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
172 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
173 matters on fast-turnover and PID-randomising systems, which were getting
174 out-of-order delivery.
176 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
177 a possibly-overlapping copy. The symptom was that "Remote host closed
178 connection in response to HELO" was logged instead of the actual 4xx
181 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
182 Previously only that bufferd was discarded, resulting in SYMTP command
185 JH/32 DKIM: when a message has multiple signatures matching an identity given
186 in dkim_verify_signers, run the dkim acl once for each. Previously only
187 one run was done. Bug 2189.
189 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
190 "panic the current process" to "deliberately defer". The panic log is
191 still written with the problem list name; the mail and reject logs now
192 get a temp-reject line for the message that was being handled, saying
193 something like "domains check lookup or other defer". The SMTP 451
194 message is still "Temporary local problem".
200 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
201 than -2003 did; needs libidn2 in addition to libidn.
203 JH/02 The path option on a pipe transport is now expanded before use.
205 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
206 Patch provided by "Björn", documentation fix added too.
208 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
209 missing a wire-to-host endian conversion.
211 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
212 close after a BDAT command line could be taken as a following command,
213 giving a synch failure. Fix by only checking for synch immediately
214 before acknowledging the chunk.
216 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
217 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
218 macro. Patches provided by Josh Soref.
220 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
221 Previously we did not; the RFC seems ambiguous and VRFY is not listed
222 by IANA as a service extension. However, John Klensin suggests that we
225 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
226 the dkim code may be unix-mode line endings rather than smtp wire-format
227 CRLF, so prepend a CR to any bare LF.
229 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
230 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
232 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
233 the first were themselves being wrongly included in the feed into dkim
234 processing; with most chunk sizes in use this resulted in an incorrect
235 body hash calculated value.
237 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
238 DKIM signature block, for verification. Although advised against by
239 standards it is specifically not ruled illegal.
241 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
243 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
244 missing a body hash (the bh= tag).
246 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
247 It seems that HAProxy sends the Proxy Protocol information in clear and
248 only then does a TLS startup, so do the same.
250 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
251 TCP connections (such as for Spamd) unless the daemon successfully set
252 Fast Open mode on its listening sockets. This fixes breakage seen on
253 too-old kernels or those not configured for Fast Open, at the cost of
254 requiring both directions being enabled for TFO, and TFO never being used
255 by non-daemon-related Exim processes.
257 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
258 endings, at least on the first header line. Try to canonify any that get
259 past that check, despite the cost.
261 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
262 now limited to an arbitrary five deep, while parsing addresses with the
263 strip_excess_angle_brackets option enabled.
265 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
266 instead leave the unprompted TLS handshake in socket buffer for the
267 TLS library to consume.
269 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
271 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
273 JH/16 Drop variables when they go out of scope. Memory management drops a whole
274 region in one operation, for speed, and this leaves assigned pointers
275 dangling. Add checks run only under the testsuite which checks all
276 variables at a store-reset and panics on a dangling pointer; add code
277 explicitly nulling out all the variables discovered. Fixes one known
278 bug: a transport crash, where a dangling pointer for $sending_ip_address
279 originally assigned in a verify callout, is re-used.
281 PP/06 Drop '.' from @INC in various Perl scripts.
283 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
285 PP/08 Reduce a number of compilation warnings under clang; building with
286 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
287 should be warning-free.
289 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
291 HS/01 Fix portability problems introduced by PP/08 for platforms where
292 realloc(NULL) is not equivalent to malloc() [SunOS et al].
294 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
295 chunk. This allows us to accept broken chunked messages. We need a more
296 general solution here.
298 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
299 already-broken messages in the queue.
301 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
303 JH/19 Fix reference counting bug in routing-generated-address tracking.
309 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
310 supports it and a size is available (ie. the sending peer gave us one).
312 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
313 years of being deprecated). The replacements are the ACLs
314 acl_smtp_mime and acl_not_smtp_mime.
316 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
317 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
318 or even an in-clear connection were permitted. Now, if the host lookup
319 was dnssec and dane was requested then the host is only used if the
320 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
321 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
322 if one fails this test.
323 This means that a poorly-configured remote DNS will make it incommunicado;
324 but it protects against a DNS-interception attack on it.
326 JH/04 Bug 1810: make continued-use of an open smtp transport connection
327 non-noisy when a race steals the message being considered.
329 JH/05 If main configuration option tls_certificate is unset, generate a
330 self-signed certificate for inbound TLS connections.
332 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
333 in rewrites and routers.
335 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
336 and logged a warning sing 4.83; now they are a configuration file error.
338 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
339 (lacking @domain). Apply the same qualification processing as RCPT.
341 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
343 JH/10 Support ${sha256:} applied to a string (as well as the previous
346 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
347 a cutthrough deliver is pending, as we always want to make a connection.
348 This also avoids re-routing the message when later placing the cutthrough
349 connection after a verify cache hit.
350 Do not update it with the verify result either.
352 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
353 when routing results in more than one destination address.
355 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
356 signing (which inhibits the cutthrough capability). Previously only
357 the presence of an option was tested; now an expansion evaluating as
358 empty is permissible (obviously it should depend only on data available
359 when the cutthrough connection is made).
361 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
362 the relevant preceding SMTP command did not note the pipelining mode.
364 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
365 Previously they were not counted.
367 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
368 as one having no matching records. Previously we deferred the message
369 that needed the lookup.
371 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
372 distinguished as "(=".
374 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
375 for missing MX records. Previously it only worked for missing A records.
377 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
379 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
380 after the data-go-ahead and data-ack. Patch from Jason Betts.
382 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
383 even for a "none" policy. Patch from Tony Meyer.
385 JH/22 Fix continued use of a connection for further deliveries. If a port was
386 specified by a router, it must also match for the delivery to be
389 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
390 When one of the recipients of a message was unsuitable for the connection
391 (has no matching addresses), we lost track of needing to mark it
392 deferred. As a result mail would be lost.
394 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
396 JH/25 Decoding ACL controls is now done using a binary search; the source code
397 takes up less space and should be simpler to maintain. Merge the ACL
398 condition decode tables also, with similar effect.
400 JH/26 Fix problem with one_time used on a redirect router which returned the
401 parent address unchanged. A retry would see the parent address marked as
402 delivered, so not attempt the (identical) child. As a result mail would
405 JH/27 Fix a possible security hole, wherein a process operating with the Exim
406 UID can gain a root shell. Credit to http://www.halfdog.net/ for
407 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
410 JH/28 Enable {spool,log} filesystem space and inode checks as default.
411 Main config options check_{log,spool}_{inodes,space} are now
412 100 inodes, 10MB unless set otherwise in the configuration.
414 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
415 Previously it only applied to the main-section connection policy
418 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
420 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
421 by me. Added RFC7919 DH primes as an alternative.
423 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
424 are not in the system include path.
426 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
427 GnuTLS, when a session startup failed (eg because the client disconnected)
428 Exim did stdio operations after fclose. This was exposed by a recent
429 change which nulled out the file handle after the fclose.
431 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
432 signed directly by the cert-signing cert, rather than an intermediate
433 OCSP-signing cert. This is the model used by LetsEncrypt.
435 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
437 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
438 an incoming connection.
440 HS/02 Bug 1802: Do not half-close the connection after sending a request
443 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
444 fallback to "prime256v1".
446 JH/34 SECURITY: Use proper copy of DATA command in error message.
447 Could leak key material. Remotely exploitable. CVE-2016-9963.
453 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
454 and 3.4.4 - once the server is enabled to respond to an OCSP request
455 it does even when not requested, resulting in a stapling non-aware
456 client dropping the TLS connection.
458 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
459 support variable-length bit vectors. No functional change.
461 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
462 The I= interface field on outgoing lines is now after the H= remote
463 host field, same as incoming lines. There is a separate
464 outgoing_interface log selector which allows you to disable the
467 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
468 If not running log_selector +smtp_connection the mainlog would be held
469 open indefinitely after a "too many connections" event, including to a
470 deleted file after a log rotate. Leave the per net connection logging
471 leaving it open for efficiency as that will be quickly detected by the
472 check on the next write.
474 HS/01 Bug 1671: Fix post transport crash.
475 Processing the wait-<transport> messages could crash the delivery
476 process if the message IDs didn't exist for some reason. When
477 using 'split_spool_directory=yes' the construction of the spool
478 file name failed already, exposing the same netto behaviour.
480 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
481 mime_regex ACL conditions.
483 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
484 to DSN fail messages (bounces): remote IP, remote greeting, remote response
485 to HELO, local diagnostic string.
487 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
488 log line to debug. Even when configured with a tls authenticator many
489 client connections are expected to not authenticate in this way, so
490 an authenticate fail is not an error.
492 HS/02 Add the Exim version string to the process info. This way exiwhat
493 gives some more detail about the running daemon.
495 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
496 matter for fast-change records such as DNSBLs.
498 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
499 retry record, even if constant. There may be multiple transports with
500 different interface settings and the retry behaviour needs to be kept
503 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
505 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
507 JH/10 Bug 840: fix log_defer_output option of pipe transport
509 JH/11 Bug 830: use same host for all RCPTS of a message, even under
510 hosts_randomize. This matters a lot when combined with mua_wrapper.
512 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
513 ${quote_pgsql:<string>} operator.
515 JH/13 Bug 1708: avoid misaligned access in cached lookup.
517 JH/14 Change header file name for freeradius-client. Relevant if compiling
518 with Radius support; from the Gentoo tree and checked under Fedora.
520 JH/15 Bug 1712: Introduce $prdr_requested flag variable
522 JH/16 Bug 1714: Permit an empty string as expansion result for transport
523 option transport_filter, meaning no filtering.
525 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
527 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
528 defaults to "*" (all hosts). The variable is now available when not built
529 with TLS, default unset, mainly to enable keeping the testsuite sane.
530 If a server certificate is not supplied (via tls_certificate) an error is
531 logged, and clients will find TLS connections fail on startup. Presumably
532 they will retry in-clear.
533 Packagers of Exim are strongly encouraged to create a server certificate
534 at installation time.
536 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
537 with the $config_file variable.
539 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
540 in transport context, after the attempt, and per-recipient. The latter type
541 is per host attempted. The event data is the error message, and the errno
542 information encodes the lookup type (A vs. MX) used for the (first) host,
543 and the trailing two digits of the smtp 4xx response.
545 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
546 to write to mainlog (or rejectlog, paniclog) in the window between file
547 creation and permissions/ownership being changed. Particularly affects
548 installations where exicyclog is run as root, rather than exim user;
549 result is that the running daemon panics and dies.
551 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
553 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
554 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
555 "pri" and "weight". Note that the previous implicit priority given by the
556 list order is no longer honoured.
558 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
561 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
562 by defining SUPPORT_SOCKS.
564 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
565 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
566 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
567 variables are renamed to proxy_{local,external}_{address,port}.
569 JH/27 Move Internationalisation support from Experimental to mainline, enabled
570 for a build by defining SUPPORT_I18N
572 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
573 of the query string, and make ${quote_redis:} do that quoting.
575 JH/29 Move Events support from Experimental to mainline, enabled by default
576 and removable for a build by defining DISABLE_EVENT.
578 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
580 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
581 cached by the daemon.
583 JH/32 Move Redis support from Experimental to mainline, enabled for a build
584 by defining LOOKUP_REDIS. The libhiredis library is required.
586 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
587 keys are given for lookup.
589 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
590 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
591 only supported when built with TLS support. The PolarSSL SHA routines
592 are still used when the TLS library is too old for convenient support.
594 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
595 openssl_options), for security. OpenSSL forces this from version 1.1.0
596 server-side so match that on older versions.
598 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
599 allocation for $value could be released as the expansion processing
600 concluded, but leaving the global pointer active for it.
602 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
603 and to use the domains and local_parts ACL conditions.
605 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
606 incorrectly not doubled on cutthrough transmission, hence seen as a
607 body-termination at the receiving system - resulting in truncated mails.
608 Commonly the sender saw a TCP-level error, and retransmitted the message
609 via the normal store-and-forward channel. This could result in duplicates
610 received - but deduplicating mailstores were liable to retain only the
611 initial truncated version.
613 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
615 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
617 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
618 we're in there, support oversigning also; bug 1309.
620 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
622 HS/04 Add support for keep_environment and add_environment options.
624 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
625 either intentional arithmetic overflow during PRNG, or testing config-
628 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
629 delivery resulted in actual delivery. Cancel cutthrough before DATA
632 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
633 rejects a recipient: pass the reject to the originator.
635 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
636 Many were false-positives and ignorable, but it's worth fixing the
639 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
640 for the new environment-manipulation done at startup. Move the routines
641 from being local to tls.c to being global via the os.c file.
643 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
644 an extract embedded as result-arg for a map, the first arg for extract
645 is unavailable so we cannot tell if this is a numbered or keyed
646 extraction. Accept either.
652 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
655 JH/02 The smtp transport option "multi_domain" is now expanded.
657 JH/03 The smtp transport now requests PRDR by default, if the server offers
660 JH/04 Certificate name checking on server certificates, when exim is a client,
661 is now done by default. The transport option tls_verify_cert_hostnames
662 can be used to disable this per-host. The build option
663 EXPERIMENTAL_CERTNAMES is withdrawn.
665 JH/05 The value of the tls_verify_certificates smtp transport and main options
666 default to the word "system" to access the system default CA bundle.
667 For GnuTLS, only version 3.0.20 or later.
669 JH/06 Verification of the server certificate for a TLS connection is now tried
670 (but not required) by default. The verification status is now logged by
671 default, for both outbound TLS and client-certificate supplying inbound
674 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
677 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
678 Status Notification (bounce) messages are now MIME format per RFC 3464.
679 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
680 under the control of the dsn_advertise_hosts option, and routers may
681 have a dsn_lasthop option.
683 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
684 default, modifiable by a malware= option. The list separator for
685 the options can now be changed in the usual way. Bug 68.
687 JH/10 The smtp_receive_timeout main option is now expanded before use.
689 JH/11 The incoming_interface log option now also enables logging of the
690 local interface on delivery outgoing connections.
692 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
693 if the interface and destination host and port all match.
695 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
698 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
699 Patch from Andrew Lewis.
701 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
702 now supports optional time-restrictions, weighting, and priority
703 modifiers per server. Patch originally by <rommer@active.by>.
705 JH/16 The spamd_address main option now supports a mixed list of local
706 and remote servers. Remote servers can be IPv6 addresses, and
707 specify a port-range.
709 JH/17 Bug 68: The spamd_address main option now supports an optional
710 timeout value per server.
712 JH/18 Bug 1581: Router and transport options headers_add/remove can
713 now have the list separator specified.
715 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
718 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
721 JH/21 Support for the A6 type of dns record is withdrawn.
723 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
724 rather than the verbs used.
726 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
727 from 255 to 1024 chars.
729 JH/24 Verification callouts now attempt to use TLS by default.
731 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
732 are generic router options now. The defaults didn't change.
734 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
735 Original patch from Alexander Shikoff, worked over by JH.
737 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
738 files if ZCAT_COMMAND is not executable.
740 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
742 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
744 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
745 Normally benign, it bites when the pair was led to by a CNAME;
746 modern usage is to not canonicalize the domain to a CNAME target
747 (and we were inconsistent anyway for A-only vs AAAA+A).
749 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
751 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
752 when evaluating $sender_host_dnssec.
754 JH/31 Check the HELO verification lookup for DNSSEC, adding new
755 $sender_helo_dnssec variable.
757 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
759 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
761 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
763 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
764 documented as working, but never had. Support all but $spam_report.
766 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
767 added for tls authenticator.
769 HS/03 Add perl_taintmode main config option
775 TL/01 When running the test suite, the README says that variables such as
776 no_msglog_check are global and can be placed anywhere in a specific
777 test's script, however it was observed that placement needed to be near
778 the beginning for it to behave that way. Changed the runtest perl
779 script to read through the entire script once to detect and set these
780 variables, reset to the beginning of the script, and then run through
781 the script parsing/test process like normal.
783 TL/02 The BSD's have an arc4random API. One of the functions to induce
784 adding randomness was arc4random_stir(), but it has been removed in
785 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
786 function when detected.
788 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
789 cause callback expansion.
791 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
792 syntax errors in an expansion can be treated as a string instead of
793 logging or causing an error, due to the internal use of bool_lax
794 instead of bool when processing it.
796 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
797 server certificates when making smtp deliveries.
799 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
801 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
803 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
805 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
806 Merged patch from Sebastian Wiedenroth.
808 JH/05 Fix results-pipe from transport process. Several recipients, combined
809 with certificate use, exposed issues where response data items split
810 over buffer boundaries were not parsed properly. This eventually
811 resulted in duplicates being sent. This issue only became common enough
812 to notice due to the introduction of connection certificate information,
813 the item size being so much larger. Found and fixed by Wolfgang Breyha.
815 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
816 size buffer was used, resulting in syntax errors when an expansion
819 JH/07 Add support for directories of certificates when compiled with a GnuTLS
820 version 3.3.6 or later.
822 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
823 is EXPERIMENTAL_EVENT, the main-configuration and transport options
824 both become "event_action", the variables become $event_name, $event_data
825 and $event_defer_errno. There is a new variable $verify_mode, usable in
826 routers, transports and related events. The tls:cert event is now also
827 raised for inbound connections, if the main configuration event_action
830 TL/06 In test suite, disable OCSP for old versions of openssl which contained
831 early OCSP support, but no stapling (appears to be less than 1.0.0).
833 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
834 server certificate names available under the smtp transport option
835 "tls_verify_cert_hostname" now do not permit multi-component wildcard
838 JH/10 Time-related extraction expansions from certificates now use the main
839 option "timezone" setting for output formatting, and are consistent
840 between OpenSSL and GnuTLS compilations. Bug 1541.
842 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
843 encoded parameter in the incoming message. Bug 1558.
845 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
846 include certificate info, eximon was claiming there were spoolfile
849 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
851 JH/14 Log delivery-related information more consistently, using the sequence
852 "H=<name> [<ip>]" wherever possible.
854 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
855 are problematic for Debian distribution, omit them from the release
858 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
860 JH/16 Fix string representation of time values on 64bit time_t architectures.
863 JH/17 Fix a null-indirection in certextract expansions when a nondefault
864 output list separator was used.
869 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
870 checkers that were complaining about end of non-void function with no
873 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
874 This was a regression introduced in 4.83 by another bugfix.
876 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
878 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
879 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
885 TF/01 Correctly close the server side of TLS when forking for delivery.
887 When a message was received over SMTP with TLS, Exim failed to clear up
888 the incoming connection properly after forking off the child process to
889 deliver the message. In some situations the subsequent outgoing
890 delivery connection happened to have the same fd number as the incoming
891 connection previously had. Exim would try to use TLS and fail, logging
892 a "Bad file descriptor" error.
894 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
895 utilities have not been installed.
897 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
898 temporary space as the ACL may create new global variables.
900 TL/01 LDAP support uses per connection or global context settings, depending
901 upon the detected version of the libraries at build time.
903 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
904 to extract and use the src ip:port in logging and expansions as if it
905 were a direct connection from the outside internet. PPv2 support was
906 updated based on HAProxy spec change in May 2014.
908 JH/02 Add ${listextract {number}{list}{success}{fail}}.
910 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
911 Properly escape header and check for NULL return.
913 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
916 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
918 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
919 characters in header names, implemented as a verify condition.
920 Contributed by Michael Fischer v. Mollard.
922 TL/05 Rename SPF condition results err_perm and err_temp to standardized
923 results permerror and temperror. Previous values are deprecated but
924 still accepted. In a future release, err_perm and err_temp will be
925 completely removed, which will be a backward incompatibility if the
926 ACL tests for either of these two old results. Patch contributed by
927 user bes-internal on the mailing list.
929 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
931 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
932 selectors, in both main and reject logs.
934 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
937 JH/07 Add malware type "sock" for talking to simple daemon.
939 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
941 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
942 routers/transports under cutthrough routing.
944 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
945 numbers. Touch up "bool" conditional to keep the same definition.
947 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
949 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
951 JH/12 Expand items in router/transport headers_add or headers_remove lists
952 individually rather than the list as a whole. Bug 1452.
954 Required for reasonable handling of multiple headers_ options when
955 they may be empty; requires that headers_remove items with embedded
956 colons must have them doubled (or the list-separator changed).
958 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
959 view the policy declared in the DMARC record. Currently, $dmarc_status
960 is a combined value of both the record presence and the result of the
963 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
965 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
966 dnslookup router and the smtp transport (applying to the forward
969 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
970 of ldap servers used for a specific lookup. Patch provided by Heiko
973 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
974 New variable $lookup_dnssec_authenticated for observability.
976 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
977 Patch submitted by Lars Timman.
979 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
981 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
982 Requires trusted mode and valid format message id, aborts otherwise.
983 Patch contributed by Heiko Schlichting.
985 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
986 certextract with support for various fields. Bug 1358.
988 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
989 is requested by default, modifiable by smtp transport option
992 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
993 operate on certificate variables to give certificate fingerprints
994 Also new ${sha256:cert_variable}.
996 JH/23 The PRDR feature is moved from being Experimental into the mainline.
998 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
999 Christian Aistleitner.
1001 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1003 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1004 file. Patch from Wolfgang Breyha.
1006 JH/25 Expand the coverage of the delivery $host and $host_address to
1007 client authenticators run in verify callout. Bug 1476.
1009 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1010 align with daemon_smtp_ports. Bug 72.
1012 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1013 support and error reporting did not work properly.
1015 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1016 and is readable. Patch from Andrew Colin Kissa.
1018 TL/14 Enhance documentation of ${run expansion and how it parses the
1019 commandline after expansion, particularly in the case when an
1020 unquoted variable expansion results in an empty value.
1022 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1024 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1025 by renaming away from it. Addresses GH issue 15, reported by
1028 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1030 TL/15 SECURITY: prevent double expansion in math comparison functions
1031 (can expand unsanitized data). Not remotely exploitable.
1038 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1040 PP/02 Make -n do something, by making it not do something.
1041 When combined with -bP, the name of an option is not output.
1043 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1046 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1047 $sender_host_name and config options to manage this, and basic check
1050 PP/05 DSCP support for outbound connections and control modifier for inbound.
1052 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1053 (Only plugin which currently uses this is kerberos4, which nobody should
1054 be using, but we should make it available and other future plugins might
1055 conceivably use it, even though it would break NAT; stuff *should* be
1056 using channel bindings instead).
1058 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1059 name; added for Sendmail compatibility; requires admin caller.
1060 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1061 just ignore it); requires trusted caller.
1062 Also parse but ignore: -Ac -Am -X<logfile>
1065 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1067 TL/02 Add +smtp_confirmation as a default logging option.
1069 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1070 Patch by Magnus Holmgren from 2007-02-20.
1072 TL/04 Bugzilla 1281 - Spec typo.
1073 Bugzilla 1283 - Spec typo.
1074 Bugzilla 1290 - Spec grammar fixes.
1076 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1078 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1080 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1081 mailing list by by Dmitry Isaikin.
1083 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1085 JH/02 Support "G" suffix to numbers in ${if comparisons.
1087 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1089 NM/01 Bugzilla 1197 - Spec typo
1090 Bugzilla 1196 - Spec examples corrections
1092 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1094 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1095 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1098 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1099 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1101 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1102 "acl {{name}{arg}...}", and optional args on acl condition
1105 JH/05 Permit multiple router/transport headers_add/remove lines.
1107 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1109 JH/07 Avoid using a waiting database for a single-message-only transport.
1110 Performance patch from Paul Fisher. Bugzilla 1262.
1112 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1115 JH/09 Add $headers_added variable, with content from use of ACL modifier
1116 add_header (but not yet added to the message). Bugzilla 199.
1118 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1119 Pulled from Bugzilla 817 by Wolfgang Breyha.
1121 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1123 (nb: this is the same fix as in Exim 4.80.1)
1125 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1128 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1131 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1133 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1134 advertise SMTP AUTH mechanism to us, instead of a generic
1135 protocol violation error. Also, make Exim more robust to bad
1136 data from the Dovecot auth socket.
1138 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1140 When a queue runner is handling a message, Exim first routes the
1141 recipient addresses, during which it prunes them based on the retry
1142 hints database. After that it attempts to deliver the message to
1143 any remaining recipients. It then updates the hints database using
1146 So if a recipient address works intermittently, it can get repeatedly
1147 deferred at routing time. The retry hints record remains fresh so the
1148 address never reaches the final cutoff time.
1150 This is a fairly common occurrence when a user is bumping up against
1151 their storage quota. Exim had some logic in its local delivery code
1152 to deal with this. However it did not apply to per-recipient defers
1153 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1155 This change adds a proper retry rule check during routing so that the
1156 final cutoff time is checked against the message's age. We only do
1157 this check if there is an address retry record and there is not a
1158 domain retry record; this implies that previous attempts to handle
1159 the address had the retry_use_local_parts option turned on. We use
1160 this as an approximation for the destination being like a local
1161 delivery, as in LMTP.
1163 I suspect this new check makes the old local delivery cutoff check
1164 redundant, but I have not verified this so I left the code in place.
1166 TF/02 Correct gecos expansion when From: is a prefix of the username.
1168 Test 0254 submits a message to Exim with the header
1172 When I ran the test suite under the user fanf2, Exim expanded
1173 the header to contain my full name, whereas it should have added
1174 a Resent-Sender: header. It erroneously treats any prefix of the
1175 username as equal to the username.
1177 This change corrects that bug.
1179 GF/01 DCC debug and logging tidyup
1180 Error conditions log to paniclog rather than rejectlog.
1181 Debug lines prefixed by "DCC: " to remove any ambiguity.
1183 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1185 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1186 Bug spotted by Jeremy Harris; was flawed since initial commit.
1187 Would have resulted in OCSP responses post-SNI triggering an Exim
1188 NULL dereference and crash.
1190 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1192 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1193 Bug detection, analysis and fix by Samuel Thibault.
1194 Bugzilla 1331, Debian bug #698092.
1196 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1198 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1199 Server implementation by Todd Lyons, client by JH.
1200 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1201 config variable "prdr_enable" controls whether the server
1202 advertises the facility. If the client requests PRDR a new
1203 acl_data_smtp_prdr ACL is called once for each recipient, after
1204 the body content is received and before the acl_smtp_data ACL.
1205 The client is controlled by both of: a hosts_try_prdr option
1206 on the smtp transport, and the server advertisement.
1207 Default client logging of deliveries and rejections involving
1208 PRDR are flagged with the string "PRDR".
1210 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1211 fclose(). Diagnosis by Todd Lyons.
1213 PP/17 Update configure.default to handle IPv6 localhost better.
1214 Patch by Alain Williams (plus minor tweaks).
1217 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1218 This is now consistent with GnuTLS, and is now documented: the
1219 previous undocumented portable approach to treating the option as
1220 unset was to force an expansion failure. That still works, and
1221 an empty string is now equivalent.
1223 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1224 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1225 not performing validation itself.
1227 PP/20 Added force_command boolean option to pipe transport.
1228 Patch from Nick Koston, of cPanel Inc.
1230 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1233 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1235 PP/21 Fix eximon continuous updating with timestamped log-files.
1236 Broken in a format-string cleanup in 4.80, missed when I repaired the
1237 other false fix of the same issue.
1238 Report and fix from Heiko Schlichting.
1241 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1242 Report from Prashanth Katuri.
1244 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1245 It's SecureTransport, so affects any MacOS clients which use the
1246 system-integrated TLS libraries, including email clients.
1248 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1249 using a MIME ACL for non-SMTP local injection.
1250 Report and assistance in diagnosis by Warren Baker.
1252 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1254 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1256 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1257 last id that failed so it may be referenced in subsequent ACL's.
1259 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1262 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1263 ldap library initialization, allowing self-signed CA's to be
1264 used. Also properly sets require_cert option later in code by
1265 using NULL (global ldap config) instead of ldap handle (per
1266 session). Bug diagnosis and testing by alxgomz.
1268 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1269 the src/util/ subdirectory.
1271 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1272 renamed to Transport Post Delivery Action by Jeremy Harris, as
1275 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1276 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1277 redis_servers = needs to be configured which will be used by the redis
1278 lookup. Patch from Warren Baker, of The Packet Hub.
1280 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1282 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1283 hostname or reverse DNS when processing a host list. Used suggestions
1284 from multiple comments on this bug.
1286 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1288 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1289 Missed a few lines, added it to make the runtest require no keyboard
1292 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1293 contains upper case chars. Make router use caseful_local_part.
1295 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1296 support when GnuTLS has been built with p11-kit.
1302 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1304 This, or similar/improved, will also be change PP/11 of 4.82.
1310 PP/01 Handle short writes when writing local log-files.
1311 In practice, only affects FreeBSD (8 onwards).
1312 Bugzilla 1053, with thanks to Dmitry Isaikin.
1314 NM/01 Bugzilla 949 - Documentation tweak
1316 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1319 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1321 PP/02 Implemented gsasl authenticator.
1323 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1325 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1326 `pkg-config foo` for cflags/libs.
1328 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1329 with rest of GSASL and with heimdal_gssapi.
1331 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1332 `pkg-config foo` for cflags/libs for the TLS implementation.
1334 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1335 properties get this fed in as external SSF. A number of robustness
1336 and debugging improvements to the cyrus_sasl authenticator.
1338 PP/08 cyrus_sasl server now expands the server_realm option.
1340 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1341 Patch by Jeremy Harris.
1343 PP/10 Added dbmjz lookup type.
1345 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1347 PP/12 MAIL args handles TAB as well as SP, for better interop with
1348 non-compliant senders.
1349 Analysis and variant patch by Todd Lyons.
1351 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1352 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1353 Patch from Dirk Mueller <dmueller@suse.com>
1355 PP/13 tls_peerdn now print-escaped for spool files.
1356 Observed some $tls_peerdn in wild which contained \n, which resulted
1357 in spool file corruption.
1359 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1360 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1361 or write after TLS renegotiation, which otherwise led to messages
1364 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1365 as a tracking header (ie: a signed header comes before the signature).
1366 Patch from Wolfgang Breyha.
1368 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1369 comma-sep list; embedded commas doubled.
1371 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1373 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1375 Report and patch from Dmitry Banschikov.
1377 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1378 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1379 failures. We appear to now support TLS1.1+ with Exim.
1381 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1382 lets Exim select keys and certificates based upon TLS SNI from client.
1383 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1384 before an outbound SMTP session. New log_selector, +tls_sni.
1386 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1387 NULL dereference. Report and patch from Alun Jones.
1389 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1390 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1391 Not seeing resolver debug output on NetBSD, but suspect this is a
1392 resolver implementation change.
1394 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1395 Left warnings. Added "eximon gdb" invocation mode.
1397 PP/21 Defaulting "accept_8bitmime" to true, not false.
1399 PP/22 Added -bw for inetd wait mode support.
1401 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1402 locate the relevant includes and libraries. Made this the default.
1404 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1405 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1407 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1408 This may cause build issues on older platforms.
1410 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1411 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1412 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1413 Added SNI support via GnuTLS too.
1414 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1416 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1418 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1419 Applied second patch from Janne, implementing suggestion to default
1420 multiple-strings-in-record handling to match SPF spec.
1422 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1424 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1425 read-only, out of scope).
1426 Patch from Wolfgang Breyha, report from Stuart Northfield.
1428 PP/29 Fix three issues highlighted by clang analyser static analysis.
1429 Only crash-plausible issue would require the Cambridge-specific
1430 iplookup router and a misconfiguration.
1431 Report from Marcin Mirosław.
1433 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1435 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1436 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1437 As part of this, removing so much warning spew let me fix some minor
1438 real issues in debug logging.
1440 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1441 assignment on my part. Fixed.
1443 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1444 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1445 Janne Snabb (who went above and beyond: thank you).
1447 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1448 string otherwise requires a connection and a bunch more work and it's
1449 relatively easy to get wrong. Should also expose TLS library linkage
1452 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1453 64-bit ${eval} (JH/03).
1455 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1456 GNU libc to support some of the 64-bit stuff, should not lead to
1457 conflicts. Defined before os.h is pulled in, so if a given platform
1458 needs to override this, it can.
1460 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1461 protection layer was required, which is not implemented.
1462 Bugzilla 1254, patch from Wolfgang Breyha.
1464 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1465 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1466 tls_dhparam take prime identifiers. Also unbreak combination of
1467 OpenSSL+DH_params+TLSSNI.
1469 PP/39 Disable SSLv2 by default in OpenSSL support.
1475 PP/01 Solaris build fix for Oracle's LDAP libraries.
1476 Bugzilla 1109, patch from Stephen Usher.
1478 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1480 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1483 TF/02 Fix a couple more cases where we did not log the error message
1484 when unlink() failed. See also change 4.74-TF/03.
1486 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1487 lock up or crash if it happened to be inside a call to libc when it
1488 got a SIGUSR1 from exiwhat.
1490 The SIGUSR1 handler appends the current process status to the process
1491 log which is later printed by exiwhat. It used to use the general
1492 purpose logging code to do this, but several functions it calls are
1493 not safe for signals.
1495 The new output code in the SIGUSR1 handler is specific to the process
1496 log, and simple enough that it's easy to inspect for signal safety.
1497 Removing some special cases also simplifies the general logging code.
1498 Removing the spurious timestamps from the process log simplifies
1501 TF/04 Improved ratelimit ACL condition.
1503 The /noupdate option has been deprecated in favour of /readonly which
1504 has clearer semantics. The /leaky, /strict, and /readonly update modes
1505 are mutually exclusive. The update mode is no longer included in the
1506 database key; it just determines when the database is updated. (This
1507 means that when you upgrade Exim will forget old rate measurements.)
1509 Exim now checks that the per_* options are used with an update mode that
1510 makes sense for the current ACL. For example, when Exim is processing a
1511 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1512 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1513 must specify per_mail/readonly. If you omit the update mode it defaults to
1514 /leaky where that makes sense (as before) or /readonly where required.
1516 The /noupdate option is now undocumented but still supported for
1517 backwards compatibility. It is equivalent to /readonly except that in
1518 ACLs where /readonly is required you may specify /leaky/noupdate or
1519 /strict/noupdate which are treated the same as /readonly.
1521 A useful new feature is the /count= option. This is a generalization
1522 of the per_byte option, so that you can measure the throughput of other
1523 aggregate values. For example, the per_byte option is now equivalent
1524 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1526 The per_rcpt option has been generalized using the /count= mechanism
1527 (though it's more complicated than the per_byte equivalence). When it is
1528 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1529 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1530 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1531 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1532 Note that using per_rcpt with a non-readonly update mode in more than
1533 one ACL will cause the recipients to be double-counted. (The per_mail
1534 and per_byte options don't have this problem.)
1536 The handling of very low rates has changed slightly. If the computed rate
1537 is less than the event's count (usually one) then this event is the first
1538 after a long gap. In this case the rate is set to the same as this event's
1539 count, so that the first message of a spam run is counted properly.
1541 The major new feature is a mechanism for counting the rate of unique
1542 events. The new per_addr option counts the number of different
1543 recipients that someone has sent messages to in the last time period. It
1544 behaves like per_rcpt if all the recipient addresses are different, but
1545 duplicate recipient addresses do not increase the measured rate. Like
1546 the /count= option this is a general mechanism, so the per_addr option
1547 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1548 example, measure the rate that a client uses different sender addresses
1549 with the options per_mail/unique=$sender_address. There are further
1550 details in the main documentation.
1552 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1554 TF/06 Removed a few PCRE remnants.
1556 TF/07 Automatically extract Exim's version number from tags in the git
1557 repository when doing development or release builds.
1559 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1560 Bugzilla 879. Patch from Paul Fisher.
1562 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1563 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1566 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1568 PP/05 Variable $av_failed, true if the AV scanner deferred.
1569 Bugzilla 1078. Patch from John Horne.
1571 PP/06 Stop make process more reliably on build failure.
1572 Bugzilla 1087. Patch from Heiko Schlittermann.
1574 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1575 Bugzilla 1089. Patch from Heiko Schlittermann.
1577 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1578 Bugzilla 1131. Patch from Holger Weiß.
1580 PP/09 Handle IPv6 addresses with SPF.
1581 Bugzilla 860. Patch from Wolfgang Breyha.
1583 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1585 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1588 PP/11 match_* no longer expand right-hand-side by default.
1589 New compile-time build option, EXPAND_LISTMATCH_RHS.
1590 New expansion conditions, "inlist", "inlisti".
1592 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1594 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1596 PP/14 fix log_write() format string regression from TF/03.
1597 Bugzilla 1152. Patch from Dmitry Isaikin.
1603 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1605 PP/02 Harmonised TLS library version reporting; only show if debugging.
1606 Layout now matches that introduced for other libraries in 4.74 PP/03.
1608 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1610 PP/04 New "dns_use_edns0" global option.
1612 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1615 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1616 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1618 TK/01 Updated PolarSSL code to 0.14.2.
1619 Bugzilla 1097. Patch from Andreas Metzler.
1621 PP/07 Catch divide-by-zero in ${eval:...}.
1622 Fixes bugzilla 1102.
1624 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1627 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1628 format-string attack -- SECURITY: remote arbitrary code execution.
1630 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1631 time unintentionally subject to list matching rules, letting the header
1632 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1633 arbitrary string expansion). This allowed for information disclosure.
1635 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1636 INT_MIN/-1 -- value coerced to INT_MAX.
1642 NM/01 Workaround for PCRE version dependency in version reporting
1645 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1646 This fixes portability to compilers other than gcc, notably
1647 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1649 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1650 makefiles for portability to HP-UX and POSIX correctness.
1652 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1653 Also via indented variable definition in the Makefile.
1654 (Debugging by Oliver Heesakkers).
1656 PP/02 Restore caching of spamd results with expanded spamd_address.
1657 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1659 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1660 Improves build reliability. Fix from: Frank Elsner
1662 NM/02 Fix wide character breakage in the rfc2047 coding
1663 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1665 NM/03 Allow underscore in dnslist lookups
1666 Fixes bug 1026. Patch from Graeme Fowler
1668 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1669 Code patches from Adam Ciarcinski of NetBSD.
1671 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1674 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1675 is logged, to avoid truncation. Patch from John Horne.
1677 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1678 Patch from Jakob Hirsch.
1680 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1681 SQL string expansion failure details.
1682 Patch from Andrey Oktyabrski.
1684 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1685 Patch from Simon Arlott.
1687 PP/09 New lookups functionality failed to compile on old gcc which rejects
1688 extern declarations in function scope.
1689 Patch from Oliver Fleischmann
1691 PP/10 Use sig_atomic_t for flags set from signal handlers.
1692 Check getgroups() return and improve debugging.
1693 Fixed developed for diagnosis in bug 927 (which turned out to be
1696 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1697 Patch from Mark Zealey.
1699 PP/12 Bugzilla 1056: Improved spamd server selection.
1700 Patch from Mark Zealey.
1702 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1703 Based on patch from Heiko Schlittermann.
1705 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1706 Patch from Uwe Doering, sign-off by Michael Haardt.
1708 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1709 variable declaration deep within a block. Bug and patch from
1712 PP/15 lookups-Makefile IRIX compatibility coercion.
1714 PP/16 Make DISABLE_DKIM build knob functional.
1716 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1717 Patch by Simon Arlott
1719 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1720 variable argument macros. Our copy now differs from upstream.
1726 TF/01 Failure to get a lock on a hints database can have serious
1727 consequences so log it to the panic log.
1729 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1730 controlled using the smtp_confirmation log selector.
1732 TF/03 Include the error message when we fail to unlink a spool file.
1734 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1735 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1736 for maintaining out-of-tree patches for some time.
1738 PP/01 Bugzilla 139: Documentation and portability issues.
1739 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1740 Handle per-OS dynamic-module compilation flags.
1742 PP/02 Let /dev/null have normal permissions.
1743 The 4.73 fixes were a little too stringent and complained about the
1744 permissions on /dev/null. Exempt it from some checks.
1745 Reported by Andreas M. Kirchwitz.
1747 PP/03 Report version information for many libraries, including
1748 Exim version information for dynamically loaded libraries. Created
1749 version.h, now support a version extension string for distributors
1750 who patch heavily. Dynamic module ABI change.
1752 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1753 privilege escalation vulnerability whereby the Exim run-time user
1754 can cause root to append content of the attacker's choosing to
1757 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1760 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1761 If dropping privileges for untrusted macros, we disabled normal logging
1762 on the basis that it would fail; for the Exim run-time user, this is not
1763 the case, and it resulted in successful deliveries going unlogged.
1764 Fixed. Reported by Andreas Metzler.
1770 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1771 only prepend for the Resent-* case. Fixes regression introduced in
1772 Exim 4.70 by NM/22 for Bugzilla 607.
1774 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1775 increasing numbers of administrators be bitten by this.
1777 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1779 PP/03 Bugzilla 994: added openssl_options main configuration option.
1781 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1783 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1785 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1787 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1788 without a peer certificate, leading to a segfault because of an
1789 assumption that peers always have certificates. Be a little more
1790 paranoid. Problem reported by Martin Tscholak.
1792 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1793 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1794 NB: ClamAV planning to remove STREAM in "middle of 2010".
1795 CL also introduces -bmalware, various -d+acl logging additions and
1796 more caution in buffer sizes.
1798 PP/09 Implemented reverse_ip expansion operator.
1800 PP/10 Bugzilla 937: provide a "debug" ACL control.
1802 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1804 PP/12 Bugzilla 973: Implement --version.
1806 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1808 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1810 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1812 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1813 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1814 ignore trailing whitespace.
1816 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1818 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1821 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1822 Notification from Dr Andrew Aitchison.
1824 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1825 ExtendedDetectionInfo response format.
1826 Notification from John Horne.
1828 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1831 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1832 XSL and documented dependency on system catalogs, with examples of how
1835 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1838 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1839 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1840 configuration file which is writeable by the Exim user or group.
1842 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1843 of configuration files to cover files specified with the -C option if
1844 they are going to be used with root privileges, not just the default
1847 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1848 option (effectively making it always true).
1850 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1851 files to be used while preserving root privileges.
1853 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1854 that rogue child processes cannot use them.
1856 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1857 run-time user, instead of root.
1859 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1860 Exim run-time user without dropping privileges.
1862 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1863 result string, instead of calling string_vformat() twice with the same
1866 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1867 for other users. Others should always drop root privileges if they use
1868 -C on the command line, even for a whitelisted configure file.
1870 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1872 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1878 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1879 $data_path, and $header_path variables; fixed documentation bugs and
1882 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1883 exipick to access non-standard spools, including the "frozen" queue
1886 NM/01 Bugzilla 965: Support mysql stored procedures.
1887 Patch from Alain Williams
1889 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1891 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1892 Patch from Andreas Metzler
1894 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1895 Patch from Kirill Miazine
1897 NM/05 Bugzilla 671: Added umask to procmail example.
1899 JJ/03 installed exipick 20100323.0, fixing doc bug
1901 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1902 directory. Notification and patch from Dan Rosenberg.
1904 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1906 TK/02 Improve log output when DKIM signing operation fails.
1908 MH/01 Treat the transport option dkim_domain as a colon separated
1909 list, not as a single string, and sign the message with each element,
1910 omitting multiple occurences of the same signer.
1912 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1913 Bugzilla 985, 986. Patch by Simon Arlott
1915 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1916 Patch by Simon Arlott
1918 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1919 MBX locking. Notification from Dan Rosenberg.
1925 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1927 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1929 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1931 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1933 NM/04 Bugzilla 915: Flag broken perl installation during build.
1939 TK/01 Added patch by Johannes Berg that expands the main option
1940 "spamd_address" if it starts with a dollar sign.
1942 TK/02 Write list of recipients to X-Envelope-Sender header when building
1943 the mbox-format spool file for content scanning (suggested by Jakob
1946 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1947 (http://www.dcc-servers.net/) support via dccifd. Activated by
1948 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1950 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1951 by Mark Daniel Reidel <mr@df.eu>.
1953 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1954 When building exim an external PCRE library is now needed -
1955 PCRE is a system library on the majority of modern systems.
1956 See entry on PCRE_LIBS in EDITME file.
1958 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1959 conversation. Added nologin parameter to request.
1960 Patch contributed by Kirill Miazine.
1962 TF/01 Do not log submission mode rewrites if they do not change the address.
1964 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1966 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1967 log files in place. Contributed by Roberto Lima.
1969 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1971 TF/03 Bugzilla 615: When checking the local_parts router precondition
1972 after a local_part_suffix or local_part_prefix option, Exim now
1973 does not use the address's named list lookup cache, since this
1974 contains cached lookups for the whole local part.
1976 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1977 Robert Millan. Documentation is in experimental-spec.txt.
1979 TF/04 Bugzilla 668: Fix parallel build (make -j).
1981 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1983 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1984 Patch provided by Jan Srzednicki.
1986 TF/05 Leading white space used to be stripped from $spam_report which
1987 wrecked the formatting. Now it is preserved.
1989 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1990 that they are available at delivery time.
1992 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1994 TF/08 TLS error reporting now respects the incoming_interface and
1995 incoming_port log selectors.
1997 TF/09 Produce a more useful error message if an SMTP transport's hosts
1998 setting expands to an empty string.
2000 NM/06 Bugzilla 744: EXPN did not work under TLS.
2001 Patch provided by Phil Pennock.
2003 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2004 Patch provided by Richard Godbee.
2006 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2007 acl_smtp_notquit, added index entry.
2009 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2010 Patch provided by Eugene Bujak.
2012 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2013 accept(). Patch provided by Maxim Dounin.
2015 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2016 Patch provided by Phil Pennock.
2018 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2020 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2021 Patch provided by Brad "anomie" Jorsch.
2023 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2024 Patch provided by Dean Brooks.
2026 TK/05 Add native DKIM support (does not depend on external libraries).
2028 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2029 Patch provided by Graeme Fowler.
2031 NM/16 Bugzilla 851: Documentation example syntax fix.
2033 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2035 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2038 NM/19 Bugzilla 745: TLS version reporting.
2039 Patch provided by Phil Pennock.
2041 NM/20 Bugzilla 167: bool: condition support.
2042 Patch provided by Phil Pennock.
2044 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2045 clients. Patch provided by Phil Pennock.
2047 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2048 Patch provided by Brad "anomie" Jorsch.
2050 NM/23 Bugzilla 687: Fix misparses in eximstats.
2051 Patch provided by Heiko Schlittermann.
2053 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2054 Patch provided by Heiko Schlittermann.
2056 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2057 plus update to original patch.
2059 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2061 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2062 Patch provided by David Brownlee.
2064 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2066 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2068 NM/30 Bugzilla 888: TLS documentation bugfixes.
2070 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2072 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2073 Unlike the original bugzilla I have changed all shell scripts in src tree.
2075 NM/33 Bugzilla 898: Transport filter timeout fix.
2076 Patch by Todd Rinaldo.
2078 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2079 Patch by Serge Demonchaux.
2081 NM/35 Bugzilla 39: Base64 decode bug fixes.
2082 Patch by Jakob Hirsch.
2084 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2086 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2088 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2090 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2096 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2097 ALT-N's libdkim that I have put here:
2098 http://duncanthrax.net/exim-experimental/
2100 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2101 were called 'true' and it seems that C99 defines that as a reserved
2102 keyword to be used with 'bool' variable types. That means you could
2103 not include C99-style headers which use bools without triggering
2104 build errors in sieve.c.
2106 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2107 as mailq or other aliases. Changed the --help handling significantly
2108 to do whats expected. exim_usage() emits usage/help information.
2110 SC/01 Added the -bylocaldomain option to eximstats.
2112 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2114 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2116 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2122 PH/01 Another patch from the Sieve maintainer.
2124 PH/02 When an IPv6 address is converted to a string for single-key lookup
2125 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2126 dots are used instead of colons so that keys in lsearch files need not
2127 contain colons. This was done some time before quoting was made available
2128 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2129 (notated using the quote facility) so as to distinguish them from IPv4
2130 keys. This meant that lookups for IP addresses in host lists did not work
2131 for iplsearch lookups.
2133 This has been fixed by arranging for IPv6 addresses to be expressed with
2134 colons if the lookup type is iplsearch. This is not incompatible, because
2135 previously such lookups could never work.
2137 The situation is now rather anomalous, since one *can* have colons in
2138 ordinary lsearch keys. However, making the change in all cases is
2139 incompatible and would probably break a number of configurations.
2141 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2144 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2145 conversion specification without a maximum field width, thereby enabling
2146 a rogue spamd server to cause a buffer overflow. While nobody in their
2147 right mind would setup Exim to query an untrusted spamd server, an
2148 attacker that gains access to a server running spamd could potentially
2149 exploit this vulnerability to run arbitrary code as the Exim user.
2151 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2152 $primary_hostname instead of what libspf2 thinks the hosts name is.
2154 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2155 a directory entry by the name of the lookup key. Previously, if a
2156 symlink pointed to a non-existing file or a file in a directory that
2157 Exim lacked permissions to read, a lookup for a key matching that
2158 symlink would fail. Now it is enough that a matching directory entry
2159 exists, symlink or not. (Bugzilla 503.)
2161 PH/03 The body_linecount and body_zerocount variables are now exported in the
2164 PH/04 Added the $dnslist_matched variable.
2166 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2167 This means they are set thereafter only if the connection becomes
2170 PH/06 Added the client_condition to authenticators so that some can be skipped
2171 by clients under certain conditions.
2173 PH/07 The error message for a badly-placed control=no_multiline_responses left
2174 "_responses" off the end of the name.
2176 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2178 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2179 (without spaces) instead of just copying the configuration text.
2181 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2183 PH/11 Added $max_received_linelength.
2185 PH/12 Added +ignore_defer and +include_defer to host lists.
2187 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2188 way in which PCRE > 7.0 is built.
2190 PH/14 Implemented queue_only_load_latch.
2192 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2193 MAIL command. The effect was to mangle the value on 64-bit systems.
2195 PH/16 Another patch from the Sieve maintainer.
2197 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2199 PH/18 If a system quota error occurred while trying to create the file for
2200 a maildir delivery, the message "Mailbox is full" was not appended to the
2201 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2202 only to a quota excession during the actual writing of the file.
2204 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2205 characters?) which causes problems in log lines. The DN values are now
2206 passed through string_printing() before being added to log lines.
2208 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2209 and InterBase are left for another time.)
2211 PH/21 Added message_body_newlines option.
2213 PH/22 Guard against possible overflow in moan_check_errorcopy().
2215 PH/23 POSIX allows open() to be a macro; guard against that.
2217 PH/24 If the recipient of an error message contained an @ in the local part
2218 (suitably quoted, of course), incorrect values were put in $domain and
2219 $local_part during the evaluation of errors_copy.
2225 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2226 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2229 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2230 issue a MAIL command.
2232 PH/02 In an ACL statement such as
2234 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2236 if a client was not listed at all, or was listed with a value other than
2237 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2238 the condition was not true (as it should be), so access was not denied.
2239 The bug was that the ! inversion was incorrectly passed on to the second
2240 item. This has been fixed.
2242 PH/03 Added additional dnslists conditions == and =& which are different from
2243 = and & when the dns lookup returns more than one IP address.
2245 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2246 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2248 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2249 FSYNC, which compiles an option called disable_fsync that allows for
2250 bypassing fsync(). The documentation is heavily laced with warnings.
2252 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2254 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2255 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2256 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2257 including adding "make clean"; (3) Added -fPIC when compiling the test
2258 dynamically loaded module, to get rid of a warning.
2260 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2261 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2262 = 0s. The bug is otherwise harmless.
2264 PH/07 There was a bug in the dovecot authenticator such that the value of
2265 $auth1 could be overwritten, and so not correctly preserved, after a
2266 successful authentication. This usually meant that the value preserved by
2267 the server_setid option was incorrect.
2269 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2271 PH/09 Installed PCRE release 7.0.
2273 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2274 run for batched SMTP input. It is now run at the start of every message
2275 in the batch. While fixing this I discovered that the process information
2276 (output by running exiwhat) was not always getting set for -bs and -bS
2277 input. This is fixed, and it now also says "batched" for BSMTP.
2279 PH/11 Added control=no_pipelining.
2281 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2282 patch, slightly modified), and move the expansion of helo_data till after
2283 the connection is made in the smtp transport (so it can use these
2286 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2288 PH/14 Added log_selector = +pid.
2290 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2292 PH/16 Add ${if forany and ${if forall.
2294 PH/17 Added dsn_from option to vary the From: line in DSNs.
2296 PH/18 Flush SMTP output before performing a callout, unless control =
2297 no_callout_flush is set.
2299 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2300 was true (the default) a successful delivery failed to delete the retry
2301 item, thus causing premature timeout of the address. The bug is now
2304 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2306 PH/21 Long custom messages for fakedefer and fakereject are now split up
2307 into multiline responses in the same way that messages for "deny" and
2308 other ACL rejections are.
2310 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2311 with slight modification.
2313 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2314 draft, changing the syntax and factoring some duplicate code".
2316 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2317 for deliveries of the second and subsequent messages over the same SMTP
2320 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2321 ${reduce, with only minor "tidies".
2323 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2325 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2326 expansion side effects.
2328 PH/27 When a message times out after an over-quota error from an Exim-imposed
2329 quota, the bounce message says "mailbox is full". This message was not
2330 being given when it was a system quota that was exceeded. It now should
2333 MH/03 Made $recipients available in local_scan(). local_scan() already has
2334 better access to the recipient list through recipients_list[], but
2335 $recipients can be useful in postmaster-provided expansion strings.
2337 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2338 in the case of a MAIL command with additional options following the
2339 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2340 were accidentally chopped off.
2342 PH/29 SMTP synchronization checks are implemented when a command is read -
2343 there is a check that no more input is waiting when there shouldn't be
2344 any. However, for some commands, a delay in an ACL can mean that it is
2345 some time before the response is written. In this time, more input might
2346 arrive, invalidly. So now there are extra checks after an ACL has run for
2347 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2348 pipelining has not been advertised.
2350 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2352 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2353 returned in the SMTP response when smtp_return_error_details was false.
2354 This has been fixed.
2356 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2357 instead of the C I/O that was originally supplied, because problems were
2358 reported on Solaris.
2360 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2361 Exim which did not show up earlier: it was assuming that a call to
2362 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2363 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2364 was a macro that became an assignment, so it seemed to work. This has
2365 changed to a proper function call with a void return, hence the compile
2366 error. Exim's code has been fixed.
2368 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2371 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2373 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2375 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2376 because it turns out to be too big or there is a timeout) there is no
2377 "Completed" line in the log. When some messages of this type were
2378 selected by exigrep, they were listed as "not completed". Others were
2379 picked up by some special patterns. I have improved the selection
2380 criteria to be more general.
2382 PH/38 The host_find_failed option in the manualroute router can now be set
2383 to "ignore", to completely ignore a host whose IP address cannot be
2384 found. If all hosts are ignored, the behaviour is controlled by the new
2385 host_all_ignored option.
2387 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2388 homing or because of multiple MX records with /mx) generated more than
2389 one IP address, and the following item turned out to be the local host,
2390 all the secondary addresses of the first item were incorrectly removed
2391 from the list, along with the local host and any following hosts (which
2392 is what is supposed to happen).
2394 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2395 whoever called Exim into the -H file. In the case of the daemon it was
2396 behaving confusingly. When first started, it used values for whoever
2397 started the daemon, but after a SIGHUP it used the Exim user (because it
2398 calls itself on a restart). I have changed the code so that it now always
2401 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2402 message are rejected with the same error (e.g. no authentication or bad
2403 sender address), and a DATA command is nevertheless sent (as can happen
2404 with PIPELINING or a stupid MUA), the error message that was given to the
2405 RCPT commands is included in the rejection of the DATA command. This is
2406 intended to be helpful for MUAs that show only the final error to their
2409 PH/42 Another patch from the Sieve maintainer.
2411 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2412 Eximstats - Fixed some broken HTML links and added missing column headers
2414 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2415 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2417 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2419 PH/43 Yet another patch from the Sieve maintainer.
2421 PH/44 I found a way to check for a TCP/IP connection going away before sending
2422 the response to the final '.' that terminates a message, but only in the
2423 case where the client has not sent further data following the '.'
2424 (unfortunately, this is allowed). However, in many cases there won't be
2425 any further data because there won't be any more messages to send. A call
2426 to select() can be used: if it shows that the input is "ready", there is
2427 either input waiting, or the socket has been closed. An attempt to read
2428 the next input character can distinguish the two cases. Previously, Exim
2429 would have sent an OK response which the client would never have see.
2430 This could lead to message repetition. This fix should cure that, at
2431 least in a lot of common cases.
2433 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2434 advertised in response to EHLO.
2440 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2441 fixed by 4.65/MH/01 (is this a record?) are fixed:
2443 (i) An empty string was always treated as zero by the numeric comparison
2444 operators. This behaviour has been restored.
2446 (ii) It is documented that the numeric comparison operators always treat
2447 their arguments as decimal numbers. This was broken in that numbers
2448 starting with 0 were being interpreted as octal.
2450 While fixing these problems I realized that there was another issue that
2451 hadn't been noticed. Values of message_size_limit (both the global option
2452 and the transport option) were treated as octal if they started with 0.
2453 The documentation was vague. These values are now always treated as
2454 decimal, and I will make that clear in the documentation.
2460 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2461 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2464 MH/01 Don't check that the operands of numeric comparison operators are
2465 integers when their expansion is in "skipping" mode (fixes bug
2466 introduced by 4.64-PH/07).
2468 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2469 child addresses, Exim now panics and dies. Previously, because the count
2470 is held in a short int, deliveries were likely to be lost. As such a
2471 large number of recipients for a single message is ridiculous
2472 (performance will be very, very poor), I have chosen to impose a limit
2473 rather than extend the field.
2479 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2480 leftover -K file (the existence of which was triggered by #402).
2481 While we were at it, introduced process PID as part of the -K
2482 filename. This should rule out race conditions when creating
2485 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2486 processing considerably. Previous code took too long for large mails,
2487 triggering a timeout which in turn triggers #401.
2489 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2490 in the DK code in transports.c. sendfile() is not really portable,
2491 hence the _LINUX specificness.
2493 TF/01 In the add_headers option to the mail command in an Exim filter,
2494 there was a bug that Exim would claim a syntax error in any
2495 header after the first one which had an odd number of characters
2498 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2499 callout verification, Exim cached a "reject" for the entire domain. This
2500 is correct for most verifications, but it is not correct for a recipient
2501 verification with use_sender or use_postmaster set, because in that case
2502 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2503 case of MAIL FROM:<> rejection from other early rejections (e.g.
2504 rejection of HELO). When verifying a recipient using a non-null MAIL
2505 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2506 Whatever the result of the callout, the value of the domain cache is
2507 left unchanged (for any other kind of callout, getting as far as trying
2508 RCPT means that the domain itself is ok).
2510 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2513 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2514 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2515 session, the daemon ignores ECONNECT errors and logs others; it now
2516 ignores EPIPE as well.
2518 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2519 (quoted-printable decoding).
2521 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2522 later the small subsequent patch to fix an introduced bug.
2524 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2526 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2528 PH/08 An error is now given if message_size_limit is specified negative.
2530 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2531 to be given (somewhat) arbitrary names.
2533 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2536 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2537 miscellaneous code fixes
2539 PH/10 Added the log_reject_target ACL modifier to specify where to log
2542 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2543 hostname. This is wrong, because it relates to the incoming message (and
2544 probably the interface on which it is arriving) and not to the outgoing
2545 callout (which could be using a different interface). This has been
2546 changed to use the value of the helo_data option from the smtp transport
2547 instead - this is what is used when a message is actually being sent. If
2548 there is no remote transport (possible with a router that sets up host
2549 addresses), $smtp_active_hostname is used.
2551 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2552 tweaks were necessary in order to get it to work (see also 21 below):
2553 (a) The code assumed that strncpy() returns a negative number on buffer
2554 overflow, which isn't the case. Replaced with Exim's string_format()
2556 (b) There were several signed/unsigned issues. I just did the minimum
2557 hacking in of casts. There is scope for a larger refactoring.
2558 (c) The code used strcasecmp() which is not a standard C function.
2559 Replaced with Exim's strcmpic() function.
2560 (d) The code set only $1; it now sets $auth1 as well.
2561 (e) A simple test gave the error "authentication client didn't specify
2562 service in request". It would seem that Dovecot has changed its
2563 interface. Fortunately there's a specification; I followed it and
2564 changed what the client sends and it appears to be working now.
2566 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2569 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2570 address A is aliased to B and C, where B exists and C does not. Without
2571 -v the output is "A verified" because verification stops after a
2572 successful redirection if more than one address is generated. However,
2573 with -v the child addresses are also verified. Exim was outputting "A
2574 failed to verify" and then showing the successful verification for C,
2575 with its parentage. It now outputs "B failed to verify", showing B's
2576 parentage before showing the successful verification of C.
2578 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2579 look up a TXT record in a specific list after matching in a combined
2582 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2583 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2584 they consult the DNS. I had assumed they would set it the way they
2585 wanted; and indeed my experiments on Linux seem to show that in some
2586 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2587 To be on the safe side, however, I have now made the interface to
2588 host_find_byname() similar to host_find_bydns(), with an argument
2589 containing the DNS resolver options. The host_find_byname() function now
2590 sets these options at its start, just as host_find_bydns() does. The smtp
2591 transport options dns_qualify_single and dns_search_parents are passed to
2592 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2593 of host_find_byname() use the default settings of RES_DEFNAMES
2594 (qualify_single) but not RES_DNSRCH (search_parents).
2596 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2597 spool_read_header() do less string testing, by means of a preliminary
2598 switch on the second character of optional "-foo" lines. (This is
2599 overdue, caused by the large number of possibilities that now exist.
2600 Originally there were few.) While I was there, I also converted the
2601 str(n)cmp tests so they don't re-test the leading "-" and the first
2602 character, in the hope this might squeeze out yet more improvement.
2604 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2605 flag allowing group syntax was set by the header_syntax check but not
2606 turned off, possible causing trouble later; (2) The flag was not being
2607 set at all for the header_verify test, causing "group"-style headers to
2608 be rejected. I have now set it in this case, and also caused header_
2609 verify to ignore an empty address taken from a group. While doing this, I
2610 came across some other cases where the code for allowing group syntax
2611 while scanning a header line wasn't quite right (mostly, not resetting
2612 the flag correctly in the right place). These bugs could have caused
2613 trouble for malformed header lines. I hope it is now all correct.
2615 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2616 with the "reply" argument non-NULL. The code, however (which originally
2617 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2618 but it didn't always do it. This confused somebody who was copying the
2619 code for some other use. I have removed all the tests.
2621 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2622 feature that was used to support insecure browsers during the U.S. crypto
2623 embargo. It requires special client support, and Exim is probably the
2624 only MTA that supported it -- and would never use it because real RSA is
2625 always available. This code has been removed, because it had the bad
2626 effect of slowing Exim down by computing (never used) parameters for the
2627 RSA_EXPORT functionality.
2629 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2630 authenticator to fail if there's a tab character in the incoming data
2631 (there should never be unless someone is messing about, as it's supposed
2632 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2633 if the connection is using TLS or if the remote IP is the same as the
2634 local IP, and the "valid-client-cert option" if a client certificate has
2637 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2638 authenticators. This can be used for authorization after authentication
2639 succeeds. (In the case of plaintext, it servers for both authentication
2642 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2643 if any retry times were supplied.
2645 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2646 connection, where there is no client IP address to check. In this
2647 situation, the verify now always succeeds.
2649 PH/25 Applied John Jetmore's -Mset patch.
2651 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2653 PH/27 In a string expansion for a processed (not raw) header when multiple
2654 headers of the same name were present, leading whitespace was being
2655 removed from all of them, but trailing whitespace was being removed only
2656 from the last one. Now trailing whitespace is removed from each header
2657 before concatenation. Completely empty headers in a concatenation (as
2658 before) are ignored.
2660 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2661 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2663 PH/29 [Removed. This was a change that I later backed out, and forgot to
2664 correct the ChangeLog entry (that I had efficiently created) before
2665 committing the later change.]
2667 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2668 address errors (4xx response to RCPT) over the same connection as other
2669 messages routed to the same hosts. Such deliveries are always "forced",
2670 so retry times are not inspected. This resulted in far too many retries
2671 for the affected addresses. The effect occurred only when there were more
2672 hosts than the hosts_max_try setting in the smtp transport when it had
2673 the 4xx errors. Those hosts that it had tried were not added to the list
2674 of hosts for which the message was waiting, so if all were tried, there
2675 was no problem. Two fixes have been applied:
2677 (i) If there are any address or message errors in an SMTP delivery, none
2678 of the hosts (tried or untried) are now added to the list of hosts
2679 for which the message is waiting, so the message should not be a
2680 candidate for sending over the same connection that was used for a
2681 successful delivery of some other message. This seems entirely
2682 reasonable: after all the message is NOT "waiting for some host".
2683 This is so "obvious" that I'm not sure why it wasn't done
2684 previously. Hope I haven't missed anything, but it can't do any
2685 harm, as the worst effect is to miss an optimization.
2687 (ii) If, despite (i), such a delivery is accidentally attempted, the
2688 routing retry time is respected, so at least it doesn't keep
2689 hammering the server.
2691 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2692 in ${readsocket because some servers need this prod.
2694 PH/32 Added some extra debug output when updating a wait-xxx database.
2696 PH/33 The hint "could be header name not terminated by colon", which has been
2697 given for certain expansion errors for a long time, was not being given
2698 for the ${if def:h_colon_omitted{... case.
2700 PH/34 The spec says: "With one important exception, whenever a domain list is
2701 being scanned, $domain contains the subject domain." There was at least
2702 one case where this was not true.
2704 PH/35 The error "getsockname() failed: connection reset by peer" was being
2705 written to the panic log as well as the main log, but it isn't really
2706 panic-worthy as it just means the connection died rather early on. I have
2707 removed the panic log writing for the ECONNRESET error when getsockname()
2710 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2711 runs only) independently of the message's sender address. This meant
2712 that, if the 4xx error was in fact related to the sender, a different
2713 message to the same recipient with a different sender could confuse
2714 things. In particular, this can happen when sending to a greylisting
2715 server, but other circumstances could also provoke similar problems.
2716 I have changed the default so that the retry time for these errors is now
2717 based a combination of the sender and recipient addresses. This change
2718 can be overridden by setting address_retry_include_sender=false in the
2721 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2722 remote server are returned as part of bounce messages. This was not
2723 happening for LMTP over a pipe (the lmtp transport), but now it is the
2724 same for both kinds of LMTP.
2726 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2727 in header lines that were in fact CNAMEs. This is no longer the case.
2729 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2730 and queue runs started by the daemon processed all messages. This has
2731 been fixed so that -R and -S can now usefully be given with -q<time>.
2733 PH/40 Import PCRE release 6.7 (fixes some bugs).
2735 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2737 PH/42 Give an error if -q is specified more than once.
2739 PH/43 Renamed the variables $interface_address and $interface_port as
2740 $received_ip_address and $received_port, to make it clear that these
2741 values apply to message reception, and not to the outgoing interface when
2742 a message is delivered. (The old names remain recognized, of course.)
2744 PH/44 There was no timeout on the connect() call when using a Unix domain
2745 socket in the ${readsocket expansion. There now is.
2747 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2748 be meaningful with "accept".
2750 SC/01 Eximstats V1.43
2751 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2753 SC/02 Eximstats V1.44
2754 Use a glob alias rather than an array ref in the generated
2755 parser. This improves both readability and performance.
2757 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2758 Collect SpamAssassin and rejection statistics.
2759 Don't display local sender or destination tables unless
2760 there is data to show.
2761 Added average volumes into the top table text output.
2763 SC/04 Eximstats V1.46
2764 Collect data on the number of addresses (recipients)
2765 as well as the number of messages.
2767 SC/05 Eximstats V1.47
2768 Added 'Message too big' to the list of mail rejection
2769 reasons (thanks to Marco Gaiarin).
2771 SC/06 Eximstats V1.48
2772 Mainlog lines which have GMT offsets and are too short to
2773 have a flag are now skipped.
2775 SC/07 Eximstats V1.49 (Alain Williams)
2776 Added the -emptyok flag.
2778 SC/08 Eximstats V1.50
2779 Fixes for obtaining the IP address from reject messages.
2781 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2782 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2783 whitespace changes from 4.64-PH/27
2785 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2788 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2789 are found, allow negative numbers in numeric criteria)
2791 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2793 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2794 to match changes made in 4.64-PH/43
2796 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2798 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2799 "domain" is actually an IP address, and force a failure. This locks out
2800 those revolvers/nameservers that support "A-for-A" lookups, in
2801 contravention of the specifications.
2803 PH/48 When a host name was looked up from an IP address, and the subsequent
2804 forward lookup of the name timed out, the host name was left in
2805 $sender_host_name, contrary to the specification.
2807 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2808 restricted to single-key lookups, Exim was not diagnosing an error if
2809 * or *@ was used with a query-style lookup.
2811 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2813 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2814 long ago, but noone interested enough thought of it. Let's just say that
2815 the "1.1" means that there are some new functions that weren't there at
2816 some point in the past.
2818 PH/51 Error processing for expansion failure of helo_data from an smtp
2819 transport during callout processing was broken.
2821 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2822 tested/used via the -bh/-bhc/-bs options.
2824 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2825 bug, fixed in subsequent PCRE releases).
2827 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2828 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2830 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2836 SC/01 Use a glob alias rather than an array ref in eximstats generated
2837 parser. This improves both readability and performance.
2839 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2840 Don't display local sender or destination tables in eximstats unless
2841 there is data to show.
2842 Added average volumes into the eximstats top table text output.
2844 SC/03 Collect data on the number of addresses (recipients) as well
2845 as the number of messages in eximstats.
2847 TF/01 Correct an error in the documentation for the redirect router. Exim
2848 does (usually) call initgroups() when daemonizing.
2850 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2851 with consistent privilege compared to when running as a daemon.
2853 TF/03 Note in the spec that $authenticated_id is not set for local
2854 submissions from trusted users.
2856 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2857 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2859 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2860 by adding some example configuration directives to the default
2861 configuration file. A little bit of work is required to uncomment the
2862 directives and define how usernames and passwords are checked, but
2863 there is now a framework to start from.
2865 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2866 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2867 without this. I don't know how relevant this is to other LDAP libraries.
2869 PH/02 Add the verb name to the "unknown ACL verb" error.
2871 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2873 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2875 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2876 directory not expanded when it should be if an expanded home directory
2877 was set for the address (which is overridden by the transport).
2879 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2882 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2883 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2884 because it is too late at that time, and has no effect.
2886 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2887 security issue with \' (bugzilla #107). I could not use the
2888 PQescapeStringConn() function, because it needs a PGconn value as one of
2891 PH/08 When testing addresses using -bt, indicate those final addresses that
2892 are duplicates that would not cause an additional delivery. At least one
2893 person was confused, thinking that -bt output corresponded to deliveries.
2894 (Suppressing duplicates isn't a good idea as you lose the information
2895 about possibly different redirections that led to the duplicates.)
2897 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2898 systems where poll() doesn't work, in particular OS X.
2900 PH/10 Added more information to debugging output for retry time not reached.
2902 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2903 operations in malware.c.
2905 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2908 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2909 syslog_duplication set false, log lines that would normally be written
2910 both the the main log and to the reject log were not written to syslog at
2913 PH/14 In the default configuration, change the use of "message" in ACL warn
2914 statements to "add_header".
2916 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2917 not followed by a command (e.g. "seen endif").
2919 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2920 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2923 PH/17 Added extra conditions to the default value of delay_warning_condition
2927 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2928 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2929 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2932 The Auto-Submitted: and various List- headers are standardised, whereas I
2933 don't think Precedence: ever was.
2935 PH/18 Refactored debugging code in route_finduser() to show more information,
2936 in particular, the error code if getpwnam() issues one.
2938 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2939 This is apparently needed in addition to the PH/07 change above to avoid
2940 any possible encoding problems.
2942 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2943 but not after initializing Perl.
2945 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2946 output them only if debugging. By default they are written stderr,
2947 apparently, which is not desirable.
2949 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2952 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2955 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2957 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2958 authenticated or an ident call has been made. Suppress the default
2959 values for $authenticated_id and $authenticated_sender (but permit -oMai
2960 and -oMas) when testing with -bh.
2962 PH/24 Re-jigged the order of the tests in the default configuration so that the
2963 tests for valid domains and recipients precede the DNS black list and CSA
2964 tests, on the grounds that those ones are more expensive.
2966 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2967 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2968 This bug exists in every version of Exim that I still have, right back to
2971 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2972 However, an attempt to turn on case-sensitivity in a regex key by
2973 including (?-i) didn't work because the subject string was already
2974 lowercased, and the effects were non-intuitive. It turns out that a
2975 one-line patch can be used to allow (?-i) to work as expected.
2981 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2982 other effects) broke the use of negated acl sub-conditions.
2984 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2987 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2988 "Deny" causes Exim to reject the incoming connection with a 554 error.
2989 Unfortunately, if there is a major crisis, such as a disk failure,
2990 tcp-wrappers gives "deny", whereas what one would like would be some
2991 kind of temporary error. A kludge has been added to help with this.
2992 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2993 554 error is used if errno is still zero or contains ENOENT (which occurs
2994 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2997 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2999 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3000 errors. Otherwise a message that provokes a temporary error (when other
3001 messages do not) can cause a whole host to time out.
3003 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3004 addresses were routed directly to files or pipes from a redirect router.
3005 File deliveries just didn't batch; pipe deliveries might have suffered
3008 PH/06 A failure to get a lock for a hints database would erroneously always say
3009 "Failed to get write lock", even when it was really a read lock.
3011 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3012 of 0600. This has been changed to use the value of the lockfile_mode
3013 option (which defaults to 0600).
3015 PH/08 Applied small patch from the Sieve maintainer.
3017 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3018 folder from quota calculations, a direct delivery into this folder messed
3019 up the contents of the maildirsize file. This was because the regex was
3020 used only to exclude .Trash (or whatever) when the size of the mailbox
3021 was calculated. There was no check that a delivery was happening into an
3022 excluded directory. This bug has been fixed by ignoring all quota
3023 processing for deliveries into excluded directories.
3025 PH/10 Added the maildirfolder_create_regex option to appendfile.
3031 PH/01 The code for finding all the local interface addresses on a FreeBSD
3032 system running IPv6 was broken. This may well have applied to all BSD
3033 systems, as well as to others that have similar system calls. The broken
3034 code found IPv4 interfaces correctly, but gave incorrect values for the
3035 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3036 that it would not match correctly against @[] and not recognize the IPv6
3039 PH/02 The ipliteral router was not recognizing addresses of the form user@
3040 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3042 PH/03 Added disable_ipv6.
3044 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3045 decoded form, because it is most often used to construct To: headers
3046 lines in autoreplies, and the decoded form may well be syntactically
3047 invalid. However, $reply_address has leading white space removed, and all
3048 newlines turned into spaces so that the autoreply transport does not
3051 PH/05 If group was specified without a user on a router, and no group or user
3052 was specified on a transport, the group from the router was ignored.
3054 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3055 for visible compile-time settings that can be used to change these
3056 numbers, for those that want even more. Backwards compatibility with old
3057 spool files has been maintained. However, going back to a previous Exim
3058 release will lost any variables that are in spool files.
3060 PH/07 Two small changes when running in the test harness: increase delay when
3061 passing a TCP/IP connection to a new process, in case the original
3062 process has to generate a bounce, and remove special handling of
3063 127.0.0.2 (sic), which is no longer necessary.
3065 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3066 be the same on different OS.
3068 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3071 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3072 whether --show-vars was specified or not
3074 JJ/02 exipick: Added support for new ACL variable spool format introduced
3077 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3078 syntactically invalid From: or Reply-to: line, and a filter used this to
3079 generate an autoreply, and therefore failed to obtain an address for the
3080 autoreply, Exim could try to deliver to a non-existent relative file
3081 name, causing unrelated and misleading errors. What now happens is that
3082 it logs this as a hard delivery error, but does not attempt to create a
3085 PH/11 The exinext utility has a -C option for testing purposes, but although
3086 the given file was scanned by exinext itself; it wasn't being passed on
3087 when Exim was called.
3089 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3090 an end-of-file indication when reading a command response.
3092 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3093 compiled. In many other places in Exim, IPv6 addresses are always
3094 recognized, so I have changed this. It also means that IPv4 domain
3095 literals of the form [IPV4:n.n.n.n] are now always recognized.
3097 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3098 used if the router is not running as root, for example, when verifying at
3099 ACL time, or when using -bh. The debugging output from this situation was
3100 non-existent - all you got was a failure to exec. I have made two
3103 (a) Failures to set uid/gid, the current directory, or a process leader
3104 in a subprocess such as that created by queryprogram now generate
3105 suitable debugging output when -d is set.
3107 (b) The queryprogram router detects when it is not running as root,
3108 outputs suitable debugging information if -d is set, and then runs
3109 the subprocess without attempting to change uid/gid.
3111 PH/15 Minor change to Makefile for building test_host (undocumented testing
3114 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3115 additional section of a DNS packet that returns MX or SRV records.
3116 Instead, it always explicitly searches for A/AAAA records. This avoids
3117 major problems that occur when a DNS server includes only records of one
3118 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3119 fixed another bug: if SRV records were looked up and the corresponding
3120 address records were *not* found in the additional section, the port
3121 values from the SRV records were lost.
3123 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3124 using the correct key (the original address) when searching the retry
3125 rules in order to find which one to use for generating the retry hint.
3127 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3128 adding the default one. Similarly, if it contains a Reply-To: header, the
3129 errors_reply_to option, if set, is not used.
3131 PH/19 When calculating a retry time, Exim used to measure the "time since
3132 failure" by looking at the "first failed" field in the retry record. Now
3133 it does not use this if it is later than than the arrival time of the
3134 message. Instead it uses the arrival time. This makes for better
3135 behaviour in cases where some deliveries succeed, thus re-setting the
3136 "first failed" field. An example is a quota failure for a huge message
3137 when small messages continue to be delivered. Without this change, the
3138 "time since failure" will always be short, possible causing more frequent
3139 delivery attempts for the huge message than are intended.
3140 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3142 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3143 $1, $2, $3) because the numerical variables can be reset during some
3144 expansion items (e.g. "match"), thereby losing the authentication data.
3146 PH/21 Make -bV show the size of off_t variables so that the test suite can
3147 decide whether to run tests for quotas > 2G.
3149 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3150 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3151 filecount value is greater than 2G or if a quota value is greater than 2G
3152 on a system where the size of off_t is not greater than 4, a panic error
3155 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3156 never match. The debug and -bh output now contains an explicit error
3157 message indicating a malformed IPv4 address or mask.
3159 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3160 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3161 PH/23 above applies.
3163 PH/25 Do not write to syslog when running in the test harness. The only
3164 occasion when this arises is a failure to open the main or panic logs
3165 (for which there is an explicit test).
3167 PH/26 Added the /no_tell option to "control=freeze".
3169 PH/27 If a host name lookup failed very early in a connection, for example, if
3170 the IP address matched host_lookup and the reverse lookup yielded a name
3171 that did not have a forward lookup, an error message of the form "no IP
3172 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3173 could be logged. Now it outputs the IP address instead of "NULL".
3175 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3176 allows the sender and the authenticated sender to be set when
3177 submitting a message from within Exim. Since child_open_exim() is
3178 documented for local_scan(), the new function should be too.
3180 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3181 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3182 results in an empty string is now treated as unset.
3184 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3186 PH/31 Added sender_verify_fail logging option.
3188 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3189 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3190 tidied the source and removed it altogether.
3192 PH/33 When a queue run was abandoned because the load average was too high, a
3193 log line was always written; now it is written only if the queue_run log
3194 selector is set. In addition, the log line for abandonment now contains
3195 information about the queue run such as the pid. This is always present
3196 in "start" and "stop" lines but was omitted from the "abandon" line.
3198 PH/34 Omit spaces between a header name and the colon in the error message that
3199 is given when verify = headers_syntax fails (if there are lots of them,
3200 the message gets confusing).
3202 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3203 names, as there are now some PTR records that contain slashes. This check
3204 is only to protect against broken name servers that fall over on strange
3205 characters, so the fact that it applies to all lookups doesn't matter.
3207 PH/36 Now that the new test suite is complete, we can remove some of the
3208 special code in Exim that was needed for the old test suite. For example,
3209 sorting DNS records because real resolvers return them in an arbitrary
3210 order. The new test suite's fake resolver always returns records in the
3213 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3214 bounces) except when queue_only is set, to avoid logging races between
3215 the different processes.
3217 PH/38 Panic-die if .include specifies a non-absolute path.
3219 PH/39 A tweak to the "H" retry rule from its user.
3221 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3222 a label. They prevented compilation on older perls.
3224 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3225 a warning to be raised on newish perls.
3227 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3228 on queue. Changes to match documented behaviour of showing count of
3229 messages matching specified criteria.
3231 PH/40 Changed the default ident timeout from 30s to 5s.
3233 PH/41 Added support for the use of login_cap features, on those BSD systems
3234 that have them, for controlling the resources used by pipe deliveries.
3236 PH/42 The content-scanning code uses fopen() to create files in which to put
3237 message data. Previously it was not paying any attention to the mode of
3238 the files. Exim runs with umask(0) because the rest of the code creates
3239 files with open(), and sets the required mode explicitly. Thus, these
3240 files were ending up world-writeable. This was not a big issue, because,
3241 being within the spool directory, they were not world-accessible. I have
3242 created a function called modefopen, which takes an additional mode
3243 argument. It sets umask(777), creates the file, chmods it to the required
3244 mode, then resets the umask. All the relevant calls to fopen() in the
3245 content scanning code have been changed to use this function.
3247 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3248 to 24 hours. This avoids potential overflow problems when processing G
3249 and H retry rules. I suspect nobody ever tinkers with this value.
3251 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3253 PH/45 When the plaintext authenticator is running as a client, the server's
3254 challenges are checked to ensure they are valid base64 strings. By
3255 default, the authentication attempt is cancelled if an invalid string is
3256 received. Setting client_ignore_invalid_base64 true ignores these errors.
3257 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3258 they are received. Thus, the responses can be made to depend on the
3259 challenges. If an invalid string is ignored, an empty string is placed in
3262 PH/46 Messages that are created by the autoreply transport now contains a
3263 References: header, in accordance with RFCs 2822 and 3834.
3265 PH/47 Added authenticated_sender_force to the smtp transport.
3267 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3269 PH/49 Installed latest patch from the Sieve maintainer.
3271 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3272 was also set, the appendfile transport was unnecessarily scanning a
3273 directory of message files (e.g. for maildir delivery) to find the count
3274 of files (along with the size), even though it did not need this
3275 information. It now does the scan only if it needs to find either the
3276 size of the count of files.
3278 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3280 PH/52 Two bugs concerned with error handling when the smtp transport is
3283 (i) Exim was not creating retry information for temporary errors given
3284 for individual recipients after the DATA command when the smtp transport
3285 was used in LMTP mode. This meant that they could be retried too
3286 frequently, and not timed out correctly.
3288 (ii) Exim was setting the flag that allows error details to be returned
3289 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3290 recipients that were returned after the DATA command.
3292 PH/53 This is related to PH/52, but is more general: for any failing address,
3293 when detailed error information was permitted to be returned to the
3294 sender, but the error was temporary, then after the final timeout, only
3295 "retry timeout exceeded" was returned. Now it returns the full error as
3296 well as "retry timeout exceeded".
3298 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3299 do this, and (what is worse) MTAs that accept it.
3301 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3302 will now be deprecated.
3304 PH/56 New os.c-cygwin from the Cygwin maintainer.
3306 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3307 formats (previously only available in exim formats via -bpr, -bpru,
3308 and -bpra. Now also available in native and exiqgrep formats)
3310 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3311 with very large, slow to parse queues
3313 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3315 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3317 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3318 responses to authentication challenges, though it was showing the
3319 challenges; (ii) I've removed the CR characters from the debug output for
3322 PH/58 Allow for the insertion of a newline as well as a space when a string
3323 is turned into more than one encoded-word during RFC 2047 encoding. The
3324 Sieve code now uses this.
3326 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3327 data_4xx, lost_connection, tls_required.
3329 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3330 message was being sent as an SMTP response.
3332 PH/61 Add -l and -k options to exicyclog.
3334 PH/62 When verifying, if an address was redirected to one new address, so that
3335 verification continued, and the new address failed or deferred after
3336 having set something in $address_data, the value of $address_data was not
3337 passed back to the ACL. This was different to the case when no
3338 redirection occurred. The value is now passed back in both cases.
3340 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3341 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3342 use login_cap.h, so on its own it isn't the distinguishing feature. The
3343 new name refers directly to the setclassresources() function.
3345 PH/65 Added configuration files for NetBSD3.
3347 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3349 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3350 is preferred over IPv4.
3352 PH/68 The bounce_return_message and bounce_return_body options were not being
3353 honoured for bounces generated during the reception of non-SMTP messages.
3354 In particular, this applied to messages rejected by the ACL. This bug has
3355 been fixed. However, if bounce_return_message is true and bounce_return_
3356 body is false, the headers that are returned for a non-SMTP message
3357 include only those that have been read before the error was detected.
3358 (In the case of an ACL rejection, they have all been read.)
3360 PH/69 The HTML version of the specification is now built in a directory called
3361 spec_html instead of spec.html, because the latter looks like a path with
3362 a MIME-type, and this confuses some software.
3364 PH/70 Catch two compiler warnings in sieve.c.
3366 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3367 function verify_get_ident() calls ip_connect() to connect a socket, but
3368 if the "connect()" function timed out, ip_connect() used to close the
3369 socket. However, verify_get_ident() also closes the socket later, and in
3370 between Exim writes to the log, which may get opened at this point. When
3371 the socket was closed in ip_connect(), the log could get the same file
3372 descriptor number as the socket. This naturally causes chaos. The fix is
3373 not to close the socket in ip_connect(); the socket should be closed by
3374 the function that creates it. There was only one place in the code where
3375 this was missing, in the iplookup router, which I don't think anybody now
3376 uses, but I've fixed it anyway.
3378 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3379 well as to direct DNS lookups. Otherwise the handling of names in host
3380 lists is inconsistent and therefore confusing.
3386 PH/01 Two changes to the default runtime configuration:
3388 (1) Move the checks for relay_from_hosts and authenticated clients from
3389 after to before the (commented out) DNS black list checks.
3391 (2) Add control=submission to the relay_from_hosts and authenticated
3392 clients checks, on the grounds that messages accepted by these
3393 statements are most likely to be submissions.
3395 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3397 (1) Generate an error if the third argument for the ${prvs expansion is
3400 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3403 (3) Reset the variables that are obtained from the first argument of
3404 ${prvscheck and used in the second argument before leaving the code,
3405 because their memory is reclaimed, so using them afterwards may do
3408 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3409 one (it's much easier than Tom thought :-).
3411 (5) Because of (4), we can now allow for the use of $prvscheck_result
3412 inside the third argument.
3414 PH/03 For some reason, the default setting of PATH when running a command from
3415 a pipe transport was just "/usr/bin". I have changed it to
3418 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3419 anything to be listed in the output from -bV.
3421 PH/05 When a filter generated an autoreply, the entire To: header line was
3422 quoted in the delivery log line, like this:
3424 => >A.N.Other <ano@some.domain> <original@ddress> ...
3426 This has been changed so that it extracts the operative address. There
3427 may be more than one such address. If so, they are comma-separated, like
3430 => >ano@some.domain,ona@other.domain <original@ddress> ...
3432 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3433 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3434 not being looked up in the rDNS to get a host name, Exim was showing the
3435 IP address twice in Received: lines, even though the IP addresses were
3436 identical. For example:
3438 Received: from [1.2.3.4] (helo=[1.2.3.4])
3440 However, if the real host name was known, it was omitting the HELO data
3441 if it matched the actual IP address. This has been tidied up so that it
3442 doesn't show the same IP address twice.
3444 PH/07 When both +timestamp and +memory debugging was on, the value given by
3445 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3446 called by the debug printing, thereby overwriting the timestamp buffer.
3447 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3449 PH/08 When the original message was included in an autoreply transport, it
3450 always said "this is a copy of the message, including all the headers",
3451 even if body_only or headers_only was set. It now gives an appropriate
3454 PH/09 Applied a patch from the Sieve maintainer which:
3456 o fixes some comments
3457 o adds the (disabled) notify extension core
3458 o adds some debug output for the result of if/elsif tests
3459 o points to the current vacation draft in the documentation
3460 and documents the missing references header update
3464 o fixes a bug in processing the envelope test (when testing
3465 multiple envelope elements, the last element determined the
3468 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3469 Electronic Mail") by including:
3471 Auto-submitted: auto-generated
3473 in the messages that it generates (bounce messages and others, such as
3474 warnings). In the case of bounce messages for non-SMTP messages, there was
3475 also a typo: it was using "Auto_submitted" (underscore instead of
3476 hyphen). Since every message generated by Exim is necessarily in response
3477 to another message, thes have all been changed to:
3479 Auto-Submitted: auto-replied
3481 in accordance with these statements in the RFC:
3483 The auto-replied keyword:
3485 - SHOULD be used on messages sent in direct response to another
3486 message by an automatic process,
3488 - MUST NOT be used on manually-generated messages,
3490 - MAY be used on Delivery Status Notifications (DSNs) and Message
3491 Disposition Notifications (MDNs),
3493 - MUST NOT be used on messages generated by automatic or periodic
3494 processes, except for messages which are automatic responses to
3497 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3498 to the default Received: header definition.
3500 PH/12 Added log selector acl_warn_skipped (default on).
3502 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3503 variables because (a) they are in the wrong storage pool and (b) even if
3504 they were copied, it wouldn't work properly because of the caching.
3506 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3507 checking when decoding. Apparently there are clients that generate
3508 overlong encoded strings. Why am I not surprised?
3510 PH/15 If the first argument of "${if match_address" was not empty, but did not
3511 contain an "@" character, Exim crashed. Now it writes a panic log message
3512 and treats the condition as false.
3514 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3516 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3517 extension "envelope-auth". The code is finished and in agreement with
3518 other implementations, but there is no documentation so far and in fact,
3519 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3520 not changing the active code.
3522 Print executed "if" and "elsif" statements when debugging is used. This
3523 helps a great deal to understand what a filter does.
3525 Document more things not specified clearly in RFC3028. I had all this
3526 sorted out, when out of a sudden new issues came to my mind. Oops."
3528 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3531 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3532 canonical form (as documented). However, after a host name lookup from
3533 the IP address, check_host() was doing a simple string comparison with
3534 addresses acquired from the DNS when checking that the found name did
3535 have the original IP as one of its addresses. Since any found IPv6
3536 addresses are likely to be in abbreviated form, the comparison could
3537 fail. Luckily, there already exists a function for doing the comparison
3538 by converting both addresses to binary, so now that is used instead of
3539 the text comparison.
3541 PH/20 There was another similar case to PH/19, when a complete host name was
3542 given in a host list; looking up its IP address could give an abbreviated
3543 form, whereas the current host's name might or might not be abbreviated.
3544 The same fix has been applied.
3550 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3551 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3554 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3555 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3557 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3559 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3560 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3561 Also, for "Auto-Submitted", treat anything other than "no" as
3562 non-personal, in accordance with RFC 3834. (Previously it treated
3563 anything starting "auto-" as non-personal.)
3565 TF/01 The control=submission/name=... option had a problem with syntax
3566 errors if the name included a slash character. The /name= option
3567 now slurps the rest of the string, so it can include any characters
3568 but it must come last in the list of options (after /sender_retain
3571 PH/05 Some modifications to the interface to the fake nameserver for the new
3579 TK/01 Added the "success_on_redirect" address verification option. See
3580 NewStuff for rationale and an example.
3582 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3584 PH/02 Patch to exigrep to allow it to work on syslog lines.
3586 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3587 fread() to skip over the body file's header line, because in Cygwin the
3588 header line is locked and is inaccessible.
3590 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3591 co-exist for some time) to make it clear that it is the Exim ID that is
3592 referenced, not the Message-ID: header line.
3594 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3595 string_format() function, because snprintf() does not exist on all
3598 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3599 ${stat: expansion item.
3601 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3602 protocol synchronization error", to keep the pedants happy.
3604 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3605 well as for IRIX systems, when gcc is being used. See the host.c source
3608 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3610 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3613 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3614 transport and to the smtp transport in LMTP mode.
3616 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3618 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3619 run a filter in a subprocess. This could lead to confusion in subsequent
3620 lookups in the parent process. There should also be a search_tidyup() at
3621 the end of the subprocess.
3623 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3624 only if the host matched helo_try_verify_hosts, which caused the
3625 verification to occur when the EHLO/HELO command was issued. The ACL just
3626 tested the remembered result. Now, if a previous verification attempt has
3627 not happened, "verify = helo" does it there and then.
3629 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3631 TK/03 Fix log output including CR from clamd.
3633 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3634 exist provoked a memory error which could cause a segfault.
3636 PH/15 Installed PCRE 6.2
3638 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3640 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3641 of the problem. Specifically, suggested +O2 rather than +O1 for the
3644 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3646 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3647 record, the port was not correctly logged when the outgoing_port log
3648 selector was set (it logged the transort's default port).
3650 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3651 fallback_hosts, and "hosts" in the smtp transport.
3653 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3654 host errors such as "Connection refused".
3656 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3657 authentication with radiusclient 0.4.9:
3659 - Error returned from rc_read_config was caught wrongly
3660 - Username/password not passed on to radius server due to wrong length.
3662 The presumption is that some radiusclient API changes for 4.51/PH/17
3663 were not taken care of correctly. The code is still untested by me (my
3664 Linux distribution still has 0.3.2 of radiusclient), but it was
3665 contributed by a Radius user.
3667 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3668 expanding the "port" option of the smtp transport.
3670 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3671 while reading a MIME header. Thanks to Tom Hughes for a patch.
3673 PH/24 Include config.h inside local_scan.h so that configuration settings are
3676 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3677 that in an ACL for RCPT (for example), you can examine exactly what was
3680 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3681 commands, but it was not correctly comparing the address with the actual
3682 client host address. Thus, it would show the EHLO address in Received:
3683 header lines when this was not necessary.
3685 PH/27 Added the % operator to ${eval:}.
3687 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3688 it should be ignoring failures (because with -C, for example, it has lost
3689 privilege). It wasn't ignoring creation failures other than "already
3692 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3695 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3696 because an input line was too long, either on its own, or by virtue of
3697 too many continuations, the temporary file was not being removed, and the
3698 return code was incorrect.
3700 PH/31 Missing "BOOL" in function definition in filtertest.c.
3702 PH/32 Applied Sieve patches from the maintainer.
3704 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3706 PH/33 Added "verify = not_blind".
3708 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3709 Local/Makefile (with some defaults set). These are used in built scripts
3710 such as exicyclog, but they have never been used in the exim_install
3711 script (though there are many overriding facilities there). I have
3712 arranged that the exim_install script now takes note of these two
3715 PH/35 Installed configuration files for Dragonfly.
3717 PH/36 When a locally submitted message by a trusted user did not contain a
3718 From: header, and the sender address was obtained from -f or from an SMTP
3719 MAIL command, and the trusted user did not use -F to supply a sender
3720 name, $originator_name was incorrectly used when constructing a From:
3721 header. Furthermore, $originator_name was used for submission mode
3722 messages from external hosts without From: headers in a similar way,
3723 which is clearly wrong.
3725 PH/37 Added control=suppress_local_fixups.
3727 PH/38 When log_selector = +received_sender was set, and the addition of the
3728 sender made the log line's construction buffer exactly full, or one byte
3729 less than full, an overflow happened when the terminating "\n" was
3732 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3733 when the result of a list match is failure because a DNS lookup failed.
3735 PH/40 RM_COMMAND is now used in the building process.
3737 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3738 the "build-* directories that it finds.
3740 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3741 address in a domain literal was a prefix of an interface address.
3743 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3744 when verifying a sender address, unless rewrite_headers is false.
3746 PH/44 Wrote a long comment about why errors_to addresses are verified as
3747 recipients, not senders.
3749 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3750 the ratelimit ACL was added.
3752 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3754 PH/46 Added extra information about PostgreSQL errors to the error string.
3756 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3757 avoiding the need to install special zones in a real server. This is
3758 backwards compatible; if it can't find the fake resolver, it drops back.
3759 Thus, both old and new test suites can be run.
3761 TF/02 Added util/ratelimit.pl
3763 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3764 clock is set back in time.
3766 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3767 Candler <B.Candler@pobox.com>.
3769 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3770 OK for addresses that are the result of redirections.
3772 PH/48 A number of further additions for the benefit of the new test suite,
3773 including a fake gethostbyname() that interfaces to the fake DNS resolver
3776 TF/06 The fix for widen_domains has also been applied to qualify_single and
3777 search_parents which are the other dnslookup options that can cause
3780 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3783 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3785 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3786 service" if it encountered an MX record with an empty target hostname.
3787 The message is now "an MX or SRV record indicated no SMTP service".
3789 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3790 if the DNS of the sending site is misconfigured. This is quite a
3791 common situation. This change restores the behaviour of treating a
3792 helo verification defer as a failure.
3794 PH/51 If self=fail was set on a router, the bounce message did not include the
3795 actual error message.
3801 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3803 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3804 command itself ended in error, the underlying message about the transport
3805 filter timeout was being overwritten with the pipe command error. Now the
3806 underlying error message should be appended to the second error message.
3808 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3810 PH/02 Reduce the amount of output that "make" produces by default. Full output
3811 can still be requested.
3813 PH/03 The warning log line about a condition test deferring for a "warn" verb
3814 was being output only once per connection, rather than after each
3815 occurrence (because it was using the same function as for successful
3816 "warn" verbs). This seems wrong, so I have changed it.
3818 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3819 it should not have, which might have caused a crash in the right
3820 circumstances, but probably never did.
3822 PH/04 Installed a modified version of Tony Finch's patch to make submission
3823 mode fix the return path as well as the Sender: header line, and to
3824 add a /name= option so that you can make the user's friendly name appear
3827 TF/03 Added the control = fakedefer ACL modifier.
3829 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3830 Mark Lowes for thorough testing.
3832 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3834 TK/03 Merged latest SRS patch from Miles Wilton.
3836 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3837 with the definition in sysexits.h (which is #included earlier).
3838 Fortunately, Exim does not actually use EX_OK. The code used to try to
3839 preserve the sysexits.h value, by assuming that macro definitions were
3840 scanned for macro replacements. I have been disabused of this notion,
3841 so now the code just undefines EX_OK before #including unistd.h.
3843 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3844 in the smtp transport. When a block could not be written in a single
3845 write() function, the timeout was being re-applied to each part-write.
3846 This seems wrong - if the receiver was accepting one byte at a time it
3847 would take for ever. The timeout is now adjusted when this happens. It
3848 doesn't have to be particularly precise.
3850 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3851 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3853 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3854 without a domain if the check to <postmaster@domain> fails.
3856 SC/01 Eximstats: added -xls and the ability to specify output files
3857 (patch written by Frank Heydlauf).
3859 SC/02 Eximstats: use FileHandles for outputting results.
3861 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3863 SC/04 Eximstats: fixed display of large numbers with -nvr option
3865 SC/05 Eximstats: fixed merging of reports with empty tables.
3867 SC/06 Eximstats: added the -include_original_destination flag
3869 SC/07 Eximstats: removed tabs and trailing whitespace.
3871 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3873 TK/06 MBOX spool code: Add real "From " MBOX separator line
3874 so the .eml file is really in mbox format (even though
3875 most programs do not really care). Patch from Alex Miller.
3877 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3878 The latter is generated from $received_to and is only set if the
3879 message has one envelope recipient. SA can use these headers,
3880 obviously out-of-the-box. Patch from Alex Miller.
3882 PH/08 The ${def test on a variable was returning false if the variable's
3883 value was "0", contrary to what the specification has always said!
3884 The result should be true unless the variable is empty.
3886 PH/09 The syntax error of a character other than { following "${if
3887 def:variable_name" (after optional whitespace) was not being diagnosed.
3888 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3889 accidental colon was present, for example, could give incorrect results.
3891 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3892 result is used (not including appendfile, where other changes are about
3895 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3896 This involved changing a lot of size variables from int to off_t. It
3897 should work with maildirs and everything.
3899 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3900 spamd dying while we are connected to it.
3902 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3905 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3906 function for BDB 4.3.
3908 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3910 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3911 the sources for PCRE have been re-arranged and more files are now
3914 PH/15 The code I had for printing potentially long long variables in PH/11
3915 above was not the best (it lost precision). The length of off_t variables
3916 is now inspected at build time, and an appropriate printing format (%ld
3917 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3918 to be "long long int" or "long int". This is needed for the internal
3919 formatting function string_vformat().
3921 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3922 the configuration file to be ":syslog", then the script "guesses" where
3923 the logs files are, rather than using the compiled in default. In our
3924 case the guess is not the same as the compiled default, so the script
3925 suddenly stopped working when I started to use syslog. The patch checks
3926 to see if log_file_path is "". If so, it attempts to read it from exim
3927 with no configuration file to get the compiled in version, before it
3928 falls back to the previous guessing code."
3930 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3931 implementing BATV in an Exim configuration. See NewStuff for the gory
3934 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3935 Makefile that are specific to HP-UX.
3937 PH/18 If the "use_postmaster" option was set for a recipient callout together
3938 with the "random" option, the postmaster address was used as the MAIL
3939 FROM address for the random test, but not for the subsequent recipient
3940 test. It is now used for both.
3942 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3943 patch removes a few documentation additions to RFC 3028, because the
3944 latest draft now contains them. It adds the new en;ascii-case comparator
3945 and a new error check for 8bit text in MIME parts. Comparator and
3946 require names are now matched exactly. I enabled the subaddress
3947 extension, but it is not well tested yet (read: it works for me)."
3949 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3950 rework some of the code of TK/09 above to avoid the hardwired use of
3951 "%lld" and "long long". Replaced the call to snprintf() with a call to
3954 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3955 records point to non-existent hosts", "retry timeout exceeded", and
3956 "retry time not reached for any host after a long failure period".
3958 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3959 experimental DomainKeys support:
3961 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3962 (2) On an error such as an illegally used "control", the wrong name for
3963 the control was given.
3965 These problems did NOT occur unless DomainKeys support was compiled.
3967 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3969 PH/24 Added ${if match_ip condition.
3971 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3972 databases so that it will be absolutely obvious if a crash occurs in the
3973 DB library. This is a regular occurrence (often caused by mis-matched
3976 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3977 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3978 picked up on a user's system that detects such things. There doesn't seem
3979 to be a gcc warning option for this - only an attribute that has to be
3980 put on the function's prototype. It seems that in Fedora Core 4 they have
3981 set this on a number of new functions. No doubt there will be more in due
3984 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3985 specify a transport. However, if an address that was verified by such a
3986 router was the subject of a callout, Exim crashed because it tried to
3987 read the rcpt_include_affixes from the non-existent transport. Now it
3988 just assumes that the setting of that option is false. This bug was
3989 introduced by 4.51/PH/31.
3991 PH/28 Changed -d+all to exclude +memory, because that information is very
3992 rarely of interest, but it makes the output a lot bigger. People tend to
3993 do -d+all out of habit.
3995 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3996 code in os-type was giving problems when libc.so lives in lib64, like on
3999 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4000 aren't the modern standard, and it seems that some systems' include files
4001 don't always have them. Exim was already checking for some of the newer
4002 ones like T_AAAA, and defining it itself. I've added checks for all the
4003 record types that Exim uses.
4005 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4006 not automatically generating a new one, as it is supposed to. This
4007 prevented TLS from working. If the file did exist, but contained invalid
4008 data, a new version was generated, as expected. It was only the case of a
4009 non-existent file that was broken.
4011 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4012 with a change in libdomainkeys > 0.64.
4014 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4015 from DNS. If the selector record carries the flag, it now has
4016 precedence over the domain-wide flag.
4018 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4020 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4021 the use of a port name that isn't defined in /etc/services) occurred, the
4022 message was deferred as in a normal delivery, and thus remained on the
4023 spool, instead of being failed because of the mua_wrapper setting. This
4024 is now fixed, and I tidied up some of the mua_wrapper messages at the
4027 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4028 the messages in a hash of arrays rather than using individual hashes.
4029 This is a bit cleaner and results in dramatic memory savings, albeit
4030 at a slight CPU cost.
4032 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4033 as requested by Marc Sherman.
4035 SC/10 Eximstats: added histograms for user specified patterns as requested
4038 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4040 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4041 fopen() in the content-scanning modules that did not already have it.
4047 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4048 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4050 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4052 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4054 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4055 to test Sieve filters that use "vacation".
4057 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4058 that changes the way the GnuTLS parameters are stored in the cache file.
4059 The new format can be generated externally. For backward compatibility,
4060 if the data in the cache doesn't make sense, Exim assumes it has read an
4061 old-format file, and it generates new data and writes a new file. This
4062 means that you can't go back to an older release without removing the
4065 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4066 work if there are any delivery delays because "one_time" forces the
4067 parent to be marked "delivered", so its unseen clone is never tried
4068 again. For this reason, Exim now forbids the simultaneous setting of
4071 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4072 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4073 entry if you want to know the details. The fix, however, neglected to
4074 consider the case when local delivery batching is involved. The test for
4075 "previously delivered" was not happening when checking to see if an
4076 address could be batched with a previous (undelivered) one; under
4077 certain circumstances this could lead to multiple deliveries to the same
4080 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4081 in its include files, and this causes problems building Exim.
4083 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4084 header_syntax) but Exim was just ignoring anything given after a slash.
4085 In particular, this caused confusion with an attempt to use "verify =
4086 reverse_host_lookup/defer_ok". An error is now given when options are
4087 supplied for verify items that do not have them. (Maybe reverse_host_
4088 lookup should have a defer_ok option, but that's a different point.)
4090 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4091 defined by RFC 821) to 2048, because there were problems with some AUTH
4092 commands, and RFC 1869 says the size should be increased for extended
4093 SMTP commands that take arguments.
4095 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4098 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4099 "unknown" error; now it says that the functionality isn't in the binary.
4101 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4102 an address' error message when a string expansion fails (syntax or
4103 whatever). Otherwise the password may appear in the log. Following change
4104 PH/42 below, there is no longer a chance of it appearing in a bounce
4107 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4109 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4110 of its addresses was ever tried. (Bugzilla bug #2.)
4112 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4113 the result incorrectly in the debug output. (It correctly added a newline
4114 to what was transported.)
4116 TF/01 Added $received_time.
4118 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4119 commented out examples of how to interface to a virus scanner and to
4120 SpamAssassin. Also added commented examples of av_scanner and
4121 spamd_address settings.
4123 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4124 and controls are allowed in which ACLs. There were a couple of minor
4125 errors. Some of the entries in the conditions table (which is a table of
4126 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4127 negation of where the condition IS allowed.
4129 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4131 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4132 header file does not have a version number, so I've had to invent a new
4133 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4134 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4135 radiusclient), but it was contributed by a Radius user.
4137 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4138 files or directories, for OpenSSL.
4140 PH/19 When an Exim process that is running as root has to create an Exim log
4141 file, it does so in a subprocess that runs as exim:exim so as to get the
4142 ownership right at creation (otherwise, other Exim processes might see
4143 the file with the wrong ownership). There was no test for failure of this
4144 fork() call, which would lead to the process getting stuck as it waited
4145 for a non-existent subprocess. Forks do occasionally fail when resources
4146 run out. I reviewed all the other calls to fork(); they all seem to check
4149 PH/20 When checking for unexpected SMTP input at connect time (before writing
4150 the banner), Exim was not dealing correctly with a non-positive return
4151 from the read() function. If the client had disconnected by this time,
4152 the result was a log entry for a synchronization error with an empty
4153 string after "input=" when read() returned zero. If read() returned -1
4154 (an event I could not check), uninitialized data bytes were printed.
4155 There were reports of junk text (parts of files, etc) appearing after
4158 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4160 PH/22 Added support for macro redefinition, and (re)definition in between
4161 driver and ACL definitions.
4163 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4164 forgetting to use the resulting value; it was using the unexpanded value.
4166 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4167 hadn't been configured. The fix is from Juergen Kreileder, who
4168 understands it better than I do:
4170 "Here's what I see happening with three configured cyrus_sasl
4171 authenticators configured (plain, login, cram-md5):
4173 On startup auth_cyrus_sasl_init() gets called for each of these.
4174 This means three calls to sasl_listmech() without a specified mech_list.
4175 => SASL tests which mechs of all available mechs actually work
4176 => three warnings about OTP not working
4177 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4179 With the patch, sasl_listmech() also gets called three times. But now
4180 SASL's mech_list option is set to the server_mech specified in the the
4181 authenticator. Or in other words, the answer from sasl_listmech()
4182 gets limited to just the mech you're testing for (which is different
4184 => the return list contains just 'plain' or 'login', 'cram-md5' or
4185 nothing depending on the value of ob->server_mech.
4187 I've just tested the patch: Authentication still works fine,
4188 unavailable mechs specified in the exim configuration are still
4189 caught, and the auth.log warnings about OTP are gone."
4191 PH/25 When debugging is enabled, the contents of the command line are added
4192 to the debugging output, even when log_selector=+arguments is not
4195 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4196 answer is "GNU", and only if the return is "GNU/something" is the answer
4199 PH/27 $acl_verify_message is now set immediately after the failure of a
4200 verification in an ACL, and so is available in subsequent modifiers. In
4201 particular, the message can be preserved by coding like this:
4203 warn !verify = sender
4204 set acl_m0 = $acl_verify_message
4206 Previously, $acl_verify_message was set only while expanding "message"
4207 and "log_message" when a very denied access.
4209 PH/28 Modified OS/os.c-Linux with
4211 -#ifndef OS_LOAD_AVERAGE
4212 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4214 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4215 nomenclature these days.)
4217 PH/29 Installed patch from the Sieve maintainer that adds the options
4218 sieve_useraddress and sieve_subaddress to the redirect router.
4220 PH/30 In these circumstances:
4221 . Two addresses routed to the same list of hosts;
4222 . First host does not offer TLS;
4223 . First host accepts first address;
4224 . First host gives temporary error to second address;
4225 . Second host offers TLS and a TLS session is established;
4226 . Second host accepts second address.
4227 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4228 and peerdn, if requested) that were in fact used only for the second
4231 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4232 attention to any local part prefix or suffix that was matched by the
4233 router that accepted the address. It now behaves in the same way as it
4234 does for delivery: the affixes are removed from the local part unless
4235 rcpt_include_affixes is set on the transport.
4237 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4238 timeout during the DATA phase of an incoming message.
4240 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4241 applied a patch sanctioned by the Sieve maintainer.
4243 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4244 the uid or gid is negative. A case of a negative gid caused this to be
4245 noticed. The fix allows for either to be negative.
4247 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4248 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4251 PH/36 The change PH/12 above was broken. Fixed it.
4253 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4254 the grounds that routing the same address twice would always produce the
4255 same answer. This might have been true once, but it is certainly no
4256 longer true now. Routing a child address may depend on the previous
4257 routing that produced that child. Some complicated redirection strategies
4258 went wrong when messages had multiple recipients, and made Exim's
4259 behaviour dependent on the order in which the addresses were given.
4261 I have moved the duplicate checking until after the routing is complete.
4262 Exim scans the addresses that are assigned to local and remote
4263 transports, and removes any duplicates. This means that more work will be
4264 done, as duplicates will always all be routed, but duplicates are
4265 presumably rare, so I don't expect this is of any significance.
4267 For deliveries to pipes, files, and autoreplies, the duplicate checking
4268 still happens during the routing process, since they are not going to be
4271 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4272 It corrects a timeout issue with spamd. This is Ian's comment: "The
4273 background is that sometimes spamd either never reads data from a
4274 connection it has accepted, or it never writes response data. The exiscan
4275 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4276 blindly assumes that writes won't block so it may never time out."
4278 PH/39 Allow G after quota size as well as K and M.
4280 PH/40 The value set for $authenticated_id in an authenticator may not contain
4281 binary zeroes or newlines because the value is written to log lines and
4282 to spool files. There was no check on this. Now the value is run through
4283 the string_printing() function so that such characters are converted to
4284 printable escape sequences.
4286 PH/41 $message_linecount is a new variable that contains the total number of
4287 lines in the message. Compare $body_linecount, which is the count for the
4290 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4291 bounce and delay warning messages, except in certain special cases, which
4294 (a) An SMTP error message from a remote host;
4295 (b) A message specified in a :fail: redirection;
4296 (c) A message specified in a "fail" command in a system filter;
4297 (d) A message specified in a FAIL return from the queryprogram router;
4298 (e) A message specified by the cannot_route_message router option.
4300 In these cases only, Exim does include the error details in bounce and
4301 warning messages. There are also a few cases where bland messages such
4302 as "unrouteable address" or "local delivery error" are given.
4304 PH/43 $value is now also set for the "else" part of a ${run expansion.
4306 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4307 being worked on, but at least Exim now implements the latest version to
4310 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4311 process to complete was treated as a delivery failure, a timeout while
4312 writing the message to the pipe was logged, but erroneously treated as a
4313 successful delivery. Such timeouts include transport filter timeouts. For
4314 consistency with the overall process timeout, these timeouts are now
4315 treated as errors, giving rise to delivery failures by default. However,
4316 there is now a new Boolean option for the pipe transport called
4317 timeout_defer, which, if set TRUE, converts the failures into defers for
4318 both kinds of timeout. A transport filter timeout is now identified in
4321 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4322 systems where "make" and "gmake" are different, calling "gmake" at top
4323 level broke things. I've arranged for the value of $(MAKE) to be passed
4324 from the Makefile to this script so that it can call the same version of
4328 A note about Exim versions 4.44 and 4.50
4329 ----------------------------------------
4331 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4332 changes of various kinds. As a consequence, a big documentation update was
4333 needed. This delayed the release for rather longer than seemed good, especially
4334 in the light of a couple of (minor) security issues. Therefore, the changes
4335 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4336 release. So 4.44 and 4.50 are in effect two different branches that both start
4339 I have left the 4.50 change log unchanged; it contains all the changes since
4340 4.43. The change log for 4.44 is below; many of its items are identical to
4341 those for 4.50. This seems to be the most sensible way to preserve the
4342 historical information.
4348 1. Minor wording change to the doc/README.SIEVE file.
4350 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4351 computation of the current number of files was incorrect.
4353 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4354 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4357 4. Give more explanation in the error message when the command for a transport
4358 filter fails to execute.
4360 5. There are several places where Exim runs a non-Exim command in a
4361 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4362 was being done only for the command run by the queryprogram router. It is
4363 now done for all such subprocesses. The other cases are: ${run, transport
4364 filters, and the commands run by the lmtp and pipe transports.
4366 6. Added CONFIGURE_GROUP build-time option.
4368 7. Some older OS have a limit of 256 on the maximum number of file
4369 descriptors. Exim was using setrlimit() to set 1000 as a large value
4370 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4371 systems. I've change it so that if it can't get 1000, it tries for 256.
4373 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4374 was an oversight, and furthermore, ever since the addition of extra
4375 controls (e.g. 4.43/32), the checks on when to allow different forms of
4376 "control" were broken. There should now be diagnostics for all cases when a
4377 control that does not make sense is encountered.
4379 9. Added the /retain_sender option to "control=submission".
4381 10. $recipients is now available in the predata ACL (oversight).
4383 11. Tidy the search cache before the fork to do a delivery from a message
4384 received from the command line. Otherwise the child will trigger a lookup
4385 failure and thereby defer the delivery if it tries to use (for example) a
4386 cached ldap connection that the parent has called unbind on.
4388 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4389 of $address_data from the recipient verification was clobbered by the
4390 sender verification.
4392 13. The value of address_data from a sender verification is now available in
4393 $sender_address_data in subsequent conditions in the ACL statement.
4395 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4397 15. Added a new option "connect=<time>" to callout options, to set a different
4400 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4401 was its contents. (It was OK if the option was not defined at all.)
4403 17. A "Completed" log line is now written for messages that are removed from
4404 the spool by the -Mrm option.
4406 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4407 information about exactly what failed.
4409 19. Added -dd to debug only the daemon process.
4411 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4412 handles timeouts, both on the server side and network timeouts. Renamed the
4413 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4415 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4416 It is now set to "smtps".
4418 22. $host_address is now set to the target address during the checking of
4419 ignore_target_hosts.
4421 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4422 being passed; this would have caused $sender_host_name to have been used if
4423 matching the list had actually called for a host name (not very likely,
4424 since this list is usually IP addresses). A host name is now passed as
4427 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4428 code that specifies a non-restarting handler (typically sigaction() in
4429 modern systems) in an attempt to fix a rare and obscure crash bug.
4431 25. Narrowed the window for a race in the daemon that could cause it to ignore
4432 SIGCHLD signals. This is not a major problem, because they are used only to
4433 wake it up if nothing else does.
4435 26. A malformed maildirsize file could cause Exim to calculate negative values
4436 for the mailbox size or file count. Odd effects could occur as a result.
4437 The maildirsize information is now recalculated if the size or filecount
4440 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4441 support for a long time. Removed HAVE_SYS_VFS_H.
4443 28. Installed the latest version of exipick from John Jetmore.
4445 29. In an address list, if the pattern was not a regular expression, an empty
4446 subject address (from a bounce message) matched only if the pattern was an
4447 empty string. Non-empty patterns were not even tested. This was the wrong
4448 because it is perfectly reasonable to use an empty address as part of a
4449 database query. An empty address is now tested by patterns that are
4450 lookups. However, all the other forms of pattern expect the subject to
4451 contain a local part and a domain, and therefore, for them, an empty
4452 address still always fails if the pattern is not itself empty.
4454 30. Exim went into a mad DNS loop when attempting to do a callout where the
4455 host was specified on an smtp transport, and looking it up yielded more
4456 than one IP address.
4458 31. Re-factored the code for checking spool and log partition space into a
4459 function that finds that data and another that does the check. The former
4460 is then used to implement four new variables: $spool_space, $log_space,
4461 $spool_inodes, and $log_inodes.
4463 32. The RFC2047 encoding function was originally intended for short strings
4464 such as real names; it was not keeping to the 75-character limit for
4465 encoded words that the RFC imposes. It now respects the limit, and
4466 generates multiple encoded words if necessary. To be on the safe side, I
4467 have increased the buffer size for the ${rfc2047: expansion operator from
4470 33. It is now permitted to omit both strings after an "if" condition; if the
4471 condition is true, the result is "true". As before, when the second string
4472 is omitted, a false condition yields an empty string. This makes it less
4473 cumbersome to write custom ACL and router conditions.
4475 34. Failure to deliver a bounce message always caused it to be frozen, even if
4476 there was an errors_to setting on the router. The errors_to setting is now
4479 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4480 canonical form (fully expanded) before being placed in
4481 $sender_host_address.
4483 36. The table in the code that translates DNS record types into text (T_A to
4484 "A" for instance) was missing entries for NS and CNAME. It is just possible
4485 that this could have caused confusion if both these types were looked up
4486 for the same domain, because the text type is used as part of Exim's
4487 per-process caching. But the chance of anyone hitting this buglet seems
4490 37. The dnsdb lookup has been extended in a number of ways.
4492 (1) There is a new type, "zns", which walks up the domain tree until it
4493 finds some nameserver records. It should be used with care.
4495 (2) There is a new type, "mxh", which is like "mx" except that it returns
4496 just the host names, not the priorities.
4498 (3) It is now possible to give a list of domains (or IP addresses) to be
4499 looked up. The behaviour when one of the lookups defers can be
4500 controlled by a keyword.
4502 (4) It is now possible to specify the separator character for use when
4503 multiple records are returned.
4505 38. The dnslists ACL condition has been extended: it is now possible to supply
4506 a list of IP addresses and/or domains to be looked up in a particular DNS
4509 39. Added log_selector=+queue_time_overall.
4511 40. When running the queue in the test harness, wait just a tad after forking a
4512 delivery process, to get repeatability of debugging output.
4514 41. Include certificate and key file names in error message when GnuTLS fails
4515 to set them up, because the GnuTLS error message doesn't include the name
4516 of the failing file when there is a problem reading it.
4518 42. Allow both -bf and -bF in the same test run.
4520 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4522 44. The "Exiscan patch" is now merged into the mainline Exim source.
4524 45. Sometimes the final signoff response after QUIT could fail to get
4525 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4526 < 0 before doing a fflush(). This bug looks as though it goes back to the
4527 introduction of TLS in release 3.20, but "sometimes" must have been rare
4528 because the tests only now provoked it.
4530 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4531 (this can affect the format of dates).
4533 47. exim_tidydb, when checking for the continued existence of a message for
4534 which it has found a message-specific retry record, was not finding
4535 messages that were in split spool directories. Consequently, it was
4536 deleting retry records that should have stayed in existence.
4538 48. Steve fixed some bugs in eximstats.
4540 49. The SPA authentication driver was not abandoning authentication and moving
4541 on to the next authenticator when an expansion was forced to fail,
4542 contradicting the general specification for all authenticators. Instead it
4543 was generating a temporary error. It now behaves as specified.
4545 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4546 (the order specifies the preference for clients). The order is now AES256,
4547 AES128, 3DES, ARCFOUR128.
4549 51. Small patch to Sieve code - explicitly set From: when generating an
4552 52. Exim crashed if a remote delivery caused a very long error message to be
4553 recorded - for instance if somebody sent an entire SpamAssassin report back
4554 as a large number of 550 error lines. This bug was coincidentally fixed by
4555 increasing the size of one of Exim's internal buffers (big_buffer) that
4556 happened as part of the Exiscan merge. However, to be on the safe side, I
4557 have made the code more robust (and fixed the comments that describe what
4560 53. Now that there can be additional text after "Completed" in log lines (if
4561 the queue_time_overall log selector is set), a one-byte patch to exigrep
4562 was needed to allow it to recognize "Completed" as not the last thing in
4565 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4566 patch that reportedly fixes this has been added. I am not expert enough to
4567 create a test for it. This is what the patch creator wrote:
4569 "I found a little strange behaviour of ldap code when working with
4570 Windows 2003 AD Domain, where users was placed in more than one
4571 Organization Units. When I tried to give exim partial DN, the exit code
4572 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4573 But simultaneously result of request was absolutely normal ldap result,
4574 so I produce this patch..."
4576 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4577 REFERENCE, so I have modified the code to exclude the patch when that macro
4580 55. Some experimental protocols are using DNS PTR records for new purposes. The
4581 keys for these records are domain names, not reversed IP addresses. The
4582 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4583 leaves it alone. Component reversal etc. now happens only for IP addresses.
4586 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4588 57. Double the size of the debug message buffer (to 2048) so that more of very
4589 long debug lines gets shown.
4591 58. The exicyclog utility now does better if the number of log files to keep
4592 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4594 59. Two changes related to the smtp_active_hostname option:
4596 (1) $smtp_active_hostname is now available as a variable.
4597 (2) The default for smtp_banner uses $smtp_active_hostname instead
4598 of $primary_hostname.
4600 60. The host_aton() function is supposed to be passed a string that is known
4601 to be a valid IP address. However, in the case of IPv6 addresses, it was
4602 not checking this. This is a hostage to fortune. Exim now panics and dies
4603 if the condition is not met. A case was found where this could be provoked
4604 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4605 components; fortuitously, this particular loophole had already been fixed
4606 by change 4.50/55 above.
4608 If there are any other similar loopholes, the new check in host_aton()
4609 itself should stop them being exploited. The report I received stated that
4610 data on the command line could provoke the exploit when Exim was running as
4611 exim, but did not say which command line option was involved. All I could
4612 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4613 running as the user.
4616 61. There was a buffer overflow vulnerability in the SPA authentication code
4617 (which came originally from the Samba project). I have added a test to the
4618 spa_base64_to_bits() function which I hope fixes it.
4621 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4622 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4624 63. The daemon start-up calls getloadavg() while still root for those OS that
4625 need the first call to be done as root, but it missed one case: when
4626 deliver_queue_load_max is set with deliver_drop_privilege. This is
4627 necessary for the benefit of the queue runner, because there is no re-exec
4628 when deliver_drop_privilege is set.
4630 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4631 This has been fixed.
4633 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4634 was in use, was not putting the data itself into the right store pool;
4635 consequently, it could be overwritten for a subsequent message in the same
4636 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4639 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4641 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4642 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4643 the function were treating the return as a boolean value, which happened to
4644 work because 0=false and not-0=true, but is not correct code.
4646 68. The host_aton() function was not handling scoped IPv6 addresses (those
4647 with, for example, "%eth0" on the end) correctly.
4649 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4650 negated items (that is, ~something) in unsigned ints. Some compilers
4651 apparently mutter when there is no cast.
4653 70. If an address verification called from an ACL failed, and did not produce a
4654 user-specific message (i.e. there was only a "system" message), nothing was
4655 put in $acl_verify_message. In this situation, it now puts the system
4658 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4659 session; change 4.31/43 added the unwanted input to the log line - except
4660 that it did not do this in the start of session case. It now does.
4662 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4663 This is wrong and can cause the other end to generate a synchronization
4664 error if it is another Exim or anything else that does the synchronization
4665 check. A QUIT command is no longer sent after a timeout.
4667 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4668 during host lookups.
4670 74. The defer_ok option of callout verification was not working if it was used
4671 when verifying addresses in header lines, that is, for this case:
4673 verify = header_sender/callout=defer_ok
4675 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4676 those file descriptors could be used for SMTP connections. If anything
4677 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4678 could be sent to the SMTP client, causing chaos. The daemon now opens
4679 stdin, stdout, and stderr to /dev/null when it puts itself into the
4682 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4683 log by default. The user can override this with suitable Perl magic.
4685 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4686 the log message when discard triggers, was not working for the DATA ACL or
4687 for the non-SMTP ACL.
4689 78. Error message wording change in sieve.c.
4691 79. If smtp_accept_max_per_host was set, the number of connections could be
4692 restricted to fewer than expected, because the daemon was trying to set up
4693 a new connection before checking whether the processes handling previous
4694 connections had finished. The check for completed processes is now done
4695 earlier. On busy systems, this bug wouldn't be noticed because something
4696 else would have woken the daemon, and it would have reaped the completed
4699 80. If a message was submitted locally by a user whose login name contained one
4700 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4701 It caused a spool format error. I have fixed the spool reading code. A
4702 related problem was that the "from" clause in the Received: line became
4703 illegal because of the space(s). It is now covered by ${quote_local_part.
4705 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4708 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4710 83. Patch from Sieve maintainer to fix unterminated string problem in
4711 "vacation" handling.
4713 84. Some minor changes to the Linux configuration files to help with other
4714 OS variants using glibc.
4716 85. One more patch for Sieve to update vacation handling to latest spec.
4719 ----------------------------------------------------
4720 See the note above about the 4.44 and 4.50 releases.
4721 ----------------------------------------------------
4727 1. Change 4.43/35 introduced a bug that caused file counts to be
4728 incorrectly computed when quota_filecount was set in an appendfile
4731 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4732 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4735 3. Give more explanation in the error message when the command for a transport
4736 filter fails to execute.
4738 4. There are several places where Exim runs a non-Exim command in a
4739 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4740 was being done only for the command run by the queryprogram router. It is
4741 now done for all such subprocesses. The other cases are: ${run, transport
4742 filters, and the commands run by the lmtp and pipe transports.
4744 5. Some older OS have a limit of 256 on the maximum number of file
4745 descriptors. Exim was using setrlimit() to set 1000 as a large value
4746 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4747 systems. I've change it so that if it can't get 1000, it tries for 256.
4749 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4750 was an oversight, and furthermore, ever since the addition of extra
4751 controls (e.g. 4.43/32), the checks on when to allow different forms of
4752 "control" were broken. There should now be diagnostics for all cases when a
4753 control that does not make sense is encountered.
4755 7. $recipients is now available in the predata ACL (oversight).
4757 8. Tidy the search cache before the fork to do a delivery from a message
4758 received from the command line. Otherwise the child will trigger a lookup
4759 failure and thereby defer the delivery if it tries to use (for example) a
4760 cached ldap connection that the parent has called unbind on.
4762 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4763 of $address_data from the recipient verification was clobbered by the
4764 sender verification.
4766 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4767 was its contents. (It was OK if the option was not defined at all.)
4769 11. A "Completed" log line is now written for messages that are removed from
4770 the spool by the -Mrm option.
4772 12. $host_address is now set to the target address during the checking of
4773 ignore_target_hosts.
4775 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4776 being passed; this would have caused $sender_host_name to have been used if
4777 matching the list had actually called for a host name (not very likely,
4778 since this list is usually IP addresses). A host name is now passed as
4781 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4782 code that specifies a non-restarting handler (typically sigaction() in
4783 modern systems) in an attempt to fix a rare and obscure crash bug.
4785 15. Narrowed the window for a race in the daemon that could cause it to ignore
4786 SIGCHLD signals. This is not a major problem, because they are used only to
4787 wake it up if nothing else does.
4789 16. A malformed maildirsize file could cause Exim to calculate negative values
4790 for the mailbox size or file count. Odd effects could occur as a result.
4791 The maildirsize information is now recalculated if the size or filecount
4794 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4795 support for a long time. Removed HAVE_SYS_VFS_H.
4797 18. Updated exipick to current release from John Jetmore.
4799 19. Allow an empty sender to be matched against a lookup in an address list.
4800 Previously the only cases considered were a regular expression, or an
4803 20. Exim went into a mad DNS lookup loop when doing a callout where the
4804 host was specified on the transport, if the DNS lookup yielded more than
4807 21. The RFC2047 encoding function was originally intended for short strings
4808 such as real names; it was not keeping to the 75-character limit for
4809 encoded words that the RFC imposes. It now respects the limit, and
4810 generates multiple encoded words if necessary. To be on the safe side, I
4811 have increased the buffer size for the ${rfc2047: expansion operator from
4814 22. Failure to deliver a bounce message always caused it to be frozen, even if
4815 there was an errors_to setting on the router. The errors_to setting is now
4818 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4819 canonical form (fully expanded) before being placed in
4820 $sender_host_address.
4822 24. Updated eximstats to version 1.33
4824 25. Include certificate and key file names in error message when GnuTLS fails
4825 to set them up, because the GnuTLS error message doesn't include the name
4826 of the failing file when there is a problem reading it.
4828 26. Expand error message when OpenSSL has problems setting up cert/key files.
4831 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4832 (this can affect the format of dates).
4834 28. exim_tidydb, when checking for the continued existence of a message for
4835 which it has found a message-specific retry record, was not finding
4836 messages that were in split spool directories. Consequently, it was
4837 deleting retry records that should have stayed in existence.
4839 29. eximstats updated to version 1.35
4840 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4841 1.35 - bugfix such that pie charts by volume are generated correctly
4843 30. The SPA authentication driver was not abandoning authentication and moving
4844 on to the next authenticator when an expansion was forced to fail,
4845 contradicting the general specification for all authenticators. Instead it
4846 was generating a temporary error. It now behaves as specified.
4848 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4849 (the order specifies the preference for clients). The order is now AES256,
4850 AES128, 3DES, ARCFOUR128.
4852 31. Small patch to Sieve code - explicitly set From: when generating an
4855 32. Exim crashed if a remote delivery caused a very long error message to be
4856 recorded - for instance if somebody sent an entire SpamAssassin report back
4857 as a large number of 550 error lines. This bug was coincidentally fixed by
4858 increasing the size of one of Exim's internal buffers (big_buffer) that
4859 happened as part of the Exiscan merge. However, to be on the safe side, I
4860 have made the code more robust (and fixed the comments that describe what
4863 33. Some experimental protocols are using DNS PTR records for new purposes. The
4864 keys for these records are domain names, not reversed IP addresses. The
4865 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4866 leaves it alone. Component reversal etc. now happens only for IP addresses.
4869 34. The host_aton() function is supposed to be passed a string that is known
4870 to be a valid IP address. However, in the case of IPv6 addresses, it was
4871 not checking this. This is a hostage to fortune. Exim now panics and dies
4872 if the condition is not met. A case was found where this could be provoked
4873 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4874 components; fortuitously, this particular loophole had already been fixed
4875 by change 4.50/55 or 4.44/33 above.
4877 If there are any other similar loopholes, the new check in host_aton()
4878 itself should stop them being exploited. The report I received stated that
4879 data on the command line could provoke the exploit when Exim was running as
4880 exim, but did not say which command line option was involved. All I could
4881 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4882 running as the user.
4885 35. There was a buffer overflow vulnerability in the SPA authentication code
4886 (which came originally from the Samba project). I have added a test to the
4887 spa_base64_to_bits() function which I hope fixes it.
4890 36. The daemon start-up calls getloadavg() while still root for those OS that
4891 need the first call to be done as root, but it missed one case: when
4892 deliver_queue_load_max is set with deliver_drop_privilege. This is
4893 necessary for the benefit of the queue runner, because there is no re-exec
4894 when deliver_drop_privilege is set.
4896 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4897 was in use, was not putting the data itself into the right store pool;
4898 consequently, it could be overwritten for a subsequent message in the same
4899 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4902 38. Sometimes the final signoff response after QUIT could fail to get
4903 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4904 < 0 before doing a fflush(). This bug looks as though it goes back to the
4905 introduction of TLS in release 3.20, but "sometimes" must have been rare
4906 because the tests only now provoked it.
4912 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4913 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4914 Now it returns an integer. A number of calls were still expecting a T/F
4915 return. Fortuitously, in all cases, the tests worked in OK situations,
4916 which is the norm. However, things would have gone wrong on any write
4917 failures on the smtp file descriptor. This function is used when sending
4918 messages over SMTP and also when doing verify callouts.
4920 2. When Exim is called to do synchronous delivery of a locally submitted
4921 message (the -odf or -odi options), it no longer closes stderr before doing
4924 3. Implemented the mua_wrapper option.
4926 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4928 5. Implemented the functions header_remove(), header_testname(),
4929 header_add_at_position(), and receive_remove_recipient(), and exported them
4932 6. If an ACL "warn" statement specified the addition of headers, Exim already
4933 inserted X-ACL-Warn: at the start if there was no header name. However, it
4934 was not making this test for the second and subsequent header lines if
4935 there were newlines in the string. This meant that an invalid header could
4936 be inserted if Exim was badly configured.
4938 7. Allow an ACL "warn" statement to add header lines at the start or after all
4939 the Received: headers, as well as at the end.
4941 8. Added the rcpt_4xx retry error code.
4943 9. Added postmaster_mailfrom=xxx to callout verification option.
4945 10. Added mailfrom=xxxx to the callout verification option, for verify=
4948 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4949 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4951 12. Inserted some casts to stop certain compilers warning when using pointer
4952 differences as field lengths or precisions in printf-type calls (mostly
4953 affecting debugging statements).
4955 13. Added optional readline() support for -be (dynamically loaded).
4957 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4958 same clock tick as a message's arrival, so that its received time was the
4959 same as the "first fail" time on the retry record, and that message
4960 remained on the queue past the ultimate address timeout, every queue runner
4961 would try a delivery (because it was past the ultimate address timeout) but
4962 after another failure, the ultimate address timeout, which should have then
4963 bounced the address, did not kick in. This was a "< instead of <=" error;
4964 in most cases the first failure would have been in the next clock tick
4965 after the received time, and all would be well.
4967 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4968 being recognized when the domain list was tested by the match_domain
4969 condition in an expansion string.
4971 16. Added the ${str2b64: operator.
4973 17. Exim was always calling setrlimit() to set a large limit for the number of
4974 processes, without checking whether the existing limit was already
4975 adequate. (It did check for the limit on file descriptors.) Furthermore,
4976 errors from getrlimit() and setrlimit() were being ignored. Now they are
4977 logged to the main and panic logs, but Exim does carry on, to try to do its
4978 job under whatever limits there are.
4980 18. Imported PCRE 5.0.
4982 19. Trivial typo in log message " temporarily refused connection" (the leading
4985 20. If the log selector return_path_on_delivery was set and an address was
4986 redirected to /dev/null, the delivery process crashed because it assumed
4987 that a return path would always be set for a "successful" delivery. In this
4988 case, the whole delivery is bypassed as an optimization, and therefore no
4991 21. Internal re-arrangement: the function for sending a challenge and reading
4992 a response while authentication was assuming a zero-terminated challenge
4993 string. It's now changed to take a pointer and a length, to allow for
4994 binary data in such strings.
4996 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4998 23. Exim was not respecting finduser_retries when seeking the login of the
4999 uid under which it was called; it was always trying 10 times. (The default
5000 setting of finduser_retries is zero.) Also, it was sleeping after the final
5001 failure, which is pointless.
5003 24. Implemented tls_on_connect_ports.
5005 25. Implemented acl_smtp_predata.
5007 26. If the domain in control=submission is set empty, Exim assumes that the
5008 authenticated id is a complete email address when it generates From: or
5009 Sender: header lines.
5011 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5012 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5013 chown and chgrp in /bin and hostname in /usr/bin.
5015 28. Exim was keeping the "process log" file open after each use, just as it
5016 does for the main log. This opens the possibility of it remaining open for
5017 long periods when the USR1 signal hits a daemon. Occasional processlog
5018 errors were reported, that could have been caused by this. Anyway, it seems
5019 much more sensible not to leave this file open at all, so that is what now
5022 29. The long-running daemon process does not normally write to the log once it
5023 has entered its main loop, and it closes the log before doing so. This is
5024 so that log files can straightforwardly be renamed and moved. However,
5025 there are a couple of unusual error situations where the daemon does write
5026 log entries, and I had neglected to close the log afterwards.
5028 30. The text of an SMTP error response that was received during a remote
5029 delivery was being truncated at 512 bytes. This is too short for some of
5030 the long messages that one sometimes sees. I've increased the limit to
5033 31. It is now possible to make retry rules that apply only when a message has a
5034 specific sender, in particular, an empty sender.
5036 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5037 it possible to be selective about when SMTP synchronization is enforced.
5039 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5041 32. Implemented hosts_connection_nolog.
5043 33. Added an ACL for QUIT.
5045 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5048 35. Added mailbox_size and mailbox_filecount to appendfile.
5050 36. Added control = no_multiline_responses to ACLs.
5052 37. There was a bug in the logic of the code that waits for the clock to tick
5053 in the case where the clock went backwards by a substantial amount such
5054 that the microsecond fraction of "now" was more than the microsecond
5055 fraction of "then" (but the whole seconds number was less).
5057 38. Added support for the libradius Radius client library this is found on
5058 FreeBSD (previously only the radiusclient library was supported).
5064 1. When certain lookups returned multiple values in the form name=value, the
5065 quoting of the values was not always being done properly. Specifically:
5066 (a) If the value started with a double quote, but contained no whitespace,
5068 (b) If the value contained whitespace other than a space character (i.e.
5069 tabs or newlines or carriage returns) it was not quoted.
5070 This fix has been applied to the mysql and pgsql lookups by writing a
5071 separate quoting function and calling it from the lookup code. The fix
5072 should probably also be applied to nisplus, ibase and oracle lookups, but
5073 since I cannot test any of those, I have not disturbed their existing code.
5075 2. A hit in the callout cache for a specific address caused a log line with no
5076 reason for rejecting RCPT. Now it says "Previous (cached) callout
5077 verification failure".
5079 3. There was an off-by-one bug in the queryprogram router. An over-long
5080 return line was truncated at 256 instead of 255 characters, thereby
5081 overflowing its buffer with the terminating zero. As well as fixing this, I
5082 have increased the buffer size to 1024 (and made a note to document this).
5084 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5085 when Exim is waiting for an SMTP response from a remote server, Exim
5086 restarts its select() call on the socket, thereby resetting its timeout.
5087 This is not a problem when such interrupts are rare. Somebody set up a cron
5088 job to run exiwhat every 2 minutes, which is less than the normal select()
5089 timeout (5 or 10 minutes). This meant that the select() timeout never
5090 kicked in because it was always reset. I have fixed this by comparing the
5091 time when an interrupt arrives with the time at the start of the first call
5092 to select(). If more time than the timeout has elapsed, the interrupt is
5093 treated as a timeout.
5095 5. Some internal re-factoring in preparation for the addition of Sieve
5096 extensions (by MH). In particular, the "personal" test is moved to a
5097 separate function, and given an option for scanning Cc: and Bcc: (which is
5098 not set for Exim filters).
5100 6. When Exim created an email address using the login of the caller as the
5101 local part (e.g. when creating a From: or Sender: header line), it was not
5102 quoting the local part when it contained special characters such as @.
5104 7. Installed new OpenBSD configuration files.
5106 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5107 try to make them clearer.
5109 9. Callout options, other than the timeout value, were being ignored when
5110 verifying sender addresses in header lines. For example, when using
5112 verify = header_sender/callout=no_cache
5114 the cache was (incorrectly) being used.
5116 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5117 only the Cygwin environment.
5119 11. When return_path_on_delivery was set as a log selector, if different remote
5120 addresses in the same message used different return paths and parallel
5121 remote delivery occurred, the wrong values would sometimes be logged.
5122 (Whenever a remote delivery process finished, the return path value from
5123 the most recently started remote delivery process was logged.)
5125 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5126 lines when AUTH and/or TLS are in use. This is the "received protocol"
5127 field. Exim used to use "asmtp" for authenticated SMTP, without any
5128 indication (in the protocol name) for TLS use. Now it follows the RFC and
5129 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5130 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5131 names appear in log lines as well as in Received: header lines.
5133 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5134 extensions, and comparison tests, and to fix some bugs.
5136 14. Changes to the "personal" filter test:
5138 (1) The test was buggy in that it was just doing the equivalent of
5139 "contains" tests on header lines. For example, if a user's address was
5140 anne@some.where, the "personal" test would incorrectly be true for
5142 To: susanne@some.where
5144 This test is now done by extracting each address from the header in turn,
5145 and checking the entire address. Other tests that are part of "personal"
5146 are now done using regular expressions (for example, to check local parts
5147 of addresses in From: header lines).
5149 (2) The list of non-personal local parts in From: addresses has been
5150 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5151 taken from the Sieve specification recommendations.
5153 (3) If the message contains any header line starting with "List-" it is
5154 treated as non-personal.
5156 (4) The test for "circular" in the Subject: header line has been removed
5157 because it now seems ill-conceived.
5159 15. Minor typos in src/EDITME comments corrected.
5161 16. Installed latest exipick from John Jetmore.
5163 17. If headers_add on a router specified a text string that was too long for
5164 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5165 of string_sprintf() is now avoided.
5167 18. $message_body_size was not set (it was always zero) when running the DATA
5168 ACL and the local_scan() function.
5170 19. For the "mail" command in an Exim filter, no default was being set for
5171 the once_repeat time, causing a random time value to be used if "once" was
5172 specified. (If the value happened to be <= 0, no repeat happened.) The
5173 default is now 0s, meaning "never repeat". The "vacation" command was OK
5174 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5175 (I found it when inspecting the code).
5177 20. There is now an overall timeout for performing a callout verification. It
5178 defaults to 4 times the callout timeout, which applies to individual SMTP
5179 commands during the callout. The overall timeout applies when there is more
5180 than one host that can be tried. The timeout is checked before trying the
5181 next host. This prevents very long delays if there are a large number of
5182 hosts and all are timing out (e.g. when the network connections are timing
5183 out). The value of the overall timeout can be changed by specifying an
5184 additional sub-option for "callout", called "maxwait". For example:
5186 verify = sender/callout=5s,maxwait=20s
5188 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5189 to the end before writing, but this should make it even safer).
5191 22. Exim was forgetting that it had advertised PIPELINING for the second and
5192 subsequent messages on an SMTP connection. It was also not resetting its
5193 memory on STARTTLS and an internal HELO.
5195 23. When Exim logs an SMTP synchronization error within a session, it now
5196 records whether PIPELINING has been advertised or not.
5198 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5199 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5200 rather than long int.
5202 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5204 26. Added the never_mail option to autoreply.
5210 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5211 crash if the getsockname() call failed; this can happen if a connection is
5212 closed very soon after it is established. The problem was simply in the
5213 order in which certain operations were done, causing Exim to try to write
5214 to the SMTP stream before it had set up the file descriptor. The bug has
5215 been fixed by making things happen in the correct order.
5221 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5222 before the connection was closed, thus losing the rejection response.
5224 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5225 some early Solaris releases, but causes trouble in current releases where
5226 socklen_t is defined.
5228 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5231 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5234 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5235 that releasing the top of it at the end releases what was used for sub-
5236 expansions (unless the block got too big). However, discard this block if
5237 the first thing is a variable or header, so that we can use its block when
5238 it is dynamic (useful for very large $message_headers, for example).
5240 6. Lookups now cache *every* query, not just the most recent. A new, separate
5241 store pool is used for this. It can be recovered when all lookup caches are
5242 flushed. Lookups now release memory at the end of their result strings.
5243 This has involved some general refactoring of the lookup sources.
5245 7. Some code has been added to the store_xxx() functions to reduce the amount
5246 of flapping under certain conditions.
5248 8. log_incoming_interface used to affect only the <= reception log lines. Now
5249 it causes the local interface and port to be added to several more SMTP log
5250 lines, for example "SMTP connection from", and rejection lines.
5252 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5254 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5256 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5257 could be overwritten at the end of the current message (or the start of a
5258 new message if it was set in a HELO ACL). The value is now preserved for
5259 the duration of the SMTP connection.
5261 12. If a transport had a headers_rewrite setting, and a matching header line
5262 contained an unqualified address, that address was qualified, even if it
5263 did not match any rewriting rules. The underlying bug was that the values
5264 of the flags that permit the existence of unqualified sender and recipient
5265 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5266 non-local messages, and by -bnq for local messages) were not being
5267 preserved with the message after it was received.
5269 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5270 "next input=" as part of the text comprising the host identity instead of
5271 the correct text. The code was using the same buffer for two different
5272 strings. However, depending on which order the printing function evaluated
5273 its arguments, the bug did not always show up. Under Linux, for example, my
5274 test suite worked just fine.
5276 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5277 doesn't work with some older versions of Perl. It has been changed to "my",
5278 which in any case is probably the better facility to use.
5280 15. A really picky compiler found some instances of statements for creating
5281 error messages that either had too many or two few arguments for the format
5284 16. The size of the buffer for calls to the DNS resolver has been increased
5285 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5286 for addresses that have a lot of PTR records. This alleviates a problem; it
5287 does not fully solve it.
5289 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5290 buffer now truncates the list and logs the incident, which is the same
5291 action as happens when Exim is looking up a host name and its aliases.
5292 Previously in this situation something unpredictable would happen;
5293 sometimes it was "internal error: store_reset failed".
5295 18. If a server dropped the connection unexpectedly when an Exim client was
5296 using GnuTLS and trying to read a response, the client delivery process
5297 crashed while trying to generate an error log message.
5299 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5300 string, for example:
5302 warn message = H1: something\nH2: something
5304 the text was added as a single header line from Exim's point of view
5305 though it ended up OK in the delivered message. However, searching for the
5306 second and subsequent header lines using $h_h2: did not work. This has been
5307 fixed. Similarly, if a system filter added multiple headers in this way,
5308 the routers could not see them.
5310 20. Expanded the error message when iplsearch is called with an invalid key to
5311 suggest using net-iplsearch in a host list.
5313 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5314 ACLs are no longer actually imposed (and a message to that effect is
5317 22. If a "gecos" field in a passwd entry contained escaped characters, in
5318 particular, if it contained a \" sequence, Exim got it wrong when building
5319 a From: or a Sender: header from that name. A second bug also caused
5320 incorrect handling when an unquoted " was present following a character
5321 that needed quoting.
5323 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5324 was not being matched caselessly.
5326 24. Arranged for all hyphens in the exim.8 source to be escaped with
5329 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5330 properly. Recipient callout cache records were still being keyed to include
5331 the sender, even when use_sender was set false. This led to far more
5332 callouts that were necessary. The sender is no longer included in the key
5333 when use_sender is false.
5335 26. Added "control = submission" modifier to ACLs.
5337 27. Added the ${base62d: operator to decode base 62 numbers.
5339 28. dnsdb lookups can now access SRV records.
5341 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5342 the configuration file.
5344 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5345 (-v) mode. This makes the output for a verbose queue run more intelligible.
5347 31. Added a use_postmaster feature to recipient callouts.
5349 32. Added the $body_zerocount variable, containing the number of binary zero
5350 bytes in the message body.
5352 33. The time of last modification of the "new" subdirectory is now used as the
5353 "mailbox time last read" when there is a quota error for a maildir
5356 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5358 35. Added +ignore_unknown as a special item in host lists.
5360 36. Code for decoding IPv6 addresses in host lists is now included, even if
5361 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5362 address was recognized as an IP address, but was then not correctly decoded
5363 into binary, causing unexpected and incorrect effects when compared with
5370 1. Very minor rewording of debugging text in manualroute to say "list of
5371 hosts" instead of "hostlist".
5373 2. If verify=header_syntax was set, and a header line with an unqualified
5374 address (no domain) and a large number of spaces between the end of the
5375 name and the colon was received, the reception process suffered a buffer
5376 overflow, and (when I tested it) crashed. This was caused by some obsolete
5377 code that should have been removed. The fix is to remove it!
5379 3. When running in the test harness, delay a bit after writing a bounce
5380 message to get a bit more predictability in the log output.
5382 4. Added a call to search_tidyup() just before forking a reception process. In
5383 theory, someone could use a lookup in the expansion of smtp_accept_max_
5384 per_host which, without the tidyup, could leave open a database connection.
5386 5. Added the variables $recipient_data and $sender_data which get set from a
5387 lookup success in an ACL "recipients" or "senders" condition, or a router
5388 "senders" option, similar to $domain_data and $local_part_data.
5390 6. Moved the writing of debug_print from before to after the "senders" test
5393 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5394 problems for message scanning, either using a data ACL, or using
5395 local_scan() because the Received: header was not generated till after they
5396 were called (in order to set the time as the time of reception completion).
5397 I have revised the way this works. The header is now generated after the
5398 body is received, but before the ACL or local_scan() are called. After they
5399 are run, the timestamp in the header is updated.
5405 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5406 before starting a queue runner without re-exec. This happened only when
5407 deliver_drop_privilege was set or when the Exim user was set to root. The
5408 effect of the bug was that timeouts during subsequent deliveries caused
5409 crashes instead of being properly handled. The handler is now left at its
5410 default (and expected) setting.
5412 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5413 message, again when deliver_drop_privilege is set or Exim is run as root.
5414 The bug described in (1) was not present in this case, but the tidying up
5415 of the other signals was missing. I have made the two cases consistent.
5417 3. The ignore_target_hosts setting on a manualroute router was being ignored
5418 for hosts that were looked up using the /MX notation.
5420 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5423 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5424 operated on the sender address. After changing the $sender_address to <>
5425 for the sender address verify, Exim was re-instated it as the original
5426 (before rewriting) address, but remembering that it had rewritten it, so it
5427 wasn't rewriting it again. This bug also had the effect of breaking the
5428 sender address verification caching when the sender address was rewritten.
5430 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5431 This has been changed so that if the ip literal address matches
5432 ignore_target_hosts, the router declines.
5434 7. Added expansion conditions match_domain, match_address, and match_local_
5435 part (NOT match_host).
5437 8. The placeholder for the Received: header didn't have a length field set.
5439 9. Added code to Exim itself and to exim_lock to test for a specific race
5440 condition that could lead to file corruption when using MBX delivery. The
5441 issue is with the lockfile that is created in /tmp. If this file is removed
5442 after a process has opened it but before that process has acquired a lock,
5443 there is the potential for a second process to recreate the file and also
5444 acquire a lock. This could lead to two Exim processes writing to the file
5445 at the same time. The added code performs the same test as UW imapd; it
5446 checks after acquiring the lock that its file descriptor still refers to
5447 the same named file.
5449 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5450 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5453 11. Added the smtp_active_hostname option. If used, this will typically be made
5454 to depend on the incoming interface address. Because $interface_address is
5455 not set up until the daemon has forked a reception process, error responses
5456 that can happen earlier (such as "too many connections") no longer contain
5459 12. If an expansion in a condition on a "warn" statement fails because a lookup
5460 defers, the "warn" statement is abandoned, and the next ACL statement is
5461 processed. Previously this caused the whole ACL to be aborted.
5463 13. Added the iplsearch lookup type.
5465 14. Added ident_timeout as a log selector.
5467 15. Added tls_certificate_verified as a log selector.
5469 16. Added a global option tls_require_ciphers (compare the smtp transport
5470 option of the same name). This controls incoming TLS connections.
5472 17. I finally figured out how to make tls_require_ciphers do a similar thing
5473 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5474 before starting the TLS session.
5476 18. Tabs are now shown as \t in -bP output.
5478 19. If the log selector return_path_on_delivery was set, Exim crashed when
5479 bouncing a message because it had too many Received: header lines.
5481 20. If two routers both had headers_remove settings, and the first one included
5482 a superfluous trailing colon, the final name in the first list and the
5483 first name in the second list were incorrectly joined into one item (with a
5484 colon in the middle).
5490 1. Added -C and -D options to the exinext utility, mainly to make it easier
5491 to include in the automated testing, but these could be helpful when
5492 multiple configurations are in use.
5494 2. The exinext utility was not formatting the output nicely when there was
5495 an alternate port involved in the retry record key, nor when there was a
5496 message id as well (for retries that were specific to a specific message
5497 and a specific host). It was also confused by IPv6 addresses, because of
5498 the additional colons they contain. I have fixed the IPv4 problem, and
5499 patched it up to do a reasonable job for IPv6.
5501 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5502 delivery, the log line now contains "pipelined" if PIPELINING was used.
5504 4. An SMTP transport process used to panic and die if the bind() call to set
5505 an explicit outgoing interface failed. This has been changed; it is now
5506 treated in the same way as a connect() failure.
5508 5. A reference to $sender_host_name in the part of a conditional expansion
5509 that was being skipped was still causing a DNS lookup. This no longer
5512 6. The def: expansion condition was not recognizing references to header lines
5513 that used bh_ and bheader_.
5515 7. Added the _cache feature to named lists.
5517 8. The code for checking quota_filecount in the appendfile transport was
5518 allowing one more file than it should have been.
5520 9. For compatibility with Sendmail, the command line option
5528 and sets the incoming protocol and host name (for trusted callers). The
5529 host name and its colon can be omitted when only the protocol is to be set.
5530 Note the Exim already has two private options, -pd and -ps, that refer to
5531 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5532 "s", but I don't think that's a major issue.
5534 10. A number of refactoring changes to the code, none of which should affect
5537 (a) The number of logging options was getting close to filling up the
5538 32-bit word that was used as a bit map. I have split them into two classes:
5539 those that are passed in the argument to log_write(), and those that are
5540 only ever tested independently outside of that function. These are now in
5541 separate 32-bit words, so there is plenty of room for expansion again.
5542 There is no change in the user interface or the logging behaviour.
5544 (b) When building, for example, log lines, the code previously used a
5545 macro that called string_cat() twice, in order to add two strings. This is
5546 not really sufficiently general. Furthermore, there was one instance where
5547 it was actually wrong because one of the argument was used twice, and in
5548 one call a function was used. (As it happened, calling the function twice
5549 did not affect the overall behaviour.) The macro has been replaced by a
5550 function that can join an arbitrary number of extra strings onto a growing
5553 (c) The code for expansion conditions now uses a table and a binary chop
5554 instead of a serial search (which was left over from when there were very
5555 few conditions). Also, it now recognizes conditions like "pam" even when
5556 the relevant support is not compiled in: a suitably worded error message is
5557 given if an attempt is made to use such a condition.
5559 11. Added ${time_interval:xxxxx}.
5561 12. A bug was causing one of the ddress fields not to be passed back correctly
5562 from remote delivery subprocesses. The field in question was not being
5563 subsequently used, so this caused to problems in practice.
5565 13. Added new log selectors queue_time and deliver_time.
5567 14. Might have fixed a bug in maildirsizefile handling that threw up
5568 "unexpected character" debug warnings, and recalculated the data
5569 unnecessarily. In any case, I expanded the warning message to give more
5572 15. Added the message "Restricted characters in address" to the statements in
5573 the default ACL that block characters like @ and % in local parts.
5575 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5576 Three changes have been made:
5578 (a) There was a serious bug; a negative response to MAIL caused the whole
5579 recipient domain to be cached as invalid, thereby blocking all messages
5580 to all local parts at the same domain, from all senders. This bug has
5581 been fixed. The domain is no longer cached after a negative response to
5582 MAIL if the sender used is not empty.
5584 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5587 (c) A new callout option, "use_sender" has been added for people who want
5588 the modified behaviour.
5594 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5597 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5598 indeed breaks things for older releases.
5600 3. Added additional logging to the case where there is a problem reading data
5601 from a filter that is running in a subprocess using a pipe, in order to
5602 try to track down a specific problem.
5604 4. Testing facility fudge: when running in the test harness and attempting
5605 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5606 getting "No route to host". Convert this to a timeout.
5608 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5611 6. Some OS don't have socklen_t but use size_t instead. This affects the
5612 fifth argument of getsockopt() amongst other things. This is now
5613 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5614 can be set for individual OS. I have set it for SunOS5, OSF1, and
5615 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5616 some earlier ones do not.
5618 7. Change 4.30/15 was not doing the test caselessly.
5620 8. The standard form for an IPv6 address literal was being rejected by address
5621 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5622 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5623 this, as well as the form without the "IPv6" on the front (but only when
5624 address literals are enabled, of course).
5626 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5628 10. Exim crashed if a message with an empty sender address specified by -f
5629 encountered a router with an errors_to setting. This could be provoked only
5630 by a command such as
5634 where an empty string was supplied; "<>" did not hit this bug.
5636 11. Installed PCRE release 4.5.
5638 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5639 remained set. It is now erased.
5641 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5642 times from message ids (which are base 36 rather than the normal 62).
5644 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5645 were being counted as actual protocol errors, and logged if the log
5646 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5647 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5648 and DATA following a set of rejected RCPTs do not count as protocol errors.
5649 In other words, Exim assumes they were pipelined, though this may not
5650 actually be the case. Of course, in all cases the client gets an
5651 appropriate error code.
5653 15. If a lookup fails in an ACL condition, a message about the failure may
5654 be available; it is used if testing the ACL cannot continue, because most
5655 such messages specify what the cause of the deferral is. However, some
5656 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5657 that caused an old message to be retained and used if a later statement
5658 caused a defer, replacing the real cause of the deferral.
5660 16. If an IP address had so many PTR records that the DNS lookup buffer
5661 was not large enough to hold them, Exim could crash while trying to process
5662 the truncated data. It now detects and logs this case.
5664 17. Further to 4.21/58, another change has been made: if (and only if) the
5665 first line of a message (the first header line) ends with CRLF, a bare LF
5666 in a subsequent header line has a space inserted after it, so as not to
5667 terminate the header.
5669 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5670 unnecessarily, used atoi() instead of strtol(), and didn't check the
5671 termination when getting file sizes from file names by regex.
5673 19. Completely re-implemented the support for maildirsize files, in the light
5674 of a number of problems with the previous contributed implementation
5675 (4.30/29). In particular:
5677 . If the quota is zero, the maildirsize file is maintained, but no quota is
5680 . If the maildir directory does not exist, it is created before any attempt
5681 to write a maildirsize file.
5683 . The quota value in the file is just a cache; if the quota is changed in
5684 the transport, the new value overrides.
5686 . A regular expression is available for excluding directories from the
5689 20. The autoreply transport checks the characters in options that define the
5690 message's headers; it allows continued headers, but it was checking with
5691 isspace() after an embedded newline instead of explicitly looking for a
5694 21. If all the "regular" hosts to which an address was routed had passed their
5695 expiry times, and had not reached their retry times, the address was
5696 bounced, even if fallback hosts were defined. Now Exim should go on to try
5699 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5700 equivalent code in the SMTP transport. Some hosts send humungous responses
5701 to HELO/EHLO, more than 1024 it seems.
5703 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5704 gives compiler warnings in some environments. I've now done it "properly",
5707 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5708 (because of problems with the built-in one) was declared to return uschar *
5709 instead of char *, causing compiler failure.
5711 25. Fixed a file descriptor leak when processing alias/forward files.
5713 26. Fixed a minor format string issue in dbfn.c.
5715 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5717 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5718 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5721 29. When a sender address is verified, it is cached, to save repeating the test
5722 when there is more than one recipient in a message. However, when the
5723 verification involves a callout, it is possible for different callout
5724 options to be set for different recipients. It is too complicated to keep
5725 track of this in the cache, so now Exim always runs a verification when a
5726 callout is required, relying on the callout cache for the optimization.
5727 The overhead is duplication of the address routing, but this should not be
5730 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5731 to be verified with callout=postmaster, and the main callout worked but the
5732 postmaster check failed, the verification correctly failed. However, if a
5733 subsequent RCPT command asked for sender verification *without* the
5734 postmaster check, incorrect caching caused this verification also to fail,
5737 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5738 it was not caching the DNS options (qualify_single, search_parents) that
5739 were used when the lookup failed. A subsequent lookup with different
5740 options therefore always gave the same answer, though there were cases
5741 where it should not have. (Example: a "domains = !$mx_any" option on a
5742 dnslookup router: the "domains" option is always processed without any
5743 widening, but the router might have qualify_single set.) Now Exim uses the
5744 cached value only when the same options are set.
5746 32. Added John Jetmore's "exipick" utility to the distribution.
5748 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5749 than a timeout (e.g. a certificate is required, and is not provided), an
5750 Exim server now closes the connection immediately. Previously it waited for
5751 the client to close - but if the client is SSL, it seems that they each
5752 wait for each other, leading to a delay before one of them times out.
5754 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5755 maintained 0.8.x compatibility because I don't think many are using it, and
5756 it is clearly obsolete.
5758 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5761 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5762 client certificate was expired. A simple patch fixes this, though I don't
5763 understand the full logic of why the verify callback is called multiple
5766 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5767 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5768 which causes problems with some clients (such as the Certicom SSL Plus
5769 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5770 disables the coutermeasure allowing Eudora to connect."
5772 38. Exim was not checking that a write() to a log file succeeded. This could
5773 lead to Bad Things if a log got too big, in particular if it hit a file
5774 size limit. Exim now panics and dies if it cannot write to a log file, just
5775 as it does if it cannot open a log file.
5777 39. Modified OS/Makefile-Linux so that it now contains
5779 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5781 The two -D definitions ensure that Exim is compiled with large file
5782 support, which makes it possible to handle log files that are bigger than
5785 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5786 instance) a domain was checked against a named list that involved a lookup,
5787 causing $domain_data to be set, then another domain was checked against the
5788 same list, then the first domain was re-checked, the value of $domain_data
5789 after the final check could be wrong. In particular, if the second check
5790 failed, it could be set empty. This bug probably also applied to
5793 41. The strip_trailing_dot option was not being applied to the address given
5794 with the -f command-line option.
5796 42. The code for reading a message's header from the spool was incrementing
5797 $received_count, but never initializing it. This meant that the value was
5798 incorrect (doubled) while delivering a message in the same process in which
5799 it was received. In the most common configuration of Exim, this never
5800 happens - a fresh exec is done - but it can happen when
5801 deliver_drop_privilege is set.
5803 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5804 it now includes up to 150 characters of the unexpected data in the log
5807 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5808 and building data strings. The size of both of these buffers was 10 000
5809 bytes - far larger than anybody would *ever* want, thought I. Needless to
5810 say, somebody hit the limit. I have increased the maximum line length to
5811 20 000 and the maximum data length of concatenated lines to 100 000. I have
5812 also fixed two bugs, because there was no checking on these buffers. Tsk,
5813 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5814 buffer is too small.
5816 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5817 lsearch lookups. Now it does.
5819 46. When parsing a route_list item in a manualroute router, a fixed-length
5820 buffer was used for the list of hosts. I made this 1024 bytes long,
5821 thinking that nobody would ever have a list of hosts that long. Wrong.
5822 Somebody had a whole pile of complicated expansion conditions, and the
5823 string was silently truncated, leading to an expansion error. It turns out
5824 that it is easier to change to an unlimited length (owing to other changes
5825 that have happened since this code was originally written) than to build
5826 structure for giving a limitation error. The length of the item that
5827 expands into the list of hosts is now unlimited.
5829 47. The lsearch lookup could not handle data where the length of text line was
5830 more than 4095 characters. Such lines were truncated, leading to shortened
5831 data being returned. It should now handle lines of any length.
5833 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5834 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5837 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5838 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5839 no longer happens in comment lines. A list of replacements is now placed
5840 at the head of all of the source files, except those whose only change is
5841 to replace PERL_COMMAND in the very first #! line.
5843 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5844 messages on the queue, with a bottom-up merge sort, using code contributed
5845 by Michael Haardt. This should make operations like -bp somewhat faster on
5846 large queues. It won't affect queue runners, except when queue_run_in_order
5849 51. Installed eximstats 1.31 in the distribution.
5851 52. Added support for SRV lookups to the dnslookup router.
5853 53. If an ACL referred to $message_body or $message_body_end, the value was not
5854 reset for any messages that followed in the same SMTP session.
5856 54. The store-handling optimization for building very long strings was not
5857 differentiating between the different store pools. I don't think this
5858 actually made any difference in practice, but I've tidied it.
5860 55. While running the routers to verify a sender address, $sender_address
5861 was still set to the sender address. This is wrong, because when routing to
5862 send a bounce to the sender, it would be empty. Therefore, I have changed
5863 it so that, while verifying a sender address, $sender_address is set to <>.
5864 (There is no change to what happens when verifying a recipient address.)
5866 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5867 target A or AAAA records (if not already returned) without resetting the
5868 qualify_single or search_parents options of the DNS resolver. These are
5869 inappropriate in this case because the targets of MX and SRV records must
5870 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5871 target that, when qualified, matched something in the local domain. These
5872 two options are now turned off when doing these lookups.
5874 57. It seems that at least some releases of Reiserfs (which does not have the
5875 concept of a fixed number of inodes) returns zero and not -1 for the
5876 number of available inodes. This interacted badly with check_spool_inodes,
5877 which assumed that -1 was the "no such thing" setting. What I have done is
5878 to check that the total number of inodes is greater than zero before doing
5879 the test of how many are available.
5881 58. When a "warn" ACL statement has a log_message modifier, the message is
5882 remembered, and not repeated. This is to avoid a lot of repetition when a
5883 message has many recipients that cause the same warning to be written.
5884 However, Exim was preserving the list of already written lines for an
5885 entire SMTP session, which doesn't seem right. The memory is now reset if a
5886 new message is started.
5888 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5889 debugging output unless log_rewrite was also set.
5891 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5892 of (int)(handle) when we know that handle contains (void *)(-1).
5894 61. The Exim daemon panic-logs an error return when it closes the incoming
5895 connection. However "connection reset by peer" seems to be common, and
5896 isn't really an error worthy of noting specially, so that particular error
5899 62. When Exim is trying to find all the local interfaces, it used to panic and
5900 die if the ioctl to get the interface flags failed. However, it seems that
5901 on at least one OS (Solaris 9) it is possible to have an interface that is
5902 included in the list of interfaces, but for which you get a failure error
5903 for this call. This happens when the interface is not "plumbed" into a
5904 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5905 failure of the "get flags" call assumes that the interface is down.
5907 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5908 makes life easier for people who are doing arithmetic on fields extracted
5909 from dates, where you often get leading zeros that should not be
5910 interpreted as octal.
5912 64. Added qualify_domain to the redirect router, to override the global
5915 65. If a pathologically long header line contained very many addresses (the
5916 report of this problem mentioned 10 000) and each of them was rewritten,
5917 Exim could use up a very large amount of memory. (It kept on making new
5918 copies of the header line as it rewrote, and never released the old ones.)
5919 At the expense of a bit more processing, the header rewriting function has
5920 been changed so that it no longer eats memory in this way.
5922 66. The generation of the Received: header has been moved from the time that a
5923 message starts to be received, to the time that it finishes. The timestamp
5924 in the Received: header should now be very close to that of the <= log
5925 line. There are two side-effects of this change:
5927 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5928 logged header lines no longer include the local Received: line, because
5929 it has not yet been created. The same applies to a copy of the message
5930 that is returned to a non-SMTP sender when a message is rejected.
5932 (b) When a filter file is tested using -bf, no additional Received: header
5933 is added to the test message. After some thought, I decided that this
5936 This change does not affect the value of $received_for. It is still set
5937 after address rewriting, but before local_scan() is called.
5939 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5941 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5942 gave an unhelpful panic error message, and a defer error. I have managed to
5943 change this behaviour so that it now rejects any supplied certificate,
5944 which seems right, as the list of acceptable certificates is empty.
5946 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5947 gave an unhelpful defer error. I have not managed to make this reject any
5948 supplied certificates, but the error message it gives is "no certificate
5949 supplied", which is not helpful.
5951 70. exigrep's output now also includes lines that are not associated with any
5952 message, but which match the given pattern. Implemented by a patch from
5953 Martin Sluka, which also tidied up the Perl a bit.
5955 71. Recipient callout verification, like sender verification, was using <> in
5956 the MAIL FROM command. This isn't really the right thing, since the actual
5957 sender may affect whether the remote host accepts the recipient or not. I
5958 have changed it to use the actual sender in the callout; this means that
5959 the cache record is now keyed on a recipient/sender pair, not just the
5960 recipient address. There doesn't seem to be a real danger of callout loops,
5961 since a callout by the remote host to check the sender would use <>.
5962 [SEE ABOVE: changed after hitting problems.]
5964 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5965 temporary errors. However, in the case of such a code being given after
5966 the end of a data transmission (i.e. after ".") Exim was failing to write
5967 a retry record for the message. (Yes, there was some broken host that was
5968 actually sending 8xx at this point.)
5970 73. An unknown lookup type in a host list could cause Exim to panic-die when
5971 the list was checked. (An example that provoked this was putting <; in the
5972 middle of a list instead of at the start.) If this happened during a DATA
5973 ACL check, a -D file could be left lying around. This kind of configuration
5974 error no longer causes Exim to die; instead it causes a defer error. The
5975 incident is still logged to the main and panic logs.
5977 74. Buglet left over from Exim 3 conversion. The message "too many messages
5978 in one connection" was written to the rejectlog but not the mainlog, except
5979 when address rewriting (yes!) was being logged.
5981 75. Added write_rejectlog option.
5983 76. When a system filter was run not as root (that is, when system_filter_user
5984 was set), the values of the $n variables were not being returned to the
5985 main process; thus, they were not subsequently available in the $sn
5988 77. Added +return_path_on_delivery log selector.
5990 78. A connection timeout was being treated differently from recipients deferred
5991 when testing hosts_max_try with a message that was older than the host's
5992 retry timeout. (The host should not be counted, thus allowing all hosts to
5993 be tried at least once before bouncing.) This may have been the cause of an
5994 occasionally reported bug whereby a message would remain on the queue
5995 longer than the retry timeout, but would be bounced if a delivery was
5996 forced. I say "may" because I never totally pinned down the problem;
5997 setting up timeout/retry tests is difficult. See also the next item.
5999 79. The ultimate address timeout was not being applied to errors that involved
6000 a combination of host plus message (for example, a timeout on a MAIL
6001 command). When an address resolved to a number of possible hosts, and they
6002 were not all tried for each delivery (e.g. because of hosts_max_try), a
6003 message could remain on the queue longer than the retry timeout.
6005 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6008 81. Fixed an obscure SMTP outgoing bug which required at least the following
6009 conditions: (a) there was another message waiting for the same server;
6010 (b) the server returned 5xx to all RCPT commands in the first message so
6011 that the message was not completed; (c) the server dropped the connection
6012 or gave a negative response to the RSET that Exim sends to abort the
6013 transaction. The observed case was a dropped connection after DATA that had
6014 been sent in pipelining mode. That is, the server had advertised PIPELINING
6015 but was not implementing it correctly. The effect of the bug was incorrect
6016 behaviour, such as trying another host, and this could lead to a crash.
6022 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6023 and daemon.c were passed as pointers to ints; they should have been
6024 pointers to socklen_t variables (which are typically unsigned ints).
6026 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6029 3. Fixed a really odd bug that affected only the testing scheme; patching a
6030 certain fixed string in the binary changed the value of another string that
6031 happened to be identical to the end of the original first string.
6033 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6034 name", it returns that address as the IP address. On some operating
6035 systems (e.g. Solaris), it also passes back the IP address string as the
6036 "host name". However, on others (e.g. Linux), it passes back an empty
6037 string. Exim wasn't checking for this, and was changing the host name to an
6038 empty string, assuming it had been canonicalized.
6040 5. Although rare, it is permitted to have more than one PTR record for a given
6041 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6042 all the names associated with an address, because they do in Solaris.
6043 However, it seems that they do not in Linux for data that comes from the
6044 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6045 I found this out when I moved to a new Linux workstation and tried to run
6046 the Exim test suite.
6048 To get round this problem I have changed the code so that it now does its
6049 own call to the DNS to look up PTR records when searching for a host name.
6050 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6051 addresses that are only in /etc/hosts are still found.
6053 This behaviour is, however, controlled by an option called host_lookup_
6054 order, which defaults to "bydns:byaddr". If people want to use the other
6055 order, or indeed, just use one or the other means of lookup, they can
6056 specify it in this variable.
6058 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6059 some operating systems, this comes back from gethostbyaddr() as an empty
6060 string, and this is what Exim used to test for. However, it seems that in
6061 other systems, "." is yielded. Exim now tests for this case too.
6063 7. The values of check_spool_space and check_log_space are now held internally
6064 as a number of kilobytes instead of an absolute number of bytes. If a
6065 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6066 kilobyte. This means that much larger values can be stored.
6068 8. Exim monitor: an attempt to get the action menu when not actually pointing
6069 at a message produces an empty menu entitled "No message selected". This
6070 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6071 no entries in it ("Shell widget menu has zero width and/or height"). So I
6072 have added a single, blank menu entry in this case.
6074 9. Added ${quote_local_part.
6076 10. MIME decoding is now applied to the contents of Subject: header lines when
6079 11. Now that a reference to $sender_host_address automatically causes a reverse
6080 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6081 host lookup before query-style lookups in lists that might use this
6082 variable. This has therefore been abolished, and the "net-" prefix is no
6083 longer necessary for query-style lookups.
6085 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6086 have been a typo for LFLAGS, so it has been changed.
6088 13. The install script calls Exim with "-C /dev/null" in order to find the
6089 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6090 to be output. However, since Exim outputs its version number before the
6091 error, it didn't break the script. It just looked ugly. I fixed this by
6092 always allowing "-C /dev/null" if the caller is root.
6094 14. Ignore overlarge ACL variable number when reading spool file - insurance
6095 against a later release with more variables having written the file.
6097 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6098 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6099 this, as well as the form without the "IPv6" on the front.
6101 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6102 OS/Makefile-Darwin file.
6104 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6105 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6107 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6108 variable. However, some OS use ints for inodes. I've added cast to long int
6109 to get rid of the compiler warning.
6111 19. I had forgotten to lock out "/../" in configuration file names when
6112 ALT_CONFIG_PREFIX was set.
6114 20. Routers used for verification do not need to specify transports. However,
6115 if such a router generated a host list, and callout was configured, Exim
6116 crashed, because it could not find a port number from the (non-existent)
6117 transport. It now assumes port 25 in this circumstance.
6119 21. Added the -t option to exigrep.
6121 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6122 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6123 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6124 them is set, LOOKUP_LSEARCH is forced.
6126 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6128 24. Added sender and host information to the "rejected by local_scan()" log
6129 line; previously there was no indication of these.
6131 25. Added .include_if_exists.
6133 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6134 Linux. It turns out that not all file systems support this. Apparently some
6135 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6136 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6137 not supported on the file descriptor, is now ignored when Exim is trying to
6138 sync a directory. This applies only to Linux.
6140 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6142 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6143 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6146 29. Added support for maildirsize files from supplied patch (modified a bit).
6148 30. The use of :fail: followed by an empty string could lead Exim to respond to
6149 sender verification failures with (e.g.):
6151 550 Verification failed for <xxx>
6152 550 Sender verify failed
6154 where the first response line was missing the '-' that indicates it is not
6155 the final line of the response.
6157 31. The loop for finding the name of the user that called Exim had a hardwired
6158 limit of 10; it now uses the value of finduser_retries, which is used for
6159 all other user lookups.
6161 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6164 33. Exim was neglecting to zero errno before one call of strtol() when
6165 expanding a string and expecting an integer value. On some systems this
6166 resulted in spurious "integer overflow" errors. Also, it was casting the
6167 result into an int without checking.
6169 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6170 did not work. The code looks as if it has *never* worked, though it appears
6171 to have been documented since at least release 1.62. I have made it work.
6173 35. The "timeout_DNS" error in retry rules, also documented since at least
6174 1.62, also never worked. As it isn't clear exactly what this means, and
6175 clearly it isn't a major issue, I have abolished the feature by treating it
6176 as "timeout", and writing a warning to the main and panic logs.
6178 36. The display of retry rules for -brt wasn't always showing the error code
6181 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6182 timeout_connect_A, timeout_connect_MX.
6184 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6185 to the empty sender.
6187 39. The daemon was not analysing the content of -oX till after it had closed
6188 stderr and disconnected from the controlling terminal. This meant that any
6189 syntax errors were only noted on the panic log, and the return code from
6190 the command was 0. By re-arranging the code a little, I've made the
6191 decoding happen first, so such errors now appear on stderr, and the return
6192 code is 1. However, the actual setting up of the sockets still happens in
6193 the disconnected process, so errors there are still only recorded on the
6196 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6197 connections (as happens on some IP stacks) was logged at start up time as
6198 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6199 it from "IPv6 and IPv4", which means that two separate sockets are being
6202 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6203 says whether AF_INET or AF_INET6 was passed as an argument.
