these statements in the RFC:
3450 The auto-replied keyword:
3452 - SHOULD be used on messages sent in direct response to another
3453 message by an automatic process,
3455 - MUST NOT be used on manually-generated messages,
3457 - MAY be used on Delivery Status Notifications (DSNs) and Message
3458 Disposition Notifications (MDNs),
3460 - MUST NOT be used on messages generated by automatic or periodic
3461 processes, except for messages which are automatic responses to
3464 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3465 to the default Received: header definition.
3467 PH/12 Added log selector acl_warn_skipped (default on).
3469 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3470 variables because (a) they are in the wrong storage pool and (b) even if
3471 they were copied, it wouldn't work properly because of the caching.
3473 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3474 checking when decoding. Apparently there are clients that generate
3475 overlong encoded strings. Why am I not surprised?
3477 PH/15 If the first argument of "${if match_address" was not empty, but did not
3478 contain an "@" character, Exim crashed. Now it writes a panic log message
3479 and treats the condition as false.
3481 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3483 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3484 extension "envelope-auth". The code is finished and in agreement with
3485 other implementations, but there is no documentation so far and in fact,
3486 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3487 not changing the active code.
3489 Print executed "if" and "elsif" statements when debugging is used. This
3490 helps a great deal to understand what a filter does.
3492 Document more things not specified clearly in RFC3028. I had all this
3493 sorted out, when out of a sudden new issues came to my mind. Oops."
3495 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3498 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3499 canonical form (as documented). However, after a host name lookup from
3500 the IP address, check_host() was doing a simple string comparison with
3501 addresses acquired from the DNS when checking that the found name did
3502 have the original IP as one of its addresses. Since any found IPv6
3503 addresses are likely to be in abbreviated form, the comparison could
3504 fail. Luckily, there already exists a function for doing the comparison
3505 by converting both addresses to binary, so now that is used instead of
3506 the text comparison.
3508 PH/20 There was another similar case to PH/19, when a complete host name was
3509 given in a host list; looking up its IP address could give an abbreviated
3510 form, whereas the current host's name might or might not be abbreviated.
3511 The same fix has been applied.
3517 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3518 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3521 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3522 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3524 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3526 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3527 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3528 Also, for "Auto-Submitted", treat anything other than "no" as
3529 non-personal, in accordance with RFC 3834. (Previously it treated
3530 anything starting "auto-" as non-personal.)
3532 TF/01 The control=submission/name=... option had a problem with syntax
3533 errors if the name included a slash character. The /name= option
3534 now slurps the rest of the string, so it can include any characters
3535 but it must come last in the list of options (after /sender_retain
3538 PH/05 Some modifications to the interface to the fake nameserver for the new
3546 TK/01 Added the "success_on_redirect" address verification option. See
3547 NewStuff for rationale and an example.
3549 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3551 PH/02 Patch to exigrep to allow it to work on syslog lines.
3553 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3554 fread() to skip over the body file's header line, because in Cygwin the
3555 header line is locked and is inaccessible.
3557 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3558 co-exist for some time) to make it clear that it is the Exim ID that is
3559 referenced, not the Message-ID: header line.
3561 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3562 string_format() function, because snprintf() does not exist on all
3565 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3566 ${stat: expansion item.
3568 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3569 protocol synchronization error", to keep the pedants happy.
3571 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3572 well as for IRIX systems, when gcc is being used. See the host.c source
3575 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3577 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3580 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3581 transport and to the smtp transport in LMTP mode.
3583 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3585 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3586 run a filter in a subprocess. This could lead to confusion in subsequent
3587 lookups in the parent process. There should also be a search_tidyup() at
3588 the end of the subprocess.
3590 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3591 only if the host matched helo_try_verify_hosts, which caused the
3592 verification to occur when the EHLO/HELO command was issued. The ACL just
3593 tested the remembered result. Now, if a previous verification attempt has
3594 not happened, "verify = helo" does it there and then.
3596 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3598 TK/03 Fix log output including CR from clamd.
3600 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3601 exist provoked a memory error which could cause a segfault.
3603 PH/15 Installed PCRE 6.2
3605 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3607 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3608 of the problem. Specifically, suggested +O2 rather than +O1 for the
3611 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3613 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3614 record, the port was not correctly logged when the outgoing_port log
3615 selector was set (it logged the transort's default port).
3617 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3618 fallback_hosts, and "hosts" in the smtp transport.
3620 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3621 host errors such as "Connection refused".
3623 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3624 authentication with radiusclient 0.4.9:
3626 - Error returned from rc_read_config was caught wrongly
3627 - Username/password not passed on to radius server due to wrong length.
3629 The presumption is that some radiusclient API changes for 4.51/PH/17
3630 were not taken care of correctly. The code is still untested by me (my
3631 Linux distribution still has 0.3.2 of radiusclient), but it was
3632 contributed by a Radius user.
3634 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3635 expanding the "port" option of the smtp transport.
3637 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3638 while reading a MIME header. Thanks to Tom Hughes for a patch.
3640 PH/24 Include config.h inside local_scan.h so that configuration settings are
3643 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3644 that in an ACL for RCPT (for example), you can examine exactly what was
3647 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3648 commands, but it was not correctly comparing the address with the actual
3649 client host address. Thus, it would show the EHLO address in Received:
3650 header lines when this was not necessary.
3652 PH/27 Added the % operator to ${eval:}.
3654 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3655 it should be ignoring failures (because with -C, for example, it has lost
3656 privilege). It wasn't ignoring creation failures other than "already
3659 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3662 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3663 because an input line was too long, either on its own, or by virtue of
3664 too many continuations, the temporary file was not being removed, and the
3665 return code was incorrect.
3667 PH/31 Missing "BOOL" in function definition in filtertest.c.
3669 PH/32 Applied Sieve patches from the maintainer.
3671 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3673 PH/33 Added "verify = not_blind".
3675 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3676 Local/Makefile (with some defaults set). These are used in built scripts
3677 such as exicyclog, but they have never been used in the exim_install
3678 script (though there are many overriding facilities there). I have
3679 arranged that the exim_install script now takes note of these two
3682 PH/35 Installed configuration files for Dragonfly.
3684 PH/36 When a locally submitted message by a trusted user did not contain a
3685 From: header, and the sender address was obtained from -f or from an SMTP
3686 MAIL command, and the trusted user did not use -F to supply a sender
3687 name, $originator_name was incorrectly used when constructing a From:
3688 header. Furthermore, $originator_name was used for submission mode
3689 messages from external hosts without From: headers in a similar way,
3690 which is clearly wrong.
3692 PH/37 Added control=suppress_local_fixups.
3694 PH/38 When log_selector = +received_sender was set, and the addition of the
3695 sender made the log line's construction buffer exactly full, or one byte
3696 less than full, an overflow happened when the terminating "\n" was
3699 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3700 when the result of a list match is failure because a DNS lookup failed.
3702 PH/40 RM_COMMAND is now used in the building process.
3704 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3705 the "build-* directories that it finds.
3707 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3708 address in a domain literal was a prefix of an interface address.
3710 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3711 when verifying a sender address, unless rewrite_headers is false.
3713 PH/44 Wrote a long comment about why errors_to addresses are verified as
3714 recipients, not senders.
3716 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3717 the ratelimit ACL was added.
3719 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3721 PH/46 Added extra information about PostgreSQL errors to the error string.
3723 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3724 avoiding the need to install special zones in a real server. This is
3725 backwards compatible; if it can't find the fake resolver, it drops back.
3726 Thus, both old and new test suites can be run.
3728 TF/02 Added util/ratelimit.pl
3730 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3731 clock is set back in time.
3733 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3734 Candler <B.Candler@pobox.com>.
3736 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3737 OK for addresses that are the result of redirections.
3739 PH/48 A number of further additions for the benefit of the new test suite,
3740 including a fake gethostbyname() that interfaces to the fake DNS resolver
3743 TF/06 The fix for widen_domains has also been applied to qualify_single and
3744 search_parents which are the other dnslookup options that can cause
3747 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3750 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3752 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3753 service" if it encountered an MX record with an empty target hostname.
3754 The message is now "an MX or SRV record indicated no SMTP service".
3756 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3757 if the DNS of the sending site is misconfigured. This is quite a
3758 common situation. This change restores the behaviour of treating a
3759 helo verification defer as a failure.
3761 PH/51 If self=fail was set on a router, the bounce message did not include the
3762 actual error message.
3768 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3770 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3771 command itself ended in error, the underlying message about the transport
3772 filter timeout was being overwritten with the pipe command error. Now the
3773 underlying error message should be appended to the second error message.
3775 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3777 PH/02 Reduce the amount of output that "make" produces by default. Full output
3778 can still be requested.
3780 PH/03 The warning log line about a condition test deferring for a "warn" verb
3781 was being output only once per connection, rather than after each
3782 occurrence (because it was using the same function as for successful
3783 "warn" verbs). This seems wrong, so I have changed it.
3785 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3786 it should not have, which might have caused a crash in the right
3787 circumstances, but probably never did.
3789 PH/04 Installed a modified version of Tony Finch's patch to make submission
3790 mode fix the return path as well as the Sender: header line, and to
3791 add a /name= option so that you can make the user's friendly name appear
3794 TF/03 Added the control = fakedefer ACL modifier.
3796 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3797 Mark Lowes for thorough testing.
3799 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3801 TK/03 Merged latest SRS patch from Miles Wilton.
3803 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3804 with the definition in sysexits.h (which is #included earlier).
3805 Fortunately, Exim does not actually use EX_OK. The code used to try to
3806 preserve the sysexits.h value, by assuming that macro definitions were
3807 scanned for macro replacements. I have been disabused of this notion,
3808 so now the code just undefines EX_OK before #including unistd.h.
3810 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3811 in the smtp transport. When a block could not be written in a single
3812 write() function, the timeout was being re-applied to each part-write.
3813 This seems wrong - if the receiver was accepting one byte at a time it
3814 would take for ever. The timeout is now adjusted when this happens. It
3815 doesn't have to be particularly precise.
3817 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3818 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3820 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3821 without a domain if the check to <postmaster@domain> fails.
3823 SC/01 Eximstats: added -xls and the ability to specify output files
3824 (patch written by Frank Heydlauf).
3826 SC/02 Eximstats: use FileHandles for outputting results.
3828 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3830 SC/04 Eximstats: fixed display of large numbers with -nvr option
3832 SC/05 Eximstats: fixed merging of reports with empty tables.
3834 SC/06 Eximstats: added the -include_original_destination flag
3836 SC/07 Eximstats: removed tabs and trailing whitespace.
3838 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3840 TK/06 MBOX spool code: Add real "From " MBOX separator line
3841 so the .eml file is really in mbox format (even though
3842 most programs do not really care). Patch from Alex Miller.
3844 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3845 The latter is generated from $received_to and is only set if the
3846 message has one envelope recipient. SA can use these headers,
3847 obviously out-of-the-box. Patch from Alex Miller.
3849 PH/08 The ${def test on a variable was returning false if the variable's
3850 value was "0", contrary to what the specification has always said!
3851 The result should be true unless the variable is empty.
3853 PH/09 The syntax error of a character other than { following "${if
3854 def:variable_name" (after optional whitespace) was not being diagnosed.
3855 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3856 accidental colon was present, for example, could give incorrect results.
3858 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3859 result is used (not including appendfile, where other changes are about
3862 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3863 This involved changing a lot of size variables from int to off_t. It
3864 should work with maildirs and everything.
3866 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3867 spamd dying while we are connected to it.
3869 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3872 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3873 function for BDB 4.3.
3875 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3877 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3878 the sources for PCRE have been re-arranged and more files are now
3881 PH/15 The code I had for printing potentially long long variables in PH/11
3882 above was not the best (it lost precision). The length of off_t variables
3883 is now inspected at build time, and an appropriate printing format (%ld
3884 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3885 to be "long long int" or "long int". This is needed for the internal
3886 formatting function string_vformat().
3888 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3889 the configuration file to be ":syslog", then the script "guesses" where
3890 the logs files are, rather than using the compiled in default. In our
3891 case the guess is not the same as the compiled default, so the script
3892 suddenly stopped working when I started to use syslog. The patch checks
3893 to see if log_file_path is "". If so, it attempts to read it from exim
3894 with no configuration file to get the compiled in version, before it
3895 falls back to the previous guessing code."
3897 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3898 implementing BATV in an Exim configuration. See NewStuff for the gory
3901 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3902 Makefile that are specific to HP-UX.
3904 PH/18 If the "use_postmaster" option was set for a recipient callout together
3905 with the "random" option, the postmaster address was used as the MAIL
3906 FROM address for the random test, but not for the subsequent recipient
3907 test. It is now used for both.
3909 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3910 patch removes a few documentation additions to RFC 3028, because the
3911 latest draft now contains them. It adds the new en;ascii-case comparator
3912 and a new error check for 8bit text in MIME parts. Comparator and
3913 require names are now matched exactly. I enabled the subaddress
3914 extension, but it is not well tested yet (read: it works for me)."
3916 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3917 rework some of the code of TK/09 above to avoid the hardwired use of
3918 "%lld" and "long long". Replaced the call to snprintf() with a call to
3921 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3922 records point to non-existent hosts", "retry timeout exceeded", and
3923 "retry time not reached for any host after a long failure period".
3925 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3926 experimental DomainKeys support:
3928 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3929 (2) On an error such as an illegally used "control", the wrong name for
3930 the control was given.
3932 These problems did NOT occur unless DomainKeys support was compiled.
3934 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3936 PH/24 Added ${if match_ip condition.
3938 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3939 databases so that it will be absolutely obvious if a crash occurs in the
3940 DB library. This is a regular occurrence (often caused by mis-matched
3943 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3944 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3945 picked up on a user's system that detects such things. There doesn't seem
3946 to be a gcc warning option for this - only an attribute that has to be
3947 put on the function's prototype. It seems that in Fedora Core 4 they have
3948 set this on a number of new functions. No doubt there will be more in due
3951 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3952 specify a transport. However, if an address that was verified by such a
3953 router was the subject of a callout, Exim crashed because it tried to
3954 read the rcpt_include_affixes from the non-existent transport. Now it
3955 just assumes that the setting of that option is false. This bug was
3956 introduced by 4.51/PH/31.
3958 PH/28 Changed -d+all to exclude +memory, because that information is very
3959 rarely of interest, but it makes the output a lot bigger. People tend to
3960 do -d+all out of habit.
3962 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3963 code in os-type was giving problems when libc.so lives in lib64, like on
3966 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3967 aren't the modern standard, and it seems that some systems' include files
3968 don't always have them. Exim was already checking for some of the newer
3969 ones like T_AAAA, and defining it itself. I've added checks for all the
3970 record types that Exim uses.
3972 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3973 not automatically generating a new one, as it is supposed to. This
3974 prevented TLS from working. If the file did exist, but contained invalid
3975 data, a new version was generated, as expected. It was only the case of a
3976 non-existent file that was broken.
3978 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3979 with a change in libdomainkeys > 0.64.
3981 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3982 from DNS. If the selector record carries the flag, it now has
3983 precedence over the domain-wide flag.
3985 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3987 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3988 the use of a port name that isn't defined in /etc/services) occurred, the
3989 message was deferred as in a normal delivery, and thus remained on the
3990 spool, instead of being failed because of the mua_wrapper setting. This
3991 is now fixed, and I tidied up some of the mua_wrapper messages at the
3994 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3995 the messages in a hash of arrays rather than using individual hashes.
3996 This is a bit cleaner and results in dramatic memory savings, albeit
3997 at a slight CPU cost.
3999 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4000 as requested by Marc Sherman.
4002 SC/10 Eximstats: added histograms for user specified patterns as requested
4005 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4007 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4008 fopen() in the content-scanning modules that did not already have it.
4014 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4015 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4017 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4019 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4021 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4022 to test Sieve filters that use "vacation".
4024 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4025 that changes the way the GnuTLS parameters are stored in the cache file.
4026 The new format can be generated externally. For backward compatibility,
4027 if the data in the cache doesn't make sense, Exim assumes it has read an
4028 old-format file, and it generates new data and writes a new file. This
4029 means that you can't go back to an older release without removing the
4032 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4033 work if there are any delivery delays because "one_time" forces the
4034 parent to be marked "delivered", so its unseen clone is never tried
4035 again. For this reason, Exim now forbids the simultaneous setting of
4038 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4039 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4040 entry if you want to know the details. The fix, however, neglected to
4041 consider the case when local delivery batching is involved. The test for
4042 "previously delivered" was not happening when checking to see if an
4043 address could be batched with a previous (undelivered) one; under
4044 certain circumstances this could lead to multiple deliveries to the same
4047 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4048 in its include files, and this causes problems building Exim.
4050 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4051 header_syntax) but Exim was just ignoring anything given after a slash.
4052 In particular, this caused confusion with an attempt to use "verify =
4053 reverse_host_lookup/defer_ok". An error is now given when options are
4054 supplied for verify items that do not have them. (Maybe reverse_host_
4055 lookup should have a defer_ok option, but that's a different point.)
4057 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4058 defined by RFC 821) to 2048, because there were problems with some AUTH
4059 commands, and RFC 1869 says the size should be increased for extended
4060 SMTP commands that take arguments.
4062 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4065 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4066 "unknown" error; now it says that the functionality isn't in the binary.
4068 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4069 an address' error message when a string expansion fails (syntax or
4070 whatever). Otherwise the password may appear in the log. Following change
4071 PH/42 below, there is no longer a chance of it appearing in a bounce
4074 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4076 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4077 of its addresses was ever tried. (Bugzilla bug #2.)
4079 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4080 the result incorrectly in the debug output. (It correctly added a newline
4081 to what was transported.)
4083 TF/01 Added $received_time.
4085 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4086 commented out examples of how to interface to a virus scanner and to
4087 SpamAssassin. Also added commented examples of av_scanner and
4088 spamd_address settings.
4090 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4091 and controls are allowed in which ACLs. There were a couple of minor
4092 errors. Some of the entries in the conditions table (which is a table of
4093 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4094 negation of where the condition IS allowed.
4096 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4098 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4099 header file does not have a version number, so I've had to invent a new
4100 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4101 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4102 radiusclient), but it was contributed by a Radius user.
4104 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4105 files or directories, for OpenSSL.
4107 PH/19 When an Exim process that is running as root has to create an Exim log
4108 file, it does so in a subprocess that runs as exim:exim so as to get the
4109 ownership right at creation (otherwise, other Exim processes might see
4110 the file with the wrong ownership). There was no test for failure of this
4111 fork() call, which would lead to the process getting stuck as it waited
4112 for a non-existent subprocess. Forks do occasionally fail when resources
4113 run out. I reviewed all the other calls to fork(); they all seem to check
4116 PH/20 When checking for unexpected SMTP input at connect time (before writing
4117 the banner), Exim was not dealing correctly with a non-positive return
4118 from the read() function. If the client had disconnected by this time,
4119 the result was a log entry for a synchronization error with an empty
4120 string after "input=" when read() returned zero. If read() returned -1
4121 (an event I could not check), uninitialized data bytes were printed.
4122 There were reports of junk text (parts of files, etc) appearing after
4125 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4127 PH/22 Added support for macro redefinition, and (re)definition in between
4128 driver and ACL definitions.
4130 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4131 forgetting to use the resulting value; it was using the unexpanded value.
4133 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4134 hadn't been configured. The fix is from Juergen Kreileder, who
4135 understands it better than I do:
4137 "Here's what I see happening with three configured cyrus_sasl
4138 authenticators configured (plain, login, cram-md5):
4140 On startup auth_cyrus_sasl_init() gets called for each of these.
4141 This means three calls to sasl_listmech() without a specified mech_list.
4142 => SASL tests which mechs of all available mechs actually work
4143 => three warnings about OTP not working
4144 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4146 With the patch, sasl_listmech() also gets called three times. But now
4147 SASL's mech_list option is set to the server_mech specified in the the
4148 authenticator. Or in other words, the answer from sasl_listmech()
4149 gets limited to just the mech you're testing for (which is different
4151 => the return list contains just 'plain' or 'login', 'cram-md5' or
4152 nothing depending on the value of ob->server_mech.
4154 I've just tested the patch: Authentication still works fine,
4155 unavailable mechs specified in the exim configuration are still
4156 caught, and the auth.log warnings about OTP are gone."
4158 PH/25 When debugging is enabled, the contents of the command line are added
4159 to the debugging output, even when log_selector=+arguments is not
4162 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4163 answer is "GNU", and only if the return is "GNU/something" is the answer
4166 PH/27 $acl_verify_message is now set immediately after the failure of a
4167 verification in an ACL, and so is available in subsequent modifiers. In
4168 particular, the message can be preserved by coding like this:
4170 warn !verify = sender
4171 set acl_m0 = $acl_verify_message
4173 Previously, $acl_verify_message was set only while expanding "message"
4174 and "log_message" when a very denied access.
4176 PH/28 Modified OS/os.c-Linux with
4178 -#ifndef OS_LOAD_AVERAGE
4179 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4181 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4182 nomenclature these days.)
4184 PH/29 Installed patch from the Sieve maintainer that adds the options
4185 sieve_useraddress and sieve_subaddress to the redirect router.
4187 PH/30 In these circumstances:
4188 . Two addresses routed to the same list of hosts;
4189 . First host does not offer TLS;
4190 . First host accepts first address;
4191 . First host gives temporary error to second address;
4192 . Second host offers TLS and a TLS session is established;
4193 . Second host accepts second address.
4194 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4195 and peerdn, if requested) that were in fact used only for the second
4198 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4199 attention to any local part prefix or suffix that was matched by the
4200 router that accepted the address. It now behaves in the same way as it
4201 does for delivery: the affixes are removed from the local part unless
4202 rcpt_include_affixes is set on the transport.
4204 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4205 timeout during the DATA phase of an incoming message.
4207 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4208 applied a patch sanctioned by the Sieve maintainer.
4210 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4211 the uid or gid is negative. A case of a negative gid caused this to be
4212 noticed. The fix allows for either to be negative.
4214 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4215 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4218 PH/36 The change PH/12 above was broken. Fixed it.
4220 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4221 the grounds that routing the same address twice would always produce the
4222 same answer. This might have been true once, but it is certainly no
4223 longer true now. Routing a child address may depend on the previous
4224 routing that produced that child. Some complicated redirection strategies
4225 went wrong when messages had multiple recipients, and made Exim's
4226 behaviour dependent on the order in which the addresses were given.
4228 I have moved the duplicate checking until after the routing is complete.
4229 Exim scans the addresses that are assigned to local and remote
4230 transports, and removes any duplicates. This means that more work will be
4231 done, as duplicates will always all be routed, but duplicates are
4232 presumably rare, so I don't expect this is of any significance.
4234 For deliveries to pipes, files, and autoreplies, the duplicate checking
4235 still happens during the routing process, since they are not going to be
4238 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4239 It corrects a timeout issue with spamd. This is Ian's comment: "The
4240 background is that sometimes spamd either never reads data from a
4241 connection it has accepted, or it never writes response data. The exiscan
4242 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4243 blindly assumes that writes won't block so it may never time out."
4245 PH/39 Allow G after quota size as well as K and M.
4247 PH/40 The value set for $authenticated_id in an authenticator may not contain
4248 binary zeroes or newlines because the value is written to log lines and
4249 to spool files. There was no check on this. Now the value is run through
4250 the string_printing() function so that such characters are converted to
4251 printable escape sequences.
4253 PH/41 $message_linecount is a new variable that contains the total number of
4254 lines in the message. Compare $body_linecount, which is the count for the
4257 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4258 bounce and delay warning messages, except in certain special cases, which
4261 (a) An SMTP error message from a remote host;
4262 (b) A message specified in a :fail: redirection;
4263 (c) A message specified in a "fail" command in a system filter;
4264 (d) A message specified in a FAIL return from the queryprogram router;
4265 (e) A message specified by the cannot_route_message router option.
4267 In these cases only, Exim does include the error details in bounce and
4268 warning messages. There are also a few cases where bland messages such
4269 as "unrouteable address" or "local delivery error" are given.
4271 PH/43 $value is now also set for the "else" part of a ${run expansion.
4273 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4274 being worked on, but at least Exim now implements the latest version to
4277 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4278 process to complete was treated as a delivery failure, a timeout while
4279 writing the message to the pipe was logged, but erroneously treated as a
4280 successful delivery. Such timeouts include transport filter timeouts. For
4281 consistency with the overall process timeout, these timeouts are now
4282 treated as errors, giving rise to delivery failures by default. However,
4283 there is now a new Boolean option for the pipe transport called
4284 timeout_defer, which, if set TRUE, converts the failures into defers for
4285 both kinds of timeout. A transport filter timeout is now identified in
4288 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4289 systems where "make" and "gmake" are different, calling "gmake" at top
4290 level broke things. I've arranged for the value of $(MAKE) to be passed
4291 from the Makefile to this script so that it can call the same version of
4295 A note about Exim versions 4.44 and 4.50
4296 ----------------------------------------
4298 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4299 changes of various kinds. As a consequence, a big documentation update was
4300 needed. This delayed the release for rather longer than seemed good, especially
4301 in the light of a couple of (minor) security issues. Therefore, the changes
4302 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4303 release. So 4.44 and 4.50 are in effect two different branches that both start
4306 I have left the 4.50 change log unchanged; it contains all the changes since
4307 4.43. The change log for 4.44 is below; many of its items are identical to
4308 those for 4.50. This seems to be the most sensible way to preserve the
4309 historical information.
4315 1. Minor wording change to the doc/README.SIEVE file.
4317 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4318 computation of the current number of files was incorrect.
4320 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4321 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4324 4. Give more explanation in the error message when the command for a transport
4325 filter fails to execute.
4327 5. There are several places where Exim runs a non-Exim command in a
4328 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4329 was being done only for the command run by the queryprogram router. It is
4330 now done for all such subprocesses. The other cases are: ${run, transport
4331 filters, and the commands run by the lmtp and pipe transports.
4333 6. Added CONFIGURE_GROUP build-time option.
4335 7. Some older OS have a limit of 256 on the maximum number of file
4336 descriptors. Exim was using setrlimit() to set 1000 as a large value
4337 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4338 systems. I've change it so that if it can't get 1000, it tries for 256.
4340 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4341 was an oversight, and furthermore, ever since the addition of extra
4342 controls (e.g. 4.43/32), the checks on when to allow different forms of
4343 "control" were broken. There should now be diagnostics for all cases when a
4344 control that does not make sense is encountered.
4346 9. Added the /retain_sender option to "control=submission".
4348 10. $recipients is now available in the predata ACL (oversight).
4350 11. Tidy the search cache before the fork to do a delivery from a message
4351 received from the command line. Otherwise the child will trigger a lookup
4352 failure and thereby defer the delivery if it tries to use (for example) a
4353 cached ldap connection that the parent has called unbind on.
4355 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4356 of $address_data from the recipient verification was clobbered by the
4357 sender verification.
4359 13. The value of address_data from a sender verification is now available in
4360 $sender_address_data in subsequent conditions in the ACL statement.
4362 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4364 15. Added a new option "connect=<time>" to callout options, to set a different
4367 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4368 was its contents. (It was OK if the option was not defined at all.)
4370 17. A "Completed" log line is now written for messages that are removed from
4371 the spool by the -Mrm option.
4373 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4374 information about exactly what failed.
4376 19. Added -dd to debug only the daemon process.
4378 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4379 handles timeouts, both on the server side and network timeouts. Renamed the
4380 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4382 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4383 It is now set to "smtps".
4385 22. $host_address is now set to the target address during the checking of
4386 ignore_target_hosts.
4388 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4389 being passed; this would have caused $sender_host_name to have been used if
4390 matching the list had actually called for a host name (not very likely,
4391 since this list is usually IP addresses). A host name is now passed as
4394 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4395 code that specifies a non-restarting handler (typically sigaction() in
4396 modern systems) in an attempt to fix a rare and obscure crash bug.
4398 25. Narrowed the window for a race in the daemon that could cause it to ignore
4399 SIGCHLD signals. This is not a major problem, because they are used only to
4400 wake it up if nothing else does.
4402 26. A malformed maildirsize file could cause Exim to calculate negative values
4403 for the mailbox size or file count. Odd effects could occur as a result.
4404 The maildirsize information is now recalculated if the size or filecount
4407 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4408 support for a long time. Removed HAVE_SYS_VFS_H.
4410 28. Installed the latest version of exipick from John Jetmore.
4412 29. In an address list, if the pattern was not a regular expression, an empty
4413 subject address (from a bounce message) matched only if the pattern was an
4414 empty string. Non-empty patterns were not even tested. This was the wrong
4415 because it is perfectly reasonable to use an empty address as part of a
4416 database query. An empty address is now tested by patterns that are
4417 lookups. However, all the other forms of pattern expect the subject to
4418 contain a local part and a domain, and therefore, for them, an empty
4419 address still always fails if the pattern is not itself empty.
4421 30. Exim went into a mad DNS loop when attempting to do a callout where the
4422 host was specified on an smtp transport, and looking it up yielded more
4423 than one IP address.
4425 31. Re-factored the code for checking spool and log partition space into a
4426 function that finds that data and another that does the check. The former
4427 is then used to implement four new variables: $spool_space, $log_space,
4428 $spool_inodes, and $log_inodes.
4430 32. The RFC2047 encoding function was originally intended for short strings
4431 such as real names; it was not keeping to the 75-character limit for
4432 encoded words that the RFC imposes. It now respects the limit, and
4433 generates multiple encoded words if necessary. To be on the safe side, I
4434 have increased the buffer size for the ${rfc2047: expansion operator from
4437 33. It is now permitted to omit both strings after an "if" condition; if the
4438 condition is true, the result is "true". As before, when the second string
4439 is omitted, a false condition yields an empty string. This makes it less
4440 cumbersome to write custom ACL and router conditions.
4442 34. Failure to deliver a bounce message always caused it to be frozen, even if
4443 there was an errors_to setting on the router. The errors_to setting is now
4446 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4447 canonical form (fully expanded) before being placed in
4448 $sender_host_address.
4450 36. The table in the code that translates DNS record types into text (T_A to
4451 "A" for instance) was missing entries for NS and CNAME. It is just possible
4452 that this could have caused confusion if both these types were looked up
4453 for the same domain, because the text type is used as part of Exim's
4454 per-process caching. But the chance of anyone hitting this buglet seems
4457 37. The dnsdb lookup has been extended in a number of ways.
4459 (1) There is a new type, "zns", which walks up the domain tree until it
4460 finds some nameserver records. It should be used with care.
4462 (2) There is a new type, "mxh", which is like "mx" except that it returns
4463 just the host names, not the priorities.
4465 (3) It is now possible to give a list of domains (or IP addresses) to be
4466 looked up. The behaviour when one of the lookups defers can be
4467 controlled by a keyword.
4469 (4) It is now possible to specify the separator character for use when
4470 multiple records are returned.
4472 38. The dnslists ACL condition has been extended: it is now possible to supply
4473 a list of IP addresses and/or domains to be looked up in a particular DNS
4476 39. Added log_selector=+queue_time_overall.
4478 40. When running the queue in the test harness, wait just a tad after forking a
4479 delivery process, to get repeatability of debugging output.
4481 41. Include certificate and key file names in error message when GnuTLS fails
4482 to set them up, because the GnuTLS error message doesn't include the name
4483 of the failing file when there is a problem reading it.
4485 42. Allow both -bf and -bF in the same test run.
4487 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4489 44. The "Exiscan patch" is now merged into the mainline Exim source.
4491 45. Sometimes the final signoff response after QUIT could fail to get
4492 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4493 < 0 before doing a fflush(). This bug looks as though it goes back to the
4494 introduction of TLS in release 3.20, but "sometimes" must have been rare
4495 because the tests only now provoked it.
4497 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4498 (this can affect the format of dates).
4500 47. exim_tidydb, when checking for the continued existence of a message for
4501 which it has found a message-specific retry record, was not finding
4502 messages that were in split spool directories. Consequently, it was
4503 deleting retry records that should have stayed in existence.
4505 48. Steve fixed some bugs in eximstats.
4507 49. The SPA authentication driver was not abandoning authentication and moving
4508 on to the next authenticator when an expansion was forced to fail,
4509 contradicting the general specification for all authenticators. Instead it
4510 was generating a temporary error. It now behaves as specified.
4512 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4513 (the order specifies the preference for clients). The order is now AES256,
4514 AES128, 3DES, ARCFOUR128.
4516 51. Small patch to Sieve code - explicitly set From: when generating an
4519 52. Exim crashed if a remote delivery caused a very long error message to be
4520 recorded - for instance if somebody sent an entire SpamAssassin report back
4521 as a large number of 550 error lines. This bug was coincidentally fixed by
4522 increasing the size of one of Exim's internal buffers (big_buffer) that
4523 happened as part of the Exiscan merge. However, to be on the safe side, I
4524 have made the code more robust (and fixed the comments that describe what
4527 53. Now that there can be additional text after "Completed" in log lines (if
4528 the queue_time_overall log selector is set), a one-byte patch to exigrep
4529 was needed to allow it to recognize "Completed" as not the last thing in
4532 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4533 patch that reportedly fixes this has been added. I am not expert enough to
4534 create a test for it. This is what the patch creator wrote:
4536 "I found a little strange behaviour of ldap code when working with
4537 Windows 2003 AD Domain, where users was placed in more than one
4538 Organization Units. When I tried to give exim partial DN, the exit code
4539 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4540 But simultaneously result of request was absolutely normal ldap result,
4541 so I produce this patch..."
4543 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4544 REFERENCE, so I have modified the code to exclude the patch when that macro
4547 55. Some experimental protocols are using DNS PTR records for new purposes. The
4548 keys for these records are domain names, not reversed IP addresses. The
4549 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4550 leaves it alone. Component reversal etc. now happens only for IP addresses.
4553 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4555 57. Double the size of the debug message buffer (to 2048) so that more of very
4556 long debug lines gets shown.
4558 58. The exicyclog utility now does better if the number of log files to keep
4559 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4561 59. Two changes related to the smtp_active_hostname option:
4563 (1) $smtp_active_hostname is now available as a variable.
4564 (2) The default for smtp_banner uses $smtp_active_hostname instead
4565 of $primary_hostname.
4567 60. The host_aton() function is supposed to be passed a string that is known
4568 to be a valid IP address. However, in the case of IPv6 addresses, it was
4569 not checking this. This is a hostage to fortune. Exim now panics and dies
4570 if the condition is not met. A case was found where this could be provoked
4571 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4572 components; fortuitously, this particular loophole had already been fixed
4573 by change 4.50/55 above.
4575 If there are any other similar loopholes, the new check in host_aton()
4576 itself should stop them being exploited. The report I received stated that
4577 data on the command line could provoke the exploit when Exim was running as
4578 exim, but did not say which command line option was involved. All I could
4579 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4580 running as the user.
4583 61. There was a buffer overflow vulnerability in the SPA authentication code
4584 (which came originally from the Samba project). I have added a test to the
4585 spa_base64_to_bits() function which I hope fixes it.
4588 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4589 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4591 63. The daemon start-up calls getloadavg() while still root for those OS that
4592 need the first call to be done as root, but it missed one case: when
4593 deliver_queue_load_max is set with deliver_drop_privilege. This is
4594 necessary for the benefit of the queue runner, because there is no re-exec
4595 when deliver_drop_privilege is set.
4597 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4598 This has been fixed.
4600 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4601 was in use, was not putting the data itself into the right store pool;
4602 consequently, it could be overwritten for a subsequent message in the same
4603 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4606 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4608 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4609 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4610 the function were treating the return as a boolean value, which happened to
4611 work because 0=false and not-0=true, but is not correct code.
4613 68. The host_aton() function was not handling scoped IPv6 addresses (those
4614 with, for example, "%eth0" on the end) correctly.
4616 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4617 negated items (that is, ~something) in unsigned ints. Some compilers
4618 apparently mutter when there is no cast.
4620 70. If an address verification called from an ACL failed, and did not produce a
4621 user-specific message (i.e. there was only a "system" message), nothing was
4622 put in $acl_verify_message. In this situation, it now puts the system
4625 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4626 session; change 4.31/43 added the unwanted input to the log line - except
4627 that it did not do this in the start of session case. It now does.
4629 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4630 This is wrong and can cause the other end to generate a synchronization
4631 error if it is another Exim or anything else that does the synchronization
4632 check. A QUIT command is no longer sent after a timeout.
4634 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4635 during host lookups.
4637 74. The defer_ok option of callout verification was not working if it was used
4638 when verifying addresses in header lines, that is, for this case:
4640 verify = header_sender/callout=defer_ok
4642 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4643 those file descriptors could be used for SMTP connections. If anything
4644 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4645 could be sent to the SMTP client, causing chaos. The daemon now opens
4646 stdin, stdout, and stderr to /dev/null when it puts itself into the
4649 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4650 log by default. The user can override this with suitable Perl magic.
4652 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4653 the log message when discard triggers, was not working for the DATA ACL or
4654 for the non-SMTP ACL.
4656 78. Error message wording change in sieve.c.
4658 79. If smtp_accept_max_per_host was set, the number of connections could be
4659 restricted to fewer than expected, because the daemon was trying to set up
4660 a new connection before checking whether the processes handling previous
4661 connections had finished. The check for completed processes is now done
4662 earlier. On busy systems, this bug wouldn't be noticed because something
4663 else would have woken the daemon, and it would have reaped the completed
4666 80. If a message was submitted locally by a user whose login name contained one
4667 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4668 It caused a spool format error. I have fixed the spool reading code. A
4669 related problem was that the "from" clause in the Received: line became
4670 illegal because of the space(s). It is now covered by ${quote_local_part.
4672 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4675 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4677 83. Patch from Sieve maintainer to fix unterminated string problem in
4678 "vacation" handling.
4680 84. Some minor changes to the Linux configuration files to help with other
4681 OS variants using glibc.
4683 85. One more patch for Sieve to update vacation handling to latest spec.
4686 ----------------------------------------------------
4687 See the note above about the 4.44 and 4.50 releases.
4688 ----------------------------------------------------
4694 1. Change 4.43/35 introduced a bug that caused file counts to be
4695 incorrectly computed when quota_filecount was set in an appendfile
4698 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4699 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4702 3. Give more explanation in the error message when the command for a transport
4703 filter fails to execute.
4705 4. There are several places where Exim runs a non-Exim command in a
4706 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4707 was being done only for the command run by the queryprogram router. It is
4708 now done for all such subprocesses. The other cases are: ${run, transport
4709 filters, and the commands run by the lmtp and pipe transports.
4711 5. Some older OS have a limit of 256 on the maximum number of file
4712 descriptors. Exim was using setrlimit() to set 1000 as a large value
4713 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4714 systems. I've change it so that if it can't get 1000, it tries for 256.
4716 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4717 was an oversight, and furthermore, ever since the addition of extra
4718 controls (e.g. 4.43/32), the checks on when to allow different forms of
4719 "control" were broken. There should now be diagnostics for all cases when a
4720 control that does not make sense is encountered.
4722 7. $recipients is now available in the predata ACL (oversight).
4724 8. Tidy the search cache before the fork to do a delivery from a message
4725 received from the command line. Otherwise the child will trigger a lookup
4726 failure and thereby defer the delivery if it tries to use (for example) a
4727 cached ldap connection that the parent has called unbind on.
4729 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4730 of $address_data from the recipient verification was clobbered by the
4731 sender verification.
4733 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4734 was its contents. (It was OK if the option was not defined at all.)
4736 11. A "Completed" log line is now written for messages that are removed from
4737 the spool by the -Mrm option.
4739 12. $host_address is now set to the target address during the checking of
4740 ignore_target_hosts.
4742 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4743 being passed; this would have caused $sender_host_name to have been used if
4744 matching the list had actually called for a host name (not very likely,
4745 since this list is usually IP addresses). A host name is now passed as
4748 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4749 code that specifies a non-restarting handler (typically sigaction() in
4750 modern systems) in an attempt to fix a rare and obscure crash bug.
4752 15. Narrowed the window for a race in the daemon that could cause it to ignore
4753 SIGCHLD signals. This is not a major problem, because they are used only to
4754 wake it up if nothing else does.
4756 16. A malformed maildirsize file could cause Exim to calculate negative values
4757 for the mailbox size or file count. Odd effects could occur as a result.
4758 The maildirsize information is now recalculated if the size or filecount
4761 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4762 support for a long time. Removed HAVE_SYS_VFS_H.
4764 18. Updated exipick to current release from John Jetmore.
4766 19. Allow an empty sender to be matched against a lookup in an address list.
4767 Previously the only cases considered were a regular expression, or an
4770 20. Exim went into a mad DNS lookup loop when doing a callout where the
4771 host was specified on the transport, if the DNS lookup yielded more than
4774 21. The RFC2047 encoding function was originally intended for short strings
4775 such as real names; it was not keeping to the 75-character limit for
4776 encoded words that the RFC imposes. It now respects the limit, and
4777 generates multiple encoded words if necessary. To be on the safe side, I
4778 have increased the buffer size for the ${rfc2047: expansion operator from
4781 22. Failure to deliver a bounce message always caused it to be frozen, even if
4782 there was an errors_to setting on the router. The errors_to setting is now
4785 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4786 canonical form (fully expanded) before being placed in
4787 $sender_host_address.
4789 24. Updated eximstats to version 1.33
4791 25. Include certificate and key file names in error message when GnuTLS fails
4792 to set them up, because the GnuTLS error message doesn't include the name
4793 of the failing file when there is a problem reading it.
4795 26. Expand error message when OpenSSL has problems setting up cert/key files.
4798 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4799 (this can affect the format of dates).
4801 28. exim_tidydb, when checking for the continued existence of a message for
4802 which it has found a message-specific retry record, was not finding
4803 messages that were in split spool directories. Consequently, it was
4804 deleting retry records that should have stayed in existence.
4806 29. eximstats updated to version 1.35
4807 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4808 1.35 - bugfix such that pie charts by volume are generated correctly
4810 30. The SPA authentication driver was not abandoning authentication and moving
4811 on to the next authenticator when an expansion was forced to fail,
4812 contradicting the general specification for all authenticators. Instead it
4813 was generating a temporary error. It now behaves as specified.
4815 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4816 (the order specifies the preference for clients). The order is now AES256,
4817 AES128, 3DES, ARCFOUR128.
4819 31. Small patch to Sieve code - explicitly set From: when generating an
4822 32. Exim crashed if a remote delivery caused a very long error message to be
4823 recorded - for instance if somebody sent an entire SpamAssassin report back
4824 as a large number of 550 error lines. This bug was coincidentally fixed by
4825 increasing the size of one of Exim's internal buffers (big_buffer) that
4826 happened as part of the Exiscan merge. However, to be on the safe side, I
4827 have made the code more robust (and fixed the comments that describe what
4830 33. Some experimental protocols are using DNS PTR records for new purposes. The
4831 keys for these records are domain names, not reversed IP addresses. The
4832 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4833 leaves it alone. Component reversal etc. now happens only for IP addresses.
4836 34. The host_aton() function is supposed to be passed a string that is known
4837 to be a valid IP address. However, in the case of IPv6 addresses, it was
4838 not checking this. This is a hostage to fortune. Exim now panics and dies
4839 if the condition is not met. A case was found where this could be provoked
4840 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4841 components; fortuitously, this particular loophole had already been fixed
4842 by change 4.50/55 or 4.44/33 above.
4844 If there are any other similar loopholes, the new check in host_aton()
4845 itself should stop them being exploited. The report I received stated that
4846 data on the command line could provoke the exploit when Exim was running as
4847 exim, but did not say which command line option was involved. All I could
4848 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4849 running as the user.
4852 35. There was a buffer overflow vulnerability in the SPA authentication code
4853 (which came originally from the Samba project). I have added a test to the
4854 spa_base64_to_bits() function which I hope fixes it.
4857 36. The daemon start-up calls getloadavg() while still root for those OS that
4858 need the first call to be done as root, but it missed one case: when
4859 deliver_queue_load_max is set with deliver_drop_privilege. This is
4860 necessary for the benefit of the queue runner, because there is no re-exec
4861 when deliver_drop_privilege is set.
4863 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4864 was in use, was not putting the data itself into the right store pool;
4865 consequently, it could be overwritten for a subsequent message in the same
4866 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4869 38. Sometimes the final signoff response after QUIT could fail to get
4870 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4871 < 0 before doing a fflush(). This bug looks as though it goes back to the
4872 introduction of TLS in release 3.20, but "sometimes" must have been rare
4873 because the tests only now provoked it.
4879 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4880 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4881 Now it returns an integer. A number of calls were still expecting a T/F
4882 return. Fortuitously, in all cases, the tests worked in OK situations,
4883 which is the norm. However, things would have gone wrong on any write
4884 failures on the smtp file descriptor. This function is used when sending
4885 messages over SMTP and also when doing verify callouts.
4887 2. When Exim is called to do synchronous delivery of a locally submitted
4888 message (the -odf or -odi options), it no longer closes stderr before doing
4891 3. Implemented the mua_wrapper option.
4893 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4895 5. Implemented the functions header_remove(), header_testname(),
4896 header_add_at_position(), and receive_remove_recipient(), and exported them
4899 6. If an ACL "warn" statement specified the addition of headers, Exim already
4900 inserted X-ACL-Warn: at the start if there was no header name. However, it
4901 was not making this test for the second and subsequent header lines if
4902 there were newlines in the string. This meant that an invalid header could
4903 be inserted if Exim was badly configured.
4905 7. Allow an ACL "warn" statement to add header lines at the start or after all
4906 the Received: headers, as well as at the end.
4908 8. Added the rcpt_4xx retry error code.
4910 9. Added postmaster_mailfrom=xxx to callout verification option.
4912 10. Added mailfrom=xxxx to the callout verification option, for verify=
4915 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4916 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4918 12. Inserted some casts to stop certain compilers warning when using pointer
4919 differences as field lengths or precisions in printf-type calls (mostly
4920 affecting debugging statements).
4922 13. Added optional readline() support for -be (dynamically loaded).
4924 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4925 same clock tick as a message's arrival, so that its received time was the
4926 same as the "first fail" time on the retry record, and that message
4927 remained on the queue past the ultimate address timeout, every queue runner
4928 would try a delivery (because it was past the ultimate address timeout) but
4929 after another failure, the ultimate address timeout, which should have then
4930 bounced the address, did not kick in. This was a "< instead of <=" error;
4931 in most cases the first failure would have been in the next clock tick
4932 after the received time, and all would be well.
4934 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4935 being recognized when the domain list was tested by the match_domain
4936 condition in an expansion string.
4938 16. Added the ${str2b64: operator.
4940 17. Exim was always calling setrlimit() to set a large limit for the number of
4941 processes, without checking whether the existing limit was already
4942 adequate. (It did check for the limit on file descriptors.) Furthermore,
4943 errors from getrlimit() and setrlimit() were being ignored. Now they are
4944 logged to the main and panic logs, but Exim does carry on, to try to do its
4945 job under whatever limits there are.
4947 18. Imported PCRE 5.0.
4949 19. Trivial typo in log message " temporarily refused connection" (the leading
4952 20. If the log selector return_path_on_delivery was set and an address was
4953 redirected to /dev/null, the delivery process crashed because it assumed
4954 that a return path would always be set for a "successful" delivery. In this
4955 case, the whole delivery is bypassed as an optimization, and therefore no
4958 21. Internal re-arrangement: the function for sending a challenge and reading
4959 a response while authentication was assuming a zero-terminated challenge
4960 string. It's now changed to take a pointer and a length, to allow for
4961 binary data in such strings.
4963 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4965 23. Exim was not respecting finduser_retries when seeking the login of the
4966 uid under which it was called; it was always trying 10 times. (The default
4967 setting of finduser_retries is zero.) Also, it was sleeping after the final
4968 failure, which is pointless.
4970 24. Implemented tls_on_connect_ports.
4972 25. Implemented acl_smtp_predata.
4974 26. If the domain in control=submission is set empty, Exim assumes that the
4975 authenticated id is a complete email address when it generates From: or
4976 Sender: header lines.
4978 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4979 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4980 chown and chgrp in /bin and hostname in /usr/bin.
4982 28. Exim was keeping the "process log" file open after each use, just as it
4983 does for the main log. This opens the possibility of it remaining open for
4984 long periods when the USR1 signal hits a daemon. Occasional processlog
4985 errors were reported, that could have been caused by this. Anyway, it seems
4986 much more sensible not to leave this file open at all, so that is what now
4989 29. The long-running daemon process does not normally write to the log once it
4990 has entered its main loop, and it closes the log before doing so. This is
4991 so that log files can straightforwardly be renamed and moved. However,
4992 there are a couple of unusual error situations where the daemon does write
4993 log entries, and I had neglected to close the log afterwards.
4995 30. The text of an SMTP error response that was received during a remote
4996 delivery was being truncated at 512 bytes. This is too short for some of
4997 the long messages that one sometimes sees. I've increased the limit to
5000 31. It is now possible to make retry rules that apply only when a message has a
5001 specific sender, in particular, an empty sender.
5003 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5004 it possible to be selective about when SMTP synchronization is enforced.
5006 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5008 32. Implemented hosts_connection_nolog.
5010 33. Added an ACL for QUIT.
5012 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5015 35. Added mailbox_size and mailbox_filecount to appendfile.
5017 36. Added control = no_multiline_responses to ACLs.
5019 37. There was a bug in the logic of the code that waits for the clock to tick
5020 in the case where the clock went backwards by a substantial amount such
5021 that the microsecond fraction of "now" was more than the microsecond
5022 fraction of "then" (but the whole seconds number was less).
5024 38. Added support for the libradius Radius client library this is found on
5025 FreeBSD (previously only the radiusclient library was supported).
5031 1. When certain lookups returned multiple values in the form name=value, the
5032 quoting of the values was not always being done properly. Specifically:
5033 (a) If the value started with a double quote, but contained no whitespace,
5035 (b) If the value contained whitespace other than a space character (i.e.
5036 tabs or newlines or carriage returns) it was not quoted.
5037 This fix has been applied to the mysql and pgsql lookups by writing a
5038 separate quoting function and calling it from the lookup code. The fix
5039 should probably also be applied to nisplus, ibase and oracle lookups, but
5040 since I cannot test any of those, I have not disturbed their existing code.
5042 2. A hit in the callout cache for a specific address caused a log line with no
5043 reason for rejecting RCPT. Now it says "Previous (cached) callout
5044 verification failure".
5046 3. There was an off-by-one bug in the queryprogram router. An over-long
5047 return line was truncated at 256 instead of 255 characters, thereby
5048 overflowing its buffer with the terminating zero. As well as fixing this, I
5049 have increased the buffer size to 1024 (and made a note to document this).
5051 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5052 when Exim is waiting for an SMTP response from a remote server, Exim
5053 restarts its select() call on the socket, thereby resetting its timeout.
5054 This is not a problem when such interrupts are rare. Somebody set up a cron
5055 job to run exiwhat every 2 minutes, which is less than the normal select()
5056 timeout (5 or 10 minutes). This meant that the select() timeout never
5057 kicked in because it was always reset. I have fixed this by comparing the
5058 time when an interrupt arrives with the time at the start of the first call
5059 to select(). If more time than the timeout has elapsed, the interrupt is
5060 treated as a timeout.
5062 5. Some internal re-factoring in preparation for the addition of Sieve
5063 extensions (by MH). In particular, the "personal" test is moved to a
5064 separate function, and given an option for scanning Cc: and Bcc: (which is
5065 not set for Exim filters).
5067 6. When Exim created an email address using the login of the caller as the
5068 local part (e.g. when creating a From: or Sender: header line), it was not
5069 quoting the local part when it contained special characters such as @.
5071 7. Installed new OpenBSD configuration files.
5073 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5074 try to make them clearer.
5076 9. Callout options, other than the timeout value, were being ignored when
5077 verifying sender addresses in header lines. For example, when using
5079 verify = header_sender/callout=no_cache
5081 the cache was (incorrectly) being used.
5083 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5084 only the Cygwin environment.
5086 11. When return_path_on_delivery was set as a log selector, if different remote
5087 addresses in the same message used different return paths and parallel
5088 remote delivery occurred, the wrong values would sometimes be logged.
5089 (Whenever a remote delivery process finished, the return path value from
5090 the most recently started remote delivery process was logged.)
5092 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5093 lines when AUTH and/or TLS are in use. This is the "received protocol"
5094 field. Exim used to use "asmtp" for authenticated SMTP, without any
5095 indication (in the protocol name) for TLS use. Now it follows the RFC and
5096 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5097 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5098 names appear in log lines as well as in Received: header lines.
5100 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5101 extensions, and comparison tests, and to fix some bugs.
5103 14. Changes to the "personal" filter test:
5105 (1) The test was buggy in that it was just doing the equivalent of
5106 "contains" tests on header lines. For example, if a user's address was
5107 anne@some.where, the "personal" test would incorrectly be true for
5109 To: susanne@some.where
5111 This test is now done by extracting each address from the header in turn,
5112 and checking the entire address. Other tests that are part of "personal"
5113 are now done using regular expressions (for example, to check local parts
5114 of addresses in From: header lines).
5116 (2) The list of non-personal local parts in From: addresses has been
5117 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5118 taken from the Sieve specification recommendations.
5120 (3) If the message contains any header line starting with "List-" it is
5121 treated as non-personal.
5123 (4) The test for "circular" in the Subject: header line has been removed
5124 because it now seems ill-conceived.
5126 15. Minor typos in src/EDITME comments corrected.
5128 16. Installed latest exipick from John Jetmore.
5130 17. If headers_add on a router specified a text string that was too long for
5131 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5132 of string_sprintf() is now avoided.
5134 18. $message_body_size was not set (it was always zero) when running the DATA
5135 ACL and the local_scan() function.
5137 19. For the "mail" command in an Exim filter, no default was being set for
5138 the once_repeat time, causing a random time value to be used if "once" was
5139 specified. (If the value happened to be <= 0, no repeat happened.) The
5140 default is now 0s, meaning "never repeat". The "vacation" command was OK
5141 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5142 (I found it when inspecting the code).
5144 20. There is now an overall timeout for performing a callout verification. It
5145 defaults to 4 times the callout timeout, which applies to individual SMTP
5146 commands during the callout. The overall timeout applies when there is more
5147 than one host that can be tried. The timeout is checked before trying the
5148 next host. This prevents very long delays if there are a large number of
5149 hosts and all are timing out (e.g. when the network connections are timing
5150 out). The value of the overall timeout can be changed by specifying an
5151 additional sub-option for "callout", called "maxwait". For example:
5153 verify = sender/callout=5s,maxwait=20s
5155 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5156 to the end before writing, but this should make it even safer).
5158 22. Exim was forgetting that it had advertised PIPELINING for the second and
5159 subsequent messages on an SMTP connection. It was also not resetting its
5160 memory on STARTTLS and an internal HELO.
5162 23. When Exim logs an SMTP synchronization error within a session, it now
5163 records whether PIPELINING has been advertised or not.
5165 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5166 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5167 rather than long int.
5169 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5171 26. Added the never_mail option to autoreply.
5177 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5178 crash if the getsockname() call failed; this can happen if a connection is
5179 closed very soon after it is established. The problem was simply in the
5180 order in which certain operations were done, causing Exim to try to write
5181 to the SMTP stream before it had set up the file descriptor. The bug has
5182 been fixed by making things happen in the correct order.
5188 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5189 before the connection was closed, thus losing the rejection response.
5191 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5192 some early Solaris releases, but causes trouble in current releases where
5193 socklen_t is defined.
5195 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5198 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5201 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5202 that releasing the top of it at the end releases what was used for sub-
5203 expansions (unless the block got too big). However, discard this block if
5204 the first thing is a variable or header, so that we can use its block when
5205 it is dynamic (useful for very large $message_headers, for example).
5207 6. Lookups now cache *every* query, not just the most recent. A new, separate
5208 store pool is used for this. It can be recovered when all lookup caches are
5209 flushed. Lookups now release memory at the end of their result strings.
5210 This has involved some general refactoring of the lookup sources.
5212 7. Some code has been added to the store_xxx() functions to reduce the amount
5213 of flapping under certain conditions.
5215 8. log_incoming_interface used to affect only the <= reception log lines. Now
5216 it causes the local interface and port to be added to several more SMTP log
5217 lines, for example "SMTP connection from", and rejection lines.
5219 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5221 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5223 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5224 could be overwritten at the end of the current message (or the start of a
5225 new message if it was set in a HELO ACL). The value is now preserved for
5226 the duration of the SMTP connection.
5228 12. If a transport had a headers_rewrite setting, and a matching header line
5229 contained an unqualified address, that address was qualified, even if it
5230 did not match any rewriting rules. The underlying bug was that the values
5231 of the flags that permit the existence of unqualified sender and recipient
5232 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5233 non-local messages, and by -bnq for local messages) were not being
5234 preserved with the message after it was received.
5236 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5237 "next input=" as part of the text comprising the host identity instead of
5238 the correct text. The code was using the same buffer for two different
5239 strings. However, depending on which order the printing function evaluated
5240 its arguments, the bug did not always show up. Under Linux, for example, my
5241 test suite worked just fine.
5243 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5244 doesn't work with some older versions of Perl. It has been changed to "my",
5245 which in any case is probably the better facility to use.
5247 15. A really picky compiler found some instances of statements for creating
5248 error messages that either had too many or two few arguments for the format
5251 16. The size of the buffer for calls to the DNS resolver has been increased
5252 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5253 for addresses that have a lot of PTR records. This alleviates a problem; it
5254 does not fully solve it.
5256 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5257 buffer now truncates the list and logs the incident, which is the same
5258 action as happens when Exim is looking up a host name and its aliases.
5259 Previously in this situation something unpredictable would happen;
5260 sometimes it was "internal error: store_reset failed".
5262 18. If a server dropped the connection unexpectedly when an Exim client was
5263 using GnuTLS and trying to read a response, the client delivery process
5264 crashed while trying to generate an error log message.
5266 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5267 string, for example:
5269 warn message = H1: something\nH2: something
5271 the text was added as a single header line from Exim's point of view
5272 though it ended up OK in the delivered message. However, searching for the
5273 second and subsequent header lines using $h_h2: did not work. This has been
5274 fixed. Similarly, if a system filter added multiple headers in this way,
5275 the routers could not see them.
5277 20. Expanded the error message when iplsearch is called with an invalid key to
5278 suggest using net-iplsearch in a host list.
5280 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5281 ACLs are no longer actually imposed (and a message to that effect is
5284 22. If a "gecos" field in a passwd entry contained escaped characters, in
5285 particular, if it contained a \" sequence, Exim got it wrong when building
5286 a From: or a Sender: header from that name. A second bug also caused
5287 incorrect handling when an unquoted " was present following a character
5288 that needed quoting.
5290 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5291 was not being matched caselessly.
5293 24. Arranged for all hyphens in the exim.8 source to be escaped with
5296 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5297 properly. Recipient callout cache records were still being keyed to include
5298 the sender, even when use_sender was set false. This led to far more
5299 callouts that were necessary. The sender is no longer included in the key
5300 when use_sender is false.
5302 26. Added "control = submission" modifier to ACLs.
5304 27. Added the ${base62d: operator to decode base 62 numbers.
5306 28. dnsdb lookups can now access SRV records.
5308 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5309 the configuration file.
5311 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5312 (-v) mode. This makes the output for a verbose queue run more intelligible.
5314 31. Added a use_postmaster feature to recipient callouts.
5316 32. Added the $body_zerocount variable, containing the number of binary zero
5317 bytes in the message body.
5319 33. The time of last modification of the "new" subdirectory is now used as the
5320 "mailbox time last read" when there is a quota error for a maildir
5323 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5325 35. Added +ignore_unknown as a special item in host lists.
5327 36. Code for decoding IPv6 addresses in host lists is now included, even if
5328 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5329 address was recognized as an IP address, but was then not correctly decoded
5330 into binary, causing unexpected and incorrect effects when compared with
5337 1. Very minor rewording of debugging text in manualroute to say "list of
5338 hosts" instead of "hostlist".
5340 2. If verify=header_syntax was set, and a header line with an unqualified
5341 address (no domain) and a large number of spaces between the end of the
5342 name and the colon was received, the reception process suffered a buffer
5343 overflow, and (when I tested it) crashed. This was caused by some obsolete
5344 code that should have been removed. The fix is to remove it!
5346 3. When running in the test harness, delay a bit after writing a bounce
5347 message to get a bit more predictability in the log output.
5349 4. Added a call to search_tidyup() just before forking a reception process. In
5350 theory, someone could use a lookup in the expansion of smtp_accept_max_
5351 per_host which, without the tidyup, could leave open a database connection.
5353 5. Added the variables $recipient_data and $sender_data which get set from a
5354 lookup success in an ACL "recipients" or "senders" condition, or a router
5355 "senders" option, similar to $domain_data and $local_part_data.
5357 6. Moved the writing of debug_print from before to after the "senders" test
5360 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5361 problems for message scanning, either using a data ACL, or using
5362 local_scan() because the Received: header was not generated till after they
5363 were called (in order to set the time as the time of reception completion).
5364 I have revised the way this works. The header is now generated after the
5365 body is received, but before the ACL or local_scan() are called. After they
5366 are run, the timestamp in the header is updated.
5372 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5373 before starting a queue runner without re-exec. This happened only when
5374 deliver_drop_privilege was set or when the Exim user was set to root. The
5375 effect of the bug was that timeouts during subsequent deliveries caused
5376 crashes instead of being properly handled. The handler is now left at its
5377 default (and expected) setting.
5379 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5380 message, again when deliver_drop_privilege is set or Exim is run as root.
5381 The bug described in (1) was not present in this case, but the tidying up
5382 of the other signals was missing. I have made the two cases consistent.
5384 3. The ignore_target_hosts setting on a manualroute router was being ignored
5385 for hosts that were looked up using the /MX notation.
5387 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5390 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5391 operated on the sender address. After changing the $sender_address to <>
5392 for the sender address verify, Exim was re-instated it as the original
5393 (before rewriting) address, but remembering that it had rewritten it, so it
5394 wasn't rewriting it again. This bug also had the effect of breaking the
5395 sender address verification caching when the sender address was rewritten.
5397 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5398 This has been changed so that if the ip literal address matches
5399 ignore_target_hosts, the router declines.
5401 7. Added expansion conditions match_domain, match_address, and match_local_
5402 part (NOT match_host).
5404 8. The placeholder for the Received: header didn't have a length field set.
5406 9. Added code to Exim itself and to exim_lock to test for a specific race
5407 condition that could lead to file corruption when using MBX delivery. The
5408 issue is with the lockfile that is created in /tmp. If this file is removed
5409 after a process has opened it but before that process has acquired a lock,
5410 there is the potential for a second process to recreate the file and also
5411 acquire a lock. This could lead to two Exim processes writing to the file
5412 at the same time. The added code performs the same test as UW imapd; it
5413 checks after acquiring the lock that its file descriptor still refers to
5414 the same named file.
5416 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5417 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5420 11. Added the smtp_active_hostname option. If used, this will typically be made
5421 to depend on the incoming interface address. Because $interface_address is
5422 not set up until the daemon has forked a reception process, error responses
5423 that can happen earlier (such as "too many connections") no longer contain
5426 12. If an expansion in a condition on a "warn" statement fails because a lookup
5427 defers, the "warn" statement is abandoned, and the next ACL statement is
5428 processed. Previously this caused the whole ACL to be aborted.
5430 13. Added the iplsearch lookup type.
5432 14. Added ident_timeout as a log selector.
5434 15. Added tls_certificate_verified as a log selector.
5436 16. Added a global option tls_require_ciphers (compare the smtp transport
5437 option of the same name). This controls incoming TLS connections.
5439 17. I finally figured out how to make tls_require_ciphers do a similar thing
5440 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5441 before starting the TLS session.
5443 18. Tabs are now shown as \t in -bP output.
5445 19. If the log selector return_path_on_delivery was set, Exim crashed when
5446 bouncing a message because it had too many Received: header lines.
5448 20. If two routers both had headers_remove settings, and the first one included
5449 a superfluous trailing colon, the final name in the first list and the
5450 first name in the second list were incorrectly joined into one item (with a
5451 colon in the middle).
5457 1. Added -C and -D options to the exinext utility, mainly to make it easier
5458 to include in the automated testing, but these could be helpful when
5459 multiple configurations are in use.
5461 2. The exinext utility was not formatting the output nicely when there was
5462 an alternate port involved in the retry record key, nor when there was a
5463 message id as well (for retries that were specific to a specific message
5464 and a specific host). It was also confused by IPv6 addresses, because of
5465 the additional colons they contain. I have fixed the IPv4 problem, and
5466 patched it up to do a reasonable job for IPv6.
5468 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5469 delivery, the log line now contains "pipelined" if PIPELINING was used.
5471 4. An SMTP transport process used to panic and die if the bind() call to set
5472 an explicit outgoing interface failed. This has been changed; it is now
5473 treated in the same way as a connect() failure.
5475 5. A reference to $sender_host_name in the part of a conditional expansion
5476 that was being skipped was still causing a DNS lookup. This no longer
5479 6. The def: expansion condition was not recognizing references to header lines
5480 that used bh_ and bheader_.
5482 7. Added the _cache feature to named lists.
5484 8. The code for checking quota_filecount in the appendfile transport was
5485 allowing one more file than it should have been.
5487 9. For compatibility with Sendmail, the command line option
5495 and sets the incoming protocol and host name (for trusted callers). The
5496 host name and its colon can be omitted when only the protocol is to be set.
5497 Note the Exim already has two private options, -pd and -ps, that refer to
5498 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5499 "s", but I don't think that's a major issue.
5501 10. A number of refactoring changes to the code, none of which should affect
5504 (a) The number of logging options was getting close to filling up the
5505 32-bit word that was used as a bit map. I have split them into two classes:
5506 those that are passed in the argument to log_write(), and those that are
5507 only ever tested independently outside of that function. These are now in
5508 separate 32-bit words, so there is plenty of room for expansion again.
5509 There is no change in the user interface or the logging behaviour.
5511 (b) When building, for example, log lines, the code previously used a
5512 macro that called string_cat() twice, in order to add two strings. This is
5513 not really sufficiently general. Furthermore, there was one instance where
5514 it was actually wrong because one of the argument was used twice, and in
5515 one call a function was used. (As it happened, calling the function twice
5516 did not affect the overall behaviour.) The macro has been replaced by a
5517 function that can join an arbitrary number of extra strings onto a growing
5520 (c) The code for expansion conditions now uses a table and a binary chop
5521 instead of a serial search (which was left over from when there were very
5522 few conditions). Also, it now recognizes conditions like "pam" even when
5523 the relevant support is not compiled in: a suitably worded error message is
5524 given if an attempt is made to use such a condition.
5526 11. Added ${time_interval:xxxxx}.
5528 12. A bug was causing one of the ddress fields not to be passed back correctly
5529 from remote delivery subprocesses. The field in question was not being
5530 subsequently used, so this caused to problems in practice.
5532 13. Added new log selectors queue_time and deliver_time.
5534 14. Might have fixed a bug in maildirsizefile handling that threw up
5535 "unexpected character" debug warnings, and recalculated the data
5536 unnecessarily. In any case, I expanded the warning message to give more
5539 15. Added the message "Restricted characters in address" to the statements in
5540 the default ACL that block characters like @ and % in local parts.
5542 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5543 Three changes have been made:
5545 (a) There was a serious bug; a negative response to MAIL caused the whole
5546 recipient domain to be cached as invalid, thereby blocking all messages
5547 to all local parts at the same domain, from all senders. This bug has
5548 been fixed. The domain is no longer cached after a negative response to
5549 MAIL if the sender used is not empty.
5551 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5554 (c) A new callout option, "use_sender" has been added for people who want
5555 the modified behaviour.
5561 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5564 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5565 indeed breaks things for older releases.
5567 3. Added additional logging to the case where there is a problem reading data
5568 from a filter that is running in a subprocess using a pipe, in order to
5569 try to track down a specific problem.
5571 4. Testing facility fudge: when running in the test harness and attempting
5572 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5573 getting "No route to host". Convert this to a timeout.
5575 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5578 6. Some OS don't have socklen_t but use size_t instead. This affects the
5579 fifth argument of getsockopt() amongst other things. This is now
5580 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5581 can be set for individual OS. I have set it for SunOS5, OSF1, and
5582 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5583 some earlier ones do not.
5585 7. Change 4.30/15 was not doing the test caselessly.
5587 8. The standard form for an IPv6 address literal was being rejected by address
5588 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5589 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5590 this, as well as the form without the "IPv6" on the front (but only when
5591 address literals are enabled, of course).
5593 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5595 10. Exim crashed if a message with an empty sender address specified by -f
5596 encountered a router with an errors_to setting. This could be provoked only
5597 by a command such as
5601 where an empty string was supplied; "<>" did not hit this bug.
5603 11. Installed PCRE release 4.5.
5605 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5606 remained set. It is now erased.
5608 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5609 times from message ids (which are base 36 rather than the normal 62).
5611 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5612 were being counted as actual protocol errors, and logged if the log
5613 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5614 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5615 and DATA following a set of rejected RCPTs do not count as protocol errors.
5616 In other words, Exim assumes they were pipelined, though this may not
5617 actually be the case. Of course, in all cases the client gets an
5618 appropriate error code.
5620 15. If a lookup fails in an ACL condition, a message about the failure may
5621 be available; it is used if testing the ACL cannot continue, because most
5622 such messages specify what the cause of the deferral is. However, some
5623 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5624 that caused an old message to be retained and used if a later statement
5625 caused a defer, replacing the real cause of the deferral.
5627 16. If an IP address had so many PTR records that the DNS lookup buffer
5628 was not large enough to hold them, Exim could crash while trying to process
5629 the truncated data. It now detects and logs this case.
5631 17. Further to 4.21/58, another change has been made: if (and only if) the
5632 first line of a message (the first header line) ends with CRLF, a bare LF
5633 in a subsequent header line has a space inserted after it, so as not to
5634 terminate the header.
5636 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5637 unnecessarily, used atoi() instead of strtol(), and didn't check the
5638 termination when getting file sizes from file names by regex.
5640 19. Completely re-implemented the support for maildirsize files, in the light
5641 of a number of problems with the previous contributed implementation
5642 (4.30/29). In particular:
5644 . If the quota is zero, the maildirsize file is maintained, but no quota is
5647 . If the maildir directory does not exist, it is created before any attempt
5648 to write a maildirsize file.
5650 . The quota value in the file is just a cache; if the quota is changed in
5651 the transport, the new value overrides.
5653 . A regular expression is available for excluding directories from the
5656 20. The autoreply transport checks the characters in options that define the
5657 message's headers; it allows continued headers, but it was checking with
5658 isspace() after an embedded newline instead of explicitly looking for a
5661 21. If all the "regular" hosts to which an address was routed had passed their
5662 expiry times, and had not reached their retry times, the address was
5663 bounced, even if fallback hosts were defined. Now Exim should go on to try
5666 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5667 equivalent code in the SMTP transport. Some hosts send humungous responses
5668 to HELO/EHLO, more than 1024 it seems.
5670 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5671 gives compiler warnings in some environments. I've now done it "properly",
5674 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5675 (because of problems with the built-in one) was declared to return uschar *
5676 instead of char *, causing compiler failure.
5678 25. Fixed a file descriptor leak when processing alias/forward files.
5680 26. Fixed a minor format string issue in dbfn.c.
5682 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5684 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5685 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5688 29. When a sender address is verified, it is cached, to save repeating the test
5689 when there is more than one recipient in a message. However, when the
5690 verification involves a callout, it is possible for different callout
5691 options to be set for different recipients. It is too complicated to keep
5692 track of this in the cache, so now Exim always runs a verification when a
5693 callout is required, relying on the callout cache for the optimization.
5694 The overhead is duplication of the address routing, but this should not be
5697 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5698 to be verified with callout=postmaster, and the main callout worked but the
5699 postmaster check failed, the verification correctly failed. However, if a
5700 subsequent RCPT command asked for sender verification *without* the
5701 postmaster check, incorrect caching caused this verification also to fail,
5704 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5705 it was not caching the DNS options (qualify_single, search_parents) that
5706 were used when the lookup failed. A subsequent lookup with different
5707 options therefore always gave the same answer, though there were cases
5708 where it should not have. (Example: a "domains = !$mx_any" option on a
5709 dnslookup router: the "domains" option is always processed without any
5710 widening, but the router might have qualify_single set.) Now Exim uses the
5711 cached value only when the same options are set.
5713 32. Added John Jetmore's "exipick" utility to the distribution.
5715 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5716 than a timeout (e.g. a certificate is required, and is not provided), an
5717 Exim server now closes the connection immediately. Previously it waited for
5718 the client to close - but if the client is SSL, it seems that they each
5719 wait for each other, leading to a delay before one of them times out.
5721 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5722 maintained 0.8.x compatibility because I don't think many are using it, and
5723 it is clearly obsolete.
5725 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5728 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5729 client certificate was expired. A simple patch fixes this, though I don't
5730 understand the full logic of why the verify callback is called multiple
5733 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5734 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5735 which causes problems with some clients (such as the Certicom SSL Plus
5736 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5737 disables the coutermeasure allowing Eudora to connect."
5739 38. Exim was not checking that a write() to a log file succeeded. This could
5740 lead to Bad Things if a log got too big, in particular if it hit a file
5741 size limit. Exim now panics and dies if it cannot write to a log file, just
5742 as it does if it cannot open a log file.
5744 39. Modified OS/Makefile-Linux so that it now contains
5746 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5748 The two -D definitions ensure that Exim is compiled with large file
5749 support, which makes it possible to handle log files that are bigger than
5752 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5753 instance) a domain was checked against a named list that involved a lookup,
5754 causing $domain_data to be set, then another domain was checked against the
5755 same list, then the first domain was re-checked, the value of $domain_data
5756 after the final check could be wrong. In particular, if the second check
5757 failed, it could be set empty. This bug probably also applied to
5760 41. The strip_trailing_dot option was not being applied to the address given
5761 with the -f command-line option.
5763 42. The code for reading a message's header from the spool was incrementing
5764 $received_count, but never initializing it. This meant that the value was
5765 incorrect (doubled) while delivering a message in the same process in which
5766 it was received. In the most common configuration of Exim, this never
5767 happens - a fresh exec is done - but it can happen when
5768 deliver_drop_privilege is set.
5770 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5771 it now includes up to 150 characters of the unexpected data in the log
5774 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5775 and building data strings. The size of both of these buffers was 10 000
5776 bytes - far larger than anybody would *ever* want, thought I. Needless to
5777 say, somebody hit the limit. I have increased the maximum line length to
5778 20 000 and the maximum data length of concatenated lines to 100 000. I have
5779 also fixed two bugs, because there was no checking on these buffers. Tsk,
5780 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5781 buffer is too small.
5783 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5784 lsearch lookups. Now it does.
5786 46. When parsing a route_list item in a manualroute router, a fixed-length
5787 buffer was used for the list of hosts. I made this 1024 bytes long,
5788 thinking that nobody would ever have a list of hosts that long. Wrong.
5789 Somebody had a whole pile of complicated expansion conditions, and the
5790 string was silently truncated, leading to an expansion error. It turns out
5791 that it is easier to change to an unlimited length (owing to other changes
5792 that have happened since this code was originally written) than to build
5793 structure for giving a limitation error. The length of the item that
5794 expands into the list of hosts is now unlimited.
5796 47. The lsearch lookup could not handle data where the length of text line was
5797 more than 4095 characters. Such lines were truncated, leading to shortened
5798 data being returned. It should now handle lines of any length.
5800 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5801 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5804 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5805 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5806 no longer happens in comment lines. A list of replacements is now placed
5807 at the head of all of the source files, except those whose only change is
5808 to replace PERL_COMMAND in the very first #! line.
5810 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5811 messages on the queue, with a bottom-up merge sort, using code contributed
5812 by Michael Haardt. This should make operations like -bp somewhat faster on
5813 large queues. It won't affect queue runners, except when queue_run_in_order
5816 51. Installed eximstats 1.31 in the distribution.
5818 52. Added support for SRV lookups to the dnslookup router.
5820 53. If an ACL referred to $message_body or $message_body_end, the value was not
5821 reset for any messages that followed in the same SMTP session.
5823 54. The store-handling optimization for building very long strings was not
5824 differentiating between the different store pools. I don't think this
5825 actually made any difference in practice, but I've tidied it.
5827 55. While running the routers to verify a sender address, $sender_address
5828 was still set to the sender address. This is wrong, because when routing to
5829 send a bounce to the sender, it would be empty. Therefore, I have changed
5830 it so that, while verifying a sender address, $sender_address is set to <>.
5831 (There is no change to what happens when verifying a recipient address.)
5833 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5834 target A or AAAA records (if not already returned) without resetting the
5835 qualify_single or search_parents options of the DNS resolver. These are
5836 inappropriate in this case because the targets of MX and SRV records must
5837 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5838 target that, when qualified, matched something in the local domain. These
5839 two options are now turned off when doing these lookups.
5841 57. It seems that at least some releases of Reiserfs (which does not have the
5842 concept of a fixed number of inodes) returns zero and not -1 for the
5843 number of available inodes. This interacted badly with check_spool_inodes,
5844 which assumed that -1 was the "no such thing" setting. What I have done is
5845 to check that the total number of inodes is greater than zero before doing
5846 the test of how many are available.
5848 58. When a "warn" ACL statement has a log_message modifier, the message is
5849 remembered, and not repeated. This is to avoid a lot of repetition when a
5850 message has many recipients that cause the same warning to be written.
5851 However, Exim was preserving the list of already written lines for an
5852 entire SMTP session, which doesn't seem right. The memory is now reset if a
5853 new message is started.
5855 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5856 debugging output unless log_rewrite was also set.
5858 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5859 of (int)(handle) when we know that handle contains (void *)(-1).
5861 61. The Exim daemon panic-logs an error return when it closes the incoming
5862 connection. However "connection reset by peer" seems to be common, and
5863 isn't really an error worthy of noting specially, so that particular error
5866 62. When Exim is trying to find all the local interfaces, it used to panic and
5867 die if the ioctl to get the interface flags failed. However, it seems that
5868 on at least one OS (Solaris 9) it is possible to have an interface that is
5869 included in the list of interfaces, but for which you get a failure error
5870 for this call. This happens when the interface is not "plumbed" into a
5871 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5872 failure of the "get flags" call assumes that the interface is down.
5874 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5875 makes life easier for people who are doing arithmetic on fields extracted
5876 from dates, where you often get leading zeros that should not be
5877 interpreted as octal.
5879 64. Added qualify_domain to the redirect router, to override the global
5882 65. If a pathologically long header line contained very many addresses (the
5883 report of this problem mentioned 10 000) and each of them was rewritten,
5884 Exim could use up a very large amount of memory. (It kept on making new
5885 copies of the header line as it rewrote, and never released the old ones.)
5886 At the expense of a bit more processing, the header rewriting function has
5887 been changed so that it no longer eats memory in this way.
5889 66. The generation of the Received: header has been moved from the time that a
5890 message starts to be received, to the time that it finishes. The timestamp
5891 in the Received: header should now be very close to that of the <= log
5892 line. There are two side-effects of this change:
5894 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5895 logged header lines no longer include the local Received: line, because
5896 it has not yet been created. The same applies to a copy of the message
5897 that is returned to a non-SMTP sender when a message is rejected.
5899 (b) When a filter file is tested using -bf, no additional Received: header
5900 is added to the test message. After some thought, I decided that this
5903 This change does not affect the value of $received_for. It is still set
5904 after address rewriting, but before local_scan() is called.
5906 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5908 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5909 gave an unhelpful panic error message, and a defer error. I have managed to
5910 change this behaviour so that it now rejects any supplied certificate,
5911 which seems right, as the list of acceptable certificates is empty.
5913 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5914 gave an unhelpful defer error. I have not managed to make this reject any
5915 supplied certificates, but the error message it gives is "no certificate
5916 supplied", which is not helpful.
5918 70. exigrep's output now also includes lines that are not associated with any
5919 message, but which match the given pattern. Implemented by a patch from
5920 Martin Sluka, which also tidied up the Perl a bit.
5922 71. Recipient callout verification, like sender verification, was using <> in
5923 the MAIL FROM command. This isn't really the right thing, since the actual
5924 sender may affect whether the remote host accepts the recipient or not. I
5925 have changed it to use the actual sender in the callout; this means that
5926 the cache record is now keyed on a recipient/sender pair, not just the
5927 recipient address. There doesn't seem to be a real danger of callout loops,
5928 since a callout by the remote host to check the sender would use <>.
5929 [SEE ABOVE: changed after hitting problems.]
5931 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5932 temporary errors. However, in the case of such a code being given after
5933 the end of a data transmission (i.e. after ".") Exim was failing to write
5934 a retry record for the message. (Yes, there was some broken host that was
5935 actually sending 8xx at this point.)
5937 73. An unknown lookup type in a host list could cause Exim to panic-die when
5938 the list was checked. (An example that provoked this was putting <; in the
5939 middle of a list instead of at the start.) If this happened during a DATA
5940 ACL check, a -D file could be left lying around. This kind of configuration
5941 error no longer causes Exim to die; instead it causes a defer error. The
5942 incident is still logged to the main and panic logs.
5944 74. Buglet left over from Exim 3 conversion. The message "too many messages
5945 in one connection" was written to the rejectlog but not the mainlog, except
5946 when address rewriting (yes!) was being logged.
5948 75. Added write_rejectlog option.
5950 76. When a system filter was run not as root (that is, when system_filter_user
5951 was set), the values of the $n variables were not being returned to the
5952 main process; thus, they were not subsequently available in the $sn
5955 77. Added +return_path_on_delivery log selector.
5957 78. A connection timeout was being treated differently from recipients deferred
5958 when testing hosts_max_try with a message that was older than the host's
5959 retry timeout. (The host should not be counted, thus allowing all hosts to
5960 be tried at least once before bouncing.) This may have been the cause of an
5961 occasionally reported bug whereby a message would remain on the queue
5962 longer than the retry timeout, but would be bounced if a delivery was
5963 forced. I say "may" because I never totally pinned down the problem;
5964 setting up timeout/retry tests is difficult. See also the next item.
5966 79. The ultimate address timeout was not being applied to errors that involved
5967 a combination of host plus message (for example, a timeout on a MAIL
5968 command). When an address resolved to a number of possible hosts, and they
5969 were not all tried for each delivery (e.g. because of hosts_max_try), a
5970 message could remain on the queue longer than the retry timeout.
5972 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5975 81. Fixed an obscure SMTP outgoing bug which required at least the following
5976 conditions: (a) there was another message waiting for the same server;
5977 (b) the server returned 5xx to all RCPT commands in the first message so
5978 that the message was not completed; (c) the server dropped the connection
5979 or gave a negative response to the RSET that Exim sends to abort the
5980 transaction. The observed case was a dropped connection after DATA that had
5981 been sent in pipelining mode. That is, the server had advertised PIPELINING
5982 but was not implementing it correctly. The effect of the bug was incorrect
5983 behaviour, such as trying another host, and this could lead to a crash.
5989 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5990 and daemon.c were passed as pointers to ints; they should have been
5991 pointers to socklen_t variables (which are typically unsigned ints).
5993 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5996 3. Fixed a really odd bug that affected only the testing scheme; patching a
5997 certain fixed string in the binary changed the value of another string that
5998 happened to be identical to the end of the original first string.
6000 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6001 name", it returns that address as the IP address. On some operating
6002 systems (e.g. Solaris), it also passes back the IP address string as the
6003 "host name". However, on others (e.g. Linux), it passes back an empty
6004 string. Exim wasn't checking for this, and was changing the host name to an
6005 empty string, assuming it had been canonicalized.
6007 5. Although rare, it is permitted to have more than one PTR record for a given
6008 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6009 all the names associated with an address, because they do in Solaris.
6010 However, it seems that they do not in Linux for data that comes from the
6011 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6012 I found this out when I moved to a new Linux workstation and tried to run
6013 the Exim test suite.
6015 To get round this problem I have changed the code so that it now does its
6016 own call to the DNS to look up PTR records when searching for a host name.
6017 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6018 addresses that are only in /etc/hosts are still found.
6020 This behaviour is, however, controlled by an option called host_lookup_
6021 order, which defaults to "bydns:byaddr". If people want to use the other
6022 order, or indeed, just use one or the other means of lookup, they can
6023 specify it in this variable.
6025 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6026 some operating systems, this comes back from gethostbyaddr() as an empty
6027 string, and this is what Exim used to test for. However, it seems that in
6028 other systems, "." is yielded. Exim now tests for this case too.
6030 7. The values of check_spool_space and check_log_space are now held internally
6031 as a number of kilobytes instead of an absolute number of bytes. If a
6032 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6033 kilobyte. This means that much larger values can be stored.
6035 8. Exim monitor: an attempt to get the action menu when not actually pointing
6036 at a message produces an empty menu entitled "No message selected". This
6037 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6038 no entries in it ("Shell widget menu has zero width and/or height"). So I
6039 have added a single, blank menu entry in this case.
6041 9. Added ${quote_local_part.
6043 10. MIME decoding is now applied to the contents of Subject: header lines when
6046 11. Now that a reference to $sender_host_address automatically causes a reverse
6047 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6048 host lookup before query-style lookups in lists that might use this
6049 variable. This has therefore been abolished, and the "net-" prefix is no
6050 longer necessary for query-style lookups.
6052 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6053 have been a typo for LFLAGS, so it has been changed.
6055 13. The install script calls Exim with "-C /dev/null" in order to find the
6056 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6057 to be output. However, since Exim outputs its version number before the
6058 error, it didn't break the script. It just looked ugly. I fixed this by
6059 always allowing "-C /dev/null" if the caller is root.
6061 14. Ignore overlarge ACL variable number when reading spool file - insurance
6062 against a later release with more variables having written the file.
6064 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6065 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6066 this, as well as the form without the "IPv6" on the front.
6068 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6069 OS/Makefile-Darwin file.
6071 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6072 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6074 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6075 variable. However, some OS use ints for inodes. I've added cast to long int
6076 to get rid of the compiler warning.
6078 19. I had forgotten to lock out "/../" in configuration file names when
6079 ALT_CONFIG_PREFIX was set.
6081 20. Routers used for verification do not need to specify transports. However,
6082 if such a router generated a host list, and callout was configured, Exim
6083 crashed, because it could not find a port number from the (non-existent)
6084 transport. It now assumes port 25 in this circumstance.
6086 21. Added the -t option to exigrep.
6088 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6089 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6090 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6091 them is set, LOOKUP_LSEARCH is forced.
6093 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6095 24. Added sender and host information to the "rejected by local_scan()" log
6096 line; previously there was no indication of these.
6098 25. Added .include_if_exists.
6100 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6101 Linux. It turns out that not all file systems support this. Apparently some
6102 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6103 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6104 not supported on the file descriptor, is now ignored when Exim is trying to
6105 sync a directory. This applies only to Linux.
6107 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6109 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6110 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6113 29. Added support for maildirsize files from supplied patch (modified a bit).
6115 30. The use of :fail: followed by an empty string could lead Exim to respond to
6116 sender verification failures with (e.g.):
6118 550 Verification failed for <xxx>
6119 550 Sender verify failed
6121 where the first response line was missing the '-' that indicates it is not
6122 the final line of the response.
6124 31. The loop for finding the name of the user that called Exim had a hardwired
6125 limit of 10; it now uses the value of finduser_retries, which is used for
6126 all other user lookups.
6128 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6131 33. Exim was neglecting to zero errno before one call of strtol() when
6132 expanding a string and expecting an integer value. On some systems this
6133 resulted in spurious "integer overflow" errors. Also, it was casting the
6134 result into an int without checking.
6136 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6137 did not work. The code looks as if it has *never* worked, though it appears
6138 to have been documented since at least release 1.62. I have made it work.
6140 35. The "timeout_DNS" error in retry rules, also documented since at least
6141 1.62, also never worked. As it isn't clear exactly what this means, and
6142 clearly it isn't a major issue, I have abolished the feature by treating it
6143 as "timeout", and writing a warning to the main and panic logs.
6145 36. The display of retry rules for -brt wasn't always showing the error code
6148 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6149 timeout_connect_A, timeout_connect_MX.
6151 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6152 to the empty sender.
6154 39. The daemon was not analysing the content of -oX till after it had closed
6155 stderr and disconnected from the controlling terminal. This meant that any
6156 syntax errors were only noted on the panic log, and the return code from
6157 the command was 0. By re-arranging the code a little, I've made the
6158 decoding happen first, so such errors now appear on stderr, and the return
6159 code is 1. However, the actual setting up of the sockets still happens in
6160 the disconnected process, so errors there are still only recorded on the
6163 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6164 connections (as happens on some IP stacks) was logged at start up time as
6165 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6166 it from "IPv6 and IPv4", which means that two separate sockets are being
6169 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6170 says whether AF_INET or AF_INET6 was passed as an argument.
6172 42. Exiwhat output was messed up when time zones were included in log
6175 43. Exiwhat now gives more information about the daemon's listening ports,
6176 and whether -tls-on-connect was used.
6178 44. The "port" option of the smtp transport is now expanded.
6180 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6181 silently ignored. Now an error message is written to the main and panic
6184 46. There's a new ACL modifier called "logwrite" which writes to a log file
6185 as soon as it is encountered.
6187 47. Added $local_user_uid and $local_user_gid at routing time.
6189 48. Exim crashed when trying to verify a sender address that was being
6192 49. Exim was recognizing only a space character after ".include". It now also
6193 recognizes a tab character.
6195 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6196 extracting the relevant information from the specification. The man page no
6197 longer contains scrambled data for the -d option, and I've added a section
6198 at the front about calling Exim under different names.
6200 51. Added "extra_headers" argument to the "mail" command in filter files.
6202 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6205 53. Installed eximstats 1.29.
6207 54. Added transport_filter_timeout as a generic transport option.
6209 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6210 Cc: header lines. This was required by RFC 822, but it not required by RFC
6213 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6214 incoming messages that did not have them. Now it does so only if the
6215 message originates locally, that is, if there is no associated remote host
6216 address. When Resent- header lines are present, this applies to the Resent-
6217 lines rather than the non-Resent- lines.
6219 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6220 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6222 58. Messages for configuration errors now include the name of the main
6223 configuration file - useful now that there may be more than one file in a
6224 list (.included file names were always shown).
6226 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6227 for those rare installations that do not start the daemon as root or run it
6228 setuid root. I've cut out the call to initgroups() if the daemon is not
6231 60. The Exim user and group can now be bound into the binary as text strings
6232 that are looked up at the start of Exim's processing.
6234 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6236 62. Added $mailstore_basename variable.
6238 63. Installed patch to sieve.c from Michael Haardt.
6240 64. When Exim failed to open the panic log after failing to open the main log,
6241 the original message it was trying to log was written to stderr and debug
6242 output, but if they were not available (the usual case in production), it
6243 was lost. Now it is written to syslog before the two lines that record the
6244 failures to open the logs.
6246 65. Users' Exim filters run in subprocesses under the user's uid. It is
6247 possible for a "deliver" command or an alias in a "personal" command to
6248 provoke an address rewrite. If logging of address rewriting is configured,
6249 this fails because the process is not running as root or exim. There may be
6250 a better way of dealing with this, but for the moment (because 4.30 needs
6251 to be released), I have disabled address rewrite logging when running a
6252 filter in a non-root, non-exim process.
6258 1. The buildconfig auxiliary program wasn't quoting the value set for
6259 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6260 not defined. This bug was masked in 4.22 by the effect that was fixed in
6263 2. Some messages that were rejected after a message id was allocated were
6264 shown as "incomplete" by exigrep. It no longer does this for messages that
6265 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6267 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6268 have allow_domain_literals set, the ID did not get logged in the <= line.
6269 Domain literals are now always recognized in Message-ID: header lines.
6271 4. The first argument for a ${extract expansion item is the key name or field
6272 number. Leading and trailing spaces in this item were not being ignored,
6273 causing some misleading effects.
6275 5. When deliver_drop_privilege was set, single queue runner processes started
6276 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6277 same command in the process it spins off) were not dropping privilege.
6279 6. When the daemon running as "exim" started a queue runner, it always
6280 re-executed Exim in the spun-off process. This is a waste of effort when
6281 deliver_drop_privilege is set. The new process now just calls the
6282 queue-runner function directly.
6288 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6291 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6292 ignored. Though the use of -oP was forcing the writing of a pid file, it
6293 was always written to the default place.
6295 3. If the message "no IP address found for host xxxx" is generated during
6296 incoming verification, it is now followed by identification of the incoming
6297 connection (so you can more easily find what provoked it).
6299 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6301 5. Added some features to "harden" Exim a bit more against certain attacks:
6303 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6304 be put in Local/Makefile. This is like the never_users runtime option,
6305 but it cannot be overridden. The default setting is "root".
6307 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6308 prefix string with which any file named in a -C command line option
6311 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6312 is retained for -C and -D only if the caller of Exim is root. Without
6313 it, the exim user may also use -C and -D and retain privilege.
6315 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6316 command line option is disabled.
6318 6. Macro names set by the -D option must start with an upper case letter, just
6319 like macro names defined in the configuration file.
6321 7. Added "dereference=" facility to LDAP.
6323 8. Two instances of the typo "uknown" in the source files are fixed.
6325 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6326 the Configure-Makefile script screwed up while processing it.
6328 10. Incorporated PCRE 4.4.
6330 11. The SMTP synchronization check was not operating right at the start of an
6331 SMTP session. For example, it could not catch a HELO sent before the client
6332 waited for the greeting. There is now a check for outstanding input at the
6333 point when the greeting is written. Because of the duplex, asynchronous
6334 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6335 way, but not yet received, when the check is performed.
6337 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6338 on TCP/IP sockets, because this apparently causes some broken clients to
6341 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6342 unchanged) from the Cygwin maintainer.
6344 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6345 was supported instead of testing for mbx. Effectively a typo.
6347 15. The spa authenticator server code was not checking that the input it
6348 received was valid base64.
6350 16. The debug output line for the "set" modifier in ACLs was not showing the
6351 name of the variable that was being set.
6353 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6355 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6356 lookup on its own. Something else was needed to trigger the lookup. For
6357 example, a match in host_lookup or the need for a host name in a host list.
6358 Now, if $sender_host_name is referenced and the host name has not yet been
6359 looked up, a lookup is performed. If the lookup fails, the variable remains
6360 empty, and $host_lookup_failed is set to "1".
6362 19. Added "eqi" as a case-independent comparison operator.
6364 20. The saslauthd authentication condition could segfault if neither service
6365 nor realm was specified.
6367 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6368 error message that was logged was misleading, and incoming SMTP
6369 connections were dropped. The message is now more accurate, and temporary
6370 errors are given to SMTP connections.
6372 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6373 (except RSET) with a 421 error, until QUIT is received. However, it was
6374 failing to send a response to QUIT.
6376 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6377 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6378 ACL is now run after the helo_try_verify_hosts code.
6380 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6381 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6382 case-independent, so other case variants are also recognized). Apparently
6383 some systems use these upper case variants.
6385 25. If more than two messages were waiting for the same host, and a transport
6386 filter was specified for the transport, Exim sent two messages over the
6387 same TCP/IP connection, and then failed with "socket operation on non-
6388 socket" when it tried to send the third.
6390 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6392 27. The extern definition of crypt16() in expand.c was not being excluded when
6393 the OS had its own crypt16() function.
6395 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6396 as a preferred synonym for return_size_limit, both as an option and as an
6399 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6401 30. Changed the default configuration ACL to relax the local part checking rule
6402 for addresses that are not in any local domains. For these addresses,
6403 slashes and pipe symbols are allowed within local parts, but the sequence
6404 /../ is explicitly forbidden.
6406 31. SPA server authentication was not clearing the challenge buffer before
6409 32. log_message in a "warn" ACL statement was writing to the reject log as
6410 well as to the main log, which contradicts the documentation and doesn't
6411 seem right (because no rejection is happening). So I have stopped it.
6413 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6414 However, I am unable to do any testing of this.
6416 34. Fixed an infelicity in the appendfile transport. When checking directories
6417 for a mailbox, to see if any needed to be created, it was accidentally
6418 using path names with one or more superfluous leading slashes; tracing
6419 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6421 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6422 contents are added to the log line that is written for every discarded
6423 recipient. (Previously a log_message setting was ignored.)
6425 36. The ${quote: operator now quotes the string if it is empty.
6427 37. The install script runs exim in order to find its version number. If for
6428 some reason other than non-existence or emptiness, which it checks, it
6429 could not run './exim', it was installing it with an empty version number,
6430 i.e. as "exim-". This error state is now caught, and the installation is
6433 38. An argument was missing from the function that creates an error message
6434 when Exim fails to connect to the socket for saslauthd authentication.
6435 This could cause Exim to crash, or give a corrupted message.
6437 39. Added isip, isip4, and isip6 to ${if conditions.
6439 40. The ACL variables $acl_xx are now saved with the message, and can be
6440 accessed later in routers, transports, and filters.
6442 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6443 strings in the file are not string-expanded.
6445 42. If a MAIL command specified a SIZE value that was too large to fit into an
6446 int variable, the check against message_size_limit failed. Such values are
6447 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6448 day this will have to be increased, but I don't think I want to be around
6449 when emails are that large.
6456 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6457 iconv() is not standard in FreeBSD.
6459 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6460 IPv6 enabled. The observed symptom was a segmentation fault on return from
6461 the function os_common_find_running_interfaces() in src/os.c.
6463 3. In the check_special_case() function in daemon.c I had used "errno" as an
6464 argument name, which causes warnings on some systems. This was basically a
6465 typo, since it was named "eno" in the comments!
6467 4. The code that waits for the clock to tick (at a resolution of some fraction
6468 of a second) so as to ensure message-id uniqueness was always waiting for
6469 at least one whole tick, when it could have waited for less. [This is
6470 almost certainly not relevant at current processor speeds, where it is
6471 unlikely to ever wait at all. But we try to future-proof.]
6473 5. The function that sleeps for a time interval that includes fractions of a
6474 second contained a race. It did not block SIGALRM between setting the
6475 timer, and suspending (a couple of lines later). If the interval was short
6476 and the sigsuspend() was delayed until after it had expired, the suspension
6477 never ended. On busy systems this could lead to processes getting stuck for
6480 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6481 process, before it forks any delivery processes. The open lookup caching
6482 mechanism meant that the open file or database connection was passed into
6483 the delivery process. The problem was that delivery processes always tidy
6484 up cached lookup data. This could cause a problem for the next delivery
6485 process started by the queue runner, because the external queue runner
6486 process does not know about the closure. So the next delivery process
6487 still has data in the lookup cache. In the case of a file lookup, there was
6488 no problem because closing a file descriptor in a subprocess doesn't affect
6489 the parent. However, if the lookup was caching a connection to a database,
6490 the connection was closed, and the second delivery process was likely to
6491 see errors such as "PGSQL: query failed: server closed the connection
6492 unexpectedly". The problem has been fixed by closing all cached lookups
6493 in a queue runner before running a delivery process.
6495 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6496 seem to have the "const" qualifier which it has on other OS. I've
6499 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6500 *of the same type* (client or server) with the same public name that an
6501 error should be diagnosed.
6503 9. When Exim looked up a host name for an IP address, but failed to find the
6504 original IP address when looking up the host name (a safety check), it
6505 output the message "<ip address> does not match any IP for NULL", which was
6506 confusing, to say the least. The bug was that the host name should have
6507 appeared instead of "NULL".
6509 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6510 user that is built into the binary, and the -C or -D options is used, root
6511 privilege is dropped before the configuration file is read. In addition,
6512 logging is switched to stderr instead of the normal log files. If the
6513 configuration then re-defines the Exim user, the unprivileged environment
6514 is probably not what is expected, so Exim logs a panic warning message (but
6517 However, if deliver_drop_privilege is set, the unprivileged state may well
6518 be exactly what is intended, so the warning has been cut out in that case,
6519 and Exim is allowed to try to write to its normal log files.
6525 1. smtp_return_error_details was not giving details for temporary sender
6526 or receiver verification errors.
6528 2. Diagnose a configuration error if two authenticators have the same public
6531 3. Exim used not to create the message log file for a message until the first
6532 delivery attempt. This could be confusing when incoming messages were held
6533 for policy or load reasons. The message log file is now created at the time
6534 the message is received, and an initial "Received" line is written to it.
6536 4. The automatically generated man page for command line options had a minor
6537 bug that caused no ill effects; however, a more serious problem was that
6538 the procedure for building the man page automatically didn't always
6539 operate. Consequently, release 4.20 contains an out-of-date version. This
6540 shouldn't happen again.
6542 5. When building Exim with embedded Perl support, the script that builds the
6543 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6544 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6546 6. The freeze_tell option was not being used for messages that were frozen on
6547 arrival, either by an ACL or by local_scan().
6549 7. Added the smtp_incomplete_transaction log selector.
6551 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6552 was accepting AUTH without a new EHLO.
6554 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6555 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6558 10. Logging of TCP/IP connections (when configured) now happens in the main
6559 daemon process instead of the child process, so that the TCP/IP connection
6560 count is more accurate (but it can never be perfect).
6562 11. The use of "drop" in a nested ACL was not being handled correctly in the
6563 outer ACL. Now, if condition failure induced by the nested "drop" causes
6564 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6565 or "require"), the connection is dropped.
6567 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6568 that yield "discard" can now be used with an "accept" or a "discard" verb,
6569 but an error is generated for any others (because I can't see a useful way
6570 to define what should happen).
6572 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6573 are now processed in the same way as lines in the Exim configuration file.
6574 In particular, continuation lines are supported.
6576 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6578 15. Added -ti meaning -t -i.
6580 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6581 domains, and warn by logging if others are found.
6583 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6584 returned from GIFCONF when Exim is trying to find the list of interfaces on
6585 a host. The code in os.c has been modified to copy each ifreq to an aligned
6586 structure in all cases.
6588 Also, in some cases, the returned ifreq's were being copied to a 'struct
6589 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6590 means the last couple of bytes of an IPv6 address could be chopped if the
6591 ifreq contained only a normal sockaddr (14 bytes storage).
6593 18. Named domain lists were not supported in the hosts_treat_as_local option.
6594 An entry such as +xxxx was not recognized, and was treated as a literal
6597 19. Ensure that header lines added by a DATA ACL are included in the reject log
6598 if the ACL subsequently rejects the message.
6600 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6601 MD5 (which is deprecated).
6603 21. When testing a filter file using -bf, Exim was writing a message when it
6604 took the sender from a "From " line in the message, but it was not doing so
6605 when it took $return_path from a Return-Path: header line. It now does.
6607 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6608 with a valid header line field name (a series of printing characters
6609 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6611 23. Changed "disc" in the source to "disk" to conform to the documentation and
6612 the book and for uniformity.
6614 24. Ignore Sendmail's -Ooption=value command line item.
6616 25. When execve() failed while trying to run a command in a pipe transport,
6617 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6618 could be confused with a return value of 69 from the command itself. This
6619 has been changed to 127, the value the shell returns if it is asked to run
6620 a non-existent command. The wording for the related log line suggests a
6621 non-existent command as the problem.
6623 26. If received_header_text expands to an empty string, do not add a Received:
6624 header line to the message. (Well, it adds a token one on the spool, but
6625 marks it "old" so that it doesn't get used or transmitted.)
6627 27. Installed eximstats 1.28 (addition of -nt option).
6629 28. There was no check for failure on the call to getsockname() in the daemon
6630 code. This can fail if there is a shortage of resources on the system, with
6631 ENOMEM, for example. A temporary error is now given on failure.
6633 29. Contrary to the C standard, it seems that in some environments, the
6634 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6635 program. Exim now does this explicitly; it affects the formatting of
6636 timestamps using strftime().
6638 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6639 complaints. I've now initialized all the variables, to avoid this.
6641 32. Header lines added by a system filter were not being "seen" during
6642 transport-time rewrites.
6644 33. The info_callback() function passed to OpenSSL is set up with type void
6645 (*)(SSL *, int, int), as described somewhere. However, when calling the
6646 function (actually a macro) that sets it up, the type void(*)() is
6647 expected. I've put in a cast to prevent warnings from picky compilers.
6649 34. If a DNS black list lookup found a CNAME record, but there were no A
6650 records associated with the domain it pointed at, Exim crashed.
6652 35. If a DNS black list lookup returned more than one A record, Exim ignored
6653 all but the first. It now scans all returned addresses if a particular IP
6654 value is being sought. In this situation, the contents of the
6655 $dnslist_value variable are a list of all the addresses, separated by a
6658 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6659 to accept a host name and all its aliases if the forward lookup for any of
6660 them yielded the IP address of the incoming connection. Now it accepts only
6661 those names whose forward lookup yields the correct IP address. Any other
6662 names are discarded. This closes a loophole whereby a rogue DNS
6663 administrator could create reverse DNS records to break through a
6664 wildcarded host restriction in an ACL.
6666 37. If a user filter or a system filter that ran in a subprocess used any of
6667 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6668 the wrong values were passed to the pipe command ($thisaddress had the
6669 value of $0, $0 had the value of $1, etc). This bug was introduced by
6670 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6672 38. Improved the line breaking for long SMTP error messages from ACLs.
6673 Previously, if there was no break point between 40 and 75 characters, Exim
6674 left the rest of the message alone. Two changes have been made: (a) I've
6675 reduced the minimum length to 35 characters; (b) if it can't find a break
6676 point between 35 and 75 characters, it looks ahead and uses the first one
6677 that it finds. This may give the occasional overlong line, but at least the
6678 remaining text gets split now.
6680 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6681 file descriptors might be low, and that setting 1000 would always raise it.
6682 It turns out that in some environments, the limit is already over 1000 and
6683 that lowering it causes trouble. So now Exim takes care not to decrease it.
6685 40. When delivering a message, the value of $return_path is set to $sender_
6686 address at the start of routing (routers may change the value). By an
6687 oversight, this default was not being set up when an address was tested by
6688 -bt or -bv, which affected the outcome if any router or filter referred to
6691 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6692 log only when "message" or "log_message" are set. However, if one of the
6693 conditions was an address verification, or a call to a nested ACL, the
6694 messages generated by the underlying test were being passed through. This
6695 no longer happens. The underlying message is available in $acl_verify_
6696 message for both "message" and "log_message" expansions, so it can be
6697 passed through if needed.
6699 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6700 new expansion $bh_ to give the encoded byte string without charset
6701 translation. Translation happens only if iconv() is available; HAVE_ICONV
6702 indicates this at build time. HEADERS_CHARSET gives the charset to
6703 translate to; headers_charset can change it in the configuration, and
6704 "headers charset" can change it in an individual filter file.
6706 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6707 that creates RFC 2047 encoded "words" labels them as that charset instead
6708 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6709 expansion operator; (ii) when Exim creates a From: line for a local
6710 message; (iii) when a header line is rewritten to include a "phrase" part.
6712 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6713 buggy, causing it to skip the first lines of messages whose message ID
6714 ended in 'D'. This would not have bitten before Exim release 4.14, because
6715 message IDs were unlikely to end in 'D' before then. The effect was to have
6716 incorrect size information for certain domains.
6718 45. #include "config.h" was missing at the start of the crypt16.c module. This
6719 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6722 46. If there was a timeout during a "random" callout check, Exim treated it as
6723 a failure of the random address, and carried on sending RSET and the real
6724 address. If the delay was just some slowness somewhere, the response to the
6725 original RCPT would be taken as a response to RSET and so on, causing
6726 mayhem of various kinds.
6728 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6729 when I implemented it. It didn't allow for the fact that some option values
6730 may legitimately be negative (e.g. size_addition), and it didn't even do
6731 the right test for positive values.
6733 48. Domain names in DNS records are case-independent. Exim always looks them up
6734 in lower case. Some resolvers return domain names in exactly the case they
6735 appear in the zone file, that is, they may contain uppercase letters. Not
6736 all resolvers do this - some return always lower case. Exim was treating a
6737 change of case by a resolver as a change of domain, similar to a widening
6738 of a domain abbreviation. This triggered its re-routing code and so it was
6739 trying to route what was effectively the same domain again. This normally
6740 caused routing to fail (because the router wouldn't handle the domain
6741 twice). Now Exim checks for this case specially, and just changes the
6742 casing of the domain that it ultimately uses when it transmits the message
6745 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6748 50. If a filter generated a file delivery with a non-absolute name (possible if
6749 no home directory exists for the router), the forbid_file option was not
6752 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6753 the existing equality matching.
6755 52. Exim was using ints instead of ino_t variables in some places where it was
6756 dealing with inode numbers.
6758 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6759 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6760 called TMPDIR, and if it finds it is different, it changes its value.
6762 54. The smtp_printf() function is now made available to local_scan() so
6763 additional output lines can be written before returning. There is also an
6764 smtp_fflush() function to enable the detection of a dropped connection.
6765 The variables smtp_input and smtp_batched_input are exported to
6768 55. Changed the default runtime configuration: the message "Unknown user"
6769 has been removed from the ACL, and instead placed on the localuser router,
6770 using the cannot_route_message feature. This means that any verification
6771 failures that generate their own messages won't get overridden. Similarly,
6772 the "Unrouteable address" message that was in the ACL for unverifiable
6773 relay addresses has also been removed.
6775 56. Added hosts_avoid_esmtp to the smtp transport.
6777 57. The exicyclog script was not checking for the esoteric option
6778 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6779 will work only if exicyclog is run under the appropriate euid.
6781 58. Following a discussion on the list, the rules by which Exim recognises line
6782 endings on incoming messages have been changed. The -dropcr and drop_cr
6783 options are now no-ops, retained only for backwards compatibility. The
6784 following line terminators are recognized: LF CRLF CR. However, special
6785 processing applies to CR:
6787 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6788 nor a local message in the state where . is a terminator.
6790 (ii) If a bare CR is encountered in a header line, an extra space is added
6791 after the line terminator so as not to end the header. The reasoning
6792 behind this is that bare CRs in header lines are most likely either
6793 to be mistakes, or people trying to play silly games.
6795 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6796 was being incorrectly given as 18 bytes larger than it should have been.
6797 This is a VOB (very old bug).
6799 60. This may never have affected anything current, but just in case it has:
6800 When the local host is found other than at the start of a list of hosts,
6801 the local host, those with the same MX, and any that follow, are discarded.
6802 When the list in question was part of a longer list of hosts, the following
6803 hosts (not currently being processed) were also being discarded. This no
6804 longer happens. I'm not sure if this situation could ever has previously
6807 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6810 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6811 header. This is something that is recommended in a new Internet Draft, and
6812 is something that is documented as being done by Sendmail. There are two
6813 possible values. For messages generated by the autoreply transport, Exim
6816 Auto-Submitted: auto-replied
6818 whereas for all other generated messages (e.g. bounces) it adds
6820 Auto-Submitted: auto-generated
6822 63. The "personal" condition in filters now includes a test for the
6823 Auto-Submitted: header. If it contains the string "auto-" the message it
6824 not considered personal.
6826 64. Added rcpt_include_affixes as a generic transport option.
6828 65. Added queue_only_override (default true).
6830 66. Added the syslog_duplication option.
6832 67. If what should have been the first header line of a message consisted of
6833 a space followed by a colon, Exim was mis-interpreting it as a header line.
6834 It isn't of course - it is syntactically invalid and should therefore be
6835 treated as the start of the message body. The misbehaviour could have
6836 caused a number of strange effects, including loss of data in subsequent
6837 header lines, and spool format errors.
6839 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6840 client host had authenticated. This control can now be exercised by an ACL
6841 for more flexibility.
6843 69. By default, callouts do not happen when testing with -bh. There is now a
6844 variant, -bhc, which does actually run the callout code, including
6845 consulting and updating the callout cache.
6847 70. Added support for saslauthd authentication, courtesy of Alexander
6850 71. If statvfs() failed on the spool or log directories while checking their
6851 size for availability, Exim confusingly gave the error "space shortage".
6852 Furthermore, in debugging mode it crashed with a floating point exception.
6853 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6854 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6855 serious problem, Exim now writes to the main and panic logs when this
6856 happens, with details of the failure. It then refuses to accept the
6857 incoming message, giving the message "spool directory problem" or "log
6858 directory problem" with a 421 code for SMTP messages.
6860 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6861 0, 1, and 2 exist, because some OS complain for execs without them (see
6862 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6863 descriptors. However, the code omitted to check that the open succeeded,
6864 causing mysterious errors if for some reason the permissions on /dev/null
6865 got screwed. Now Exim writes a message to the main and panic logs, and
6866 bombs out if it can't open /dev/null.
6868 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6869 interact so that it is all more flexible. It is supposed to remain
6870 backwards compatible. Also added extra_local_interfaces.
6872 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6873 to bomb out with an assertion failure - to the client this appears as a
6874 connection drop. This problem occurs in the part of the code that was taken
6875 from the Samba project. Fortunately, the assertion is in a very simple
6876 function, so I have fixed this by reproducing the function inline in the
6877 one place where it is called, and arranging for authentication to fail
6878 instead of killing the process with assert().
6880 75. The SPA client code was not working when the server requested OEM rather
6881 than Unicode encoding.
6883 76. Added code to make require_files with a specific uid setting more usable in
6884 the case where statting the file as root fails - usually a non-root-mounted
6885 NFS file system. When this happens and the failure is EACCES, Exim now
6886 forks a subprocess and does the per-uid checking as the relevant uid.
6888 77. Added process_log_path.
6890 78. If log_file_path was not explicitly set, a setting of check_log_space or
6891 check_log_inodes was ignored.
6893 79. If a space check for the spool or log partitions fails, the incident is now
6894 logged. Of course, in the latter case the data may get lost...
6896 80. Added the %p formatting code to string_format() so that it can be used to
6897 print addresses in debug_print(). Adjusted all the address printing in the
6898 debugging in store.c to use %p rather than %d.
6900 81. There was a concern that a line of code in smtp_in.c could overflow a
6901 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6902 initially expressed, the concern was not well-founded, because trailing
6903 spaces are removed early. However, if the trailing spaces were followed by
6904 a NULL, they did not get removed, so the overflow was possible. Two fixes
6907 (a) I re-wrote the offending code in a cleaner fashion.
6908 (b) If an incoming SMTP command contains a NULL character, it is rejected
6911 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6912 runs initgroups(), so that if the Exim user is in any additional groups,
6913 they will be used during message reception.
6919 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob