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 Replace the store_release() internal interface with store_newblock(),
12 which internalises the check required to safely use the old one, plus
13 the allocate and data copy operations duplicated in both (!) of the
16 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
17 modifier. This matches the restriction on the commandline.
19 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
20 Previously only the last row was returned.
22 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
23 we assumed that tags in the header were well-formed, and parsed the
24 element content after inspecting only the first char of the tag.
25 Assumptions at that stage could crash the receive process on malformed
28 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
29 While running the DKIM ACL we operate on the Permanent memory pool so that
30 variables created with "set" persist to the DATA ACL. Also (at any time)
31 DNS lookups that fail create cache records using the Permanent pool. But
32 expansions release any allocations made on the current pool - so a dnsdb
33 lookup expansion done in the DKIM ACL releases the memory used for the
34 DNS negative-cache, and bad things result. Solution is to switch to the
35 Main pool for expansions.
36 While we're in that code, add checks on the DNS cache during store_reset,
37 active in the testsuite.
38 Problem spotted, and debugging aided, by Wolfgang Breyha.
40 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
41 When none of the hosts presented to a transport match an already-open
42 connection, close it and proceed with the list. Previously we would
43 queue the message. Spotted by Lena with Yahoo, probably involving
46 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
47 Previously a spurious "250 OK id=" response was appended to the proper
50 JH/08 The "support for" informational output now, which built with Content
51 Scanning support, has a line for the malware scanner interfaces compiled
52 in. Interface can be individually included or not at build time.
58 JH/01 Rework error string handling in TLS interface so that the caller in
59 more cases is responsible for logging. This permits library-sourced
60 string to be attached to addresses during delivery, and collapses
61 pairs of long lines into single ones.
63 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
64 during configuration. Wildcards are allowed and expanded.
66 JH/02 Rework error string handling in DKIM to pass more info back to callers.
67 This permits better logging.
69 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
70 do not close it down and have the child transport start it up again on
71 the passed-on TCP connection. Instead, proxy the child (and any
72 subsequent ones) for TLS via a unix-domain socket channel. Logging is
73 affected: the continued delivery log lines do not have any DNSSEC, TLS
74 Certificate or OCSP information. TLS cipher information is still logged.
76 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
77 identical IP addresses on different listening ports. Will also affect
80 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
81 add noisy ifdef guards to special-case this sillyness.
82 Patch from Bernd Kuhls.
84 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
85 than 255 are no longer allowed.
87 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
88 Disable the session-cache too, which might reduce our load. Since we
89 currrectly use a new context for every connection, both as server and
90 client, there is no benefit for these.
91 GnuTLS appears to not support tickets server-side by default (we don't
92 call gnutls_session_ticket_enable_server()) but client side is enabled
93 by default on recent versions (3.1.3 +) unless the PFS priority string
96 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
97 <https://reproducible-builds.org/specs/source-date-epoch/>.
99 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
100 the check for any unsuccessful recipients did not notice the limit, and
101 erroneously found still-pending ones.
103 JH/08 Pipeline CHUNKING command and data together, on kernels that support
104 MSG_MORE. Only in-clear (not on TLS connections).
106 JH/09 Avoid using a temporary file during transport using dkim. Unless a
107 transport-filter is involved we can buffer the headers in memory for
108 creating the signature, and read the spool data file once for the
109 signature and again for transmission.
111 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
112 4.77 as the kernel support then wasn't solid, having issues in 64bit
113 mode. Now, it's been long enough. Add support for FreeBSD also.
115 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
116 case where the routing stage had gathered several addresses to send to
117 a host before calling the transport for the first, we previously failed
118 to close down TLS in the old transport process before passing the TCP
119 connection to the new process. The new one sent a STARTTLS command
120 which naturally failed, giving a failed delivery and bloating the retry
121 database. Investigation and fix prototype from Wolfgang Breyha.
123 JH/12 Fix check on SMTP command input synchronisation. Previously there were
124 false-negatives in the check that the sender had not preempted a response
125 or prompt from Exim (running as a server), due to that code's lack of
126 awareness of the SMTP input buffering.
128 PP/04 Add commandline_checks_require_admin option.
129 Exim drops privileges sanely, various checks such as -be aren't a
130 security problem, as long as you trust local users with access to their
131 own account. When invoked by services which pass untrusted data to
132 Exim, this might be an issue. Set this option in main configuration
133 AND make fixes to the calling application, such as using `--` to stop
136 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
137 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
138 responses to those, into a single TLS record each way (this usually means
139 a single packet). As a side issue, smtp_enforce_sync now works on TLS
142 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
143 affects you only if you're dancing at the edge of the param size limits.
144 If you are, and this message makes sense to you, then: raise the
145 configured limit or use OpenSSL 1.1. Nothing we can do for older
148 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
149 cmdline element to get the documented default one. Previously it was
152 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
153 are unsuable for non-message-specific reasons - eg. connection timeout,
154 banner-time rejection.
156 JH/16 Fix logging of delivery remote port, when specified by router, under
159 PP/06 Repair manualroute's ability to take options in any order, even if one
160 is the name of a transport.
163 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
165 JH/17 Change the list-building routines interface to use the expanding-string
166 triplet model, for better allocation and copying behaviour.
168 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
169 Previously it was constructed the first time a possibly-matching string
170 was met in the configuration file input during startup; now it is done
173 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
174 compatible one, to avoid the (poorly documented) possibility of a config
175 file in the working directory redirecting the DB files, possibly correpting
176 some existing file. CVE-2017-10140 assigned for BDB.
178 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
179 cache-hot. Previously, although the result was properly cached, the
180 initial verify call returned a defer.
182 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
183 the main verify for receipient in uncached-mode.
185 JH/22 Retire historical build files to an "unsupported" subdir. These are
186 defined as "ones for which we have no current evidence of testing".
188 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
189 if present. Previously it was ignored.
191 JH/24 Start using specified-initialisers in C structure init coding. This is
192 a C99 feature (it's 2017, so now considered safe).
194 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
195 if was a fixed-sized field and bitmask ops via macros; it is now more
198 PP/07 GitHub PR 56: Apply MariaDB build fix.
199 Patch provided by Jaroslav Škarvada.
201 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
202 during Coverity cleanups [4.87 JH/47]
203 Diagnosis and fix provided by Michael Fischer v. Mollard.
205 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
206 the right size to place the terminating semicolon on its own folded
207 line, the header hash was calculated to an incorrect value thanks to
208 the (relaxed) space the fold became.
210 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
211 and confused the parent.
213 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
214 which could crash as a result. This could lead to undeliverable messages.
216 JH/28 Logging: "next input sent too soon" now shows where input was truncated
219 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
220 matters on fast-turnover and PID-randomising systems, which were getting
221 out-of-order delivery.
223 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
224 a possibly-overlapping copy. The symptom was that "Remote host closed
225 connection in response to HELO" was logged instead of the actual 4xx
228 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
229 Previously only that bufferd was discarded, resulting in SYMTP command
232 JH/32 DKIM: when a message has multiple signatures matching an identity given
233 in dkim_verify_signers, run the dkim acl once for each. Previously only
234 one run was done. Bug 2189.
236 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
237 "panic the current process" to "deliberately defer". The panic log is
238 still written with the problem list name; the mail and reject logs now
239 get a temp-reject line for the message that was being handled, saying
240 something like "domains check lookup or other defer". The SMTP 451
241 message is still "Temporary local problem".
243 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
244 A crafted sequence of BDAT commands could result in in-use memory beeing
245 freed. CVE-2017-16943.
247 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
248 from SMTP input. Previously it was always done; now only done for DATA
249 and not BDAT commands. CVE-2017-16944.
251 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
252 to the message (such as an overlong header line). Previously this was
253 not done and we did not exit BDAT mode. Followon from the previous item
254 though a different problem.
260 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
261 than -2003 did; needs libidn2 in addition to libidn.
263 JH/02 The path option on a pipe transport is now expanded before use.
265 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
266 Patch provided by "Björn", documentation fix added too.
268 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
269 missing a wire-to-host endian conversion.
271 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
272 close after a BDAT command line could be taken as a following command,
273 giving a synch failure. Fix by only checking for synch immediately
274 before acknowledging the chunk.
276 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
277 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
278 macro. Patches provided by Josh Soref.
280 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
281 Previously we did not; the RFC seems ambiguous and VRFY is not listed
282 by IANA as a service extension. However, John Klensin suggests that we
285 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
286 the dkim code may be unix-mode line endings rather than smtp wire-format
287 CRLF, so prepend a CR to any bare LF.
289 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
290 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
292 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
293 the first were themselves being wrongly included in the feed into dkim
294 processing; with most chunk sizes in use this resulted in an incorrect
295 body hash calculated value.
297 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
298 DKIM signature block, for verification. Although advised against by
299 standards it is specifically not ruled illegal.
301 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
303 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
304 missing a body hash (the bh= tag).
306 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
307 It seems that HAProxy sends the Proxy Protocol information in clear and
308 only then does a TLS startup, so do the same.
310 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
311 TCP connections (such as for Spamd) unless the daemon successfully set
312 Fast Open mode on its listening sockets. This fixes breakage seen on
313 too-old kernels or those not configured for Fast Open, at the cost of
314 requiring both directions being enabled for TFO, and TFO never being used
315 by non-daemon-related Exim processes.
317 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
318 endings, at least on the first header line. Try to canonify any that get
319 past that check, despite the cost.
321 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
322 now limited to an arbitrary five deep, while parsing addresses with the
323 strip_excess_angle_brackets option enabled.
325 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
326 instead leave the unprompted TLS handshake in socket buffer for the
327 TLS library to consume.
329 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
331 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
333 JH/16 Drop variables when they go out of scope. Memory management drops a whole
334 region in one operation, for speed, and this leaves assigned pointers
335 dangling. Add checks run only under the testsuite which checks all
336 variables at a store-reset and panics on a dangling pointer; add code
337 explicitly nulling out all the variables discovered. Fixes one known
338 bug: a transport crash, where a dangling pointer for $sending_ip_address
339 originally assigned in a verify callout, is re-used.
341 PP/06 Drop '.' from @INC in various Perl scripts.
343 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
345 PP/08 Reduce a number of compilation warnings under clang; building with
346 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
347 should be warning-free.
349 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
351 HS/01 Fix portability problems introduced by PP/08 for platforms where
352 realloc(NULL) is not equivalent to malloc() [SunOS et al].
354 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
355 chunk. This allows us to accept broken chunked messages. We need a more
356 general solution here.
358 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
359 already-broken messages in the queue.
361 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
363 JH/19 Fix reference counting bug in routing-generated-address tracking.
369 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
370 supports it and a size is available (ie. the sending peer gave us one).
372 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
373 years of being deprecated). The replacements are the ACLs
374 acl_smtp_mime and acl_not_smtp_mime.
376 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
377 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
378 or even an in-clear connection were permitted. Now, if the host lookup
379 was dnssec and dane was requested then the host is only used if the
380 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
381 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
382 if one fails this test.
383 This means that a poorly-configured remote DNS will make it incommunicado;
384 but it protects against a DNS-interception attack on it.
386 JH/04 Bug 1810: make continued-use of an open smtp transport connection
387 non-noisy when a race steals the message being considered.
389 JH/05 If main configuration option tls_certificate is unset, generate a
390 self-signed certificate for inbound TLS connections.
392 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
393 in rewrites and routers.
395 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
396 and logged a warning sing 4.83; now they are a configuration file error.
398 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
399 (lacking @domain). Apply the same qualification processing as RCPT.
401 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
403 JH/10 Support ${sha256:} applied to a string (as well as the previous
406 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
407 a cutthrough deliver is pending, as we always want to make a connection.
408 This also avoids re-routing the message when later placing the cutthrough
409 connection after a verify cache hit.
410 Do not update it with the verify result either.
412 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
413 when routing results in more than one destination address.
415 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
416 signing (which inhibits the cutthrough capability). Previously only
417 the presence of an option was tested; now an expansion evaluating as
418 empty is permissible (obviously it should depend only on data available
419 when the cutthrough connection is made).
421 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
422 the relevant preceding SMTP command did not note the pipelining mode.
424 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
425 Previously they were not counted.
427 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
428 as one having no matching records. Previously we deferred the message
429 that needed the lookup.
431 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
432 distinguished as "(=".
434 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
435 for missing MX records. Previously it only worked for missing A records.
437 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
439 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
440 after the data-go-ahead and data-ack. Patch from Jason Betts.
442 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
443 even for a "none" policy. Patch from Tony Meyer.
445 JH/22 Fix continued use of a connection for further deliveries. If a port was
446 specified by a router, it must also match for the delivery to be
449 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
450 When one of the recipients of a message was unsuitable for the connection
451 (has no matching addresses), we lost track of needing to mark it
452 deferred. As a result mail would be lost.
454 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
456 JH/25 Decoding ACL controls is now done using a binary search; the source code
457 takes up less space and should be simpler to maintain. Merge the ACL
458 condition decode tables also, with similar effect.
460 JH/26 Fix problem with one_time used on a redirect router which returned the
461 parent address unchanged. A retry would see the parent address marked as
462 delivered, so not attempt the (identical) child. As a result mail would
465 JH/27 Fix a possible security hole, wherein a process operating with the Exim
466 UID can gain a root shell. Credit to http://www.halfdog.net/ for
467 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
470 JH/28 Enable {spool,log} filesystem space and inode checks as default.
471 Main config options check_{log,spool}_{inodes,space} are now
472 100 inodes, 10MB unless set otherwise in the configuration.
474 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
475 Previously it only applied to the main-section connection policy
478 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
480 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
481 by me. Added RFC7919 DH primes as an alternative.
483 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
484 are not in the system include path.
486 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
487 GnuTLS, when a session startup failed (eg because the client disconnected)
488 Exim did stdio operations after fclose. This was exposed by a recent
489 change which nulled out the file handle after the fclose.
491 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
492 signed directly by the cert-signing cert, rather than an intermediate
493 OCSP-signing cert. This is the model used by LetsEncrypt.
495 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
497 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
498 an incoming connection.
500 HS/02 Bug 1802: Do not half-close the connection after sending a request
503 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
504 fallback to "prime256v1".
506 JH/34 SECURITY: Use proper copy of DATA command in error message.
507 Could leak key material. Remotely exploitable. CVE-2016-9963.
513 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
514 and 3.4.4 - once the server is enabled to respond to an OCSP request
515 it does even when not requested, resulting in a stapling non-aware
516 client dropping the TLS connection.
518 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
519 support variable-length bit vectors. No functional change.
521 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
522 The I= interface field on outgoing lines is now after the H= remote
523 host field, same as incoming lines. There is a separate
524 outgoing_interface log selector which allows you to disable the
527 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
528 If not running log_selector +smtp_connection the mainlog would be held
529 open indefinitely after a "too many connections" event, including to a
530 deleted file after a log rotate. Leave the per net connection logging
531 leaving it open for efficiency as that will be quickly detected by the
532 check on the next write.
534 HS/01 Bug 1671: Fix post transport crash.
535 Processing the wait-<transport> messages could crash the delivery
536 process if the message IDs didn't exist for some reason. When
537 using 'split_spool_directory=yes' the construction of the spool
538 file name failed already, exposing the same netto behaviour.
540 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
541 mime_regex ACL conditions.
543 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
544 to DSN fail messages (bounces): remote IP, remote greeting, remote response
545 to HELO, local diagnostic string.
547 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
548 log line to debug. Even when configured with a tls authenticator many
549 client connections are expected to not authenticate in this way, so
550 an authenticate fail is not an error.
552 HS/02 Add the Exim version string to the process info. This way exiwhat
553 gives some more detail about the running daemon.
555 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
556 matter for fast-change records such as DNSBLs.
558 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
559 retry record, even if constant. There may be multiple transports with
560 different interface settings and the retry behaviour needs to be kept
563 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
565 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
567 JH/10 Bug 840: fix log_defer_output option of pipe transport
569 JH/11 Bug 830: use same host for all RCPTS of a message, even under
570 hosts_randomize. This matters a lot when combined with mua_wrapper.
572 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
573 ${quote_pgsql:<string>} operator.
575 JH/13 Bug 1708: avoid misaligned access in cached lookup.
577 JH/14 Change header file name for freeradius-client. Relevant if compiling
578 with Radius support; from the Gentoo tree and checked under Fedora.
580 JH/15 Bug 1712: Introduce $prdr_requested flag variable
582 JH/16 Bug 1714: Permit an empty string as expansion result for transport
583 option transport_filter, meaning no filtering.
585 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
587 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
588 defaults to "*" (all hosts). The variable is now available when not built
589 with TLS, default unset, mainly to enable keeping the testsuite sane.
590 If a server certificate is not supplied (via tls_certificate) an error is
591 logged, and clients will find TLS connections fail on startup. Presumably
592 they will retry in-clear.
593 Packagers of Exim are strongly encouraged to create a server certificate
594 at installation time.
596 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
597 with the $config_file variable.
599 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
600 in transport context, after the attempt, and per-recipient. The latter type
601 is per host attempted. The event data is the error message, and the errno
602 information encodes the lookup type (A vs. MX) used for the (first) host,
603 and the trailing two digits of the smtp 4xx response.
605 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
606 to write to mainlog (or rejectlog, paniclog) in the window between file
607 creation and permissions/ownership being changed. Particularly affects
608 installations where exicyclog is run as root, rather than exim user;
609 result is that the running daemon panics and dies.
611 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
613 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
614 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
615 "pri" and "weight". Note that the previous implicit priority given by the
616 list order is no longer honoured.
618 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
621 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
622 by defining SUPPORT_SOCKS.
624 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
625 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
626 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
627 variables are renamed to proxy_{local,external}_{address,port}.
629 JH/27 Move Internationalisation support from Experimental to mainline, enabled
630 for a build by defining SUPPORT_I18N
632 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
633 of the query string, and make ${quote_redis:} do that quoting.
635 JH/29 Move Events support from Experimental to mainline, enabled by default
636 and removable for a build by defining DISABLE_EVENT.
638 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
640 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
641 cached by the daemon.
643 JH/32 Move Redis support from Experimental to mainline, enabled for a build
644 by defining LOOKUP_REDIS. The libhiredis library is required.
646 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
647 keys are given for lookup.
649 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
650 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
651 only supported when built with TLS support. The PolarSSL SHA routines
652 are still used when the TLS library is too old for convenient support.
654 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
655 openssl_options), for security. OpenSSL forces this from version 1.1.0
656 server-side so match that on older versions.
658 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
659 allocation for $value could be released as the expansion processing
660 concluded, but leaving the global pointer active for it.
662 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
663 and to use the domains and local_parts ACL conditions.
665 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
666 incorrectly not doubled on cutthrough transmission, hence seen as a
667 body-termination at the receiving system - resulting in truncated mails.
668 Commonly the sender saw a TCP-level error, and retransmitted the message
669 via the normal store-and-forward channel. This could result in duplicates
670 received - but deduplicating mailstores were liable to retain only the
671 initial truncated version.
673 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
675 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
677 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
678 we're in there, support oversigning also; bug 1309.
680 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
682 HS/04 Add support for keep_environment and add_environment options.
684 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
685 either intentional arithmetic overflow during PRNG, or testing config-
688 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
689 delivery resulted in actual delivery. Cancel cutthrough before DATA
692 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
693 rejects a recipient: pass the reject to the originator.
695 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
696 Many were false-positives and ignorable, but it's worth fixing the
699 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
700 for the new environment-manipulation done at startup. Move the routines
701 from being local to tls.c to being global via the os.c file.
703 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
704 an extract embedded as result-arg for a map, the first arg for extract
705 is unavailable so we cannot tell if this is a numbered or keyed
706 extraction. Accept either.
712 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
715 JH/02 The smtp transport option "multi_domain" is now expanded.
717 JH/03 The smtp transport now requests PRDR by default, if the server offers
720 JH/04 Certificate name checking on server certificates, when exim is a client,
721 is now done by default. The transport option tls_verify_cert_hostnames
722 can be used to disable this per-host. The build option
723 EXPERIMENTAL_CERTNAMES is withdrawn.
725 JH/05 The value of the tls_verify_certificates smtp transport and main options
726 default to the word "system" to access the system default CA bundle.
727 For GnuTLS, only version 3.0.20 or later.
729 JH/06 Verification of the server certificate for a TLS connection is now tried
730 (but not required) by default. The verification status is now logged by
731 default, for both outbound TLS and client-certificate supplying inbound
734 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
737 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
738 Status Notification (bounce) messages are now MIME format per RFC 3464.
739 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
740 under the control of the dsn_advertise_hosts option, and routers may
741 have a dsn_lasthop option.
743 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
744 default, modifiable by a malware= option. The list separator for
745 the options can now be changed in the usual way. Bug 68.
747 JH/10 The smtp_receive_timeout main option is now expanded before use.
749 JH/11 The incoming_interface log option now also enables logging of the
750 local interface on delivery outgoing connections.
752 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
753 if the interface and destination host and port all match.
755 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
758 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
759 Patch from Andrew Lewis.
761 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
762 now supports optional time-restrictions, weighting, and priority
763 modifiers per server. Patch originally by <rommer@active.by>.
765 JH/16 The spamd_address main option now supports a mixed list of local
766 and remote servers. Remote servers can be IPv6 addresses, and
767 specify a port-range.
769 JH/17 Bug 68: The spamd_address main option now supports an optional
770 timeout value per server.
772 JH/18 Bug 1581: Router and transport options headers_add/remove can
773 now have the list separator specified.
775 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
778 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
781 JH/21 Support for the A6 type of dns record is withdrawn.
783 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
784 rather than the verbs used.
786 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
787 from 255 to 1024 chars.
789 JH/24 Verification callouts now attempt to use TLS by default.
791 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
792 are generic router options now. The defaults didn't change.
794 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
795 Original patch from Alexander Shikoff, worked over by JH.
797 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
798 files if ZCAT_COMMAND is not executable.
800 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
802 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
804 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
805 Normally benign, it bites when the pair was led to by a CNAME;
806 modern usage is to not canonicalize the domain to a CNAME target
807 (and we were inconsistent anyway for A-only vs AAAA+A).
809 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
811 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
812 when evaluating $sender_host_dnssec.
814 JH/31 Check the HELO verification lookup for DNSSEC, adding new
815 $sender_helo_dnssec variable.
817 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
819 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
821 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
823 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
824 documented as working, but never had. Support all but $spam_report.
826 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
827 added for tls authenticator.
829 HS/03 Add perl_taintmode main config option
835 TL/01 When running the test suite, the README says that variables such as
836 no_msglog_check are global and can be placed anywhere in a specific
837 test's script, however it was observed that placement needed to be near
838 the beginning for it to behave that way. Changed the runtest perl
839 script to read through the entire script once to detect and set these
840 variables, reset to the beginning of the script, and then run through
841 the script parsing/test process like normal.
843 TL/02 The BSD's have an arc4random API. One of the functions to induce
844 adding randomness was arc4random_stir(), but it has been removed in
845 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
846 function when detected.
848 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
849 cause callback expansion.
851 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
852 syntax errors in an expansion can be treated as a string instead of
853 logging or causing an error, due to the internal use of bool_lax
854 instead of bool when processing it.
856 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
857 server certificates when making smtp deliveries.
859 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
861 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
863 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
865 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
866 Merged patch from Sebastian Wiedenroth.
868 JH/05 Fix results-pipe from transport process. Several recipients, combined
869 with certificate use, exposed issues where response data items split
870 over buffer boundaries were not parsed properly. This eventually
871 resulted in duplicates being sent. This issue only became common enough
872 to notice due to the introduction of connection certificate information,
873 the item size being so much larger. Found and fixed by Wolfgang Breyha.
875 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
876 size buffer was used, resulting in syntax errors when an expansion
879 JH/07 Add support for directories of certificates when compiled with a GnuTLS
880 version 3.3.6 or later.
882 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
883 is EXPERIMENTAL_EVENT, the main-configuration and transport options
884 both become "event_action", the variables become $event_name, $event_data
885 and $event_defer_errno. There is a new variable $verify_mode, usable in
886 routers, transports and related events. The tls:cert event is now also
887 raised for inbound connections, if the main configuration event_action
890 TL/06 In test suite, disable OCSP for old versions of openssl which contained
891 early OCSP support, but no stapling (appears to be less than 1.0.0).
893 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
894 server certificate names available under the smtp transport option
895 "tls_verify_cert_hostname" now do not permit multi-component wildcard
898 JH/10 Time-related extraction expansions from certificates now use the main
899 option "timezone" setting for output formatting, and are consistent
900 between OpenSSL and GnuTLS compilations. Bug 1541.
902 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
903 encoded parameter in the incoming message. Bug 1558.
905 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
906 include certificate info, eximon was claiming there were spoolfile
909 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
911 JH/14 Log delivery-related information more consistently, using the sequence
912 "H=<name> [<ip>]" wherever possible.
914 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
915 are problematic for Debian distribution, omit them from the release
918 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
920 JH/16 Fix string representation of time values on 64bit time_t architectures.
923 JH/17 Fix a null-indirection in certextract expansions when a nondefault
924 output list separator was used.
929 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
930 checkers that were complaining about end of non-void function with no
933 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
934 This was a regression introduced in 4.83 by another bugfix.
936 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
938 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
939 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
945 TF/01 Correctly close the server side of TLS when forking for delivery.
947 When a message was received over SMTP with TLS, Exim failed to clear up
948 the incoming connection properly after forking off the child process to
949 deliver the message. In some situations the subsequent outgoing
950 delivery connection happened to have the same fd number as the incoming
951 connection previously had. Exim would try to use TLS and fail, logging
952 a "Bad file descriptor" error.
954 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
955 utilities have not been installed.
957 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
958 temporary space as the ACL may create new global variables.
960 TL/01 LDAP support uses per connection or global context settings, depending
961 upon the detected version of the libraries at build time.
963 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
964 to extract and use the src ip:port in logging and expansions as if it
965 were a direct connection from the outside internet. PPv2 support was
966 updated based on HAProxy spec change in May 2014.
968 JH/02 Add ${listextract {number}{list}{success}{fail}}.
970 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
971 Properly escape header and check for NULL return.
973 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
976 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
978 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
979 characters in header names, implemented as a verify condition.
980 Contributed by Michael Fischer v. Mollard.
982 TL/05 Rename SPF condition results err_perm and err_temp to standardized
983 results permerror and temperror. Previous values are deprecated but
984 still accepted. In a future release, err_perm and err_temp will be
985 completely removed, which will be a backward incompatibility if the
986 ACL tests for either of these two old results. Patch contributed by
987 user bes-internal on the mailing list.
989 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
991 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
992 selectors, in both main and reject logs.
994 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
997 JH/07 Add malware type "sock" for talking to simple daemon.
999 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1001 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1002 routers/transports under cutthrough routing.
1004 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1005 numbers. Touch up "bool" conditional to keep the same definition.
1007 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1009 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1011 JH/12 Expand items in router/transport headers_add or headers_remove lists
1012 individually rather than the list as a whole. Bug 1452.
1014 Required for reasonable handling of multiple headers_ options when
1015 they may be empty; requires that headers_remove items with embedded
1016 colons must have them doubled (or the list-separator changed).
1018 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1019 view the policy declared in the DMARC record. Currently, $dmarc_status
1020 is a combined value of both the record presence and the result of the
1023 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1025 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1026 dnslookup router and the smtp transport (applying to the forward
1029 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1030 of ldap servers used for a specific lookup. Patch provided by Heiko
1033 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1034 New variable $lookup_dnssec_authenticated for observability.
1036 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1037 Patch submitted by Lars Timman.
1039 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1041 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1042 Requires trusted mode and valid format message id, aborts otherwise.
1043 Patch contributed by Heiko Schlichting.
1045 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1046 certextract with support for various fields. Bug 1358.
1048 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1049 is requested by default, modifiable by smtp transport option
1052 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1053 operate on certificate variables to give certificate fingerprints
1054 Also new ${sha256:cert_variable}.
1056 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1058 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1059 Christian Aistleitner.
1061 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1063 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1064 file. Patch from Wolfgang Breyha.
1066 JH/25 Expand the coverage of the delivery $host and $host_address to
1067 client authenticators run in verify callout. Bug 1476.
1069 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1070 align with daemon_smtp_ports. Bug 72.
1072 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1073 support and error reporting did not work properly.
1075 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1076 and is readable. Patch from Andrew Colin Kissa.
1078 TL/14 Enhance documentation of ${run expansion and how it parses the
1079 commandline after expansion, particularly in the case when an
1080 unquoted variable expansion results in an empty value.
1082 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1084 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1085 by renaming away from it. Addresses GH issue 15, reported by
1088 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1090 TL/15 SECURITY: prevent double expansion in math comparison functions
1091 (can expand unsanitized data). Not remotely exploitable.
1098 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1100 PP/02 Make -n do something, by making it not do something.
1101 When combined with -bP, the name of an option is not output.
1103 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1106 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1107 $sender_host_name and config options to manage this, and basic check
1110 PP/05 DSCP support for outbound connections and control modifier for inbound.
1112 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1113 (Only plugin which currently uses this is kerberos4, which nobody should
1114 be using, but we should make it available and other future plugins might
1115 conceivably use it, even though it would break NAT; stuff *should* be
1116 using channel bindings instead).
1118 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1119 name; added for Sendmail compatibility; requires admin caller.
1120 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1121 just ignore it); requires trusted caller.
1122 Also parse but ignore: -Ac -Am -X<logfile>
1125 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1127 TL/02 Add +smtp_confirmation as a default logging option.
1129 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1130 Patch by Magnus Holmgren from 2007-02-20.
1132 TL/04 Bugzilla 1281 - Spec typo.
1133 Bugzilla 1283 - Spec typo.
1134 Bugzilla 1290 - Spec grammar fixes.
1136 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1138 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1140 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1141 mailing list by by Dmitry Isaikin.
1143 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1145 JH/02 Support "G" suffix to numbers in ${if comparisons.
1147 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1149 NM/01 Bugzilla 1197 - Spec typo
1150 Bugzilla 1196 - Spec examples corrections
1152 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1154 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1155 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1158 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1159 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1161 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1162 "acl {{name}{arg}...}", and optional args on acl condition
1165 JH/05 Permit multiple router/transport headers_add/remove lines.
1167 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1169 JH/07 Avoid using a waiting database for a single-message-only transport.
1170 Performance patch from Paul Fisher. Bugzilla 1262.
1172 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1175 JH/09 Add $headers_added variable, with content from use of ACL modifier
1176 add_header (but not yet added to the message). Bugzilla 199.
1178 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1179 Pulled from Bugzilla 817 by Wolfgang Breyha.
1181 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1183 (nb: this is the same fix as in Exim 4.80.1)
1185 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1188 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1191 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1193 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1194 advertise SMTP AUTH mechanism to us, instead of a generic
1195 protocol violation error. Also, make Exim more robust to bad
1196 data from the Dovecot auth socket.
1198 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1200 When a queue runner is handling a message, Exim first routes the
1201 recipient addresses, during which it prunes them based on the retry
1202 hints database. After that it attempts to deliver the message to
1203 any remaining recipients. It then updates the hints database using
1206 So if a recipient address works intermittently, it can get repeatedly
1207 deferred at routing time. The retry hints record remains fresh so the
1208 address never reaches the final cutoff time.
1210 This is a fairly common occurrence when a user is bumping up against
1211 their storage quota. Exim had some logic in its local delivery code
1212 to deal with this. However it did not apply to per-recipient defers
1213 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1215 This change adds a proper retry rule check during routing so that the
1216 final cutoff time is checked against the message's age. We only do
1217 this check if there is an address retry record and there is not a
1218 domain retry record; this implies that previous attempts to handle
1219 the address had the retry_use_local_parts option turned on. We use
1220 this as an approximation for the destination being like a local
1221 delivery, as in LMTP.
1223 I suspect this new check makes the old local delivery cutoff check
1224 redundant, but I have not verified this so I left the code in place.
1226 TF/02 Correct gecos expansion when From: is a prefix of the username.
1228 Test 0254 submits a message to Exim with the header
1232 When I ran the test suite under the user fanf2, Exim expanded
1233 the header to contain my full name, whereas it should have added
1234 a Resent-Sender: header. It erroneously treats any prefix of the
1235 username as equal to the username.
1237 This change corrects that bug.
1239 GF/01 DCC debug and logging tidyup
1240 Error conditions log to paniclog rather than rejectlog.
1241 Debug lines prefixed by "DCC: " to remove any ambiguity.
1243 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1245 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1246 Bug spotted by Jeremy Harris; was flawed since initial commit.
1247 Would have resulted in OCSP responses post-SNI triggering an Exim
1248 NULL dereference and crash.
1250 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1252 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1253 Bug detection, analysis and fix by Samuel Thibault.
1254 Bugzilla 1331, Debian bug #698092.
1256 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1258 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1259 Server implementation by Todd Lyons, client by JH.
1260 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1261 config variable "prdr_enable" controls whether the server
1262 advertises the facility. If the client requests PRDR a new
1263 acl_data_smtp_prdr ACL is called once for each recipient, after
1264 the body content is received and before the acl_smtp_data ACL.
1265 The client is controlled by both of: a hosts_try_prdr option
1266 on the smtp transport, and the server advertisement.
1267 Default client logging of deliveries and rejections involving
1268 PRDR are flagged with the string "PRDR".
1270 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1271 fclose(). Diagnosis by Todd Lyons.
1273 PP/17 Update configure.default to handle IPv6 localhost better.
1274 Patch by Alain Williams (plus minor tweaks).
1277 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1278 This is now consistent with GnuTLS, and is now documented: the
1279 previous undocumented portable approach to treating the option as
1280 unset was to force an expansion failure. That still works, and
1281 an empty string is now equivalent.
1283 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1284 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1285 not performing validation itself.
1287 PP/20 Added force_command boolean option to pipe transport.
1288 Patch from Nick Koston, of cPanel Inc.
1290 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1293 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1295 PP/21 Fix eximon continuous updating with timestamped log-files.
1296 Broken in a format-string cleanup in 4.80, missed when I repaired the
1297 other false fix of the same issue.
1298 Report and fix from Heiko Schlichting.
1301 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1302 Report from Prashanth Katuri.
1304 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1305 It's SecureTransport, so affects any MacOS clients which use the
1306 system-integrated TLS libraries, including email clients.
1308 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1309 using a MIME ACL for non-SMTP local injection.
1310 Report and assistance in diagnosis by Warren Baker.
1312 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1314 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1316 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1317 last id that failed so it may be referenced in subsequent ACL's.
1319 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1322 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1323 ldap library initialization, allowing self-signed CA's to be
1324 used. Also properly sets require_cert option later in code by
1325 using NULL (global ldap config) instead of ldap handle (per
1326 session). Bug diagnosis and testing by alxgomz.
1328 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1329 the src/util/ subdirectory.
1331 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1332 renamed to Transport Post Delivery Action by Jeremy Harris, as
1335 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1336 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1337 redis_servers = needs to be configured which will be used by the redis
1338 lookup. Patch from Warren Baker, of The Packet Hub.
1340 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1342 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1343 hostname or reverse DNS when processing a host list. Used suggestions
1344 from multiple comments on this bug.
1346 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1348 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1349 Missed a few lines, added it to make the runtest require no keyboard
1352 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1353 contains upper case chars. Make router use caseful_local_part.
1355 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1356 support when GnuTLS has been built with p11-kit.
1362 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1364 This, or similar/improved, will also be change PP/11 of 4.82.
1370 PP/01 Handle short writes when writing local log-files.
1371 In practice, only affects FreeBSD (8 onwards).
1372 Bugzilla 1053, with thanks to Dmitry Isaikin.
1374 NM/01 Bugzilla 949 - Documentation tweak
1376 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1379 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1381 PP/02 Implemented gsasl authenticator.
1383 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1385 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1386 `pkg-config foo` for cflags/libs.
1388 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1389 with rest of GSASL and with heimdal_gssapi.
1391 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1392 `pkg-config foo` for cflags/libs for the TLS implementation.
1394 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1395 properties get this fed in as external SSF. A number of robustness
1396 and debugging improvements to the cyrus_sasl authenticator.
1398 PP/08 cyrus_sasl server now expands the server_realm option.
1400 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1401 Patch by Jeremy Harris.
1403 PP/10 Added dbmjz lookup type.
1405 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1407 PP/12 MAIL args handles TAB as well as SP, for better interop with
1408 non-compliant senders.
1409 Analysis and variant patch by Todd Lyons.
1411 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1412 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1413 Patch from Dirk Mueller <dmueller@suse.com>
1415 PP/13 tls_peerdn now print-escaped for spool files.
1416 Observed some $tls_peerdn in wild which contained \n, which resulted
1417 in spool file corruption.
1419 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1420 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1421 or write after TLS renegotiation, which otherwise led to messages
1424 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1425 as a tracking header (ie: a signed header comes before the signature).
1426 Patch from Wolfgang Breyha.
1428 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1429 comma-sep list; embedded commas doubled.
1431 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1433 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1435 Report and patch from Dmitry Banschikov.
1437 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1438 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1439 failures. We appear to now support TLS1.1+ with Exim.
1441 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1442 lets Exim select keys and certificates based upon TLS SNI from client.
1443 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1444 before an outbound SMTP session. New log_selector, +tls_sni.
1446 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1447 NULL dereference. Report and patch from Alun Jones.
1449 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1450 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1451 Not seeing resolver debug output on NetBSD, but suspect this is a
1452 resolver implementation change.
1454 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1455 Left warnings. Added "eximon gdb" invocation mode.
1457 PP/21 Defaulting "accept_8bitmime" to true, not false.
1459 PP/22 Added -bw for inetd wait mode support.
1461 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1462 locate the relevant includes and libraries. Made this the default.
1464 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1465 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1467 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1468 This may cause build issues on older platforms.
1470 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1471 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1472 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1473 Added SNI support via GnuTLS too.
1474 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1476 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1478 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1479 Applied second patch from Janne, implementing suggestion to default
1480 multiple-strings-in-record handling to match SPF spec.
1482 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1484 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1485 read-only, out of scope).
1486 Patch from Wolfgang Breyha, report from Stuart Northfield.
1488 PP/29 Fix three issues highlighted by clang analyser static analysis.
1489 Only crash-plausible issue would require the Cambridge-specific
1490 iplookup router and a misconfiguration.
1491 Report from Marcin Mirosław.
1493 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1495 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1496 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1497 As part of this, removing so much warning spew let me fix some minor
1498 real issues in debug logging.
1500 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1501 assignment on my part. Fixed.
1503 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1504 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1505 Janne Snabb (who went above and beyond: thank you).
1507 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1508 string otherwise requires a connection and a bunch more work and it's
1509 relatively easy to get wrong. Should also expose TLS library linkage
1512 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1513 64-bit ${eval} (JH/03).
1515 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1516 GNU libc to support some of the 64-bit stuff, should not lead to
1517 conflicts. Defined before os.h is pulled in, so if a given platform
1518 needs to override this, it can.
1520 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1521 protection layer was required, which is not implemented.
1522 Bugzilla 1254, patch from Wolfgang Breyha.
1524 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1525 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1526 tls_dhparam take prime identifiers. Also unbreak combination of
1527 OpenSSL+DH_params+TLSSNI.
1529 PP/39 Disable SSLv2 by default in OpenSSL support.
1535 PP/01 Solaris build fix for Oracle's LDAP libraries.
1536 Bugzilla 1109, patch from Stephen Usher.
1538 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1540 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1543 TF/02 Fix a couple more cases where we did not log the error message
1544 when unlink() failed. See also change 4.74-TF/03.
1546 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1547 lock up or crash if it happened to be inside a call to libc when it
1548 got a SIGUSR1 from exiwhat.
1550 The SIGUSR1 handler appends the current process status to the process
1551 log which is later printed by exiwhat. It used to use the general
1552 purpose logging code to do this, but several functions it calls are
1553 not safe for signals.
1555 The new output code in the SIGUSR1 handler is specific to the process
1556 log, and simple enough that it's easy to inspect for signal safety.
1557 Removing some special cases also simplifies the general logging code.
1558 Removing the spurious timestamps from the process log simplifies
1561 TF/04 Improved ratelimit ACL condition.
1563 The /noupdate option has been deprecated in favour of /readonly which
1564 has clearer semantics. The /leaky, /strict, and /readonly update modes
1565 are mutually exclusive. The update mode is no longer included in the
1566 database key; it just determines when the database is updated. (This
1567 means that when you upgrade Exim will forget old rate measurements.)
1569 Exim now checks that the per_* options are used with an update mode that
1570 makes sense for the current ACL. For example, when Exim is processing a
1571 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1572 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1573 must specify per_mail/readonly. If you omit the update mode it defaults to
1574 /leaky where that makes sense (as before) or /readonly where required.
1576 The /noupdate option is now undocumented but still supported for
1577 backwards compatibility. It is equivalent to /readonly except that in
1578 ACLs where /readonly is required you may specify /leaky/noupdate or
1579 /strict/noupdate which are treated the same as /readonly.
1581 A useful new feature is the /count= option. This is a generalization
1582 of the per_byte option, so that you can measure the throughput of other
1583 aggregate values. For example, the per_byte option is now equivalent
1584 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1586 The per_rcpt option has been generalized using the /count= mechanism
1587 (though it's more complicated than the per_byte equivalence). When it is
1588 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1589 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1590 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1591 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1592 Note that using per_rcpt with a non-readonly update mode in more than
1593 one ACL will cause the recipients to be double-counted. (The per_mail
1594 and per_byte options don't have this problem.)
1596 The handling of very low rates has changed slightly. If the computed rate
1597 is less than the event's count (usually one) then this event is the first
1598 after a long gap. In this case the rate is set to the same as this event's
1599 count, so that the first message of a spam run is counted properly.
1601 The major new feature is a mechanism for counting the rate of unique
1602 events. The new per_addr option counts the number of different
1603 recipients that someone has sent messages to in the last time period. It
1604 behaves like per_rcpt if all the recipient addresses are different, but
1605 duplicate recipient addresses do not increase the measured rate. Like
1606 the /count= option this is a general mechanism, so the per_addr option
1607 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1608 example, measure the rate that a client uses different sender addresses
1609 with the options per_mail/unique=$sender_address. There are further
1610 details in the main documentation.
1612 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1614 TF/06 Removed a few PCRE remnants.
1616 TF/07 Automatically extract Exim's version number from tags in the git
1617 repository when doing development or release builds.
1619 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1620 Bugzilla 879. Patch from Paul Fisher.
1622 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1623 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1626 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1628 PP/05 Variable $av_failed, true if the AV scanner deferred.
1629 Bugzilla 1078. Patch from John Horne.
1631 PP/06 Stop make process more reliably on build failure.
1632 Bugzilla 1087. Patch from Heiko Schlittermann.
1634 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1635 Bugzilla 1089. Patch from Heiko Schlittermann.
1637 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1638 Bugzilla 1131. Patch from Holger Weiß.
1640 PP/09 Handle IPv6 addresses with SPF.
1641 Bugzilla 860. Patch from Wolfgang Breyha.
1643 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1645 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1648 PP/11 match_* no longer expand right-hand-side by default.
1649 New compile-time build option, EXPAND_LISTMATCH_RHS.
1650 New expansion conditions, "inlist", "inlisti".
1652 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1654 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1656 PP/14 fix log_write() format string regression from TF/03.
1657 Bugzilla 1152. Patch from Dmitry Isaikin.
1663 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1665 PP/02 Harmonised TLS library version reporting; only show if debugging.
1666 Layout now matches that introduced for other libraries in 4.74 PP/03.
1668 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1670 PP/04 New "dns_use_edns0" global option.
1672 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1675 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1676 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1678 TK/01 Updated PolarSSL code to 0.14.2.
1679 Bugzilla 1097. Patch from Andreas Metzler.
1681 PP/07 Catch divide-by-zero in ${eval:...}.
1682 Fixes bugzilla 1102.
1684 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1687 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1688 format-string attack -- SECURITY: remote arbitrary code execution.
1690 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1691 time unintentionally subject to list matching rules, letting the header
1692 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1693 arbitrary string expansion). This allowed for information disclosure.
1695 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1696 INT_MIN/-1 -- value coerced to INT_MAX.
1702 NM/01 Workaround for PCRE version dependency in version reporting
1705 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1706 This fixes portability to compilers other than gcc, notably
1707 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1709 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1710 makefiles for portability to HP-UX and POSIX correctness.
1712 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1713 Also via indented variable definition in the Makefile.
1714 (Debugging by Oliver Heesakkers).
1716 PP/02 Restore caching of spamd results with expanded spamd_address.
1717 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1719 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1720 Improves build reliability. Fix from: Frank Elsner
1722 NM/02 Fix wide character breakage in the rfc2047 coding
1723 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1725 NM/03 Allow underscore in dnslist lookups
1726 Fixes bug 1026. Patch from Graeme Fowler
1728 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1729 Code patches from Adam Ciarcinski of NetBSD.
1731 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1734 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1735 is logged, to avoid truncation. Patch from John Horne.
1737 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1738 Patch from Jakob Hirsch.
1740 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1741 SQL string expansion failure details.
1742 Patch from Andrey Oktyabrski.
1744 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1745 Patch from Simon Arlott.
1747 PP/09 New lookups functionality failed to compile on old gcc which rejects
1748 extern declarations in function scope.
1749 Patch from Oliver Fleischmann
1751 PP/10 Use sig_atomic_t for flags set from signal handlers.
1752 Check getgroups() return and improve debugging.
1753 Fixed developed for diagnosis in bug 927 (which turned out to be
1756 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1757 Patch from Mark Zealey.
1759 PP/12 Bugzilla 1056: Improved spamd server selection.
1760 Patch from Mark Zealey.
1762 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1763 Based on patch from Heiko Schlittermann.
1765 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1766 Patch from Uwe Doering, sign-off by Michael Haardt.
1768 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1769 variable declaration deep within a block. Bug and patch from
1772 PP/15 lookups-Makefile IRIX compatibility coercion.
1774 PP/16 Make DISABLE_DKIM build knob functional.
1776 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1777 Patch by Simon Arlott
1779 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1780 variable argument macros. Our copy now differs from upstream.
1786 TF/01 Failure to get a lock on a hints database can have serious
1787 consequences so log it to the panic log.
1789 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1790 controlled using the smtp_confirmation log selector.
1792 TF/03 Include the error message when we fail to unlink a spool file.
1794 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1795 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1796 for maintaining out-of-tree patches for some time.
1798 PP/01 Bugzilla 139: Documentation and portability issues.
1799 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1800 Handle per-OS dynamic-module compilation flags.
1802 PP/02 Let /dev/null have normal permissions.
1803 The 4.73 fixes were a little too stringent and complained about the
1804 permissions on /dev/null. Exempt it from some checks.
1805 Reported by Andreas M. Kirchwitz.
1807 PP/03 Report version information for many libraries, including
1808 Exim version information for dynamically loaded libraries. Created
1809 version.h, now support a version extension string for distributors
1810 who patch heavily. Dynamic module ABI change.
1812 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1813 privilege escalation vulnerability whereby the Exim run-time user
1814 can cause root to append content of the attacker's choosing to
1817 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1820 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1821 If dropping privileges for untrusted macros, we disabled normal logging
1822 on the basis that it would fail; for the Exim run-time user, this is not
1823 the case, and it resulted in successful deliveries going unlogged.
1824 Fixed. Reported by Andreas Metzler.
1830 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1831 only prepend for the Resent-* case. Fixes regression introduced in
1832 Exim 4.70 by NM/22 for Bugzilla 607.
1834 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1835 increasing numbers of administrators be bitten by this.
1837 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1839 PP/03 Bugzilla 994: added openssl_options main configuration option.
1841 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1843 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1845 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1847 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1848 without a peer certificate, leading to a segfault because of an
1849 assumption that peers always have certificates. Be a little more
1850 paranoid. Problem reported by Martin Tscholak.
1852 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1853 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1854 NB: ClamAV planning to remove STREAM in "middle of 2010".
1855 CL also introduces -bmalware, various -d+acl logging additions and
1856 more caution in buffer sizes.
1858 PP/09 Implemented reverse_ip expansion operator.
1860 PP/10 Bugzilla 937: provide a "debug" ACL control.
1862 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1864 PP/12 Bugzilla 973: Implement --version.
1866 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1868 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1870 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1872 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1873 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1874 ignore trailing whitespace.
1876 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1878 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1881 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1882 Notification from Dr Andrew Aitchison.
1884 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1885 ExtendedDetectionInfo response format.
1886 Notification from John Horne.
1888 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1891 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1892 XSL and documented dependency on system catalogs, with examples of how
1895 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1898 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1899 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1900 configuration file which is writeable by the Exim user or group.
1902 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1903 of configuration files to cover files specified with the -C option if
1904 they are going to be used with root privileges, not just the default
1907 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1908 option (effectively making it always true).
1910 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1911 files to be used while preserving root privileges.
1913 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1914 that rogue child processes cannot use them.
1916 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1917 run-time user, instead of root.
1919 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1920 Exim run-time user without dropping privileges.
1922 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1923 result string, instead of calling string_vformat() twice with the same
1926 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1927 for other users. Others should always drop root privileges if they use
1928 -C on the command line, even for a whitelisted configure file.
1930 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1932 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1938 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1939 $data_path, and $header_path variables; fixed documentation bugs and
1942 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1943 exipick to access non-standard spools, including the "frozen" queue
1946 NM/01 Bugzilla 965: Support mysql stored procedures.
1947 Patch from Alain Williams
1949 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1951 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1952 Patch from Andreas Metzler
1954 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1955 Patch from Kirill Miazine
1957 NM/05 Bugzilla 671: Added umask to procmail example.
1959 JJ/03 installed exipick 20100323.0, fixing doc bug
1961 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1962 directory. Notification and patch from Dan Rosenberg.
1964 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1966 TK/02 Improve log output when DKIM signing operation fails.
1968 MH/01 Treat the transport option dkim_domain as a colon separated
1969 list, not as a single string, and sign the message with each element,
1970 omitting multiple occurences of the same signer.
1972 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1973 Bugzilla 985, 986. Patch by Simon Arlott
1975 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1976 Patch by Simon Arlott
1978 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1979 MBX locking. Notification from Dan Rosenberg.
1985 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1987 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1989 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1991 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1993 NM/04 Bugzilla 915: Flag broken perl installation during build.
1999 TK/01 Added patch by Johannes Berg that expands the main option
2000 "spamd_address" if it starts with a dollar sign.
2002 TK/02 Write list of recipients to X-Envelope-Sender header when building
2003 the mbox-format spool file for content scanning (suggested by Jakob
2006 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2007 (http://www.dcc-servers.net/) support via dccifd. Activated by
2008 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2010 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2011 by Mark Daniel Reidel <mr@df.eu>.
2013 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2014 When building exim an external PCRE library is now needed -
2015 PCRE is a system library on the majority of modern systems.
2016 See entry on PCRE_LIBS in EDITME file.
2018 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2019 conversation. Added nologin parameter to request.
2020 Patch contributed by Kirill Miazine.
2022 TF/01 Do not log submission mode rewrites if they do not change the address.
2024 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2026 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2027 log files in place. Contributed by Roberto Lima.
2029 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2031 TF/03 Bugzilla 615: When checking the local_parts router precondition
2032 after a local_part_suffix or local_part_prefix option, Exim now
2033 does not use the address's named list lookup cache, since this
2034 contains cached lookups for the whole local part.
2036 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2037 Robert Millan. Documentation is in experimental-spec.txt.
2039 TF/04 Bugzilla 668: Fix parallel build (make -j).
2041 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2043 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2044 Patch provided by Jan Srzednicki.
2046 TF/05 Leading white space used to be stripped from $spam_report which
2047 wrecked the formatting. Now it is preserved.
2049 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2050 that they are available at delivery time.
2052 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2054 TF/08 TLS error reporting now respects the incoming_interface and
2055 incoming_port log selectors.
2057 TF/09 Produce a more useful error message if an SMTP transport's hosts
2058 setting expands to an empty string.
2060 NM/06 Bugzilla 744: EXPN did not work under TLS.
2061 Patch provided by Phil Pennock.
2063 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2064 Patch provided by Richard Godbee.
2066 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2067 acl_smtp_notquit, added index entry.
2069 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2070 Patch provided by Eugene Bujak.
2072 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2073 accept(). Patch provided by Maxim Dounin.
2075 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2076 Patch provided by Phil Pennock.
2078 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2080 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2081 Patch provided by Brad "anomie" Jorsch.
2083 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2084 Patch provided by Dean Brooks.
2086 TK/05 Add native DKIM support (does not depend on external libraries).
2088 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2089 Patch provided by Graeme Fowler.
2091 NM/16 Bugzilla 851: Documentation example syntax fix.
2093 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2095 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2098 NM/19 Bugzilla 745: TLS version reporting.
2099 Patch provided by Phil Pennock.
2101 NM/20 Bugzilla 167: bool: condition support.
2102 Patch provided by Phil Pennock.
2104 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2105 clients. Patch provided by Phil Pennock.
2107 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2108 Patch provided by Brad "anomie" Jorsch.
2110 NM/23 Bugzilla 687: Fix misparses in eximstats.
2111 Patch provided by Heiko Schlittermann.
2113 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2114 Patch provided by Heiko Schlittermann.
2116 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2117 plus update to original patch.
2119 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2121 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2122 Patch provided by David Brownlee.
2124 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2126 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2128 NM/30 Bugzilla 888: TLS documentation bugfixes.
2130 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2132 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2133 Unlike the original bugzilla I have changed all shell scripts in src tree.
2135 NM/33 Bugzilla 898: Transport filter timeout fix.
2136 Patch by Todd Rinaldo.
2138 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2139 Patch by Serge Demonchaux.
2141 NM/35 Bugzilla 39: Base64 decode bug fixes.
2142 Patch by Jakob Hirsch.
2144 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2146 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2148 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2150 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2156 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2157 ALT-N's libdkim that I have put here:
2158 http://duncanthrax.net/exim-experimental/
2160 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2161 were called 'true' and it seems that C99 defines that as a reserved
2162 keyword to be used with 'bool' variable types. That means you could
2163 not include C99-style headers which use bools without triggering
2164 build errors in sieve.c.
2166 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2167 as mailq or other aliases. Changed the --help handling significantly
2168 to do whats expected. exim_usage() emits usage/help information.
2170 SC/01 Added the -bylocaldomain option to eximstats.
2172 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2174 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2176 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2182 PH/01 Another patch from the Sieve maintainer.
2184 PH/02 When an IPv6 address is converted to a string for single-key lookup
2185 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2186 dots are used instead of colons so that keys in lsearch files need not
2187 contain colons. This was done some time before quoting was made available
2188 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2189 (notated using the quote facility) so as to distinguish them from IPv4
2190 keys. This meant that lookups for IP addresses in host lists did not work
2191 for iplsearch lookups.
2193 This has been fixed by arranging for IPv6 addresses to be expressed with
2194 colons if the lookup type is iplsearch. This is not incompatible, because
2195 previously such lookups could never work.
2197 The situation is now rather anomalous, since one *can* have colons in
2198 ordinary lsearch keys. However, making the change in all cases is
2199 incompatible and would probably break a number of configurations.
2201 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2204 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2205 conversion specification without a maximum field width, thereby enabling
2206 a rogue spamd server to cause a buffer overflow. While nobody in their
2207 right mind would setup Exim to query an untrusted spamd server, an
2208 attacker that gains access to a server running spamd could potentially
2209 exploit this vulnerability to run arbitrary code as the Exim user.
2211 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2212 $primary_hostname instead of what libspf2 thinks the hosts name is.
2214 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2215 a directory entry by the name of the lookup key. Previously, if a
2216 symlink pointed to a non-existing file or a file in a directory that
2217 Exim lacked permissions to read, a lookup for a key matching that
2218 symlink would fail. Now it is enough that a matching directory entry
2219 exists, symlink or not. (Bugzilla 503.)
2221 PH/03 The body_linecount and body_zerocount variables are now exported in the
2224 PH/04 Added the $dnslist_matched variable.
2226 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2227 This means they are set thereafter only if the connection becomes
2230 PH/06 Added the client_condition to authenticators so that some can be skipped
2231 by clients under certain conditions.
2233 PH/07 The error message for a badly-placed control=no_multiline_responses left
2234 "_responses" off the end of the name.
2236 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2238 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2239 (without spaces) instead of just copying the configuration text.
2241 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2243 PH/11 Added $max_received_linelength.
2245 PH/12 Added +ignore_defer and +include_defer to host lists.
2247 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2248 way in which PCRE > 7.0 is built.
2250 PH/14 Implemented queue_only_load_latch.
2252 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2253 MAIL command. The effect was to mangle the value on 64-bit systems.
2255 PH/16 Another patch from the Sieve maintainer.
2257 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2259 PH/18 If a system quota error occurred while trying to create the file for
2260 a maildir delivery, the message "Mailbox is full" was not appended to the
2261 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2262 only to a quota excession during the actual writing of the file.
2264 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2265 characters?) which causes problems in log lines. The DN values are now
2266 passed through string_printing() before being added to log lines.
2268 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2269 and InterBase are left for another time.)
2271 PH/21 Added message_body_newlines option.
2273 PH/22 Guard against possible overflow in moan_check_errorcopy().
2275 PH/23 POSIX allows open() to be a macro; guard against that.
2277 PH/24 If the recipient of an error message contained an @ in the local part
2278 (suitably quoted, of course), incorrect values were put in $domain and
2279 $local_part during the evaluation of errors_copy.
2285 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2286 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2289 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2290 issue a MAIL command.
2292 PH/02 In an ACL statement such as
2294 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2296 if a client was not listed at all, or was listed with a value other than
2297 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2298 the condition was not true (as it should be), so access was not denied.
2299 The bug was that the ! inversion was incorrectly passed on to the second
2300 item. This has been fixed.
2302 PH/03 Added additional dnslists conditions == and =& which are different from
2303 = and & when the dns lookup returns more than one IP address.
2305 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2306 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2308 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2309 FSYNC, which compiles an option called disable_fsync that allows for
2310 bypassing fsync(). The documentation is heavily laced with warnings.
2312 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2314 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2315 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2316 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2317 including adding "make clean"; (3) Added -fPIC when compiling the test
2318 dynamically loaded module, to get rid of a warning.
2320 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2321 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2322 = 0s. The bug is otherwise harmless.
2324 PH/07 There was a bug in the dovecot authenticator such that the value of
2325 $auth1 could be overwritten, and so not correctly preserved, after a
2326 successful authentication. This usually meant that the value preserved by
2327 the server_setid option was incorrect.
2329 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2331 PH/09 Installed PCRE release 7.0.
2333 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2334 run for batched SMTP input. It is now run at the start of every message
2335 in the batch. While fixing this I discovered that the process information
2336 (output by running exiwhat) was not always getting set for -bs and -bS
2337 input. This is fixed, and it now also says "batched" for BSMTP.
2339 PH/11 Added control=no_pipelining.
2341 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2342 patch, slightly modified), and move the expansion of helo_data till after
2343 the connection is made in the smtp transport (so it can use these
2346 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2348 PH/14 Added log_selector = +pid.
2350 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2352 PH/16 Add ${if forany and ${if forall.
2354 PH/17 Added dsn_from option to vary the From: line in DSNs.
2356 PH/18 Flush SMTP output before performing a callout, unless control =
2357 no_callout_flush is set.
2359 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2360 was true (the default) a successful delivery failed to delete the retry
2361 item, thus causing premature timeout of the address. The bug is now
2364 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2366 PH/21 Long custom messages for fakedefer and fakereject are now split up
2367 into multiline responses in the same way that messages for "deny" and
2368 other ACL rejections are.
2370 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2371 with slight modification.
2373 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2374 draft, changing the syntax and factoring some duplicate code".
2376 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2377 for deliveries of the second and subsequent messages over the same SMTP
2380 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2381 ${reduce, with only minor "tidies".
2383 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2385 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2386 expansion side effects.
2388 PH/27 When a message times out after an over-quota error from an Exim-imposed
2389 quota, the bounce message says "mailbox is full". This message was not
2390 being given when it was a system quota that was exceeded. It now should
2393 MH/03 Made $recipients available in local_scan(). local_scan() already has
2394 better access to the recipient list through recipients_list[], but
2395 $recipients can be useful in postmaster-provided expansion strings.
2397 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2398 in the case of a MAIL command with additional options following the
2399 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2400 were accidentally chopped off.
2402 PH/29 SMTP synchronization checks are implemented when a command is read -
2403 there is a check that no more input is waiting when there shouldn't be
2404 any. However, for some commands, a delay in an ACL can mean that it is
2405 some time before the response is written. In this time, more input might
2406 arrive, invalidly. So now there are extra checks after an ACL has run for
2407 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2408 pipelining has not been advertised.
2410 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2412 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2413 returned in the SMTP response when smtp_return_error_details was false.
2414 This has been fixed.
2416 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2417 instead of the C I/O that was originally supplied, because problems were
2418 reported on Solaris.
2420 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2421 Exim which did not show up earlier: it was assuming that a call to
2422 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2423 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2424 was a macro that became an assignment, so it seemed to work. This has
2425 changed to a proper function call with a void return, hence the compile
2426 error. Exim's code has been fixed.
2428 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2431 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2433 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2435 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2436 because it turns out to be too big or there is a timeout) there is no
2437 "Completed" line in the log. When some messages of this type were
2438 selected by exigrep, they were listed as "not completed". Others were
2439 picked up by some special patterns. I have improved the selection
2440 criteria to be more general.
2442 PH/38 The host_find_failed option in the manualroute router can now be set
2443 to "ignore", to completely ignore a host whose IP address cannot be
2444 found. If all hosts are ignored, the behaviour is controlled by the new
2445 host_all_ignored option.
2447 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2448 homing or because of multiple MX records with /mx) generated more than
2449 one IP address, and the following item turned out to be the local host,
2450 all the secondary addresses of the first item were incorrectly removed
2451 from the list, along with the local host and any following hosts (which
2452 is what is supposed to happen).
2454 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2455 whoever called Exim into the -H file. In the case of the daemon it was
2456 behaving confusingly. When first started, it used values for whoever
2457 started the daemon, but after a SIGHUP it used the Exim user (because it
2458 calls itself on a restart). I have changed the code so that it now always
2461 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2462 message are rejected with the same error (e.g. no authentication or bad
2463 sender address), and a DATA command is nevertheless sent (as can happen
2464 with PIPELINING or a stupid MUA), the error message that was given to the
2465 RCPT commands is included in the rejection of the DATA command. This is
2466 intended to be helpful for MUAs that show only the final error to their
2469 PH/42 Another patch from the Sieve maintainer.
2471 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2472 Eximstats - Fixed some broken HTML links and added missing column headers
2474 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2475 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2477 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2479 PH/43 Yet another patch from the Sieve maintainer.
2481 PH/44 I found a way to check for a TCP/IP connection going away before sending
2482 the response to the final '.' that terminates a message, but only in the
2483 case where the client has not sent further data following the '.'
2484 (unfortunately, this is allowed). However, in many cases there won't be
2485 any further data because there won't be any more messages to send. A call
2486 to select() can be used: if it shows that the input is "ready", there is
2487 either input waiting, or the socket has been closed. An attempt to read
2488 the next input character can distinguish the two cases. Previously, Exim
2489 would have sent an OK response which the client would never have see.
2490 This could lead to message repetition. This fix should cure that, at
2491 least in a lot of common cases.
2493 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2494 advertised in response to EHLO.
2500 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2501 fixed by 4.65/MH/01 (is this a record?) are fixed:
2503 (i) An empty string was always treated as zero by the numeric comparison
2504 operators. This behaviour has been restored.
2506 (ii) It is documented that the numeric comparison operators always treat
2507 their arguments as decimal numbers. This was broken in that numbers
2508 starting with 0 were being interpreted as octal.
2510 While fixing these problems I realized that there was another issue that
2511 hadn't been noticed. Values of message_size_limit (both the global option
2512 and the transport option) were treated as octal if they started with 0.
2513 The documentation was vague. These values are now always treated as
2514 decimal, and I will make that clear in the documentation.
2520 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2521 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2524 MH/01 Don't check that the operands of numeric comparison operators are
2525 integers when their expansion is in "skipping" mode (fixes bug
2526 introduced by 4.64-PH/07).
2528 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2529 child addresses, Exim now panics and dies. Previously, because the count
2530 is held in a short int, deliveries were likely to be lost. As such a
2531 large number of recipients for a single message is ridiculous
2532 (performance will be very, very poor), I have chosen to impose a limit
2533 rather than extend the field.
2539 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2540 leftover -K file (the existence of which was triggered by #402).
2541 While we were at it, introduced process PID as part of the -K
2542 filename. This should rule out race conditions when creating
2545 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2546 processing considerably. Previous code took too long for large mails,
2547 triggering a timeout which in turn triggers #401.
2549 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2550 in the DK code in transports.c. sendfile() is not really portable,
2551 hence the _LINUX specificness.
2553 TF/01 In the add_headers option to the mail command in an Exim filter,
2554 there was a bug that Exim would claim a syntax error in any
2555 header after the first one which had an odd number of characters
2558 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2559 callout verification, Exim cached a "reject" for the entire domain. This
2560 is correct for most verifications, but it is not correct for a recipient
2561 verification with use_sender or use_postmaster set, because in that case
2562 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2563 case of MAIL FROM:<> rejection from other early rejections (e.g.
2564 rejection of HELO). When verifying a recipient using a non-null MAIL
2565 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2566 Whatever the result of the callout, the value of the domain cache is
2567 left unchanged (for any other kind of callout, getting as far as trying
2568 RCPT means that the domain itself is ok).
2570 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2573 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2574 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2575 session, the daemon ignores ECONNECT errors and logs others; it now
2576 ignores EPIPE as well.
2578 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2579 (quoted-printable decoding).
2581 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2582 later the small subsequent patch to fix an introduced bug.
2584 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2586 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2588 PH/08 An error is now given if message_size_limit is specified negative.
2590 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2591 to be given (somewhat) arbitrary names.
2593 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2596 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2597 miscellaneous code fixes
2599 PH/10 Added the log_reject_target ACL modifier to specify where to log
2602 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2603 hostname. This is wrong, because it relates to the incoming message (and
2604 probably the interface on which it is arriving) and not to the outgoing
2605 callout (which could be using a different interface). This has been
2606 changed to use the value of the helo_data option from the smtp transport
2607 instead - this is what is used when a message is actually being sent. If
2608 there is no remote transport (possible with a router that sets up host
2609 addresses), $smtp_active_hostname is used.
2611 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2612 tweaks were necessary in order to get it to work (see also 21 below):
2613 (a) The code assumed that strncpy() returns a negative number on buffer
2614 overflow, which isn't the case. Replaced with Exim's string_format()
2616 (b) There were several signed/unsigned issues. I just did the minimum
2617 hacking in of casts. There is scope for a larger refactoring.
2618 (c) The code used strcasecmp() which is not a standard C function.
2619 Replaced with Exim's strcmpic() function.
2620 (d) The code set only $1; it now sets $auth1 as well.
2621 (e) A simple test gave the error "authentication client didn't specify
2622 service in request". It would seem that Dovecot has changed its
2623 interface. Fortunately there's a specification; I followed it and
2624 changed what the client sends and it appears to be working now.
2626 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2629 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2630 address A is aliased to B and C, where B exists and C does not. Without
2631 -v the output is "A verified" because verification stops after a
2632 successful redirection if more than one address is generated. However,
2633 with -v the child addresses are also verified. Exim was outputting "A
2634 failed to verify" and then showing the successful verification for C,
2635 with its parentage. It now outputs "B failed to verify", showing B's
2636 parentage before showing the successful verification of C.
2638 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2639 look up a TXT record in a specific list after matching in a combined
2642 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2643 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2644 they consult the DNS. I had assumed they would set it the way they
2645 wanted; and indeed my experiments on Linux seem to show that in some
2646 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2647 To be on the safe side, however, I have now made the interface to
2648 host_find_byname() similar to host_find_bydns(), with an argument
2649 containing the DNS resolver options. The host_find_byname() function now
2650 sets these options at its start, just as host_find_bydns() does. The smtp
2651 transport options dns_qualify_single and dns_search_parents are passed to
2652 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2653 of host_find_byname() use the default settings of RES_DEFNAMES
2654 (qualify_single) but not RES_DNSRCH (search_parents).
2656 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2657 spool_read_header() do less string testing, by means of a preliminary
2658 switch on the second character of optional "-foo" lines. (This is
2659 overdue, caused by the large number of possibilities that now exist.
2660 Originally there were few.) While I was there, I also converted the
2661 str(n)cmp tests so they don't re-test the leading "-" and the first
2662 character, in the hope this might squeeze out yet more improvement.
2664 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2665 flag allowing group syntax was set by the header_syntax check but not
2666 turned off, possible causing trouble later; (2) The flag was not being
2667 set at all for the header_verify test, causing "group"-style headers to
2668 be rejected. I have now set it in this case, and also caused header_
2669 verify to ignore an empty address taken from a group. While doing this, I
2670 came across some other cases where the code for allowing group syntax
2671 while scanning a header line wasn't quite right (mostly, not resetting
2672 the flag correctly in the right place). These bugs could have caused
2673 trouble for malformed header lines. I hope it is now all correct.
2675 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2676 with the "reply" argument non-NULL. The code, however (which originally
2677 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2678 but it didn't always do it. This confused somebody who was copying the
2679 code for some other use. I have removed all the tests.
2681 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2682 feature that was used to support insecure browsers during the U.S. crypto
2683 embargo. It requires special client support, and Exim is probably the
2684 only MTA that supported it -- and would never use it because real RSA is
2685 always available. This code has been removed, because it had the bad
2686 effect of slowing Exim down by computing (never used) parameters for the
2687 RSA_EXPORT functionality.
2689 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2690 authenticator to fail if there's a tab character in the incoming data
2691 (there should never be unless someone is messing about, as it's supposed
2692 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2693 if the connection is using TLS or if the remote IP is the same as the
2694 local IP, and the "valid-client-cert option" if a client certificate has
2697 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2698 authenticators. This can be used for authorization after authentication
2699 succeeds. (In the case of plaintext, it servers for both authentication
2702 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2703 if any retry times were supplied.
2705 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2706 connection, where there is no client IP address to check. In this
2707 situation, the verify now always succeeds.
2709 PH/25 Applied John Jetmore's -Mset patch.
2711 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2713 PH/27 In a string expansion for a processed (not raw) header when multiple
2714 headers of the same name were present, leading whitespace was being
2715 removed from all of them, but trailing whitespace was being removed only
2716 from the last one. Now trailing whitespace is removed from each header
2717 before concatenation. Completely empty headers in a concatenation (as
2718 before) are ignored.
2720 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2721 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2723 PH/29 [Removed. This was a change that I later backed out, and forgot to
2724 correct the ChangeLog entry (that I had efficiently created) before
2725 committing the later change.]
2727 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2728 address errors (4xx response to RCPT) over the same connection as other
2729 messages routed to the same hosts. Such deliveries are always "forced",
2730 so retry times are not inspected. This resulted in far too many retries
2731 for the affected addresses. The effect occurred only when there were more
2732 hosts than the hosts_max_try setting in the smtp transport when it had
2733 the 4xx errors. Those hosts that it had tried were not added to the list
2734 of hosts for which the message was waiting, so if all were tried, there
2735 was no problem. Two fixes have been applied:
2737 (i) If there are any address or message errors in an SMTP delivery, none
2738 of the hosts (tried or untried) are now added to the list of hosts
2739 for which the message is waiting, so the message should not be a
2740 candidate for sending over the same connection that was used for a
2741 successful delivery of some other message. This seems entirely
2742 reasonable: after all the message is NOT "waiting for some host".
2743 This is so "obvious" that I'm not sure why it wasn't done
2744 previously. Hope I haven't missed anything, but it can't do any
2745 harm, as the worst effect is to miss an optimization.
2747 (ii) If, despite (i), such a delivery is accidentally attempted, the
2748 routing retry time is respected, so at least it doesn't keep
2749 hammering the server.
2751 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2752 in ${readsocket because some servers need this prod.
2754 PH/32 Added some extra debug output when updating a wait-xxx database.
2756 PH/33 The hint "could be header name not terminated by colon", which has been
2757 given for certain expansion errors for a long time, was not being given
2758 for the ${if def:h_colon_omitted{... case.
2760 PH/34 The spec says: "With one important exception, whenever a domain list is
2761 being scanned, $domain contains the subject domain." There was at least
2762 one case where this was not true.
2764 PH/35 The error "getsockname() failed: connection reset by peer" was being
2765 written to the panic log as well as the main log, but it isn't really
2766 panic-worthy as it just means the connection died rather early on. I have
2767 removed the panic log writing for the ECONNRESET error when getsockname()
2770 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2771 runs only) independently of the message's sender address. This meant
2772 that, if the 4xx error was in fact related to the sender, a different
2773 message to the same recipient with a different sender could confuse
2774 things. In particular, this can happen when sending to a greylisting
2775 server, but other circumstances could also provoke similar problems.
2776 I have changed the default so that the retry time for these errors is now
2777 based a combination of the sender and recipient addresses. This change
2778 can be overridden by setting address_retry_include_sender=false in the
2781 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2782 remote server are returned as part of bounce messages. This was not
2783 happening for LMTP over a pipe (the lmtp transport), but now it is the
2784 same for both kinds of LMTP.
2786 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2787 in header lines that were in fact CNAMEs. This is no longer the case.
2789 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2790 and queue runs started by the daemon processed all messages. This has
2791 been fixed so that -R and -S can now usefully be given with -q<time>.
2793 PH/40 Import PCRE release 6.7 (fixes some bugs).
2795 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2797 PH/42 Give an error if -q is specified more than once.
2799 PH/43 Renamed the variables $interface_address and $interface_port as
2800 $received_ip_address and $received_port, to make it clear that these
2801 values apply to message reception, and not to the outgoing interface when
2802 a message is delivered. (The old names remain recognized, of course.)
2804 PH/44 There was no timeout on the connect() call when using a Unix domain
2805 socket in the ${readsocket expansion. There now is.
2807 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2808 be meaningful with "accept".
2810 SC/01 Eximstats V1.43
2811 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2813 SC/02 Eximstats V1.44
2814 Use a glob alias rather than an array ref in the generated
2815 parser. This improves both readability and performance.
2817 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2818 Collect SpamAssassin and rejection statistics.
2819 Don't display local sender or destination tables unless
2820 there is data to show.
2821 Added average volumes into the top table text output.
2823 SC/04 Eximstats V1.46
2824 Collect data on the number of addresses (recipients)
2825 as well as the number of messages.
2827 SC/05 Eximstats V1.47
2828 Added 'Message too big' to the list of mail rejection
2829 reasons (thanks to Marco Gaiarin).
2831 SC/06 Eximstats V1.48
2832 Mainlog lines which have GMT offsets and are too short to
2833 have a flag are now skipped.
2835 SC/07 Eximstats V1.49 (Alain Williams)
2836 Added the -emptyok flag.
2838 SC/08 Eximstats V1.50
2839 Fixes for obtaining the IP address from reject messages.
2841 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2842 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2843 whitespace changes from 4.64-PH/27
2845 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2848 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2849 are found, allow negative numbers in numeric criteria)
2851 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2853 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2854 to match changes made in 4.64-PH/43
2856 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2858 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2859 "domain" is actually an IP address, and force a failure. This locks out
2860 those revolvers/nameservers that support "A-for-A" lookups, in
2861 contravention of the specifications.
2863 PH/48 When a host name was looked up from an IP address, and the subsequent
2864 forward lookup of the name timed out, the host name was left in
2865 $sender_host_name, contrary to the specification.
2867 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2868 restricted to single-key lookups, Exim was not diagnosing an error if
2869 * or *@ was used with a query-style lookup.
2871 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2873 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2874 long ago, but noone interested enough thought of it. Let's just say that
2875 the "1.1" means that there are some new functions that weren't there at
2876 some point in the past.
2878 PH/51 Error processing for expansion failure of helo_data from an smtp
2879 transport during callout processing was broken.
2881 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2882 tested/used via the -bh/-bhc/-bs options.
2884 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2885 bug, fixed in subsequent PCRE releases).
2887 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2888 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2890 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2896 SC/01 Use a glob alias rather than an array ref in eximstats generated
2897 parser. This improves both readability and performance.
2899 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2900 Don't display local sender or destination tables in eximstats unless
2901 there is data to show.
2902 Added average volumes into the eximstats top table text output.
2904 SC/03 Collect data on the number of addresses (recipients) as well
2905 as the number of messages in eximstats.
2907 TF/01 Correct an error in the documentation for the redirect router. Exim
2908 does (usually) call initgroups() when daemonizing.
2910 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2911 with consistent privilege compared to when running as a daemon.
2913 TF/03 Note in the spec that $authenticated_id is not set for local
2914 submissions from trusted users.
2916 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2917 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2919 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2920 by adding some example configuration directives to the default
2921 configuration file. A little bit of work is required to uncomment the
2922 directives and define how usernames and passwords are checked, but
2923 there is now a framework to start from.
2925 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2926 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2927 without this. I don't know how relevant this is to other LDAP libraries.
2929 PH/02 Add the verb name to the "unknown ACL verb" error.
2931 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2933 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2935 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2936 directory not expanded when it should be if an expanded home directory
2937 was set for the address (which is overridden by the transport).
2939 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2942 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2943 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2944 because it is too late at that time, and has no effect.
2946 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2947 security issue with \' (bugzilla #107). I could not use the
2948 PQescapeStringConn() function, because it needs a PGconn value as one of
2951 PH/08 When testing addresses using -bt, indicate those final addresses that
2952 are duplicates that would not cause an additional delivery. At least one
2953 person was confused, thinking that -bt output corresponded to deliveries.
2954 (Suppressing duplicates isn't a good idea as you lose the information
2955 about possibly different redirections that led to the duplicates.)
2957 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2958 systems where poll() doesn't work, in particular OS X.
2960 PH/10 Added more information to debugging output for retry time not reached.
2962 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2963 operations in malware.c.
2965 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2968 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2969 syslog_duplication set false, log lines that would normally be written
2970 both the the main log and to the reject log were not written to syslog at
2973 PH/14 In the default configuration, change the use of "message" in ACL warn
2974 statements to "add_header".
2976 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2977 not followed by a command (e.g. "seen endif").
2979 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2980 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2983 PH/17 Added extra conditions to the default value of delay_warning_condition
2987 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2988 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2989 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2992 The Auto-Submitted: and various List- headers are standardised, whereas I
2993 don't think Precedence: ever was.
2995 PH/18 Refactored debugging code in route_finduser() to show more information,
2996 in particular, the error code if getpwnam() issues one.
2998 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2999 This is apparently needed in addition to the PH/07 change above to avoid
3000 any possible encoding problems.
3002 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3003 but not after initializing Perl.
3005 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3006 output them only if debugging. By default they are written stderr,
3007 apparently, which is not desirable.
3009 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3012 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3015 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3017 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3018 authenticated or an ident call has been made. Suppress the default
3019 values for $authenticated_id and $authenticated_sender (but permit -oMai
3020 and -oMas) when testing with -bh.
3022 PH/24 Re-jigged the order of the tests in the default configuration so that the
3023 tests for valid domains and recipients precede the DNS black list and CSA
3024 tests, on the grounds that those ones are more expensive.
3026 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3027 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3028 This bug exists in every version of Exim that I still have, right back to
3031 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3032 However, an attempt to turn on case-sensitivity in a regex key by
3033 including (?-i) didn't work because the subject string was already
3034 lowercased, and the effects were non-intuitive. It turns out that a
3035 one-line patch can be used to allow (?-i) to work as expected.
3041 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3042 other effects) broke the use of negated acl sub-conditions.
3044 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3047 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3048 "Deny" causes Exim to reject the incoming connection with a 554 error.
3049 Unfortunately, if there is a major crisis, such as a disk failure,
3050 tcp-wrappers gives "deny", whereas what one would like would be some
3051 kind of temporary error. A kludge has been added to help with this.
3052 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3053 554 error is used if errno is still zero or contains ENOENT (which occurs
3054 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3057 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3059 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3060 errors. Otherwise a message that provokes a temporary error (when other
3061 messages do not) can cause a whole host to time out.
3063 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3064 addresses were routed directly to files or pipes from a redirect router.
3065 File deliveries just didn't batch; pipe deliveries might have suffered
3068 PH/06 A failure to get a lock for a hints database would erroneously always say
3069 "Failed to get write lock", even when it was really a read lock.
3071 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3072 of 0600. This has been changed to use the value of the lockfile_mode
3073 option (which defaults to 0600).
3075 PH/08 Applied small patch from the Sieve maintainer.
3077 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3078 folder from quota calculations, a direct delivery into this folder messed
3079 up the contents of the maildirsize file. This was because the regex was
3080 used only to exclude .Trash (or whatever) when the size of the mailbox
3081 was calculated. There was no check that a delivery was happening into an
3082 excluded directory. This bug has been fixed by ignoring all quota
3083 processing for deliveries into excluded directories.
3085 PH/10 Added the maildirfolder_create_regex option to appendfile.
3091 PH/01 The code for finding all the local interface addresses on a FreeBSD
3092 system running IPv6 was broken. This may well have applied to all BSD
3093 systems, as well as to others that have similar system calls. The broken
3094 code found IPv4 interfaces correctly, but gave incorrect values for the
3095 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3096 that it would not match correctly against @[] and not recognize the IPv6
3099 PH/02 The ipliteral router was not recognizing addresses of the form user@
3100 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3102 PH/03 Added disable_ipv6.
3104 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3105 decoded form, because it is most often used to construct To: headers
3106 lines in autoreplies, and the decoded form may well be syntactically
3107 invalid. However, $reply_address has leading white space removed, and all
3108 newlines turned into spaces so that the autoreply transport does not
3111 PH/05 If group was specified without a user on a router, and no group or user
3112 was specified on a transport, the group from the router was ignored.
3114 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3115 for visible compile-time settings that can be used to change these
3116 numbers, for those that want even more. Backwards compatibility with old
3117 spool files has been maintained. However, going back to a previous Exim
3118 release will lost any variables that are in spool files.
3120 PH/07 Two small changes when running in the test harness: increase delay when
3121 passing a TCP/IP connection to a new process, in case the original
3122 process has to generate a bounce, and remove special handling of
3123 127.0.0.2 (sic), which is no longer necessary.
3125 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3126 be the same on different OS.
3128 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3131 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3132 whether --show-vars was specified or not
3134 JJ/02 exipick: Added support for new ACL variable spool format introduced
3137 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3138 syntactically invalid From: or Reply-to: line, and a filter used this to
3139 generate an autoreply, and therefore failed to obtain an address for the
3140 autoreply, Exim could try to deliver to a non-existent relative file
3141 name, causing unrelated and misleading errors. What now happens is that
3142 it logs this as a hard delivery error, but does not attempt to create a
3145 PH/11 The exinext utility has a -C option for testing purposes, but although
3146 the given file was scanned by exinext itself; it wasn't being passed on
3147 when Exim was called.
3149 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3150 an end-of-file indication when reading a command response.
3152 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3153 compiled. In many other places in Exim, IPv6 addresses are always
3154 recognized, so I have changed this. It also means that IPv4 domain
3155 literals of the form [IPV4:n.n.n.n] are now always recognized.
3157 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3158 used if the router is not running as root, for example, when verifying at
3159 ACL time, or when using -bh. The debugging output from this situation was
3160 non-existent - all you got was a failure to exec. I have made two
3163 (a) Failures to set uid/gid, the current directory, or a process leader
3164 in a subprocess such as that created by queryprogram now generate
3165 suitable debugging output when -d is set.
3167 (b) The queryprogram router detects when it is not running as root,
3168 outputs suitable debugging information if -d is set, and then runs
3169 the subprocess without attempting to change uid/gid.
3171 PH/15 Minor change to Makefile for building test_host (undocumented testing
3174 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3175 additional section of a DNS packet that returns MX or SRV records.
3176 Instead, it always explicitly searches for A/AAAA records. This avoids
3177 major problems that occur when a DNS server includes only records of one
3178 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3179 fixed another bug: if SRV records were looked up and the corresponding
3180 address records were *not* found in the additional section, the port
3181 values from the SRV records were lost.
3183 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3184 using the correct key (the original address) when searching the retry
3185 rules in order to find which one to use for generating the retry hint.
3187 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3188 adding the default one. Similarly, if it contains a Reply-To: header, the
3189 errors_reply_to option, if set, is not used.
3191 PH/19 When calculating a retry time, Exim used to measure the "time since
3192 failure" by looking at the "first failed" field in the retry record. Now
3193 it does not use this if it is later than than the arrival time of the
3194 message. Instead it uses the arrival time. This makes for better
3195 behaviour in cases where some deliveries succeed, thus re-setting the
3196 "first failed" field. An example is a quota failure for a huge message
3197 when small messages continue to be delivered. Without this change, the
3198 "time since failure" will always be short, possible causing more frequent
3199 delivery attempts for the huge message than are intended.
3200 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3202 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3203 $1, $2, $3) because the numerical variables can be reset during some
3204 expansion items (e.g. "match"), thereby losing the authentication data.
3206 PH/21 Make -bV show the size of off_t variables so that the test suite can
3207 decide whether to run tests for quotas > 2G.
3209 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3210 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3211 filecount value is greater than 2G or if a quota value is greater than 2G
3212 on a system where the size of off_t is not greater than 4, a panic error
3215 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3216 never match. The debug and -bh output now contains an explicit error
3217 message indicating a malformed IPv4 address or mask.
3219 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3220 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3221 PH/23 above applies.
3223 PH/25 Do not write to syslog when running in the test harness. The only
3224 occasion when this arises is a failure to open the main or panic logs
3225 (for which there is an explicit test).
3227 PH/26 Added the /no_tell option to "control=freeze".
3229 PH/27 If a host name lookup failed very early in a connection, for example, if
3230 the IP address matched host_lookup and the reverse lookup yielded a name
3231 that did not have a forward lookup, an error message of the form "no IP
3232 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3233 could be logged. Now it outputs the IP address instead of "NULL".
3235 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3236 allows the sender and the authenticated sender to be set when
3237 submitting a message from within Exim. Since child_open_exim() is
3238 documented for local_scan(), the new function should be too.
3240 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3241 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3242 results in an empty string is now treated as unset.
3244 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3246 PH/31 Added sender_verify_fail logging option.
3248 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3249 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3250 tidied the source and removed it altogether.
3252 PH/33 When a queue run was abandoned because the load average was too high, a
3253 log line was always written; now it is written only if the queue_run log
3254 selector is set. In addition, the log line for abandonment now contains
3255 information about the queue run such as the pid. This is always present
3256 in "start" and "stop" lines but was omitted from the "abandon" line.
3258 PH/34 Omit spaces between a header name and the colon in the error message that
3259 is given when verify = headers_syntax fails (if there are lots of them,
3260 the message gets confusing).
3262 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3263 names, as there are now some PTR records that contain slashes. This check
3264 is only to protect against broken name servers that fall over on strange
3265 characters, so the fact that it applies to all lookups doesn't matter.
3267 PH/36 Now that the new test suite is complete, we can remove some of the
3268 special code in Exim that was needed for the old test suite. For example,
3269 sorting DNS records because real resolvers return them in an arbitrary
3270 order. The new test suite's fake resolver always returns records in the
3273 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3274 bounces) except when queue_only is set, to avoid logging races between
3275 the different processes.
3277 PH/38 Panic-die if .include specifies a non-absolute path.
3279 PH/39 A tweak to the "H" retry rule from its user.
3281 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3282 a label. They prevented compilation on older perls.
3284 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3285 a warning to be raised on newish perls.
3287 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3288 on queue. Changes to match documented behaviour of showing count of
3289 messages matching specified criteria.
3291 PH/40 Changed the default ident timeout from 30s to 5s.
3293 PH/41 Added support for the use of login_cap features, on those BSD systems
3294 that have them, for controlling the resources used by pipe deliveries.
3296 PH/42 The content-scanning code uses fopen() to create files in which to put
3297 message data. Previously it was not paying any attention to the mode of
3298 the files. Exim runs with umask(0) because the rest of the code creates
3299 files with open(), and sets the required mode explicitly. Thus, these
3300 files were ending up world-writeable. This was not a big issue, because,
3301 being within the spool directory, they were not world-accessible. I have
3302 created a function called modefopen, which takes an additional mode
3303 argument. It sets umask(777), creates the file, chmods it to the required
3304 mode, then resets the umask. All the relevant calls to fopen() in the
3305 content scanning code have been changed to use this function.
3307 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3308 to 24 hours. This avoids potential overflow problems when processing G
3309 and H retry rules. I suspect nobody ever tinkers with this value.
3311 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3313 PH/45 When the plaintext authenticator is running as a client, the server's
3314 challenges are checked to ensure they are valid base64 strings. By
3315 default, the authentication attempt is cancelled if an invalid string is
3316 received. Setting client_ignore_invalid_base64 true ignores these errors.
3317 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3318 they are received. Thus, the responses can be made to depend on the
3319 challenges. If an invalid string is ignored, an empty string is placed in
3322 PH/46 Messages that are created by the autoreply transport now contains a
3323 References: header, in accordance with RFCs 2822 and 3834.
3325 PH/47 Added authenticated_sender_force to the smtp transport.
3327 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3329 PH/49 Installed latest patch from the Sieve maintainer.
3331 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3332 was also set, the appendfile transport was unnecessarily scanning a
3333 directory of message files (e.g. for maildir delivery) to find the count
3334 of files (along with the size), even though it did not need this
3335 information. It now does the scan only if it needs to find either the
3336 size of the count of files.
3338 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3340 PH/52 Two bugs concerned with error handling when the smtp transport is
3343 (i) Exim was not creating retry information for temporary errors given
3344 for individual recipients after the DATA command when the smtp transport
3345 was used in LMTP mode. This meant that they could be retried too
3346 frequently, and not timed out correctly.
3348 (ii) Exim was setting the flag that allows error details to be returned
3349 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3350 recipients that were returned after the DATA command.
3352 PH/53 This is related to PH/52, but is more general: for any failing address,
3353 when detailed error information was permitted to be returned to the
3354 sender, but the error was temporary, then after the final timeout, only
3355 "retry timeout exceeded" was returned. Now it returns the full error as
3356 well as "retry timeout exceeded".
3358 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3359 do this, and (what is worse) MTAs that accept it.
3361 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3362 will now be deprecated.
3364 PH/56 New os.c-cygwin from the Cygwin maintainer.
3366 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3367 formats (previously only available in exim formats via -bpr, -bpru,
3368 and -bpra. Now also available in native and exiqgrep formats)
3370 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3371 with very large, slow to parse queues
3373 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3375 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3377 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3378 responses to authentication challenges, though it was showing the
3379 challenges; (ii) I've removed the CR characters from the debug output for
3382 PH/58 Allow for the insertion of a newline as well as a space when a string
3383 is turned into more than one encoded-word during RFC 2047 encoding. The
3384 Sieve code now uses this.
3386 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3387 data_4xx, lost_connection, tls_required.
3389 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3390 message was being sent as an SMTP response.
3392 PH/61 Add -l and -k options to exicyclog.
3394 PH/62 When verifying, if an address was redirected to one new address, so that
3395 verification continued, and the new address failed or deferred after
3396 having set something in $address_data, the value of $address_data was not
3397 passed back to the ACL. This was different to the case when no
3398 redirection occurred. The value is now passed back in both cases.
3400 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3401 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3402 use login_cap.h, so on its own it isn't the distinguishing feature. The
3403 new name refers directly to the setclassresources() function.
3405 PH/65 Added configuration files for NetBSD3.
3407 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3409 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3410 is preferred over IPv4.
3412 PH/68 The bounce_return_message and bounce_return_body options were not being
3413 honoured for bounces generated during the reception of non-SMTP messages.
3414 In particular, this applied to messages rejected by the ACL. This bug has
3415 been fixed. However, if bounce_return_message is true and bounce_return_
3416 body is false, the headers that are returned for a non-SMTP message
3417 include only those that have been read before the error was detected.
3418 (In the case of an ACL rejection, they have all been read.)
3420 PH/69 The HTML version of the specification is now built in a directory called
3421 spec_html instead of spec.html, because the latter looks like a path with
3422 a MIME-type, and this confuses some software.
3424 PH/70 Catch two compiler warnings in sieve.c.
3426 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3427 function verify_get_ident() calls ip_connect() to connect a socket, but
3428 if the "connect()" function timed out, ip_connect() used to close the
3429 socket. However, verify_get_ident() also closes the socket later, and in
3430 between Exim writes to the log, which may get opened at this point. When
3431 the socket was closed in ip_connect(), the log could get the same file
3432 descriptor number as the socket. This naturally causes chaos. The fix is
3433 not to close the socket in ip_connect(); the socket should be closed by
3434 the function that creates it. There was only one place in the code where
3435 this was missing, in the iplookup router, which I don't think anybody now
3436 uses, but I've fixed it anyway.
3438 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3439 well as to direct DNS lookups. Otherwise the handling of names in host
3440 lists is inconsistent and therefore confusing.
3446 PH/01 Two changes to the default runtime configuration:
3448 (1) Move the checks for relay_from_hosts and authenticated clients from
3449 after to before the (commented out) DNS black list checks.
3451 (2) Add control=submission to the relay_from_hosts and authenticated
3452 clients checks, on the grounds that messages accepted by these
3453 statements are most likely to be submissions.
3455 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3457 (1) Generate an error if the third argument for the ${prvs expansion is
3460 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3463 (3) Reset the variables that are obtained from the first argument of
3464 ${prvscheck and used in the second argument before leaving the code,
3465 because their memory is reclaimed, so using them afterwards may do
3468 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3469 one (it's much easier than Tom thought :-).
3471 (5) Because of (4), we can now allow for the use of $prvscheck_result
3472 inside the third argument.
3474 PH/03 For some reason, the default setting of PATH when running a command from
3475 a pipe transport was just "/usr/bin". I have changed it to
3478 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3479 anything to be listed in the output from -bV.
3481 PH/05 When a filter generated an autoreply, the entire To: header line was
3482 quoted in the delivery log line, like this:
3484 => >A.N.Other <ano@some.domain> <original@ddress> ...
3486 This has been changed so that it extracts the operative address. There
3487 may be more than one such address. If so, they are comma-separated, like
3490 => >ano@some.domain,ona@other.domain <original@ddress> ...
3492 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3493 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3494 not being looked up in the rDNS to get a host name, Exim was showing the
3495 IP address twice in Received: lines, even though the IP addresses were
3496 identical. For example:
3498 Received: from [1.2.3.4] (helo=[1.2.3.4])
3500 However, if the real host name was known, it was omitting the HELO data
3501 if it matched the actual IP address. This has been tidied up so that it
3502 doesn't show the same IP address twice.
3504 PH/07 When both +timestamp and +memory debugging was on, the value given by
3505 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3506 called by the debug printing, thereby overwriting the timestamp buffer.
3507 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3509 PH/08 When the original message was included in an autoreply transport, it
3510 always said "this is a copy of the message, including all the headers",
3511 even if body_only or headers_only was set. It now gives an appropriate
3514 PH/09 Applied a patch from the Sieve maintainer which:
3516 o fixes some comments
3517 o adds the (disabled) notify extension core
3518 o adds some debug output for the result of if/elsif tests
3519 o points to the current vacation draft in the documentation
3520 and documents the missing references header update
3524 o fixes a bug in processing the envelope test (when testing
3525 multiple envelope elements, the last element determined the
3528 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3529 Electronic Mail") by including:
3531 Auto-submitted: auto-generated
3533 in the messages that it generates (bounce messages and others, such as
3534 warnings). In the case of bounce messages for non-SMTP messages, there was
3535 also a typo: it was using "Auto_submitted" (underscore instead of
3536 hyphen). Since every message generated by Exim is necessarily in response
3537 to another message, thes have all been changed to:
3539 Auto-Submitted: auto-replied
3541 in accordance with these statements in the RFC:
3543 The auto-replied keyword:
3545 - SHOULD be used on messages sent in direct response to another
3546 message by an automatic process,
3548 - MUST NOT be used on manually-generated messages,
3550 - MAY be used on Delivery Status Notifications (DSNs) and Message
3551 Disposition Notifications (MDNs),
3553 - MUST NOT be used on messages generated by automatic or periodic
3554 processes, except for messages which are automatic responses to
3557 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3558 to the default Received: header definition.
3560 PH/12 Added log selector acl_warn_skipped (default on).
3562 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3563 variables because (a) they are in the wrong storage pool and (b) even if
3564 they were copied, it wouldn't work properly because of the caching.
3566 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3567 checking when decoding. Apparently there are clients that generate
3568 overlong encoded strings. Why am I not surprised?
3570 PH/15 If the first argument of "${if match_address" was not empty, but did not
3571 contain an "@" character, Exim crashed. Now it writes a panic log message
3572 and treats the condition as false.
3574 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3576 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3577 extension "envelope-auth". The code is finished and in agreement with
3578 other implementations, but there is no documentation so far and in fact,
3579 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3580 not changing the active code.
3582 Print executed "if" and "elsif" statements when debugging is used. This
3583 helps a great deal to understand what a filter does.
3585 Document more things not specified clearly in RFC3028. I had all this
3586 sorted out, when out of a sudden new issues came to my mind. Oops."
3588 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3591 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3592 canonical form (as documented). However, after a host name lookup from
3593 the IP address, check_host() was doing a simple string comparison with
3594 addresses acquired from the DNS when checking that the found name did
3595 have the original IP as one of its addresses. Since any found IPv6
3596 addresses are likely to be in abbreviated form, the comparison could
3597 fail. Luckily, there already exists a function for doing the comparison
3598 by converting both addresses to binary, so now that is used instead of
3599 the text comparison.
3601 PH/20 There was another similar case to PH/19, when a complete host name was
3602 given in a host list; looking up its IP address could give an abbreviated
3603 form, whereas the current host's name might or might not be abbreviated.
3604 The same fix has been applied.
3610 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3611 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3614 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3615 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3617 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3619 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3620 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3621 Also, for "Auto-Submitted", treat anything other than "no" as
3622 non-personal, in accordance with RFC 3834. (Previously it treated
3623 anything starting "auto-" as non-personal.)
3625 TF/01 The control=submission/name=... option had a problem with syntax
3626 errors if the name included a slash character. The /name= option
3627 now slurps the rest of the string, so it can include any characters
3628 but it must come last in the list of options (after /sender_retain
3631 PH/05 Some modifications to the interface to the fake nameserver for the new
3639 TK/01 Added the "success_on_redirect" address verification option. See
3640 NewStuff for rationale and an example.
3642 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3644 PH/02 Patch to exigrep to allow it to work on syslog lines.
3646 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3647 fread() to skip over the body file's header line, because in Cygwin the
3648 header line is locked and is inaccessible.
3650 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3651 co-exist for some time) to make it clear that it is the Exim ID that is
3652 referenced, not the Message-ID: header line.
3654 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3655 string_format() function, because snprintf() does not exist on all
3658 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3659 ${stat: expansion item.
3661 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3662 protocol synchronization error", to keep the pedants happy.
3664 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3665 well as for IRIX systems, when gcc is being used. See the host.c source
3668 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3670 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3673 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3674 transport and to the smtp transport in LMTP mode.
3676 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3678 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3679 run a filter in a subprocess. This could lead to confusion in subsequent
3680 lookups in the parent process. There should also be a search_tidyup() at
3681 the end of the subprocess.
3683 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3684 only if the host matched helo_try_verify_hosts, which caused the
3685 verification to occur when the EHLO/HELO command was issued. The ACL just
3686 tested the remembered result. Now, if a previous verification attempt has
3687 not happened, "verify = helo" does it there and then.
3689 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3691 TK/03 Fix log output including CR from clamd.
3693 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3694 exist provoked a memory error which could cause a segfault.
3696 PH/15 Installed PCRE 6.2
3698 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3700 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3701 of the problem. Specifically, suggested +O2 rather than +O1 for the
3704 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3706 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3707 record, the port was not correctly logged when the outgoing_port log
3708 selector was set (it logged the transort's default port).
3710 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3711 fallback_hosts, and "hosts" in the smtp transport.
3713 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3714 host errors such as "Connection refused".
3716 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3717 authentication with radiusclient 0.4.9:
3719 - Error returned from rc_read_config was caught wrongly
3720 - Username/password not passed on to radius server due to wrong length.
3722 The presumption is that some radiusclient API changes for 4.51/PH/17
3723 were not taken care of correctly. The code is still untested by me (my
3724 Linux distribution still has 0.3.2 of radiusclient), but it was
3725 contributed by a Radius user.
3727 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3728 expanding the "port" option of the smtp transport.
3730 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3731 while reading a MIME header. Thanks to Tom Hughes for a patch.
3733 PH/24 Include config.h inside local_scan.h so that configuration settings are
3736 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3737 that in an ACL for RCPT (for example), you can examine exactly what was
3740 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3741 commands, but it was not correctly comparing the address with the actual
3742 client host address. Thus, it would show the EHLO address in Received:
3743 header lines when this was not necessary.
3745 PH/27 Added the % operator to ${eval:}.
3747 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3748 it should be ignoring failures (because with -C, for example, it has lost
3749 privilege). It wasn't ignoring creation failures other than "already
3752 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3755 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3756 because an input line was too long, either on its own, or by virtue of
3757 too many continuations, the temporary file was not being removed, and the
3758 return code was incorrect.
3760 PH/31 Missing "BOOL" in function definition in filtertest.c.
3762 PH/32 Applied Sieve patches from the maintainer.
3764 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3766 PH/33 Added "verify = not_blind".
3768 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3769 Local/Makefile (with some defaults set). These are used in built scripts
3770 such as exicyclog, but they have never been used in the exim_install
3771 script (though there are many overriding facilities there). I have
3772 arranged that the exim_install script now takes note of these two
3775 PH/35 Installed configuration files for Dragonfly.
3777 PH/36 When a locally submitted message by a trusted user did not contain a
3778 From: header, and the sender address was obtained from -f or from an SMTP
3779 MAIL command, and the trusted user did not use -F to supply a sender
3780 name, $originator_name was incorrectly used when constructing a From:
3781 header. Furthermore, $originator_name was used for submission mode
3782 messages from external hosts without From: headers in a similar way,
3783 which is clearly wrong.
3785 PH/37 Added control=suppress_local_fixups.
3787 PH/38 When log_selector = +received_sender was set, and the addition of the
3788 sender made the log line's construction buffer exactly full, or one byte
3789 less than full, an overflow happened when the terminating "\n" was
3792 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3793 when the result of a list match is failure because a DNS lookup failed.
3795 PH/40 RM_COMMAND is now used in the building process.
3797 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3798 the "build-* directories that it finds.
3800 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3801 address in a domain literal was a prefix of an interface address.
3803 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3804 when verifying a sender address, unless rewrite_headers is false.
3806 PH/44 Wrote a long comment about why errors_to addresses are verified as
3807 recipients, not senders.
3809 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3810 the ratelimit ACL was added.
3812 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3814 PH/46 Added extra information about PostgreSQL errors to the error string.
3816 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3817 avoiding the need to install special zones in a real server. This is
3818 backwards compatible; if it can't find the fake resolver, it drops back.
3819 Thus, both old and new test suites can be run.
3821 TF/02 Added util/ratelimit.pl
3823 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3824 clock is set back in time.
3826 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3827 Candler <B.Candler@pobox.com>.
3829 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3830 OK for addresses that are the result of redirections.
3832 PH/48 A number of further additions for the benefit of the new test suite,
3833 including a fake gethostbyname() that interfaces to the fake DNS resolver
3836 TF/06 The fix for widen_domains has also been applied to qualify_single and
3837 search_parents which are the other dnslookup options that can cause
3840 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3843 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3845 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3846 service" if it encountered an MX record with an empty target hostname.
3847 The message is now "an MX or SRV record indicated no SMTP service".
3849 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3850 if the DNS of the sending site is misconfigured. This is quite a
3851 common situation. This change restores the behaviour of treating a
3852 helo verification defer as a failure.
3854 PH/51 If self=fail was set on a router, the bounce message did not include the
3855 actual error message.
3861 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3863 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3864 command itself ended in error, the underlying message about the transport
3865 filter timeout was being overwritten with the pipe command error. Now the
3866 underlying error message should be appended to the second error message.
3868 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3870 PH/02 Reduce the amount of output that "make" produces by default. Full output
3871 can still be requested.
3873 PH/03 The warning log line about a condition test deferring for a "warn" verb
3874 was being output only once per connection, rather than after each
3875 occurrence (because it was using the same function as for successful
3876 "warn" verbs). This seems wrong, so I have changed it.
3878 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3879 it should not have, which might have caused a crash in the right
3880 circumstances, but probably never did.
3882 PH/04 Installed a modified version of Tony Finch's patch to make submission
3883 mode fix the return path as well as the Sender: header line, and to
3884 add a /name= option so that you can make the user's friendly name appear
3887 TF/03 Added the control = fakedefer ACL modifier.
3889 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3890 Mark Lowes for thorough testing.
3892 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3894 TK/03 Merged latest SRS patch from Miles Wilton.
3896 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3897 with the definition in sysexits.h (which is #included earlier).
3898 Fortunately, Exim does not actually use EX_OK. The code used to try to
3899 preserve the sysexits.h value, by assuming that macro definitions were
3900 scanned for macro replacements. I have been disabused of this notion,
3901 so now the code just undefines EX_OK before #including unistd.h.
3903 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3904 in the smtp transport. When a block could not be written in a single
3905 write() function, the timeout was being re-applied to each part-write.
3906 This seems wrong - if the receiver was accepting one byte at a time it
3907 would take for ever. The timeout is now adjusted when this happens. It
3908 doesn't have to be particularly precise.
3910 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3911 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3913 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3914 without a domain if the check to <postmaster@domain> fails.
3916 SC/01 Eximstats: added -xls and the ability to specify output files
3917 (patch written by Frank Heydlauf).
3919 SC/02 Eximstats: use FileHandles for outputting results.
3921 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3923 SC/04 Eximstats: fixed display of large numbers with -nvr option
3925 SC/05 Eximstats: fixed merging of reports with empty tables.
3927 SC/06 Eximstats: added the -include_original_destination flag
3929 SC/07 Eximstats: removed tabs and trailing whitespace.
3931 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3933 TK/06 MBOX spool code: Add real "From " MBOX separator line
3934 so the .eml file is really in mbox format (even though
3935 most programs do not really care). Patch from Alex Miller.
3937 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3938 The latter is generated from $received_to and is only set if the
3939 message has one envelope recipient. SA can use these headers,
3940 obviously out-of-the-box. Patch from Alex Miller.
3942 PH/08 The ${def test on a variable was returning false if the variable's
3943 value was "0", contrary to what the specification has always said!
3944 The result should be true unless the variable is empty.
3946 PH/09 The syntax error of a character other than { following "${if
3947 def:variable_name" (after optional whitespace) was not being diagnosed.
3948 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3949 accidental colon was present, for example, could give incorrect results.
3951 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3952 result is used (not including appendfile, where other changes are about
3955 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3956 This involved changing a lot of size variables from int to off_t. It
3957 should work with maildirs and everything.
3959 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3960 spamd dying while we are connected to it.
3962 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3965 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3966 function for BDB 4.3.
3968 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3970 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3971 the sources for PCRE have been re-arranged and more files are now
3974 PH/15 The code I had for printing potentially long long variables in PH/11
3975 above was not the best (it lost precision). The length of off_t variables
3976 is now inspected at build time, and an appropriate printing format (%ld
3977 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3978 to be "long long int" or "long int". This is needed for the internal
3979 formatting function string_vformat().
3981 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3982 the configuration file to be ":syslog", then the script "guesses" where
3983 the logs files are, rather than using the compiled in default. In our
3984 case the guess is not the same as the compiled default, so the script
3985 suddenly stopped working when I started to use syslog. The patch checks
3986 to see if log_file_path is "". If so, it attempts to read it from exim
3987 with no configuration file to get the compiled in version, before it
3988 falls back to the previous guessing code."
3990 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3991 implementing BATV in an Exim configuration. See NewStuff for the gory
3994 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3995 Makefile that are specific to HP-UX.
3997 PH/18 If the "use_postmaster" option was set for a recipient callout together
3998 with the "random" option, the postmaster address was used as the MAIL
3999 FROM address for the random test, but not for the subsequent recipient
4000 test. It is now used for both.
4002 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4003 patch removes a few documentation additions to RFC 3028, because the
4004 latest draft now contains them. It adds the new en;ascii-case comparator
4005 and a new error check for 8bit text in MIME parts. Comparator and
4006 require names are now matched exactly. I enabled the subaddress
4007 extension, but it is not well tested yet (read: it works for me)."
4009 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4010 rework some of the code of TK/09 above to avoid the hardwired use of
4011 "%lld" and "long long". Replaced the call to snprintf() with a call to
4014 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4015 records point to non-existent hosts", "retry timeout exceeded", and
4016 "retry time not reached for any host after a long failure period".
4018 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4019 experimental DomainKeys support:
4021 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4022 (2) On an error such as an illegally used "control", the wrong name for
4023 the control was given.
4025 These problems did NOT occur unless DomainKeys support was compiled.
4027 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4029 PH/24 Added ${if match_ip condition.
4031 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4032 databases so that it will be absolutely obvious if a crash occurs in the
4033 DB library. This is a regular occurrence (often caused by mis-matched
4036 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4037 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4038 picked up on a user's system that detects such things. There doesn't seem
4039 to be a gcc warning option for this - only an attribute that has to be
4040 put on the function's prototype. It seems that in Fedora Core 4 they have
4041 set this on a number of new functions. No doubt there will be more in due
4044 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4045 specify a transport. However, if an address that was verified by such a
4046 router was the subject of a callout, Exim crashed because it tried to
4047 read the rcpt_include_affixes from the non-existent transport. Now it
4048 just assumes that the setting of that option is false. This bug was
4049 introduced by 4.51/PH/31.
4051 PH/28 Changed -d+all to exclude +memory, because that information is very
4052 rarely of interest, but it makes the output a lot bigger. People tend to
4053 do -d+all out of habit.
4055 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4056 code in os-type was giving problems when libc.so lives in lib64, like on
4059 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4060 aren't the modern standard, and it seems that some systems' include files
4061 don't always have them. Exim was already checking for some of the newer
4062 ones like T_AAAA, and defining it itself. I've added checks for all the
4063 record types that Exim uses.
4065 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4066 not automatically generating a new one, as it is supposed to. This
4067 prevented TLS from working. If the file did exist, but contained invalid
4068 data, a new version was generated, as expected. It was only the case of a
4069 non-existent file that was broken.
4071 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4072 with a change in libdomainkeys > 0.64.
4074 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4075 from DNS. If the selector record carries the flag, it now has
4076 precedence over the domain-wide flag.
4078 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4080 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4081 the use of a port name that isn't defined in /etc/services) occurred, the
4082 message was deferred as in a normal delivery, and thus remained on the
4083 spool, instead of being failed because of the mua_wrapper setting. This
4084 is now fixed, and I tidied up some of the mua_wrapper messages at the
4087 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4088 the messages in a hash of arrays rather than using individual hashes.
4089 This is a bit cleaner and results in dramatic memory savings, albeit
4090 at a slight CPU cost.
4092 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4093 as requested by Marc Sherman.
4095 SC/10 Eximstats: added histograms for user specified patterns as requested
4098 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4100 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4101 fopen() in the content-scanning modules that did not already have it.
4107 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4108 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4110 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4112 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4114 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4115 to test Sieve filters that use "vacation".
4117 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4118 that changes the way the GnuTLS parameters are stored in the cache file.
4119 The new format can be generated externally. For backward compatibility,
4120 if the data in the cache doesn't make sense, Exim assumes it has read an
4121 old-format file, and it generates new data and writes a new file. This
4122 means that you can't go back to an older release without removing the
4125 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4126 work if there are any delivery delays because "one_time" forces the
4127 parent to be marked "delivered", so its unseen clone is never tried
4128 again. For this reason, Exim now forbids the simultaneous setting of
4131 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4132 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4133 entry if you want to know the details. The fix, however, neglected to
4134 consider the case when local delivery batching is involved. The test for
4135 "previously delivered" was not happening when checking to see if an
4136 address could be batched with a previous (undelivered) one; under
4137 certain circumstances this could lead to multiple deliveries to the same
4140 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4141 in its include files, and this causes problems building Exim.
4143 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4144 header_syntax) but Exim was just ignoring anything given after a slash.
4145 In particular, this caused confusion with an attempt to use "verify =
4146 reverse_host_lookup/defer_ok". An error is now given when options are
4147 supplied for verify items that do not have them. (Maybe reverse_host_
4148 lookup should have a defer_ok option, but that's a different point.)
4150 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4151 defined by RFC 821) to 2048, because there were problems with some AUTH
4152 commands, and RFC 1869 says the size should be increased for extended
4153 SMTP commands that take arguments.
4155 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4158 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4159 "unknown" error; now it says that the functionality isn't in the binary.
4161 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4162 an address' error message when a string expansion fails (syntax or
4163 whatever). Otherwise the password may appear in the log. Following change
4164 PH/42 below, there is no longer a chance of it appearing in a bounce
4167 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4169 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4170 of its addresses was ever tried. (Bugzilla bug #2.)
4172 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4173 the result incorrectly in the debug output. (It correctly added a newline
4174 to what was transported.)
4176 TF/01 Added $received_time.
4178 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4179 commented out examples of how to interface to a virus scanner and to
4180 SpamAssassin. Also added commented examples of av_scanner and
4181 spamd_address settings.
4183 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4184 and controls are allowed in which ACLs. There were a couple of minor
4185 errors. Some of the entries in the conditions table (which is a table of
4186 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4187 negation of where the condition IS allowed.
4189 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4191 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4192 header file does not have a version number, so I've had to invent a new
4193 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4194 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4195 radiusclient), but it was contributed by a Radius user.
4197 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4198 files or directories, for OpenSSL.
4200 PH/19 When an Exim process that is running as root has to create an Exim log
4201 file, it does so in a subprocess that runs as exim:exim so as to get the
4202 ownership right at creation (otherwise, other Exim processes might see
4203 the file with the wrong ownership). There was no test for failure of this
4204 fork() call, which would lead to the process getting stuck as it waited
4205 for a non-existent subprocess. Forks do occasionally fail when resources
4206 run out. I reviewed all the other calls to fork(); they all seem to check
4209 PH/20 When checking for unexpected SMTP input at connect time (before writing
4210 the banner), Exim was not dealing correctly with a non-positive return
4211 from the read() function. If the client had disconnected by this time,
4212 the result was a log entry for a synchronization error with an empty
4213 string after "input=" when read() returned zero. If read() returned -1
4214 (an event I could not check), uninitialized data bytes were printed.
4215 There were reports of junk text (parts of files, etc) appearing after
4218 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4220 PH/22 Added support for macro redefinition, and (re)definition in between
4221 driver and ACL definitions.
4223 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4224 forgetting to use the resulting value; it was using the unexpanded value.
4226 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4227 hadn't been configured. The fix is from Juergen Kreileder, who
4228 understands it better than I do:
4230 "Here's what I see happening with three configured cyrus_sasl
4231 authenticators configured (plain, login, cram-md5):
4233 On startup auth_cyrus_sasl_init() gets called for each of these.
4234 This means three calls to sasl_listmech() without a specified mech_list.
4235 => SASL tests which mechs of all available mechs actually work
4236 => three warnings about OTP not working
4237 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4239 With the patch, sasl_listmech() also gets called three times. But now
4240 SASL's mech_list option is set to the server_mech specified in the the
4241 authenticator. Or in other words, the answer from sasl_listmech()
4242 gets limited to just the mech you're testing for (which is different
4244 => the return list contains just 'plain' or 'login', 'cram-md5' or
4245 nothing depending on the value of ob->server_mech.
4247 I've just tested the patch: Authentication still works fine,
4248 unavailable mechs specified in the exim configuration are still
4249 caught, and the auth.log warnings about OTP are gone."
4251 PH/25 When debugging is enabled, the contents of the command line are added
4252 to the debugging output, even when log_selector=+arguments is not
4255 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4256 answer is "GNU", and only if the return is "GNU/something" is the answer
4259 PH/27 $acl_verify_message is now set immediately after the failure of a
4260 verification in an ACL, and so is available in subsequent modifiers. In
4261 particular, the message can be preserved by coding like this:
4263 warn !verify = sender
4264 set acl_m0 = $acl_verify_message
4266 Previously, $acl_verify_message was set only while expanding "message"
4267 and "log_message" when a very denied access.
4269 PH/28 Modified OS/os.c-Linux with
4271 -#ifndef OS_LOAD_AVERAGE
4272 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4274 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4275 nomenclature these days.)
4277 PH/29 Installed patch from the Sieve maintainer that adds the options
4278 sieve_useraddress and sieve_subaddress to the redirect router.
4280 PH/30 In these circumstances:
4281 . Two addresses routed to the same list of hosts;
4282 . First host does not offer TLS;
4283 . First host accepts first address;
4284 . First host gives temporary error to second address;
4285 . Second host offers TLS and a TLS session is established;
4286 . Second host accepts second address.
4287 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4288 and peerdn, if requested) that were in fact used only for the second
4291 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4292 attention to any local part prefix or suffix that was matched by the
4293 router that accepted the address. It now behaves in the same way as it
4294 does for delivery: the affixes are removed from the local part unless
4295 rcpt_include_affixes is set on the transport.
4297 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4298 timeout during the DATA phase of an incoming message.
4300 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4301 applied a patch sanctioned by the Sieve maintainer.
4303 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4304 the uid or gid is negative. A case of a negative gid caused this to be
4305 noticed. The fix allows for either to be negative.
4307 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4308 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4311 PH/36 The change PH/12 above was broken. Fixed it.
4313 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4314 the grounds that routing the same address twice would always produce the
4315 same answer. This might have been true once, but it is certainly no
4316 longer true now. Routing a child address may depend on the previous
4317 routing that produced that child. Some complicated redirection strategies
4318 went wrong when messages had multiple recipients, and made Exim's
4319 behaviour dependent on the order in which the addresses were given.
4321 I have moved the duplicate checking until after the routing is complete.
4322 Exim scans the addresses that are assigned to local and remote
4323 transports, and removes any duplicates. This means that more work will be
4324 done, as duplicates will always all be routed, but duplicates are
4325 presumably rare, so I don't expect this is of any significance.
4327 For deliveries to pipes, files, and autoreplies, the duplicate checking
4328 still happens during the routing process, since they are not going to be
4331 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4332 It corrects a timeout issue with spamd. This is Ian's comment: "The
4333 background is that sometimes spamd either never reads data from a
4334 connection it has accepted, or it never writes response data. The exiscan
4335 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4336 blindly assumes that writes won't block so it may never time out."
4338 PH/39 Allow G after quota size as well as K and M.
4340 PH/40 The value set for $authenticated_id in an authenticator may not contain
4341 binary zeroes or newlines because the value is written to log lines and
4342 to spool files. There was no check on this. Now the value is run through
4343 the string_printing() function so that such characters are converted to
4344 printable escape sequences.
4346 PH/41 $message_linecount is a new variable that contains the total number of
4347 lines in the message. Compare $body_linecount, which is the count for the
4350 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4351 bounce and delay warning messages, except in certain special cases, which
4354 (a) An SMTP error message from a remote host;
4355 (b) A message specified in a :fail: redirection;
4356 (c) A message specified in a "fail" command in a system filter;
4357 (d) A message specified in a FAIL return from the queryprogram router;
4358 (e) A message specified by the cannot_route_message router option.
4360 In these cases only, Exim does include the error details in bounce and
4361 warning messages. There are also a few cases where bland messages such
4362 as "unrouteable address" or "local delivery error" are given.
4364 PH/43 $value is now also set for the "else" part of a ${run expansion.
4366 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4367 being worked on, but at least Exim now implements the latest version to
4370 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4371 process to complete was treated as a delivery failure, a timeout while
4372 writing the message to the pipe was logged, but erroneously treated as a
4373 successful delivery. Such timeouts include transport filter timeouts. For
4374 consistency with the overall process timeout, these timeouts are now
4375 treated as errors, giving rise to delivery failures by default. However,
4376 there is now a new Boolean option for the pipe transport called
4377 timeout_defer, which, if set TRUE, converts the failures into defers for
4378 both kinds of timeout. A transport filter timeout is now identified in
4381 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4382 systems where "make" and "gmake" are different, calling "gmake" at top
4383 level broke things. I've arranged for the value of $(MAKE) to be passed
4384 from the Makefile to this script so that it can call the same version of
4388 A note about Exim versions 4.44 and 4.50
4389 ----------------------------------------
4391 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4392 changes of various kinds. As a consequence, a big documentation update was
4393 needed. This delayed the release for rather longer than seemed good, especially
4394 in the light of a couple of (minor) security issues. Therefore, the changes
4395 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4396 release. So 4.44 and 4.50 are in effect two different branches that both start
4399 I have left the 4.50 change log unchanged; it contains all the changes since
4400 4.43. The change log for 4.44 is below; many of its items are identical to
4401 those for 4.50. This seems to be the most sensible way to preserve the
4402 historical information.
4408 1. Minor wording change to the doc/README.SIEVE file.
4410 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4411 computation of the current number of files was incorrect.
4413 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4414 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4417 4. Give more explanation in the error message when the command for a transport
4418 filter fails to execute.
4420 5. There are several places where Exim runs a non-Exim command in a
4421 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4422 was being done only for the command run by the queryprogram router. It is
4423 now done for all such subprocesses. The other cases are: ${run, transport
4424 filters, and the commands run by the lmtp and pipe transports.
4426 6. Added CONFIGURE_GROUP build-time option.
4428 7. Some older OS have a limit of 256 on the maximum number of file
4429 descriptors. Exim was using setrlimit() to set 1000 as a large value
4430 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4431 systems. I've change it so that if it can't get 1000, it tries for 256.
4433 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4434 was an oversight, and furthermore, ever since the addition of extra
4435 controls (e.g. 4.43/32), the checks on when to allow different forms of
4436 "control" were broken. There should now be diagnostics for all cases when a
4437 control that does not make sense is encountered.
4439 9. Added the /retain_sender option to "control=submission".
4441 10. $recipients is now available in the predata ACL (oversight).
4443 11. Tidy the search cache before the fork to do a delivery from a message
4444 received from the command line. Otherwise the child will trigger a lookup
4445 failure and thereby defer the delivery if it tries to use (for example) a
4446 cached ldap connection that the parent has called unbind on.
4448 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4449 of $address_data from the recipient verification was clobbered by the
4450 sender verification.
4452 13. The value of address_data from a sender verification is now available in
4453 $sender_address_data in subsequent conditions in the ACL statement.
4455 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4457 15. Added a new option "connect=<time>" to callout options, to set a different
4460 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4461 was its contents. (It was OK if the option was not defined at all.)
4463 17. A "Completed" log line is now written for messages that are removed from
4464 the spool by the -Mrm option.
4466 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4467 information about exactly what failed.
4469 19. Added -dd to debug only the daemon process.
4471 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4472 handles timeouts, both on the server side and network timeouts. Renamed the
4473 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4475 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4476 It is now set to "smtps".
4478 22. $host_address is now set to the target address during the checking of
4479 ignore_target_hosts.
4481 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4482 being passed; this would have caused $sender_host_name to have been used if
4483 matching the list had actually called for a host name (not very likely,
4484 since this list is usually IP addresses). A host name is now passed as
4487 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4488 code that specifies a non-restarting handler (typically sigaction() in
4489 modern systems) in an attempt to fix a rare and obscure crash bug.
4491 25. Narrowed the window for a race in the daemon that could cause it to ignore
4492 SIGCHLD signals. This is not a major problem, because they are used only to
4493 wake it up if nothing else does.
4495 26. A malformed maildirsize file could cause Exim to calculate negative values
4496 for the mailbox size or file count. Odd effects could occur as a result.
4497 The maildirsize information is now recalculated if the size or filecount
4500 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4501 support for a long time. Removed HAVE_SYS_VFS_H.
4503 28. Installed the latest version of exipick from John Jetmore.
4505 29. In an address list, if the pattern was not a regular expression, an empty
4506 subject address (from a bounce message) matched only if the pattern was an
4507 empty string. Non-empty patterns were not even tested. This was the wrong
4508 because it is perfectly reasonable to use an empty address as part of a
4509 database query. An empty address is now tested by patterns that are
4510 lookups. However, all the other forms of pattern expect the subject to
4511 contain a local part and a domain, and therefore, for them, an empty
4512 address still always fails if the pattern is not itself empty.
4514 30. Exim went into a mad DNS loop when attempting to do a callout where the
4515 host was specified on an smtp transport, and looking it up yielded more
4516 than one IP address.
4518 31. Re-factored the code for checking spool and log partition space into a
4519 function that finds that data and another that does the check. The former
4520 is then used to implement four new variables: $spool_space, $log_space,
4521 $spool_inodes, and $log_inodes.
4523 32. The RFC2047 encoding function was originally intended for short strings
4524 such as real names; it was not keeping to the 75-character limit for
4525 encoded words that the RFC imposes. It now respects the limit, and
4526 generates multiple encoded words if necessary. To be on the safe side, I
4527 have increased the buffer size for the ${rfc2047: expansion operator from
4530 33. It is now permitted to omit both strings after an "if" condition; if the
4531 condition is true, the result is "true". As before, when the second string
4532 is omitted, a false condition yields an empty string. This makes it less
4533 cumbersome to write custom ACL and router conditions.
4535 34. Failure to deliver a bounce message always caused it to be frozen, even if
4536 there was an errors_to setting on the router. The errors_to setting is now
4539 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4540 canonical form (fully expanded) before being placed in
4541 $sender_host_address.
4543 36. The table in the code that translates DNS record types into text (T_A to
4544 "A" for instance) was missing entries for NS and CNAME. It is just possible
4545 that this could have caused confusion if both these types were looked up
4546 for the same domain, because the text type is used as part of Exim's
4547 per-process caching. But the chance of anyone hitting this buglet seems
4550 37. The dnsdb lookup has been extended in a number of ways.
4552 (1) There is a new type, "zns", which walks up the domain tree until it
4553 finds some nameserver records. It should be used with care.
4555 (2) There is a new type, "mxh", which is like "mx" except that it returns
4556 just the host names, not the priorities.
4558 (3) It is now possible to give a list of domains (or IP addresses) to be
4559 looked up. The behaviour when one of the lookups defers can be
4560 controlled by a keyword.
4562 (4) It is now possible to specify the separator character for use when
4563 multiple records are returned.
4565 38. The dnslists ACL condition has been extended: it is now possible to supply
4566 a list of IP addresses and/or domains to be looked up in a particular DNS
4569 39. Added log_selector=+queue_time_overall.
4571 40. When running the queue in the test harness, wait just a tad after forking a
4572 delivery process, to get repeatability of debugging output.
4574 41. Include certificate and key file names in error message when GnuTLS fails
4575 to set them up, because the GnuTLS error message doesn't include the name
4576 of the failing file when there is a problem reading it.
4578 42. Allow both -bf and -bF in the same test run.
4580 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4582 44. The "Exiscan patch" is now merged into the mainline Exim source.
4584 45. Sometimes the final signoff response after QUIT could fail to get
4585 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4586 < 0 before doing a fflush(). This bug looks as though it goes back to the
4587 introduction of TLS in release 3.20, but "sometimes" must have been rare
4588 because the tests only now provoked it.
4590 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4591 (this can affect the format of dates).
4593 47. exim_tidydb, when checking for the continued existence of a message for
4594 which it has found a message-specific retry record, was not finding
4595 messages that were in split spool directories. Consequently, it was
4596 deleting retry records that should have stayed in existence.
4598 48. Steve fixed some bugs in eximstats.
4600 49. The SPA authentication driver was not abandoning authentication and moving
4601 on to the next authenticator when an expansion was forced to fail,
4602 contradicting the general specification for all authenticators. Instead it
4603 was generating a temporary error. It now behaves as specified.
4605 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4606 (the order specifies the preference for clients). The order is now AES256,
4607 AES128, 3DES, ARCFOUR128.
4609 51. Small patch to Sieve code - explicitly set From: when generating an
4612 52. Exim crashed if a remote delivery caused a very long error message to be
4613 recorded - for instance if somebody sent an entire SpamAssassin report back
4614 as a large number of 550 error lines. This bug was coincidentally fixed by
4615 increasing the size of one of Exim's internal buffers (big_buffer) that
4616 happened as part of the Exiscan merge. However, to be on the safe side, I
4617 have made the code more robust (and fixed the comments that describe what
4620 53. Now that there can be additional text after "Completed" in log lines (if
4621 the queue_time_overall log selector is set), a one-byte patch to exigrep
4622 was needed to allow it to recognize "Completed" as not the last thing in
4625 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4626 patch that reportedly fixes this has been added. I am not expert enough to
4627 create a test for it. This is what the patch creator wrote:
4629 "I found a little strange behaviour of ldap code when working with
4630 Windows 2003 AD Domain, where users was placed in more than one
4631 Organization Units. When I tried to give exim partial DN, the exit code
4632 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4633 But simultaneously result of request was absolutely normal ldap result,
4634 so I produce this patch..."
4636 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4637 REFERENCE, so I have modified the code to exclude the patch when that macro
4640 55. Some experimental protocols are using DNS PTR records for new purposes. The
4641 keys for these records are domain names, not reversed IP addresses. The
4642 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4643 leaves it alone. Component reversal etc. now happens only for IP addresses.
4646 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4648 57. Double the size of the debug message buffer (to 2048) so that more of very
4649 long debug lines gets shown.
4651 58. The exicyclog utility now does better if the number of log files to keep
4652 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4654 59. Two changes related to the smtp_active_hostname option:
4656 (1) $smtp_active_hostname is now available as a variable.
4657 (2) The default for smtp_banner uses $smtp_active_hostname instead
4658 of $primary_hostname.
4660 60. The host_aton() function is supposed to be passed a string that is known
4661 to be a valid IP address. However, in the case of IPv6 addresses, it was
4662 not checking this. This is a hostage to fortune. Exim now panics and dies
4663 if the condition is not met. A case was found where this could be provoked
4664 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4665 components; fortuitously, this particular loophole had already been fixed
4666 by change 4.50/55 above.
4668 If there are any other similar loopholes, the new check in host_aton()
4669 itself should stop them being exploited. The report I received stated that
4670 data on the command line could provoke the exploit when Exim was running as
4671 exim, but did not say which command line option was involved. All I could
4672 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4673 running as the user.
4676 61. There was a buffer overflow vulnerability in the SPA authentication code
4677 (which came originally from the Samba project). I have added a test to the
4678 spa_base64_to_bits() function which I hope fixes it.
4681 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4682 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4684 63. The daemon start-up calls getloadavg() while still root for those OS that
4685 need the first call to be done as root, but it missed one case: when
4686 deliver_queue_load_max is set with deliver_drop_privilege. This is
4687 necessary for the benefit of the queue runner, because there is no re-exec
4688 when deliver_drop_privilege is set.
4690 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4691 This has been fixed.
4693 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4694 was in use, was not putting the data itself into the right store pool;
4695 consequently, it could be overwritten for a subsequent message in the same
4696 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4699 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4701 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4702 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4703 the function were treating the return as a boolean value, which happened to
4704 work because 0=false and not-0=true, but is not correct code.
4706 68. The host_aton() function was not handling scoped IPv6 addresses (those
4707 with, for example, "%eth0" on the end) correctly.
4709 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4710 negated items (that is, ~something) in unsigned ints. Some compilers
4711 apparently mutter when there is no cast.
4713 70. If an address verification called from an ACL failed, and did not produce a
4714 user-specific message (i.e. there was only a "system" message), nothing was
4715 put in $acl_verify_message. In this situation, it now puts the system
4718 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4719 session; change 4.31/43 added the unwanted input to the log line - except
4720 that it did not do this in the start of session case. It now does.
4722 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4723 This is wrong and can cause the other end to generate a synchronization
4724 error if it is another Exim or anything else that does the synchronization
4725 check. A QUIT command is no longer sent after a timeout.
4727 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4728 during host lookups.
4730 74. The defer_ok option of callout verification was not working if it was used
4731 when verifying addresses in header lines, that is, for this case:
4733 verify = header_sender/callout=defer_ok
4735 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4736 those file descriptors could be used for SMTP connections. If anything
4737 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4738 could be sent to the SMTP client, causing chaos. The daemon now opens
4739 stdin, stdout, and stderr to /dev/null when it puts itself into the
4742 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4743 log by default. The user can override this with suitable Perl magic.
4745 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4746 the log message when discard triggers, was not working for the DATA ACL or
4747 for the non-SMTP ACL.
4749 78. Error message wording change in sieve.c.
4751 79. If smtp_accept_max_per_host was set, the number of connections could be
4752 restricted to fewer than expected, because the daemon was trying to set up
4753 a new connection before checking whether the processes handling previous
4754 connections had finished. The check for completed processes is now done
4755 earlier. On busy systems, this bug wouldn't be noticed because something
4756 else would have woken the daemon, and it would have reaped the completed
4759 80. If a message was submitted locally by a user whose login name contained one
4760 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4761 It caused a spool format error. I have fixed the spool reading code. A
4762 related problem was that the "from" clause in the Received: line became
4763 illegal because of the space(s). It is now covered by ${quote_local_part.
4765 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4768 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4770 83. Patch from Sieve maintainer to fix unterminated string problem in
4771 "vacation" handling.
4773 84. Some minor changes to the Linux configuration files to help with other
4774 OS variants using glibc.
4776 85. One more patch for Sieve to update vacation handling to latest spec.
4779 ----------------------------------------------------
4780 See the note above about the 4.44 and 4.50 releases.
4781 ----------------------------------------------------
4787 1. Change 4.43/35 introduced a bug that caused file counts to be
4788 incorrectly computed when quota_filecount was set in an appendfile
4791 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4792 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4795 3. Give more explanation in the error message when the command for a transport
4796 filter fails to execute.
4798 4. There are several places where Exim runs a non-Exim command in a
4799 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4800 was being done only for the command run by the queryprogram router. It is
4801 now done for all such subprocesses. The other cases are: ${run, transport
4802 filters, and the commands run by the lmtp and pipe transports.
4804 5. Some older OS have a limit of 256 on the maximum number of file
4805 descriptors. Exim was using setrlimit() to set 1000 as a large value
4806 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4807 systems. I've change it so that if it can't get 1000, it tries for 256.
4809 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4810 was an oversight, and furthermore, ever since the addition of extra
4811 controls (e.g. 4.43/32), the checks on when to allow different forms of
4812 "control" were broken. There should now be diagnostics for all cases when a
4813 control that does not make sense is encountered.
4815 7. $recipients is now available in the predata ACL (oversight).
4817 8. Tidy the search cache before the fork to do a delivery from a message
4818 received from the command line. Otherwise the child will trigger a lookup
4819 failure and thereby defer the delivery if it tries to use (for example) a
4820 cached ldap connection that the parent has called unbind on.
4822 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4823 of $address_data from the recipient verification was clobbered by the
4824 sender verification.
4826 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4827 was its contents. (It was OK if the option was not defined at all.)
4829 11. A "Completed" log line is now written for messages that are removed from
4830 the spool by the -Mrm option.
4832 12. $host_address is now set to the target address during the checking of
4833 ignore_target_hosts.
4835 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4836 being passed; this would have caused $sender_host_name to have been used if
4837 matching the list had actually called for a host name (not very likely,
4838 since this list is usually IP addresses). A host name is now passed as
4841 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4842 code that specifies a non-restarting handler (typically sigaction() in
4843 modern systems) in an attempt to fix a rare and obscure crash bug.
4845 15. Narrowed the window for a race in the daemon that could cause it to ignore
4846 SIGCHLD signals. This is not a major problem, because they are used only to
4847 wake it up if nothing else does.
4849 16. A malformed maildirsize file could cause Exim to calculate negative values
4850 for the mailbox size or file count. Odd effects could occur as a result.
4851 The maildirsize information is now recalculated if the size or filecount
4854 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4855 support for a long time. Removed HAVE_SYS_VFS_H.
4857 18. Updated exipick to current release from John Jetmore.
4859 19. Allow an empty sender to be matched against a lookup in an address list.
4860 Previously the only cases considered were a regular expression, or an
4863 20. Exim went into a mad DNS lookup loop when doing a callout where the
4864 host was specified on the transport, if the DNS lookup yielded more than
4867 21. The RFC2047 encoding function was originally intended for short strings
4868 such as real names; it was not keeping to the 75-character limit for
4869 encoded words that the RFC imposes. It now respects the limit, and
4870 generates multiple encoded words if necessary. To be on the safe side, I
4871 have increased the buffer size for the ${rfc2047: expansion operator from
4874 22. Failure to deliver a bounce message always caused it to be frozen, even if
4875 there was an errors_to setting on the router. The errors_to setting is now
4878 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4879 canonical form (fully expanded) before being placed in
4880 $sender_host_address.
4882 24. Updated eximstats to version 1.33
4884 25. Include certificate and key file names in error message when GnuTLS fails
4885 to set them up, because the GnuTLS error message doesn't include the name
4886 of the failing file when there is a problem reading it.
4888 26. Expand error message when OpenSSL has problems setting up cert/key files.
4891 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4892 (this can affect the format of dates).
4894 28. exim_tidydb, when checking for the continued existence of a message for
4895 which it has found a message-specific retry record, was not finding
4896 messages that were in split spool directories. Consequently, it was
4897 deleting retry records that should have stayed in existence.
4899 29. eximstats updated to version 1.35
4900 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4901 1.35 - bugfix such that pie charts by volume are generated correctly
4903 30. The SPA authentication driver was not abandoning authentication and moving
4904 on to the next authenticator when an expansion was forced to fail,
4905 contradicting the general specification for all authenticators. Instead it
4906 was generating a temporary error. It now behaves as specified.
4908 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4909 (the order specifies the preference for clients). The order is now AES256,
4910 AES128, 3DES, ARCFOUR128.
4912 31. Small patch to Sieve code - explicitly set From: when generating an
4915 32. Exim crashed if a remote delivery caused a very long error message to be
4916 recorded - for instance if somebody sent an entire SpamAssassin report back
4917 as a large number of 550 error lines. This bug was coincidentally fixed by
4918 increasing the size of one of Exim's internal buffers (big_buffer) that
4919 happened as part of the Exiscan merge. However, to be on the safe side, I
4920 have made the code more robust (and fixed the comments that describe what
4923 33. Some experimental protocols are using DNS PTR records for new purposes. The
4924 keys for these records are domain names, not reversed IP addresses. The
4925 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4926 leaves it alone. Component reversal etc. now happens only for IP addresses.
4929 34. The host_aton() function is supposed to be passed a string that is known
4930 to be a valid IP address. However, in the case of IPv6 addresses, it was
4931 not checking this. This is a hostage to fortune. Exim now panics and dies
4932 if the condition is not met. A case was found where this could be provoked
4933 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4934 components; fortuitously, this particular loophole had already been fixed
4935 by change 4.50/55 or 4.44/33 above.
4937 If there are any other similar loopholes, the new check in host_aton()
4938 itself should stop them being exploited. The report I received stated that
4939 data on the command line could provoke the exploit when Exim was running as
4940 exim, but did not say which command line option was involved. All I could
4941 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4942 running as the user.
4945 35. There was a buffer overflow vulnerability in the SPA authentication code
4946 (which came originally from the Samba project). I have added a test to the
4947 spa_base64_to_bits() function which I hope fixes it.
4950 36. The daemon start-up calls getloadavg() while still root for those OS that
4951 need the first call to be done as root, but it missed one case: when
4952 deliver_queue_load_max is set with deliver_drop_privilege. This is
4953 necessary for the benefit of the queue runner, because there is no re-exec
4954 when deliver_drop_privilege is set.
4956 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4957 was in use, was not putting the data itself into the right store pool;
4958 consequently, it could be overwritten for a subsequent message in the same
4959 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4962 38. Sometimes the final signoff response after QUIT could fail to get
4963 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4964 < 0 before doing a fflush(). This bug looks as though it goes back to the
4965 introduction of TLS in release 3.20, but "sometimes" must have been rare
4966 because the tests only now provoked it.
4972 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4973 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4974 Now it returns an integer. A number of calls were still expecting a T/F
4975 return. Fortuitously, in all cases, the tests worked in OK situations,
4976 which is the norm. However, things would have gone wrong on any write
4977 failures on the smtp file descriptor. This function is used when sending
4978 messages over SMTP and also when doing verify callouts.
4980 2. When Exim is called to do synchronous delivery of a locally submitted
4981 message (the -odf or -odi options), it no longer closes stderr before doing
4984 3. Implemented the mua_wrapper option.
4986 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4988 5. Implemented the functions header_remove(), header_testname(),
4989 header_add_at_position(), and receive_remove_recipient(), and exported them
4992 6. If an ACL "warn" statement specified the addition of headers, Exim already
4993 inserted X-ACL-Warn: at the start if there was no header name. However, it
4994 was not making this test for the second and subsequent header lines if
4995 there were newlines in the string. This meant that an invalid header could
4996 be inserted if Exim was badly configured.
4998 7. Allow an ACL "warn" statement to add header lines at the start or after all
4999 the Received: headers, as well as at the end.
5001 8. Added the rcpt_4xx retry error code.
5003 9. Added postmaster_mailfrom=xxx to callout verification option.
5005 10. Added mailfrom=xxxx to the callout verification option, for verify=
5008 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5009 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5011 12. Inserted some casts to stop certain compilers warning when using pointer
5012 differences as field lengths or precisions in printf-type calls (mostly
5013 affecting debugging statements).
5015 13. Added optional readline() support for -be (dynamically loaded).
5017 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5018 same clock tick as a message's arrival, so that its received time was the
5019 same as the "first fail" time on the retry record, and that message
5020 remained on the queue past the ultimate address timeout, every queue runner
5021 would try a delivery (because it was past the ultimate address timeout) but
5022 after another failure, the ultimate address timeout, which should have then
5023 bounced the address, did not kick in. This was a "< instead of <=" error;
5024 in most cases the first failure would have been in the next clock tick
5025 after the received time, and all would be well.
5027 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5028 being recognized when the domain list was tested by the match_domain
5029 condition in an expansion string.
5031 16. Added the ${str2b64: operator.
5033 17. Exim was always calling setrlimit() to set a large limit for the number of
5034 processes, without checking whether the existing limit was already
5035 adequate. (It did check for the limit on file descriptors.) Furthermore,
5036 errors from getrlimit() and setrlimit() were being ignored. Now they are
5037 logged to the main and panic logs, but Exim does carry on, to try to do its
5038 job under whatever limits there are.
5040 18. Imported PCRE 5.0.
5042 19. Trivial typo in log message " temporarily refused connection" (the leading
5045 20. If the log selector return_path_on_delivery was set and an address was
5046 redirected to /dev/null, the delivery process crashed because it assumed
5047 that a return path would always be set for a "successful" delivery. In this
5048 case, the whole delivery is bypassed as an optimization, and therefore no
5051 21. Internal re-arrangement: the function for sending a challenge and reading
5052 a response while authentication was assuming a zero-terminated challenge
5053 string. It's now changed to take a pointer and a length, to allow for
5054 binary data in such strings.
5056 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5058 23. Exim was not respecting finduser_retries when seeking the login of the
5059 uid under which it was called; it was always trying 10 times. (The default
5060 setting of finduser_retries is zero.) Also, it was sleeping after the final
5061 failure, which is pointless.
5063 24. Implemented tls_on_connect_ports.
5065 25. Implemented acl_smtp_predata.
5067 26. If the domain in control=submission is set empty, Exim assumes that the
5068 authenticated id is a complete email address when it generates From: or
5069 Sender: header lines.
5071 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5072 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5073 chown and chgrp in /bin and hostname in /usr/bin.
5075 28. Exim was keeping the "process log" file open after each use, just as it
5076 does for the main log. This opens the possibility of it remaining open for
5077 long periods when the USR1 signal hits a daemon. Occasional processlog
5078 errors were reported, that could have been caused by this. Anyway, it seems
5079 much more sensible not to leave this file open at all, so that is what now
5082 29. The long-running daemon process does not normally write to the log once it
5083 has entered its main loop, and it closes the log before doing so. This is
5084 so that log files can straightforwardly be renamed and moved. However,
5085 there are a couple of unusual error situations where the daemon does write
5086 log entries, and I had neglected to close the log afterwards.
5088 30. The text of an SMTP error response that was received during a remote
5089 delivery was being truncated at 512 bytes. This is too short for some of
5090 the long messages that one sometimes sees. I've increased the limit to
5093 31. It is now possible to make retry rules that apply only when a message has a
5094 specific sender, in particular, an empty sender.
5096 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5097 it possible to be selective about when SMTP synchronization is enforced.
5099 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5101 32. Implemented hosts_connection_nolog.
5103 33. Added an ACL for QUIT.
5105 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5108 35. Added mailbox_size and mailbox_filecount to appendfile.
5110 36. Added control = no_multiline_responses to ACLs.
5112 37. There was a bug in the logic of the code that waits for the clock to tick
5113 in the case where the clock went backwards by a substantial amount such
5114 that the microsecond fraction of "now" was more than the microsecond
5115 fraction of "then" (but the whole seconds number was less).
5117 38. Added support for the libradius Radius client library this is found on
5118 FreeBSD (previously only the radiusclient library was supported).
5124 1. When certain lookups returned multiple values in the form name=value, the
5125 quoting of the values was not always being done properly. Specifically:
5126 (a) If the value started with a double quote, but contained no whitespace,
5128 (b) If the value contained whitespace other than a space character (i.e.
5129 tabs or newlines or carriage returns) it was not quoted.
5130 This fix has been applied to the mysql and pgsql lookups by writing a
5131 separate quoting function and calling it from the lookup code. The fix
5132 should probably also be applied to nisplus, ibase and oracle lookups, but
5133 since I cannot test any of those, I have not disturbed their existing code.
5135 2. A hit in the callout cache for a specific address caused a log line with no
5136 reason for rejecting RCPT. Now it says "Previous (cached) callout
5137 verification failure".
5139 3. There was an off-by-one bug in the queryprogram router. An over-long
5140 return line was truncated at 256 instead of 255 characters, thereby
5141 overflowing its buffer with the terminating zero. As well as fixing this, I
5142 have increased the buffer size to 1024 (and made a note to document this).
5144 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5145 when Exim is waiting for an SMTP response from a remote server, Exim
5146 restarts its select() call on the socket, thereby resetting its timeout.
5147 This is not a problem when such interrupts are rare. Somebody set up a cron
5148 job to run exiwhat every 2 minutes, which is less than the normal select()
5149 timeout (5 or 10 minutes). This meant that the select() timeout never
5150 kicked in because it was always reset. I have fixed this by comparing the
5151 time when an interrupt arrives with the time at the start of the first call
5152 to select(). If more time than the timeout has elapsed, the interrupt is
5153 treated as a timeout.
5155 5. Some internal re-factoring in preparation for the addition of Sieve
5156 extensions (by MH). In particular, the "personal" test is moved to a
5157 separate function, and given an option for scanning Cc: and Bcc: (which is
5158 not set for Exim filters).
5160 6. When Exim created an email address using the login of the caller as the
5161 local part (e.g. when creating a From: or Sender: header line), it was not
5162 quoting the local part when it contained special characters such as @.
5164 7. Installed new OpenBSD configuration files.
5166 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5167 try to make them clearer.
5169 9. Callout options, other than the timeout value, were being ignored when
5170 verifying sender addresses in header lines. For example, when using
5172 verify = header_sender/callout=no_cache
5174 the cache was (incorrectly) being used.
5176 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5177 only the Cygwin environment.
5179 11. When return_path_on_delivery was set as a log selector, if different remote
5180 addresses in the same message used different return paths and parallel
5181 remote delivery occurred, the wrong values would sometimes be logged.
5182 (Whenever a remote delivery process finished, the return path value from
5183 the most recently started remote delivery process was logged.)
5185 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5186 lines when AUTH and/or TLS are in use. This is the "received protocol"
5187 field. Exim used to use "asmtp" for authenticated SMTP, without any
5188 indication (in the protocol name) for TLS use. Now it follows the RFC and
5189 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5190 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5191 names appear in log lines as well as in Received: header lines.
5193 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5194 extensions, and comparison tests, and to fix some bugs.
5196 14. Changes to the "personal" filter test:
5198 (1) The test was buggy in that it was just doing the equivalent of
5199 "contains" tests on header lines. For example, if a user's address was
5200 anne@some.where, the "personal" test would incorrectly be true for
5202 To: susanne@some.where
5204 This test is now done by extracting each address from the header in turn,
5205 and checking the entire address. Other tests that are part of "personal"
5206 are now done using regular expressions (for example, to check local parts
5207 of addresses in From: header lines).
5209 (2) The list of non-personal local parts in From: addresses has been
5210 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5211 taken from the Sieve specification recommendations.
5213 (3) If the message contains any header line starting with "List-" it is
5214 treated as non-personal.
5216 (4) The test for "circular" in the Subject: header line has been removed
5217 because it now seems ill-conceived.
5219 15. Minor typos in src/EDITME comments corrected.
5221 16. Installed latest exipick from John Jetmore.
5223 17. If headers_add on a router specified a text string that was too long for
5224 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5225 of string_sprintf() is now avoided.
5227 18. $message_body_size was not set (it was always zero) when running the DATA
5228 ACL and the local_scan() function.
5230 19. For the "mail" command in an Exim filter, no default was being set for
5231 the once_repeat time, causing a random time value to be used if "once" was
5232 specified. (If the value happened to be <= 0, no repeat happened.) The
5233 default is now 0s, meaning "never repeat". The "vacation" command was OK
5234 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5235 (I found it when inspecting the code).
5237 20. There is now an overall timeout for performing a callout verification. It
5238 defaults to 4 times the callout timeout, which applies to individual SMTP
5239 commands during the callout. The overall timeout applies when there is more
5240 than one host that can be tried. The timeout is checked before trying the
5241 next host. This prevents very long delays if there are a large number of
5242 hosts and all are timing out (e.g. when the network connections are timing
5243 out). The value of the overall timeout can be changed by specifying an
5244 additional sub-option for "callout", called "maxwait". For example:
5246 verify = sender/callout=5s,maxwait=20s
5248 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5249 to the end before writing, but this should make it even safer).
5251 22. Exim was forgetting that it had advertised PIPELINING for the second and
5252 subsequent messages on an SMTP connection. It was also not resetting its
5253 memory on STARTTLS and an internal HELO.
5255 23. When Exim logs an SMTP synchronization error within a session, it now
5256 records whether PIPELINING has been advertised or not.
5258 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5259 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5260 rather than long int.
5262 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5264 26. Added the never_mail option to autoreply.
5270 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5271 crash if the getsockname() call failed; this can happen if a connection is
5272 closed very soon after it is established. The problem was simply in the
5273 order in which certain operations were done, causing Exim to try to write
5274 to the SMTP stream before it had set up the file descriptor. The bug has
5275 been fixed by making things happen in the correct order.
5281 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5282 before the connection was closed, thus losing the rejection response.
5284 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5285 some early Solaris releases, but causes trouble in current releases where
5286 socklen_t is defined.
5288 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5291 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5294 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5295 that releasing the top of it at the end releases what was used for sub-
5296 expansions (unless the block got too big). However, discard this block if
5297 the first thing is a variable or header, so that we can use its block when
5298 it is dynamic (useful for very large $message_headers, for example).
5300 6. Lookups now cache *every* query, not just the most recent. A new, separate
5301 store pool is used for this. It can be recovered when all lookup caches are
5302 flushed. Lookups now release memory at the end of their result strings.
5303 This has involved some general refactoring of the lookup sources.
5305 7. Some code has been added to the store_xxx() functions to reduce the amount
5306 of flapping under certain conditions.
5308 8. log_incoming_interface used to affect only the <= reception log lines. Now
5309 it causes the local interface and port to be added to several more SMTP log
5310 lines, for example "SMTP connection from", and rejection lines.
5312 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5314 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5316 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5317 could be overwritten at the end of the current message (or the start of a
5318 new message if it was set in a HELO ACL). The value is now preserved for
5319 the duration of the SMTP connection.
5321 12. If a transport had a headers_rewrite setting, and a matching header line
5322 contained an unqualified address, that address was qualified, even if it
5323 did not match any rewriting rules. The underlying bug was that the values
5324 of the flags that permit the existence of unqualified sender and recipient
5325 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5326 non-local messages, and by -bnq for local messages) were not being
5327 preserved with the message after it was received.
5329 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5330 "next input=" as part of the text comprising the host identity instead of
5331 the correct text. The code was using the same buffer for two different
5332 strings. However, depending on which order the printing function evaluated
5333 its arguments, the bug did not always show up. Under Linux, for example, my
5334 test suite worked just fine.
5336 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5337 doesn't work with some older versions of Perl. It has been changed to "my",
5338 which in any case is probably the better facility to use.
5340 15. A really picky compiler found some instances of statements for creating
5341 error messages that either had too many or two few arguments for the format
5344 16. The size of the buffer for calls to the DNS resolver has been increased
5345 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5346 for addresses that have a lot of PTR records. This alleviates a problem; it
5347 does not fully solve it.
5349 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5350 buffer now truncates the list and logs the incident, which is the same
5351 action as happens when Exim is looking up a host name and its aliases.
5352 Previously in this situation something unpredictable would happen;
5353 sometimes it was "internal error: store_reset failed".
5355 18. If a server dropped the connection unexpectedly when an Exim client was
5356 using GnuTLS and trying to read a response, the client delivery process
5357 crashed while trying to generate an error log message.
5359 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5360 string, for example:
5362 warn message = H1: something\nH2: something
5364 the text was added as a single header line from Exim's point of view
5365 though it ended up OK in the delivered message. However, searching for the
5366 second and subsequent header lines using $h_h2: did not work. This has been
5367 fixed. Similarly, if a system filter added multiple headers in this way,
5368 the routers could not see them.
5370 20. Expanded the error message when iplsearch is called with an invalid key to
5371 suggest using net-iplsearch in a host list.
5373 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5374 ACLs are no longer actually imposed (and a message to that effect is
5377 22. If a "gecos" field in a passwd entry contained escaped characters, in
5378 particular, if it contained a \" sequence, Exim got it wrong when building
5379 a From: or a Sender: header from that name. A second bug also caused
5380 incorrect handling when an unquoted " was present following a character
5381 that needed quoting.
5383 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5384 was not being matched caselessly.
5386 24. Arranged for all hyphens in the exim.8 source to be escaped with
5389 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5390 properly. Recipient callout cache records were still being keyed to include
5391 the sender, even when use_sender was set false. This led to far more
5392 callouts that were necessary. The sender is no longer included in the key
5393 when use_sender is false.
5395 26. Added "control = submission" modifier to ACLs.
5397 27. Added the ${base62d: operator to decode base 62 numbers.
5399 28. dnsdb lookups can now access SRV records.
5401 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5402 the configuration file.
5404 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5405 (-v) mode. This makes the output for a verbose queue run more intelligible.
5407 31. Added a use_postmaster feature to recipient callouts.
5409 32. Added the $body_zerocount variable, containing the number of binary zero
5410 bytes in the message body.
5412 33. The time of last modification of the "new" subdirectory is now used as the
5413 "mailbox time last read" when there is a quota error for a maildir
5416 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5418 35. Added +ignore_unknown as a special item in host lists.
5420 36. Code for decoding IPv6 addresses in host lists is now included, even if
5421 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5422 address was recognized as an IP address, but was then not correctly decoded
5423 into binary, causing unexpected and incorrect effects when compared with
5430 1. Very minor rewording of debugging text in manualroute to say "list of
5431 hosts" instead of "hostlist".
5433 2. If verify=header_syntax was set, and a header line with an unqualified
5434 address (no domain) and a large number of spaces between the end of the
5435 name and the colon was received, the reception process suffered a buffer
5436 overflow, and (when I tested it) crashed. This was caused by some obsolete
5437 code that should have been removed. The fix is to remove it!
5439 3. When running in the test harness, delay a bit after writing a bounce
5440 message to get a bit more predictability in the log output.
5442 4. Added a call to search_tidyup() just before forking a reception process. In
5443 theory, someone could use a lookup in the expansion of smtp_accept_max_
5444 per_host which, without the tidyup, could leave open a database connection.
5446 5. Added the variables $recipient_data and $sender_data which get set from a
5447 lookup success in an ACL "recipients" or "senders" condition, or a router
5448 "senders" option, similar to $domain_data and $local_part_data.
5450 6. Moved the writing of debug_print from before to after the "senders" test
5453 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5454 problems for message scanning, either using a data ACL, or using
5455 local_scan() because the Received: header was not generated till after they
5456 were called (in order to set the time as the time of reception completion).
5457 I have revised the way this works. The header is now generated after the
5458 body is received, but before the ACL or local_scan() are called. After they
5459 are run, the timestamp in the header is updated.
5465 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5466 before starting a queue runner without re-exec. This happened only when
5467 deliver_drop_privilege was set or when the Exim user was set to root. The
5468 effect of the bug was that timeouts during subsequent deliveries caused
5469 crashes instead of being properly handled. The handler is now left at its
5470 default (and expected) setting.
5472 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5473 message, again when deliver_drop_privilege is set or Exim is run as root.
5474 The bug described in (1) was not present in this case, but the tidying up
5475 of the other signals was missing. I have made the two cases consistent.
5477 3. The ignore_target_hosts setting on a manualroute router was being ignored
5478 for hosts that were looked up using the /MX notation.
5480 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5483 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5484 operated on the sender address. After changing the $sender_address to <>
5485 for the sender address verify, Exim was re-instated it as the original
5486 (before rewriting) address, but remembering that it had rewritten it, so it
5487 wasn't rewriting it again. This bug also had the effect of breaking the
5488 sender address verification caching when the sender address was rewritten.
5490 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5491 This has been changed so that if the ip literal address matches
5492 ignore_target_hosts, the router declines.
5494 7. Added expansion conditions match_domain, match_address, and match_local_
5495 part (NOT match_host).
5497 8. The placeholder for the Received: header didn't have a length field set.
5499 9. Added code to Exim itself and to exim_lock to test for a specific race
5500 condition that could lead to file corruption when using MBX delivery. The
5501 issue is with the lockfile that is created in /tmp. If this file is removed
5502 after a process has opened it but before that process has acquired a lock,
5503 there is the potential for a second process to recreate the file and also
5504 acquire a lock. This could lead to two Exim processes writing to the file
5505 at the same time. The added code performs the same test as UW imapd; it
5506 checks after acquiring the lock that its file descriptor still refers to
5507 the same named file.
5509 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5510 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5513 11. Added the smtp_active_hostname option. If used, this will typically be made
5514 to depend on the incoming interface address. Because $interface_address is
5515 not set up until the daemon has forked a reception process, error responses
5516 that can happen earlier (such as "too many connections") no longer contain
5519 12. If an expansion in a condition on a "warn" statement fails because a lookup
5520 defers, the "warn" statement is abandoned, and the next ACL statement is
5521 processed. Previously this caused the whole ACL to be aborted.
5523 13. Added the iplsearch lookup type.
5525 14. Added ident_timeout as a log selector.
5527 15. Added tls_certificate_verified as a log selector.
5529 16. Added a global option tls_require_ciphers (compare the smtp transport
5530 option of the same name). This controls incoming TLS connections.
5532 17. I finally figured out how to make tls_require_ciphers do a similar thing
5533 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5534 before starting the TLS session.
5536 18. Tabs are now shown as \t in -bP output.
5538 19. If the log selector return_path_on_delivery was set, Exim crashed when
5539 bouncing a message because it had too many Received: header lines.
5541 20. If two routers both had headers_remove settings, and the first one included
5542 a superfluous trailing colon, the final name in the first list and the
5543 first name in the second list were incorrectly joined into one item (with a
5544 colon in the middle).
5550 1. Added -C and -D options to the exinext utility, mainly to make it easier
5551 to include in the automated testing, but these could be helpful when
5552 multiple configurations are in use.
5554 2. The exinext utility was not formatting the output nicely when there was
5555 an alternate port involved in the retry record key, nor when there was a
5556 message id as well (for retries that were specific to a specific message
5557 and a specific host). It was also confused by IPv6 addresses, because of
5558 the additional colons they contain. I have fixed the IPv4 problem, and
5559 patched it up to do a reasonable job for IPv6.
5561 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5562 delivery, the log line now contains "pipelined" if PIPELINING was used.
5564 4. An SMTP transport process used to panic and die if the bind() call to set
5565 an explicit outgoing interface failed. This has been changed; it is now
5566 treated in the same way as a connect() failure.
5568 5. A reference to $sender_host_name in the part of a conditional expansion
5569 that was being skipped was still causing a DNS lookup. This no longer
5572 6. The def: expansion condition was not recognizing references to header lines
5573 that used bh_ and bheader_.
5575 7. Added the _cache feature to named lists.
5577 8. The code for checking quota_filecount in the appendfile transport was
5578 allowing one more file than it should have been.
5580 9. For compatibility with Sendmail, the command line option
5588 and sets the incoming protocol and host name (for trusted callers). The
5589 host name and its colon can be omitted when only the protocol is to be set.
5590 Note the Exim already has two private options, -pd and -ps, that refer to
5591 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5592 "s", but I don't think that's a major issue.
5594 10. A number of refactoring changes to the code, none of which should affect
5597 (a) The number of logging options was getting close to filling up the
5598 32-bit word that was used as a bit map. I have split them into two classes:
5599 those that are passed in the argument to log_write(), and those that are
5600 only ever tested independently outside of that function. These are now in
5601 separate 32-bit words, so there is plenty of room for expansion again.
5602 There is no change in the user interface or the logging behaviour.
5604 (b) When building, for example, log lines, the code previously used a
5605 macro that called string_cat() twice, in order to add two strings. This is
5606 not really sufficiently general. Furthermore, there was one instance where
5607 it was actually wrong because one of the argument was used twice, and in
5608 one call a function was used. (As it happened, calling the function twice
5609 did not affect the overall behaviour.) The macro has been replaced by a
5610 function that can join an arbitrary number of extra strings onto a growing
5613 (c) The code for expansion conditions now uses a table and a binary chop
5614 instead of a serial search (which was left over from when there were very
5615 few conditions). Also, it now recognizes conditions like "pam" even when
5616 the relevant support is not compiled in: a suitably worded error message is
5617 given if an attempt is made to use such a condition.
5619 11. Added ${time_interval:xxxxx}.
5621 12. A bug was causing one of the ddress fields not to be passed back correctly
5622 from remote delivery subprocesses. The field in question was not being
5623 subsequently used, so this caused to problems in practice.
5625 13. Added new log selectors queue_time and deliver_time.
5627 14. Might have fixed a bug in maildirsizefile handling that threw up
5628 "unexpected character" debug warnings, and recalculated the data
5629 unnecessarily. In any case, I expanded the warning message to give more
5632 15. Added the message "Restricted characters in address" to the statements in
5633 the default ACL that block characters like @ and % in local parts.
5635 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5636 Three changes have been made:
5638 (a) There was a serious bug; a negative response to MAIL caused the whole
5639 recipient domain to be cached as invalid, thereby blocking all messages
5640 to all local parts at the same domain, from all senders. This bug has
5641 been fixed. The domain is no longer cached after a negative response to
5642 MAIL if the sender used is not empty.
5644 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5647 (c) A new callout option, "use_sender" has been added for people who want
5648 the modified behaviour.
5654 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5657 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5658 indeed breaks things for older releases.
5660 3. Added additional logging to the case where there is a problem reading data
5661 from a filter that is running in a subprocess using a pipe, in order to
5662 try to track down a specific problem.
5664 4. Testing facility fudge: when running in the test harness and attempting
5665 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5666 getting "No route to host". Convert this to a timeout.
5668 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5671 6. Some OS don't have socklen_t but use size_t instead. This affects the
5672 fifth argument of getsockopt() amongst other things. This is now
5673 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5674 can be set for individual OS. I have set it for SunOS5, OSF1, and
5675 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5676 some earlier ones do not.
5678 7. Change 4.30/15 was not doing the test caselessly.
5680 8. The standard form for an IPv6 address literal was being rejected by address
5681 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5682 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5683 this, as well as the form without the "IPv6" on the front (but only when
5684 address literals are enabled, of course).
5686 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5688 10. Exim crashed if a message with an empty sender address specified by -f
5689 encountered a router with an errors_to setting. This could be provoked only
5690 by a command such as
5694 where an empty string was supplied; "<>" did not hit this bug.
5696 11. Installed PCRE release 4.5.
5698 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5699 remained set. It is now erased.
5701 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5702 times from message ids (which are base 36 rather than the normal 62).
5704 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5705 were being counted as actual protocol errors, and logged if the log
5706 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5707 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5708 and DATA following a set of rejected RCPTs do not count as protocol errors.
5709 In other words, Exim assumes they were pipelined, though this may not
5710 actually be the case. Of course, in all cases the client gets an
5711 appropriate error code.
5713 15. If a lookup fails in an ACL condition, a message about the failure may
5714 be available; it is used if testing the ACL cannot continue, because most
5715 such messages specify what the cause of the deferral is. However, some
5716 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5717 that caused an old message to be retained and used if a later statement
5718 caused a defer, replacing the real cause of the deferral.
5720 16. If an IP address had so many PTR records that the DNS lookup buffer
5721 was not large enough to hold them, Exim could crash while trying to process
5722 the truncated data. It now detects and logs this case.
5724 17. Further to 4.21/58, another change has been made: if (and only if) the
5725 first line of a message (the first header line) ends with CRLF, a bare LF
5726 in a subsequent header line has a space inserted after it, so as not to
5727 terminate the header.
5729 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5730 unnecessarily, used atoi() instead of strtol(), and didn't check the
5731 termination when getting file sizes from file names by regex.
5733 19. Completely re-implemented the support for maildirsize files, in the light
5734 of a number of problems with the previous contributed implementation
5735 (4.30/29). In particular:
5737 . If the quota is zero, the maildirsize file is maintained, but no quota is
5740 . If the maildir directory does not exist, it is created before any attempt
5741 to write a maildirsize file.
5743 . The quota value in the file is just a cache; if the quota is changed in
5744 the transport, the new value overrides.
5746 . A regular expression is available for excluding directories from the
5749 20. The autoreply transport checks the characters in options that define the
5750 message's headers; it allows continued headers, but it was checking with
5751 isspace() after an embedded newline instead of explicitly looking for a
5754 21. If all the "regular" hosts to which an address was routed had passed their
5755 expiry times, and had not reached their retry times, the address was
5756 bounced, even if fallback hosts were defined. Now Exim should go on to try
5759 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5760 equivalent code in the SMTP transport. Some hosts send humungous responses
5761 to HELO/EHLO, more than 1024 it seems.
5763 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5764 gives compiler warnings in some environments. I've now done it "properly",
5767 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5768 (because of problems with the built-in one) was declared to return uschar *
5769 instead of char *, causing compiler failure.
5771 25. Fixed a file descriptor leak when processing alias/forward files.
5773 26. Fixed a minor format string issue in dbfn.c.
5775 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5777 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5778 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5781 29. When a sender address is verified, it is cached, to save repeating the test
5782 when there is more than one recipient in a message. However, when the
5783 verification involves a callout, it is possible for different callout
5784 options to be set for different recipients. It is too complicated to keep
5785 track of this in the cache, so now Exim always runs a verification when a
5786 callout is required, relying on the callout cache for the optimization.
5787 The overhead is duplication of the address routing, but this should not be
5790 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5791 to be verified with callout=postmaster, and the main callout worked but the
5792 postmaster check failed, the verification correctly failed. However, if a
5793 subsequent RCPT command asked for sender verification *without* the
5794 postmaster check, incorrect caching caused this verification also to fail,
5797 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5798 it was not caching the DNS options (qualify_single, search_parents) that
5799 were used when the lookup failed. A subsequent lookup with different
5800 options therefore always gave the same answer, though there were cases
5801 where it should not have. (Example: a "domains = !$mx_any" option on a
5802 dnslookup router: the "domains" option is always processed without any
5803 widening, but the router might have qualify_single set.) Now Exim uses the
5804 cached value only when the same options are set.
5806 32. Added John Jetmore's "exipick" utility to the distribution.
5808 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5809 than a timeout (e.g. a certificate is required, and is not provided), an
5810 Exim server now closes the connection immediately. Previously it waited for
5811 the client to close - but if the client is SSL, it seems that they each
5812 wait for each other, leading to a delay before one of them times out.
5814 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5815 maintained 0.8.x compatibility because I don't think many are using it, and
5816 it is clearly obsolete.
5818 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5821 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5822 client certificate was expired. A simple patch fixes this, though I don't
5823 understand the full logic of why the verify callback is called multiple
5826 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5827 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5828 which causes problems with some clients (such as the Certicom SSL Plus
5829 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5830 disables the coutermeasure allowing Eudora to connect."
5832 38. Exim was not checking that a write() to a log file succeeded. This could
5833 lead to Bad Things if a log got too big, in particular if it hit a file
5834 size limit. Exim now panics and dies if it cannot write to a log file, just
5835 as it does if it cannot open a log file.
5837 39. Modified OS/Makefile-Linux so that it now contains
5839 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5841 The two -D definitions ensure that Exim is compiled with large file
5842 support, which makes it possible to handle log files that are bigger than
5845 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5846 instance) a domain was checked against a named list that involved a lookup,
5847 causing $domain_data to be set, then another domain was checked against the
5848 same list, then the first domain was re-checked, the value of $domain_data
5849 after the final check could be wrong. In particular, if the second check
5850 failed, it could be set empty. This bug probably also applied to
5853 41. The strip_trailing_dot option was not being applied to the address given
5854 with the -f command-line option.
5856 42. The code for reading a message's header from the spool was incrementing
5857 $received_count, but never initializing it. This meant that the value was
5858 incorrect (doubled) while delivering a message in the same process in which
5859 it was received. In the most common configuration of Exim, this never
5860 happens - a fresh exec is done - but it can happen when
5861 deliver_drop_privilege is set.
5863 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5864 it now includes up to 150 characters of the unexpected data in the log
5867 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5868 and building data strings. The size of both of these buffers was 10 000
5869 bytes - far larger than anybody would *ever* want, thought I. Needless to
5870 say, somebody hit the limit. I have increased the maximum line length to
5871 20 000 and the maximum data length of concatenated lines to 100 000. I have
5872 also fixed two bugs, because there was no checking on these buffers. Tsk,
5873 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5874 buffer is too small.
5876 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5877 lsearch lookups. Now it does.
5879 46. When parsing a route_list item in a manualroute router, a fixed-length
5880 buffer was used for the list of hosts. I made this 1024 bytes long,
5881 thinking that nobody would ever have a list of hosts that long. Wrong.
5882 Somebody had a whole pile of complicated expansion conditions, and the
5883 string was silently truncated, leading to an expansion error. It turns out
5884 that it is easier to change to an unlimited length (owing to other changes
5885 that have happened since this code was originally written) than to build
5886 structure for giving a limitation error. The length of the item that
5887 expands into the list of hosts is now unlimited.
5889 47. The lsearch lookup could not handle data where the length of text line was
5890 more than 4095 characters. Such lines were truncated, leading to shortened
5891 data being returned. It should now handle lines of any length.
5893 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5894 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5897 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5898 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5899 no longer happens in comment lines. A list of replacements is now placed
5900 at the head of all of the source files, except those whose only change is
5901 to replace PERL_COMMAND in the very first #! line.
5903 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5904 messages on the queue, with a bottom-up merge sort, using code contributed
5905 by Michael Haardt. This should make operations like -bp somewhat faster on
5906 large queues. It won't affect queue runners, except when queue_run_in_order
5909 51. Installed eximstats 1.31 in the distribution.
5911 52. Added support for SRV lookups to the dnslookup router.
5913 53. If an ACL referred to $message_body or $message_body_end, the value was not
5914 reset for any messages that followed in the same SMTP session.
5916 54. The store-handling optimization for building very long strings was not
5917 differentiating between the different store pools. I don't think this
5918 actually made any difference in practice, but I've tidied it.
5920 55. While running the routers to verify a sender address, $sender_address
5921 was still set to the sender address. This is wrong, because when routing to
5922 send a bounce to the sender, it would be empty. Therefore, I have changed
5923 it so that, while verifying a sender address, $sender_address is set to <>.
5924 (There is no change to what happens when verifying a recipient address.)
5926 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5927 target A or AAAA records (if not already returned) without resetting the
5928 qualify_single or search_parents options of the DNS resolver. These are
5929 inappropriate in this case because the targets of MX and SRV records must
5930 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5931 target that, when qualified, matched something in the local domain. These
5932 two options are now turned off when doing these lookups.
5934 57. It seems that at least some releases of Reiserfs (which does not have the
5935 concept of a fixed number of inodes) returns zero and not -1 for the
5936 number of available inodes. This interacted badly with check_spool_inodes,
5937 which assumed that -1 was the "no such thing" setting. What I have done is
5938 to check that the total number of inodes is greater than zero before doing
5939 the test of how many are available.
5941 58. When a "warn" ACL statement has a log_message modifier, the message is
5942 remembered, and not repeated. This is to avoid a lot of repetition when a
5943 message has many recipients that cause the same warning to be written.
5944 However, Exim was preserving the list of already written lines for an
5945 entire SMTP session, which doesn't seem right. The memory is now reset if a
5946 new message is started.
5948 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5949 debugging output unless log_rewrite was also set.
5951 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5952 of (int)(handle) when we know that handle contains (void *)(-1).
5954 61. The Exim daemon panic-logs an error return when it closes the incoming
5955 connection. However "connection reset by peer" seems to be common, and
5956 isn't really an error worthy of noting specially, so that particular error
5959 62. When Exim is trying to find all the local interfaces, it used to panic and
5960 die if the ioctl to get the interface flags failed. However, it seems that
5961 on at least one OS (Solaris 9) it is possible to have an interface that is
5962 included in the list of interfaces, but for which you get a failure error
5963 for this call. This happens when the interface is not "plumbed" into a
5964 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5965 failure of the "get flags" call assumes that the interface is down.
5967 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5968 makes life easier for people who are doing arithmetic on fields extracted
5969 from dates, where you often get leading zeros that should not be
5970 interpreted as octal.
5972 64. Added qualify_domain to the redirect router, to override the global
5975 65. If a pathologically long header line contained very many addresses (the
5976 report of this problem mentioned 10 000) and each of them was rewritten,
5977 Exim could use up a very large amount of memory. (It kept on making new
5978 copies of the header line as it rewrote, and never released the old ones.)
5979 At the expense of a bit more processing, the header rewriting function has
5980 been changed so that it no longer eats memory in this way.
5982 66. The generation of the Received: header has been moved from the time that a
5983 message starts to be received, to the time that it finishes. The timestamp
5984 in the Received: header should now be very close to that of the <= log
5985 line. There are two side-effects of this change:
5987 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5988 logged header lines no longer include the local Received: line, because
5989 it has not yet been created. The same applies to a copy of the message
5990 that is returned to a non-SMTP sender when a message is rejected.
5992 (b) When a filter file is tested using -bf, no additional Received: header
5993 is added to the test message. After some thought, I decided that this
5996 This change does not affect the value of $received_for. It is still set
5997 after address rewriting, but before local_scan() is called.
5999 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6001 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6002 gave an unhelpful panic error message, and a defer error. I have managed to
6003 change this behaviour so that it now rejects any supplied certificate,
6004 which seems right, as the list of acceptable certificates is empty.
6006 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6007 gave an unhelpful defer error. I have not managed to make this reject any
6008 supplied certificates, but the error message it gives is "no certificate
6009 supplied", which is not helpful.
6011 70. exigrep's output now also includes lines that are not associated with any
6012 message, but which match the given pattern. Implemented by a patch from
6013 Martin Sluka, which also tidied up the Perl a bit.
6015 71. Recipient callout verification, like sender verification, was using <> in
6016 the MAIL FROM command. This isn't really the right thing, since the actual
6017 sender may affect whether the remote host accepts the recipient or not. I
6018 have changed it to use the actual sender in the callout; this means that
6019 the cache record is now keyed on a recipient/sender pair, not just the
6020 recipient address. There doesn't seem to be a real danger of callout loops,
6021 since a callout by the remote host to check the sender would use <>.
6022 [SEE ABOVE: changed after hitting problems.]
6024 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6025 temporary errors. However, in the case of such a code being given after
6026 the end of a data transmission (i.e. after ".") Exim was failing to write
6027 a retry record for the message. (Yes, there was some broken host that was
6028 actually sending 8xx at this point.)
6030 73. An unknown lookup type in a host list could cause Exim to panic-die when
6031 the list was checked. (An example that provoked this was putting <; in the
6032 middle of a list instead of at the start.) If this happened during a DATA
6033 ACL check, a -D file could be left lying around. This kind of configuration
6034 error no longer causes Exim to die; instead it causes a defer error. The
6035 incident is still logged to the main and panic logs.
6037 74. Buglet left over from Exim 3 conversion. The message "too many messages
6038 in one connection" was written to the rejectlog but not the mainlog, except
6039 when address rewriting (yes!) was being logged.
6041 75. Added write_rejectlog option.
6043 76. When a system filter was run not as root (that is, when system_filter_user
6044 was set), the values of the $n variables were not being returned to the
6045 main process; thus, they were not subsequently available in the $sn
6048 77. Added +return_path_on_delivery log selector.
6050 78. A connection timeout was being treated differently from recipients deferred
6051 when testing hosts_max_try with a message that was older than the host's
6052 retry timeout. (The host should not be counted, thus allowing all hosts to
6053 be tried at least once before bouncing.) This may have been the cause of an
6054 occasionally reported bug whereby a message would remain on the queue
6055 longer than the retry timeout, but would be bounced if a delivery was
6056 forced. I say "may" because I never totally pinned down the problem;
6057 setting up timeout/retry tests is difficult. See also the next item.
6059 79. The ultimate address timeout was not being applied to errors that involved
6060 a combination of host plus message (for example, a timeout on a MAIL
6061 command). When an address resolved to a number of possible hosts, and they
6062 were not all tried for each delivery (e.g. because of hosts_max_try), a
6063 message could remain on the queue longer than the retry timeout.
6065 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6068 81. Fixed an obscure SMTP outgoing bug which required at least the following
6069 conditions: (a) there was another message waiting for the same server;
6070 (b) the server returned 5xx to all RCPT commands in the first message so
6071 that the message was not completed; (c) the server dropped the connection
6072 or gave a negative response to the RSET that Exim sends to abort the
6073 transaction. The observed case was a dropped connection after DATA that had
6074 been sent in pipelining mode. That is, the server had advertised PIPELINING
6075 but was not implementing it correctly. The effect of the bug was incorrect
6076 behaviour, such as trying another host, and this could lead to a crash.
6082 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6083 and daemon.c were passed as pointers to ints; they should have been
6084 pointers to socklen_t variables (which are typically unsigned ints).
6086 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6089 3. Fixed a really odd bug that affected only the testing scheme; patching a
6090 certain fixed string in the binary changed the value of another string that
6091 happened to be identical to the end of the original first string.
6093 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6094 name", it returns that address as the IP address. On some operating
6095 systems (e.g. Solaris), it also passes back the IP address string as the
6096 "host name". However, on others (e.g. Linux), it passes back an empty
6097 string. Exim wasn't checking for this, and was changing the host name to an
6098 empty string, assuming it had been canonicalized.
6100 5. Although rare, it is permitted to have more than one PTR record for a given
6101 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6102 all the names associated with an address, because they do in Solaris.
6103 However, it seems that they do not in Linux for data that comes from the
6104 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6105 I found this out when I moved to a new Linux workstation and tried to run
6106 the Exim test suite.
6108 To get round this problem I have changed the code so that it now does its
6109 own call to the DNS to look up PTR records when searching for a host name.
6110 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6111 addresses that are only in /etc/hosts are still found.
6113 This behaviour is, however, controlled by an option called host_lookup_
6114 order, which defaults to "bydns:byaddr". If people want to use the other
6115 order, or indeed, just use one or the other means of lookup, they can
6116 specify it in this variable.
6118 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6119 some operating systems, this comes back from gethostbyaddr() as an empty
6120 string, and this is what Exim used to test for. However, it seems that in
6121 other systems, "." is yielded. Exim now tests for this case too.
6123 7. The values of check_spool_space and check_log_space are now held internally
6124 as a number of kilobytes instead of an absolute number of bytes. If a
6125 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6126 kilobyte. This means that much larger values can be stored.
6128 8. Exim monitor: an attempt to get the action menu when not actually pointing
6129 at a message produces an empty menu entitled "No message selected". This
6130 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6131 no entries in it ("Shell widget menu has zero width and/or height"). So I
6132 have added a single, blank menu entry in this case.
6134 9. Added ${quote_local_part.
6136 10. MIME decoding is now applied to the contents of Subject: header lines when
6139 11. Now that a reference to $sender_host_address automatically causes a reverse
6140 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6141 host lookup before query-style lookups in lists that might use this
6142 variable. This has therefore been abolished, and the "net-" prefix is no
6143 longer necessary for query-style lookups.
6145 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6146 have been a typo for LFLAGS, so it has been changed.
6148 13. The install script calls Exim with "-C /dev/null" in order to find the
6149 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6150 to be output. However, since Exim outputs its version number before the
6151 error, it didn't break the script. It just looked ugly. I fixed this by
6152 always allowing "-C /dev/null" if the caller is root.
6154 14. Ignore overlarge ACL variable number when reading spool file - insurance
6155 against a later release with more variables having written the file.
6157 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6158 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6159 this, as well as the form without the "IPv6" on the front.
6161 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6162 OS/Makefile-Darwin file.
6164 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6165 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6167 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6168 variable. However, some OS use ints for inodes. I've added cast to long int
6169 to get rid of the compiler warning.
6171 19. I had forgotten to lock out "/../" in configuration file names when
6172 ALT_CONFIG_PREFIX was set.
6174 20. Routers used for verification do not need to specify transports. However,
6175 if such a router generated a host list, and callout was configured, Exim
6176 crashed, because it could not find a port number from the (non-existent)
6177 transport. It now assumes port 25 in this circumstance.
6179 21. Added the -t option to exigrep.
6181 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6182 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6183 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6184 them is set, LOOKUP_LSEARCH is forced.
6186 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6188 24. Added sender and host information to the "rejected by local_scan()" log
6189 line; previously there was no indication of these.
6191 25. Added .include_if_exists.
6193 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6194 Linux. It turns out that not all file systems support this. Apparently some
6195 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6196 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6197 not supported on the file descriptor, is now ignored when Exim is trying to
6198 sync a directory. This applies only to Linux.
6200 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6202 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6203 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6206 29. Added support for maildirsize files from supplied patch (modified a bit).
6208 30. The use of :fail: followed by an empty string could lead Exim to respond to
6209 sender verification failures with (e.g.):
6211 550 Verification failed for <xxx>
6212 550 Sender verify failed
6214 where the first response line was missing the '-' that indicates it is not
6215 the final line of the response.
6217 31. The loop for finding the name of the user that called Exim had a hardwired
6218 limit of 10; it now uses the value of finduser_retries, which is used for
6219 all other user lookups.
6221 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6224 33. Exim was neglecting to zero errno before one call of strtol() when
6225 expanding a string and expecting an integer value. On some systems this
6226 resulted in spurious "integer overflow" errors. Also, it was casting the
6227 result into an int without checking.
6229 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6230 did not work. The code looks as if it has *never* worked, though it appears
6231 to have been documented since at least release 1.62. I have made it work.
6233 35. The "timeout_DNS" error in retry rules, also documented since at least
6234 1.62, also never worked. As it isn't clear exactly what this means, and
6235 clearly it isn't a major issue, I have abolished the feature by treating it
6236 as "timeout", and writing a warning to the main and panic logs.
6238 36. The display of retry rules for -brt wasn't always showing the error code
6241 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6242 timeout_connect_A, timeout_connect_MX.
6244 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6245 to the empty sender.
6247 39. The daemon was not analysing the content of -oX till after it had closed
6248 stderr and disconnected from the controlling terminal. This meant that any
6249 syntax errors were only noted on the panic log, and the return code from
6250 the command was 0. By re-arranging the code a little, I've made the
6251 decoding happen first, so such errors now appear on stderr, and the return
6252 code is 1. However, the actual setting up of the sockets still happens in
6253 the disconnected process, so errors there are still only recorded on the
6256 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6257 connections (as happens on some IP stacks) was logged at start up time as
6258 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6259 it from "IPv6 and IPv4", which means that two separate sockets are being
6262 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6263 says whether AF_INET or AF_INET6 was passed as an argument.
6265 42. Exiwhat output was messed up when time zones were included in log
6268 43. Exiwhat now gives more information about the daemon's listening ports,
6269 and whether -tls-on-connect was used.
6271 44. The "port" option of the smtp transport is now expanded.
6273 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6274 silently ignored. Now an error message is written to the main and panic
6277 46. There's a new ACL modifier called "logwrite" which writes to a log file
6278 as soon as it is encountered.
6280 47. Added $local_user_uid and $local_user_gid at routing time.
6282 48. Exim crashed when trying to verify a sender address that was being
6285 49. Exim was recognizing only a space character after ".include". It now also
6286 recognizes a tab character.
6288 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6289 extracting the relevant information from the specification. The man page no
6290 longer contains scrambled data for the -d option, and I've added a section
6291 at the front about calling Exim under different names.
6293 51. Added "extra_headers" argument to the "mail" command in filter files.
6295 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6298 53. Installed eximstats 1.29.
6300 54. Added transport_filter_timeout as a generic transport option.
6302 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6303 Cc: header lines. This was required by RFC 822, but it not required by RFC
6306 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6307 incoming messages that did not have them. Now it does so only if the
6308 message originates locally, that is, if there is no associated remote host
6309 address. When Resent- header lines are present, this applies to the Resent-
6310 lines rather than the non-Resent- lines.
6312 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6313 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6315 58. Messages for configuration errors now include the name of the main
6316 configuration file - useful now that there may be more than one file in a
6317 list (.included file names were always shown).
6319 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6320 for those rare installations that do not start the daemon as root or run it
6321 setuid root. I've cut out the call to initgroups() if the daemon is not
6324 60. The Exim user and group can now be bound into the binary as text strings
6325 that are looked up at the start of Exim's processing.
6327 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6329 62. Added $mailstore_basename variable.
6331 63. Installed patch to sieve.c from Michael Haardt.
6333 64. When Exim failed to open the panic log after failing to open the main log,
6334 the original message it was trying to log was written to stderr and debug
6335 output, but if they were not available (the usual case in production), it
6336 was lost. Now it is written to syslog before the two lines that record the
6337 failures to open the logs.
6339 65. Users' Exim filters run in subprocesses under the user's uid. It is
6340 possible for a "deliver" command or an alias in a "personal" command to
6341 provoke an address rewrite. If logging of address rewriting is configured,
6342 this fails because the process is not running as root or exim. There may be
6343 a better way of dealing with this, but for the moment (because 4.30 needs
6344 to be released), I have disabled address rewrite logging when running a
6345 filter in a non-root, non-exim process.
6351 1. The buildconfig auxiliary program wasn't quoting the value set for
6352 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6353 not defined. This bug was masked in 4.22 by the effect that was fixed in
6356 2. Some messages that were rejected after a message id was allocated were
6357 shown as "incomplete" by exigrep. It no longer does this for messages that
6358 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6360 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6361 have allow_domain_literals set, the ID did not get logged in the <= line.
6362 Domain literals are now always recognized in Message-ID: header lines.
6364 4. The first argument for a ${extract expansion item is the key name or field
6365 number. Leading and trailing spaces in this item were not being ignored,
6366 causing some misleading effects.
6368 5. When deliver_drop_privilege was set, single queue runner processes started
6369 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6370 same command in the process it spins off) were not dropping privilege.
6372 6. When the daemon running as "exim" started a queue runner, it always
6373 re-executed Exim in the spun-off process. This is a waste of effort when
6374 deliver_drop_privilege is set. The new process now just calls the
6375 queue-runner function directly.
6381 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6384 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6385 ignored. Though the use of -oP was forcing the writing of a pid file, it
6386 was always written to the default place.
6388 3. If the message "no IP address found for host xxxx" is generated during
6389 incoming verification, it is now followed by identification of the incoming
6390 connection (so you can more easily find what provoked it).
6392 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6394 5. Added some features to "harden" Exim a bit more against certain attacks:
6396 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6397 be put in Local/Makefile. This is like the never_users runtime option,
6398 but it cannot be overridden. The default setting is "root".
6400 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6401 prefix string with which any file named in a -C command line option
6404 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6405 is retained for -C and -D only if the caller of Exim is root. Without
6406 it, the exim user may also use -C and -D and retain privilege.
6408 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6409 command line option is disabled.
6411 6. Macro names set by the -D option must start with an upper case letter, just
6412 like macro names defined in the configuration file.
6414 7. Added "dereference=" facility to LDAP.
6416 8. Two instances of the typo "uknown" in the source files are fixed.
6418 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6419 the Configure-Makefile script screwed up while processing it.
6421 10. Incorporated PCRE 4.4.
6423 11. The SMTP synchronization check was not operating right at the start of an
6424 SMTP session. For example, it could not catch a HELO sent before the client
6425 waited for the greeting. There is now a check for outstanding input at the
6426 point when the greeting is written. Because of the duplex, asynchronous
6427 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6428 way, but not yet received, when the check is performed.
6430 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6431 on TCP/IP sockets, because this apparently causes some broken clients to
6434 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6435 unchanged) from the Cygwin maintainer.
6437 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6438 was supported instead of testing for mbx. Effectively a typo.
6440 15. The spa authenticator server code was not checking that the input it
6441 received was valid base64.
6443 16. The debug output line for the "set" modifier in ACLs was not showing the
6444 name of the variable that was being set.
6446 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6448 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6449 lookup on its own. Something else was needed to trigger the lookup. For
6450 example, a match in host_lookup or the need for a host name in a host list.
6451 Now, if $sender_host_name is referenced and the host name has not yet been
6452 looked up, a lookup is performed. If the lookup fails, the variable remains
6453 empty, and $host_lookup_failed is set to "1".
6455 19. Added "eqi" as a case-independent comparison operator.
6457 20. The saslauthd authentication condition could segfault if neither service
6458 nor realm was specified.
6460 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6461 error message that was logged was misleading, and incoming SMTP
6462 connections were dropped. The message is now more accurate, and temporary
6463 errors are given to SMTP connections.
6465 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6466 (except RSET) with a 421 error, until QUIT is received. However, it was
6467 failing to send a response to QUIT.
6469 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6470 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6471 ACL is now run after the helo_try_verify_hosts code.
6473 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6474 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6475 case-independent, so other case variants are also recognized). Apparently
6476 some systems use these upper case variants.
6478 25. If more than two messages were waiting for the same host, and a transport
6479 filter was specified for the transport, Exim sent two messages over the
6480 same TCP/IP connection, and then failed with "socket operation on non-
6481 socket" when it tried to send the third.
6483 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6485 27. The extern definition of crypt16() in expand.c was not being excluded when
6486 the OS had its own crypt16() function.
6488 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6489 as a preferred synonym for return_size_limit, both as an option and as an
6492 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6494 30. Changed the default configuration ACL to relax the local part checking rule
6495 for addresses that are not in any local domains. For these addresses,
6496 slashes and pipe symbols are allowed within local parts, but the sequence
6497 /../ is explicitly forbidden.
6499 31. SPA server authentication was not clearing the challenge buffer before
6502 32. log_message in a "warn" ACL statement was writing to the reject log as
6503 well as to the main log, which contradicts the documentation and doesn't
6504 seem right (because no rejection is happening). So I have stopped it.
6506 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6507 However, I am unable to do any testing of this.
6509 34. Fixed an infelicity in the appendfile transport. When checking directories
6510 for a mailbox, to see if any needed to be created, it was accidentally
6511 using path names with one or more superfluous leading slashes; tracing
6512 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6514 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6515 contents are added to the log line that is written for every discarded
6516 recipient. (Previously a log_message setting was ignored.)
6518 36. The ${quote: operator now quotes the string if it is empty.
6520 37. The install script runs exim in order to find its version number. If for
6521 some reason other than non-existence or emptiness, which it checks, it
6522 could not run './exim', it was installing it with an empty version number,
6523 i.e. as "exim-". This error state is now caught, and the installation is
6526 38. An argument was missing from the function that creates an error message
6527 when Exim fails to connect to the socket for saslauthd authentication.
6528 This could cause Exim to crash, or give a corrupted message.
6530 39. Added isip, isip4, and isip6 to ${if conditions.
6532 40. The ACL variables $acl_xx are now saved with the message, and can be
6533 accessed later in routers, transports, and filters.
6535 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6536 strings in the file are not string-expanded.
6538 42. If a MAIL command specified a SIZE value that was too large to fit into an
6539 int variable, the check against message_size_limit failed. Such values are
6540 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6541 day this will have to be increased, but I don't think I want to be around
6542 when emails are that large.
6549 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6550 iconv() is not standard in FreeBSD.
6552 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6553 IPv6 enabled. The observed symptom was a segmentation fault on return from
6554 the function os_common_find_running_interfaces() in src/os.c.
6556 3. In the check_special_case() function in daemon.c I had used "errno" as an
6557 argument name, which causes warnings on some systems. This was basically a
6558 typo, since it was named "eno" in the comments!
6560 4. The code that waits for the clock to tick (at a resolution of some fraction
6561 of a second) so as to ensure message-id uniqueness was always waiting for
6562 at least one whole tick, when it could have waited for less. [This is
6563 almost certainly not relevant at current processor speeds, where it is
6564 unlikely to ever wait at all. But we try to future-proof.]
6566 5. The function that sleeps for a time interval that includes fractions of a
6567 second contained a race. It did not block SIGALRM between setting the
6568 timer, and suspending (a couple of lines later). If the interval was short
6569 and the sigsuspend() was delayed until after it had expired, the suspension
6570 never ended. On busy systems this could lead to processes getting stuck for
6573 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6574 process, before it forks any delivery processes. The open lookup caching
6575 mechanism meant that the open file or database connection was passed into
6576 the delivery process. The problem was that delivery processes always tidy
6577 up cached lookup data. This could cause a problem for the next delivery
6578 process started by the queue runner, because the external queue runner
6579 process does not know about the closure. So the next delivery process
6580 still has data in the lookup cache. In the case of a file lookup, there was
6581 no problem because closing a file descriptor in a subprocess doesn't affect
6582 the parent. However, if the lookup was caching a connection to a database,
6583 the connection was closed, and the second delivery process was likely to
6584 see errors such as "PGSQL: query failed: server closed the connection
6585 unexpectedly". The problem has been fixed by closing all cached lookups
6586 in a queue runner before running a delivery process.
6588 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6589 seem to have the "const" qualifier which it has on other OS. I've
6592 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6593 *of the same type* (client or server) with the same public name that an
6594 error should be diagnosed.
6596 9. When Exim looked up a host name for an IP address, but failed to find the
6597 original IP address when looking up the host name (a safety check), it
6598 output the message "<ip address> does not match any IP for NULL", which was
6599 confusing, to say the least. The bug was that the host name should have
6600 appeared instead of "NULL".
6602 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6603 user that is built into the binary, and the -C or -D options is used, root
6604 privilege is dropped before the configuration file is read. In addition,
6605 logging is switched to stderr instead of the normal log files. If the
6606 configuration then re-defines the Exim user, the unprivileged environment
6607 is probably not what is expected, so Exim logs a panic warning message (but
6610 However, if deliver_drop_privilege is set, the unprivileged state may well
6611 be exactly what is intended, so the warning has been cut out in that case,
6612 and Exim is allowed to try to write to its normal log files.
6618 1. smtp_return_error_details was not giving details for temporary sender
6619 or receiver verification errors.
6621 2. Diagnose a configuration error if two authenticators have the same public
6624 3. Exim used not to create the message log file for a message until the first
6625 delivery attempt. This could be confusing when incoming messages were held
6626 for policy or load reasons. The message log file is now created at the time
6627 the message is received, and an initial "Received" line is written to it.
6629 4. The automatically generated man page for command line options had a minor
6630 bug that caused no ill effects; however, a more serious problem was that
6631 the procedure for building the man page automatically didn't always
6632 operate. Consequently, release 4.20 contains an out-of-date version. This
6633 shouldn't happen again.
6635 5. When building Exim with embedded Perl support, the script that builds the
6636 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6637 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6639 6. The freeze_tell option was not being used for messages that were frozen on
6640 arrival, either by an ACL or by local_scan().
6642 7. Added the smtp_incomplete_transaction log selector.
6644 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6645 was accepting AUTH without a new EHLO.
6647 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6648 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6651 10. Logging of TCP/IP connections (when configured) now happens in the main
6652 daemon process instead of the child process, so that the TCP/IP connection
6653 count is more accurate (but it can never be perfect).
6655 11. The use of "drop" in a nested ACL was not being handled correctly in the
6656 outer ACL. Now, if condition failure induced by the nested "drop" causes
6657 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6658 or "require"), the connection is dropped.
6660 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6661 that yield "discard" can now be used with an "accept" or a "discard" verb,
6662 but an error is generated for any others (because I can't see a useful way
6663 to define what should happen).
6665 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6666 are now processed in the same way as lines in the Exim configuration file.
6667 In particular, continuation lines are supported.
6669 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6671 15. Added -ti meaning -t -i.
6673 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6674 domains, and warn by logging if others are found.
6676 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6677 returned from GIFCONF when Exim is trying to find the list of interfaces on
6678 a host. The code in os.c has been modified to copy each ifreq to an aligned
6679 structure in all cases.
6681 Also, in some cases, the returned ifreq's were being copied to a 'struct
6682 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6683 means the last couple of bytes of an IPv6 address could be chopped if the
6684 ifreq contained only a normal sockaddr (14 bytes storage).
6686 18. Named domain lists were not supported in the hosts_treat_as_local option.
6687 An entry such as +xxxx was not recognized, and was treated as a literal
6690 19. Ensure that header lines added by a DATA ACL are included in the reject log
6691 if the ACL subsequently rejects the message.
6693 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6694 MD5 (which is deprecated).
6696 21. When testing a filter file using -bf, Exim was writing a message when it
6697 took the sender from a "From " line in the message, but it was not doing so
6698 when it took $return_path from a Return-Path: header line. It now does.
6700 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6701 with a valid header line field name (a series of printing characters
6702 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6704 23. Changed "disc" in the source to "disk" to conform to the documentation and
6705 the book and for uniformity.
6707 24. Ignore Sendmail's -Ooption=value command line item.
6709 25. When execve() failed while trying to run a command in a pipe transport,
6710 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6711 could be confused with a return value of 69 from the command itself. This
6712 has been changed to 127, the value the shell returns if it is asked to run
6713 a non-existent command. The wording for the related log line suggests a
6714 non-existent command as the problem.
6716 26. If received_header_text expands to an empty string, do not add a Received:
6717 header line to the message. (Well, it adds a token one on the spool, but
6718 marks it "old" so that it doesn't get used or transmitted.)
6720 27. Installed eximstats 1.28 (addition of -nt option).
6722 28. There was no check for failure on the call to getsockname() in the daemon
6723 code. This can fail if there is a shortage of resources on the system, with
6724 ENOMEM, for example. A temporary error is now given on failure.
6726 29. Contrary to the C standard, it seems that in some environments, the
6727 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6728 program. Exim now does this explicitly; it affects the formatting of
6729 timestamps using strftime().
6731 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6732 complaints. I've now initialized all the variables, to avoid this.
6734 32. Header lines added by a system filter were not being "seen" during
6735 transport-time rewrites.
6737 33. The info_callback() function passed to OpenSSL is set up with type void
6738 (*)(SSL *, int, int), as described somewhere. However, when calling the
6739 function (actually a macro) that sets it up, the type void(*)() is
6740 expected. I've put in a cast to prevent warnings from picky compilers.
6742 34. If a DNS black list lookup found a CNAME record, but there were no A
6743 records associated with the domain it pointed at, Exim crashed.
6745 35. If a DNS black list lookup returned more than one A record, Exim ignored
6746 all but the first. It now scans all returned addresses if a particular IP
6747 value is being sought. In this situation, the contents of the
6748 $dnslist_value variable are a list of all the addresses, separated by a
6751 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6752 to accept a host name and all its aliases if the forward lookup for any of
6753 them yielded the IP address of the incoming connection. Now it accepts only
6754 those names whose forward lookup yields the correct IP address. Any other
6755 names are discarded. This closes a loophole whereby a rogue DNS
6756 administrator could create reverse DNS records to break through a
6757 wildcarded host restriction in an ACL.
6759 37. If a user filter or a system filter that ran in a subprocess used any of
6760 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6761 the wrong values were passed to the pipe command ($thisaddress had the
6762 value of $0, $0 had the value of $1, etc). This bug was introduced by
6763 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6765 38. Improved the line breaking for long SMTP error messages from ACLs.
6766 Previously, if there was no break point between 40 and 75 characters, Exim
6767 left the rest of the message alone. Two changes have been made: (a) I've
6768 reduced the minimum length to 35 characters; (b) if it can't find a break
6769 point between 35 and 75 characters, it looks ahead and uses the first one
6770 that it finds. This may give the occasional overlong line, but at least the
6771 remaining text gets split now.
6773 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6774 file descriptors might be low, and that setting 1000 would always raise it.
6775 It turns out that in some environments, the limit is already over 1000 and
6776 that lowering it causes trouble. So now Exim takes care not to decrease it.
6778 40. When delivering a message, the value of $return_path is set to $sender_
6779 address at the start of routing (routers may change the value). By an
6780 oversight, this default was not being set up when an address was tested by
6781 -bt or -bv, which affected the outcome if any router or filter referred to
6784 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6785 log only when "message" or "log_message" are set. However, if one of the
6786 conditions was an address verification, or a call to a nested ACL, the
6787 messages generated by the underlying test were being passed through. This
6788 no longer happens. The underlying message is available in $acl_verify_
6789 message for both "message" and "log_message" expansions, so it can be
6790 passed through if needed.
6792 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6793 new expansion $bh_ to give the encoded byte string without charset
6794 translation. Translation happens only if iconv() is available; HAVE_ICONV
6795 indicates this at build time. HEADERS_CHARSET gives the charset to
6796 translate to; headers_charset can change it in the configuration, and
6797 "headers charset" can change it in an individual filter file.
6799 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6800 that creates RFC 2047 encoded "words" labels them as that charset instead
6801 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6802 expansion operator; (ii) when Exim creates a From: line for a local
6803 message; (iii) when a header line is rewritten to include a "phrase" part.
6805 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6806 buggy, causing it to skip the first lines of messages whose message ID
6807 ended in 'D'. This would not have bitten before Exim release 4.14, because
6808 message IDs were unlikely to end in 'D' before then. The effect was to have
6809 incorrect size information for certain domains.
6811 45. #include "config.h" was missing at the start of the crypt16.c module. This
6812 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6815 46. If there was a timeout during a "random" callout check, Exim treated it as
6816 a failure of the random address, and carried on sending RSET and the real
6817 address. If the delay was just some slowness somewhere, the response to the
6818 original RCPT would be taken as a response to RSET and so on, causing
6819 mayhem of various kinds.
6821 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6822 when I implemented it. It didn't allow for the fact that some option values
6823 may legitimately be negative (e.g. size_addition), and it didn't even do
6824 the right test for positive values.
6826 48. Domain names in DNS records are case-independent. Exim always looks them up
6827 in lower case. Some resolvers return domain names in exactly the case they
6828 appear in the zone file, that is, they may contain uppercase letters. Not
6829 all resolvers do this - some return always lower case. Exim was treating a
6830 change of case by a resolver as a change of domain, similar to a widening
6831 of a domain abbreviation. This triggered its re-routing code and so it was
6832 trying to route what was effectively the same domain again. This normally
6833 caused routing to fail (because the router wouldn't handle the domain
6834 twice). Now Exim checks for this case specially, and just changes the
6835 casing of the domain that it ultimately uses when it transmits the message
6838 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6841 50. If a filter generated a file delivery with a non-absolute name (possible if
6842 no home directory exists for the router), the forbid_file option was not
6845 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6846 the existing equality matching.
6848 52. Exim was using ints instead of ino_t variables in some places where it was
6849 dealing with inode numbers.
6851 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6852 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6853 called TMPDIR, and if it finds it is different, it changes its value.
6855 54. The smtp_printf() function is now made available to local_scan() so
6856 additional output lines can be written before returning. There is also an
6857 smtp_fflush() function to enable the detection of a dropped connection.
6858 The variables smtp_input and smtp_batched_input are exported to
6861 55. Changed the default runtime configuration: the message "Unknown user"
6862 has been removed from the ACL, and instead placed on the localuser router,
6863 using the cannot_route_message feature. This means that any verification
6864 failures that generate their own messages won't get overridden. Similarly,
6865 the "Unrouteable address" message that was in the ACL for unverifiable
6866 relay addresses has also been removed.
6868 56. Added hosts_avoid_esmtp to the smtp transport.
6870 57. The exicyclog script was not checking for the esoteric option
6871 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6872 will work only if exicyclog is run under the appropriate euid.
6874 58. Following a discussion on the list, the rules by which Exim recognises line
6875 endings on incoming messages have been changed. The -dropcr and drop_cr
6876 options are now no-ops, retained only for backwards compatibility. The
6877 following line terminators are recognized: LF CRLF CR. However, special
6878 processing applies to CR:
6880 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6881 nor a local message in the state where . is a terminator.
6883 (ii) If a bare CR is encountered in a header line, an extra space is added
6884 after the line terminator so as not to end the header. The reasoning
6885 behind this is that bare CRs in header lines are most likely either
6886 to be mistakes, or people trying to play silly games.
6888 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6889 was being incorrectly given as 18 bytes larger than it should have been.
6890 This is a VOB (very old bug).
6892 60. This may never have affected anything current, but just in case it has:
6893 When the local host is found other than at the start of a list of hosts,
6894 the local host, those with the same MX, and any that follow, are discarded.
6895 When the list in question was part of a longer list of hosts, the following
6896 hosts (not currently being processed) were also being discarded. This no
6897 longer happens. I'm not sure if this situation could ever has previously
6900 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6903 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6904 header. This is something that is recommended in a new Internet Draft, and
6905 is something that is documented as being done by Sendmail. There are two
6906 possible values. For messages generated by the autoreply transport, Exim
6909 Auto-Submitted: auto-replied
6911 whereas for all other generated messages (e.g. bounces) it adds
6913 Auto-Submitted: auto-generated
6915 63. The "personal" condition in filters now includes a test for the
6916 Auto-Submitted: header. If it contains the string "auto-" the message it
6917 not considered personal.
6919 64. Added rcpt_include_affixes as a generic transport option.
6921 65. Added queue_only_override (default true).
6923 66. Added the syslog_duplication option.
6925 67. If what should have been the first header line of a message consisted of
6926 a space followed by a colon, Exim was mis-interpreting it as a header line.
6927 It isn't of course - it is syntactically invalid and should therefore be
6928 treated as the start of the message body. The misbehaviour could have
6929 caused a number of strange effects, including loss of data in subsequent
6930 header lines, and spool format errors.
6932 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6933 client host had authenticated. This control can now be exercised by an ACL
6934 for more flexibility.
6936 69. By default, callouts do not happen when testing with -bh. There is now a
6937 variant, -bhc, which does actually run the callout code, including
6938 consulting and updating the callout cache.
6940 70. Added support for saslauthd authentication, courtesy of Alexander
6943 71. If statvfs() failed on the spool or log directories while checking their
6944 size for availability, Exim confusingly gave the error "space shortage".
6945 Furthermore, in debugging mode it crashed with a floating point exception.
6946 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6947 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6948 serious problem, Exim now writes to the main and panic logs when this
6949 happens, with details of the failure. It then refuses to accept the
6950 incoming message, giving the message "spool directory problem" or "log
6951 directory problem" with a 421 code for SMTP messages.
6953 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6954 0, 1, and 2 exist, because some OS complain for execs without them (see
6955 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6956 descriptors. However, the code omitted to check that the open succeeded,
6957 causing mysterious errors if for some reason the permissions on /dev/null
6958 got screwed. Now Exim writes a message to the main and panic logs, and
6959 bombs out if it can't open /dev/null.
6961 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6962 interact so that it is all more flexible. It is supposed to remain
6963 backwards compatible. Also added extra_local_interfaces.
6965 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6966 to bomb out with an assertion failure - to the client this appears as a
6967 connection drop. This problem occurs in the part of the code that was taken
6968 from the Samba project. Fortunately, the assertion is in a very simple
6969 function, so I have fixed this by reproducing the function inline in the
6970 one place where it is called, and arranging for authentication to fail
6971 instead of killing the process with assert().
6973 75. The SPA client code was not working when the server requested OEM rather
6974 than Unicode encoding.
6976 76. Added code to make require_files with a specific uid setting more usable in
6977 the case where statting the file as root fails - usually a non-root-mounted
6978 NFS file system. When this happens and the failure is EACCES, Exim now
6979 forks a subprocess and does the per-uid checking as the relevant uid.
6981 77. Added process_log_path.
6983 78. If log_file_path was not explicitly set, a setting of check_log_space or
6984 check_log_inodes was ignored.
6986 79. If a space check for the spool or log partitions fails, the incident is now
6987 logged. Of course, in the latter case the data may get lost...
6989 80. Added the %p formatting code to string_format() so that it can be used to
6990 print addresses in debug_print(). Adjusted all the address printing in the
6991 debugging in store.c to use %p rather than %d.
6993 81. There was a concern that a line of code in smtp_in.c could overflow a
6994 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6995 initially expressed, the concern was not well-founded, because trailing
6996 spaces are removed early. However, if the trailing spaces were followed by
6997 a NULL, they did not get removed, so the overflow was possible. Two fixes
7000 (a) I re-wrote the offending code in a cleaner fashion.
7001 (b) If an incoming SMTP command contains a NULL character, it is rejected
7004 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7005 runs initgroups(), so that if the Exim user is in any additional groups,
7006 they will be used during message reception.
7012 The change log for 4.20 and earlier releases has been archived.