6205 42. Exiwhat output was messed up when time zones were included in log
6208 43. Exiwhat now gives more information about the daemon's listening ports,
6209 and whether -tls-on-connect was used.
6211 44. The "port" option of the smtp transport is now expanded.
6213 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6214 silently ignored. Now an error message is written to the main and panic
6217 46. There's a new ACL modifier called "logwrite" which writes to a log file
6218 as soon as it is encountered.
6220 47. Added $local_user_uid and $local_user_gid at routing time.
6222 48. Exim crashed when trying to verify a sender address that was being
6225 49. Exim was recognizing only a space character after ".include". It now also
6226 recognizes a tab character.
6228 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6229 extracting the relevant information from the specification. The man page no
6230 longer contains scrambled data for the -d option, and I've added a section
6231 at the front about calling Exim under different names.
6233 51. Added "extra_headers" argument to the "mail" command in filter files.
6235 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6238 53. Installed eximstats 1.29.
6240 54. Added transport_filter_timeout as a generic transport option.
6242 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6243 Cc: header lines. This was required by RFC 822, but it not required by RFC
6246 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6247 incoming messages that did not have them. Now it does so only if the
6248 message originates locally, that is, if there is no associated remote host
6249 address. When Resent- header lines are present, this applies to the Resent-
6250 lines rather than the non-Resent- lines.
6252 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6253 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6255 58. Messages for configuration errors now include the name of the main
6256 configuration file - useful now that there may be more than one file in a
6257 list (.included file names were always shown).
6259 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6260 for those rare installations that do not start the daemon as root or run it
6261 setuid root. I've cut out the call to initgroups() if the daemon is not
6264 60. The Exim user and group can now be bound into the binary as text strings
6265 that are looked up at the start of Exim's processing.
6267 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6269 62. Added $mailstore_basename variable.
6271 63. Installed patch to sieve.c from Michael Haardt.
6273 64. When Exim failed to open the panic log after failing to open the main log,
6274 the original message it was trying to log was written to stderr and debug
6275 output, but if they were not available (the usual case in production), it
6276 was lost. Now it is written to syslog before the two lines that record the
6277 failures to open the logs.
6279 65. Users' Exim filters run in subprocesses under the user's uid. It is
6280 possible for a "deliver" command or an alias in a "personal" command to
6281 provoke an address rewrite. If logging of address rewriting is configured,
6282 this fails because the process is not running as root or exim. There may be
6283 a better way of dealing with this, but for the moment (because 4.30 needs
6284 to be released), I have disabled address rewrite logging when running a
6285 filter in a non-root, non-exim process.
6291 1. The buildconfig auxiliary program wasn't quoting the value set for
6292 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6293 not defined. This bug was masked in 4.22 by the effect that was fixed in
6296 2. Some messages that were rejected after a message id was allocated were
6297 shown as "incomplete" by exigrep. It no longer does this for messages that
6298 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6300 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6301 have allow_domain_literals set, the ID did not get logged in the <= line.
6302 Domain literals are now always recognized in Message-ID: header lines.
6304 4. The first argument for a ${extract expansion item is the key name or field
6305 number. Leading and trailing spaces in this item were not being ignored,
6306 causing some misleading effects.
6308 5. When deliver_drop_privilege was set, single queue runner processes started
6309 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6310 same command in the process it spins off) were not dropping privilege.
6312 6. When the daemon running as "exim" started a queue runner, it always
6313 re-executed Exim in the spun-off process. This is a waste of effort when
6314 deliver_drop_privilege is set. The new process now just calls the
6315 queue-runner function directly.
6321 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6324 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6325 ignored. Though the use of -oP was forcing the writing of a pid file, it
6326 was always written to the default place.
6328 3. If the message "no IP address found for host xxxx" is generated during
6329 incoming verification, it is now followed by identification of the incoming
6330 connection (so you can more easily find what provoked it).
6332 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6334 5. Added some features to "harden" Exim a bit more against certain attacks:
6336 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6337 be put in Local/Makefile. This is like the never_users runtime option,
6338 but it cannot be overridden. The default setting is "root".
6340 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6341 prefix string with which any file named in a -C command line option
6344 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6345 is retained for -C and -D only if the caller of Exim is root. Without
6346 it, the exim user may also use -C and -D and retain privilege.
6348 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6349 command line option is disabled.
6351 6. Macro names set by the -D option must start with an upper case letter, just
6352 like macro names defined in the configuration file.
6354 7. Added "dereference=" facility to LDAP.
6356 8. Two instances of the typo "uknown" in the source files are fixed.
6358 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6359 the Configure-Makefile script screwed up while processing it.
6361 10. Incorporated PCRE 4.4.
6363 11. The SMTP synchronization check was not operating right at the start of an
6364 SMTP session. For example, it could not catch a HELO sent before the client
6365 waited for the greeting. There is now a check for outstanding input at the
6366 point when the greeting is written. Because of the duplex, asynchronous
6367 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6368 way, but not yet received, when the check is performed.
6370 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6371 on TCP/IP sockets, because this apparently causes some broken clients to
6374 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6375 unchanged) from the Cygwin maintainer.
6377 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6378 was supported instead of testing for mbx. Effectively a typo.
6380 15. The spa authenticator server code was not checking that the input it
6381 received was valid base64.
6383 16. The debug output line for the "set" modifier in ACLs was not showing the
6384 name of the variable that was being set.
6386 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6388 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6389 lookup on its own. Something else was needed to trigger the lookup. For
6390 example, a match in host_lookup or the need for a host name in a host list.
6391 Now, if $sender_host_name is referenced and the host name has not yet been
6392 looked up, a lookup is performed. If the lookup fails, the variable remains
6393 empty, and $host_lookup_failed is set to "1".
6395 19. Added "eqi" as a case-independent comparison operator.
6397 20. The saslauthd authentication condition could segfault if neither service
6398 nor realm was specified.
6400 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6401 error message that was logged was misleading, and incoming SMTP
6402 connections were dropped. The message is now more accurate, and temporary
6403 errors are given to SMTP connections.
6405 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6406 (except RSET) with a 421 error, until QUIT is received. However, it was
6407 failing to send a response to QUIT.
6409 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6410 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6411 ACL is now run after the helo_try_verify_hosts code.
6413 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6414 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6415 case-independent, so other case variants are also recognized). Apparently
6416 some systems use these upper case variants.
6418 25. If more than two messages were waiting for the same host, and a transport
6419 filter was specified for the transport, Exim sent two messages over the
6420 same TCP/IP connection, and then failed with "socket operation on non-
6421 socket" when it tried to send the third.
6423 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6425 27. The extern definition of crypt16() in expand.c was not being excluded when
6426 the OS had its own crypt16() function.
6428 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6429 as a preferred synonym for return_size_limit, both as an option and as an
6432 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6434 30. Changed the default configuration ACL to relax the local part checking rule
6435 for addresses that are not in any local domains. For these addresses,
6436 slashes and pipe symbols are allowed within local parts, but the sequence
6437 /../ is explicitly forbidden.
6439 31. SPA server authentication was not clearing the challenge buffer before
6442 32. log_message in a "warn" ACL statement was writing to the reject log as
6443 well as to the main log, which contradicts the documentation and doesn't
6444 seem right (because no rejection is happening). So I have stopped it.
6446 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6447 However, I am unable to do any testing of this.
6449 34. Fixed an infelicity in the appendfile transport. When checking directories
6450 for a mailbox, to see if any needed to be created, it was accidentally
6451 using path names with one or more superfluous leading slashes; tracing
6452 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6454 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6455 contents are added to the log line that is written for every discarded
6456 recipient. (Previously a log_message setting was ignored.)
6458 36. The ${quote: operator now quotes the string if it is empty.
6460 37. The install script runs exim in order to find its version number. If for
6461 some reason other than non-existence or emptiness, which it checks, it
6462 could not run './exim', it was installing it with an empty version number,
6463 i.e. as "exim-". This error state is now caught, and the installation is
6466 38. An argument was missing from the function that creates an error message
6467 when Exim fails to connect to the socket for saslauthd authentication.
6468 This could cause Exim to crash, or give a corrupted message.
6470 39. Added isip, isip4, and isip6 to ${if conditions.
6472 40. The ACL variables $acl_xx are now saved with the message, and can be
6473 accessed later in routers, transports, and filters.
6475 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6476 strings in the file are not string-expanded.
6478 42. If a MAIL command specified a SIZE value that was too large to fit into an
6479 int variable, the check against message_size_limit failed. Such values are
6480 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6481 day this will have to be increased, but I don't think I want to be around
6482 when emails are that large.
6489 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6490 iconv() is not standard in FreeBSD.
6492 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6493 IPv6 enabled. The observed symptom was a segmentation fault on return from
6494 the function os_common_find_running_interfaces() in src/os.c.
6496 3. In the check_special_case() function in daemon.c I had used "errno" as an
6497 argument name, which causes warnings on some systems. This was basically a
6498 typo, since it was named "eno" in the comments!
6500 4. The code that waits for the clock to tick (at a resolution of some fraction
6501 of a second) so as to ensure message-id uniqueness was always waiting for
6502 at least one whole tick, when it could have waited for less. [This is
6503 almost certainly not relevant at current processor speeds, where it is
6504 unlikely to ever wait at all. But we try to future-proof.]
6506 5. The function that sleeps for a time interval that includes fractions of a
6507 second contained a race. It did not block SIGALRM between setting the
6508 timer, and suspending (a couple of lines later). If the interval was short
6509 and the sigsuspend() was delayed until after it had expired, the suspension
6510 never ended. On busy systems this could lead to processes getting stuck for
6513 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6514 process, before it forks any delivery processes. The open lookup caching
6515 mechanism meant that the open file or database connection was passed into
6516 the delivery process. The problem was that delivery processes always tidy
6517 up cached lookup data. This could cause a problem for the next delivery
6518 process started by the queue runner, because the external queue runner
6519 process does not know about the closure. So the next delivery process
6520 still has data in the lookup cache. In the case of a file lookup, there was
6521 no problem because closing a file descriptor in a subprocess doesn't affect
6522 the parent. However, if the lookup was caching a connection to a database,
6523 the connection was closed, and the second delivery process was likely to
6524 see errors such as "PGSQL: query failed: server closed the connection
6525 unexpectedly". The problem has been fixed by closing all cached lookups
6526 in a queue runner before running a delivery process.
6528 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6529 seem to have the "const" qualifier which it has on other OS. I've
6532 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6533 *of the same type* (client or server) with the same public name that an
6534 error should be diagnosed.
6536 9. When Exim looked up a host name for an IP address, but failed to find the
6537 original IP address when looking up the host name (a safety check), it
6538 output the message "<ip address> does not match any IP for NULL", which was
6539 confusing, to say the least. The bug was that the host name should have
6540 appeared instead of "NULL".
6542 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6543 user that is built into the binary, and the -C or -D options is used, root
6544 privilege is dropped before the configuration file is read. In addition,
6545 logging is switched to stderr instead of the normal log files. If the
6546 configuration then re-defines the Exim user, the unprivileged environment
6547 is probably not what is expected, so Exim logs a panic warning message (but
6550 However, if deliver_drop_privilege is set, the unprivileged state may well
6551 be exactly what is intended, so the warning has been cut out in that case,
6552 and Exim is allowed to try to write to its normal log files.
6558 1. smtp_return_error_details was not giving details for temporary sender
6559 or receiver verification errors.
6561 2. Diagnose a configuration error if two authenticators have the same public
6564 3. Exim used not to create the message log file for a message until the first
6565 delivery attempt. This could be confusing when incoming messages were held
6566 for policy or load reasons. The message log file is now created at the time
6567 the message is received, and an initial "Received" line is written to it.
6569 4. The automatically generated man page for command line options had a minor
6570 bug that caused no ill effects; however, a more serious problem was that
6571 the procedure for building the man page automatically didn't always
6572 operate. Consequently, release 4.20 contains an out-of-date version. This
6573 shouldn't happen again.
6575 5. When building Exim with embedded Perl support, the script that builds the
6576 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6577 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6579 6. The freeze_tell option was not being used for messages that were frozen on
6580 arrival, either by an ACL or by local_scan().
6582 7. Added the smtp_incomplete_transaction log selector.
6584 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6585 was accepting AUTH without a new EHLO.
6587 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6588 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6591 10. Logging of TCP/IP connections (when configured) now happens in the main
6592 daemon process instead of the child process, so that the TCP/IP connection
6593 count is more accurate (but it can never be perfect).
6595 11. The use of "drop" in a nested ACL was not being handled correctly in the
6596 outer ACL. Now, if condition failure induced by the nested "drop" causes
6597 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6598 or "require"), the connection is dropped.
6600 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6601 that yield "discard" can now be used with an "accept" or a "discard" verb,
6602 but an error is generated for any others (because I can't see a useful way
6603 to define what should happen).
6605 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6606 are now processed in the same way as lines in the Exim configuration file.
6607 In particular, continuation lines are supported.
6609 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6611 15. Added -ti meaning -t -i.
6613 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6614 domains, and warn by logging if others are found.
6616 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6617 returned from GIFCONF when Exim is trying to find the list of interfaces on
6618 a host. The code in os.c has been modified to copy each ifreq to an aligned
6619 structure in all cases.
6621 Also, in some cases, the returned ifreq's were being copied to a 'struct
6622 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6623 means the last couple of bytes of an IPv6 address could be chopped if the
6624 ifreq contained only a normal sockaddr (14 bytes storage).
6626 18. Named domain lists were not supported in the hosts_treat_as_local option.
6627 An entry such as +xxxx was not recognized, and was treated as a literal
6630 19. Ensure that header lines added by a DATA ACL are included in the reject log
6631 if the ACL subsequently rejects the message.
6633 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6634 MD5 (which is deprecated).
6636 21. When testing a filter file using -bf, Exim was writing a message when it
6637 took the sender from a "From " line in the message, but it was not doing so
6638 when it took $return_path from a Return-Path: header line. It now does.
6640 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6641 with a valid header line field name (a series of printing characters
6642 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6644 23. Changed "disc" in the source to "disk" to conform to the documentation and
6645 the book and for uniformity.
6647 24. Ignore Sendmail's -Ooption=value command line item.
6649 25. When execve() failed while trying to run a command in a pipe transport,
6650 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6651 could be confused with a return value of 69 from the command itself. This
6652 has been changed to 127, the value the shell returns if it is asked to run
6653 a non-existent command. The wording for the related log line suggests a
6654 non-existent command as the problem.
6656 26. If received_header_text expands to an empty string, do not add a Received:
6657 header line to the message. (Well, it adds a token one on the spool, but
6658 marks it "old" so that it doesn't get used or transmitted.)
6660 27. Installed eximstats 1.28 (addition of -nt option).
6662 28. There was no check for failure on the call to getsockname() in the daemon
6663 code. This can fail if there is a shortage of resources on the system, with
6664 ENOMEM, for example. A temporary error is now given on failure.
6666 29. Contrary to the C standard, it seems that in some environments, the
6667 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6668 program. Exim now does this explicitly; it affects the formatting of
6669 timestamps using strftime().
6671 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6672 complaints. I've now initialized all the variables, to avoid this.
6674 32. Header lines added by a system filter were not being "seen" during
6675 transport-time rewrites.
6677 33. The info_callback() function passed to OpenSSL is set up with type void
6678 (*)(SSL *, int, int), as described somewhere. However, when calling the
6679 function (actually a macro) that sets it up, the type void(*)() is
6680 expected. I've put in a cast to prevent warnings from picky compilers.
6682 34. If a DNS black list lookup found a CNAME record, but there were no A
6683 records associated with the domain it pointed at, Exim crashed.
6685 35. If a DNS black list lookup returned more than one A record, Exim ignored
6686 all but the first. It now scans all returned addresses if a particular IP
6687 value is being sought. In this situation, the contents of the
6688 $dnslist_value variable are a list of all the addresses, separated by a
6691 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6692 to accept a host name and all its aliases if the forward lookup for any of
6693 them yielded the IP address of the incoming connection. Now it accepts only
6694 those names whose forward lookup yields the correct IP address. Any other
6695 names are discarded. This closes a loophole whereby a rogue DNS
6696 administrator could create reverse DNS records to break through a
6697 wildcarded host restriction in an ACL.
6699 37. If a user filter or a system filter that ran in a subprocess used any of
6700 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6701 the wrong values were passed to the pipe command ($thisaddress had the
6702 value of $0, $0 had the value of $1, etc). This bug was introduced by
6703 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6705 38. Improved the line breaking for long SMTP error messages from ACLs.
6706 Previously, if there was no break point between 40 and 75 characters, Exim
6707 left the rest of the message alone. Two changes have been made: (a) I've
6708 reduced the minimum length to 35 characters; (b) if it can't find a break
6709 point between 35 and 75 characters, it looks ahead and uses the first one
6710 that it finds. This may give the occasional overlong line, but at least the
6711 remaining text gets split now.
6713 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6714 file descriptors might be low, and that setting 1000 would always raise it.
6715 It turns out that in some environments, the limit is already over 1000 and
6716 that lowering it causes trouble. So now Exim takes care not to decrease it.
6718 40. When delivering a message, the value of $return_path is set to $sender_
6719 address at the start of routing (routers may change the value). By an
6720 oversight, this default was not being set up when an address was tested by
6721 -bt or -bv, which affected the outcome if any router or filter referred to
6724 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6725 log only when "message" or "log_message" are set. However, if one of the
6726 conditions was an address verification, or a call to a nested ACL, the
6727 messages generated by the underlying test were being passed through. This
6728 no longer happens. The underlying message is available in $acl_verify_
6729 message for both "message" and "log_message" expansions, so it can be
6730 passed through if needed.
6732 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6733 new expansion $bh_ to give the encoded byte string without charset
6734 translation. Translation happens only if iconv() is available; HAVE_ICONV
6735 indicates this at build time. HEADERS_CHARSET gives the charset to
6736 translate to; headers_charset can change it in the configuration, and
6737 "headers charset" can change it in an individual filter file.
6739 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6740 that creates RFC 2047 encoded "words" labels them as that charset instead
6741 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6742 expansion operator; (ii) when Exim creates a From: line for a local
6743 message; (iii) when a header line is rewritten to include a "phrase" part.
6745 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6746 buggy, causing it to skip the first lines of messages whose message ID
6747 ended in 'D'. This would not have bitten before Exim release 4.14, because
6748 message IDs were unlikely to end in 'D' before then. The effect was to have
6749 incorrect size information for certain domains.
6751 45. #include "config.h" was missing at the start of the crypt16.c module. This
6752 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6755 46. If there was a timeout during a "random" callout check, Exim treated it as
6756 a failure of the random address, and carried on sending RSET and the real
6757 address. If the delay was just some slowness somewhere, the response to the
6758 original RCPT would be taken as a response to RSET and so on, causing
6759 mayhem of various kinds.
6761 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6762 when I implemented it. It didn't allow for the fact that some option values
6763 may legitimately be negative (e.g. size_addition), and it didn't even do
6764 the right test for positive values.
6766 48. Domain names in DNS records are case-independent. Exim always looks them up
6767 in lower case. Some resolvers return domain names in exactly the case they
6768 appear in the zone file, that is, they may contain uppercase letters. Not
6769 all resolvers do this - some return always lower case. Exim was treating a
6770 change of case by a resolver as a change of domain, similar to a widening
6771 of a domain abbreviation. This triggered its re-routing code and so it was
6772 trying to route what was effectively the same domain again. This normally
6773 caused routing to fail (because the router wouldn't handle the domain
6774 twice). Now Exim checks for this case specially, and just changes the
6775 casing of the domain that it ultimately uses when it transmits the message
6778 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6781 50. If a filter generated a file delivery with a non-absolute name (possible if
6782 no home directory exists for the router), the forbid_file option was not
6785 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6786 the existing equality matching.
6788 52. Exim was using ints instead of ino_t variables in some places where it was
6789 dealing with inode numbers.
6791 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6792 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6793 called TMPDIR, and if it finds it is different, it changes its value.
6795 54. The smtp_printf() function is now made available to local_scan() so
6796 additional output lines can be written before returning. There is also an
6797 smtp_fflush() function to enable the detection of a dropped connection.
6798 The variables smtp_input and smtp_batched_input are exported to
6801 55. Changed the default runtime configuration: the message "Unknown user"
6802 has been removed from the ACL, and instead placed on the localuser router,
6803 using the cannot_route_message feature. This means that any verification
6804 failures that generate their own messages won't get overridden. Similarly,
6805 the "Unrouteable address" message that was in the ACL for unverifiable
6806 relay addresses has also been removed.
6808 56. Added hosts_avoid_esmtp to the smtp transport.
6810 57. The exicyclog script was not checking for the esoteric option
6811 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6812 will work only if exicyclog is run under the appropriate euid.
6814 58. Following a discussion on the list, the rules by which Exim recognises line
6815 endings on incoming messages have been changed. The -dropcr and drop_cr
6816 options are now no-ops, retained only for backwards compatibility. The
6817 following line terminators are recognized: LF CRLF CR. However, special
6818 processing applies to CR:
6820 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6821 nor a local message in the state where . is a terminator.
6823 (ii) If a bare CR is encountered in a header line, an extra space is added
6824 after the line terminator so as not to end the header. The reasoning
6825 behind this is that bare CRs in header lines are most likely either
6826 to be mistakes, or people trying to play silly games.
6828 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6829 was being incorrectly given as 18 bytes larger than it should have been.
6830 This is a VOB (very old bug).
6832 60. This may never have affected anything current, but just in case it has:
6833 When the local host is found other than at the start of a list of hosts,
6834 the local host, those with the same MX, and any that follow, are discarded.
6835 When the list in question was part of a longer list of hosts, the following
6836 hosts (not currently being processed) were also being discarded. This no
6837 longer happens. I'm not sure if this situation could ever has previously
6840 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6843 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6844 header. This is something that is recommended in a new Internet Draft, and
6845 is something that is documented as being done by Sendmail. There are two
6846 possible values. For messages generated by the autoreply transport, Exim
6849 Auto-Submitted: auto-replied
6851 whereas for all other generated messages (e.g. bounces) it adds
6853 Auto-Submitted: auto-generated
6855 63. The "personal" condition in filters now includes a test for the
6856 Auto-Submitted: header. If it contains the string "auto-" the message it
6857 not considered personal.
6859 64. Added rcpt_include_affixes as a generic transport option.
6861 65. Added queue_only_override (default true).
6863 66. Added the syslog_duplication option.
6865 67. If what should have been the first header line of a message consisted of
6866 a space followed by a colon, Exim was mis-interpreting it as a header line.
6867 It isn't of course - it is syntactically invalid and should therefore be
6868 treated as the start of the message body. The misbehaviour could have
6869 caused a number of strange effects, including loss of data in subsequent
6870 header lines, and spool format errors.
6872 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6873 client host had authenticated. This control can now be exercised by an ACL
6874 for more flexibility.
6876 69. By default, callouts do not happen when testing with -bh. There is now a
6877 variant, -bhc, which does actually run the callout code, including
6878 consulting and updating the callout cache.
6880 70. Added support for saslauthd authentication, courtesy of Alexander
6883 71. If statvfs() failed on the spool or log directories while checking their
6884 size for availability, Exim confusingly gave the error "space shortage".
6885 Furthermore, in debugging mode it crashed with a floating point exception.
6886 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6887 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6888 serious problem, Exim now writes to the main and panic logs when this
6889 happens, with details of the failure. It then refuses to accept the
6890 incoming message, giving the message "spool directory problem" or "log
6891 directory problem" with a 421 code for SMTP messages.
6893 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6894 0, 1, and 2 exist, because some OS complain for execs without them (see
6895 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6896 descriptors. However, the code omitted to check that the open succeeded,
6897 causing mysterious errors if for some reason the permissions on /dev/null
6898 got screwed. Now Exim writes a message to the main and panic logs, and
6899 bombs out if it can't open /dev/null.
6901 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6902 interact so that it is all more flexible. It is supposed to remain
6903 backwards compatible. Also added extra_local_interfaces.
6905 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6906 to bomb out with an assertion failure - to the client this appears as a
6907 connection drop. This problem occurs in the part of the code that was taken
6908 from the Samba project. Fortunately, the assertion is in a very simple
6909 function, so I have fixed this by reproducing the function inline in the
6910 one place where it is called, and arranging for authentication to fail
6911 instead of killing the process with assert().
6913 75. The SPA client code was not working when the server requested OEM rather
6914 than Unicode encoding.
6916 76. Added code to make require_files with a specific uid setting more usable in
6917 the case where statting the file as root fails - usually a non-root-mounted
6918 NFS file system. When this happens and the failure is EACCES, Exim now
6919 forks a subprocess and does the per-uid checking as the relevant uid.
6921 77. Added process_log_path.
6923 78. If log_file_path was not explicitly set, a setting of check_log_space or
6924 check_log_inodes was ignored.
6926 79. If a space check for the spool or log partitions fails, the incident is now
6927 logged. Of course, in the latter case the data may get lost...
6929 80. Added the %p formatting code to string_format() so that it can be used to
6930 print addresses in debug_print(). Adjusted all the address printing in the
6931 debugging in store.c to use %p rather than %d.
6933 81. There was a concern that a line of code in smtp_in.c could overflow a
6934 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6935 initially expressed, the concern was not well-founded, because trailing
6936 spaces are removed early. However, if the trailing spaces were followed by
6937 a NULL, they did not get removed, so the overflow was possible. Two fixes
6940 (a) I re-wrote the offending code in a cleaner fashion.
6941 (b) If an incoming SMTP command contains a NULL character, it is rejected
6944 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6945 runs initgroups(), so that if the Exim user is in any additional groups,
6946 they will be used during message reception.
6952 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob