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.
163 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
164 and confused the parent.
166 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
167 which could crash as a result. This could lead to undeliverable messages.
169 JH/28 Logging: "next input sent too soon" now shows where input was truncated
176 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
177 than -2003 did; needs libidn2 in addition to libidn.
179 JH/02 The path option on a pipe transport is now expanded before use.
181 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
182 Patch provided by "Björn", documentation fix added too.
184 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
185 missing a wire-to-host endian conversion.
187 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
188 close after a BDAT command line could be taken as a following command,
189 giving a synch failure. Fix by only checking for synch immediately
190 before acknowledging the chunk.
192 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
193 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
194 macro. Patches provided by Josh Soref.
196 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
197 Previously we did not; the RFC seems ambiguous and VRFY is not listed
198 by IANA as a service extension. However, John Klensin suggests that we
201 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
202 the dkim code may be unix-mode line endings rather than smtp wire-format
203 CRLF, so prepend a CR to any bare LF.
205 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
206 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
208 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
209 the first were themselves being wrongly included in the feed into dkim
210 processing; with most chunk sizes in use this resulted in an incorrect
211 body hash calculated value.
213 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
214 DKIM signature block, for verification. Although advised against by
215 standards it is specifically not ruled illegal.
217 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
219 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
220 missing a body hash (the bh= tag).
222 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
223 It seems that HAProxy sends the Proxy Protocol information in clear and
224 only then does a TLS startup, so do the same.
226 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
227 TCP connections (such as for Spamd) unless the daemon successfully set
228 Fast Open mode on its listening sockets. This fixes breakage seen on
229 too-old kernels or those not configured for Fast Open, at the cost of
230 requiring both directions being enabled for TFO, and TFO never being used
231 by non-daemon-related Exim processes.
233 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
234 endings, at least on the first header line. Try to canonify any that get
235 past that check, despite the cost.
237 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
238 now limited to an arbitrary five deep, while parsing addresses with the
239 strip_excess_angle_brackets option enabled.
241 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
242 instead leave the unprompted TLS handshake in socket buffer for the
243 TLS library to consume.
245 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
247 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
249 JH/16 Drop variables when they go out of scope. Memory management drops a whole
250 region in one operation, for speed, and this leaves assigned pointers
251 dangling. Add checks run only under the testsuite which checks all
252 variables at a store-reset and panics on a dangling pointer; add code
253 explicitly nulling out all the variables discovered. Fixes one known
254 bug: a transport crash, where a dangling pointer for $sending_ip_address
255 originally assigned in a verify callout, is re-used.
257 PP/06 Drop '.' from @INC in various Perl scripts.
259 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
261 PP/08 Reduce a number of compilation warnings under clang; building with
262 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
263 should be warning-free.
265 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
267 HS/01 Fix portability problems introduced by PP/08 for platforms where
268 realloc(NULL) is not equivalent to malloc() [SunOS et al].
270 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
271 chunk. This allows us to accept broken chunked messages. We need a more
272 general solution here.
274 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
275 already-broken messages in the queue.
277 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
279 JH/19 Fix reference counting bug in routing-generated-address tracking.
285 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
286 supports it and a size is available (ie. the sending peer gave us one).
288 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
289 years of being deprecated). The replacements are the ACLs
290 acl_smtp_mime and acl_not_smtp_mime.
292 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
293 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
294 or even an in-clear connection were permitted. Now, if the host lookup
295 was dnssec and dane was requested then the host is only used if the
296 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
297 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
298 if one fails this test.
299 This means that a poorly-configured remote DNS will make it incommunicado;
300 but it protects against a DNS-interception attack on it.
302 JH/04 Bug 1810: make continued-use of an open smtp transport connection
303 non-noisy when a race steals the message being considered.
305 JH/05 If main configuration option tls_certificate is unset, generate a
306 self-signed certificate for inbound TLS connections.
308 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
309 in rewrites and routers.
311 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
312 and logged a warning sing 4.83; now they are a configuration file error.
314 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
315 (lacking @domain). Apply the same qualification processing as RCPT.
317 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
319 JH/10 Support ${sha256:} applied to a string (as well as the previous
322 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
323 a cutthrough deliver is pending, as we always want to make a connection.
324 This also avoids re-routing the message when later placing the cutthrough
325 connection after a verify cache hit.
326 Do not update it with the verify result either.
328 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
329 when routing results in more than one destination address.
331 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
332 signing (which inhibits the cutthrough capability). Previously only
333 the presence of an option was tested; now an expansion evaluating as
334 empty is permissible (obviously it should depend only on data available
335 when the cutthrough connection is made).
337 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
338 the relevant preceding SMTP command did not note the pipelining mode.
340 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
341 Previously they were not counted.
343 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
344 as one having no matching records. Previously we deferred the message
345 that needed the lookup.
347 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
348 distinguished as "(=".
350 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
351 for missing MX records. Previously it only worked for missing A records.
353 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
355 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
356 after the data-go-ahead and data-ack. Patch from Jason Betts.
358 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
359 even for a "none" policy. Patch from Tony Meyer.
361 JH/22 Fix continued use of a connection for further deliveries. If a port was
362 specified by a router, it must also match for the delivery to be
365 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
366 When one of the recipients of a message was unsuitable for the connection
367 (has no matching addresses), we lost track of needing to mark it
368 deferred. As a result mail would be lost.
370 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
372 JH/25 Decoding ACL controls is now done using a binary search; the source code
373 takes up less space and should be simpler to maintain. Merge the ACL
374 condition decode tables also, with similar effect.
376 JH/26 Fix problem with one_time used on a redirect router which returned the
377 parent address unchanged. A retry would see the parent address marked as
378 delivered, so not attempt the (identical) child. As a result mail would
381 JH/27 Fix a possible security hole, wherein a process operating with the Exim
382 UID can gain a root shell. Credit to http://www.halfdog.net/ for
383 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
386 JH/28 Enable {spool,log} filesystem space and inode checks as default.
387 Main config options check_{log,spool}_{inodes,space} are now
388 100 inodes, 10MB unless set otherwise in the configuration.
390 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
391 Previously it only applied to the main-section connection policy
394 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
396 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
397 by me. Added RFC7919 DH primes as an alternative.
399 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
400 are not in the system include path.
402 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
403 GnuTLS, when a session startup failed (eg because the client disconnected)
404 Exim did stdio operations after fclose. This was exposed by a recent
405 change which nulled out the file handle after the fclose.
407 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
408 signed directly by the cert-signing cert, rather than an intermediate
409 OCSP-signing cert. This is the model used by LetsEncrypt.
411 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
413 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
414 an incoming connection.
416 HS/02 Bug 1802: Do not half-close the connection after sending a request
419 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
420 fallback to "prime256v1".
422 JH/34 SECURITY: Use proper copy of DATA command in error message.
423 Could leak key material. Remotely exploitable. CVE-2016-9963.
429 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
430 and 3.4.4 - once the server is enabled to respond to an OCSP request
431 it does even when not requested, resulting in a stapling non-aware
432 client dropping the TLS connection.
434 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
435 support variable-length bit vectors. No functional change.
437 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
438 The I= interface field on outgoing lines is now after the H= remote
439 host field, same as incoming lines. There is a separate
440 outgoing_interface log selector which allows you to disable the
443 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
444 If not running log_selector +smtp_connection the mainlog would be held
445 open indefinitely after a "too many connections" event, including to a
446 deleted file after a log rotate. Leave the per net connection logging
447 leaving it open for efficiency as that will be quickly detected by the
448 check on the next write.
450 HS/01 Bug 1671: Fix post transport crash.
451 Processing the wait-<transport> messages could crash the delivery
452 process if the message IDs didn't exist for some reason. When
453 using 'split_spool_directory=yes' the construction of the spool
454 file name failed already, exposing the same netto behaviour.
456 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
457 mime_regex ACL conditions.
459 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
460 to DSN fail messages (bounces): remote IP, remote greeting, remote response
461 to HELO, local diagnostic string.
463 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
464 log line to debug. Even when configured with a tls authenticator many
465 client connections are expected to not authenticate in this way, so
466 an authenticate fail is not an error.
468 HS/02 Add the Exim version string to the process info. This way exiwhat
469 gives some more detail about the running daemon.
471 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
472 matter for fast-change records such as DNSBLs.
474 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
475 retry record, even if constant. There may be multiple transports with
476 different interface settings and the retry behaviour needs to be kept
479 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
481 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
483 JH/10 Bug 840: fix log_defer_output option of pipe transport
485 JH/11 Bug 830: use same host for all RCPTS of a message, even under
486 hosts_randomize. This matters a lot when combined with mua_wrapper.
488 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
489 ${quote_pgsql:<string>} operator.
491 JH/13 Bug 1708: avoid misaligned access in cached lookup.
493 JH/14 Change header file name for freeradius-client. Relevant if compiling
494 with Radius support; from the Gentoo tree and checked under Fedora.
496 JH/15 Bug 1712: Introduce $prdr_requested flag variable
498 JH/16 Bug 1714: Permit an empty string as expansion result for transport
499 option transport_filter, meaning no filtering.
501 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
503 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
504 defaults to "*" (all hosts). The variable is now available when not built
505 with TLS, default unset, mainly to enable keeping the testsuite sane.
506 If a server certificate is not supplied (via tls_certificate) an error is
507 logged, and clients will find TLS connections fail on startup. Presumably
508 they will retry in-clear.
509 Packagers of Exim are strongly encouraged to create a server certificate
510 at installation time.
512 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
513 with the $config_file variable.
515 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
516 in transport context, after the attempt, and per-recipient. The latter type
517 is per host attempted. The event data is the error message, and the errno
518 information encodes the lookup type (A vs. MX) used for the (first) host,
519 and the trailing two digits of the smtp 4xx response.
521 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
522 to write to mainlog (or rejectlog, paniclog) in the window between file
523 creation and permissions/ownership being changed. Particularly affects
524 installations where exicyclog is run as root, rather than exim user;
525 result is that the running daemon panics and dies.
527 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
529 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
530 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
531 "pri" and "weight". Note that the previous implicit priority given by the
532 list order is no longer honoured.
534 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
537 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
538 by defining SUPPORT_SOCKS.
540 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
541 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
542 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
543 variables are renamed to proxy_{local,external}_{address,port}.
545 JH/27 Move Internationalisation support from Experimental to mainline, enabled
546 for a build by defining SUPPORT_I18N
548 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
549 of the query string, and make ${quote_redis:} do that quoting.
551 JH/29 Move Events support from Experimental to mainline, enabled by default
552 and removable for a build by defining DISABLE_EVENT.
554 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
556 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
557 cached by the daemon.
559 JH/32 Move Redis support from Experimental to mainline, enabled for a build
560 by defining LOOKUP_REDIS. The libhiredis library is required.
562 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
563 keys are given for lookup.
565 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
566 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
567 only supported when built with TLS support. The PolarSSL SHA routines
568 are still used when the TLS library is too old for convenient support.
570 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
571 openssl_options), for security. OpenSSL forces this from version 1.1.0
572 server-side so match that on older versions.
574 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
575 allocation for $value could be released as the expansion processing
576 concluded, but leaving the global pointer active for it.
578 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
579 and to use the domains and local_parts ACL conditions.
581 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
582 incorrectly not doubled on cutthrough transmission, hence seen as a
583 body-termination at the receiving system - resulting in truncated mails.
584 Commonly the sender saw a TCP-level error, and retransmitted the message
585 via the normal store-and-forward channel. This could result in duplicates
586 received - but deduplicating mailstores were liable to retain only the
587 initial truncated version.
589 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
591 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
593 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
594 we're in there, support oversigning also; bug 1309.
596 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
598 HS/04 Add support for keep_environment and add_environment options.
600 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
601 either intentional arithmetic overflow during PRNG, or testing config-
604 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
605 delivery resulted in actual delivery. Cancel cutthrough before DATA
608 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
609 rejects a recipient: pass the reject to the originator.
611 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
612 Many were false-positives and ignorable, but it's worth fixing the
615 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
616 for the new environment-manipulation done at startup. Move the routines
617 from being local to tls.c to being global via the os.c file.
619 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
620 an extract embedded as result-arg for a map, the first arg for extract
621 is unavailable so we cannot tell if this is a numbered or keyed
622 extraction. Accept either.
628 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
631 JH/02 The smtp transport option "multi_domain" is now expanded.
633 JH/03 The smtp transport now requests PRDR by default, if the server offers
636 JH/04 Certificate name checking on server certificates, when exim is a client,
637 is now done by default. The transport option tls_verify_cert_hostnames
638 can be used to disable this per-host. The build option
639 EXPERIMENTAL_CERTNAMES is withdrawn.
641 JH/05 The value of the tls_verify_certificates smtp transport and main options
642 default to the word "system" to access the system default CA bundle.
643 For GnuTLS, only version 3.0.20 or later.
645 JH/06 Verification of the server certificate for a TLS connection is now tried
646 (but not required) by default. The verification status is now logged by
647 default, for both outbound TLS and client-certificate supplying inbound
650 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
653 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
654 Status Notification (bounce) messages are now MIME format per RFC 3464.
655 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
656 under the control of the dsn_advertise_hosts option, and routers may
657 have a dsn_lasthop option.
659 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
660 default, modifiable by a malware= option. The list separator for
661 the options can now be changed in the usual way. Bug 68.
663 JH/10 The smtp_receive_timeout main option is now expanded before use.
665 JH/11 The incoming_interface log option now also enables logging of the
666 local interface on delivery outgoing connections.
668 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
669 if the interface and destination host and port all match.
671 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
674 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
675 Patch from Andrew Lewis.
677 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
678 now supports optional time-restrictions, weighting, and priority
679 modifiers per server. Patch originally by <rommer@active.by>.
681 JH/16 The spamd_address main option now supports a mixed list of local
682 and remote servers. Remote servers can be IPv6 addresses, and
683 specify a port-range.
685 JH/17 Bug 68: The spamd_address main option now supports an optional
686 timeout value per server.
688 JH/18 Bug 1581: Router and transport options headers_add/remove can
689 now have the list separator specified.
691 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
694 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
697 JH/21 Support for the A6 type of dns record is withdrawn.
699 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
700 rather than the verbs used.
702 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
703 from 255 to 1024 chars.
705 JH/24 Verification callouts now attempt to use TLS by default.
707 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
708 are generic router options now. The defaults didn't change.
710 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
711 Original patch from Alexander Shikoff, worked over by JH.
713 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
714 files if ZCAT_COMMAND is not executable.
716 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
718 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
720 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
721 Normally benign, it bites when the pair was led to by a CNAME;
722 modern usage is to not canonicalize the domain to a CNAME target
723 (and we were inconsistent anyway for A-only vs AAAA+A).
725 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
727 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
728 when evaluating $sender_host_dnssec.
730 JH/31 Check the HELO verification lookup for DNSSEC, adding new
731 $sender_helo_dnssec variable.
733 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
735 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
737 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
739 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
740 documented as working, but never had. Support all but $spam_report.
742 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
743 added for tls authenticator.
745 HS/03 Add perl_taintmode main config option
751 TL/01 When running the test suite, the README says that variables such as
752 no_msglog_check are global and can be placed anywhere in a specific
753 test's script, however it was observed that placement needed to be near
754 the beginning for it to behave that way. Changed the runtest perl
755 script to read through the entire script once to detect and set these
756 variables, reset to the beginning of the script, and then run through
757 the script parsing/test process like normal.
759 TL/02 The BSD's have an arc4random API. One of the functions to induce
760 adding randomness was arc4random_stir(), but it has been removed in
761 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
762 function when detected.
764 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
765 cause callback expansion.
767 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
768 syntax errors in an expansion can be treated as a string instead of
769 logging or causing an error, due to the internal use of bool_lax
770 instead of bool when processing it.
772 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
773 server certificates when making smtp deliveries.
775 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
777 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
779 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
781 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
782 Merged patch from Sebastian Wiedenroth.
784 JH/05 Fix results-pipe from transport process. Several recipients, combined
785 with certificate use, exposed issues where response data items split
786 over buffer boundaries were not parsed properly. This eventually
787 resulted in duplicates being sent. This issue only became common enough
788 to notice due to the introduction of connection certificate information,
789 the item size being so much larger. Found and fixed by Wolfgang Breyha.
791 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
792 size buffer was used, resulting in syntax errors when an expansion
795 JH/07 Add support for directories of certificates when compiled with a GnuTLS
796 version 3.3.6 or later.
798 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
799 is EXPERIMENTAL_EVENT, the main-configuration and transport options
800 both become "event_action", the variables become $event_name, $event_data
801 and $event_defer_errno. There is a new variable $verify_mode, usable in
802 routers, transports and related events. The tls:cert event is now also
803 raised for inbound connections, if the main configuration event_action
806 TL/06 In test suite, disable OCSP for old versions of openssl which contained
807 early OCSP support, but no stapling (appears to be less than 1.0.0).
809 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
810 server certificate names available under the smtp transport option
811 "tls_verify_cert_hostname" now do not permit multi-component wildcard
814 JH/10 Time-related extraction expansions from certificates now use the main
815 option "timezone" setting for output formatting, and are consistent
816 between OpenSSL and GnuTLS compilations. Bug 1541.
818 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
819 encoded parameter in the incoming message. Bug 1558.
821 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
822 include certificate info, eximon was claiming there were spoolfile
825 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
827 JH/14 Log delivery-related information more consistently, using the sequence
828 "H=<name> [<ip>]" wherever possible.
830 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
831 are problematic for Debian distribution, omit them from the release
834 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
836 JH/16 Fix string representation of time values on 64bit time_t architectures.
839 JH/17 Fix a null-indirection in certextract expansions when a nondefault
840 output list separator was used.
845 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
846 checkers that were complaining about end of non-void function with no
849 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
850 This was a regression introduced in 4.83 by another bugfix.
852 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
854 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
855 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
861 TF/01 Correctly close the server side of TLS when forking for delivery.
863 When a message was received over SMTP with TLS, Exim failed to clear up
864 the incoming connection properly after forking off the child process to
865 deliver the message. In some situations the subsequent outgoing
866 delivery connection happened to have the same fd number as the incoming
867 connection previously had. Exim would try to use TLS and fail, logging
868 a "Bad file descriptor" error.
870 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
871 utilities have not been installed.
873 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
874 temporary space as the ACL may create new global variables.
876 TL/01 LDAP support uses per connection or global context settings, depending
877 upon the detected version of the libraries at build time.
879 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
880 to extract and use the src ip:port in logging and expansions as if it
881 were a direct connection from the outside internet. PPv2 support was
882 updated based on HAProxy spec change in May 2014.
884 JH/02 Add ${listextract {number}{list}{success}{fail}}.
886 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
887 Properly escape header and check for NULL return.
889 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
892 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
894 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
895 characters in header names, implemented as a verify condition.
896 Contributed by Michael Fischer v. Mollard.
898 TL/05 Rename SPF condition results err_perm and err_temp to standardized
899 results permerror and temperror. Previous values are deprecated but
900 still accepted. In a future release, err_perm and err_temp will be
901 completely removed, which will be a backward incompatibility if the
902 ACL tests for either of these two old results. Patch contributed by
903 user bes-internal on the mailing list.
905 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
907 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
908 selectors, in both main and reject logs.
910 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
913 JH/07 Add malware type "sock" for talking to simple daemon.
915 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
917 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
918 routers/transports under cutthrough routing.
920 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
921 numbers. Touch up "bool" conditional to keep the same definition.
923 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
925 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
927 JH/12 Expand items in router/transport headers_add or headers_remove lists
928 individually rather than the list as a whole. Bug 1452.
930 Required for reasonable handling of multiple headers_ options when
931 they may be empty; requires that headers_remove items with embedded
932 colons must have them doubled (or the list-separator changed).
934 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
935 view the policy declared in the DMARC record. Currently, $dmarc_status
936 is a combined value of both the record presence and the result of the
939 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
941 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
942 dnslookup router and the smtp transport (applying to the forward
945 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
946 of ldap servers used for a specific lookup. Patch provided by Heiko
949 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
950 New variable $lookup_dnssec_authenticated for observability.
952 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
953 Patch submitted by Lars Timman.
955 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
957 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
958 Requires trusted mode and valid format message id, aborts otherwise.
959 Patch contributed by Heiko Schlichting.
961 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
962 certextract with support for various fields. Bug 1358.
964 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
965 is requested by default, modifiable by smtp transport option
968 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
969 operate on certificate variables to give certificate fingerprints
970 Also new ${sha256:cert_variable}.
972 JH/23 The PRDR feature is moved from being Experimental into the mainline.
974 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
975 Christian Aistleitner.
977 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
979 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
980 file. Patch from Wolfgang Breyha.
982 JH/25 Expand the coverage of the delivery $host and $host_address to
983 client authenticators run in verify callout. Bug 1476.
985 JH/26 Port service names are now accepted for tls_on_connect_ports, to
986 align with daemon_smtp_ports. Bug 72.
988 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
989 support and error reporting did not work properly.
991 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
992 and is readable. Patch from Andrew Colin Kissa.
994 TL/14 Enhance documentation of ${run expansion and how it parses the
995 commandline after expansion, particularly in the case when an
996 unquoted variable expansion results in an empty value.
998 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1000 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1001 by renaming away from it. Addresses GH issue 15, reported by
1004 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1006 TL/15 SECURITY: prevent double expansion in math comparison functions
1007 (can expand unsanitized data). Not remotely exploitable.
1014 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1016 PP/02 Make -n do something, by making it not do something.
1017 When combined with -bP, the name of an option is not output.
1019 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1022 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1023 $sender_host_name and config options to manage this, and basic check
1026 PP/05 DSCP support for outbound connections and control modifier for inbound.
1028 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1029 (Only plugin which currently uses this is kerberos4, which nobody should
1030 be using, but we should make it available and other future plugins might
1031 conceivably use it, even though it would break NAT; stuff *should* be
1032 using channel bindings instead).
1034 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1035 name; added for Sendmail compatibility; requires admin caller.
1036 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1037 just ignore it); requires trusted caller.
1038 Also parse but ignore: -Ac -Am -X<logfile>
1041 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1043 TL/02 Add +smtp_confirmation as a default logging option.
1045 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1046 Patch by Magnus Holmgren from 2007-02-20.
1048 TL/04 Bugzilla 1281 - Spec typo.
1049 Bugzilla 1283 - Spec typo.
1050 Bugzilla 1290 - Spec grammar fixes.
1052 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1054 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1056 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1057 mailing list by by Dmitry Isaikin.
1059 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1061 JH/02 Support "G" suffix to numbers in ${if comparisons.
1063 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1065 NM/01 Bugzilla 1197 - Spec typo
1066 Bugzilla 1196 - Spec examples corrections
1068 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1070 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1071 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1074 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1075 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1077 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1078 "acl {{name}{arg}...}", and optional args on acl condition
1081 JH/05 Permit multiple router/transport headers_add/remove lines.
1083 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1085 JH/07 Avoid using a waiting database for a single-message-only transport.
1086 Performance patch from Paul Fisher. Bugzilla 1262.
1088 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1091 JH/09 Add $headers_added variable, with content from use of ACL modifier
1092 add_header (but not yet added to the message). Bugzilla 199.
1094 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1095 Pulled from Bugzilla 817 by Wolfgang Breyha.
1097 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1099 (nb: this is the same fix as in Exim 4.80.1)
1101 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1104 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1107 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1109 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1110 advertise SMTP AUTH mechanism to us, instead of a generic
1111 protocol violation error. Also, make Exim more robust to bad
1112 data from the Dovecot auth socket.
1114 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1116 When a queue runner is handling a message, Exim first routes the
1117 recipient addresses, during which it prunes them based on the retry
1118 hints database. After that it attempts to deliver the message to
1119 any remaining recipients. It then updates the hints database using
1122 So if a recipient address works intermittently, it can get repeatedly
1123 deferred at routing time. The retry hints record remains fresh so the
1124 address never reaches the final cutoff time.
1126 This is a fairly common occurrence when a user is bumping up against
1127 their storage quota. Exim had some logic in its local delivery code
1128 to deal with this. However it did not apply to per-recipient defers
1129 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1131 This change adds a proper retry rule check during routing so that the
1132 final cutoff time is checked against the message's age. We only do
1133 this check if there is an address retry record and there is not a
1134 domain retry record; this implies that previous attempts to handle
1135 the address had the retry_use_local_parts option turned on. We use
1136 this as an approximation for the destination being like a local
1137 delivery, as in LMTP.
1139 I suspect this new check makes the old local delivery cutoff check
1140 redundant, but I have not verified this so I left the code in place.
1142 TF/02 Correct gecos expansion when From: is a prefix of the username.
1144 Test 0254 submits a message to Exim with the header
1148 When I ran the test suite under the user fanf2, Exim expanded
1149 the header to contain my full name, whereas it should have added
1150 a Resent-Sender: header. It erroneously treats any prefix of the
1151 username as equal to the username.
1153 This change corrects that bug.
1155 GF/01 DCC debug and logging tidyup
1156 Error conditions log to paniclog rather than rejectlog.
1157 Debug lines prefixed by "DCC: " to remove any ambiguity.
1159 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1161 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1162 Bug spotted by Jeremy Harris; was flawed since initial commit.
1163 Would have resulted in OCSP responses post-SNI triggering an Exim
1164 NULL dereference and crash.
1166 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1168 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1169 Bug detection, analysis and fix by Samuel Thibault.
1170 Bugzilla 1331, Debian bug #698092.
1172 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1174 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1175 Server implementation by Todd Lyons, client by JH.
1176 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1177 config variable "prdr_enable" controls whether the server
1178 advertises the facility. If the client requests PRDR a new
1179 acl_data_smtp_prdr ACL is called once for each recipient, after
1180 the body content is received and before the acl_smtp_data ACL.
1181 The client is controlled by both of: a hosts_try_prdr option
1182 on the smtp transport, and the server advertisement.
1183 Default client logging of deliveries and rejections involving
1184 PRDR are flagged with the string "PRDR".
1186 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1187 fclose(). Diagnosis by Todd Lyons.
1189 PP/17 Update configure.default to handle IPv6 localhost better.
1190 Patch by Alain Williams (plus minor tweaks).
1193 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1194 This is now consistent with GnuTLS, and is now documented: the
1195 previous undocumented portable approach to treating the option as
1196 unset was to force an expansion failure. That still works, and
1197 an empty string is now equivalent.
1199 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1200 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1201 not performing validation itself.
1203 PP/20 Added force_command boolean option to pipe transport.
1204 Patch from Nick Koston, of cPanel Inc.
1206 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1209 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1211 PP/21 Fix eximon continuous updating with timestamped log-files.
1212 Broken in a format-string cleanup in 4.80, missed when I repaired the
1213 other false fix of the same issue.
1214 Report and fix from Heiko Schlichting.
1217 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1218 Report from Prashanth Katuri.
1220 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1221 It's SecureTransport, so affects any MacOS clients which use the
1222 system-integrated TLS libraries, including email clients.
1224 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1225 using a MIME ACL for non-SMTP local injection.
1226 Report and assistance in diagnosis by Warren Baker.
1228 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1230 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1232 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1233 last id that failed so it may be referenced in subsequent ACL's.
1235 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1238 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1239 ldap library initialization, allowing self-signed CA's to be
1240 used. Also properly sets require_cert option later in code by
1241 using NULL (global ldap config) instead of ldap handle (per
1242 session). Bug diagnosis and testing by alxgomz.
1244 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1245 the src/util/ subdirectory.
1247 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1248 renamed to Transport Post Delivery Action by Jeremy Harris, as
1251 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1252 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1253 redis_servers = needs to be configured which will be used by the redis
1254 lookup. Patch from Warren Baker, of The Packet Hub.
1256 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1258 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1259 hostname or reverse DNS when processing a host list. Used suggestions
1260 from multiple comments on this bug.
1262 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1264 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1265 Missed a few lines, added it to make the runtest require no keyboard
1268 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1269 contains upper case chars. Make router use caseful_local_part.
1271 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1272 support when GnuTLS has been built with p11-kit.
1278 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1280 This, or similar/improved, will also be change PP/11 of 4.82.
1286 PP/01 Handle short writes when writing local log-files.
1287 In practice, only affects FreeBSD (8 onwards).
1288 Bugzilla 1053, with thanks to Dmitry Isaikin.
1290 NM/01 Bugzilla 949 - Documentation tweak
1292 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1295 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1297 PP/02 Implemented gsasl authenticator.
1299 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1301 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1302 `pkg-config foo` for cflags/libs.
1304 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1305 with rest of GSASL and with heimdal_gssapi.
1307 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1308 `pkg-config foo` for cflags/libs for the TLS implementation.
1310 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1311 properties get this fed in as external SSF. A number of robustness
1312 and debugging improvements to the cyrus_sasl authenticator.
1314 PP/08 cyrus_sasl server now expands the server_realm option.
1316 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1317 Patch by Jeremy Harris.
1319 PP/10 Added dbmjz lookup type.
1321 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1323 PP/12 MAIL args handles TAB as well as SP, for better interop with
1324 non-compliant senders.
1325 Analysis and variant patch by Todd Lyons.
1327 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1328 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1329 Patch from Dirk Mueller <dmueller@suse.com>
1331 PP/13 tls_peerdn now print-escaped for spool files.
1332 Observed some $tls_peerdn in wild which contained \n, which resulted
1333 in spool file corruption.
1335 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1336 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1337 or write after TLS renegotiation, which otherwise led to messages
1340 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1341 as a tracking header (ie: a signed header comes before the signature).
1342 Patch from Wolfgang Breyha.
1344 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1345 comma-sep list; embedded commas doubled.
1347 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1349 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1351 Report and patch from Dmitry Banschikov.
1353 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1354 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1355 failures. We appear to now support TLS1.1+ with Exim.
1357 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1358 lets Exim select keys and certificates based upon TLS SNI from client.
1359 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1360 before an outbound SMTP session. New log_selector, +tls_sni.
1362 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1363 NULL dereference. Report and patch from Alun Jones.
1365 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1366 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1367 Not seeing resolver debug output on NetBSD, but suspect this is a
1368 resolver implementation change.
1370 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1371 Left warnings. Added "eximon gdb" invocation mode.
1373 PP/21 Defaulting "accept_8bitmime" to true, not false.
1375 PP/22 Added -bw for inetd wait mode support.
1377 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1378 locate the relevant includes and libraries. Made this the default.
1380 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1381 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1383 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1384 This may cause build issues on older platforms.
1386 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1387 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1388 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1389 Added SNI support via GnuTLS too.
1390 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1392 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1394 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1395 Applied second patch from Janne, implementing suggestion to default
1396 multiple-strings-in-record handling to match SPF spec.
1398 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1400 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1401 read-only, out of scope).
1402 Patch from Wolfgang Breyha, report from Stuart Northfield.
1404 PP/29 Fix three issues highlighted by clang analyser static analysis.
1405 Only crash-plausible issue would require the Cambridge-specific
1406 iplookup router and a misconfiguration.
1407 Report from Marcin Mirosław.
1409 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1411 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1412 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1413 As part of this, removing so much warning spew let me fix some minor
1414 real issues in debug logging.
1416 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1417 assignment on my part. Fixed.
1419 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1420 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1421 Janne Snabb (who went above and beyond: thank you).
1423 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1424 string otherwise requires a connection and a bunch more work and it's
1425 relatively easy to get wrong. Should also expose TLS library linkage
1428 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1429 64-bit ${eval} (JH/03).
1431 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1432 GNU libc to support some of the 64-bit stuff, should not lead to
1433 conflicts. Defined before os.h is pulled in, so if a given platform
1434 needs to override this, it can.
1436 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1437 protection layer was required, which is not implemented.
1438 Bugzilla 1254, patch from Wolfgang Breyha.
1440 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1441 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1442 tls_dhparam take prime identifiers. Also unbreak combination of
1443 OpenSSL+DH_params+TLSSNI.
1445 PP/39 Disable SSLv2 by default in OpenSSL support.
1451 PP/01 Solaris build fix for Oracle's LDAP libraries.
1452 Bugzilla 1109, patch from Stephen Usher.
1454 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1456 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1459 TF/02 Fix a couple more cases where we did not log the error message
1460 when unlink() failed. See also change 4.74-TF/03.
1462 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1463 lock up or crash if it happened to be inside a call to libc when it
1464 got a SIGUSR1 from exiwhat.
1466 The SIGUSR1 handler appends the current process status to the process
1467 log which is later printed by exiwhat. It used to use the general
1468 purpose logging code to do this, but several functions it calls are
1469 not safe for signals.
1471 The new output code in the SIGUSR1 handler is specific to the process
1472 log, and simple enough that it's easy to inspect for signal safety.
1473 Removing some special cases also simplifies the general logging code.
1474 Removing the spurious timestamps from the process log simplifies
1477 TF/04 Improved ratelimit ACL condition.
1479 The /noupdate option has been deprecated in favour of /readonly which
1480 has clearer semantics. The /leaky, /strict, and /readonly update modes
1481 are mutually exclusive. The update mode is no longer included in the
1482 database key; it just determines when the database is updated. (This
1483 means that when you upgrade Exim will forget old rate measurements.)
1485 Exim now checks that the per_* options are used with an update mode that
1486 makes sense for the current ACL. For example, when Exim is processing a
1487 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1488 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1489 must specify per_mail/readonly. If you omit the update mode it defaults to
1490 /leaky where that makes sense (as before) or /readonly where required.
1492 The /noupdate option is now undocumented but still supported for
1493 backwards compatibility. It is equivalent to /readonly except that in
1494 ACLs where /readonly is required you may specify /leaky/noupdate or
1495 /strict/noupdate which are treated the same as /readonly.
1497 A useful new feature is the /count= option. This is a generalization
1498 of the per_byte option, so that you can measure the throughput of other
1499 aggregate values. For example, the per_byte option is now equivalent
1500 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1502 The per_rcpt option has been generalized using the /count= mechanism
1503 (though it's more complicated than the per_byte equivalence). When it is
1504 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1505 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1506 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1507 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1508 Note that using per_rcpt with a non-readonly update mode in more than
1509 one ACL will cause the recipients to be double-counted. (The per_mail
1510 and per_byte options don't have this problem.)
1512 The handling of very low rates has changed slightly. If the computed rate
1513 is less than the event's count (usually one) then this event is the first
1514 after a long gap. In this case the rate is set to the same as this event's
1515 count, so that the first message of a spam run is counted properly.
1517 The major new feature is a mechanism for counting the rate of unique
1518 events. The new per_addr option counts the number of different
1519 recipients that someone has sent messages to in the last time period. It
1520 behaves like per_rcpt if all the recipient addresses are different, but
1521 duplicate recipient addresses do not increase the measured rate. Like
1522 the /count= option this is a general mechanism, so the per_addr option
1523 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1524 example, measure the rate that a client uses different sender addresses
1525 with the options per_mail/unique=$sender_address. There are further
1526 details in the main documentation.
1528 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1530 TF/06 Removed a few PCRE remnants.
1532 TF/07 Automatically extract Exim's version number from tags in the git
1533 repository when doing development or release builds.
1535 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1536 Bugzilla 879. Patch from Paul Fisher.
1538 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1539 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1542 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1544 PP/05 Variable $av_failed, true if the AV scanner deferred.
1545 Bugzilla 1078. Patch from John Horne.
1547 PP/06 Stop make process more reliably on build failure.
1548 Bugzilla 1087. Patch from Heiko Schlittermann.
1550 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1551 Bugzilla 1089. Patch from Heiko Schlittermann.
1553 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1554 Bugzilla 1131. Patch from Holger Weiß.
1556 PP/09 Handle IPv6 addresses with SPF.
1557 Bugzilla 860. Patch from Wolfgang Breyha.
1559 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1561 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1564 PP/11 match_* no longer expand right-hand-side by default.
1565 New compile-time build option, EXPAND_LISTMATCH_RHS.
1566 New expansion conditions, "inlist", "inlisti".
1568 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1570 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1572 PP/14 fix log_write() format string regression from TF/03.
1573 Bugzilla 1152. Patch from Dmitry Isaikin.
1579 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1581 PP/02 Harmonised TLS library version reporting; only show if debugging.
1582 Layout now matches that introduced for other libraries in 4.74 PP/03.
1584 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1586 PP/04 New "dns_use_edns0" global option.
1588 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1591 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1592 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1594 TK/01 Updated PolarSSL code to 0.14.2.
1595 Bugzilla 1097. Patch from Andreas Metzler.
1597 PP/07 Catch divide-by-zero in ${eval:...}.
1598 Fixes bugzilla 1102.
1600 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1603 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1604 format-string attack -- SECURITY: remote arbitrary code execution.
1606 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1607 time unintentionally subject to list matching rules, letting the header
1608 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1609 arbitrary string expansion). This allowed for information disclosure.
1611 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1612 INT_MIN/-1 -- value coerced to INT_MAX.
1618 NM/01 Workaround for PCRE version dependency in version reporting
1621 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1622 This fixes portability to compilers other than gcc, notably
1623 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1625 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1626 makefiles for portability to HP-UX and POSIX correctness.
1628 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1629 Also via indented variable definition in the Makefile.
1630 (Debugging by Oliver Heesakkers).
1632 PP/02 Restore caching of spamd results with expanded spamd_address.
1633 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1635 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1636 Improves build reliability. Fix from: Frank Elsner
1638 NM/02 Fix wide character breakage in the rfc2047 coding
1639 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1641 NM/03 Allow underscore in dnslist lookups
1642 Fixes bug 1026. Patch from Graeme Fowler
1644 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1645 Code patches from Adam Ciarcinski of NetBSD.
1647 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1650 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1651 is logged, to avoid truncation. Patch from John Horne.
1653 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1654 Patch from Jakob Hirsch.
1656 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1657 SQL string expansion failure details.
1658 Patch from Andrey Oktyabrski.
1660 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1661 Patch from Simon Arlott.
1663 PP/09 New lookups functionality failed to compile on old gcc which rejects
1664 extern declarations in function scope.
1665 Patch from Oliver Fleischmann
1667 PP/10 Use sig_atomic_t for flags set from signal handlers.
1668 Check getgroups() return and improve debugging.
1669 Fixed developed for diagnosis in bug 927 (which turned out to be
1672 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1673 Patch from Mark Zealey.
1675 PP/12 Bugzilla 1056: Improved spamd server selection.
1676 Patch from Mark Zealey.
1678 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1679 Based on patch from Heiko Schlittermann.
1681 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1682 Patch from Uwe Doering, sign-off by Michael Haardt.
1684 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1685 variable declaration deep within a block. Bug and patch from
1688 PP/15 lookups-Makefile IRIX compatibility coercion.
1690 PP/16 Make DISABLE_DKIM build knob functional.
1692 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1693 Patch by Simon Arlott
1695 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1696 variable argument macros. Our copy now differs from upstream.
1702 TF/01 Failure to get a lock on a hints database can have serious
1703 consequences so log it to the panic log.
1705 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1706 controlled using the smtp_confirmation log selector.
1708 TF/03 Include the error message when we fail to unlink a spool file.
1710 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1711 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1712 for maintaining out-of-tree patches for some time.
1714 PP/01 Bugzilla 139: Documentation and portability issues.
1715 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1716 Handle per-OS dynamic-module compilation flags.
1718 PP/02 Let /dev/null have normal permissions.
1719 The 4.73 fixes were a little too stringent and complained about the
1720 permissions on /dev/null. Exempt it from some checks.
1721 Reported by Andreas M. Kirchwitz.
1723 PP/03 Report version information for many libraries, including
1724 Exim version information for dynamically loaded libraries. Created
1725 version.h, now support a version extension string for distributors
1726 who patch heavily. Dynamic module ABI change.
1728 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1729 privilege escalation vulnerability whereby the Exim run-time user
1730 can cause root to append content of the attacker's choosing to
1733 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1736 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1737 If dropping privileges for untrusted macros, we disabled normal logging
1738 on the basis that it would fail; for the Exim run-time user, this is not
1739 the case, and it resulted in successful deliveries going unlogged.
1740 Fixed. Reported by Andreas Metzler.
1746 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1747 only prepend for the Resent-* case. Fixes regression introduced in
1748 Exim 4.70 by NM/22 for Bugzilla 607.
1750 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1751 increasing numbers of administrators be bitten by this.
1753 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1755 PP/03 Bugzilla 994: added openssl_options main configuration option.
1757 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1759 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1761 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1763 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1764 without a peer certificate, leading to a segfault because of an
1765 assumption that peers always have certificates. Be a little more
1766 paranoid. Problem reported by Martin Tscholak.
1768 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1769 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1770 NB: ClamAV planning to remove STREAM in "middle of 2010".
1771 CL also introduces -bmalware, various -d+acl logging additions and
1772 more caution in buffer sizes.
1774 PP/09 Implemented reverse_ip expansion operator.
1776 PP/10 Bugzilla 937: provide a "debug" ACL control.
1778 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1780 PP/12 Bugzilla 973: Implement --version.
1782 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1784 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1786 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1788 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1789 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1790 ignore trailing whitespace.
1792 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1794 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1797 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1798 Notification from Dr Andrew Aitchison.
1800 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1801 ExtendedDetectionInfo response format.
1802 Notification from John Horne.
1804 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1807 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1808 XSL and documented dependency on system catalogs, with examples of how
1811 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1814 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1815 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1816 configuration file which is writeable by the Exim user or group.
1818 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1819 of configuration files to cover files specified with the -C option if
1820 they are going to be used with root privileges, not just the default
1823 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1824 option (effectively making it always true).
1826 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1827 files to be used while preserving root privileges.
1829 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1830 that rogue child processes cannot use them.
1832 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1833 run-time user, instead of root.
1835 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1836 Exim run-time user without dropping privileges.
1838 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1839 result string, instead of calling string_vformat() twice with the same
1842 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1843 for other users. Others should always drop root privileges if they use
1844 -C on the command line, even for a whitelisted configure file.
1846 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1848 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1854 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1855 $data_path, and $header_path variables; fixed documentation bugs and
1858 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1859 exipick to access non-standard spools, including the "frozen" queue
1862 NM/01 Bugzilla 965: Support mysql stored procedures.
1863 Patch from Alain Williams
1865 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1867 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1868 Patch from Andreas Metzler
1870 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1871 Patch from Kirill Miazine
1873 NM/05 Bugzilla 671: Added umask to procmail example.
1875 JJ/03 installed exipick 20100323.0, fixing doc bug
1877 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1878 directory. Notification and patch from Dan Rosenberg.
1880 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1882 TK/02 Improve log output when DKIM signing operation fails.
1884 MH/01 Treat the transport option dkim_domain as a colon separated
1885 list, not as a single string, and sign the message with each element,
1886 omitting multiple occurences of the same signer.
1888 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1889 Bugzilla 985, 986. Patch by Simon Arlott
1891 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1892 Patch by Simon Arlott
1894 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1895 MBX locking. Notification from Dan Rosenberg.
1901 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1903 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1905 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1907 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1909 NM/04 Bugzilla 915: Flag broken perl installation during build.
1915 TK/01 Added patch by Johannes Berg that expands the main option
1916 "spamd_address" if it starts with a dollar sign.
1918 TK/02 Write list of recipients to X-Envelope-Sender header when building
1919 the mbox-format spool file for content scanning (suggested by Jakob
1922 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1923 (http://www.dcc-servers.net/) support via dccifd. Activated by
1924 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1926 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1927 by Mark Daniel Reidel <mr@df.eu>.
1929 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1930 When building exim an external PCRE library is now needed -
1931 PCRE is a system library on the majority of modern systems.
1932 See entry on PCRE_LIBS in EDITME file.
1934 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1935 conversation. Added nologin parameter to request.
1936 Patch contributed by Kirill Miazine.
1938 TF/01 Do not log submission mode rewrites if they do not change the address.
1940 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1942 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1943 log files in place. Contributed by Roberto Lima.
1945 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1947 TF/03 Bugzilla 615: When checking the local_parts router precondition
1948 after a local_part_suffix or local_part_prefix option, Exim now
1949 does not use the address's named list lookup cache, since this
1950 contains cached lookups for the whole local part.
1952 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1953 Robert Millan. Documentation is in experimental-spec.txt.
1955 TF/04 Bugzilla 668: Fix parallel build (make -j).
1957 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1959 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1960 Patch provided by Jan Srzednicki.
1962 TF/05 Leading white space used to be stripped from $spam_report which
1963 wrecked the formatting. Now it is preserved.
1965 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1966 that they are available at delivery time.
1968 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1970 TF/08 TLS error reporting now respects the incoming_interface and
1971 incoming_port log selectors.
1973 TF/09 Produce a more useful error message if an SMTP transport's hosts
1974 setting expands to an empty string.
1976 NM/06 Bugzilla 744: EXPN did not work under TLS.
1977 Patch provided by Phil Pennock.
1979 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1980 Patch provided by Richard Godbee.
1982 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1983 acl_smtp_notquit, added index entry.
1985 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1986 Patch provided by Eugene Bujak.
1988 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1989 accept(). Patch provided by Maxim Dounin.
1991 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1992 Patch provided by Phil Pennock.
1994 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1996 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1997 Patch provided by Brad "anomie" Jorsch.
1999 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2000 Patch provided by Dean Brooks.
2002 TK/05 Add native DKIM support (does not depend on external libraries).
2004 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2005 Patch provided by Graeme Fowler.
2007 NM/16 Bugzilla 851: Documentation example syntax fix.
2009 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2011 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2014 NM/19 Bugzilla 745: TLS version reporting.
2015 Patch provided by Phil Pennock.
2017 NM/20 Bugzilla 167: bool: condition support.
2018 Patch provided by Phil Pennock.
2020 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2021 clients. Patch provided by Phil Pennock.
2023 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2024 Patch provided by Brad "anomie" Jorsch.
2026 NM/23 Bugzilla 687: Fix misparses in eximstats.
2027 Patch provided by Heiko Schlittermann.
2029 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2030 Patch provided by Heiko Schlittermann.
2032 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2033 plus update to original patch.
2035 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2037 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2038 Patch provided by David Brownlee.
2040 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2042 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2044 NM/30 Bugzilla 888: TLS documentation bugfixes.
2046 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2048 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2049 Unlike the original bugzilla I have changed all shell scripts in src tree.
2051 NM/33 Bugzilla 898: Transport filter timeout fix.
2052 Patch by Todd Rinaldo.
2054 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2055 Patch by Serge Demonchaux.
2057 NM/35 Bugzilla 39: Base64 decode bug fixes.
2058 Patch by Jakob Hirsch.
2060 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2062 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2064 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2066 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2072 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2073 ALT-N's libdkim that I have put here:
2074 http://duncanthrax.net/exim-experimental/
2076 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2077 were called 'true' and it seems that C99 defines that as a reserved
2078 keyword to be used with 'bool' variable types. That means you could
2079 not include C99-style headers which use bools without triggering
2080 build errors in sieve.c.
2082 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2083 as mailq or other aliases. Changed the --help handling significantly
2084 to do whats expected. exim_usage() emits usage/help information.
2086 SC/01 Added the -bylocaldomain option to eximstats.
2088 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2090 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2092 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2098 PH/01 Another patch from the Sieve maintainer.
2100 PH/02 When an IPv6 address is converted to a string for single-key lookup
2101 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2102 dots are used instead of colons so that keys in lsearch files need not
2103 contain colons. This was done some time before quoting was made available
2104 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2105 (notated using the quote facility) so as to distinguish them from IPv4
2106 keys. This meant that lookups for IP addresses in host lists did not work
2107 for iplsearch lookups.
2109 This has been fixed by arranging for IPv6 addresses to be expressed with
2110 colons if the lookup type is iplsearch. This is not incompatible, because
2111 previously such lookups could never work.
2113 The situation is now rather anomalous, since one *can* have colons in
2114 ordinary lsearch keys. However, making the change in all cases is
2115 incompatible and would probably break a number of configurations.
2117 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2120 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2121 conversion specification without a maximum field width, thereby enabling
2122 a rogue spamd server to cause a buffer overflow. While nobody in their
2123 right mind would setup Exim to query an untrusted spamd server, an
2124 attacker that gains access to a server running spamd could potentially
2125 exploit this vulnerability to run arbitrary code as the Exim user.
2127 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2128 $primary_hostname instead of what libspf2 thinks the hosts name is.
2130 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2131 a directory entry by the name of the lookup key. Previously, if a
2132 symlink pointed to a non-existing file or a file in a directory that
2133 Exim lacked permissions to read, a lookup for a key matching that
2134 symlink would fail. Now it is enough that a matching directory entry
2135 exists, symlink or not. (Bugzilla 503.)
2137 PH/03 The body_linecount and body_zerocount variables are now exported in the
2140 PH/04 Added the $dnslist_matched variable.
2142 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2143 This means they are set thereafter only if the connection becomes
2146 PH/06 Added the client_condition to authenticators so that some can be skipped
2147 by clients under certain conditions.
2149 PH/07 The error message for a badly-placed control=no_multiline_responses left
2150 "_responses" off the end of the name.
2152 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2154 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2155 (without spaces) instead of just copying the configuration text.
2157 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2159 PH/11 Added $max_received_linelength.
2161 PH/12 Added +ignore_defer and +include_defer to host lists.
2163 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2164 way in which PCRE > 7.0 is built.
2166 PH/14 Implemented queue_only_load_latch.
2168 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2169 MAIL command. The effect was to mangle the value on 64-bit systems.
2171 PH/16 Another patch from the Sieve maintainer.
2173 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2175 PH/18 If a system quota error occurred while trying to create the file for
2176 a maildir delivery, the message "Mailbox is full" was not appended to the
2177 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2178 only to a quota excession during the actual writing of the file.
2180 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2181 characters?) which causes problems in log lines. The DN values are now
2182 passed through string_printing() before being added to log lines.
2184 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2185 and InterBase are left for another time.)
2187 PH/21 Added message_body_newlines option.
2189 PH/22 Guard against possible overflow in moan_check_errorcopy().
2191 PH/23 POSIX allows open() to be a macro; guard against that.
2193 PH/24 If the recipient of an error message contained an @ in the local part
2194 (suitably quoted, of course), incorrect values were put in $domain and
2195 $local_part during the evaluation of errors_copy.
2201 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2202 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2205 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2206 issue a MAIL command.
2208 PH/02 In an ACL statement such as
2210 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2212 if a client was not listed at all, or was listed with a value other than
2213 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2214 the condition was not true (as it should be), so access was not denied.
2215 The bug was that the ! inversion was incorrectly passed on to the second
2216 item. This has been fixed.
2218 PH/03 Added additional dnslists conditions == and =& which are different from
2219 = and & when the dns lookup returns more than one IP address.
2221 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2222 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2224 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2225 FSYNC, which compiles an option called disable_fsync that allows for
2226 bypassing fsync(). The documentation is heavily laced with warnings.
2228 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2230 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2231 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2232 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2233 including adding "make clean"; (3) Added -fPIC when compiling the test
2234 dynamically loaded module, to get rid of a warning.
2236 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2237 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2238 = 0s. The bug is otherwise harmless.
2240 PH/07 There was a bug in the dovecot authenticator such that the value of
2241 $auth1 could be overwritten, and so not correctly preserved, after a
2242 successful authentication. This usually meant that the value preserved by
2243 the server_setid option was incorrect.
2245 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2247 PH/09 Installed PCRE release 7.0.
2249 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2250 run for batched SMTP input. It is now run at the start of every message
2251 in the batch. While fixing this I discovered that the process information
2252 (output by running exiwhat) was not always getting set for -bs and -bS
2253 input. This is fixed, and it now also says "batched" for BSMTP.
2255 PH/11 Added control=no_pipelining.
2257 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2258 patch, slightly modified), and move the expansion of helo_data till after
2259 the connection is made in the smtp transport (so it can use these
2262 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2264 PH/14 Added log_selector = +pid.
2266 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2268 PH/16 Add ${if forany and ${if forall.
2270 PH/17 Added dsn_from option to vary the From: line in DSNs.
2272 PH/18 Flush SMTP output before performing a callout, unless control =
2273 no_callout_flush is set.
2275 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2276 was true (the default) a successful delivery failed to delete the retry
2277 item, thus causing premature timeout of the address. The bug is now
2280 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2282 PH/21 Long custom messages for fakedefer and fakereject are now split up
2283 into multiline responses in the same way that messages for "deny" and
2284 other ACL rejections are.
2286 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2287 with slight modification.
2289 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2290 draft, changing the syntax and factoring some duplicate code".
2292 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2293 for deliveries of the second and subsequent messages over the same SMTP
2296 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2297 ${reduce, with only minor "tidies".
2299 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2301 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2302 expansion side effects.
2304 PH/27 When a message times out after an over-quota error from an Exim-imposed
2305 quota, the bounce message says "mailbox is full". This message was not
2306 being given when it was a system quota that was exceeded. It now should
2309 MH/03 Made $recipients available in local_scan(). local_scan() already has
2310 better access to the recipient list through recipients_list[], but
2311 $recipients can be useful in postmaster-provided expansion strings.
2313 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2314 in the case of a MAIL command with additional options following the
2315 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2316 were accidentally chopped off.
2318 PH/29 SMTP synchronization checks are implemented when a command is read -
2319 there is a check that no more input is waiting when there shouldn't be
2320 any. However, for some commands, a delay in an ACL can mean that it is
2321 some time before the response is written. In this time, more input might
2322 arrive, invalidly. So now there are extra checks after an ACL has run for
2323 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2324 pipelining has not been advertised.
2326 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2328 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2329 returned in the SMTP response when smtp_return_error_details was false.
2330 This has been fixed.
2332 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2333 instead of the C I/O that was originally supplied, because problems were
2334 reported on Solaris.
2336 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2337 Exim which did not show up earlier: it was assuming that a call to
2338 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2339 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2340 was a macro that became an assignment, so it seemed to work. This has
2341 changed to a proper function call with a void return, hence the compile
2342 error. Exim's code has been fixed.
2344 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2347 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2349 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2351 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2352 because it turns out to be too big or there is a timeout) there is no
2353 "Completed" line in the log. When some messages of this type were
2354 selected by exigrep, they were listed as "not completed". Others were
2355 picked up by some special patterns. I have improved the selection
2356 criteria to be more general.
2358 PH/38 The host_find_failed option in the manualroute router can now be set
2359 to "ignore", to completely ignore a host whose IP address cannot be
2360 found. If all hosts are ignored, the behaviour is controlled by the new
2361 host_all_ignored option.
2363 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2364 homing or because of multiple MX records with /mx) generated more than
2365 one IP address, and the following item turned out to be the local host,
2366 all the secondary addresses of the first item were incorrectly removed
2367 from the list, along with the local host and any following hosts (which
2368 is what is supposed to happen).
2370 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2371 whoever called Exim into the -H file. In the case of the daemon it was
2372 behaving confusingly. When first started, it used values for whoever
2373 started the daemon, but after a SIGHUP it used the Exim user (because it
2374 calls itself on a restart). I have changed the code so that it now always
2377 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2378 message are rejected with the same error (e.g. no authentication or bad
2379 sender address), and a DATA command is nevertheless sent (as can happen
2380 with PIPELINING or a stupid MUA), the error message that was given to the
2381 RCPT commands is included in the rejection of the DATA command. This is
2382 intended to be helpful for MUAs that show only the final error to their
2385 PH/42 Another patch from the Sieve maintainer.
2387 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2388 Eximstats - Fixed some broken HTML links and added missing column headers
2390 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2391 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2393 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2395 PH/43 Yet another patch from the Sieve maintainer.
2397 PH/44 I found a way to check for a TCP/IP connection going away before sending
2398 the response to the final '.' that terminates a message, but only in the
2399 case where the client has not sent further data following the '.'
2400 (unfortunately, this is allowed). However, in many cases there won't be
2401 any further data because there won't be any more messages to send. A call
2402 to select() can be used: if it shows that the input is "ready", there is
2403 either input waiting, or the socket has been closed. An attempt to read
2404 the next input character can distinguish the two cases. Previously, Exim
2405 would have sent an OK response which the client would never have see.
2406 This could lead to message repetition. This fix should cure that, at
2407 least in a lot of common cases.
2409 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2410 advertised in response to EHLO.
2416 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2417 fixed by 4.65/MH/01 (is this a record?) are fixed:
2419 (i) An empty string was always treated as zero by the numeric comparison
2420 operators. This behaviour has been restored.
2422 (ii) It is documented that the numeric comparison operators always treat
2423 their arguments as decimal numbers. This was broken in that numbers
2424 starting with 0 were being interpreted as octal.
2426 While fixing these problems I realized that there was another issue that
2427 hadn't been noticed. Values of message_size_limit (both the global option
2428 and the transport option) were treated as octal if they started with 0.
2429 The documentation was vague. These values are now always treated as
2430 decimal, and I will make that clear in the documentation.
2436 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2437 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2440 MH/01 Don't check that the operands of numeric comparison operators are
2441 integers when their expansion is in "skipping" mode (fixes bug
2442 introduced by 4.64-PH/07).
2444 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2445 child addresses, Exim now panics and dies. Previously, because the count
2446 is held in a short int, deliveries were likely to be lost. As such a
2447 large number of recipients for a single message is ridiculous
2448 (performance will be very, very poor), I have chosen to impose a limit
2449 rather than extend the field.
2455 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2456 leftover -K file (the existence of which was triggered by #402).
2457 While we were at it, introduced process PID as part of the -K
2458 filename. This should rule out race conditions when creating
2461 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2462 processing considerably. Previous code took too long for large mails,
2463 triggering a timeout which in turn triggers #401.
2465 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2466 in the DK code in transports.c. sendfile() is not really portable,
2467 hence the _LINUX specificness.
2469 TF/01 In the add_headers option to the mail command in an Exim filter,
2470 there was a bug that Exim would claim a syntax error in any
2471 header after the first one which had an odd number of characters
2474 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2475 callout verification, Exim cached a "reject" for the entire domain. This
2476 is correct for most verifications, but it is not correct for a recipient
2477 verification with use_sender or use_postmaster set, because in that case
2478 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2479 case of MAIL FROM:<> rejection from other early rejections (e.g.
2480 rejection of HELO). When verifying a recipient using a non-null MAIL
2481 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2482 Whatever the result of the callout, the value of the domain cache is
2483 left unchanged (for any other kind of callout, getting as far as trying
2484 RCPT means that the domain itself is ok).
2486 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2489 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2490 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2491 session, the daemon ignores ECONNECT errors and logs others; it now
2492 ignores EPIPE as well.
2494 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2495 (quoted-printable decoding).
2497 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2498 later the small subsequent patch to fix an introduced bug.
2500 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2502 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2504 PH/08 An error is now given if message_size_limit is specified negative.
2506 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2507 to be given (somewhat) arbitrary names.
2509 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2512 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2513 miscellaneous code fixes
2515 PH/10 Added the log_reject_target ACL modifier to specify where to log
2518 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2519 hostname. This is wrong, because it relates to the incoming message (and
2520 probably the interface on which it is arriving) and not to the outgoing
2521 callout (which could be using a different interface). This has been
2522 changed to use the value of the helo_data option from the smtp transport
2523 instead - this is what is used when a message is actually being sent. If
2524 there is no remote transport (possible with a router that sets up host
2525 addresses), $smtp_active_hostname is used.
2527 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2528 tweaks were necessary in order to get it to work (see also 21 below):
2529 (a) The code assumed that strncpy() returns a negative number on buffer
2530 overflow, which isn't the case. Replaced with Exim's string_format()
2532 (b) There were several signed/unsigned issues. I just did the minimum
2533 hacking in of casts. There is scope for a larger refactoring.
2534 (c) The code used strcasecmp() which is not a standard C function.
2535 Replaced with Exim's strcmpic() function.
2536 (d) The code set only $1; it now sets $auth1 as well.
2537 (e) A simple test gave the error "authentication client didn't specify
2538 service in request". It would seem that Dovecot has changed its
2539 interface. Fortunately there's a specification; I followed it and
2540 changed what the client sends and it appears to be working now.
2542 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2545 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2546 address A is aliased to B and C, where B exists and C does not. Without
2547 -v the output is "A verified" because verification stops after a
2548 successful redirection if more than one address is generated. However,
2549 with -v the child addresses are also verified. Exim was outputting "A
2550 failed to verify" and then showing the successful verification for C,
2551 with its parentage. It now outputs "B failed to verify", showing B's
2552 parentage before showing the successful verification of C.
2554 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2555 look up a TXT record in a specific list after matching in a combined
2558 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2559 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2560 they consult the DNS. I had assumed they would set it the way they
2561 wanted; and indeed my experiments on Linux seem to show that in some
2562 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2563 To be on the safe side, however, I have now made the interface to
2564 host_find_byname() similar to host_find_bydns(), with an argument
2565 containing the DNS resolver options. The host_find_byname() function now
2566 sets these options at its start, just as host_find_bydns() does. The smtp
2567 transport options dns_qualify_single and dns_search_parents are passed to
2568 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2569 of host_find_byname() use the default settings of RES_DEFNAMES
2570 (qualify_single) but not RES_DNSRCH (search_parents).
2572 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2573 spool_read_header() do less string testing, by means of a preliminary
2574 switch on the second character of optional "-foo" lines. (This is
2575 overdue, caused by the large number of possibilities that now exist.
2576 Originally there were few.) While I was there, I also converted the
2577 str(n)cmp tests so they don't re-test the leading "-" and the first
2578 character, in the hope this might squeeze out yet more improvement.
2580 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2581 flag allowing group syntax was set by the header_syntax check but not
2582 turned off, possible causing trouble later; (2) The flag was not being
2583 set at all for the header_verify test, causing "group"-style headers to
2584 be rejected. I have now set it in this case, and also caused header_
2585 verify to ignore an empty address taken from a group. While doing this, I
2586 came across some other cases where the code for allowing group syntax
2587 while scanning a header line wasn't quite right (mostly, not resetting
2588 the flag correctly in the right place). These bugs could have caused
2589 trouble for malformed header lines. I hope it is now all correct.
2591 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2592 with the "reply" argument non-NULL. The code, however (which originally
2593 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2594 but it didn't always do it. This confused somebody who was copying the
2595 code for some other use. I have removed all the tests.
2597 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2598 feature that was used to support insecure browsers during the U.S. crypto
2599 embargo. It requires special client support, and Exim is probably the
2600 only MTA that supported it -- and would never use it because real RSA is
2601 always available. This code has been removed, because it had the bad
2602 effect of slowing Exim down by computing (never used) parameters for the
2603 RSA_EXPORT functionality.
2605 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2606 authenticator to fail if there's a tab character in the incoming data
2607 (there should never be unless someone is messing about, as it's supposed
2608 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2609 if the connection is using TLS or if the remote IP is the same as the
2610 local IP, and the "valid-client-cert option" if a client certificate has
2613 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2614 authenticators. This can be used for authorization after authentication
2615 succeeds. (In the case of plaintext, it servers for both authentication
2618 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2619 if any retry times were supplied.
2621 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2622 connection, where there is no client IP address to check. In this
2623 situation, the verify now always succeeds.
2625 PH/25 Applied John Jetmore's -Mset patch.
2627 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2629 PH/27 In a string expansion for a processed (not raw) header when multiple
2630 headers of the same name were present, leading whitespace was being
2631 removed from all of them, but trailing whitespace was being removed only
2632 from the last one. Now trailing whitespace is removed from each header
2633 before concatenation. Completely empty headers in a concatenation (as
2634 before) are ignored.
2636 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2637 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2639 PH/29 [Removed. This was a change that I later backed out, and forgot to
2640 correct the ChangeLog entry (that I had efficiently created) before
2641 committing the later change.]
2643 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2644 address errors (4xx response to RCPT) over the same connection as other
2645 messages routed to the same hosts. Such deliveries are always "forced",
2646 so retry times are not inspected. This resulted in far too many retries
2647 for the affected addresses. The effect occurred only when there were more
2648 hosts than the hosts_max_try setting in the smtp transport when it had
2649 the 4xx errors. Those hosts that it had tried were not added to the list
2650 of hosts for which the message was waiting, so if all were tried, there
2651 was no problem. Two fixes have been applied:
2653 (i) If there are any address or message errors in an SMTP delivery, none
2654 of the hosts (tried or untried) are now added to the list of hosts
2655 for which the message is waiting, so the message should not be a
2656 candidate for sending over the same connection that was used for a
2657 successful delivery of some other message. This seems entirely
2658 reasonable: after all the message is NOT "waiting for some host".
2659 This is so "obvious" that I'm not sure why it wasn't done
2660 previously. Hope I haven't missed anything, but it can't do any
2661 harm, as the worst effect is to miss an optimization.
2663 (ii) If, despite (i), such a delivery is accidentally attempted, the
2664 routing retry time is respected, so at least it doesn't keep
2665 hammering the server.
2667 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2668 in ${readsocket because some servers need this prod.
2670 PH/32 Added some extra debug output when updating a wait-xxx database.
2672 PH/33 The hint "could be header name not terminated by colon", which has been
2673 given for certain expansion errors for a long time, was not being given
2674 for the ${if def:h_colon_omitted{... case.
2676 PH/34 The spec says: "With one important exception, whenever a domain list is
2677 being scanned, $domain contains the subject domain." There was at least
2678 one case where this was not true.
2680 PH/35 The error "getsockname() failed: connection reset by peer" was being
2681 written to the panic log as well as the main log, but it isn't really
2682 panic-worthy as it just means the connection died rather early on. I have
2683 removed the panic log writing for the ECONNRESET error when getsockname()
2686 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2687 runs only) independently of the message's sender address. This meant
2688 that, if the 4xx error was in fact related to the sender, a different
2689 message to the same recipient with a different sender could confuse
2690 things. In particular, this can happen when sending to a greylisting
2691 server, but other circumstances could also provoke similar problems.
2692 I have changed the default so that the retry time for these errors is now
2693 based a combination of the sender and recipient addresses. This change
2694 can be overridden by setting address_retry_include_sender=false in the
2697 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2698 remote server are returned as part of bounce messages. This was not
2699 happening for LMTP over a pipe (the lmtp transport), but now it is the
2700 same for both kinds of LMTP.
2702 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2703 in header lines that were in fact CNAMEs. This is no longer the case.
2705 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2706 and queue runs started by the daemon processed all messages. This has
2707 been fixed so that -R and -S can now usefully be given with -q<time>.
2709 PH/40 Import PCRE release 6.7 (fixes some bugs).
2711 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2713 PH/42 Give an error if -q is specified more than once.
2715 PH/43 Renamed the variables $interface_address and $interface_port as
2716 $received_ip_address and $received_port, to make it clear that these
2717 values apply to message reception, and not to the outgoing interface when
2718 a message is delivered. (The old names remain recognized, of course.)
2720 PH/44 There was no timeout on the connect() call when using a Unix domain
2721 socket in the ${readsocket expansion. There now is.
2723 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2724 be meaningful with "accept".
2726 SC/01 Eximstats V1.43
2727 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2729 SC/02 Eximstats V1.44
2730 Use a glob alias rather than an array ref in the generated
2731 parser. This improves both readability and performance.
2733 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2734 Collect SpamAssassin and rejection statistics.
2735 Don't display local sender or destination tables unless
2736 there is data to show.
2737 Added average volumes into the top table text output.
2739 SC/04 Eximstats V1.46
2740 Collect data on the number of addresses (recipients)
2741 as well as the number of messages.
2743 SC/05 Eximstats V1.47
2744 Added 'Message too big' to the list of mail rejection
2745 reasons (thanks to Marco Gaiarin).
2747 SC/06 Eximstats V1.48
2748 Mainlog lines which have GMT offsets and are too short to
2749 have a flag are now skipped.
2751 SC/07 Eximstats V1.49 (Alain Williams)
2752 Added the -emptyok flag.
2754 SC/08 Eximstats V1.50
2755 Fixes for obtaining the IP address from reject messages.
2757 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2758 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2759 whitespace changes from 4.64-PH/27
2761 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2764 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2765 are found, allow negative numbers in numeric criteria)
2767 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2769 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2770 to match changes made in 4.64-PH/43
2772 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2774 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2775 "domain" is actually an IP address, and force a failure. This locks out
2776 those revolvers/nameservers that support "A-for-A" lookups, in
2777 contravention of the specifications.
2779 PH/48 When a host name was looked up from an IP address, and the subsequent
2780 forward lookup of the name timed out, the host name was left in
2781 $sender_host_name, contrary to the specification.
2783 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2784 restricted to single-key lookups, Exim was not diagnosing an error if
2785 * or *@ was used with a query-style lookup.
2787 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2789 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2790 long ago, but noone interested enough thought of it. Let's just say that
2791 the "1.1" means that there are some new functions that weren't there at
2792 some point in the past.
2794 PH/51 Error processing for expansion failure of helo_data from an smtp
2795 transport during callout processing was broken.
2797 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2798 tested/used via the -bh/-bhc/-bs options.
2800 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2801 bug, fixed in subsequent PCRE releases).
2803 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2804 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2806 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2812 SC/01 Use a glob alias rather than an array ref in eximstats generated
2813 parser. This improves both readability and performance.
2815 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2816 Don't display local sender or destination tables in eximstats unless
2817 there is data to show.
2818 Added average volumes into the eximstats top table text output.
2820 SC/03 Collect data on the number of addresses (recipients) as well
2821 as the number of messages in eximstats.
2823 TF/01 Correct an error in the documentation for the redirect router. Exim
2824 does (usually) call initgroups() when daemonizing.
2826 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2827 with consistent privilege compared to when running as a daemon.
2829 TF/03 Note in the spec that $authenticated_id is not set for local
2830 submissions from trusted users.
2832 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2833 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2835 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2836 by adding some example configuration directives to the default
2837 configuration file. A little bit of work is required to uncomment the
2838 directives and define how usernames and passwords are checked, but
2839 there is now a framework to start from.
2841 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2842 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2843 without this. I don't know how relevant this is to other LDAP libraries.
2845 PH/02 Add the verb name to the "unknown ACL verb" error.
2847 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2849 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2851 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2852 directory not expanded when it should be if an expanded home directory
2853 was set for the address (which is overridden by the transport).
2855 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2858 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2859 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2860 because it is too late at that time, and has no effect.
2862 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2863 security issue with \' (bugzilla #107). I could not use the
2864 PQescapeStringConn() function, because it needs a PGconn value as one of
2867 PH/08 When testing addresses using -bt, indicate those final addresses that
2868 are duplicates that would not cause an additional delivery. At least one
2869 person was confused, thinking that -bt output corresponded to deliveries.
2870 (Suppressing duplicates isn't a good idea as you lose the information
2871 about possibly different redirections that led to the duplicates.)
2873 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2874 systems where poll() doesn't work, in particular OS X.
2876 PH/10 Added more information to debugging output for retry time not reached.
2878 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2879 operations in malware.c.
2881 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2884 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2885 syslog_duplication set false, log lines that would normally be written
2886 both the the main log and to the reject log were not written to syslog at
2889 PH/14 In the default configuration, change the use of "message" in ACL warn
2890 statements to "add_header".
2892 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2893 not followed by a command (e.g. "seen endif").
2895 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2896 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2899 PH/17 Added extra conditions to the default value of delay_warning_condition
2903 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2904 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2905 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2908 The Auto-Submitted: and various List- headers are standardised, whereas I
2909 don't think Precedence: ever was.
2911 PH/18 Refactored debugging code in route_finduser() to show more information,
2912 in particular, the error code if getpwnam() issues one.
2914 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2915 This is apparently needed in addition to the PH/07 change above to avoid
2916 any possible encoding problems.
2918 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2919 but not after initializing Perl.
2921 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2922 output them only if debugging. By default they are written stderr,
2923 apparently, which is not desirable.
2925 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2928 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2931 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2933 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2934 authenticated or an ident call has been made. Suppress the default
2935 values for $authenticated_id and $authenticated_sender (but permit -oMai
2936 and -oMas) when testing with -bh.
2938 PH/24 Re-jigged the order of the tests in the default configuration so that the
2939 tests for valid domains and recipients precede the DNS black list and CSA
2940 tests, on the grounds that those ones are more expensive.
2942 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2943 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2944 This bug exists in every version of Exim that I still have, right back to
2947 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2948 However, an attempt to turn on case-sensitivity in a regex key by
2949 including (?-i) didn't work because the subject string was already
2950 lowercased, and the effects were non-intuitive. It turns out that a
2951 one-line patch can be used to allow (?-i) to work as expected.
2957 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2958 other effects) broke the use of negated acl sub-conditions.
2960 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2963 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2964 "Deny" causes Exim to reject the incoming connection with a 554 error.
2965 Unfortunately, if there is a major crisis, such as a disk failure,
2966 tcp-wrappers gives "deny", whereas what one would like would be some
2967 kind of temporary error. A kludge has been added to help with this.
2968 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2969 554 error is used if errno is still zero or contains ENOENT (which occurs
2970 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2973 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2975 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2976 errors. Otherwise a message that provokes a temporary error (when other
2977 messages do not) can cause a whole host to time out.
2979 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2980 addresses were routed directly to files or pipes from a redirect router.
2981 File deliveries just didn't batch; pipe deliveries might have suffered
2984 PH/06 A failure to get a lock for a hints database would erroneously always say
2985 "Failed to get write lock", even when it was really a read lock.
2987 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2988 of 0600. This has been changed to use the value of the lockfile_mode
2989 option (which defaults to 0600).
2991 PH/08 Applied small patch from the Sieve maintainer.
2993 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2994 folder from quota calculations, a direct delivery into this folder messed
2995 up the contents of the maildirsize file. This was because the regex was
2996 used only to exclude .Trash (or whatever) when the size of the mailbox
2997 was calculated. There was no check that a delivery was happening into an
2998 excluded directory. This bug has been fixed by ignoring all quota
2999 processing for deliveries into excluded directories.
3001 PH/10 Added the maildirfolder_create_regex option to appendfile.
3007 PH/01 The code for finding all the local interface addresses on a FreeBSD
3008 system running IPv6 was broken. This may well have applied to all BSD
3009 systems, as well as to others that have similar system calls. The broken
3010 code found IPv4 interfaces correctly, but gave incorrect values for the
3011 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3012 that it would not match correctly against @[] and not recognize the IPv6
3015 PH/02 The ipliteral router was not recognizing addresses of the form user@
3016 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3018 PH/03 Added disable_ipv6.
3020 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3021 decoded form, because it is most often used to construct To: headers
3022 lines in autoreplies, and the decoded form may well be syntactically
3023 invalid. However, $reply_address has leading white space removed, and all
3024 newlines turned into spaces so that the autoreply transport does not
3027 PH/05 If group was specified without a user on a router, and no group or user
3028 was specified on a transport, the group from the router was ignored.
3030 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3031 for visible compile-time settings that can be used to change these
3032 numbers, for those that want even more. Backwards compatibility with old
3033 spool files has been maintained. However, going back to a previous Exim
3034 release will lost any variables that are in spool files.
3036 PH/07 Two small changes when running in the test harness: increase delay when
3037 passing a TCP/IP connection to a new process, in case the original
3038 process has to generate a bounce, and remove special handling of
3039 127.0.0.2 (sic), which is no longer necessary.
3041 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3042 be the same on different OS.
3044 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3047 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3048 whether --show-vars was specified or not
3050 JJ/02 exipick: Added support for new ACL variable spool format introduced
3053 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3054 syntactically invalid From: or Reply-to: line, and a filter used this to
3055 generate an autoreply, and therefore failed to obtain an address for the
3056 autoreply, Exim could try to deliver to a non-existent relative file
3057 name, causing unrelated and misleading errors. What now happens is that
3058 it logs this as a hard delivery error, but does not attempt to create a
3061 PH/11 The exinext utility has a -C option for testing purposes, but although
3062 the given file was scanned by exinext itself; it wasn't being passed on
3063 when Exim was called.
3065 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3066 an end-of-file indication when reading a command response.
3068 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3069 compiled. In many other places in Exim, IPv6 addresses are always
3070 recognized, so I have changed this. It also means that IPv4 domain
3071 literals of the form [IPV4:n.n.n.n] are now always recognized.
3073 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3074 used if the router is not running as root, for example, when verifying at
3075 ACL time, or when using -bh. The debugging output from this situation was
3076 non-existent - all you got was a failure to exec. I have made two
3079 (a) Failures to set uid/gid, the current directory, or a process leader
3080 in a subprocess such as that created by queryprogram now generate
3081 suitable debugging output when -d is set.
3083 (b) The queryprogram router detects when it is not running as root,
3084 outputs suitable debugging information if -d is set, and then runs
3085 the subprocess without attempting to change uid/gid.
3087 PH/15 Minor change to Makefile for building test_host (undocumented testing
3090 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3091 additional section of a DNS packet that returns MX or SRV records.
3092 Instead, it always explicitly searches for A/AAAA records. This avoids
3093 major problems that occur when a DNS server includes only records of one
3094 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3095 fixed another bug: if SRV records were looked up and the corresponding
3096 address records were *not* found in the additional section, the port
3097 values from the SRV records were lost.
3099 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3100 using the correct key (the original address) when searching the retry
3101 rules in order to find which one to use for generating the retry hint.
3103 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3104 adding the default one. Similarly, if it contains a Reply-To: header, the
3105 errors_reply_to option, if set, is not used.
3107 PH/19 When calculating a retry time, Exim used to measure the "time since
3108 failure" by looking at the "first failed" field in the retry record. Now
3109 it does not use this if it is later than than the arrival time of the
3110 message. Instead it uses the arrival time. This makes for better
3111 behaviour in cases where some deliveries succeed, thus re-setting the
3112 "first failed" field. An example is a quota failure for a huge message
3113 when small messages continue to be delivered. Without this change, the
3114 "time since failure" will always be short, possible causing more frequent
3115 delivery attempts for the huge message than are intended.
3116 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3118 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3119 $1, $2, $3) because the numerical variables can be reset during some
3120 expansion items (e.g. "match"), thereby losing the authentication data.
3122 PH/21 Make -bV show the size of off_t variables so that the test suite can
3123 decide whether to run tests for quotas > 2G.
3125 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3126 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3127 filecount value is greater than 2G or if a quota value is greater than 2G
3128 on a system where the size of off_t is not greater than 4, a panic error
3131 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3132 never match. The debug and -bh output now contains an explicit error
3133 message indicating a malformed IPv4 address or mask.
3135 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3136 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3137 PH/23 above applies.
3139 PH/25 Do not write to syslog when running in the test harness. The only
3140 occasion when this arises is a failure to open the main or panic logs
3141 (for which there is an explicit test).
3143 PH/26 Added the /no_tell option to "control=freeze".
3145 PH/27 If a host name lookup failed very early in a connection, for example, if
3146 the IP address matched host_lookup and the reverse lookup yielded a name
3147 that did not have a forward lookup, an error message of the form "no IP
3148 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3149 could be logged. Now it outputs the IP address instead of "NULL".
3151 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3152 allows the sender and the authenticated sender to be set when
3153 submitting a message from within Exim. Since child_open_exim() is
3154 documented for local_scan(), the new function should be too.
3156 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3157 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3158 results in an empty string is now treated as unset.
3160 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3162 PH/31 Added sender_verify_fail logging option.
3164 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3165 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3166 tidied the source and removed it altogether.
3168 PH/33 When a queue run was abandoned because the load average was too high, a
3169 log line was always written; now it is written only if the queue_run log
3170 selector is set. In addition, the log line for abandonment now contains
3171 information about the queue run such as the pid. This is always present
3172 in "start" and "stop" lines but was omitted from the "abandon" line.
3174 PH/34 Omit spaces between a header name and the colon in the error message that
3175 is given when verify = headers_syntax fails (if there are lots of them,
3176 the message gets confusing).
3178 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3179 names, as there are now some PTR records that contain slashes. This check
3180 is only to protect against broken name servers that fall over on strange
3181 characters, so the fact that it applies to all lookups doesn't matter.
3183 PH/36 Now that the new test suite is complete, we can remove some of the
3184 special code in Exim that was needed for the old test suite. For example,
3185 sorting DNS records because real resolvers return them in an arbitrary
3186 order. The new test suite's fake resolver always returns records in the
3189 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3190 bounces) except when queue_only is set, to avoid logging races between
3191 the different processes.
3193 PH/38 Panic-die if .include specifies a non-absolute path.
3195 PH/39 A tweak to the "H" retry rule from its user.
3197 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3198 a label. They prevented compilation on older perls.
3200 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3201 a warning to be raised on newish perls.
3203 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3204 on queue. Changes to match documented behaviour of showing count of
3205 messages matching specified criteria.
3207 PH/40 Changed the default ident timeout from 30s to 5s.
3209 PH/41 Added support for the use of login_cap features, on those BSD systems
3210 that have them, for controlling the resources used by pipe deliveries.
3212 PH/42 The content-scanning code uses fopen() to create files in which to put
3213 message data. Previously it was not paying any attention to the mode of
3214 the files. Exim runs with umask(0) because the rest of the code creates
3215 files with open(), and sets the required mode explicitly. Thus, these
3216 files were ending up world-writeable. This was not a big issue, because,
3217 being within the spool directory, they were not world-accessible. I have
3218 created a function called modefopen, which takes an additional mode
3219 argument. It sets umask(777), creates the file, chmods it to the required
3220 mode, then resets the umask. All the relevant calls to fopen() in the
3221 content scanning code have been changed to use this function.
3223 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3224 to 24 hours. This avoids potential overflow problems when processing G
3225 and H retry rules. I suspect nobody ever tinkers with this value.
3227 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3229 PH/45 When the plaintext authenticator is running as a client, the server's
3230 challenges are checked to ensure they are valid base64 strings. By
3231 default, the authentication attempt is cancelled if an invalid string is
3232 received. Setting client_ignore_invalid_base64 true ignores these errors.
3233 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3234 they are received. Thus, the responses can be made to depend on the
3235 challenges. If an invalid string is ignored, an empty string is placed in
3238 PH/46 Messages that are created by the autoreply transport now contains a
3239 References: header, in accordance with RFCs 2822 and 3834.
3241 PH/47 Added authenticated_sender_force to the smtp transport.
3243 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3245 PH/49 Installed latest patch from the Sieve maintainer.
3247 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3248 was also set, the appendfile transport was unnecessarily scanning a
3249 directory of message files (e.g. for maildir delivery) to find the count
3250 of files (along with the size), even though it did not need this
3251 information. It now does the scan only if it needs to find either the
3252 size of the count of files.
3254 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3256 PH/52 Two bugs concerned with error handling when the smtp transport is
3259 (i) Exim was not creating retry information for temporary errors given
3260 for individual recipients after the DATA command when the smtp transport
3261 was used in LMTP mode. This meant that they could be retried too
3262 frequently, and not timed out correctly.
3264 (ii) Exim was setting the flag that allows error details to be returned
3265 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3266 recipients that were returned after the DATA command.
3268 PH/53 This is related to PH/52, but is more general: for any failing address,
3269 when detailed error information was permitted to be returned to the
3270 sender, but the error was temporary, then after the final timeout, only
3271 "retry timeout exceeded" was returned. Now it returns the full error as
3272 well as "retry timeout exceeded".
3274 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3275 do this, and (what is worse) MTAs that accept it.
3277 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3278 will now be deprecated.
3280 PH/56 New os.c-cygwin from the Cygwin maintainer.
3282 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3283 formats (previously only available in exim formats via -bpr, -bpru,
3284 and -bpra. Now also available in native and exiqgrep formats)
3286 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3287 with very large, slow to parse queues
3289 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3291 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3293 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3294 responses to authentication challenges, though it was showing the
3295 challenges; (ii) I've removed the CR characters from the debug output for
3298 PH/58 Allow for the insertion of a newline as well as a space when a string
3299 is turned into more than one encoded-word during RFC 2047 encoding. The
3300 Sieve code now uses this.
3302 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3303 data_4xx, lost_connection, tls_required.
3305 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3306 message was being sent as an SMTP response.
3308 PH/61 Add -l and -k options to exicyclog.
3310 PH/62 When verifying, if an address was redirected to one new address, so that
3311 verification continued, and the new address failed or deferred after
3312 having set something in $address_data, the value of $address_data was not
3313 passed back to the ACL. This was different to the case when no
3314 redirection occurred. The value is now passed back in both cases.
3316 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3317 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3318 use login_cap.h, so on its own it isn't the distinguishing feature. The
3319 new name refers directly to the setclassresources() function.
3321 PH/65 Added configuration files for NetBSD3.
3323 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3325 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3326 is preferred over IPv4.
3328 PH/68 The bounce_return_message and bounce_return_body options were not being
3329 honoured for bounces generated during the reception of non-SMTP messages.
3330 In particular, this applied to messages rejected by the ACL. This bug has
3331 been fixed. However, if bounce_return_message is true and bounce_return_
3332 body is false, the headers that are returned for a non-SMTP message
3333 include only those that have been read before the error was detected.
3334 (In the case of an ACL rejection, they have all been read.)
3336 PH/69 The HTML version of the specification is now built in a directory called
3337 spec_html instead of spec.html, because the latter looks like a path with
3338 a MIME-type, and this confuses some software.
3340 PH/70 Catch two compiler warnings in sieve.c.
3342 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3343 function verify_get_ident() calls ip_connect() to connect a socket, but
3344 if the "connect()" function timed out, ip_connect() used to close the
3345 socket. However, verify_get_ident() also closes the socket later, and in
3346 between Exim writes to the log, which may get opened at this point. When
3347 the socket was closed in ip_connect(), the log could get the same file
3348 descriptor number as the socket. This naturally causes chaos. The fix is
3349 not to close the socket in ip_connect(); the socket should be closed by
3350 the function that creates it. There was only one place in the code where
3351 this was missing, in the iplookup router, which I don't think anybody now
3352 uses, but I've fixed it anyway.
3354 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3355 well as to direct DNS lookups. Otherwise the handling of names in host
3356 lists is inconsistent and therefore confusing.
3362 PH/01 Two changes to the default runtime configuration:
3364 (1) Move the checks for relay_from_hosts and authenticated clients from
3365 after to before the (commented out) DNS black list checks.
3367 (2) Add control=submission to the relay_from_hosts and authenticated
3368 clients checks, on the grounds that messages accepted by these
3369 statements are most likely to be submissions.
3371 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3373 (1) Generate an error if the third argument for the ${prvs expansion is
3376 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3379 (3) Reset the variables that are obtained from the first argument of
3380 ${prvscheck and used in the second argument before leaving the code,
3381 because their memory is reclaimed, so using them afterwards may do
3384 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3385 one (it's much easier than Tom thought :-).
3387 (5) Because of (4), we can now allow for the use of $prvscheck_result
3388 inside the third argument.
3390 PH/03 For some reason, the default setting of PATH when running a command from
3391 a pipe transport was just "/usr/bin". I have changed it to
3394 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3395 anything to be listed in the output from -bV.
3397 PH/05 When a filter generated an autoreply, the entire To: header line was
3398 quoted in the delivery log line, like this:
3400 => >A.N.Other <ano@some.domain> <original@ddress> ...
3402 This has been changed so that it extracts the operative address. There
3403 may be more than one such address. If so, they are comma-separated, like
3406 => >ano@some.domain,ona@other.domain <original@ddress> ...
3408 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3409 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3410 not being looked up in the rDNS to get a host name, Exim was showing the
3411 IP address twice in Received: lines, even though the IP addresses were
3412 identical. For example:
3414 Received: from [1.2.3.4] (helo=[1.2.3.4])
3416 However, if the real host name was known, it was omitting the HELO data
3417 if it matched the actual IP address. This has been tidied up so that it
3418 doesn't show the same IP address twice.
3420 PH/07 When both +timestamp and +memory debugging was on, the value given by
3421 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3422 called by the debug printing, thereby overwriting the timestamp buffer.
3423 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3425 PH/08 When the original message was included in an autoreply transport, it
3426 always said "this is a copy of the message, including all the headers",
3427 even if body_only or headers_only was set. It now gives an appropriate
3430 PH/09 Applied a patch from the Sieve maintainer which:
3432 o fixes some comments
3433 o adds the (disabled) notify extension core
3434 o adds some debug output for the result of if/elsif tests
3435 o points to the current vacation draft in the documentation
3436 and documents the missing references header update
3440 o fixes a bug in processing the envelope test (when testing
3441 multiple envelope elements, the last element determined the
3444 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3445 Electronic Mail") by including:
3447 Auto-submitted: auto-generated
3449 in the messages that it generates (bounce messages and others, such as
3450 warnings). In the case of bounce messages for non-SMTP messages, there was
3451 also a typo: it was using "Auto_submitted" (underscore instead of
3452 hyphen). Since every message generated by Exim is necessarily in response
3453 to another message, thes have all been changed to:
3455 Auto-Submitted: auto-replied
3457 in accordance with these statements in the RFC:
3459 The auto-replied keyword:
3461 - SHOULD be used on messages sent in direct response to another
3462 message by an automatic process,
3464 - MUST NOT be used on manually-generated messages,
3466 - MAY be used on Delivery Status Notifications (DSNs) and Message
3467 Disposition Notifications (MDNs),
3469 - MUST NOT be used on messages generated by automatic or periodic
3470 processes, except for messages which are automatic responses to
3473 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3474 to the default Received: header definition.
3476 PH/12 Added log selector acl_warn_skipped (default on).
3478 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3479 variables because (a) they are in the wrong storage pool and (b) even if
3480 they were copied, it wouldn't work properly because of the caching.
3482 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3483 checking when decoding. Apparently there are clients that generate
3484 overlong encoded strings. Why am I not surprised?
3486 PH/15 If the first argument of "${if match_address" was not empty, but did not
3487 contain an "@" character, Exim crashed. Now it writes a panic log message
3488 and treats the condition as false.
3490 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3492 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3493 extension "envelope-auth". The code is finished and in agreement with
3494 other implementations, but there is no documentation so far and in fact,
3495 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3496 not changing the active code.
3498 Print executed "if" and "elsif" statements when debugging is used. This
3499 helps a great deal to understand what a filter does.
3501 Document more things not specified clearly in RFC3028. I had all this
3502 sorted out, when out of a sudden new issues came to my mind. Oops."
3504 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3507 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3508 canonical form (as documented). However, after a host name lookup from
3509 the IP address, check_host() was doing a simple string comparison with
3510 addresses acquired from the DNS when checking that the found name did
3511 have the original IP as one of its addresses. Since any found IPv6
3512 addresses are likely to be in abbreviated form, the comparison could
3513 fail. Luckily, there already exists a function for doing the comparison
3514 by converting both addresses to binary, so now that is used instead of
3515 the text comparison.
3517 PH/20 There was another similar case to PH/19, when a complete host name was
3518 given in a host list; looking up its IP address could give an abbreviated
3519 form, whereas the current host's name might or might not be abbreviated.
3520 The same fix has been applied.
3526 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3527 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3530 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3531 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3533 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3535 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3536 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3537 Also, for "Auto-Submitted", treat anything other than "no" as
3538 non-personal, in accordance with RFC 3834. (Previously it treated
3539 anything starting "auto-" as non-personal.)
3541 TF/01 The control=submission/name=... option had a problem with syntax
3542 errors if the name included a slash character. The /name= option
3543 now slurps the rest of the string, so it can include any characters
3544 but it must come last in the list of options (after /sender_retain
3547 PH/05 Some modifications to the interface to the fake nameserver for the new
3555 TK/01 Added the "success_on_redirect" address verification option. See
3556 NewStuff for rationale and an example.
3558 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3560 PH/02 Patch to exigrep to allow it to work on syslog lines.
3562 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3563 fread() to skip over the body file's header line, because in Cygwin the
3564 header line is locked and is inaccessible.
3566 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3567 co-exist for some time) to make it clear that it is the Exim ID that is
3568 referenced, not the Message-ID: header line.
3570 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3571 string_format() function, because snprintf() does not exist on all
3574 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3575 ${stat: expansion item.
3577 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3578 protocol synchronization error", to keep the pedants happy.
3580 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3581 well as for IRIX systems, when gcc is being used. See the host.c source
3584 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3586 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3589 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3590 transport and to the smtp transport in LMTP mode.
3592 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3594 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3595 run a filter in a subprocess. This could lead to confusion in subsequent
3596 lookups in the parent process. There should also be a search_tidyup() at
3597 the end of the subprocess.
3599 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3600 only if the host matched helo_try_verify_hosts, which caused the
3601 verification to occur when the EHLO/HELO command was issued. The ACL just
3602 tested the remembered result. Now, if a previous verification attempt has
3603 not happened, "verify = helo" does it there and then.
3605 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3607 TK/03 Fix log output including CR from clamd.
3609 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3610 exist provoked a memory error which could cause a segfault.
3612 PH/15 Installed PCRE 6.2
3614 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3616 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3617 of the problem. Specifically, suggested +O2 rather than +O1 for the
3620 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3622 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3623 record, the port was not correctly logged when the outgoing_port log
3624 selector was set (it logged the transort's default port).
3626 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3627 fallback_hosts, and "hosts" in the smtp transport.
3629 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3630 host errors such as "Connection refused".
3632 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3633 authentication with radiusclient 0.4.9:
3635 - Error returned from rc_read_config was caught wrongly
3636 - Username/password not passed on to radius server due to wrong length.
3638 The presumption is that some radiusclient API changes for 4.51/PH/17
3639 were not taken care of correctly. The code is still untested by me (my
3640 Linux distribution still has 0.3.2 of radiusclient), but it was
3641 contributed by a Radius user.
3643 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3644 expanding the "port" option of the smtp transport.
3646 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3647 while reading a MIME header. Thanks to Tom Hughes for a patch.
3649 PH/24 Include config.h inside local_scan.h so that configuration settings are
3652 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3653 that in an ACL for RCPT (for example), you can examine exactly what was
3656 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3657 commands, but it was not correctly comparing the address with the actual
3658 client host address. Thus, it would show the EHLO address in Received:
3659 header lines when this was not necessary.
3661 PH/27 Added the % operator to ${eval:}.
3663 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3664 it should be ignoring failures (because with -C, for example, it has lost
3665 privilege). It wasn't ignoring creation failures other than "already
3668 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3671 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3672 because an input line was too long, either on its own, or by virtue of
3673 too many continuations, the temporary file was not being removed, and the
3674 return code was incorrect.
3676 PH/31 Missing "BOOL" in function definition in filtertest.c.
3678 PH/32 Applied Sieve patches from the maintainer.
3680 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3682 PH/33 Added "verify = not_blind".
3684 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3685 Local/Makefile (with some defaults set). These are used in built scripts
3686 such as exicyclog, but they have never been used in the exim_install
3687 script (though there are many overriding facilities there). I have
3688 arranged that the exim_install script now takes note of these two
3691 PH/35 Installed configuration files for Dragonfly.
3693 PH/36 When a locally submitted message by a trusted user did not contain a
3694 From: header, and the sender address was obtained from -f or from an SMTP
3695 MAIL command, and the trusted user did not use -F to supply a sender
3696 name, $originator_name was incorrectly used when constructing a From:
3697 header. Furthermore, $originator_name was used for submission mode
3698 messages from external hosts without From: headers in a similar way,
3699 which is clearly wrong.
3701 PH/37 Added control=suppress_local_fixups.
3703 PH/38 When log_selector = +received_sender was set, and the addition of the
3704 sender made the log line's construction buffer exactly full, or one byte
3705 less than full, an overflow happened when the terminating "\n" was
3708 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3709 when the result of a list match is failure because a DNS lookup failed.
3711 PH/40 RM_COMMAND is now used in the building process.
3713 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3714 the "build-* directories that it finds.
3716 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3717 address in a domain literal was a prefix of an interface address.
3719 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3720 when verifying a sender address, unless rewrite_headers is false.
3722 PH/44 Wrote a long comment about why errors_to addresses are verified as
3723 recipients, not senders.
3725 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3726 the ratelimit ACL was added.
3728 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3730 PH/46 Added extra information about PostgreSQL errors to the error string.
3732 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3733 avoiding the need to install special zones in a real server. This is
3734 backwards compatible; if it can't find the fake resolver, it drops back.
3735 Thus, both old and new test suites can be run.
3737 TF/02 Added util/ratelimit.pl
3739 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3740 clock is set back in time.
3742 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3743 Candler <B.Candler@pobox.com>.
3745 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3746 OK for addresses that are the result of redirections.
3748 PH/48 A number of further additions for the benefit of the new test suite,
3749 including a fake gethostbyname() that interfaces to the fake DNS resolver
3752 TF/06 The fix for widen_domains has also been applied to qualify_single and
3753 search_parents which are the other dnslookup options that can cause
3756 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3759 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3761 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3762 service" if it encountered an MX record with an empty target hostname.
3763 The message is now "an MX or SRV record indicated no SMTP service".
3765 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3766 if the DNS of the sending site is misconfigured. This is quite a
3767 common situation. This change restores the behaviour of treating a
3768 helo verification defer as a failure.
3770 PH/51 If self=fail was set on a router, the bounce message did not include the
3771 actual error message.
3777 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3779 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3780 command itself ended in error, the underlying message about the transport
3781 filter timeout was being overwritten with the pipe command error. Now the
3782 underlying error message should be appended to the second error message.
3784 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3786 PH/02 Reduce the amount of output that "make" produces by default. Full output
3787 can still be requested.
3789 PH/03 The warning log line about a condition test deferring for a "warn" verb
3790 was being output only once per connection, rather than after each
3791 occurrence (because it was using the same function as for successful
3792 "warn" verbs). This seems wrong, so I have changed it.
3794 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3795 it should not have, which might have caused a crash in the right
3796 circumstances, but probably never did.
3798 PH/04 Installed a modified version of Tony Finch's patch to make submission
3799 mode fix the return path as well as the Sender: header line, and to
3800 add a /name= option so that you can make the user's friendly name appear
3803 TF/03 Added the control = fakedefer ACL modifier.
3805 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3806 Mark Lowes for thorough testing.
3808 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3810 TK/03 Merged latest SRS patch from Miles Wilton.
3812 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3813 with the definition in sysexits.h (which is #included earlier).
3814 Fortunately, Exim does not actually use EX_OK. The code used to try to
3815 preserve the sysexits.h value, by assuming that macro definitions were
3816 scanned for macro replacements. I have been disabused of this notion,
3817 so now the code just undefines EX_OK before #including unistd.h.
3819 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3820 in the smtp transport. When a block could not be written in a single
3821 write() function, the timeout was being re-applied to each part-write.
3822 This seems wrong - if the receiver was accepting one byte at a time it
3823 would take for ever. The timeout is now adjusted when this happens. It
3824 doesn't have to be particularly precise.
3826 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3827 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3829 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3830 without a domain if the check to <postmaster@domain> fails.
3832 SC/01 Eximstats: added -xls and the ability to specify output files
3833 (patch written by Frank Heydlauf).
3835 SC/02 Eximstats: use FileHandles for outputting results.
3837 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3839 SC/04 Eximstats: fixed display of large numbers with -nvr option
3841 SC/05 Eximstats: fixed merging of reports with empty tables.
3843 SC/06 Eximstats: added the -include_original_destination flag
3845 SC/07 Eximstats: removed tabs and trailing whitespace.
3847 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3849 TK/06 MBOX spool code: Add real "From " MBOX separator line
3850 so the .eml file is really in mbox format (even though
3851 most programs do not really care). Patch from Alex Miller.
3853 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3854 The latter is generated from $received_to and is only set if the
3855 message has one envelope recipient. SA can use these headers,
3856 obviously out-of-the-box. Patch from Alex Miller.
3858 PH/08 The ${def test on a variable was returning false if the variable's
3859 value was "0", contrary to what the specification has always said!
3860 The result should be true unless the variable is empty.
3862 PH/09 The syntax error of a character other than { following "${if
3863 def:variable_name" (after optional whitespace) was not being diagnosed.
3864 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3865 accidental colon was present, for example, could give incorrect results.
3867 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3868 result is used (not including appendfile, where other changes are about
3871 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3872 This involved changing a lot of size variables from int to off_t. It
3873 should work with maildirs and everything.
3875 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3876 spamd dying while we are connected to it.
3878 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3881 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3882 function for BDB 4.3.
3884 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3886 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3887 the sources for PCRE have been re-arranged and more files are now
3890 PH/15 The code I had for printing potentially long long variables in PH/11
3891 above was not the best (it lost precision). The length of off_t variables
3892 is now inspected at build time, and an appropriate printing format (%ld
3893 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3894 to be "long long int" or "long int". This is needed for the internal
3895 formatting function string_vformat().
3897 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3898 the configuration file to be ":syslog", then the script "guesses" where
3899 the logs files are, rather than using the compiled in default. In our
3900 case the guess is not the same as the compiled default, so the script
3901 suddenly stopped working when I started to use syslog. The patch checks
3902 to see if log_file_path is "". If so, it attempts to read it from exim
3903 with no configuration file to get the compiled in version, before it
3904 falls back to the previous guessing code."
3906 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3907 implementing BATV in an Exim configuration. See NewStuff for the gory
3910 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3911 Makefile that are specific to HP-UX.
3913 PH/18 If the "use_postmaster" option was set for a recipient callout together
3914 with the "random" option, the postmaster address was used as the MAIL
3915 FROM address for the random test, but not for the subsequent recipient
3916 test. It is now used for both.
3918 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3919 patch removes a few documentation additions to RFC 3028, because the
3920 latest draft now contains them. It adds the new en;ascii-case comparator
3921 and a new error check for 8bit text in MIME parts. Comparator and
3922 require names are now matched exactly. I enabled the subaddress
3923 extension, but it is not well tested yet (read: it works for me)."
3925 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3926 rework some of the code of TK/09 above to avoid the hardwired use of
3927 "%lld" and "long long". Replaced the call to snprintf() with a call to
3930 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3931 records point to non-existent hosts", "retry timeout exceeded", and
3932 "retry time not reached for any host after a long failure period".
3934 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3935 experimental DomainKeys support:
3937 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3938 (2) On an error such as an illegally used "control", the wrong name for
3939 the control was given.
3941 These problems did NOT occur unless DomainKeys support was compiled.
3943 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3945 PH/24 Added ${if match_ip condition.
3947 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3948 databases so that it will be absolutely obvious if a crash occurs in the
3949 DB library. This is a regular occurrence (often caused by mis-matched
3952 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3953 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3954 picked up on a user's system that detects such things. There doesn't seem
3955 to be a gcc warning option for this - only an attribute that has to be
3956 put on the function's prototype. It seems that in Fedora Core 4 they have
3957 set this on a number of new functions. No doubt there will be more in due
3960 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3961 specify a transport. However, if an address that was verified by such a
3962 router was the subject of a callout, Exim crashed because it tried to
3963 read the rcpt_include_affixes from the non-existent transport. Now it
3964 just assumes that the setting of that option is false. This bug was
3965 introduced by 4.51/PH/31.
3967 PH/28 Changed -d+all to exclude +memory, because that information is very
3968 rarely of interest, but it makes the output a lot bigger. People tend to
3969 do -d+all out of habit.
3971 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3972 code in os-type was giving problems when libc.so lives in lib64, like on
3975 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3976 aren't the modern standard, and it seems that some systems' include files
3977 don't always have them. Exim was already checking for some of the newer
3978 ones like T_AAAA, and defining it itself. I've added checks for all the
3979 record types that Exim uses.
3981 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3982 not automatically generating a new one, as it is supposed to. This
3983 prevented TLS from working. If the file did exist, but contained invalid
3984 data, a new version was generated, as expected. It was only the case of a
3985 non-existent file that was broken.
3987 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3988 with a change in libdomainkeys > 0.64.
3990 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3991 from DNS. If the selector record carries the flag, it now has
3992 precedence over the domain-wide flag.
3994 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3996 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3997 the use of a port name that isn't defined in /etc/services) occurred, the
3998 message was deferred as in a normal delivery, and thus remained on the
3999 spool, instead of being failed because of the mua_wrapper setting. This
4000 is now fixed, and I tidied up some of the mua_wrapper messages at the
4003 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4004 the messages in a hash of arrays rather than using individual hashes.
4005 This is a bit cleaner and results in dramatic memory savings, albeit
4006 at a slight CPU cost.
4008 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4009 as requested by Marc Sherman.
4011 SC/10 Eximstats: added histograms for user specified patterns as requested
4014 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4016 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4017 fopen() in the content-scanning modules that did not already have it.
4023 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4024 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4026 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4028 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4030 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4031 to test Sieve filters that use "vacation".
4033 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4034 that changes the way the GnuTLS parameters are stored in the cache file.
4035 The new format can be generated externally. For backward compatibility,
4036 if the data in the cache doesn't make sense, Exim assumes it has read an
4037 old-format file, and it generates new data and writes a new file. This
4038 means that you can't go back to an older release without removing the
4041 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4042 work if there are any delivery delays because "one_time" forces the
4043 parent to be marked "delivered", so its unseen clone is never tried
4044 again. For this reason, Exim now forbids the simultaneous setting of
4047 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4048 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4049 entry if you want to know the details. The fix, however, neglected to
4050 consider the case when local delivery batching is involved. The test for
4051 "previously delivered" was not happening when checking to see if an
4052 address could be batched with a previous (undelivered) one; under
4053 certain circumstances this could lead to multiple deliveries to the same
4056 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4057 in its include files, and this causes problems building Exim.
4059 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4060 header_syntax) but Exim was just ignoring anything given after a slash.
4061 In particular, this caused confusion with an attempt to use "verify =
4062 reverse_host_lookup/defer_ok". An error is now given when options are
4063 supplied for verify items that do not have them. (Maybe reverse_host_
4064 lookup should have a defer_ok option, but that's a different point.)
4066 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4067 defined by RFC 821) to 2048, because there were problems with some AUTH
4068 commands, and RFC 1869 says the size should be increased for extended
4069 SMTP commands that take arguments.
4071 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4074 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4075 "unknown" error; now it says that the functionality isn't in the binary.
4077 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4078 an address' error message when a string expansion fails (syntax or
4079 whatever). Otherwise the password may appear in the log. Following change
4080 PH/42 below, there is no longer a chance of it appearing in a bounce
4083 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4085 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4086 of its addresses was ever tried. (Bugzilla bug #2.)
4088 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4089 the result incorrectly in the debug output. (It correctly added a newline
4090 to what was transported.)
4092 TF/01 Added $received_time.
4094 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4095 commented out examples of how to interface to a virus scanner and to
4096 SpamAssassin. Also added commented examples of av_scanner and
4097 spamd_address settings.
4099 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4100 and controls are allowed in which ACLs. There were a couple of minor
4101 errors. Some of the entries in the conditions table (which is a table of
4102 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4103 negation of where the condition IS allowed.
4105 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4107 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4108 header file does not have a version number, so I've had to invent a new
4109 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4110 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4111 radiusclient), but it was contributed by a Radius user.
4113 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4114 files or directories, for OpenSSL.
4116 PH/19 When an Exim process that is running as root has to create an Exim log
4117 file, it does so in a subprocess that runs as exim:exim so as to get the
4118 ownership right at creation (otherwise, other Exim processes might see
4119 the file with the wrong ownership). There was no test for failure of this
4120 fork() call, which would lead to the process getting stuck as it waited
4121 for a non-existent subprocess. Forks do occasionally fail when resources
4122 run out. I reviewed all the other calls to fork(); they all seem to check
4125 PH/20 When checking for unexpected SMTP input at connect time (before writing
4126 the banner), Exim was not dealing correctly with a non-positive return
4127 from the read() function. If the client had disconnected by this time,
4128 the result was a log entry for a synchronization error with an empty
4129 string after "input=" when read() returned zero. If read() returned -1
4130 (an event I could not check), uninitialized data bytes were printed.
4131 There were reports of junk text (parts of files, etc) appearing after
4134 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4136 PH/22 Added support for macro redefinition, and (re)definition in between
4137 driver and ACL definitions.
4139 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4140 forgetting to use the resulting value; it was using the unexpanded value.
4142 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4143 hadn't been configured. The fix is from Juergen Kreileder, who
4144 understands it better than I do:
4146 "Here's what I see happening with three configured cyrus_sasl
4147 authenticators configured (plain, login, cram-md5):
4149 On startup auth_cyrus_sasl_init() gets called for each of these.
4150 This means three calls to sasl_listmech() without a specified mech_list.
4151 => SASL tests which mechs of all available mechs actually work
4152 => three warnings about OTP not working
4153 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4155 With the patch, sasl_listmech() also gets called three times. But now
4156 SASL's mech_list option is set to the server_mech specified in the the
4157 authenticator. Or in other words, the answer from sasl_listmech()
4158 gets limited to just the mech you're testing for (which is different
4160 => the return list contains just 'plain' or 'login', 'cram-md5' or
4161 nothing depending on the value of ob->server_mech.
4163 I've just tested the patch: Authentication still works fine,
4164 unavailable mechs specified in the exim configuration are still
4165 caught, and the auth.log warnings about OTP are gone."
4167 PH/25 When debugging is enabled, the contents of the command line are added
4168 to the debugging output, even when log_selector=+arguments is not
4171 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4172 answer is "GNU", and only if the return is "GNU/something" is the answer
4175 PH/27 $acl_verify_message is now set immediately after the failure of a
4176 verification in an ACL, and so is available in subsequent modifiers. In
4177 particular, the message can be preserved by coding like this:
4179 warn !verify = sender
4180 set acl_m0 = $acl_verify_message
4182 Previously, $acl_verify_message was set only while expanding "message"
4183 and "log_message" when a very denied access.
4185 PH/28 Modified OS/os.c-Linux with
4187 -#ifndef OS_LOAD_AVERAGE
4188 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4190 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4191 nomenclature these days.)
4193 PH/29 Installed patch from the Sieve maintainer that adds the options
4194 sieve_useraddress and sieve_subaddress to the redirect router.
4196 PH/30 In these circumstances:
4197 . Two addresses routed to the same list of hosts;
4198 . First host does not offer TLS;
4199 . First host accepts first address;
4200 . First host gives temporary error to second address;
4201 . Second host offers TLS and a TLS session is established;
4202 . Second host accepts second address.
4203 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4204 and peerdn, if requested) that were in fact used only for the second
4207 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4208 attention to any local part prefix or suffix that was matched by the
4209 router that accepted the address. It now behaves in the same way as it
4210 does for delivery: the affixes are removed from the local part unless
4211 rcpt_include_affixes is set on the transport.
4213 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4214 timeout during the DATA phase of an incoming message.
4216 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4217 applied a patch sanctioned by the Sieve maintainer.
4219 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4220 the uid or gid is negative. A case of a negative gid caused this to be
4221 noticed. The fix allows for either to be negative.
4223 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4224 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4227 PH/36 The change PH/12 above was broken. Fixed it.
4229 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4230 the grounds that routing the same address twice would always produce the
4231 same answer. This might have been true once, but it is certainly no
4232 longer true now. Routing a child address may depend on the previous
4233 routing that produced that child. Some complicated redirection strategies
4234 went wrong when messages had multiple recipients, and made Exim's
4235 behaviour dependent on the order in which the addresses were given.
4237 I have moved the duplicate checking until after the routing is complete.
4238 Exim scans the addresses that are assigned to local and remote
4239 transports, and removes any duplicates. This means that more work will be
4240 done, as duplicates will always all be routed, but duplicates are
4241 presumably rare, so I don't expect this is of any significance.
4243 For deliveries to pipes, files, and autoreplies, the duplicate checking
4244 still happens during the routing process, since they are not going to be
4247 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4248 It corrects a timeout issue with spamd. This is Ian's comment: "The
4249 background is that sometimes spamd either never reads data from a
4250 connection it has accepted, or it never writes response data. The exiscan
4251 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4252 blindly assumes that writes won't block so it may never time out."
4254 PH/39 Allow G after quota size as well as K and M.
4256 PH/40 The value set for $authenticated_id in an authenticator may not contain
4257 binary zeroes or newlines because the value is written to log lines and
4258 to spool files. There was no check on this. Now the value is run through
4259 the string_printing() function so that such characters are converted to
4260 printable escape sequences.
4262 PH/41 $message_linecount is a new variable that contains the total number of
4263 lines in the message. Compare $body_linecount, which is the count for the
4266 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4267 bounce and delay warning messages, except in certain special cases, which
4270 (a) An SMTP error message from a remote host;
4271 (b) A message specified in a :fail: redirection;
4272 (c) A message specified in a "fail" command in a system filter;
4273 (d) A message specified in a FAIL return from the queryprogram router;
4274 (e) A message specified by the cannot_route_message router option.
4276 In these cases only, Exim does include the error details in bounce and
4277 warning messages. There are also a few cases where bland messages such
4278 as "unrouteable address" or "local delivery error" are given.
4280 PH/43 $value is now also set for the "else" part of a ${run expansion.
4282 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4283 being worked on, but at least Exim now implements the latest version to
4286 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4287 process to complete was treated as a delivery failure, a timeout while
4288 writing the message to the pipe was logged, but erroneously treated as a
4289 successful delivery. Such timeouts include transport filter timeouts. For
4290 consistency with the overall process timeout, these timeouts are now
4291 treated as errors, giving rise to delivery failures by default. However,
4292 there is now a new Boolean option for the pipe transport called
4293 timeout_defer, which, if set TRUE, converts the failures into defers for
4294 both kinds of timeout. A transport filter timeout is now identified in
4297 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4298 systems where "make" and "gmake" are different, calling "gmake" at top
4299 level broke things. I've arranged for the value of $(MAKE) to be passed
4300 from the Makefile to this script so that it can call the same version of
4304 A note about Exim versions 4.44 and 4.50
4305 ----------------------------------------
4307 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4308 changes of various kinds. As a consequence, a big documentation update was
4309 needed. This delayed the release for rather longer than seemed good, especially
4310 in the light of a couple of (minor) security issues. Therefore, the changes
4311 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4312 release. So 4.44 and 4.50 are in effect two different branches that both start
4315 I have left the 4.50 change log unchanged; it contains all the changes since
4316 4.43. The change log for 4.44 is below; many of its items are identical to
4317 those for 4.50. This seems to be the most sensible way to preserve the
4318 historical information.
4324 1. Minor wording change to the doc/README.SIEVE file.
4326 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4327 computation of the current number of files was incorrect.
4329 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4330 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4333 4. Give more explanation in the error message when the command for a transport
4334 filter fails to execute.
4336 5. There are several places where Exim runs a non-Exim command in a
4337 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4338 was being done only for the command run by the queryprogram router. It is
4339 now done for all such subprocesses. The other cases are: ${run, transport
4340 filters, and the commands run by the lmtp and pipe transports.
4342 6. Added CONFIGURE_GROUP build-time option.
4344 7. Some older OS have a limit of 256 on the maximum number of file
4345 descriptors. Exim was using setrlimit() to set 1000 as a large value
4346 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4347 systems. I've change it so that if it can't get 1000, it tries for 256.
4349 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4350 was an oversight, and furthermore, ever since the addition of extra
4351 controls (e.g. 4.43/32), the checks on when to allow different forms of
4352 "control" were broken. There should now be diagnostics for all cases when a
4353 control that does not make sense is encountered.
4355 9. Added the /retain_sender option to "control=submission".
4357 10. $recipients is now available in the predata ACL (oversight).
4359 11. Tidy the search cache before the fork to do a delivery from a message
4360 received from the command line. Otherwise the child will trigger a lookup
4361 failure and thereby defer the delivery if it tries to use (for example) a
4362 cached ldap connection that the parent has called unbind on.
4364 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4365 of $address_data from the recipient verification was clobbered by the
4366 sender verification.
4368 13. The value of address_data from a sender verification is now available in
4369 $sender_address_data in subsequent conditions in the ACL statement.
4371 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4373 15. Added a new option "connect=<time>" to callout options, to set a different
4376 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4377 was its contents. (It was OK if the option was not defined at all.)
4379 17. A "Completed" log line is now written for messages that are removed from
4380 the spool by the -Mrm option.
4382 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4383 information about exactly what failed.
4385 19. Added -dd to debug only the daemon process.
4387 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4388 handles timeouts, both on the server side and network timeouts. Renamed the
4389 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4391 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4392 It is now set to "smtps".
4394 22. $host_address is now set to the target address during the checking of
4395 ignore_target_hosts.
4397 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4398 being passed; this would have caused $sender_host_name to have been used if
4399 matching the list had actually called for a host name (not very likely,
4400 since this list is usually IP addresses). A host name is now passed as
4403 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4404 code that specifies a non-restarting handler (typically sigaction() in
4405 modern systems) in an attempt to fix a rare and obscure crash bug.
4407 25. Narrowed the window for a race in the daemon that could cause it to ignore
4408 SIGCHLD signals. This is not a major problem, because they are used only to
4409 wake it up if nothing else does.
4411 26. A malformed maildirsize file could cause Exim to calculate negative values
4412 for the mailbox size or file count. Odd effects could occur as a result.
4413 The maildirsize information is now recalculated if the size or filecount
4416 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4417 support for a long time. Removed HAVE_SYS_VFS_H.
4419 28. Installed the latest version of exipick from John Jetmore.
4421 29. In an address list, if the pattern was not a regular expression, an empty
4422 subject address (from a bounce message) matched only if the pattern was an
4423 empty string. Non-empty patterns were not even tested. This was the wrong
4424 because it is perfectly reasonable to use an empty address as part of a
4425 database query. An empty address is now tested by patterns that are
4426 lookups. However, all the other forms of pattern expect the subject to
4427 contain a local part and a domain, and therefore, for them, an empty
4428 address still always fails if the pattern is not itself empty.
4430 30. Exim went into a mad DNS loop when attempting to do a callout where the
4431 host was specified on an smtp transport, and looking it up yielded more
4432 than one IP address.
4434 31. Re-factored the code for checking spool and log partition space into a
4435 function that finds that data and another that does the check. The former
4436 is then used to implement four new variables: $spool_space, $log_space,
4437 $spool_inodes, and $log_inodes.
4439 32. The RFC2047 encoding function was originally intended for short strings
4440 such as real names; it was not keeping to the 75-character limit for
4441 encoded words that the RFC imposes. It now respects the limit, and
4442 generates multiple encoded words if necessary. To be on the safe side, I
4443 have increased the buffer size for the ${rfc2047: expansion operator from
4446 33. It is now permitted to omit both strings after an "if" condition; if the
4447 condition is true, the result is "true". As before, when the second string
4448 is omitted, a false condition yields an empty string. This makes it less
4449 cumbersome to write custom ACL and router conditions.
4451 34. Failure to deliver a bounce message always caused it to be frozen, even if
4452 there was an errors_to setting on the router. The errors_to setting is now
4455 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4456 canonical form (fully expanded) before being placed in
4457 $sender_host_address.
4459 36. The table in the code that translates DNS record types into text (T_A to
4460 "A" for instance) was missing entries for NS and CNAME. It is just possible
4461 that this could have caused confusion if both these types were looked up
4462 for the same domain, because the text type is used as part of Exim's
4463 per-process caching. But the chance of anyone hitting this buglet seems
4466 37. The dnsdb lookup has been extended in a number of ways.
4468 (1) There is a new type, "zns", which walks up the domain tree until it
4469 finds some nameserver records. It should be used with care.
4471 (2) There is a new type, "mxh", which is like "mx" except that it returns
4472 just the host names, not the priorities.
4474 (3) It is now possible to give a list of domains (or IP addresses) to be
4475 looked up. The behaviour when one of the lookups defers can be
4476 controlled by a keyword.
4478 (4) It is now possible to specify the separator character for use when
4479 multiple records are returned.
4481 38. The dnslists ACL condition has been extended: it is now possible to supply
4482 a list of IP addresses and/or domains to be looked up in a particular DNS
4485 39. Added log_selector=+queue_time_overall.
4487 40. When running the queue in the test harness, wait just a tad after forking a
4488 delivery process, to get repeatability of debugging output.
4490 41. Include certificate and key file names in error message when GnuTLS fails
4491 to set them up, because the GnuTLS error message doesn't include the name
4492 of the failing file when there is a problem reading it.
4494 42. Allow both -bf and -bF in the same test run.
4496 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4498 44. The "Exiscan patch" is now merged into the mainline Exim source.
4500 45. Sometimes the final signoff response after QUIT could fail to get
4501 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4502 < 0 before doing a fflush(). This bug looks as though it goes back to the
4503 introduction of TLS in release 3.20, but "sometimes" must have been rare
4504 because the tests only now provoked it.
4506 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4507 (this can affect the format of dates).
4509 47. exim_tidydb, when checking for the continued existence of a message for
4510 which it has found a message-specific retry record, was not finding
4511 messages that were in split spool directories. Consequently, it was
4512 deleting retry records that should have stayed in existence.
4514 48. Steve fixed some bugs in eximstats.
4516 49. The SPA authentication driver was not abandoning authentication and moving
4517 on to the next authenticator when an expansion was forced to fail,
4518 contradicting the general specification for all authenticators. Instead it
4519 was generating a temporary error. It now behaves as specified.
4521 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4522 (the order specifies the preference for clients). The order is now AES256,
4523 AES128, 3DES, ARCFOUR128.
4525 51. Small patch to Sieve code - explicitly set From: when generating an
4528 52. Exim crashed if a remote delivery caused a very long error message to be
4529 recorded - for instance if somebody sent an entire SpamAssassin report back
4530 as a large number of 550 error lines. This bug was coincidentally fixed by
4531 increasing the size of one of Exim's internal buffers (big_buffer) that
4532 happened as part of the Exiscan merge. However, to be on the safe side, I
4533 have made the code more robust (and fixed the comments that describe what
4536 53. Now that there can be additional text after "Completed" in log lines (if
4537 the queue_time_overall log selector is set), a one-byte patch to exigrep
4538 was needed to allow it to recognize "Completed" as not the last thing in
4541 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4542 patch that reportedly fixes this has been added. I am not expert enough to
4543 create a test for it. This is what the patch creator wrote:
4545 "I found a little strange behaviour of ldap code when working with
4546 Windows 2003 AD Domain, where users was placed in more than one
4547 Organization Units. When I tried to give exim partial DN, the exit code
4548 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4549 But simultaneously result of request was absolutely normal ldap result,
4550 so I produce this patch..."
4552 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4553 REFERENCE, so I have modified the code to exclude the patch when that macro
4556 55. Some experimental protocols are using DNS PTR records for new purposes. The
4557 keys for these records are domain names, not reversed IP addresses. The
4558 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4559 leaves it alone. Component reversal etc. now happens only for IP addresses.
4562 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4564 57. Double the size of the debug message buffer (to 2048) so that more of very
4565 long debug lines gets shown.
4567 58. The exicyclog utility now does better if the number of log files to keep
4568 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4570 59. Two changes related to the smtp_active_hostname option:
4572 (1) $smtp_active_hostname is now available as a variable.
4573 (2) The default for smtp_banner uses $smtp_active_hostname instead
4574 of $primary_hostname.
4576 60. The host_aton() function is supposed to be passed a string that is known
4577 to be a valid IP address. However, in the case of IPv6 addresses, it was
4578 not checking this. This is a hostage to fortune. Exim now panics and dies
4579 if the condition is not met. A case was found where this could be provoked
4580 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4581 components; fortuitously, this particular loophole had already been fixed
4582 by change 4.50/55 above.
4584 If there are any other similar loopholes, the new check in host_aton()
4585 itself should stop them being exploited. The report I received stated that
4586 data on the command line could provoke the exploit when Exim was running as
4587 exim, but did not say which command line option was involved. All I could
4588 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4589 running as the user.
4592 61. There was a buffer overflow vulnerability in the SPA authentication code
4593 (which came originally from the Samba project). I have added a test to the
4594 spa_base64_to_bits() function which I hope fixes it.
4597 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4598 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4600 63. The daemon start-up calls getloadavg() while still root for those OS that
4601 need the first call to be done as root, but it missed one case: when
4602 deliver_queue_load_max is set with deliver_drop_privilege. This is
4603 necessary for the benefit of the queue runner, because there is no re-exec
4604 when deliver_drop_privilege is set.
4606 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4607 This has been fixed.
4609 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4610 was in use, was not putting the data itself into the right store pool;
4611 consequently, it could be overwritten for a subsequent message in the same
4612 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4615 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4617 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4618 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4619 the function were treating the return as a boolean value, which happened to
4620 work because 0=false and not-0=true, but is not correct code.
4622 68. The host_aton() function was not handling scoped IPv6 addresses (those
4623 with, for example, "%eth0" on the end) correctly.
4625 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4626 negated items (that is, ~something) in unsigned ints. Some compilers
4627 apparently mutter when there is no cast.
4629 70. If an address verification called from an ACL failed, and did not produce a
4630 user-specific message (i.e. there was only a "system" message), nothing was
4631 put in $acl_verify_message. In this situation, it now puts the system
4634 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4635 session; change 4.31/43 added the unwanted input to the log line - except
4636 that it did not do this in the start of session case. It now does.
4638 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4639 This is wrong and can cause the other end to generate a synchronization
4640 error if it is another Exim or anything else that does the synchronization
4641 check. A QUIT command is no longer sent after a timeout.
4643 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4644 during host lookups.
4646 74. The defer_ok option of callout verification was not working if it was used
4647 when verifying addresses in header lines, that is, for this case:
4649 verify = header_sender/callout=defer_ok
4651 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4652 those file descriptors could be used for SMTP connections. If anything
4653 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4654 could be sent to the SMTP client, causing chaos. The daemon now opens
4655 stdin, stdout, and stderr to /dev/null when it puts itself into the
4658 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4659 log by default. The user can override this with suitable Perl magic.
4661 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4662 the log message when discard triggers, was not working for the DATA ACL or
4663 for the non-SMTP ACL.
4665 78. Error message wording change in sieve.c.
4667 79. If smtp_accept_max_per_host was set, the number of connections could be
4668 restricted to fewer than expected, because the daemon was trying to set up
4669 a new connection before checking whether the processes handling previous
4670 connections had finished. The check for completed processes is now done
4671 earlier. On busy systems, this bug wouldn't be noticed because something
4672 else would have woken the daemon, and it would have reaped the completed
4675 80. If a message was submitted locally by a user whose login name contained one
4676 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4677 It caused a spool format error. I have fixed the spool reading code. A
4678 related problem was that the "from" clause in the Received: line became
4679 illegal because of the space(s). It is now covered by ${quote_local_part.
4681 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4684 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4686 83. Patch from Sieve maintainer to fix unterminated string problem in
4687 "vacation" handling.
4689 84. Some minor changes to the Linux configuration files to help with other
4690 OS variants using glibc.
4692 85. One more patch for Sieve to update vacation handling to latest spec.
4695 ----------------------------------------------------
4696 See the note above about the 4.44 and 4.50 releases.
4697 ----------------------------------------------------
4703 1. Change 4.43/35 introduced a bug that caused file counts to be
4704 incorrectly computed when quota_filecount was set in an appendfile
4707 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4708 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4711 3. Give more explanation in the error message when the command for a transport
4712 filter fails to execute.
4714 4. There are several places where Exim runs a non-Exim command in a
4715 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4716 was being done only for the command run by the queryprogram router. It is
4717 now done for all such subprocesses. The other cases are: ${run, transport
4718 filters, and the commands run by the lmtp and pipe transports.
4720 5. Some older OS have a limit of 256 on the maximum number of file
4721 descriptors. Exim was using setrlimit() to set 1000 as a large value
4722 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4723 systems. I've change it so that if it can't get 1000, it tries for 256.
4725 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4726 was an oversight, and furthermore, ever since the addition of extra
4727 controls (e.g. 4.43/32), the checks on when to allow different forms of
4728 "control" were broken. There should now be diagnostics for all cases when a
4729 control that does not make sense is encountered.
4731 7. $recipients is now available in the predata ACL (oversight).
4733 8. Tidy the search cache before the fork to do a delivery from a message
4734 received from the command line. Otherwise the child will trigger a lookup
4735 failure and thereby defer the delivery if it tries to use (for example) a
4736 cached ldap connection that the parent has called unbind on.
4738 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4739 of $address_data from the recipient verification was clobbered by the
4740 sender verification.
4742 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4743 was its contents. (It was OK if the option was not defined at all.)
4745 11. A "Completed" log line is now written for messages that are removed from
4746 the spool by the -Mrm option.
4748 12. $host_address is now set to the target address during the checking of
4749 ignore_target_hosts.
4751 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4752 being passed; this would have caused $sender_host_name to have been used if
4753 matching the list had actually called for a host name (not very likely,
4754 since this list is usually IP addresses). A host name is now passed as
4757 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4758 code that specifies a non-restarting handler (typically sigaction() in
4759 modern systems) in an attempt to fix a rare and obscure crash bug.
4761 15. Narrowed the window for a race in the daemon that could cause it to ignore
4762 SIGCHLD signals. This is not a major problem, because they are used only to
4763 wake it up if nothing else does.
4765 16. A malformed maildirsize file could cause Exim to calculate negative values
4766 for the mailbox size or file count. Odd effects could occur as a result.
4767 The maildirsize information is now recalculated if the size or filecount
4770 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4771 support for a long time. Removed HAVE_SYS_VFS_H.
4773 18. Updated exipick to current release from John Jetmore.
4775 19. Allow an empty sender to be matched against a lookup in an address list.
4776 Previously the only cases considered were a regular expression, or an
4779 20. Exim went into a mad DNS lookup loop when doing a callout where the
4780 host was specified on the transport, if the DNS lookup yielded more than
4783 21. The RFC2047 encoding function was originally intended for short strings
4784 such as real names; it was not keeping to the 75-character limit for
4785 encoded words that the RFC imposes. It now respects the limit, and
4786 generates multiple encoded words if necessary. To be on the safe side, I
4787 have increased the buffer size for the ${rfc2047: expansion operator from
4790 22. Failure to deliver a bounce message always caused it to be frozen, even if
4791 there was an errors_to setting on the router. The errors_to setting is now
4794 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4795 canonical form (fully expanded) before being placed in
4796 $sender_host_address.
4798 24. Updated eximstats to version 1.33
4800 25. Include certificate and key file names in error message when GnuTLS fails
4801 to set them up, because the GnuTLS error message doesn't include the name
4802 of the failing file when there is a problem reading it.
4804 26. Expand error message when OpenSSL has problems setting up cert/key files.
4807 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4808 (this can affect the format of dates).
4810 28. exim_tidydb, when checking for the continued existence of a message for
4811 which it has found a message-specific retry record, was not finding
4812 messages that were in split spool directories. Consequently, it was
4813 deleting retry records that should have stayed in existence.
4815 29. eximstats updated to version 1.35
4816 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4817 1.35 - bugfix such that pie charts by volume are generated correctly
4819 30. The SPA authentication driver was not abandoning authentication and moving
4820 on to the next authenticator when an expansion was forced to fail,
4821 contradicting the general specification for all authenticators. Instead it
4822 was generating a temporary error. It now behaves as specified.
4824 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4825 (the order specifies the preference for clients). The order is now AES256,
4826 AES128, 3DES, ARCFOUR128.
4828 31. Small patch to Sieve code - explicitly set From: when generating an
4831 32. Exim crashed if a remote delivery caused a very long error message to be
4832 recorded - for instance if somebody sent an entire SpamAssassin report back
4833 as a large number of 550 error lines. This bug was coincidentally fixed by
4834 increasing the size of one of Exim's internal buffers (big_buffer) that
4835 happened as part of the Exiscan merge. However, to be on the safe side, I
4836 have made the code more robust (and fixed the comments that describe what
4839 33. Some experimental protocols are using DNS PTR records for new purposes. The
4840 keys for these records are domain names, not reversed IP addresses. The
4841 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4842 leaves it alone. Component reversal etc. now happens only for IP addresses.
4845 34. The host_aton() function is supposed to be passed a string that is known
4846 to be a valid IP address. However, in the case of IPv6 addresses, it was
4847 not checking this. This is a hostage to fortune. Exim now panics and dies
4848 if the condition is not met. A case was found where this could be provoked
4849 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4850 components; fortuitously, this particular loophole had already been fixed
4851 by change 4.50/55 or 4.44/33 above.
4853 If there are any other similar loopholes, the new check in host_aton()
4854 itself should stop them being exploited. The report I received stated that
4855 data on the command line could provoke the exploit when Exim was running as
4856 exim, but did not say which command line option was involved. All I could
4857 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4858 running as the user.
4861 35. There was a buffer overflow vulnerability in the SPA authentication code
4862 (which came originally from the Samba project). I have added a test to the
4863 spa_base64_to_bits() function which I hope fixes it.
4866 36. The daemon start-up calls getloadavg() while still root for those OS that
4867 need the first call to be done as root, but it missed one case: when
4868 deliver_queue_load_max is set with deliver_drop_privilege. This is
4869 necessary for the benefit of the queue runner, because there is no re-exec
4870 when deliver_drop_privilege is set.
4872 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4873 was in use, was not putting the data itself into the right store pool;
4874 consequently, it could be overwritten for a subsequent message in the same
4875 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4878 38. Sometimes the final signoff response after QUIT could fail to get
4879 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4880 < 0 before doing a fflush(). This bug looks as though it goes back to the
4881 introduction of TLS in release 3.20, but "sometimes" must have been rare
4882 because the tests only now provoked it.
4888 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4889 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4890 Now it returns an integer. A number of calls were still expecting a T/F
4891 return. Fortuitously, in all cases, the tests worked in OK situations,
4892 which is the norm. However, things would have gone wrong on any write
4893 failures on the smtp file descriptor. This function is used when sending
4894 messages over SMTP and also when doing verify callouts.
4896 2. When Exim is called to do synchronous delivery of a locally submitted
4897 message (the -odf or -odi options), it no longer closes stderr before doing
4900 3. Implemented the mua_wrapper option.
4902 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4904 5. Implemented the functions header_remove(), header_testname(),
4905 header_add_at_position(), and receive_remove_recipient(), and exported them
4908 6. If an ACL "warn" statement specified the addition of headers, Exim already
4909 inserted X-ACL-Warn: at the start if there was no header name. However, it
4910 was not making this test for the second and subsequent header lines if
4911 there were newlines in the string. This meant that an invalid header could
4912 be inserted if Exim was badly configured.
4914 7. Allow an ACL "warn" statement to add header lines at the start or after all
4915 the Received: headers, as well as at the end.
4917 8. Added the rcpt_4xx retry error code.
4919 9. Added postmaster_mailfrom=xxx to callout verification option.
4921 10. Added mailfrom=xxxx to the callout verification option, for verify=
4924 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4925 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4927 12. Inserted some casts to stop certain compilers warning when using pointer
4928 differences as field lengths or precisions in printf-type calls (mostly
4929 affecting debugging statements).
4931 13. Added optional readline() support for -be (dynamically loaded).
4933 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4934 same clock tick as a message's arrival, so that its received time was the
4935 same as the "first fail" time on the retry record, and that message
4936 remained on the queue past the ultimate address timeout, every queue runner
4937 would try a delivery (because it was past the ultimate address timeout) but
4938 after another failure, the ultimate address timeout, which should have then
4939 bounced the address, did not kick in. This was a "< instead of <=" error;
4940 in most cases the first failure would have been in the next clock tick
4941 after the received time, and all would be well.
4943 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4944 being recognized when the domain list was tested by the match_domain
4945 condition in an expansion string.
4947 16. Added the ${str2b64: operator.
4949 17. Exim was always calling setrlimit() to set a large limit for the number of
4950 processes, without checking whether the existing limit was already
4951 adequate. (It did check for the limit on file descriptors.) Furthermore,
4952 errors from getrlimit() and setrlimit() were being ignored. Now they are
4953 logged to the main and panic logs, but Exim does carry on, to try to do its
4954 job under whatever limits there are.
4956 18. Imported PCRE 5.0.
4958 19. Trivial typo in log message " temporarily refused connection" (the leading
4961 20. If the log selector return_path_on_delivery was set and an address was
4962 redirected to /dev/null, the delivery process crashed because it assumed
4963 that a return path would always be set for a "successful" delivery. In this
4964 case, the whole delivery is bypassed as an optimization, and therefore no
4967 21. Internal re-arrangement: the function for sending a challenge and reading
4968 a response while authentication was assuming a zero-terminated challenge
4969 string. It's now changed to take a pointer and a length, to allow for
4970 binary data in such strings.
4972 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4974 23. Exim was not respecting finduser_retries when seeking the login of the
4975 uid under which it was called; it was always trying 10 times. (The default
4976 setting of finduser_retries is zero.) Also, it was sleeping after the final
4977 failure, which is pointless.
4979 24. Implemented tls_on_connect_ports.
4981 25. Implemented acl_smtp_predata.
4983 26. If the domain in control=submission is set empty, Exim assumes that the
4984 authenticated id is a complete email address when it generates From: or
4985 Sender: header lines.
4987 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4988 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4989 chown and chgrp in /bin and hostname in /usr/bin.
4991 28. Exim was keeping the "process log" file open after each use, just as it
4992 does for the main log. This opens the possibility of it remaining open for
4993 long periods when the USR1 signal hits a daemon. Occasional processlog
4994 errors were reported, that could have been caused by this. Anyway, it seems
4995 much more sensible not to leave this file open at all, so that is what now
4998 29. The long-running daemon process does not normally write to the log once it
4999 has entered its main loop, and it closes the log before doing so. This is
5000 so that log files can straightforwardly be renamed and moved. However,
5001 there are a couple of unusual error situations where the daemon does write
5002 log entries, and I had neglected to close the log afterwards.
5004 30. The text of an SMTP error response that was received during a remote
5005 delivery was being truncated at 512 bytes. This is too short for some of
5006 the long messages that one sometimes sees. I've increased the limit to
5009 31. It is now possible to make retry rules that apply only when a message has a
5010 specific sender, in particular, an empty sender.
5012 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5013 it possible to be selective about when SMTP synchronization is enforced.
5015 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5017 32. Implemented hosts_connection_nolog.
5019 33. Added an ACL for QUIT.
5021 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5024 35. Added mailbox_size and mailbox_filecount to appendfile.
5026 36. Added control = no_multiline_responses to ACLs.
5028 37. There was a bug in the logic of the code that waits for the clock to tick
5029 in the case where the clock went backwards by a substantial amount such
5030 that the microsecond fraction of "now" was more than the microsecond
5031 fraction of "then" (but the whole seconds number was less).
5033 38. Added support for the libradius Radius client library this is found on
5034 FreeBSD (previously only the radiusclient library was supported).
5040 1. When certain lookups returned multiple values in the form name=value, the
5041 quoting of the values was not always being done properly. Specifically:
5042 (a) If the value started with a double quote, but contained no whitespace,
5044 (b) If the value contained whitespace other than a space character (i.e.
5045 tabs or newlines or carriage returns) it was not quoted.
5046 This fix has been applied to the mysql and pgsql lookups by writing a
5047 separate quoting function and calling it from the lookup code. The fix
5048 should probably also be applied to nisplus, ibase and oracle lookups, but
5049 since I cannot test any of those, I have not disturbed their existing code.
5051 2. A hit in the callout cache for a specific address caused a log line with no
5052 reason for rejecting RCPT. Now it says "Previous (cached) callout
5053 verification failure".
5055 3. There was an off-by-one bug in the queryprogram router. An over-long
5056 return line was truncated at 256 instead of 255 characters, thereby
5057 overflowing its buffer with the terminating zero. As well as fixing this, I
5058 have increased the buffer size to 1024 (and made a note to document this).
5060 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5061 when Exim is waiting for an SMTP response from a remote server, Exim
5062 restarts its select() call on the socket, thereby resetting its timeout.
5063 This is not a problem when such interrupts are rare. Somebody set up a cron
5064 job to run exiwhat every 2 minutes, which is less than the normal select()
5065 timeout (5 or 10 minutes). This meant that the select() timeout never
5066 kicked in because it was always reset. I have fixed this by comparing the
5067 time when an interrupt arrives with the time at the start of the first call
5068 to select(). If more time than the timeout has elapsed, the interrupt is
5069 treated as a timeout.
5071 5. Some internal re-factoring in preparation for the addition of Sieve
5072 extensions (by MH). In particular, the "personal" test is moved to a
5073 separate function, and given an option for scanning Cc: and Bcc: (which is
5074 not set for Exim filters).
5076 6. When Exim created an email address using the login of the caller as the
5077 local part (e.g. when creating a From: or Sender: header line), it was not
5078 quoting the local part when it contained special characters such as @.
5080 7. Installed new OpenBSD configuration files.
5082 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5083 try to make them clearer.
5085 9. Callout options, other than the timeout value, were being ignored when
5086 verifying sender addresses in header lines. For example, when using
5088 verify = header_sender/callout=no_cache
5090 the cache was (incorrectly) being used.
5092 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5093 only the Cygwin environment.
5095 11. When return_path_on_delivery was set as a log selector, if different remote
5096 addresses in the same message used different return paths and parallel
5097 remote delivery occurred, the wrong values would sometimes be logged.
5098 (Whenever a remote delivery process finished, the return path value from
5099 the most recently started remote delivery process was logged.)
5101 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5102 lines when AUTH and/or TLS are in use. This is the "received protocol"
5103 field. Exim used to use "asmtp" for authenticated SMTP, without any
5104 indication (in the protocol name) for TLS use. Now it follows the RFC and
5105 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5106 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5107 names appear in log lines as well as in Received: header lines.
5109 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5110 extensions, and comparison tests, and to fix some bugs.
5112 14. Changes to the "personal" filter test:
5114 (1) The test was buggy in that it was just doing the equivalent of
5115 "contains" tests on header lines. For example, if a user's address was
5116 anne@some.where, the "personal" test would incorrectly be true for
5118 To: susanne@some.where
5120 This test is now done by extracting each address from the header in turn,
5121 and checking the entire address. Other tests that are part of "personal"
5122 are now done using regular expressions (for example, to check local parts
5123 of addresses in From: header lines).
5125 (2) The list of non-personal local parts in From: addresses has been
5126 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5127 taken from the Sieve specification recommendations.
5129 (3) If the message contains any header line starting with "List-" it is
5130 treated as non-personal.
5132 (4) The test for "circular" in the Subject: header line has been removed
5133 because it now seems ill-conceived.
5135 15. Minor typos in src/EDITME comments corrected.
5137 16. Installed latest exipick from John Jetmore.
5139 17. If headers_add on a router specified a text string that was too long for
5140 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5141 of string_sprintf() is now avoided.
5143 18. $message_body_size was not set (it was always zero) when running the DATA
5144 ACL and the local_scan() function.
5146 19. For the "mail" command in an Exim filter, no default was being set for
5147 the once_repeat time, causing a random time value to be used if "once" was
5148 specified. (If the value happened to be <= 0, no repeat happened.) The
5149 default is now 0s, meaning "never repeat". The "vacation" command was OK
5150 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5151 (I found it when inspecting the code).
5153 20. There is now an overall timeout for performing a callout verification. It
5154 defaults to 4 times the callout timeout, which applies to individual SMTP
5155 commands during the callout. The overall timeout applies when there is more
5156 than one host that can be tried. The timeout is checked before trying the
5157 next host. This prevents very long delays if there are a large number of
5158 hosts and all are timing out (e.g. when the network connections are timing
5159 out). The value of the overall timeout can be changed by specifying an
5160 additional sub-option for "callout", called "maxwait". For example:
5162 verify = sender/callout=5s,maxwait=20s
5164 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5165 to the end before writing, but this should make it even safer).
5167 22. Exim was forgetting that it had advertised PIPELINING for the second and
5168 subsequent messages on an SMTP connection. It was also not resetting its
5169 memory on STARTTLS and an internal HELO.
5171 23. When Exim logs an SMTP synchronization error within a session, it now
5172 records whether PIPELINING has been advertised or not.
5174 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5175 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5176 rather than long int.
5178 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5180 26. Added the never_mail option to autoreply.
5186 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5187 crash if the getsockname() call failed; this can happen if a connection is
5188 closed very soon after it is established. The problem was simply in the
5189 order in which certain operations were done, causing Exim to try to write
5190 to the SMTP stream before it had set up the file descriptor. The bug has
5191 been fixed by making things happen in the correct order.
5197 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5198 before the connection was closed, thus losing the rejection response.
5200 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5201 some early Solaris releases, but causes trouble in current releases where
5202 socklen_t is defined.
5204 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5207 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5210 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5211 that releasing the top of it at the end releases what was used for sub-
5212 expansions (unless the block got too big). However, discard this block if
5213 the first thing is a variable or header, so that we can use its block when
5214 it is dynamic (useful for very large $message_headers, for example).
5216 6. Lookups now cache *every* query, not just the most recent. A new, separate
5217 store pool is used for this. It can be recovered when all lookup caches are
5218 flushed. Lookups now release memory at the end of their result strings.
5219 This has involved some general refactoring of the lookup sources.
5221 7. Some code has been added to the store_xxx() functions to reduce the amount
5222 of flapping under certain conditions.
5224 8. log_incoming_interface used to affect only the <= reception log lines. Now
5225 it causes the local interface and port to be added to several more SMTP log
5226 lines, for example "SMTP connection from", and rejection lines.
5228 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5230 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5232 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5233 could be overwritten at the end of the current message (or the start of a
5234 new message if it was set in a HELO ACL). The value is now preserved for
5235 the duration of the SMTP connection.
5237 12. If a transport had a headers_rewrite setting, and a matching header line
5238 contained an unqualified address, that address was qualified, even if it
5239 did not match any rewriting rules. The underlying bug was that the values
5240 of the flags that permit the existence of unqualified sender and recipient
5241 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5242 non-local messages, and by -bnq for local messages) were not being
5243 preserved with the message after it was received.
5245 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5246 "next input=" as part of the text comprising the host identity instead of
5247 the correct text. The code was using the same buffer for two different
5248 strings. However, depending on which order the printing function evaluated
5249 its arguments, the bug did not always show up. Under Linux, for example, my
5250 test suite worked just fine.
5252 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5253 doesn't work with some older versions of Perl. It has been changed to "my",
5254 which in any case is probably the better facility to use.
5256 15. A really picky compiler found some instances of statements for creating
5257 error messages that either had too many or two few arguments for the format
5260 16. The size of the buffer for calls to the DNS resolver has been increased
5261 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5262 for addresses that have a lot of PTR records. This alleviates a problem; it
5263 does not fully solve it.
5265 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5266 buffer now truncates the list and logs the incident, which is the same
5267 action as happens when Exim is looking up a host name and its aliases.
5268 Previously in this situation something unpredictable would happen;
5269 sometimes it was "internal error: store_reset failed".
5271 18. If a server dropped the connection unexpectedly when an Exim client was
5272 using GnuTLS and trying to read a response, the client delivery process
5273 crashed while trying to generate an error log message.
5275 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5276 string, for example:
5278 warn message = H1: something\nH2: something
5280 the text was added as a single header line from Exim's point of view
5281 though it ended up OK in the delivered message. However, searching for the
5282 second and subsequent header lines using $h_h2: did not work. This has been
5283 fixed. Similarly, if a system filter added multiple headers in this way,
5284 the routers could not see them.
5286 20. Expanded the error message when iplsearch is called with an invalid key to
5287 suggest using net-iplsearch in a host list.
5289 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5290 ACLs are no longer actually imposed (and a message to that effect is
5293 22. If a "gecos" field in a passwd entry contained escaped characters, in
5294 particular, if it contained a \" sequence, Exim got it wrong when building
5295 a From: or a Sender: header from that name. A second bug also caused
5296 incorrect handling when an unquoted " was present following a character
5297 that needed quoting.
5299 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5300 was not being matched caselessly.
5302 24. Arranged for all hyphens in the exim.8 source to be escaped with
5305 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5306 properly. Recipient callout cache records were still being keyed to include
5307 the sender, even when use_sender was set false. This led to far more
5308 callouts that were necessary. The sender is no longer included in the key
5309 when use_sender is false.
5311 26. Added "control = submission" modifier to ACLs.
5313 27. Added the ${base62d: operator to decode base 62 numbers.
5315 28. dnsdb lookups can now access SRV records.
5317 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5318 the configuration file.
5320 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5321 (-v) mode. This makes the output for a verbose queue run more intelligible.
5323 31. Added a use_postmaster feature to recipient callouts.
5325 32. Added the $body_zerocount variable, containing the number of binary zero
5326 bytes in the message body.
5328 33. The time of last modification of the "new" subdirectory is now used as the
5329 "mailbox time last read" when there is a quota error for a maildir
5332 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5334 35. Added +ignore_unknown as a special item in host lists.
5336 36. Code for decoding IPv6 addresses in host lists is now included, even if
5337 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5338 address was recognized as an IP address, but was then not correctly decoded
5339 into binary, causing unexpected and incorrect effects when compared with
5346 1. Very minor rewording of debugging text in manualroute to say "list of
5347 hosts" instead of "hostlist".
5349 2. If verify=header_syntax was set, and a header line with an unqualified
5350 address (no domain) and a large number of spaces between the end of the
5351 name and the colon was received, the reception process suffered a buffer
5352 overflow, and (when I tested it) crashed. This was caused by some obsolete
5353 code that should have been removed. The fix is to remove it!
5355 3. When running in the test harness, delay a bit after writing a bounce
5356 message to get a bit more predictability in the log output.
5358 4. Added a call to search_tidyup() just before forking a reception process. In
5359 theory, someone could use a lookup in the expansion of smtp_accept_max_
5360 per_host which, without the tidyup, could leave open a database connection.
5362 5. Added the variables $recipient_data and $sender_data which get set from a
5363 lookup success in an ACL "recipients" or "senders" condition, or a router
5364 "senders" option, similar to $domain_data and $local_part_data.
5366 6. Moved the writing of debug_print from before to after the "senders" test
5369 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5370 problems for message scanning, either using a data ACL, or using
5371 local_scan() because the Received: header was not generated till after they
5372 were called (in order to set the time as the time of reception completion).
5373 I have revised the way this works. The header is now generated after the
5374 body is received, but before the ACL or local_scan() are called. After they
5375 are run, the timestamp in the header is updated.
5381 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5382 before starting a queue runner without re-exec. This happened only when
5383 deliver_drop_privilege was set or when the Exim user was set to root. The
5384 effect of the bug was that timeouts during subsequent deliveries caused
5385 crashes instead of being properly handled. The handler is now left at its
5386 default (and expected) setting.
5388 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5389 message, again when deliver_drop_privilege is set or Exim is run as root.
5390 The bug described in (1) was not present in this case, but the tidying up
5391 of the other signals was missing. I have made the two cases consistent.
5393 3. The ignore_target_hosts setting on a manualroute router was being ignored
5394 for hosts that were looked up using the /MX notation.
5396 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5399 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5400 operated on the sender address. After changing the $sender_address to <>
5401 for the sender address verify, Exim was re-instated it as the original
5402 (before rewriting) address, but remembering that it had rewritten it, so it
5403 wasn't rewriting it again. This bug also had the effect of breaking the
5404 sender address verification caching when the sender address was rewritten.
5406 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5407 This has been changed so that if the ip literal address matches
5408 ignore_target_hosts, the router declines.
5410 7. Added expansion conditions match_domain, match_address, and match_local_
5411 part (NOT match_host).
5413 8. The placeholder for the Received: header didn't have a length field set.
5415 9. Added code to Exim itself and to exim_lock to test for a specific race
5416 condition that could lead to file corruption when using MBX delivery. The
5417 issue is with the lockfile that is created in /tmp. If this file is removed
5418 after a process has opened it but before that process has acquired a lock,
5419 there is the potential for a second process to recreate the file and also
5420 acquire a lock. This could lead to two Exim processes writing to the file
5421 at the same time. The added code performs the same test as UW imapd; it
5422 checks after acquiring the lock that its file descriptor still refers to
5423 the same named file.
5425 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5426 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5429 11. Added the smtp_active_hostname option. If used, this will typically be made
5430 to depend on the incoming interface address. Because $interface_address is
5431 not set up until the daemon has forked a reception process, error responses
5432 that can happen earlier (such as "too many connections") no longer contain
5435 12. If an expansion in a condition on a "warn" statement fails because a lookup
5436 defers, the "warn" statement is abandoned, and the next ACL statement is
5437 processed. Previously this caused the whole ACL to be aborted.
5439 13. Added the iplsearch lookup type.
5441 14. Added ident_timeout as a log selector.
5443 15. Added tls_certificate_verified as a log selector.
5445 16. Added a global option tls_require_ciphers (compare the smtp transport
5446 option of the same name). This controls incoming TLS connections.
5448 17. I finally figured out how to make tls_require_ciphers do a similar thing
5449 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5450 before starting the TLS session.
5452 18. Tabs are now shown as \t in -bP output.
5454 19. If the log selector return_path_on_delivery was set, Exim crashed when
5455 bouncing a message because it had too many Received: header lines.
5457 20. If two routers both had headers_remove settings, and the first one included
5458 a superfluous trailing colon, the final name in the first list and the
5459 first name in the second list were incorrectly joined into one item (with a
5460 colon in the middle).
5466 1. Added -C and -D options to the exinext utility, mainly to make it easier
5467 to include in the automated testing, but these could be helpful when
5468 multiple configurations are in use.
5470 2. The exinext utility was not formatting the output nicely when there was
5471 an alternate port involved in the retry record key, nor when there was a
5472 message id as well (for retries that were specific to a specific message
5473 and a specific host). It was also confused by IPv6 addresses, because of
5474 the additional colons they contain. I have fixed the IPv4 problem, and
5475 patched it up to do a reasonable job for IPv6.
5477 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5478 delivery, the log line now contains "pipelined" if PIPELINING was used.
5480 4. An SMTP transport process used to panic and die if the bind() call to set
5481 an explicit outgoing interface failed. This has been changed; it is now
5482 treated in the same way as a connect() failure.
5484 5. A reference to $sender_host_name in the part of a conditional expansion
5485 that was being skipped was still causing a DNS lookup. This no longer
5488 6. The def: expansion condition was not recognizing references to header lines
5489 that used bh_ and bheader_.
5491 7. Added the _cache feature to named lists.
5493 8. The code for checking quota_filecount in the appendfile transport was
5494 allowing one more file than it should have been.
5496 9. For compatibility with Sendmail, the command line option
5504 and sets the incoming protocol and host name (for trusted callers). The
5505 host name and its colon can be omitted when only the protocol is to be set.
5506 Note the Exim already has two private options, -pd and -ps, that refer to
5507 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5508 "s", but I don't think that's a major issue.
5510 10. A number of refactoring changes to the code, none of which should affect
5513 (a) The number of logging options was getting close to filling up the
5514 32-bit word that was used as a bit map. I have split them into two classes:
5515 those that are passed in the argument to log_write(), and those that are
5516 only ever tested independently outside of that function. These are now in
5517 separate 32-bit words, so there is plenty of room for expansion again.
5518 There is no change in the user interface or the logging behaviour.
5520 (b) When building, for example, log lines, the code previously used a
5521 macro that called string_cat() twice, in order to add two strings. This is
5522 not really sufficiently general. Furthermore, there was one instance where
5523 it was actually wrong because one of the argument was used twice, and in
5524 one call a function was used. (As it happened, calling the function twice
5525 did not affect the overall behaviour.) The macro has been replaced by a
5526 function that can join an arbitrary number of extra strings onto a growing
5529 (c) The code for expansion conditions now uses a table and a binary chop
5530 instead of a serial search (which was left over from when there were very
5531 few conditions). Also, it now recognizes conditions like "pam" even when
5532 the relevant support is not compiled in: a suitably worded error message is
5533 given if an attempt is made to use such a condition.
5535 11. Added ${time_interval:xxxxx}.
5537 12. A bug was causing one of the ddress fields not to be passed back correctly
5538 from remote delivery subprocesses. The field in question was not being
5539 subsequently used, so this caused to problems in practice.
5541 13. Added new log selectors queue_time and deliver_time.
5543 14. Might have fixed a bug in maildirsizefile handling that threw up
5544 "unexpected character" debug warnings, and recalculated the data
5545 unnecessarily. In any case, I expanded the warning message to give more
5548 15. Added the message "Restricted characters in address" to the statements in
5549 the default ACL that block characters like @ and % in local parts.
5551 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5552 Three changes have been made:
5554 (a) There was a serious bug; a negative response to MAIL caused the whole
5555 recipient domain to be cached as invalid, thereby blocking all messages
5556 to all local parts at the same domain, from all senders. This bug has
5557 been fixed. The domain is no longer cached after a negative response to
5558 MAIL if the sender used is not empty.
5560 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5563 (c) A new callout option, "use_sender" has been added for people who want
5564 the modified behaviour.
5570 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5573 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5574 indeed breaks things for older releases.
5576 3. Added additional logging to the case where there is a problem reading data
5577 from a filter that is running in a subprocess using a pipe, in order to
5578 try to track down a specific problem.
5580 4. Testing facility fudge: when running in the test harness and attempting
5581 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5582 getting "No route to host". Convert this to a timeout.
5584 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5587 6. Some OS don't have socklen_t but use size_t instead. This affects the
5588 fifth argument of getsockopt() amongst other things. This is now
5589 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5590 can be set for individual OS. I have set it for SunOS5, OSF1, and
5591 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5592 some earlier ones do not.
5594 7. Change 4.30/15 was not doing the test caselessly.
5596 8. The standard form for an IPv6 address literal was being rejected by address
5597 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5598 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5599 this, as well as the form without the "IPv6" on the front (but only when
5600 address literals are enabled, of course).
5602 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5604 10. Exim crashed if a message with an empty sender address specified by -f
5605 encountered a router with an errors_to setting. This could be provoked only
5606 by a command such as
5610 where an empty string was supplied; "<>" did not hit this bug.
5612 11. Installed PCRE release 4.5.
5614 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5615 remained set. It is now erased.
5617 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5618 times from message ids (which are base 36 rather than the normal 62).
5620 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5621 were being counted as actual protocol errors, and logged if the log
5622 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5623 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5624 and DATA following a set of rejected RCPTs do not count as protocol errors.
5625 In other words, Exim assumes they were pipelined, though this may not
5626 actually be the case. Of course, in all cases the client gets an
5627 appropriate error code.
5629 15. If a lookup fails in an ACL condition, a message about the failure may
5630 be available; it is used if testing the ACL cannot continue, because most
5631 such messages specify what the cause of the deferral is. However, some
5632 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5633 that caused an old message to be retained and used if a later statement
5634 caused a defer, replacing the real cause of the deferral.
5636 16. If an IP address had so many PTR records that the DNS lookup buffer
5637 was not large enough to hold them, Exim could crash while trying to process
5638 the truncated data. It now detects and logs this case.
5640 17. Further to 4.21/58, another change has been made: if (and only if) the
5641 first line of a message (the first header line) ends with CRLF, a bare LF
5642 in a subsequent header line has a space inserted after it, so as not to
5643 terminate the header.
5645 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5646 unnecessarily, used atoi() instead of strtol(), and didn't check the
5647 termination when getting file sizes from file names by regex.
5649 19. Completely re-implemented the support for maildirsize files, in the light
5650 of a number of problems with the previous contributed implementation
5651 (4.30/29). In particular:
5653 . If the quota is zero, the maildirsize file is maintained, but no quota is
5656 . If the maildir directory does not exist, it is created before any attempt
5657 to write a maildirsize file.
5659 . The quota value in the file is just a cache; if the quota is changed in
5660 the transport, the new value overrides.
5662 . A regular expression is available for excluding directories from the
5665 20. The autoreply transport checks the characters in options that define the
5666 message's headers; it allows continued headers, but it was checking with
5667 isspace() after an embedded newline instead of explicitly looking for a
5670 21. If all the "regular" hosts to which an address was routed had passed their
5671 expiry times, and had not reached their retry times, the address was
5672 bounced, even if fallback hosts were defined. Now Exim should go on to try
5675 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5676 equivalent code in the SMTP transport. Some hosts send humungous responses
5677 to HELO/EHLO, more than 1024 it seems.
5679 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5680 gives compiler warnings in some environments. I've now done it "properly",
5683 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5684 (because of problems with the built-in one) was declared to return uschar *
5685 instead of char *, causing compiler failure.
5687 25. Fixed a file descriptor leak when processing alias/forward files.
5689 26. Fixed a minor format string issue in dbfn.c.
5691 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5693 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5694 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5697 29. When a sender address is verified, it is cached, to save repeating the test
5698 when there is more than one recipient in a message. However, when the
5699 verification involves a callout, it is possible for different callout
5700 options to be set for different recipients. It is too complicated to keep
5701 track of this in the cache, so now Exim always runs a verification when a
5702 callout is required, relying on the callout cache for the optimization.
5703 The overhead is duplication of the address routing, but this should not be
5706 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5707 to be verified with callout=postmaster, and the main callout worked but the
5708 postmaster check failed, the verification correctly failed. However, if a
5709 subsequent RCPT command asked for sender verification *without* the
5710 postmaster check, incorrect caching caused this verification also to fail,
5713 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5714 it was not caching the DNS options (qualify_single, search_parents) that
5715 were used when the lookup failed. A subsequent lookup with different
5716 options therefore always gave the same answer, though there were cases
5717 where it should not have. (Example: a "domains = !$mx_any" option on a
5718 dnslookup router: the "domains" option is always processed without any
5719 widening, but the router might have qualify_single set.) Now Exim uses the
5720 cached value only when the same options are set.
5722 32. Added John Jetmore's "exipick" utility to the distribution.
5724 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5725 than a timeout (e.g. a certificate is required, and is not provided), an
5726 Exim server now closes the connection immediately. Previously it waited for
5727 the client to close - but if the client is SSL, it seems that they each
5728 wait for each other, leading to a delay before one of them times out.
5730 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5731 maintained 0.8.x compatibility because I don't think many are using it, and
5732 it is clearly obsolete.
5734 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5737 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5738 client certificate was expired. A simple patch fixes this, though I don't
5739 understand the full logic of why the verify callback is called multiple
5742 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5743 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5744 which causes problems with some clients (such as the Certicom SSL Plus
5745 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5746 disables the coutermeasure allowing Eudora to connect."
5748 38. Exim was not checking that a write() to a log file succeeded. This could
5749 lead to Bad Things if a log got too big, in particular if it hit a file
5750 size limit. Exim now panics and dies if it cannot write to a log file, just
5751 as it does if it cannot open a log file.
5753 39. Modified OS/Makefile-Linux so that it now contains
5755 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5757 The two -D definitions ensure that Exim is compiled with large file
5758 support, which makes it possible to handle log files that are bigger than
5761 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5762 instance) a domain was checked against a named list that involved a lookup,
5763 causing $domain_data to be set, then another domain was checked against the
5764 same list, then the first domain was re-checked, the value of $domain_data
5765 after the final check could be wrong. In particular, if the second check
5766 failed, it could be set empty. This bug probably also applied to
5769 41. The strip_trailing_dot option was not being applied to the address given
5770 with the -f command-line option.
5772 42. The code for reading a message's header from the spool was incrementing
5773 $received_count, but never initializing it. This meant that the value was
5774 incorrect (doubled) while delivering a message in the same process in which
5775 it was received. In the most common configuration of Exim, this never
5776 happens - a fresh exec is done - but it can happen when
5777 deliver_drop_privilege is set.
5779 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5780 it now includes up to 150 characters of the unexpected data in the log
5783 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5784 and building data strings. The size of both of these buffers was 10 000
5785 bytes - far larger than anybody would *ever* want, thought I. Needless to
5786 say, somebody hit the limit. I have increased the maximum line length to
5787 20 000 and the maximum data length of concatenated lines to 100 000. I have
5788 also fixed two bugs, because there was no checking on these buffers. Tsk,
5789 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5790 buffer is too small.
5792 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5793 lsearch lookups. Now it does.
5795 46. When parsing a route_list item in a manualroute router, a fixed-length
5796 buffer was used for the list of hosts. I made this 1024 bytes long,
5797 thinking that nobody would ever have a list of hosts that long. Wrong.
5798 Somebody had a whole pile of complicated expansion conditions, and the
5799 string was silently truncated, leading to an expansion error. It turns out
5800 that it is easier to change to an unlimited length (owing to other changes
5801 that have happened since this code was originally written) than to build
5802 structure for giving a limitation error. The length of the item that
5803 expands into the list of hosts is now unlimited.
5805 47. The lsearch lookup could not handle data where the length of text line was
5806 more than 4095 characters. Such lines were truncated, leading to shortened
5807 data being returned. It should now handle lines of any length.
5809 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5810 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5813 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5814 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5815 no longer happens in comment lines. A list of replacements is now placed
5816 at the head of all of the source files, except those whose only change is
5817 to replace PERL_COMMAND in the very first #! line.
5819 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5820 messages on the queue, with a bottom-up merge sort, using code contributed
5821 by Michael Haardt. This should make operations like -bp somewhat faster on
5822 large queues. It won't affect queue runners, except when queue_run_in_order
5825 51. Installed eximstats 1.31 in the distribution.
5827 52. Added support for SRV lookups to the dnslookup router.
5829 53. If an ACL referred to $message_body or $message_body_end, the value was not
5830 reset for any messages that followed in the same SMTP session.
5832 54. The store-handling optimization for building very long strings was not
5833 differentiating between the different store pools. I don't think this
5834 actually made any difference in practice, but I've tidied it.
5836 55. While running the routers to verify a sender address, $sender_address
5837 was still set to the sender address. This is wrong, because when routing to
5838 send a bounce to the sender, it would be empty. Therefore, I have changed
5839 it so that, while verifying a sender address, $sender_address is set to <>.
5840 (There is no change to what happens when verifying a recipient address.)
5842 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5843 target A or AAAA records (if not already returned) without resetting the
5844 qualify_single or search_parents options of the DNS resolver. These are
5845 inappropriate in this case because the targets of MX and SRV records must
5846 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5847 target that, when qualified, matched something in the local domain. These
5848 two options are now turned off when doing these lookups.
5850 57. It seems that at least some releases of Reiserfs (which does not have the
5851 concept of a fixed number of inodes) returns zero and not -1 for the
5852 number of available inodes. This interacted badly with check_spool_inodes,
5853 which assumed that -1 was the "no such thing" setting. What I have done is
5854 to check that the total number of inodes is greater than zero before doing
5855 the test of how many are available.
5857 58. When a "warn" ACL statement has a log_message modifier, the message is
5858 remembered, and not repeated. This is to avoid a lot of repetition when a
5859 message has many recipients that cause the same warning to be written.
5860 However, Exim was preserving the list of already written lines for an
5861 entire SMTP session, which doesn't seem right. The memory is now reset if a
5862 new message is started.
5864 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5865 debugging output unless log_rewrite was also set.
5867 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5868 of (int)(handle) when we know that handle contains (void *)(-1).
5870 61. The Exim daemon panic-logs an error return when it closes the incoming
5871 connection. However "connection reset by peer" seems to be common, and
5872 isn't really an error worthy of noting specially, so that particular error
5875 62. When Exim is trying to find all the local interfaces, it used to panic and
5876 die if the ioctl to get the interface flags failed. However, it seems that
5877 on at least one OS (Solaris 9) it is possible to have an interface that is
5878 included in the list of interfaces, but for which you get a failure error
5879 for this call. This happens when the interface is not "plumbed" into a
5880 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5881 failure of the "get flags" call assumes that the interface is down.
5883 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5884 makes life easier for people who are doing arithmetic on fields extracted
5885 from dates, where you often get leading zeros that should not be
5886 interpreted as octal.
5888 64. Added qualify_domain to the redirect router, to override the global
5891 65. If a pathologically long header line contained very many addresses (the
5892 report of this problem mentioned 10 000) and each of them was rewritten,
5893 Exim could use up a very large amount of memory. (It kept on making new
5894 copies of the header line as it rewrote, and never released the old ones.)
5895 At the expense of a bit more processing, the header rewriting function has
5896 been changed so that it no longer eats memory in this way.
5898 66. The generation of the Received: header has been moved from the time that a
5899 message starts to be received, to the time that it finishes. The timestamp
5900 in the Received: header should now be very close to that of the <= log
5901 line. There are two side-effects of this change:
5903 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5904 logged header lines no longer include the local Received: line, because
5905 it has not yet been created. The same applies to a copy of the message
5906 that is returned to a non-SMTP sender when a message is rejected.
5908 (b) When a filter file is tested using -bf, no additional Received: header
5909 is added to the test message. After some thought, I decided that this
5912 This change does not affect the value of $received_for. It is still set
5913 after address rewriting, but before local_scan() is called.
5915 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5917 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5918 gave an unhelpful panic error message, and a defer error. I have managed to
5919 change this behaviour so that it now rejects any supplied certificate,
5920 which seems right, as the list of acceptable certificates is empty.
5922 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5923 gave an unhelpful defer error. I have not managed to make this reject any
5924 supplied certificates, but the error message it gives is "no certificate
5925 supplied", which is not helpful.
5927 70. exigrep's output now also includes lines that are not associated with any
5928 message, but which match the given pattern. Implemented by a patch from
5929 Martin Sluka, which also tidied up the Perl a bit.
5931 71. Recipient callout verification, like sender verification, was using <> in
5932 the MAIL FROM command. This isn't really the right thing, since the actual
5933 sender may affect whether the remote host accepts the recipient or not. I
5934 have changed it to use the actual sender in the callout; this means that
5935 the cache record is now keyed on a recipient/sender pair, not just the
5936 recipient address. There doesn't seem to be a real danger of callout loops,
5937 since a callout by the remote host to check the sender would use <>.
5938 [SEE ABOVE: changed after hitting problems.]
5940 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5941 temporary errors. However, in the case of such a code being given after
5942 the end of a data transmission (i.e. after ".") Exim was failing to write
5943 a retry record for the message. (Yes, there was some broken host that was
5944 actually sending 8xx at this point.)
5946 73. An unknown lookup type in a host list could cause Exim to panic-die when
5947 the list was checked. (An example that provoked this was putting <; in the
5948 middle of a list instead of at the start.) If this happened during a DATA
5949 ACL check, a -D file could be left lying around. This kind of configuration
5950 error no longer causes Exim to die; instead it causes a defer error. The
5951 incident is still logged to the main and panic logs.
5953 74. Buglet left over from Exim 3 conversion. The message "too many messages
5954 in one connection" was written to the rejectlog but not the mainlog, except
5955 when address rewriting (yes!) was being logged.
5957 75. Added write_rejectlog option.
5959 76. When a system filter was run not as root (that is, when system_filter_user
5960 was set), the values of the $n variables were not being returned to the
5961 main process; thus, they were not subsequently available in the $sn
5964 77. Added +return_path_on_delivery log selector.
5966 78. A connection timeout was being treated differently from recipients deferred
5967 when testing hosts_max_try with a message that was older than the host's
5968 retry timeout. (The host should not be counted, thus allowing all hosts to
5969 be tried at least once before bouncing.) This may have been the cause of an
5970 occasionally reported bug whereby a message would remain on the queue
5971 longer than the retry timeout, but would be bounced if a delivery was
5972 forced. I say "may" because I never totally pinned down the problem;
5973 setting up timeout/retry tests is difficult. See also the next item.
5975 79. The ultimate address timeout was not being applied to errors that involved
5976 a combination of host plus message (for example, a timeout on a MAIL
5977 command). When an address resolved to a number of possible hosts, and they
5978 were not all tried for each delivery (e.g. because of hosts_max_try), a
5979 message could remain on the queue longer than the retry timeout.
5981 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5984 81. Fixed an obscure SMTP outgoing bug which required at least the following
5985 conditions: (a) there was another message waiting for the same server;
5986 (b) the server returned 5xx to all RCPT commands in the first message so
5987 that the message was not completed; (c) the server dropped the connection
5988 or gave a negative response to the RSET that Exim sends to abort the
5989 transaction. The observed case was a dropped connection after DATA that had
5990 been sent in pipelining mode. That is, the server had advertised PIPELINING
5991 but was not implementing it correctly. The effect of the bug was incorrect
5992 behaviour, such as trying another host, and this could lead to a crash.
5998 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5999 and daemon.c were passed as pointers to ints; they should have been
6000 pointers to socklen_t variables (which are typically unsigned ints).
6002 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6005 3. Fixed a really odd bug that affected only the testing scheme; patching a
6006 certain fixed string in the binary changed the value of another string that
6007 happened to be identical to the end of the original first string.
6009 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6010 name", it returns that address as the IP address. On some operating
6011 systems (e.g. Solaris), it also passes back the IP address string as the
6012 "host name". However, on others (e.g. Linux), it passes back an empty
6013 string. Exim wasn't checking for this, and was changing the host name to an
6014 empty string, assuming it had been canonicalized.
6016 5. Although rare, it is permitted to have more than one PTR record for a given
6017 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6018 all the names associated with an address, because they do in Solaris.
6019 However, it seems that they do not in Linux for data that comes from the
6020 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6021 I found this out when I moved to a new Linux workstation and tried to run
6022 the Exim test suite.
6024 To get round this problem I have changed the code so that it now does its
6025 own call to the DNS to look up PTR records when searching for a host name.
6026 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6027 addresses that are only in /etc/hosts are still found.
6029 This behaviour is, however, controlled by an option called host_lookup_
6030 order, which defaults to "bydns:byaddr". If people want to use the other
6031 order, or indeed, just use one or the other means of lookup, they can
6032 specify it in this variable.
6034 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6035 some operating systems, this comes back from gethostbyaddr() as an empty
6036 string, and this is what Exim used to test for. However, it seems that in
6037 other systems, "." is yielded. Exim now tests for this case too.
6039 7. The values of check_spool_space and check_log_space are now held internally
6040 as a number of kilobytes instead of an absolute number of bytes. If a
6041 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6042 kilobyte. This means that much larger values can be stored.
6044 8. Exim monitor: an attempt to get the action menu when not actually pointing
6045 at a message produces an empty menu entitled "No message selected". This
6046 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6047 no entries in it ("Shell widget menu has zero width and/or height"). So I
6048 have added a single, blank menu entry in this case.
6050 9. Added ${quote_local_part.
6052 10. MIME decoding is now applied to the contents of Subject: header lines when
6055 11. Now that a reference to $sender_host_address automatically causes a reverse
6056 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6057 host lookup before query-style lookups in lists that might use this
6058 variable. This has therefore been abolished, and the "net-" prefix is no
6059 longer necessary for query-style lookups.
6061 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6062 have been a typo for LFLAGS, so it has been changed.
6064 13. The install script calls Exim with "-C /dev/null" in order to find the
6065 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6066 to be output. However, since Exim outputs its version number before the
6067 error, it didn't break the script. It just looked ugly. I fixed this by
6068 always allowing "-C /dev/null" if the caller is root.
6070 14. Ignore overlarge ACL variable number when reading spool file - insurance
6071 against a later release with more variables having written the file.
6073 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6074 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6075 this, as well as the form without the "IPv6" on the front.
6077 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6078 OS/Makefile-Darwin file.
6080 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6081 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6083 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6084 variable. However, some OS use ints for inodes. I've added cast to long int
6085 to get rid of the compiler warning.
6087 19. I had forgotten to lock out "/../" in configuration file names when
6088 ALT_CONFIG_PREFIX was set.
6090 20. Routers used for verification do not need to specify transports. However,
6091 if such a router generated a host list, and callout was configured, Exim
6092 crashed, because it could not find a port number from the (non-existent)
6093 transport. It now assumes port 25 in this circumstance.
6095 21. Added the -t option to exigrep.
6097 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6098 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6099 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6100 them is set, LOOKUP_LSEARCH is forced.
6102 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6104 24. Added sender and host information to the "rejected by local_scan()" log
6105 line; previously there was no indication of these.
6107 25. Added .include_if_exists.
6109 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6110 Linux. It turns out that not all file systems support this. Apparently some
6111 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6112 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6113 not supported on the file descriptor, is now ignored when Exim is trying to
6114 sync a directory. This applies only to Linux.
6116 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6118 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6119 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6122 29. Added support for maildirsize files from supplied patch (modified a bit).
6124 30. The use of :fail: followed by an empty string could lead Exim to respond to
6125 sender verification failures with (e.g.):
6127 550 Verification failed for <xxx>
6128 550 Sender verify failed
6130 where the first response line was missing the '-' that indicates it is not
6131 the final line of the response.
6133 31. The loop for finding the name of the user that called Exim had a hardwired
6134 limit of 10; it now uses the value of finduser_retries, which is used for
6135 all other user lookups.
6137 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6140 33. Exim was neglecting to zero errno before one call of strtol() when
6141 expanding a string and expecting an integer value. On some systems this
6142 resulted in spurious "integer overflow" errors. Also, it was casting the
6143 result into an int without checking.
6145 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6146 did not work. The code looks as if it has *never* worked, though it appears
6147 to have been documented since at least release 1.62. I have made it work.
6149 35. The "timeout_DNS" error in retry rules, also documented since at least
6150 1.62, also never worked. As it isn't clear exactly what this means, and
6151 clearly it isn't a major issue, I have abolished the feature by treating it
6152 as "timeout", and writing a warning to the main and panic logs.
6154 36. The display of retry rules for -brt wasn't always showing the error code
6157 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6158 timeout_connect_A, timeout_connect_MX.
6160 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6161 to the empty sender.
6163 39. The daemon was not analysing the content of -oX till after it had closed
6164 stderr and disconnected from the controlling terminal. This meant that any
6165 syntax errors were only noted on the panic log, and the return code from
6166 the command was 0. By re-arranging the code a little, I've made the
6167 decoding happen first, so such errors now appear on stderr, and the return
6168 code is 1. However, the actual setting up of the sockets still happens in
6169 the disconnected process, so errors there are still only recorded on the
6172 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6173 connections (as happens on some IP stacks) was logged at start up time as
6174 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6175 it from "IPv6 and IPv4", which means that two separate sockets are being
6178 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6179 says whether AF_INET or AF_INET6 was passed as an argument.
6181 42. Exiwhat output was messed up when time zones were included in log
6184 43. Exiwhat now gives more information about the daemon's listening ports,
6185 and whether -tls-on-connect was used.
6187 44. The "port" option of the smtp transport is now expanded.
6189 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6190 silently ignored. Now an error message is written to the main and panic
6193 46. There's a new ACL modifier called "logwrite" which writes to a log file
6194 as soon as it is encountered.
6196 47. Added $local_user_uid and $local_user_gid at routing time.
6198 48. Exim crashed when trying to verify a sender address that was being
6201 49. Exim was recognizing only a space character after ".include". It now also
6202 recognizes a tab character.
6204 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6205 extracting the relevant information from the specification. The man page no
6206 longer contains scrambled data for the -d option, and I've added a section
6207 at the front about calling Exim under different names.
6209 51. Added "extra_headers" argument to the "mail" command in filter files.
6211 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6214 53. Installed eximstats 1.29.
6216 54. Added transport_filter_timeout as a generic transport option.
6218 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6219 Cc: header lines. This was required by RFC 822, but it not required by RFC
6222 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6223 incoming messages that did not have them. Now it does so only if the
6224 message originates locally, that is, if there is no associated remote host
6225 address. When Resent- header lines are present, this applies to the Resent-
6226 lines rather than the non-Resent- lines.
6228 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6229 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6231 58. Messages for configuration errors now include the name of the main
6232 configuration file - useful now that there may be more than one file in a
6233 list (.included file names were always shown).
6235 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6236 for those rare installations that do not start the daemon as root or run it
6237 setuid root. I've cut out the call to initgroups() if the daemon is not
6240 60. The Exim user and group can now be bound into the binary as text strings
6241 that are looked up at the start of Exim's processing.
6243 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6245 62. Added $mailstore_basename variable.
6247 63. Installed patch to sieve.c from Michael Haardt.
6249 64. When Exim failed to open the panic log after failing to open the main log,
6250 the original message it was trying to log was written to stderr and debug
6251 output, but if they were not available (the usual case in production), it
6252 was lost. Now it is written to syslog before the two lines that record the
6253 failures to open the logs.
6255 65. Users' Exim filters run in subprocesses under the user's uid. It is
6256 possible for a "deliver" command or an alias in a "personal" command to
6257 provoke an address rewrite. If logging of address rewriting is configured,
6258 this fails because the process is not running as root or exim. There may be
6259 a better way of dealing with this, but for the moment (because 4.30 needs
6260 to be released), I have disabled address rewrite logging when running a
6261 filter in a non-root, non-exim process.
6267 1. The buildconfig auxiliary program wasn't quoting the value set for
6268 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6269 not defined. This bug was masked in 4.22 by the effect that was fixed in
6272 2. Some messages that were rejected after a message id was allocated were
6273 shown as "incomplete" by exigrep. It no longer does this for messages that
6274 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6276 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6277 have allow_domain_literals set, the ID did not get logged in the <= line.
6278 Domain literals are now always recognized in Message-ID: header lines.
6280 4. The first argument for a ${extract expansion item is the key name or field
6281 number. Leading and trailing spaces in this item were not being ignored,
6282 causing some misleading effects.
6284 5. When deliver_drop_privilege was set, single queue runner processes started
6285 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6286 same command in the process it spins off) were not dropping privilege.
6288 6. When the daemon running as "exim" started a queue runner, it always
6289 re-executed Exim in the spun-off process. This is a waste of effort when
6290 deliver_drop_privilege is set. The new process now just calls the
6291 queue-runner function directly.
6297 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6300 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6301 ignored. Though the use of -oP was forcing the writing of a pid file, it
6302 was always written to the default place.
6304 3. If the message "no IP address found for host xxxx" is generated during
6305 incoming verification, it is now followed by identification of the incoming
6306 connection (so you can more easily find what provoked it).
6308 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6310 5. Added some features to "harden" Exim a bit more against certain attacks:
6312 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6313 be put in Local/Makefile. This is like the never_users runtime option,
6314 but it cannot be overridden. The default setting is "root".
6316 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6317 prefix string with which any file named in a -C command line option
6320 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6321 is retained for -C and -D only if the caller of Exim is root. Without
6322 it, the exim user may also use -C and -D and retain privilege.
6324 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6325 command line option is disabled.
6327 6. Macro names set by the -D option must start with an upper case letter, just
6328 like macro names defined in the configuration file.
6330 7. Added "dereference=" facility to LDAP.
6332 8. Two instances of the typo "uknown" in the source files are fixed.
6334 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6335 the Configure-Makefile script screwed up while processing it.
6337 10. Incorporated PCRE 4.4.
6339 11. The SMTP synchronization check was not operating right at the start of an
6340 SMTP session. For example, it could not catch a HELO sent before the client
6341 waited for the greeting. There is now a check for outstanding input at the
6342 point when the greeting is written. Because of the duplex, asynchronous
6343 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6344 way, but not yet received, when the check is performed.
6346 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6347 on TCP/IP sockets, because this apparently causes some broken clients to
6350 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6351 unchanged) from the Cygwin maintainer.
6353 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6354 was supported instead of testing for mbx. Effectively a typo.
6356 15. The spa authenticator server code was not checking that the input it
6357 received was valid base64.
6359 16. The debug output line for the "set" modifier in ACLs was not showing the
6360 name of the variable that was being set.
6362 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6364 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6365 lookup on its own. Something else was needed to trigger the lookup. For
6366 example, a match in host_lookup or the need for a host name in a host list.
6367 Now, if $sender_host_name is referenced and the host name has not yet been
6368 looked up, a lookup is performed. If the lookup fails, the variable remains
6369 empty, and $host_lookup_failed is set to "1".
6371 19. Added "eqi" as a case-independent comparison operator.
6373 20. The saslauthd authentication condition could segfault if neither service
6374 nor realm was specified.
6376 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6377 error message that was logged was misleading, and incoming SMTP
6378 connections were dropped. The message is now more accurate, and temporary
6379 errors are given to SMTP connections.
6381 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6382 (except RSET) with a 421 error, until QUIT is received. However, it was
6383 failing to send a response to QUIT.
6385 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6386 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6387 ACL is now run after the helo_try_verify_hosts code.
6389 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6390 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6391 case-independent, so other case variants are also recognized). Apparently
6392 some systems use these upper case variants.
6394 25. If more than two messages were waiting for the same host, and a transport
6395 filter was specified for the transport, Exim sent two messages over the
6396 same TCP/IP connection, and then failed with "socket operation on non-
6397 socket" when it tried to send the third.
6399 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6401 27. The extern definition of crypt16() in expand.c was not being excluded when
6402 the OS had its own crypt16() function.
6404 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6405 as a preferred synonym for return_size_limit, both as an option and as an
6408 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6410 30. Changed the default configuration ACL to relax the local part checking rule
6411 for addresses that are not in any local domains. For these addresses,
6412 slashes and pipe symbols are allowed within local parts, but the sequence
6413 /../ is explicitly forbidden.
6415 31. SPA server authentication was not clearing the challenge buffer before
6418 32. log_message in a "warn" ACL statement was writing to the reject log as
6419 well as to the main log, which contradicts the documentation and doesn't
6420 seem right (because no rejection is happening). So I have stopped it.
6422 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6423 However, I am unable to do any testing of this.
6425 34. Fixed an infelicity in the appendfile transport. When checking directories
6426 for a mailbox, to see if any needed to be created, it was accidentally
6427 using path names with one or more superfluous leading slashes; tracing
6428 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6430 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6431 contents are added to the log line that is written for every discarded
6432 recipient. (Previously a log_message setting was ignored.)
6434 36. The ${quote: operator now quotes the string if it is empty.
6436 37. The install script runs exim in order to find its version number. If for
6437 some reason other than non-existence or emptiness, which it checks, it
6438 could not run './exim', it was installing it with an empty version number,
6439 i.e. as "exim-". This error state is now caught, and the installation is
6442 38. An argument was missing from the function that creates an error message
6443 when Exim fails to connect to the socket for saslauthd authentication.
6444 This could cause Exim to crash, or give a corrupted message.
6446 39. Added isip, isip4, and isip6 to ${if conditions.
6448 40. The ACL variables $acl_xx are now saved with the message, and can be
6449 accessed later in routers, transports, and filters.
6451 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6452 strings in the file are not string-expanded.
6454 42. If a MAIL command specified a SIZE value that was too large to fit into an
6455 int variable, the check against message_size_limit failed. Such values are
6456 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6457 day this will have to be increased, but I don't think I want to be around
6458 when emails are that large.
6465 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6466 iconv() is not standard in FreeBSD.
6468 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6469 IPv6 enabled. The observed symptom was a segmentation fault on return from
6470 the function os_common_find_running_interfaces() in src/os.c.
6472 3. In the check_special_case() function in daemon.c I had used "errno" as an
6473 argument name, which causes warnings on some systems. This was basically a
6474 typo, since it was named "eno" in the comments!
6476 4. The code that waits for the clock to tick (at a resolution of some fraction
6477 of a second) so as to ensure message-id uniqueness was always waiting for
6478 at least one whole tick, when it could have waited for less. [This is
6479 almost certainly not relevant at current processor speeds, where it is
6480 unlikely to ever wait at all. But we try to future-proof.]
6482 5. The function that sleeps for a time interval that includes fractions of a
6483 second contained a race. It did not block SIGALRM between setting the
6484 timer, and suspending (a couple of lines later). If the interval was short
6485 and the sigsuspend() was delayed until after it had expired, the suspension
6486 never ended. On busy systems this could lead to processes getting stuck for
6489 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6490 process, before it forks any delivery processes. The open lookup caching
6491 mechanism meant that the open file or database connection was passed into
6492 the delivery process. The problem was that delivery processes always tidy
6493 up cached lookup data. This could cause a problem for the next delivery
6494 process started by the queue runner, because the external queue runner
6495 process does not know about the closure. So the next delivery process
6496 still has data in the lookup cache. In the case of a file lookup, there was
6497 no problem because closing a file descriptor in a subprocess doesn't affect
6498 the parent. However, if the lookup was caching a connection to a database,
6499 the connection was closed, and the second delivery process was likely to
6500 see errors such as "PGSQL: query failed: server closed the connection
6501 unexpectedly". The problem has been fixed by closing all cached lookups
6502 in a queue runner before running a delivery process.
6504 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6505 seem to have the "const" qualifier which it has on other OS. I've
6508 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6509 *of the same type* (client or server) with the same public name that an
6510 error should be diagnosed.
6512 9. When Exim looked up a host name for an IP address, but failed to find the
6513 original IP address when looking up the host name (a safety check), it
6514 output the message "<ip address> does not match any IP for NULL", which was
6515 confusing, to say the least. The bug was that the host name should have
6516 appeared instead of "NULL".
6518 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6519 user that is built into the binary, and the -C or -D options is used, root
6520 privilege is dropped before the configuration file is read. In addition,
6521 logging is switched to stderr instead of the normal log files. If the
6522 configuration then re-defines the Exim user, the unprivileged environment
6523 is probably not what is expected, so Exim logs a panic warning message (but
6526 However, if deliver_drop_privilege is set, the unprivileged state may well
6527 be exactly what is intended, so the warning has been cut out in that case,
6528 and Exim is allowed to try to write to its normal log files.
6534 1. smtp_return_error_details was not giving details for temporary sender
6535 or receiver verification errors.
6537 2. Diagnose a configuration error if two authenticators have the same public
6540 3. Exim used not to create the message log file for a message until the first
6541 delivery attempt. This could be confusing when incoming messages were held
6542 for policy or load reasons. The message log file is now created at the time
6543 the message is received, and an initial "Received" line is written to it.
6545 4. The automatically generated man page for command line options had a minor
6546 bug that caused no ill effects; however, a more serious problem was that
6547 the procedure for building the man page automatically didn't always
6548 operate. Consequently, release 4.20 contains an out-of-date version. This
6549 shouldn't happen again.
6551 5. When building Exim with embedded Perl support, the script that builds the
6552 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6553 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6555 6. The freeze_tell option was not being used for messages that were frozen on
6556 arrival, either by an ACL or by local_scan().
6558 7. Added the smtp_incomplete_transaction log selector.
6560 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6561 was accepting AUTH without a new EHLO.
6563 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6564 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6567 10. Logging of TCP/IP connections (when configured) now happens in the main
6568 daemon process instead of the child process, so that the TCP/IP connection
6569 count is more accurate (but it can never be perfect).
6571 11. The use of "drop" in a nested ACL was not being handled correctly in the
6572 outer ACL. Now, if condition failure induced by the nested "drop" causes
6573 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6574 or "require"), the connection is dropped.
6576 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6577 that yield "discard" can now be used with an "accept" or a "discard" verb,
6578 but an error is generated for any others (because I can't see a useful way
6579 to define what should happen).
6581 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6582 are now processed in the same way as lines in the Exim configuration file.
6583 In particular, continuation lines are supported.
6585 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6587 15. Added -ti meaning -t -i.
6589 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6590 domains, and warn by logging if others are found.
6592 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6593 returned from GIFCONF when Exim is trying to find the list of interfaces on
6594 a host. The code in os.c has been modified to copy each ifreq to an aligned
6595 structure in all cases.
6597 Also, in some cases, the returned ifreq's were being copied to a 'struct
6598 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6599 means the last couple of bytes of an IPv6 address could be chopped if the
6600 ifreq contained only a normal sockaddr (14 bytes storage).
6602 18. Named domain lists were not supported in the hosts_treat_as_local option.
6603 An entry such as +xxxx was not recognized, and was treated as a literal
6606 19. Ensure that header lines added by a DATA ACL are included in the reject log
6607 if the ACL subsequently rejects the message.
6609 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6610 MD5 (which is deprecated).
6612 21. When testing a filter file using -bf, Exim was writing a message when it
6613 took the sender from a "From " line in the message, but it was not doing so
6614 when it took $return_path from a Return-Path: header line. It now does.
6616 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6617 with a valid header line field name (a series of printing characters
6618 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6620 23. Changed "disc" in the source to "disk" to conform to the documentation and
6621 the book and for uniformity.
6623 24. Ignore Sendmail's -Ooption=value command line item.
6625 25. When execve() failed while trying to run a command in a pipe transport,
6626 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6627 could be confused with a return value of 69 from the command itself. This
6628 has been changed to 127, the value the shell returns if it is asked to run
6629 a non-existent command. The wording for the related log line suggests a
6630 non-existent command as the problem.
6632 26. If received_header_text expands to an empty string, do not add a Received:
6633 header line to the message. (Well, it adds a token one on the spool, but
6634 marks it "old" so that it doesn't get used or transmitted.)
6636 27. Installed eximstats 1.28 (addition of -nt option).
6638 28. There was no check for failure on the call to getsockname() in the daemon
6639 code. This can fail if there is a shortage of resources on the system, with
6640 ENOMEM, for example. A temporary error is now given on failure.
6642 29. Contrary to the C standard, it seems that in some environments, the
6643 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6644 program. Exim now does this explicitly; it affects the formatting of
6645 timestamps using strftime().
6647 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6648 complaints. I've now initialized all the variables, to avoid this.
6650 32. Header lines added by a system filter were not being "seen" during
6651 transport-time rewrites.
6653 33. The info_callback() function passed to OpenSSL is set up with type void
6654 (*)(SSL *, int, int), as described somewhere. However, when calling the
6655 function (actually a macro) that sets it up, the type void(*)() is
6656 expected. I've put in a cast to prevent warnings from picky compilers.
6658 34. If a DNS black list lookup found a CNAME record, but there were no A
6659 records associated with the domain it pointed at, Exim crashed.
6661 35. If a DNS black list lookup returned more than one A record, Exim ignored
6662 all but the first. It now scans all returned addresses if a particular IP
6663 value is being sought. In this situation, the contents of the
6664 $dnslist_value variable are a list of all the addresses, separated by a
6667 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6668 to accept a host name and all its aliases if the forward lookup for any of
6669 them yielded the IP address of the incoming connection. Now it accepts only
6670 those names whose forward lookup yields the correct IP address. Any other
6671 names are discarded. This closes a loophole whereby a rogue DNS
6672 administrator could create reverse DNS records to break through a
6673 wildcarded host restriction in an ACL.
6675 37. If a user filter or a system filter that ran in a subprocess used any of
6676 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6677 the wrong values were passed to the pipe command ($thisaddress had the
6678 value of $0, $0 had the value of $1, etc). This bug was introduced by
6679 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6681 38. Improved the line breaking for long SMTP error messages from ACLs.
6682 Previously, if there was no break point between 40 and 75 characters, Exim
6683 left the rest of the message alone. Two changes have been made: (a) I've
6684 reduced the minimum length to 35 characters; (b) if it can't find a break
6685 point between 35 and 75 characters, it looks ahead and uses the first one
6686 that it finds. This may give the occasional overlong line, but at least the
6687 remaining text gets split now.
6689 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6690 file descriptors might be low, and that setting 1000 would always raise it.
6691 It turns out that in some environments, the limit is already over 1000 and
6692 that lowering it causes trouble. So now Exim takes care not to decrease it.
6694 40. When delivering a message, the value of $return_path is set to $sender_
6695 address at the start of routing (routers may change the value). By an
6696 oversight, this default was not being set up when an address was tested by
6697 -bt or -bv, which affected the outcome if any router or filter referred to
6700 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6701 log only when "message" or "log_message" are set. However, if one of the
6702 conditions was an address verification, or a call to a nested ACL, the
6703 messages generated by the underlying test were being passed through. This
6704 no longer happens. The underlying message is available in $acl_verify_
6705 message for both "message" and "log_message" expansions, so it can be
6706 passed through if needed.
6708 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6709 new expansion $bh_ to give the encoded byte string without charset
6710 translation. Translation happens only if iconv() is available; HAVE_ICONV
6711 indicates this at build time. HEADERS_CHARSET gives the charset to
6712 translate to; headers_charset can change it in the configuration, and
6713 "headers charset" can change it in an individual filter file.
6715 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6716 that creates RFC 2047 encoded "words" labels them as that charset instead
6717 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6718 expansion operator; (ii) when Exim creates a From: line for a local
6719 message; (iii) when a header line is rewritten to include a "phrase" part.
6721 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6722 buggy, causing it to skip the first lines of messages whose message ID
6723 ended in 'D'. This would not have bitten before Exim release 4.14, because
6724 message IDs were unlikely to end in 'D' before then. The effect was to have
6725 incorrect size information for certain domains.
6727 45. #include "config.h" was missing at the start of the crypt16.c module. This
6728 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6731 46. If there was a timeout during a "random" callout check, Exim treated it as
6732 a failure of the random address, and carried on sending RSET and the real
6733 address. If the delay was just some slowness somewhere, the response to the
6734 original RCPT would be taken as a response to RSET and so on, causing
6735 mayhem of various kinds.
6737 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6738 when I implemented it. It didn't allow for the fact that some option values
6739 may legitimately be negative (e.g. size_addition), and it didn't even do
6740 the right test for positive values.
6742 48. Domain names in DNS records are case-independent. Exim always looks them up
6743 in lower case. Some resolvers return domain names in exactly the case they
6744 appear in the zone file, that is, they may contain uppercase letters. Not
6745 all resolvers do this - some return always lower case. Exim was treating a
6746 change of case by a resolver as a change of domain, similar to a widening
6747 of a domain abbreviation. This triggered its re-routing code and so it was
6748 trying to route what was effectively the same domain again. This normally
6749 caused routing to fail (because the router wouldn't handle the domain
6750 twice). Now Exim checks for this case specially, and just changes the
6751 casing of the domain that it ultimately uses when it transmits the message
6754 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6757 50. If a filter generated a file delivery with a non-absolute name (possible if
6758 no home directory exists for the router), the forbid_file option was not
6761 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6762 the existing equality matching.
6764 52. Exim was using ints instead of ino_t variables in some places where it was
6765 dealing with inode numbers.
6767 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6768 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6769 called TMPDIR, and if it finds it is different, it changes its value.
6771 54. The smtp_printf() function is now made available to local_scan() so
6772 additional output lines can be written before returning. There is also an
6773 smtp_fflush() function to enable the detection of a dropped connection.
6774 The variables smtp_input and smtp_batched_input are exported to
6777 55. Changed the default runtime configuration: the message "Unknown user"
6778 has been removed from the ACL, and instead placed on the localuser router,
6779 using the cannot_route_message feature. This means that any verification
6780 failures that generate their own messages won't get overridden. Similarly,
6781 the "Unrouteable address" message that was in the ACL for unverifiable
6782 relay addresses has also been removed.
6784 56. Added hosts_avoid_esmtp to the smtp transport.
6786 57. The exicyclog script was not checking for the esoteric option
6787 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6788 will work only if exicyclog is run under the appropriate euid.
6790 58. Following a discussion on the list, the rules by which Exim recognises line
6791 endings on incoming messages have been changed. The -dropcr and drop_cr
6792 options are now no-ops, retained only for backwards compatibility. The
6793 following line terminators are recognized: LF CRLF CR. However, special
6794 processing applies to CR:
6796 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6797 nor a local message in the state where . is a terminator.
6799 (ii) If a bare CR is encountered in a header line, an extra space is added
6800 after the line terminator so as not to end the header. The reasoning
6801 behind this is that bare CRs in header lines are most likely either
6802 to be mistakes, or people trying to play silly games.
6804 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6805 was being incorrectly given as 18 bytes larger than it should have been.
6806 This is a VOB (very old bug).
6808 60. This may never have affected anything current, but just in case it has:
6809 When the local host is found other than at the start of a list of hosts,
6810 the local host, those with the same MX, and any that follow, are discarded.
6811 When the list in question was part of a longer list of hosts, the following
6812 hosts (not currently being processed) were also being discarded. This no
6813 longer happens. I'm not sure if this situation could ever has previously
6816 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6819 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6820 header. This is something that is recommended in a new Internet Draft, and
6821 is something that is documented as being done by Sendmail. There are two
6822 possible values. For messages generated by the autoreply transport, Exim
6825 Auto-Submitted: auto-replied
6827 whereas for all other generated messages (e.g. bounces) it adds
6829 Auto-Submitted: auto-generated
6831 63. The "personal" condition in filters now includes a test for the
6832 Auto-Submitted: header. If it contains the string "auto-" the message it
6833 not considered personal.
6835 64. Added rcpt_include_affixes as a generic transport option.
6837 65. Added queue_only_override (default true).
6839 66. Added the syslog_duplication option.
6841 67. If what should have been the first header line of a message consisted of
6842 a space followed by a colon, Exim was mis-interpreting it as a header line.
6843 It isn't of course - it is syntactically invalid and should therefore be
6844 treated as the start of the message body. The misbehaviour could have
6845 caused a number of strange effects, including loss of data in subsequent
6846 header lines, and spool format errors.
6848 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6849 client host had authenticated. This control can now be exercised by an ACL
6850 for more flexibility.
6852 69. By default, callouts do not happen when testing with -bh. There is now a
6853 variant, -bhc, which does actually run the callout code, including
6854 consulting and updating the callout cache.
6856 70. Added support for saslauthd authentication, courtesy of Alexander
6859 71. If statvfs() failed on the spool or log directories while checking their
6860 size for availability, Exim confusingly gave the error "space shortage".
6861 Furthermore, in debugging mode it crashed with a floating point exception.
6862 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6863 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6864 serious problem, Exim now writes to the main and panic logs when this
6865 happens, with details of the failure. It then refuses to accept the
6866 incoming message, giving the message "spool directory problem" or "log
6867 directory problem" with a 421 code for SMTP messages.
6869 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6870 0, 1, and 2 exist, because some OS complain for execs without them (see
6871 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6872 descriptors. However, the code omitted to check that the open succeeded,
6873 causing mysterious errors if for some reason the permissions on /dev/null
6874 got screwed. Now Exim writes a message to the main and panic logs, and
6875 bombs out if it can't open /dev/null.
6877 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6878 interact so that it is all more flexible. It is supposed to remain
6879 backwards compatible. Also added extra_local_interfaces.
6881 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6882 to bomb out with an assertion failure - to the client this appears as a
6883 connection drop. This problem occurs in the part of the code that was taken
6884 from the Samba project. Fortunately, the assertion is in a very simple
6885 function, so I have fixed this by reproducing the function inline in the
6886 one place where it is called, and arranging for authentication to fail
6887 instead of killing the process with assert().
6889 75. The SPA client code was not working when the server requested OEM rather
6890 than Unicode encoding.
6892 76. Added code to make require_files with a specific uid setting more usable in
6893 the case where statting the file as root fails - usually a non-root-mounted
6894 NFS file system. When this happens and the failure is EACCES, Exim now
6895 forks a subprocess and does the per-uid checking as the relevant uid.
6897 77. Added process_log_path.
6899 78. If log_file_path was not explicitly set, a setting of check_log_space or
6900 check_log_inodes was ignored.
6902 79. If a space check for the spool or log partitions fails, the incident is now
6903 logged. Of course, in the latter case the data may get lost...
6905 80. Added the %p formatting code to string_format() so that it can be used to
6906 print addresses in debug_print(). Adjusted all the address printing in the
6907 debugging in store.c to use %p rather than %d.
6909 81. There was a concern that a line of code in smtp_in.c could overflow a
6910 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6911 initially expressed, the concern was not well-founded, because trailing
6912 spaces are removed early. However, if the trailing spaces were followed by
6913 a NULL, they did not get removed, so the overflow was possible. Two fixes
6916 (a) I re-wrote the offending code in a cleaner fashion.
6917 (b) If an incoming SMTP command contains a NULL character, it is rejected
6920 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6921 runs initgroups(), so that if the Exim user is in any additional groups,
6922 they will be used during message reception.
6928 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob