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.
8 Since Exim version 4.90
9 -----------------------
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.
54 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
55 by the template makefile "src/EDITME". The "STREAM" support for an older
56 ClamAV interface method is removed.
58 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
59 rows affected is given instead).
61 JH/11 The runtime Berkeley DB library version is now additionally output by
62 "exim -d -bV". Previously only the compile-time version was shown.
64 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
65 SMTP connection. Previously, when one had more receipients than the
66 first, an abortive onward connection was made. Move to full support for
67 multiple onward connections in sequence, handling cutthrough connection
68 for all multi-message initiating connections.
70 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
71 routers. Previously, a multi-recipient message would fail to match the
72 onward-connection opened for the first recipient, and cause its closure.
74 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
75 a timeout on read on a GnuTLS initiating connection, resulting in the
76 initiating connection being dropped. This mattered most when the callout
77 was marked defer_ok. Fix to keep the two timeout-detection methods
80 JH/15 Relax results from ACL control request to enable cutthrough, in
81 unsupported situations, from error to silently (except under debug)
82 ignoring. This covers use with PRDR, frozen messages, queue-only and
85 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
87 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
88 metadata, resulting in a crash in free().
90 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
91 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
92 Broken also in d185889f4, with init system revamp.
94 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
95 Previously we abruptly closed the connection after reading a malware-
96 found indication; now we go on to read the "scan ok" response line,
99 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
100 ACL. Previously, a crash would result.
102 JH/19 Speed up macro lookups during configuration file read, by skipping non-
103 macro text after a replacement (previously it was only once per line) and
104 by skipping builtin macros when searching for an uppercase lead character.
106 JH/20 DANE support moved from Experimental to mainline. The Makefile control
107 for the build is renamed.
109 JH/21 Fix memory leak during multi-message connections using STARTTLS. A buffer
110 was allocated for every new TLS startup, meaning one per message. Fix
111 by only allocating once (OpenSSL) or freeing on TLS-close (GnuTLS).
117 JH/01 Rework error string handling in TLS interface so that the caller in
118 more cases is responsible for logging. This permits library-sourced
119 string to be attached to addresses during delivery, and collapses
120 pairs of long lines into single ones.
122 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
123 during configuration. Wildcards are allowed and expanded.
125 JH/02 Rework error string handling in DKIM to pass more info back to callers.
126 This permits better logging.
128 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
129 do not close it down and have the child transport start it up again on
130 the passed-on TCP connection. Instead, proxy the child (and any
131 subsequent ones) for TLS via a unix-domain socket channel. Logging is
132 affected: the continued delivery log lines do not have any DNSSEC, TLS
133 Certificate or OCSP information. TLS cipher information is still logged.
135 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
136 identical IP addresses on different listening ports. Will also affect
139 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
140 add noisy ifdef guards to special-case this sillyness.
141 Patch from Bernd Kuhls.
143 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
144 than 255 are no longer allowed.
146 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
147 Disable the session-cache too, which might reduce our load. Since we
148 currrectly use a new context for every connection, both as server and
149 client, there is no benefit for these.
150 GnuTLS appears to not support tickets server-side by default (we don't
151 call gnutls_session_ticket_enable_server()) but client side is enabled
152 by default on recent versions (3.1.3 +) unless the PFS priority string
155 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
156 <https://reproducible-builds.org/specs/source-date-epoch/>.
158 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
159 the check for any unsuccessful recipients did not notice the limit, and
160 erroneously found still-pending ones.
162 JH/08 Pipeline CHUNKING command and data together, on kernels that support
163 MSG_MORE. Only in-clear (not on TLS connections).
165 JH/09 Avoid using a temporary file during transport using dkim. Unless a
166 transport-filter is involved we can buffer the headers in memory for
167 creating the signature, and read the spool data file once for the
168 signature and again for transmission.
170 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
171 4.77 as the kernel support then wasn't solid, having issues in 64bit
172 mode. Now, it's been long enough. Add support for FreeBSD also.
174 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
175 case where the routing stage had gathered several addresses to send to
176 a host before calling the transport for the first, we previously failed
177 to close down TLS in the old transport process before passing the TCP
178 connection to the new process. The new one sent a STARTTLS command
179 which naturally failed, giving a failed delivery and bloating the retry
180 database. Investigation and fix prototype from Wolfgang Breyha.
182 JH/12 Fix check on SMTP command input synchronisation. Previously there were
183 false-negatives in the check that the sender had not preempted a response
184 or prompt from Exim (running as a server), due to that code's lack of
185 awareness of the SMTP input buffering.
187 PP/04 Add commandline_checks_require_admin option.
188 Exim drops privileges sanely, various checks such as -be aren't a
189 security problem, as long as you trust local users with access to their
190 own account. When invoked by services which pass untrusted data to
191 Exim, this might be an issue. Set this option in main configuration
192 AND make fixes to the calling application, such as using `--` to stop
195 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
196 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
197 responses to those, into a single TLS record each way (this usually means
198 a single packet). As a side issue, smtp_enforce_sync now works on TLS
201 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
202 affects you only if you're dancing at the edge of the param size limits.
203 If you are, and this message makes sense to you, then: raise the
204 configured limit or use OpenSSL 1.1. Nothing we can do for older
207 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
208 cmdline element to get the documented default one. Previously it was
211 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
212 are unsuable for non-message-specific reasons - eg. connection timeout,
213 banner-time rejection.
215 JH/16 Fix logging of delivery remote port, when specified by router, under
218 PP/06 Repair manualroute's ability to take options in any order, even if one
219 is the name of a transport.
222 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
224 JH/17 Change the list-building routines interface to use the expanding-string
225 triplet model, for better allocation and copying behaviour.
227 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
228 Previously it was constructed the first time a possibly-matching string
229 was met in the configuration file input during startup; now it is done
232 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
233 compatible one, to avoid the (poorly documented) possibility of a config
234 file in the working directory redirecting the DB files, possibly correpting
235 some existing file. CVE-2017-10140 assigned for BDB.
237 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
238 cache-hot. Previously, although the result was properly cached, the
239 initial verify call returned a defer.
241 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
242 the main verify for receipient in uncached-mode.
244 JH/22 Retire historical build files to an "unsupported" subdir. These are
245 defined as "ones for which we have no current evidence of testing".
247 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
248 if present. Previously it was ignored.
250 JH/24 Start using specified-initialisers in C structure init coding. This is
251 a C99 feature (it's 2017, so now considered safe).
253 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
254 if was a fixed-sized field and bitmask ops via macros; it is now more
257 PP/07 GitHub PR 56: Apply MariaDB build fix.
258 Patch provided by Jaroslav Škarvada.
260 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
261 during Coverity cleanups [4.87 JH/47]
262 Diagnosis and fix provided by Michael Fischer v. Mollard.
264 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
265 the right size to place the terminating semicolon on its own folded
266 line, the header hash was calculated to an incorrect value thanks to
267 the (relaxed) space the fold became.
269 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
270 and confused the parent.
272 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
273 which could crash as a result. This could lead to undeliverable messages.
275 JH/28 Logging: "next input sent too soon" now shows where input was truncated
278 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
279 matters on fast-turnover and PID-randomising systems, which were getting
280 out-of-order delivery.
282 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
283 a possibly-overlapping copy. The symptom was that "Remote host closed
284 connection in response to HELO" was logged instead of the actual 4xx
287 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
288 Previously only that bufferd was discarded, resulting in SYMTP command
291 JH/32 DKIM: when a message has multiple signatures matching an identity given
292 in dkim_verify_signers, run the dkim acl once for each. Previously only
293 one run was done. Bug 2189.
295 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
296 "panic the current process" to "deliberately defer". The panic log is
297 still written with the problem list name; the mail and reject logs now
298 get a temp-reject line for the message that was being handled, saying
299 something like "domains check lookup or other defer". The SMTP 451
300 message is still "Temporary local problem".
302 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
303 A crafted sequence of BDAT commands could result in in-use memory beeing
304 freed. CVE-2017-16943.
306 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
307 from SMTP input. Previously it was always done; now only done for DATA
308 and not BDAT commands. CVE-2017-16944.
310 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
311 to the message (such as an overlong header line). Previously this was
312 not done and we did not exit BDAT mode. Followon from the previous item
313 though a different problem.
319 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
320 than -2003 did; needs libidn2 in addition to libidn.
322 JH/02 The path option on a pipe transport is now expanded before use.
324 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
325 Patch provided by "Björn", documentation fix added too.
327 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
328 missing a wire-to-host endian conversion.
330 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
331 close after a BDAT command line could be taken as a following command,
332 giving a synch failure. Fix by only checking for synch immediately
333 before acknowledging the chunk.
335 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
336 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
337 macro. Patches provided by Josh Soref.
339 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
340 Previously we did not; the RFC seems ambiguous and VRFY is not listed
341 by IANA as a service extension. However, John Klensin suggests that we
344 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
345 the dkim code may be unix-mode line endings rather than smtp wire-format
346 CRLF, so prepend a CR to any bare LF.
348 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
349 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
351 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
352 the first were themselves being wrongly included in the feed into dkim
353 processing; with most chunk sizes in use this resulted in an incorrect
354 body hash calculated value.
356 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
357 DKIM signature block, for verification. Although advised against by
358 standards it is specifically not ruled illegal.
360 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
362 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
363 missing a body hash (the bh= tag).
365 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
366 It seems that HAProxy sends the Proxy Protocol information in clear and
367 only then does a TLS startup, so do the same.
369 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
370 TCP connections (such as for Spamd) unless the daemon successfully set
371 Fast Open mode on its listening sockets. This fixes breakage seen on
372 too-old kernels or those not configured for Fast Open, at the cost of
373 requiring both directions being enabled for TFO, and TFO never being used
374 by non-daemon-related Exim processes.
376 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
377 endings, at least on the first header line. Try to canonify any that get
378 past that check, despite the cost.
380 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
381 now limited to an arbitrary five deep, while parsing addresses with the
382 strip_excess_angle_brackets option enabled.
384 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
385 instead leave the unprompted TLS handshake in socket buffer for the
386 TLS library to consume.
388 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
390 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
392 JH/16 Drop variables when they go out of scope. Memory management drops a whole
393 region in one operation, for speed, and this leaves assigned pointers
394 dangling. Add checks run only under the testsuite which checks all
395 variables at a store-reset and panics on a dangling pointer; add code
396 explicitly nulling out all the variables discovered. Fixes one known
397 bug: a transport crash, where a dangling pointer for $sending_ip_address
398 originally assigned in a verify callout, is re-used.
400 PP/06 Drop '.' from @INC in various Perl scripts.
402 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
404 PP/08 Reduce a number of compilation warnings under clang; building with
405 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
406 should be warning-free.
408 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
410 HS/01 Fix portability problems introduced by PP/08 for platforms where
411 realloc(NULL) is not equivalent to malloc() [SunOS et al].
413 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
414 chunk. This allows us to accept broken chunked messages. We need a more
415 general solution here.
417 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
418 already-broken messages in the queue.
420 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
422 JH/19 Fix reference counting bug in routing-generated-address tracking.
428 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
429 supports it and a size is available (ie. the sending peer gave us one).
431 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
432 years of being deprecated). The replacements are the ACLs
433 acl_smtp_mime and acl_not_smtp_mime.
435 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
436 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
437 or even an in-clear connection were permitted. Now, if the host lookup
438 was dnssec and dane was requested then the host is only used if the
439 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
440 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
441 if one fails this test.
442 This means that a poorly-configured remote DNS will make it incommunicado;
443 but it protects against a DNS-interception attack on it.
445 JH/04 Bug 1810: make continued-use of an open smtp transport connection
446 non-noisy when a race steals the message being considered.
448 JH/05 If main configuration option tls_certificate is unset, generate a
449 self-signed certificate for inbound TLS connections.
451 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
452 in rewrites and routers.
454 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
455 and logged a warning sing 4.83; now they are a configuration file error.
457 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
458 (lacking @domain). Apply the same qualification processing as RCPT.
460 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
462 JH/10 Support ${sha256:} applied to a string (as well as the previous
465 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
466 a cutthrough deliver is pending, as we always want to make a connection.
467 This also avoids re-routing the message when later placing the cutthrough
468 connection after a verify cache hit.
469 Do not update it with the verify result either.
471 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
472 when routing results in more than one destination address.
474 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
475 signing (which inhibits the cutthrough capability). Previously only
476 the presence of an option was tested; now an expansion evaluating as
477 empty is permissible (obviously it should depend only on data available
478 when the cutthrough connection is made).
480 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
481 the relevant preceding SMTP command did not note the pipelining mode.
483 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
484 Previously they were not counted.
486 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
487 as one having no matching records. Previously we deferred the message
488 that needed the lookup.
490 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
491 distinguished as "(=".
493 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
494 for missing MX records. Previously it only worked for missing A records.
496 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
498 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
499 after the data-go-ahead and data-ack. Patch from Jason Betts.
501 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
502 even for a "none" policy. Patch from Tony Meyer.
504 JH/22 Fix continued use of a connection for further deliveries. If a port was
505 specified by a router, it must also match for the delivery to be
508 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
509 When one of the recipients of a message was unsuitable for the connection
510 (has no matching addresses), we lost track of needing to mark it
511 deferred. As a result mail would be lost.
513 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
515 JH/25 Decoding ACL controls is now done using a binary search; the source code
516 takes up less space and should be simpler to maintain. Merge the ACL
517 condition decode tables also, with similar effect.
519 JH/26 Fix problem with one_time used on a redirect router which returned the
520 parent address unchanged. A retry would see the parent address marked as
521 delivered, so not attempt the (identical) child. As a result mail would
524 JH/27 Fix a possible security hole, wherein a process operating with the Exim
525 UID can gain a root shell. Credit to http://www.halfdog.net/ for
526 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
529 JH/28 Enable {spool,log} filesystem space and inode checks as default.
530 Main config options check_{log,spool}_{inodes,space} are now
531 100 inodes, 10MB unless set otherwise in the configuration.
533 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
534 Previously it only applied to the main-section connection policy
537 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
539 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
540 by me. Added RFC7919 DH primes as an alternative.
542 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
543 are not in the system include path.
545 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
546 GnuTLS, when a session startup failed (eg because the client disconnected)
547 Exim did stdio operations after fclose. This was exposed by a recent
548 change which nulled out the file handle after the fclose.
550 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
551 signed directly by the cert-signing cert, rather than an intermediate
552 OCSP-signing cert. This is the model used by LetsEncrypt.
554 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
556 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
557 an incoming connection.
559 HS/02 Bug 1802: Do not half-close the connection after sending a request
562 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
563 fallback to "prime256v1".
565 JH/34 SECURITY: Use proper copy of DATA command in error message.
566 Could leak key material. Remotely exploitable. CVE-2016-9963.
572 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
573 and 3.4.4 - once the server is enabled to respond to an OCSP request
574 it does even when not requested, resulting in a stapling non-aware
575 client dropping the TLS connection.
577 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
578 support variable-length bit vectors. No functional change.
580 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
581 The I= interface field on outgoing lines is now after the H= remote
582 host field, same as incoming lines. There is a separate
583 outgoing_interface log selector which allows you to disable the
586 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
587 If not running log_selector +smtp_connection the mainlog would be held
588 open indefinitely after a "too many connections" event, including to a
589 deleted file after a log rotate. Leave the per net connection logging
590 leaving it open for efficiency as that will be quickly detected by the
591 check on the next write.
593 HS/01 Bug 1671: Fix post transport crash.
594 Processing the wait-<transport> messages could crash the delivery
595 process if the message IDs didn't exist for some reason. When
596 using 'split_spool_directory=yes' the construction of the spool
597 file name failed already, exposing the same netto behaviour.
599 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
600 mime_regex ACL conditions.
602 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
603 to DSN fail messages (bounces): remote IP, remote greeting, remote response
604 to HELO, local diagnostic string.
606 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
607 log line to debug. Even when configured with a tls authenticator many
608 client connections are expected to not authenticate in this way, so
609 an authenticate fail is not an error.
611 HS/02 Add the Exim version string to the process info. This way exiwhat
612 gives some more detail about the running daemon.
614 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
615 matter for fast-change records such as DNSBLs.
617 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
618 retry record, even if constant. There may be multiple transports with
619 different interface settings and the retry behaviour needs to be kept
622 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
624 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
626 JH/10 Bug 840: fix log_defer_output option of pipe transport
628 JH/11 Bug 830: use same host for all RCPTS of a message, even under
629 hosts_randomize. This matters a lot when combined with mua_wrapper.
631 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
632 ${quote_pgsql:<string>} operator.
634 JH/13 Bug 1708: avoid misaligned access in cached lookup.
636 JH/14 Change header file name for freeradius-client. Relevant if compiling
637 with Radius support; from the Gentoo tree and checked under Fedora.
639 JH/15 Bug 1712: Introduce $prdr_requested flag variable
641 JH/16 Bug 1714: Permit an empty string as expansion result for transport
642 option transport_filter, meaning no filtering.
644 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
646 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
647 defaults to "*" (all hosts). The variable is now available when not built
648 with TLS, default unset, mainly to enable keeping the testsuite sane.
649 If a server certificate is not supplied (via tls_certificate) an error is
650 logged, and clients will find TLS connections fail on startup. Presumably
651 they will retry in-clear.
652 Packagers of Exim are strongly encouraged to create a server certificate
653 at installation time.
655 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
656 with the $config_file variable.
658 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
659 in transport context, after the attempt, and per-recipient. The latter type
660 is per host attempted. The event data is the error message, and the errno
661 information encodes the lookup type (A vs. MX) used for the (first) host,
662 and the trailing two digits of the smtp 4xx response.
664 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
665 to write to mainlog (or rejectlog, paniclog) in the window between file
666 creation and permissions/ownership being changed. Particularly affects
667 installations where exicyclog is run as root, rather than exim user;
668 result is that the running daemon panics and dies.
670 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
672 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
673 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
674 "pri" and "weight". Note that the previous implicit priority given by the
675 list order is no longer honoured.
677 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
680 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
681 by defining SUPPORT_SOCKS.
683 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
684 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
685 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
686 variables are renamed to proxy_{local,external}_{address,port}.
688 JH/27 Move Internationalisation support from Experimental to mainline, enabled
689 for a build by defining SUPPORT_I18N
691 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
692 of the query string, and make ${quote_redis:} do that quoting.
694 JH/29 Move Events support from Experimental to mainline, enabled by default
695 and removable for a build by defining DISABLE_EVENT.
697 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
699 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
700 cached by the daemon.
702 JH/32 Move Redis support from Experimental to mainline, enabled for a build
703 by defining LOOKUP_REDIS. The libhiredis library is required.
705 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
706 keys are given for lookup.
708 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
709 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
710 only supported when built with TLS support. The PolarSSL SHA routines
711 are still used when the TLS library is too old for convenient support.
713 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
714 openssl_options), for security. OpenSSL forces this from version 1.1.0
715 server-side so match that on older versions.
717 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
718 allocation for $value could be released as the expansion processing
719 concluded, but leaving the global pointer active for it.
721 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
722 and to use the domains and local_parts ACL conditions.
724 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
725 incorrectly not doubled on cutthrough transmission, hence seen as a
726 body-termination at the receiving system - resulting in truncated mails.
727 Commonly the sender saw a TCP-level error, and retransmitted the message
728 via the normal store-and-forward channel. This could result in duplicates
729 received - but deduplicating mailstores were liable to retain only the
730 initial truncated version.
732 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
734 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
736 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
737 we're in there, support oversigning also; bug 1309.
739 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
741 HS/04 Add support for keep_environment and add_environment options.
743 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
744 either intentional arithmetic overflow during PRNG, or testing config-
747 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
748 delivery resulted in actual delivery. Cancel cutthrough before DATA
751 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
752 rejects a recipient: pass the reject to the originator.
754 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
755 Many were false-positives and ignorable, but it's worth fixing the
758 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
759 for the new environment-manipulation done at startup. Move the routines
760 from being local to tls.c to being global via the os.c file.
762 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
763 an extract embedded as result-arg for a map, the first arg for extract
764 is unavailable so we cannot tell if this is a numbered or keyed
765 extraction. Accept either.
771 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
774 JH/02 The smtp transport option "multi_domain" is now expanded.
776 JH/03 The smtp transport now requests PRDR by default, if the server offers
779 JH/04 Certificate name checking on server certificates, when exim is a client,
780 is now done by default. The transport option tls_verify_cert_hostnames
781 can be used to disable this per-host. The build option
782 EXPERIMENTAL_CERTNAMES is withdrawn.
784 JH/05 The value of the tls_verify_certificates smtp transport and main options
785 default to the word "system" to access the system default CA bundle.
786 For GnuTLS, only version 3.0.20 or later.
788 JH/06 Verification of the server certificate for a TLS connection is now tried
789 (but not required) by default. The verification status is now logged by
790 default, for both outbound TLS and client-certificate supplying inbound
793 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
796 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
797 Status Notification (bounce) messages are now MIME format per RFC 3464.
798 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
799 under the control of the dsn_advertise_hosts option, and routers may
800 have a dsn_lasthop option.
802 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
803 default, modifiable by a malware= option. The list separator for
804 the options can now be changed in the usual way. Bug 68.
806 JH/10 The smtp_receive_timeout main option is now expanded before use.
808 JH/11 The incoming_interface log option now also enables logging of the
809 local interface on delivery outgoing connections.
811 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
812 if the interface and destination host and port all match.
814 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
817 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
818 Patch from Andrew Lewis.
820 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
821 now supports optional time-restrictions, weighting, and priority
822 modifiers per server. Patch originally by <rommer@active.by>.
824 JH/16 The spamd_address main option now supports a mixed list of local
825 and remote servers. Remote servers can be IPv6 addresses, and
826 specify a port-range.
828 JH/17 Bug 68: The spamd_address main option now supports an optional
829 timeout value per server.
831 JH/18 Bug 1581: Router and transport options headers_add/remove can
832 now have the list separator specified.
834 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
837 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
840 JH/21 Support for the A6 type of dns record is withdrawn.
842 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
843 rather than the verbs used.
845 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
846 from 255 to 1024 chars.
848 JH/24 Verification callouts now attempt to use TLS by default.
850 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
851 are generic router options now. The defaults didn't change.
853 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
854 Original patch from Alexander Shikoff, worked over by JH.
856 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
857 files if ZCAT_COMMAND is not executable.
859 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
861 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
863 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
864 Normally benign, it bites when the pair was led to by a CNAME;
865 modern usage is to not canonicalize the domain to a CNAME target
866 (and we were inconsistent anyway for A-only vs AAAA+A).
868 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
870 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
871 when evaluating $sender_host_dnssec.
873 JH/31 Check the HELO verification lookup for DNSSEC, adding new
874 $sender_helo_dnssec variable.
876 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
878 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
880 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
882 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
883 documented as working, but never had. Support all but $spam_report.
885 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
886 added for tls authenticator.
888 HS/03 Add perl_taintmode main config option
894 TL/01 When running the test suite, the README says that variables such as
895 no_msglog_check are global and can be placed anywhere in a specific
896 test's script, however it was observed that placement needed to be near
897 the beginning for it to behave that way. Changed the runtest perl
898 script to read through the entire script once to detect and set these
899 variables, reset to the beginning of the script, and then run through
900 the script parsing/test process like normal.
902 TL/02 The BSD's have an arc4random API. One of the functions to induce
903 adding randomness was arc4random_stir(), but it has been removed in
904 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
905 function when detected.
907 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
908 cause callback expansion.
910 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
911 syntax errors in an expansion can be treated as a string instead of
912 logging or causing an error, due to the internal use of bool_lax
913 instead of bool when processing it.
915 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
916 server certificates when making smtp deliveries.
918 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
920 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
922 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
924 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
925 Merged patch from Sebastian Wiedenroth.
927 JH/05 Fix results-pipe from transport process. Several recipients, combined
928 with certificate use, exposed issues where response data items split
929 over buffer boundaries were not parsed properly. This eventually
930 resulted in duplicates being sent. This issue only became common enough
931 to notice due to the introduction of connection certificate information,
932 the item size being so much larger. Found and fixed by Wolfgang Breyha.
934 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
935 size buffer was used, resulting in syntax errors when an expansion
938 JH/07 Add support for directories of certificates when compiled with a GnuTLS
939 version 3.3.6 or later.
941 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
942 is EXPERIMENTAL_EVENT, the main-configuration and transport options
943 both become "event_action", the variables become $event_name, $event_data
944 and $event_defer_errno. There is a new variable $verify_mode, usable in
945 routers, transports and related events. The tls:cert event is now also
946 raised for inbound connections, if the main configuration event_action
949 TL/06 In test suite, disable OCSP for old versions of openssl which contained
950 early OCSP support, but no stapling (appears to be less than 1.0.0).
952 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
953 server certificate names available under the smtp transport option
954 "tls_verify_cert_hostname" now do not permit multi-component wildcard
957 JH/10 Time-related extraction expansions from certificates now use the main
958 option "timezone" setting for output formatting, and are consistent
959 between OpenSSL and GnuTLS compilations. Bug 1541.
961 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
962 encoded parameter in the incoming message. Bug 1558.
964 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
965 include certificate info, eximon was claiming there were spoolfile
968 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
970 JH/14 Log delivery-related information more consistently, using the sequence
971 "H=<name> [<ip>]" wherever possible.
973 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
974 are problematic for Debian distribution, omit them from the release
977 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
979 JH/16 Fix string representation of time values on 64bit time_t architectures.
982 JH/17 Fix a null-indirection in certextract expansions when a nondefault
983 output list separator was used.
988 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
989 checkers that were complaining about end of non-void function with no
992 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
993 This was a regression introduced in 4.83 by another bugfix.
995 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
997 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
998 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
1004 TF/01 Correctly close the server side of TLS when forking for delivery.
1006 When a message was received over SMTP with TLS, Exim failed to clear up
1007 the incoming connection properly after forking off the child process to
1008 deliver the message. In some situations the subsequent outgoing
1009 delivery connection happened to have the same fd number as the incoming
1010 connection previously had. Exim would try to use TLS and fail, logging
1011 a "Bad file descriptor" error.
1013 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1014 utilities have not been installed.
1016 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1017 temporary space as the ACL may create new global variables.
1019 TL/01 LDAP support uses per connection or global context settings, depending
1020 upon the detected version of the libraries at build time.
1022 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1023 to extract and use the src ip:port in logging and expansions as if it
1024 were a direct connection from the outside internet. PPv2 support was
1025 updated based on HAProxy spec change in May 2014.
1027 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1029 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1030 Properly escape header and check for NULL return.
1032 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1035 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1037 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1038 characters in header names, implemented as a verify condition.
1039 Contributed by Michael Fischer v. Mollard.
1041 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1042 results permerror and temperror. Previous values are deprecated but
1043 still accepted. In a future release, err_perm and err_temp will be
1044 completely removed, which will be a backward incompatibility if the
1045 ACL tests for either of these two old results. Patch contributed by
1046 user bes-internal on the mailing list.
1048 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1050 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1051 selectors, in both main and reject logs.
1053 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1056 JH/07 Add malware type "sock" for talking to simple daemon.
1058 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1060 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1061 routers/transports under cutthrough routing.
1063 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1064 numbers. Touch up "bool" conditional to keep the same definition.
1066 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1068 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1070 JH/12 Expand items in router/transport headers_add or headers_remove lists
1071 individually rather than the list as a whole. Bug 1452.
1073 Required for reasonable handling of multiple headers_ options when
1074 they may be empty; requires that headers_remove items with embedded
1075 colons must have them doubled (or the list-separator changed).
1077 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1078 view the policy declared in the DMARC record. Currently, $dmarc_status
1079 is a combined value of both the record presence and the result of the
1082 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1084 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1085 dnslookup router and the smtp transport (applying to the forward
1088 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1089 of ldap servers used for a specific lookup. Patch provided by Heiko
1092 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1093 New variable $lookup_dnssec_authenticated for observability.
1095 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1096 Patch submitted by Lars Timman.
1098 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1100 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1101 Requires trusted mode and valid format message id, aborts otherwise.
1102 Patch contributed by Heiko Schlichting.
1104 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1105 certextract with support for various fields. Bug 1358.
1107 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1108 is requested by default, modifiable by smtp transport option
1111 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1112 operate on certificate variables to give certificate fingerprints
1113 Also new ${sha256:cert_variable}.
1115 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1117 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1118 Christian Aistleitner.
1120 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1122 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1123 file. Patch from Wolfgang Breyha.
1125 JH/25 Expand the coverage of the delivery $host and $host_address to
1126 client authenticators run in verify callout. Bug 1476.
1128 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1129 align with daemon_smtp_ports. Bug 72.
1131 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1132 support and error reporting did not work properly.
1134 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1135 and is readable. Patch from Andrew Colin Kissa.
1137 TL/14 Enhance documentation of ${run expansion and how it parses the
1138 commandline after expansion, particularly in the case when an
1139 unquoted variable expansion results in an empty value.
1141 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1143 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1144 by renaming away from it. Addresses GH issue 15, reported by
1147 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1149 TL/15 SECURITY: prevent double expansion in math comparison functions
1150 (can expand unsanitized data). Not remotely exploitable.
1157 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1159 PP/02 Make -n do something, by making it not do something.
1160 When combined with -bP, the name of an option is not output.
1162 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1165 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1166 $sender_host_name and config options to manage this, and basic check
1169 PP/05 DSCP support for outbound connections and control modifier for inbound.
1171 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1172 (Only plugin which currently uses this is kerberos4, which nobody should
1173 be using, but we should make it available and other future plugins might
1174 conceivably use it, even though it would break NAT; stuff *should* be
1175 using channel bindings instead).
1177 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1178 name; added for Sendmail compatibility; requires admin caller.
1179 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1180 just ignore it); requires trusted caller.
1181 Also parse but ignore: -Ac -Am -X<logfile>
1184 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1186 TL/02 Add +smtp_confirmation as a default logging option.
1188 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1189 Patch by Magnus Holmgren from 2007-02-20.
1191 TL/04 Bugzilla 1281 - Spec typo.
1192 Bugzilla 1283 - Spec typo.
1193 Bugzilla 1290 - Spec grammar fixes.
1195 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1197 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1199 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1200 mailing list by by Dmitry Isaikin.
1202 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1204 JH/02 Support "G" suffix to numbers in ${if comparisons.
1206 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1208 NM/01 Bugzilla 1197 - Spec typo
1209 Bugzilla 1196 - Spec examples corrections
1211 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1213 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1214 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1217 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1218 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1220 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1221 "acl {{name}{arg}...}", and optional args on acl condition
1224 JH/05 Permit multiple router/transport headers_add/remove lines.
1226 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1228 JH/07 Avoid using a waiting database for a single-message-only transport.
1229 Performance patch from Paul Fisher. Bugzilla 1262.
1231 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1234 JH/09 Add $headers_added variable, with content from use of ACL modifier
1235 add_header (but not yet added to the message). Bugzilla 199.
1237 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1238 Pulled from Bugzilla 817 by Wolfgang Breyha.
1240 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1242 (nb: this is the same fix as in Exim 4.80.1)
1244 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1247 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1250 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1252 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1253 advertise SMTP AUTH mechanism to us, instead of a generic
1254 protocol violation error. Also, make Exim more robust to bad
1255 data from the Dovecot auth socket.
1257 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1259 When a queue runner is handling a message, Exim first routes the
1260 recipient addresses, during which it prunes them based on the retry
1261 hints database. After that it attempts to deliver the message to
1262 any remaining recipients. It then updates the hints database using
1265 So if a recipient address works intermittently, it can get repeatedly
1266 deferred at routing time. The retry hints record remains fresh so the
1267 address never reaches the final cutoff time.
1269 This is a fairly common occurrence when a user is bumping up against
1270 their storage quota. Exim had some logic in its local delivery code
1271 to deal with this. However it did not apply to per-recipient defers
1272 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1274 This change adds a proper retry rule check during routing so that the
1275 final cutoff time is checked against the message's age. We only do
1276 this check if there is an address retry record and there is not a
1277 domain retry record; this implies that previous attempts to handle
1278 the address had the retry_use_local_parts option turned on. We use
1279 this as an approximation for the destination being like a local
1280 delivery, as in LMTP.
1282 I suspect this new check makes the old local delivery cutoff check
1283 redundant, but I have not verified this so I left the code in place.
1285 TF/02 Correct gecos expansion when From: is a prefix of the username.
1287 Test 0254 submits a message to Exim with the header
1291 When I ran the test suite under the user fanf2, Exim expanded
1292 the header to contain my full name, whereas it should have added
1293 a Resent-Sender: header. It erroneously treats any prefix of the
1294 username as equal to the username.
1296 This change corrects that bug.
1298 GF/01 DCC debug and logging tidyup
1299 Error conditions log to paniclog rather than rejectlog.
1300 Debug lines prefixed by "DCC: " to remove any ambiguity.
1302 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1304 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1305 Bug spotted by Jeremy Harris; was flawed since initial commit.
1306 Would have resulted in OCSP responses post-SNI triggering an Exim
1307 NULL dereference and crash.
1309 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1311 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1312 Bug detection, analysis and fix by Samuel Thibault.
1313 Bugzilla 1331, Debian bug #698092.
1315 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1317 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1318 Server implementation by Todd Lyons, client by JH.
1319 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1320 config variable "prdr_enable" controls whether the server
1321 advertises the facility. If the client requests PRDR a new
1322 acl_data_smtp_prdr ACL is called once for each recipient, after
1323 the body content is received and before the acl_smtp_data ACL.
1324 The client is controlled by both of: a hosts_try_prdr option
1325 on the smtp transport, and the server advertisement.
1326 Default client logging of deliveries and rejections involving
1327 PRDR are flagged with the string "PRDR".
1329 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1330 fclose(). Diagnosis by Todd Lyons.
1332 PP/17 Update configure.default to handle IPv6 localhost better.
1333 Patch by Alain Williams (plus minor tweaks).
1336 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1337 This is now consistent with GnuTLS, and is now documented: the
1338 previous undocumented portable approach to treating the option as
1339 unset was to force an expansion failure. That still works, and
1340 an empty string is now equivalent.
1342 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1343 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1344 not performing validation itself.
1346 PP/20 Added force_command boolean option to pipe transport.
1347 Patch from Nick Koston, of cPanel Inc.
1349 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1352 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1354 PP/21 Fix eximon continuous updating with timestamped log-files.
1355 Broken in a format-string cleanup in 4.80, missed when I repaired the
1356 other false fix of the same issue.
1357 Report and fix from Heiko Schlichting.
1360 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1361 Report from Prashanth Katuri.
1363 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1364 It's SecureTransport, so affects any MacOS clients which use the
1365 system-integrated TLS libraries, including email clients.
1367 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1368 using a MIME ACL for non-SMTP local injection.
1369 Report and assistance in diagnosis by Warren Baker.
1371 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1373 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1375 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1376 last id that failed so it may be referenced in subsequent ACL's.
1378 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1381 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1382 ldap library initialization, allowing self-signed CA's to be
1383 used. Also properly sets require_cert option later in code by
1384 using NULL (global ldap config) instead of ldap handle (per
1385 session). Bug diagnosis and testing by alxgomz.
1387 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1388 the src/util/ subdirectory.
1390 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1391 renamed to Transport Post Delivery Action by Jeremy Harris, as
1394 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1395 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1396 redis_servers = needs to be configured which will be used by the redis
1397 lookup. Patch from Warren Baker, of The Packet Hub.
1399 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1401 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1402 hostname or reverse DNS when processing a host list. Used suggestions
1403 from multiple comments on this bug.
1405 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1407 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1408 Missed a few lines, added it to make the runtest require no keyboard
1411 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1412 contains upper case chars. Make router use caseful_local_part.
1414 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1415 support when GnuTLS has been built with p11-kit.
1421 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1423 This, or similar/improved, will also be change PP/11 of 4.82.
1429 PP/01 Handle short writes when writing local log-files.
1430 In practice, only affects FreeBSD (8 onwards).
1431 Bugzilla 1053, with thanks to Dmitry Isaikin.
1433 NM/01 Bugzilla 949 - Documentation tweak
1435 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1438 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1440 PP/02 Implemented gsasl authenticator.
1442 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1444 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1445 `pkg-config foo` for cflags/libs.
1447 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1448 with rest of GSASL and with heimdal_gssapi.
1450 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1451 `pkg-config foo` for cflags/libs for the TLS implementation.
1453 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1454 properties get this fed in as external SSF. A number of robustness
1455 and debugging improvements to the cyrus_sasl authenticator.
1457 PP/08 cyrus_sasl server now expands the server_realm option.
1459 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1460 Patch by Jeremy Harris.
1462 PP/10 Added dbmjz lookup type.
1464 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1466 PP/12 MAIL args handles TAB as well as SP, for better interop with
1467 non-compliant senders.
1468 Analysis and variant patch by Todd Lyons.
1470 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1471 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1472 Patch from Dirk Mueller <dmueller@suse.com>
1474 PP/13 tls_peerdn now print-escaped for spool files.
1475 Observed some $tls_peerdn in wild which contained \n, which resulted
1476 in spool file corruption.
1478 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1479 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1480 or write after TLS renegotiation, which otherwise led to messages
1483 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1484 as a tracking header (ie: a signed header comes before the signature).
1485 Patch from Wolfgang Breyha.
1487 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1488 comma-sep list; embedded commas doubled.
1490 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1492 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1494 Report and patch from Dmitry Banschikov.
1496 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1497 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1498 failures. We appear to now support TLS1.1+ with Exim.
1500 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1501 lets Exim select keys and certificates based upon TLS SNI from client.
1502 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1503 before an outbound SMTP session. New log_selector, +tls_sni.
1505 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1506 NULL dereference. Report and patch from Alun Jones.
1508 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1509 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1510 Not seeing resolver debug output on NetBSD, but suspect this is a
1511 resolver implementation change.
1513 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1514 Left warnings. Added "eximon gdb" invocation mode.
1516 PP/21 Defaulting "accept_8bitmime" to true, not false.
1518 PP/22 Added -bw for inetd wait mode support.
1520 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1521 locate the relevant includes and libraries. Made this the default.
1523 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1524 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1526 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1527 This may cause build issues on older platforms.
1529 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1530 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1531 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1532 Added SNI support via GnuTLS too.
1533 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1535 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1537 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1538 Applied second patch from Janne, implementing suggestion to default
1539 multiple-strings-in-record handling to match SPF spec.
1541 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1543 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1544 read-only, out of scope).
1545 Patch from Wolfgang Breyha, report from Stuart Northfield.
1547 PP/29 Fix three issues highlighted by clang analyser static analysis.
1548 Only crash-plausible issue would require the Cambridge-specific
1549 iplookup router and a misconfiguration.
1550 Report from Marcin Mirosław.
1552 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1554 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1555 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1556 As part of this, removing so much warning spew let me fix some minor
1557 real issues in debug logging.
1559 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1560 assignment on my part. Fixed.
1562 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1563 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1564 Janne Snabb (who went above and beyond: thank you).
1566 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1567 string otherwise requires a connection and a bunch more work and it's
1568 relatively easy to get wrong. Should also expose TLS library linkage
1571 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1572 64-bit ${eval} (JH/03).
1574 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1575 GNU libc to support some of the 64-bit stuff, should not lead to
1576 conflicts. Defined before os.h is pulled in, so if a given platform
1577 needs to override this, it can.
1579 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1580 protection layer was required, which is not implemented.
1581 Bugzilla 1254, patch from Wolfgang Breyha.
1583 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1584 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1585 tls_dhparam take prime identifiers. Also unbreak combination of
1586 OpenSSL+DH_params+TLSSNI.
1588 PP/39 Disable SSLv2 by default in OpenSSL support.
1594 PP/01 Solaris build fix for Oracle's LDAP libraries.
1595 Bugzilla 1109, patch from Stephen Usher.
1597 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1599 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1602 TF/02 Fix a couple more cases where we did not log the error message
1603 when unlink() failed. See also change 4.74-TF/03.
1605 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1606 lock up or crash if it happened to be inside a call to libc when it
1607 got a SIGUSR1 from exiwhat.
1609 The SIGUSR1 handler appends the current process status to the process
1610 log which is later printed by exiwhat. It used to use the general
1611 purpose logging code to do this, but several functions it calls are
1612 not safe for signals.
1614 The new output code in the SIGUSR1 handler is specific to the process
1615 log, and simple enough that it's easy to inspect for signal safety.
1616 Removing some special cases also simplifies the general logging code.
1617 Removing the spurious timestamps from the process log simplifies
1620 TF/04 Improved ratelimit ACL condition.
1622 The /noupdate option has been deprecated in favour of /readonly which
1623 has clearer semantics. The /leaky, /strict, and /readonly update modes
1624 are mutually exclusive. The update mode is no longer included in the
1625 database key; it just determines when the database is updated. (This
1626 means that when you upgrade Exim will forget old rate measurements.)
1628 Exim now checks that the per_* options are used with an update mode that
1629 makes sense for the current ACL. For example, when Exim is processing a
1630 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1631 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1632 must specify per_mail/readonly. If you omit the update mode it defaults to
1633 /leaky where that makes sense (as before) or /readonly where required.
1635 The /noupdate option is now undocumented but still supported for
1636 backwards compatibility. It is equivalent to /readonly except that in
1637 ACLs where /readonly is required you may specify /leaky/noupdate or
1638 /strict/noupdate which are treated the same as /readonly.
1640 A useful new feature is the /count= option. This is a generalization
1641 of the per_byte option, so that you can measure the throughput of other
1642 aggregate values. For example, the per_byte option is now equivalent
1643 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1645 The per_rcpt option has been generalized using the /count= mechanism
1646 (though it's more complicated than the per_byte equivalence). When it is
1647 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1648 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1649 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1650 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1651 Note that using per_rcpt with a non-readonly update mode in more than
1652 one ACL will cause the recipients to be double-counted. (The per_mail
1653 and per_byte options don't have this problem.)
1655 The handling of very low rates has changed slightly. If the computed rate
1656 is less than the event's count (usually one) then this event is the first
1657 after a long gap. In this case the rate is set to the same as this event's
1658 count, so that the first message of a spam run is counted properly.
1660 The major new feature is a mechanism for counting the rate of unique
1661 events. The new per_addr option counts the number of different
1662 recipients that someone has sent messages to in the last time period. It
1663 behaves like per_rcpt if all the recipient addresses are different, but
1664 duplicate recipient addresses do not increase the measured rate. Like
1665 the /count= option this is a general mechanism, so the per_addr option
1666 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1667 example, measure the rate that a client uses different sender addresses
1668 with the options per_mail/unique=$sender_address. There are further
1669 details in the main documentation.
1671 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1673 TF/06 Removed a few PCRE remnants.
1675 TF/07 Automatically extract Exim's version number from tags in the git
1676 repository when doing development or release builds.
1678 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1679 Bugzilla 879. Patch from Paul Fisher.
1681 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1682 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1685 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1687 PP/05 Variable $av_failed, true if the AV scanner deferred.
1688 Bugzilla 1078. Patch from John Horne.
1690 PP/06 Stop make process more reliably on build failure.
1691 Bugzilla 1087. Patch from Heiko Schlittermann.
1693 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1694 Bugzilla 1089. Patch from Heiko Schlittermann.
1696 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1697 Bugzilla 1131. Patch from Holger Weiß.
1699 PP/09 Handle IPv6 addresses with SPF.
1700 Bugzilla 860. Patch from Wolfgang Breyha.
1702 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1704 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1707 PP/11 match_* no longer expand right-hand-side by default.
1708 New compile-time build option, EXPAND_LISTMATCH_RHS.
1709 New expansion conditions, "inlist", "inlisti".
1711 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1713 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1715 PP/14 fix log_write() format string regression from TF/03.
1716 Bugzilla 1152. Patch from Dmitry Isaikin.
1722 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1724 PP/02 Harmonised TLS library version reporting; only show if debugging.
1725 Layout now matches that introduced for other libraries in 4.74 PP/03.
1727 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1729 PP/04 New "dns_use_edns0" global option.
1731 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1734 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1735 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1737 TK/01 Updated PolarSSL code to 0.14.2.
1738 Bugzilla 1097. Patch from Andreas Metzler.
1740 PP/07 Catch divide-by-zero in ${eval:...}.
1741 Fixes bugzilla 1102.
1743 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1746 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1747 format-string attack -- SECURITY: remote arbitrary code execution.
1749 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1750 time unintentionally subject to list matching rules, letting the header
1751 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1752 arbitrary string expansion). This allowed for information disclosure.
1754 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1755 INT_MIN/-1 -- value coerced to INT_MAX.
1761 NM/01 Workaround for PCRE version dependency in version reporting
1764 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1765 This fixes portability to compilers other than gcc, notably
1766 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1768 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1769 makefiles for portability to HP-UX and POSIX correctness.
1771 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1772 Also via indented variable definition in the Makefile.
1773 (Debugging by Oliver Heesakkers).
1775 PP/02 Restore caching of spamd results with expanded spamd_address.
1776 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1778 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1779 Improves build reliability. Fix from: Frank Elsner
1781 NM/02 Fix wide character breakage in the rfc2047 coding
1782 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1784 NM/03 Allow underscore in dnslist lookups
1785 Fixes bug 1026. Patch from Graeme Fowler
1787 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1788 Code patches from Adam Ciarcinski of NetBSD.
1790 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1793 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1794 is logged, to avoid truncation. Patch from John Horne.
1796 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1797 Patch from Jakob Hirsch.
1799 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1800 SQL string expansion failure details.
1801 Patch from Andrey Oktyabrski.
1803 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1804 Patch from Simon Arlott.
1806 PP/09 New lookups functionality failed to compile on old gcc which rejects
1807 extern declarations in function scope.
1808 Patch from Oliver Fleischmann
1810 PP/10 Use sig_atomic_t for flags set from signal handlers.
1811 Check getgroups() return and improve debugging.
1812 Fixed developed for diagnosis in bug 927 (which turned out to be
1815 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1816 Patch from Mark Zealey.
1818 PP/12 Bugzilla 1056: Improved spamd server selection.
1819 Patch from Mark Zealey.
1821 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1822 Based on patch from Heiko Schlittermann.
1824 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1825 Patch from Uwe Doering, sign-off by Michael Haardt.
1827 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1828 variable declaration deep within a block. Bug and patch from
1831 PP/15 lookups-Makefile IRIX compatibility coercion.
1833 PP/16 Make DISABLE_DKIM build knob functional.
1835 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1836 Patch by Simon Arlott
1838 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1839 variable argument macros. Our copy now differs from upstream.
1845 TF/01 Failure to get a lock on a hints database can have serious
1846 consequences so log it to the panic log.
1848 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1849 controlled using the smtp_confirmation log selector.
1851 TF/03 Include the error message when we fail to unlink a spool file.
1853 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1854 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1855 for maintaining out-of-tree patches for some time.
1857 PP/01 Bugzilla 139: Documentation and portability issues.
1858 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1859 Handle per-OS dynamic-module compilation flags.
1861 PP/02 Let /dev/null have normal permissions.
1862 The 4.73 fixes were a little too stringent and complained about the
1863 permissions on /dev/null. Exempt it from some checks.
1864 Reported by Andreas M. Kirchwitz.
1866 PP/03 Report version information for many libraries, including
1867 Exim version information for dynamically loaded libraries. Created
1868 version.h, now support a version extension string for distributors
1869 who patch heavily. Dynamic module ABI change.
1871 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1872 privilege escalation vulnerability whereby the Exim run-time user
1873 can cause root to append content of the attacker's choosing to
1876 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1879 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1880 If dropping privileges for untrusted macros, we disabled normal logging
1881 on the basis that it would fail; for the Exim run-time user, this is not
1882 the case, and it resulted in successful deliveries going unlogged.
1883 Fixed. Reported by Andreas Metzler.
1889 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1890 only prepend for the Resent-* case. Fixes regression introduced in
1891 Exim 4.70 by NM/22 for Bugzilla 607.
1893 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1894 increasing numbers of administrators be bitten by this.
1896 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1898 PP/03 Bugzilla 994: added openssl_options main configuration option.
1900 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1902 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1904 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1906 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1907 without a peer certificate, leading to a segfault because of an
1908 assumption that peers always have certificates. Be a little more
1909 paranoid. Problem reported by Martin Tscholak.
1911 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1912 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1913 NB: ClamAV planning to remove STREAM in "middle of 2010".
1914 CL also introduces -bmalware, various -d+acl logging additions and
1915 more caution in buffer sizes.
1917 PP/09 Implemented reverse_ip expansion operator.
1919 PP/10 Bugzilla 937: provide a "debug" ACL control.
1921 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1923 PP/12 Bugzilla 973: Implement --version.
1925 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1927 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1929 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1931 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1932 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1933 ignore trailing whitespace.
1935 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1937 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1940 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1941 Notification from Dr Andrew Aitchison.
1943 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1944 ExtendedDetectionInfo response format.
1945 Notification from John Horne.
1947 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1950 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1951 XSL and documented dependency on system catalogs, with examples of how
1954 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1957 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1958 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1959 configuration file which is writeable by the Exim user or group.
1961 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1962 of configuration files to cover files specified with the -C option if
1963 they are going to be used with root privileges, not just the default
1966 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1967 option (effectively making it always true).
1969 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1970 files to be used while preserving root privileges.
1972 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1973 that rogue child processes cannot use them.
1975 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1976 run-time user, instead of root.
1978 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1979 Exim run-time user without dropping privileges.
1981 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1982 result string, instead of calling string_vformat() twice with the same
1985 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1986 for other users. Others should always drop root privileges if they use
1987 -C on the command line, even for a whitelisted configure file.
1989 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1991 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1997 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1998 $data_path, and $header_path variables; fixed documentation bugs and
2001 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
2002 exipick to access non-standard spools, including the "frozen" queue
2005 NM/01 Bugzilla 965: Support mysql stored procedures.
2006 Patch from Alain Williams
2008 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2010 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2011 Patch from Andreas Metzler
2013 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2014 Patch from Kirill Miazine
2016 NM/05 Bugzilla 671: Added umask to procmail example.
2018 JJ/03 installed exipick 20100323.0, fixing doc bug
2020 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2021 directory. Notification and patch from Dan Rosenberg.
2023 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2025 TK/02 Improve log output when DKIM signing operation fails.
2027 MH/01 Treat the transport option dkim_domain as a colon separated
2028 list, not as a single string, and sign the message with each element,
2029 omitting multiple occurences of the same signer.
2031 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2032 Bugzilla 985, 986. Patch by Simon Arlott
2034 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2035 Patch by Simon Arlott
2037 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2038 MBX locking. Notification from Dan Rosenberg.
2044 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2046 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2048 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2050 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2052 NM/04 Bugzilla 915: Flag broken perl installation during build.
2058 TK/01 Added patch by Johannes Berg that expands the main option
2059 "spamd_address" if it starts with a dollar sign.
2061 TK/02 Write list of recipients to X-Envelope-Sender header when building
2062 the mbox-format spool file for content scanning (suggested by Jakob
2065 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2066 (http://www.dcc-servers.net/) support via dccifd. Activated by
2067 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2069 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2070 by Mark Daniel Reidel <mr@df.eu>.
2072 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2073 When building exim an external PCRE library is now needed -
2074 PCRE is a system library on the majority of modern systems.
2075 See entry on PCRE_LIBS in EDITME file.
2077 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2078 conversation. Added nologin parameter to request.
2079 Patch contributed by Kirill Miazine.
2081 TF/01 Do not log submission mode rewrites if they do not change the address.
2083 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2085 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2086 log files in place. Contributed by Roberto Lima.
2088 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2090 TF/03 Bugzilla 615: When checking the local_parts router precondition
2091 after a local_part_suffix or local_part_prefix option, Exim now
2092 does not use the address's named list lookup cache, since this
2093 contains cached lookups for the whole local part.
2095 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2096 Robert Millan. Documentation is in experimental-spec.txt.
2098 TF/04 Bugzilla 668: Fix parallel build (make -j).
2100 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2102 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2103 Patch provided by Jan Srzednicki.
2105 TF/05 Leading white space used to be stripped from $spam_report which
2106 wrecked the formatting. Now it is preserved.
2108 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2109 that they are available at delivery time.
2111 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2113 TF/08 TLS error reporting now respects the incoming_interface and
2114 incoming_port log selectors.
2116 TF/09 Produce a more useful error message if an SMTP transport's hosts
2117 setting expands to an empty string.
2119 NM/06 Bugzilla 744: EXPN did not work under TLS.
2120 Patch provided by Phil Pennock.
2122 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2123 Patch provided by Richard Godbee.
2125 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2126 acl_smtp_notquit, added index entry.
2128 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2129 Patch provided by Eugene Bujak.
2131 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2132 accept(). Patch provided by Maxim Dounin.
2134 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2135 Patch provided by Phil Pennock.
2137 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2139 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2140 Patch provided by Brad "anomie" Jorsch.
2142 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2143 Patch provided by Dean Brooks.
2145 TK/05 Add native DKIM support (does not depend on external libraries).
2147 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2148 Patch provided by Graeme Fowler.
2150 NM/16 Bugzilla 851: Documentation example syntax fix.
2152 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2154 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2157 NM/19 Bugzilla 745: TLS version reporting.
2158 Patch provided by Phil Pennock.
2160 NM/20 Bugzilla 167: bool: condition support.
2161 Patch provided by Phil Pennock.
2163 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2164 clients. Patch provided by Phil Pennock.
2166 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2167 Patch provided by Brad "anomie" Jorsch.
2169 NM/23 Bugzilla 687: Fix misparses in eximstats.
2170 Patch provided by Heiko Schlittermann.
2172 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2173 Patch provided by Heiko Schlittermann.
2175 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2176 plus update to original patch.
2178 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2180 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2181 Patch provided by David Brownlee.
2183 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2185 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2187 NM/30 Bugzilla 888: TLS documentation bugfixes.
2189 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2191 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2192 Unlike the original bugzilla I have changed all shell scripts in src tree.
2194 NM/33 Bugzilla 898: Transport filter timeout fix.
2195 Patch by Todd Rinaldo.
2197 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2198 Patch by Serge Demonchaux.
2200 NM/35 Bugzilla 39: Base64 decode bug fixes.
2201 Patch by Jakob Hirsch.
2203 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2205 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2207 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2209 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2215 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2216 ALT-N's libdkim that I have put here:
2217 http://duncanthrax.net/exim-experimental/
2219 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2220 were called 'true' and it seems that C99 defines that as a reserved
2221 keyword to be used with 'bool' variable types. That means you could
2222 not include C99-style headers which use bools without triggering
2223 build errors in sieve.c.
2225 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2226 as mailq or other aliases. Changed the --help handling significantly
2227 to do whats expected. exim_usage() emits usage/help information.
2229 SC/01 Added the -bylocaldomain option to eximstats.
2231 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2233 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2235 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2241 PH/01 Another patch from the Sieve maintainer.
2243 PH/02 When an IPv6 address is converted to a string for single-key lookup
2244 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2245 dots are used instead of colons so that keys in lsearch files need not
2246 contain colons. This was done some time before quoting was made available
2247 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2248 (notated using the quote facility) so as to distinguish them from IPv4
2249 keys. This meant that lookups for IP addresses in host lists did not work
2250 for iplsearch lookups.
2252 This has been fixed by arranging for IPv6 addresses to be expressed with
2253 colons if the lookup type is iplsearch. This is not incompatible, because
2254 previously such lookups could never work.
2256 The situation is now rather anomalous, since one *can* have colons in
2257 ordinary lsearch keys. However, making the change in all cases is
2258 incompatible and would probably break a number of configurations.
2260 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2263 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2264 conversion specification without a maximum field width, thereby enabling
2265 a rogue spamd server to cause a buffer overflow. While nobody in their
2266 right mind would setup Exim to query an untrusted spamd server, an
2267 attacker that gains access to a server running spamd could potentially
2268 exploit this vulnerability to run arbitrary code as the Exim user.
2270 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2271 $primary_hostname instead of what libspf2 thinks the hosts name is.
2273 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2274 a directory entry by the name of the lookup key. Previously, if a
2275 symlink pointed to a non-existing file or a file in a directory that
2276 Exim lacked permissions to read, a lookup for a key matching that
2277 symlink would fail. Now it is enough that a matching directory entry
2278 exists, symlink or not. (Bugzilla 503.)
2280 PH/03 The body_linecount and body_zerocount variables are now exported in the
2283 PH/04 Added the $dnslist_matched variable.
2285 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2286 This means they are set thereafter only if the connection becomes
2289 PH/06 Added the client_condition to authenticators so that some can be skipped
2290 by clients under certain conditions.
2292 PH/07 The error message for a badly-placed control=no_multiline_responses left
2293 "_responses" off the end of the name.
2295 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2297 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2298 (without spaces) instead of just copying the configuration text.
2300 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2302 PH/11 Added $max_received_linelength.
2304 PH/12 Added +ignore_defer and +include_defer to host lists.
2306 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2307 way in which PCRE > 7.0 is built.
2309 PH/14 Implemented queue_only_load_latch.
2311 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2312 MAIL command. The effect was to mangle the value on 64-bit systems.
2314 PH/16 Another patch from the Sieve maintainer.
2316 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2318 PH/18 If a system quota error occurred while trying to create the file for
2319 a maildir delivery, the message "Mailbox is full" was not appended to the
2320 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2321 only to a quota excession during the actual writing of the file.
2323 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2324 characters?) which causes problems in log lines. The DN values are now
2325 passed through string_printing() before being added to log lines.
2327 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2328 and InterBase are left for another time.)
2330 PH/21 Added message_body_newlines option.
2332 PH/22 Guard against possible overflow in moan_check_errorcopy().
2334 PH/23 POSIX allows open() to be a macro; guard against that.
2336 PH/24 If the recipient of an error message contained an @ in the local part
2337 (suitably quoted, of course), incorrect values were put in $domain and
2338 $local_part during the evaluation of errors_copy.
2344 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2345 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2348 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2349 issue a MAIL command.
2351 PH/02 In an ACL statement such as
2353 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2355 if a client was not listed at all, or was listed with a value other than
2356 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2357 the condition was not true (as it should be), so access was not denied.
2358 The bug was that the ! inversion was incorrectly passed on to the second
2359 item. This has been fixed.
2361 PH/03 Added additional dnslists conditions == and =& which are different from
2362 = and & when the dns lookup returns more than one IP address.
2364 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2365 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2367 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2368 FSYNC, which compiles an option called disable_fsync that allows for
2369 bypassing fsync(). The documentation is heavily laced with warnings.
2371 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2373 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2374 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2375 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2376 including adding "make clean"; (3) Added -fPIC when compiling the test
2377 dynamically loaded module, to get rid of a warning.
2379 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2380 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2381 = 0s. The bug is otherwise harmless.
2383 PH/07 There was a bug in the dovecot authenticator such that the value of
2384 $auth1 could be overwritten, and so not correctly preserved, after a
2385 successful authentication. This usually meant that the value preserved by
2386 the server_setid option was incorrect.
2388 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2390 PH/09 Installed PCRE release 7.0.
2392 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2393 run for batched SMTP input. It is now run at the start of every message
2394 in the batch. While fixing this I discovered that the process information
2395 (output by running exiwhat) was not always getting set for -bs and -bS
2396 input. This is fixed, and it now also says "batched" for BSMTP.
2398 PH/11 Added control=no_pipelining.
2400 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2401 patch, slightly modified), and move the expansion of helo_data till after
2402 the connection is made in the smtp transport (so it can use these
2405 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2407 PH/14 Added log_selector = +pid.
2409 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2411 PH/16 Add ${if forany and ${if forall.
2413 PH/17 Added dsn_from option to vary the From: line in DSNs.
2415 PH/18 Flush SMTP output before performing a callout, unless control =
2416 no_callout_flush is set.
2418 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2419 was true (the default) a successful delivery failed to delete the retry
2420 item, thus causing premature timeout of the address. The bug is now
2423 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2425 PH/21 Long custom messages for fakedefer and fakereject are now split up
2426 into multiline responses in the same way that messages for "deny" and
2427 other ACL rejections are.
2429 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2430 with slight modification.
2432 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2433 draft, changing the syntax and factoring some duplicate code".
2435 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2436 for deliveries of the second and subsequent messages over the same SMTP
2439 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2440 ${reduce, with only minor "tidies".
2442 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2444 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2445 expansion side effects.
2447 PH/27 When a message times out after an over-quota error from an Exim-imposed
2448 quota, the bounce message says "mailbox is full". This message was not
2449 being given when it was a system quota that was exceeded. It now should
2452 MH/03 Made $recipients available in local_scan(). local_scan() already has
2453 better access to the recipient list through recipients_list[], but
2454 $recipients can be useful in postmaster-provided expansion strings.
2456 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2457 in the case of a MAIL command with additional options following the
2458 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2459 were accidentally chopped off.
2461 PH/29 SMTP synchronization checks are implemented when a command is read -
2462 there is a check that no more input is waiting when there shouldn't be
2463 any. However, for some commands, a delay in an ACL can mean that it is
2464 some time before the response is written. In this time, more input might
2465 arrive, invalidly. So now there are extra checks after an ACL has run for
2466 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2467 pipelining has not been advertised.
2469 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2471 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2472 returned in the SMTP response when smtp_return_error_details was false.
2473 This has been fixed.
2475 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2476 instead of the C I/O that was originally supplied, because problems were
2477 reported on Solaris.
2479 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2480 Exim which did not show up earlier: it was assuming that a call to
2481 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2482 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2483 was a macro that became an assignment, so it seemed to work. This has
2484 changed to a proper function call with a void return, hence the compile
2485 error. Exim's code has been fixed.
2487 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2490 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2492 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2494 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2495 because it turns out to be too big or there is a timeout) there is no
2496 "Completed" line in the log. When some messages of this type were
2497 selected by exigrep, they were listed as "not completed". Others were
2498 picked up by some special patterns. I have improved the selection
2499 criteria to be more general.
2501 PH/38 The host_find_failed option in the manualroute router can now be set
2502 to "ignore", to completely ignore a host whose IP address cannot be
2503 found. If all hosts are ignored, the behaviour is controlled by the new
2504 host_all_ignored option.
2506 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2507 homing or because of multiple MX records with /mx) generated more than
2508 one IP address, and the following item turned out to be the local host,
2509 all the secondary addresses of the first item were incorrectly removed
2510 from the list, along with the local host and any following hosts (which
2511 is what is supposed to happen).
2513 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2514 whoever called Exim into the -H file. In the case of the daemon it was
2515 behaving confusingly. When first started, it used values for whoever
2516 started the daemon, but after a SIGHUP it used the Exim user (because it
2517 calls itself on a restart). I have changed the code so that it now always
2520 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2521 message are rejected with the same error (e.g. no authentication or bad
2522 sender address), and a DATA command is nevertheless sent (as can happen
2523 with PIPELINING or a stupid MUA), the error message that was given to the
2524 RCPT commands is included in the rejection of the DATA command. This is
2525 intended to be helpful for MUAs that show only the final error to their
2528 PH/42 Another patch from the Sieve maintainer.
2530 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2531 Eximstats - Fixed some broken HTML links and added missing column headers
2533 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2534 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2536 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2538 PH/43 Yet another patch from the Sieve maintainer.
2540 PH/44 I found a way to check for a TCP/IP connection going away before sending
2541 the response to the final '.' that terminates a message, but only in the
2542 case where the client has not sent further data following the '.'
2543 (unfortunately, this is allowed). However, in many cases there won't be
2544 any further data because there won't be any more messages to send. A call
2545 to select() can be used: if it shows that the input is "ready", there is
2546 either input waiting, or the socket has been closed. An attempt to read
2547 the next input character can distinguish the two cases. Previously, Exim
2548 would have sent an OK response which the client would never have see.
2549 This could lead to message repetition. This fix should cure that, at
2550 least in a lot of common cases.
2552 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2553 advertised in response to EHLO.
2559 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2560 fixed by 4.65/MH/01 (is this a record?) are fixed:
2562 (i) An empty string was always treated as zero by the numeric comparison
2563 operators. This behaviour has been restored.
2565 (ii) It is documented that the numeric comparison operators always treat
2566 their arguments as decimal numbers. This was broken in that numbers
2567 starting with 0 were being interpreted as octal.
2569 While fixing these problems I realized that there was another issue that
2570 hadn't been noticed. Values of message_size_limit (both the global option
2571 and the transport option) were treated as octal if they started with 0.
2572 The documentation was vague. These values are now always treated as
2573 decimal, and I will make that clear in the documentation.
2579 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2580 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2583 MH/01 Don't check that the operands of numeric comparison operators are
2584 integers when their expansion is in "skipping" mode (fixes bug
2585 introduced by 4.64-PH/07).
2587 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2588 child addresses, Exim now panics and dies. Previously, because the count
2589 is held in a short int, deliveries were likely to be lost. As such a
2590 large number of recipients for a single message is ridiculous
2591 (performance will be very, very poor), I have chosen to impose a limit
2592 rather than extend the field.
2598 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2599 leftover -K file (the existence of which was triggered by #402).
2600 While we were at it, introduced process PID as part of the -K
2601 filename. This should rule out race conditions when creating
2604 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2605 processing considerably. Previous code took too long for large mails,
2606 triggering a timeout which in turn triggers #401.
2608 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2609 in the DK code in transports.c. sendfile() is not really portable,
2610 hence the _LINUX specificness.
2612 TF/01 In the add_headers option to the mail command in an Exim filter,
2613 there was a bug that Exim would claim a syntax error in any
2614 header after the first one which had an odd number of characters
2617 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2618 callout verification, Exim cached a "reject" for the entire domain. This
2619 is correct for most verifications, but it is not correct for a recipient
2620 verification with use_sender or use_postmaster set, because in that case
2621 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2622 case of MAIL FROM:<> rejection from other early rejections (e.g.
2623 rejection of HELO). When verifying a recipient using a non-null MAIL
2624 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2625 Whatever the result of the callout, the value of the domain cache is
2626 left unchanged (for any other kind of callout, getting as far as trying
2627 RCPT means that the domain itself is ok).
2629 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2632 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2633 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2634 session, the daemon ignores ECONNECT errors and logs others; it now
2635 ignores EPIPE as well.
2637 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2638 (quoted-printable decoding).
2640 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2641 later the small subsequent patch to fix an introduced bug.
2643 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2645 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2647 PH/08 An error is now given if message_size_limit is specified negative.
2649 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2650 to be given (somewhat) arbitrary names.
2652 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2655 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2656 miscellaneous code fixes
2658 PH/10 Added the log_reject_target ACL modifier to specify where to log
2661 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2662 hostname. This is wrong, because it relates to the incoming message (and
2663 probably the interface on which it is arriving) and not to the outgoing
2664 callout (which could be using a different interface). This has been
2665 changed to use the value of the helo_data option from the smtp transport
2666 instead - this is what is used when a message is actually being sent. If
2667 there is no remote transport (possible with a router that sets up host
2668 addresses), $smtp_active_hostname is used.
2670 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2671 tweaks were necessary in order to get it to work (see also 21 below):
2672 (a) The code assumed that strncpy() returns a negative number on buffer
2673 overflow, which isn't the case. Replaced with Exim's string_format()
2675 (b) There were several signed/unsigned issues. I just did the minimum
2676 hacking in of casts. There is scope for a larger refactoring.
2677 (c) The code used strcasecmp() which is not a standard C function.
2678 Replaced with Exim's strcmpic() function.
2679 (d) The code set only $1; it now sets $auth1 as well.
2680 (e) A simple test gave the error "authentication client didn't specify
2681 service in request". It would seem that Dovecot has changed its
2682 interface. Fortunately there's a specification; I followed it and
2683 changed what the client sends and it appears to be working now.
2685 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2688 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2689 address A is aliased to B and C, where B exists and C does not. Without
2690 -v the output is "A verified" because verification stops after a
2691 successful redirection if more than one address is generated. However,
2692 with -v the child addresses are also verified. Exim was outputting "A
2693 failed to verify" and then showing the successful verification for C,
2694 with its parentage. It now outputs "B failed to verify", showing B's
2695 parentage before showing the successful verification of C.
2697 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2698 look up a TXT record in a specific list after matching in a combined
2701 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2702 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2703 they consult the DNS. I had assumed they would set it the way they
2704 wanted; and indeed my experiments on Linux seem to show that in some
2705 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2706 To be on the safe side, however, I have now made the interface to
2707 host_find_byname() similar to host_find_bydns(), with an argument
2708 containing the DNS resolver options. The host_find_byname() function now
2709 sets these options at its start, just as host_find_bydns() does. The smtp
2710 transport options dns_qualify_single and dns_search_parents are passed to
2711 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2712 of host_find_byname() use the default settings of RES_DEFNAMES
2713 (qualify_single) but not RES_DNSRCH (search_parents).
2715 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2716 spool_read_header() do less string testing, by means of a preliminary
2717 switch on the second character of optional "-foo" lines. (This is
2718 overdue, caused by the large number of possibilities that now exist.
2719 Originally there were few.) While I was there, I also converted the
2720 str(n)cmp tests so they don't re-test the leading "-" and the first
2721 character, in the hope this might squeeze out yet more improvement.
2723 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2724 flag allowing group syntax was set by the header_syntax check but not
2725 turned off, possible causing trouble later; (2) The flag was not being
2726 set at all for the header_verify test, causing "group"-style headers to
2727 be rejected. I have now set it in this case, and also caused header_
2728 verify to ignore an empty address taken from a group. While doing this, I
2729 came across some other cases where the code for allowing group syntax
2730 while scanning a header line wasn't quite right (mostly, not resetting
2731 the flag correctly in the right place). These bugs could have caused
2732 trouble for malformed header lines. I hope it is now all correct.
2734 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2735 with the "reply" argument non-NULL. The code, however (which originally
2736 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2737 but it didn't always do it. This confused somebody who was copying the
2738 code for some other use. I have removed all the tests.
2740 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2741 feature that was used to support insecure browsers during the U.S. crypto
2742 embargo. It requires special client support, and Exim is probably the
2743 only MTA that supported it -- and would never use it because real RSA is
2744 always available. This code has been removed, because it had the bad
2745 effect of slowing Exim down by computing (never used) parameters for the
2746 RSA_EXPORT functionality.
2748 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2749 authenticator to fail if there's a tab character in the incoming data
2750 (there should never be unless someone is messing about, as it's supposed
2751 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2752 if the connection is using TLS or if the remote IP is the same as the
2753 local IP, and the "valid-client-cert option" if a client certificate has
2756 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2757 authenticators. This can be used for authorization after authentication
2758 succeeds. (In the case of plaintext, it servers for both authentication
2761 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2762 if any retry times were supplied.
2764 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2765 connection, where there is no client IP address to check. In this
2766 situation, the verify now always succeeds.
2768 PH/25 Applied John Jetmore's -Mset patch.
2770 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2772 PH/27 In a string expansion for a processed (not raw) header when multiple
2773 headers of the same name were present, leading whitespace was being
2774 removed from all of them, but trailing whitespace was being removed only
2775 from the last one. Now trailing whitespace is removed from each header
2776 before concatenation. Completely empty headers in a concatenation (as
2777 before) are ignored.
2779 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2780 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2782 PH/29 [Removed. This was a change that I later backed out, and forgot to
2783 correct the ChangeLog entry (that I had efficiently created) before
2784 committing the later change.]
2786 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2787 address errors (4xx response to RCPT) over the same connection as other
2788 messages routed to the same hosts. Such deliveries are always "forced",
2789 so retry times are not inspected. This resulted in far too many retries
2790 for the affected addresses. The effect occurred only when there were more
2791 hosts than the hosts_max_try setting in the smtp transport when it had
2792 the 4xx errors. Those hosts that it had tried were not added to the list
2793 of hosts for which the message was waiting, so if all were tried, there
2794 was no problem. Two fixes have been applied:
2796 (i) If there are any address or message errors in an SMTP delivery, none
2797 of the hosts (tried or untried) are now added to the list of hosts
2798 for which the message is waiting, so the message should not be a
2799 candidate for sending over the same connection that was used for a
2800 successful delivery of some other message. This seems entirely
2801 reasonable: after all the message is NOT "waiting for some host".
2802 This is so "obvious" that I'm not sure why it wasn't done
2803 previously. Hope I haven't missed anything, but it can't do any
2804 harm, as the worst effect is to miss an optimization.
2806 (ii) If, despite (i), such a delivery is accidentally attempted, the
2807 routing retry time is respected, so at least it doesn't keep
2808 hammering the server.
2810 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2811 in ${readsocket because some servers need this prod.
2813 PH/32 Added some extra debug output when updating a wait-xxx database.
2815 PH/33 The hint "could be header name not terminated by colon", which has been
2816 given for certain expansion errors for a long time, was not being given
2817 for the ${if def:h_colon_omitted{... case.
2819 PH/34 The spec says: "With one important exception, whenever a domain list is
2820 being scanned, $domain contains the subject domain." There was at least
2821 one case where this was not true.
2823 PH/35 The error "getsockname() failed: connection reset by peer" was being
2824 written to the panic log as well as the main log, but it isn't really
2825 panic-worthy as it just means the connection died rather early on. I have
2826 removed the panic log writing for the ECONNRESET error when getsockname()
2829 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2830 runs only) independently of the message's sender address. This meant
2831 that, if the 4xx error was in fact related to the sender, a different
2832 message to the same recipient with a different sender could confuse
2833 things. In particular, this can happen when sending to a greylisting
2834 server, but other circumstances could also provoke similar problems.
2835 I have changed the default so that the retry time for these errors is now
2836 based a combination of the sender and recipient addresses. This change
2837 can be overridden by setting address_retry_include_sender=false in the
2840 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2841 remote server are returned as part of bounce messages. This was not
2842 happening for LMTP over a pipe (the lmtp transport), but now it is the
2843 same for both kinds of LMTP.
2845 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2846 in header lines that were in fact CNAMEs. This is no longer the case.
2848 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2849 and queue runs started by the daemon processed all messages. This has
2850 been fixed so that -R and -S can now usefully be given with -q<time>.
2852 PH/40 Import PCRE release 6.7 (fixes some bugs).
2854 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2856 PH/42 Give an error if -q is specified more than once.
2858 PH/43 Renamed the variables $interface_address and $interface_port as
2859 $received_ip_address and $received_port, to make it clear that these
2860 values apply to message reception, and not to the outgoing interface when
2861 a message is delivered. (The old names remain recognized, of course.)
2863 PH/44 There was no timeout on the connect() call when using a Unix domain
2864 socket in the ${readsocket expansion. There now is.
2866 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2867 be meaningful with "accept".
2869 SC/01 Eximstats V1.43
2870 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2872 SC/02 Eximstats V1.44
2873 Use a glob alias rather than an array ref in the generated
2874 parser. This improves both readability and performance.
2876 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2877 Collect SpamAssassin and rejection statistics.
2878 Don't display local sender or destination tables unless
2879 there is data to show.
2880 Added average volumes into the top table text output.
2882 SC/04 Eximstats V1.46
2883 Collect data on the number of addresses (recipients)
2884 as well as the number of messages.
2886 SC/05 Eximstats V1.47
2887 Added 'Message too big' to the list of mail rejection
2888 reasons (thanks to Marco Gaiarin).
2890 SC/06 Eximstats V1.48
2891 Mainlog lines which have GMT offsets and are too short to
2892 have a flag are now skipped.
2894 SC/07 Eximstats V1.49 (Alain Williams)
2895 Added the -emptyok flag.
2897 SC/08 Eximstats V1.50
2898 Fixes for obtaining the IP address from reject messages.
2900 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2901 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2902 whitespace changes from 4.64-PH/27
2904 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2907 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2908 are found, allow negative numbers in numeric criteria)
2910 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2912 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2913 to match changes made in 4.64-PH/43
2915 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2917 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2918 "domain" is actually an IP address, and force a failure. This locks out
2919 those revolvers/nameservers that support "A-for-A" lookups, in
2920 contravention of the specifications.
2922 PH/48 When a host name was looked up from an IP address, and the subsequent
2923 forward lookup of the name timed out, the host name was left in
2924 $sender_host_name, contrary to the specification.
2926 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2927 restricted to single-key lookups, Exim was not diagnosing an error if
2928 * or *@ was used with a query-style lookup.
2930 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2932 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2933 long ago, but noone interested enough thought of it. Let's just say that
2934 the "1.1" means that there are some new functions that weren't there at
2935 some point in the past.
2937 PH/51 Error processing for expansion failure of helo_data from an smtp
2938 transport during callout processing was broken.
2940 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2941 tested/used via the -bh/-bhc/-bs options.
2943 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2944 bug, fixed in subsequent PCRE releases).
2946 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2947 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2949 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2955 SC/01 Use a glob alias rather than an array ref in eximstats generated
2956 parser. This improves both readability and performance.
2958 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2959 Don't display local sender or destination tables in eximstats unless
2960 there is data to show.
2961 Added average volumes into the eximstats top table text output.
2963 SC/03 Collect data on the number of addresses (recipients) as well
2964 as the number of messages in eximstats.
2966 TF/01 Correct an error in the documentation for the redirect router. Exim
2967 does (usually) call initgroups() when daemonizing.
2969 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2970 with consistent privilege compared to when running as a daemon.
2972 TF/03 Note in the spec that $authenticated_id is not set for local
2973 submissions from trusted users.
2975 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2976 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2978 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2979 by adding some example configuration directives to the default
2980 configuration file. A little bit of work is required to uncomment the
2981 directives and define how usernames and passwords are checked, but
2982 there is now a framework to start from.
2984 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2985 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2986 without this. I don't know how relevant this is to other LDAP libraries.
2988 PH/02 Add the verb name to the "unknown ACL verb" error.
2990 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2992 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2994 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2995 directory not expanded when it should be if an expanded home directory
2996 was set for the address (which is overridden by the transport).
2998 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
3001 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
3002 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
3003 because it is too late at that time, and has no effect.
3005 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
3006 security issue with \' (bugzilla #107). I could not use the
3007 PQescapeStringConn() function, because it needs a PGconn value as one of
3010 PH/08 When testing addresses using -bt, indicate those final addresses that
3011 are duplicates that would not cause an additional delivery. At least one
3012 person was confused, thinking that -bt output corresponded to deliveries.
3013 (Suppressing duplicates isn't a good idea as you lose the information
3014 about possibly different redirections that led to the duplicates.)
3016 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3017 systems where poll() doesn't work, in particular OS X.
3019 PH/10 Added more information to debugging output for retry time not reached.
3021 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3022 operations in malware.c.
3024 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3027 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3028 syslog_duplication set false, log lines that would normally be written
3029 both the the main log and to the reject log were not written to syslog at
3032 PH/14 In the default configuration, change the use of "message" in ACL warn
3033 statements to "add_header".
3035 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3036 not followed by a command (e.g. "seen endif").
3038 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3039 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3042 PH/17 Added extra conditions to the default value of delay_warning_condition
3046 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3047 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3048 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3051 The Auto-Submitted: and various List- headers are standardised, whereas I
3052 don't think Precedence: ever was.
3054 PH/18 Refactored debugging code in route_finduser() to show more information,
3055 in particular, the error code if getpwnam() issues one.
3057 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3058 This is apparently needed in addition to the PH/07 change above to avoid
3059 any possible encoding problems.
3061 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3062 but not after initializing Perl.
3064 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3065 output them only if debugging. By default they are written stderr,
3066 apparently, which is not desirable.
3068 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3071 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3074 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3076 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3077 authenticated or an ident call has been made. Suppress the default
3078 values for $authenticated_id and $authenticated_sender (but permit -oMai
3079 and -oMas) when testing with -bh.
3081 PH/24 Re-jigged the order of the tests in the default configuration so that the
3082 tests for valid domains and recipients precede the DNS black list and CSA
3083 tests, on the grounds that those ones are more expensive.
3085 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3086 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3087 This bug exists in every version of Exim that I still have, right back to
3090 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3091 However, an attempt to turn on case-sensitivity in a regex key by
3092 including (?-i) didn't work because the subject string was already
3093 lowercased, and the effects were non-intuitive. It turns out that a
3094 one-line patch can be used to allow (?-i) to work as expected.
3100 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3101 other effects) broke the use of negated acl sub-conditions.
3103 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3106 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3107 "Deny" causes Exim to reject the incoming connection with a 554 error.
3108 Unfortunately, if there is a major crisis, such as a disk failure,
3109 tcp-wrappers gives "deny", whereas what one would like would be some
3110 kind of temporary error. A kludge has been added to help with this.
3111 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3112 554 error is used if errno is still zero or contains ENOENT (which occurs
3113 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3116 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3118 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3119 errors. Otherwise a message that provokes a temporary error (when other
3120 messages do not) can cause a whole host to time out.
3122 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3123 addresses were routed directly to files or pipes from a redirect router.
3124 File deliveries just didn't batch; pipe deliveries might have suffered
3127 PH/06 A failure to get a lock for a hints database would erroneously always say
3128 "Failed to get write lock", even when it was really a read lock.
3130 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3131 of 0600. This has been changed to use the value of the lockfile_mode
3132 option (which defaults to 0600).
3134 PH/08 Applied small patch from the Sieve maintainer.
3136 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3137 folder from quota calculations, a direct delivery into this folder messed
3138 up the contents of the maildirsize file. This was because the regex was
3139 used only to exclude .Trash (or whatever) when the size of the mailbox
3140 was calculated. There was no check that a delivery was happening into an
3141 excluded directory. This bug has been fixed by ignoring all quota
3142 processing for deliveries into excluded directories.
3144 PH/10 Added the maildirfolder_create_regex option to appendfile.
3150 PH/01 The code for finding all the local interface addresses on a FreeBSD
3151 system running IPv6 was broken. This may well have applied to all BSD
3152 systems, as well as to others that have similar system calls. The broken
3153 code found IPv4 interfaces correctly, but gave incorrect values for the
3154 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3155 that it would not match correctly against @[] and not recognize the IPv6
3158 PH/02 The ipliteral router was not recognizing addresses of the form user@
3159 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3161 PH/03 Added disable_ipv6.
3163 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3164 decoded form, because it is most often used to construct To: headers
3165 lines in autoreplies, and the decoded form may well be syntactically
3166 invalid. However, $reply_address has leading white space removed, and all
3167 newlines turned into spaces so that the autoreply transport does not
3170 PH/05 If group was specified without a user on a router, and no group or user
3171 was specified on a transport, the group from the router was ignored.
3173 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3174 for visible compile-time settings that can be used to change these
3175 numbers, for those that want even more. Backwards compatibility with old
3176 spool files has been maintained. However, going back to a previous Exim
3177 release will lost any variables that are in spool files.
3179 PH/07 Two small changes when running in the test harness: increase delay when
3180 passing a TCP/IP connection to a new process, in case the original
3181 process has to generate a bounce, and remove special handling of
3182 127.0.0.2 (sic), which is no longer necessary.
3184 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3185 be the same on different OS.
3187 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3190 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3191 whether --show-vars was specified or not
3193 JJ/02 exipick: Added support for new ACL variable spool format introduced
3196 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3197 syntactically invalid From: or Reply-to: line, and a filter used this to
3198 generate an autoreply, and therefore failed to obtain an address for the
3199 autoreply, Exim could try to deliver to a non-existent relative file
3200 name, causing unrelated and misleading errors. What now happens is that
3201 it logs this as a hard delivery error, but does not attempt to create a
3204 PH/11 The exinext utility has a -C option for testing purposes, but although
3205 the given file was scanned by exinext itself; it wasn't being passed on
3206 when Exim was called.
3208 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3209 an end-of-file indication when reading a command response.
3211 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3212 compiled. In many other places in Exim, IPv6 addresses are always
3213 recognized, so I have changed this. It also means that IPv4 domain
3214 literals of the form [IPV4:n.n.n.n] are now always recognized.
3216 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3217 used if the router is not running as root, for example, when verifying at
3218 ACL time, or when using -bh. The debugging output from this situation was
3219 non-existent - all you got was a failure to exec. I have made two
3222 (a) Failures to set uid/gid, the current directory, or a process leader
3223 in a subprocess such as that created by queryprogram now generate
3224 suitable debugging output when -d is set.
3226 (b) The queryprogram router detects when it is not running as root,
3227 outputs suitable debugging information if -d is set, and then runs
3228 the subprocess without attempting to change uid/gid.
3230 PH/15 Minor change to Makefile for building test_host (undocumented testing
3233 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3234 additional section of a DNS packet that returns MX or SRV records.
3235 Instead, it always explicitly searches for A/AAAA records. This avoids
3236 major problems that occur when a DNS server includes only records of one
3237 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3238 fixed another bug: if SRV records were looked up and the corresponding
3239 address records were *not* found in the additional section, the port
3240 values from the SRV records were lost.
3242 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3243 using the correct key (the original address) when searching the retry
3244 rules in order to find which one to use for generating the retry hint.
3246 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3247 adding the default one. Similarly, if it contains a Reply-To: header, the
3248 errors_reply_to option, if set, is not used.
3250 PH/19 When calculating a retry time, Exim used to measure the "time since
3251 failure" by looking at the "first failed" field in the retry record. Now
3252 it does not use this if it is later than than the arrival time of the
3253 message. Instead it uses the arrival time. This makes for better
3254 behaviour in cases where some deliveries succeed, thus re-setting the
3255 "first failed" field. An example is a quota failure for a huge message
3256 when small messages continue to be delivered. Without this change, the
3257 "time since failure" will always be short, possible causing more frequent
3258 delivery attempts for the huge message than are intended.
3259 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3261 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3262 $1, $2, $3) because the numerical variables can be reset during some
3263 expansion items (e.g. "match"), thereby losing the authentication data.
3265 PH/21 Make -bV show the size of off_t variables so that the test suite can
3266 decide whether to run tests for quotas > 2G.
3268 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3269 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3270 filecount value is greater than 2G or if a quota value is greater than 2G
3271 on a system where the size of off_t is not greater than 4, a panic error
3274 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3275 never match. The debug and -bh output now contains an explicit error
3276 message indicating a malformed IPv4 address or mask.
3278 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3279 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3280 PH/23 above applies.
3282 PH/25 Do not write to syslog when running in the test harness. The only
3283 occasion when this arises is a failure to open the main or panic logs
3284 (for which there is an explicit test).
3286 PH/26 Added the /no_tell option to "control=freeze".
3288 PH/27 If a host name lookup failed very early in a connection, for example, if
3289 the IP address matched host_lookup and the reverse lookup yielded a name
3290 that did not have a forward lookup, an error message of the form "no IP
3291 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3292 could be logged. Now it outputs the IP address instead of "NULL".
3294 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3295 allows the sender and the authenticated sender to be set when
3296 submitting a message from within Exim. Since child_open_exim() is
3297 documented for local_scan(), the new function should be too.
3299 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3300 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3301 results in an empty string is now treated as unset.
3303 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3305 PH/31 Added sender_verify_fail logging option.
3307 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3308 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3309 tidied the source and removed it altogether.
3311 PH/33 When a queue run was abandoned because the load average was too high, a
3312 log line was always written; now it is written only if the queue_run log
3313 selector is set. In addition, the log line for abandonment now contains
3314 information about the queue run such as the pid. This is always present
3315 in "start" and "stop" lines but was omitted from the "abandon" line.
3317 PH/34 Omit spaces between a header name and the colon in the error message that
3318 is given when verify = headers_syntax fails (if there are lots of them,
3319 the message gets confusing).
3321 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3322 names, as there are now some PTR records that contain slashes. This check
3323 is only to protect against broken name servers that fall over on strange
3324 characters, so the fact that it applies to all lookups doesn't matter.
3326 PH/36 Now that the new test suite is complete, we can remove some of the
3327 special code in Exim that was needed for the old test suite. For example,
3328 sorting DNS records because real resolvers return them in an arbitrary
3329 order. The new test suite's fake resolver always returns records in the
3332 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3333 bounces) except when queue_only is set, to avoid logging races between
3334 the different processes.
3336 PH/38 Panic-die if .include specifies a non-absolute path.
3338 PH/39 A tweak to the "H" retry rule from its user.
3340 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3341 a label. They prevented compilation on older perls.
3343 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3344 a warning to be raised on newish perls.
3346 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3347 on queue. Changes to match documented behaviour of showing count of
3348 messages matching specified criteria.
3350 PH/40 Changed the default ident timeout from 30s to 5s.
3352 PH/41 Added support for the use of login_cap features, on those BSD systems
3353 that have them, for controlling the resources used by pipe deliveries.
3355 PH/42 The content-scanning code uses fopen() to create files in which to put
3356 message data. Previously it was not paying any attention to the mode of
3357 the files. Exim runs with umask(0) because the rest of the code creates
3358 files with open(), and sets the required mode explicitly. Thus, these
3359 files were ending up world-writeable. This was not a big issue, because,
3360 being within the spool directory, they were not world-accessible. I have
3361 created a function called modefopen, which takes an additional mode
3362 argument. It sets umask(777), creates the file, chmods it to the required
3363 mode, then resets the umask. All the relevant calls to fopen() in the
3364 content scanning code have been changed to use this function.
3366 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3367 to 24 hours. This avoids potential overflow problems when processing G
3368 and H retry rules. I suspect nobody ever tinkers with this value.
3370 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3372 PH/45 When the plaintext authenticator is running as a client, the server's
3373 challenges are checked to ensure they are valid base64 strings. By
3374 default, the authentication attempt is cancelled if an invalid string is
3375 received. Setting client_ignore_invalid_base64 true ignores these errors.
3376 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3377 they are received. Thus, the responses can be made to depend on the
3378 challenges. If an invalid string is ignored, an empty string is placed in
3381 PH/46 Messages that are created by the autoreply transport now contains a
3382 References: header, in accordance with RFCs 2822 and 3834.
3384 PH/47 Added authenticated_sender_force to the smtp transport.
3386 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3388 PH/49 Installed latest patch from the Sieve maintainer.
3390 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3391 was also set, the appendfile transport was unnecessarily scanning a
3392 directory of message files (e.g. for maildir delivery) to find the count
3393 of files (along with the size), even though it did not need this
3394 information. It now does the scan only if it needs to find either the
3395 size of the count of files.
3397 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3399 PH/52 Two bugs concerned with error handling when the smtp transport is
3402 (i) Exim was not creating retry information for temporary errors given
3403 for individual recipients after the DATA command when the smtp transport
3404 was used in LMTP mode. This meant that they could be retried too
3405 frequently, and not timed out correctly.
3407 (ii) Exim was setting the flag that allows error details to be returned
3408 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3409 recipients that were returned after the DATA command.
3411 PH/53 This is related to PH/52, but is more general: for any failing address,
3412 when detailed error information was permitted to be returned to the
3413 sender, but the error was temporary, then after the final timeout, only
3414 "retry timeout exceeded" was returned. Now it returns the full error as
3415 well as "retry timeout exceeded".
3417 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3418 do this, and (what is worse) MTAs that accept it.
3420 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3421 will now be deprecated.
3423 PH/56 New os.c-cygwin from the Cygwin maintainer.
3425 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3426 formats (previously only available in exim formats via -bpr, -bpru,
3427 and -bpra. Now also available in native and exiqgrep formats)
3429 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3430 with very large, slow to parse queues
3432 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3434 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3436 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3437 responses to authentication challenges, though it was showing the
3438 challenges; (ii) I've removed the CR characters from the debug output for
3441 PH/58 Allow for the insertion of a newline as well as a space when a string
3442 is turned into more than one encoded-word during RFC 2047 encoding. The
3443 Sieve code now uses this.
3445 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3446 data_4xx, lost_connection, tls_required.
3448 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3449 message was being sent as an SMTP response.
3451 PH/61 Add -l and -k options to exicyclog.
3453 PH/62 When verifying, if an address was redirected to one new address, so that
3454 verification continued, and the new address failed or deferred after
3455 having set something in $address_data, the value of $address_data was not
3456 passed back to the ACL. This was different to the case when no
3457 redirection occurred. The value is now passed back in both cases.
3459 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3460 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3461 use login_cap.h, so on its own it isn't the distinguishing feature. The
3462 new name refers directly to the setclassresources() function.
3464 PH/65 Added configuration files for NetBSD3.
3466 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3468 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3469 is preferred over IPv4.
3471 PH/68 The bounce_return_message and bounce_return_body options were not being
3472 honoured for bounces generated during the reception of non-SMTP messages.
3473 In particular, this applied to messages rejected by the ACL. This bug has
3474 been fixed. However, if bounce_return_message is true and bounce_return_
3475 body is false, the headers that are returned for a non-SMTP message
3476 include only those that have been read before the error was detected.
3477 (In the case of an ACL rejection, they have all been read.)
3479 PH/69 The HTML version of the specification is now built in a directory called
3480 spec_html instead of spec.html, because the latter looks like a path with
3481 a MIME-type, and this confuses some software.
3483 PH/70 Catch two compiler warnings in sieve.c.
3485 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3486 function verify_get_ident() calls ip_connect() to connect a socket, but
3487 if the "connect()" function timed out, ip_connect() used to close the
3488 socket. However, verify_get_ident() also closes the socket later, and in
3489 between Exim writes to the log, which may get opened at this point. When
3490 the socket was closed in ip_connect(), the log could get the same file
3491 descriptor number as the socket. This naturally causes chaos. The fix is
3492 not to close the socket in ip_connect(); the socket should be closed by
3493 the function that creates it. There was only one place in the code where
3494 this was missing, in the iplookup router, which I don't think anybody now
3495 uses, but I've fixed it anyway.
3497 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3498 well as to direct DNS lookups. Otherwise the handling of names in host
3499 lists is inconsistent and therefore confusing.
3505 PH/01 Two changes to the default runtime configuration:
3507 (1) Move the checks for relay_from_hosts and authenticated clients from
3508 after to before the (commented out) DNS black list checks.
3510 (2) Add control=submission to the relay_from_hosts and authenticated
3511 clients checks, on the grounds that messages accepted by these
3512 statements are most likely to be submissions.
3514 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3516 (1) Generate an error if the third argument for the ${prvs expansion is
3519 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3522 (3) Reset the variables that are obtained from the first argument of
3523 ${prvscheck and used in the second argument before leaving the code,
3524 because their memory is reclaimed, so using them afterwards may do
3527 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3528 one (it's much easier than Tom thought :-).
3530 (5) Because of (4), we can now allow for the use of $prvscheck_result
3531 inside the third argument.
3533 PH/03 For some reason, the default setting of PATH when running a command from
3534 a pipe transport was just "/usr/bin". I have changed it to
3537 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3538 anything to be listed in the output from -bV.
3540 PH/05 When a filter generated an autoreply, the entire To: header line was
3541 quoted in the delivery log line, like this:
3543 => >A.N.Other <ano@some.domain> <original@ddress> ...
3545 This has been changed so that it extracts the operative address. There
3546 may be more than one such address. If so, they are comma-separated, like
3549 => >ano@some.domain,ona@other.domain <original@ddress> ...
3551 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3552 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3553 not being looked up in the rDNS to get a host name, Exim was showing the
3554 IP address twice in Received: lines, even though the IP addresses were
3555 identical. For example:
3557 Received: from [1.2.3.4] (helo=[1.2.3.4])
3559 However, if the real host name was known, it was omitting the HELO data
3560 if it matched the actual IP address. This has been tidied up so that it
3561 doesn't show the same IP address twice.
3563 PH/07 When both +timestamp and +memory debugging was on, the value given by
3564 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3565 called by the debug printing, thereby overwriting the timestamp buffer.
3566 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3568 PH/08 When the original message was included in an autoreply transport, it
3569 always said "this is a copy of the message, including all the headers",
3570 even if body_only or headers_only was set. It now gives an appropriate
3573 PH/09 Applied a patch from the Sieve maintainer which:
3575 o fixes some comments
3576 o adds the (disabled) notify extension core
3577 o adds some debug output for the result of if/elsif tests
3578 o points to the current vacation draft in the documentation
3579 and documents the missing references header update
3583 o fixes a bug in processing the envelope test (when testing
3584 multiple envelope elements, the last element determined the
3587 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3588 Electronic Mail") by including:
3590 Auto-submitted: auto-generated
3592 in the messages that it generates (bounce messages and others, such as
3593 warnings). In the case of bounce messages for non-SMTP messages, there was
3594 also a typo: it was using "Auto_submitted" (underscore instead of
3595 hyphen). Since every message generated by Exim is necessarily in response
3596 to another message, thes have all been changed to:
3598 Auto-Submitted: auto-replied
3600 in accordance with these statements in the RFC:
3602 The auto-replied keyword:
3604 - SHOULD be used on messages sent in direct response to another
3605 message by an automatic process,
3607 - MUST NOT be used on manually-generated messages,
3609 - MAY be used on Delivery Status Notifications (DSNs) and Message
3610 Disposition Notifications (MDNs),
3612 - MUST NOT be used on messages generated by automatic or periodic
3613 processes, except for messages which are automatic responses to
3616 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3617 to the default Received: header definition.
3619 PH/12 Added log selector acl_warn_skipped (default on).
3621 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3622 variables because (a) they are in the wrong storage pool and (b) even if
3623 they were copied, it wouldn't work properly because of the caching.
3625 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3626 checking when decoding. Apparently there are clients that generate
3627 overlong encoded strings. Why am I not surprised?
3629 PH/15 If the first argument of "${if match_address" was not empty, but did not
3630 contain an "@" character, Exim crashed. Now it writes a panic log message
3631 and treats the condition as false.
3633 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3635 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3636 extension "envelope-auth". The code is finished and in agreement with
3637 other implementations, but there is no documentation so far and in fact,
3638 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3639 not changing the active code.
3641 Print executed "if" and "elsif" statements when debugging is used. This
3642 helps a great deal to understand what a filter does.
3644 Document more things not specified clearly in RFC3028. I had all this
3645 sorted out, when out of a sudden new issues came to my mind. Oops."
3647 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3650 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3651 canonical form (as documented). However, after a host name lookup from
3652 the IP address, check_host() was doing a simple string comparison with
3653 addresses acquired from the DNS when checking that the found name did
3654 have the original IP as one of its addresses. Since any found IPv6
3655 addresses are likely to be in abbreviated form, the comparison could
3656 fail. Luckily, there already exists a function for doing the comparison
3657 by converting both addresses to binary, so now that is used instead of
3658 the text comparison.
3660 PH/20 There was another similar case to PH/19, when a complete host name was
3661 given in a host list; looking up its IP address could give an abbreviated
3662 form, whereas the current host's name might or might not be abbreviated.
3663 The same fix has been applied.
3669 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3670 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3673 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3674 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3676 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3678 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3679 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3680 Also, for "Auto-Submitted", treat anything other than "no" as
3681 non-personal, in accordance with RFC 3834. (Previously it treated
3682 anything starting "auto-" as non-personal.)
3684 TF/01 The control=submission/name=... option had a problem with syntax
3685 errors if the name included a slash character. The /name= option
3686 now slurps the rest of the string, so it can include any characters
3687 but it must come last in the list of options (after /sender_retain
3690 PH/05 Some modifications to the interface to the fake nameserver for the new
3698 TK/01 Added the "success_on_redirect" address verification option. See
3699 NewStuff for rationale and an example.
3701 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3703 PH/02 Patch to exigrep to allow it to work on syslog lines.
3705 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3706 fread() to skip over the body file's header line, because in Cygwin the
3707 header line is locked and is inaccessible.
3709 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3710 co-exist for some time) to make it clear that it is the Exim ID that is
3711 referenced, not the Message-ID: header line.
3713 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3714 string_format() function, because snprintf() does not exist on all
3717 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3718 ${stat: expansion item.
3720 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3721 protocol synchronization error", to keep the pedants happy.
3723 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3724 well as for IRIX systems, when gcc is being used. See the host.c source
3727 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3729 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3732 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3733 transport and to the smtp transport in LMTP mode.
3735 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3737 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3738 run a filter in a subprocess. This could lead to confusion in subsequent
3739 lookups in the parent process. There should also be a search_tidyup() at
3740 the end of the subprocess.
3742 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3743 only if the host matched helo_try_verify_hosts, which caused the
3744 verification to occur when the EHLO/HELO command was issued. The ACL just
3745 tested the remembered result. Now, if a previous verification attempt has
3746 not happened, "verify = helo" does it there and then.
3748 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3750 TK/03 Fix log output including CR from clamd.
3752 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3753 exist provoked a memory error which could cause a segfault.
3755 PH/15 Installed PCRE 6.2
3757 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3759 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3760 of the problem. Specifically, suggested +O2 rather than +O1 for the
3763 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3765 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3766 record, the port was not correctly logged when the outgoing_port log
3767 selector was set (it logged the transort's default port).
3769 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3770 fallback_hosts, and "hosts" in the smtp transport.
3772 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3773 host errors such as "Connection refused".
3775 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3776 authentication with radiusclient 0.4.9:
3778 - Error returned from rc_read_config was caught wrongly
3779 - Username/password not passed on to radius server due to wrong length.
3781 The presumption is that some radiusclient API changes for 4.51/PH/17
3782 were not taken care of correctly. The code is still untested by me (my
3783 Linux distribution still has 0.3.2 of radiusclient), but it was
3784 contributed by a Radius user.
3786 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3787 expanding the "port" option of the smtp transport.
3789 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3790 while reading a MIME header. Thanks to Tom Hughes for a patch.
3792 PH/24 Include config.h inside local_scan.h so that configuration settings are
3795 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3796 that in an ACL for RCPT (for example), you can examine exactly what was
3799 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3800 commands, but it was not correctly comparing the address with the actual
3801 client host address. Thus, it would show the EHLO address in Received:
3802 header lines when this was not necessary.
3804 PH/27 Added the % operator to ${eval:}.
3806 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3807 it should be ignoring failures (because with -C, for example, it has lost
3808 privilege). It wasn't ignoring creation failures other than "already
3811 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3814 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3815 because an input line was too long, either on its own, or by virtue of
3816 too many continuations, the temporary file was not being removed, and the
3817 return code was incorrect.
3819 PH/31 Missing "BOOL" in function definition in filtertest.c.
3821 PH/32 Applied Sieve patches from the maintainer.
3823 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3825 PH/33 Added "verify = not_blind".
3827 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3828 Local/Makefile (with some defaults set). These are used in built scripts
3829 such as exicyclog, but they have never been used in the exim_install
3830 script (though there are many overriding facilities there). I have
3831 arranged that the exim_install script now takes note of these two
3834 PH/35 Installed configuration files for Dragonfly.
3836 PH/36 When a locally submitted message by a trusted user did not contain a
3837 From: header, and the sender address was obtained from -f or from an SMTP
3838 MAIL command, and the trusted user did not use -F to supply a sender
3839 name, $originator_name was incorrectly used when constructing a From:
3840 header. Furthermore, $originator_name was used for submission mode
3841 messages from external hosts without From: headers in a similar way,
3842 which is clearly wrong.
3844 PH/37 Added control=suppress_local_fixups.
3846 PH/38 When log_selector = +received_sender was set, and the addition of the
3847 sender made the log line's construction buffer exactly full, or one byte
3848 less than full, an overflow happened when the terminating "\n" was
3851 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3852 when the result of a list match is failure because a DNS lookup failed.
3854 PH/40 RM_COMMAND is now used in the building process.
3856 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3857 the "build-* directories that it finds.
3859 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3860 address in a domain literal was a prefix of an interface address.
3862 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3863 when verifying a sender address, unless rewrite_headers is false.
3865 PH/44 Wrote a long comment about why errors_to addresses are verified as
3866 recipients, not senders.
3868 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3869 the ratelimit ACL was added.
3871 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3873 PH/46 Added extra information about PostgreSQL errors to the error string.
3875 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3876 avoiding the need to install special zones in a real server. This is
3877 backwards compatible; if it can't find the fake resolver, it drops back.
3878 Thus, both old and new test suites can be run.
3880 TF/02 Added util/ratelimit.pl
3882 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3883 clock is set back in time.
3885 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3886 Candler <B.Candler@pobox.com>.
3888 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3889 OK for addresses that are the result of redirections.
3891 PH/48 A number of further additions for the benefit of the new test suite,
3892 including a fake gethostbyname() that interfaces to the fake DNS resolver
3895 TF/06 The fix for widen_domains has also been applied to qualify_single and
3896 search_parents which are the other dnslookup options that can cause
3899 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3902 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3904 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3905 service" if it encountered an MX record with an empty target hostname.
3906 The message is now "an MX or SRV record indicated no SMTP service".
3908 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3909 if the DNS of the sending site is misconfigured. This is quite a
3910 common situation. This change restores the behaviour of treating a
3911 helo verification defer as a failure.
3913 PH/51 If self=fail was set on a router, the bounce message did not include the
3914 actual error message.
3920 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3922 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3923 command itself ended in error, the underlying message about the transport
3924 filter timeout was being overwritten with the pipe command error. Now the
3925 underlying error message should be appended to the second error message.
3927 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3929 PH/02 Reduce the amount of output that "make" produces by default. Full output
3930 can still be requested.
3932 PH/03 The warning log line about a condition test deferring for a "warn" verb
3933 was being output only once per connection, rather than after each
3934 occurrence (because it was using the same function as for successful
3935 "warn" verbs). This seems wrong, so I have changed it.
3937 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3938 it should not have, which might have caused a crash in the right
3939 circumstances, but probably never did.
3941 PH/04 Installed a modified version of Tony Finch's patch to make submission
3942 mode fix the return path as well as the Sender: header line, and to
3943 add a /name= option so that you can make the user's friendly name appear
3946 TF/03 Added the control = fakedefer ACL modifier.
3948 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3949 Mark Lowes for thorough testing.
3951 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3953 TK/03 Merged latest SRS patch from Miles Wilton.
3955 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3956 with the definition in sysexits.h (which is #included earlier).
3957 Fortunately, Exim does not actually use EX_OK. The code used to try to
3958 preserve the sysexits.h value, by assuming that macro definitions were
3959 scanned for macro replacements. I have been disabused of this notion,
3960 so now the code just undefines EX_OK before #including unistd.h.
3962 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3963 in the smtp transport. When a block could not be written in a single
3964 write() function, the timeout was being re-applied to each part-write.
3965 This seems wrong - if the receiver was accepting one byte at a time it
3966 would take for ever. The timeout is now adjusted when this happens. It
3967 doesn't have to be particularly precise.
3969 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3970 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3972 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3973 without a domain if the check to <postmaster@domain> fails.
3975 SC/01 Eximstats: added -xls and the ability to specify output files
3976 (patch written by Frank Heydlauf).
3978 SC/02 Eximstats: use FileHandles for outputting results.
3980 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3982 SC/04 Eximstats: fixed display of large numbers with -nvr option
3984 SC/05 Eximstats: fixed merging of reports with empty tables.
3986 SC/06 Eximstats: added the -include_original_destination flag
3988 SC/07 Eximstats: removed tabs and trailing whitespace.
3990 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3992 TK/06 MBOX spool code: Add real "From " MBOX separator line
3993 so the .eml file is really in mbox format (even though
3994 most programs do not really care). Patch from Alex Miller.
3996 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3997 The latter is generated from $received_to and is only set if the
3998 message has one envelope recipient. SA can use these headers,
3999 obviously out-of-the-box. Patch from Alex Miller.
4001 PH/08 The ${def test on a variable was returning false if the variable's
4002 value was "0", contrary to what the specification has always said!
4003 The result should be true unless the variable is empty.
4005 PH/09 The syntax error of a character other than { following "${if
4006 def:variable_name" (after optional whitespace) was not being diagnosed.
4007 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4008 accidental colon was present, for example, could give incorrect results.
4010 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4011 result is used (not including appendfile, where other changes are about
4014 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4015 This involved changing a lot of size variables from int to off_t. It
4016 should work with maildirs and everything.
4018 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4019 spamd dying while we are connected to it.
4021 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4024 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4025 function for BDB 4.3.
4027 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4029 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4030 the sources for PCRE have been re-arranged and more files are now
4033 PH/15 The code I had for printing potentially long long variables in PH/11
4034 above was not the best (it lost precision). The length of off_t variables
4035 is now inspected at build time, and an appropriate printing format (%ld
4036 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4037 to be "long long int" or "long int". This is needed for the internal
4038 formatting function string_vformat().
4040 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4041 the configuration file to be ":syslog", then the script "guesses" where
4042 the logs files are, rather than using the compiled in default. In our
4043 case the guess is not the same as the compiled default, so the script
4044 suddenly stopped working when I started to use syslog. The patch checks
4045 to see if log_file_path is "". If so, it attempts to read it from exim
4046 with no configuration file to get the compiled in version, before it
4047 falls back to the previous guessing code."
4049 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4050 implementing BATV in an Exim configuration. See NewStuff for the gory
4053 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4054 Makefile that are specific to HP-UX.
4056 PH/18 If the "use_postmaster" option was set for a recipient callout together
4057 with the "random" option, the postmaster address was used as the MAIL
4058 FROM address for the random test, but not for the subsequent recipient
4059 test. It is now used for both.
4061 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4062 patch removes a few documentation additions to RFC 3028, because the
4063 latest draft now contains them. It adds the new en;ascii-case comparator
4064 and a new error check for 8bit text in MIME parts. Comparator and
4065 require names are now matched exactly. I enabled the subaddress
4066 extension, but it is not well tested yet (read: it works for me)."
4068 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4069 rework some of the code of TK/09 above to avoid the hardwired use of
4070 "%lld" and "long long". Replaced the call to snprintf() with a call to
4073 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4074 records point to non-existent hosts", "retry timeout exceeded", and
4075 "retry time not reached for any host after a long failure period".
4077 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4078 experimental DomainKeys support:
4080 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4081 (2) On an error such as an illegally used "control", the wrong name for
4082 the control was given.
4084 These problems did NOT occur unless DomainKeys support was compiled.
4086 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4088 PH/24 Added ${if match_ip condition.
4090 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4091 databases so that it will be absolutely obvious if a crash occurs in the
4092 DB library. This is a regular occurrence (often caused by mis-matched
4095 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4096 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4097 picked up on a user's system that detects such things. There doesn't seem
4098 to be a gcc warning option for this - only an attribute that has to be
4099 put on the function's prototype. It seems that in Fedora Core 4 they have
4100 set this on a number of new functions. No doubt there will be more in due
4103 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4104 specify a transport. However, if an address that was verified by such a
4105 router was the subject of a callout, Exim crashed because it tried to
4106 read the rcpt_include_affixes from the non-existent transport. Now it
4107 just assumes that the setting of that option is false. This bug was
4108 introduced by 4.51/PH/31.
4110 PH/28 Changed -d+all to exclude +memory, because that information is very
4111 rarely of interest, but it makes the output a lot bigger. People tend to
4112 do -d+all out of habit.
4114 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4115 code in os-type was giving problems when libc.so lives in lib64, like on
4118 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4119 aren't the modern standard, and it seems that some systems' include files
4120 don't always have them. Exim was already checking for some of the newer
4121 ones like T_AAAA, and defining it itself. I've added checks for all the
4122 record types that Exim uses.
4124 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4125 not automatically generating a new one, as it is supposed to. This
4126 prevented TLS from working. If the file did exist, but contained invalid
4127 data, a new version was generated, as expected. It was only the case of a
4128 non-existent file that was broken.
4130 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4131 with a change in libdomainkeys > 0.64.
4133 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4134 from DNS. If the selector record carries the flag, it now has
4135 precedence over the domain-wide flag.
4137 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4139 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4140 the use of a port name that isn't defined in /etc/services) occurred, the
4141 message was deferred as in a normal delivery, and thus remained on the
4142 spool, instead of being failed because of the mua_wrapper setting. This
4143 is now fixed, and I tidied up some of the mua_wrapper messages at the
4146 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4147 the messages in a hash of arrays rather than using individual hashes.
4148 This is a bit cleaner and results in dramatic memory savings, albeit
4149 at a slight CPU cost.
4151 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4152 as requested by Marc Sherman.
4154 SC/10 Eximstats: added histograms for user specified patterns as requested
4157 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4159 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4160 fopen() in the content-scanning modules that did not already have it.
4166 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4167 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4169 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4171 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4173 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4174 to test Sieve filters that use "vacation".
4176 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4177 that changes the way the GnuTLS parameters are stored in the cache file.
4178 The new format can be generated externally. For backward compatibility,
4179 if the data in the cache doesn't make sense, Exim assumes it has read an
4180 old-format file, and it generates new data and writes a new file. This
4181 means that you can't go back to an older release without removing the
4184 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4185 work if there are any delivery delays because "one_time" forces the
4186 parent to be marked "delivered", so its unseen clone is never tried
4187 again. For this reason, Exim now forbids the simultaneous setting of
4190 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4191 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4192 entry if you want to know the details. The fix, however, neglected to
4193 consider the case when local delivery batching is involved. The test for
4194 "previously delivered" was not happening when checking to see if an
4195 address could be batched with a previous (undelivered) one; under
4196 certain circumstances this could lead to multiple deliveries to the same
4199 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4200 in its include files, and this causes problems building Exim.
4202 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4203 header_syntax) but Exim was just ignoring anything given after a slash.
4204 In particular, this caused confusion with an attempt to use "verify =
4205 reverse_host_lookup/defer_ok". An error is now given when options are
4206 supplied for verify items that do not have them. (Maybe reverse_host_
4207 lookup should have a defer_ok option, but that's a different point.)
4209 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4210 defined by RFC 821) to 2048, because there were problems with some AUTH
4211 commands, and RFC 1869 says the size should be increased for extended
4212 SMTP commands that take arguments.
4214 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4217 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4218 "unknown" error; now it says that the functionality isn't in the binary.
4220 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4221 an address' error message when a string expansion fails (syntax or
4222 whatever). Otherwise the password may appear in the log. Following change
4223 PH/42 below, there is no longer a chance of it appearing in a bounce
4226 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4228 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4229 of its addresses was ever tried. (Bugzilla bug #2.)
4231 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4232 the result incorrectly in the debug output. (It correctly added a newline
4233 to what was transported.)
4235 TF/01 Added $received_time.
4237 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4238 commented out examples of how to interface to a virus scanner and to
4239 SpamAssassin. Also added commented examples of av_scanner and
4240 spamd_address settings.
4242 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4243 and controls are allowed in which ACLs. There were a couple of minor
4244 errors. Some of the entries in the conditions table (which is a table of
4245 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4246 negation of where the condition IS allowed.
4248 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4250 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4251 header file does not have a version number, so I've had to invent a new
4252 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4253 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4254 radiusclient), but it was contributed by a Radius user.
4256 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4257 files or directories, for OpenSSL.
4259 PH/19 When an Exim process that is running as root has to create an Exim log
4260 file, it does so in a subprocess that runs as exim:exim so as to get the
4261 ownership right at creation (otherwise, other Exim processes might see
4262 the file with the wrong ownership). There was no test for failure of this
4263 fork() call, which would lead to the process getting stuck as it waited
4264 for a non-existent subprocess. Forks do occasionally fail when resources
4265 run out. I reviewed all the other calls to fork(); they all seem to check
4268 PH/20 When checking for unexpected SMTP input at connect time (before writing
4269 the banner), Exim was not dealing correctly with a non-positive return
4270 from the read() function. If the client had disconnected by this time,
4271 the result was a log entry for a synchronization error with an empty
4272 string after "input=" when read() returned zero. If read() returned -1
4273 (an event I could not check), uninitialized data bytes were printed.
4274 There were reports of junk text (parts of files, etc) appearing after
4277 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4279 PH/22 Added support for macro redefinition, and (re)definition in between
4280 driver and ACL definitions.
4282 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4283 forgetting to use the resulting value; it was using the unexpanded value.
4285 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4286 hadn't been configured. The fix is from Juergen Kreileder, who
4287 understands it better than I do:
4289 "Here's what I see happening with three configured cyrus_sasl
4290 authenticators configured (plain, login, cram-md5):
4292 On startup auth_cyrus_sasl_init() gets called for each of these.
4293 This means three calls to sasl_listmech() without a specified mech_list.
4294 => SASL tests which mechs of all available mechs actually work
4295 => three warnings about OTP not working
4296 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4298 With the patch, sasl_listmech() also gets called three times. But now
4299 SASL's mech_list option is set to the server_mech specified in the the
4300 authenticator. Or in other words, the answer from sasl_listmech()
4301 gets limited to just the mech you're testing for (which is different
4303 => the return list contains just 'plain' or 'login', 'cram-md5' or
4304 nothing depending on the value of ob->server_mech.
4306 I've just tested the patch: Authentication still works fine,
4307 unavailable mechs specified in the exim configuration are still
4308 caught, and the auth.log warnings about OTP are gone."
4310 PH/25 When debugging is enabled, the contents of the command line are added
4311 to the debugging output, even when log_selector=+arguments is not
4314 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4315 answer is "GNU", and only if the return is "GNU/something" is the answer
4318 PH/27 $acl_verify_message is now set immediately after the failure of a
4319 verification in an ACL, and so is available in subsequent modifiers. In
4320 particular, the message can be preserved by coding like this:
4322 warn !verify = sender
4323 set acl_m0 = $acl_verify_message
4325 Previously, $acl_verify_message was set only while expanding "message"
4326 and "log_message" when a very denied access.
4328 PH/28 Modified OS/os.c-Linux with
4330 -#ifndef OS_LOAD_AVERAGE
4331 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4333 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4334 nomenclature these days.)
4336 PH/29 Installed patch from the Sieve maintainer that adds the options
4337 sieve_useraddress and sieve_subaddress to the redirect router.
4339 PH/30 In these circumstances:
4340 . Two addresses routed to the same list of hosts;
4341 . First host does not offer TLS;
4342 . First host accepts first address;
4343 . First host gives temporary error to second address;
4344 . Second host offers TLS and a TLS session is established;
4345 . Second host accepts second address.
4346 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4347 and peerdn, if requested) that were in fact used only for the second
4350 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4351 attention to any local part prefix or suffix that was matched by the
4352 router that accepted the address. It now behaves in the same way as it
4353 does for delivery: the affixes are removed from the local part unless
4354 rcpt_include_affixes is set on the transport.
4356 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4357 timeout during the DATA phase of an incoming message.
4359 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4360 applied a patch sanctioned by the Sieve maintainer.
4362 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4363 the uid or gid is negative. A case of a negative gid caused this to be
4364 noticed. The fix allows for either to be negative.
4366 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4367 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4370 PH/36 The change PH/12 above was broken. Fixed it.
4372 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4373 the grounds that routing the same address twice would always produce the
4374 same answer. This might have been true once, but it is certainly no
4375 longer true now. Routing a child address may depend on the previous
4376 routing that produced that child. Some complicated redirection strategies
4377 went wrong when messages had multiple recipients, and made Exim's
4378 behaviour dependent on the order in which the addresses were given.
4380 I have moved the duplicate checking until after the routing is complete.
4381 Exim scans the addresses that are assigned to local and remote
4382 transports, and removes any duplicates. This means that more work will be
4383 done, as duplicates will always all be routed, but duplicates are
4384 presumably rare, so I don't expect this is of any significance.
4386 For deliveries to pipes, files, and autoreplies, the duplicate checking
4387 still happens during the routing process, since they are not going to be
4390 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4391 It corrects a timeout issue with spamd. This is Ian's comment: "The
4392 background is that sometimes spamd either never reads data from a
4393 connection it has accepted, or it never writes response data. The exiscan
4394 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4395 blindly assumes that writes won't block so it may never time out."
4397 PH/39 Allow G after quota size as well as K and M.
4399 PH/40 The value set for $authenticated_id in an authenticator may not contain
4400 binary zeroes or newlines because the value is written to log lines and
4401 to spool files. There was no check on this. Now the value is run through
4402 the string_printing() function so that such characters are converted to
4403 printable escape sequences.
4405 PH/41 $message_linecount is a new variable that contains the total number of
4406 lines in the message. Compare $body_linecount, which is the count for the
4409 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4410 bounce and delay warning messages, except in certain special cases, which
4413 (a) An SMTP error message from a remote host;
4414 (b) A message specified in a :fail: redirection;
4415 (c) A message specified in a "fail" command in a system filter;
4416 (d) A message specified in a FAIL return from the queryprogram router;
4417 (e) A message specified by the cannot_route_message router option.
4419 In these cases only, Exim does include the error details in bounce and
4420 warning messages. There are also a few cases where bland messages such
4421 as "unrouteable address" or "local delivery error" are given.
4423 PH/43 $value is now also set for the "else" part of a ${run expansion.
4425 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4426 being worked on, but at least Exim now implements the latest version to
4429 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4430 process to complete was treated as a delivery failure, a timeout while
4431 writing the message to the pipe was logged, but erroneously treated as a
4432 successful delivery. Such timeouts include transport filter timeouts. For
4433 consistency with the overall process timeout, these timeouts are now
4434 treated as errors, giving rise to delivery failures by default. However,
4435 there is now a new Boolean option for the pipe transport called
4436 timeout_defer, which, if set TRUE, converts the failures into defers for
4437 both kinds of timeout. A transport filter timeout is now identified in
4440 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4441 systems where "make" and "gmake" are different, calling "gmake" at top
4442 level broke things. I've arranged for the value of $(MAKE) to be passed
4443 from the Makefile to this script so that it can call the same version of
4447 A note about Exim versions 4.44 and 4.50
4448 ----------------------------------------
4450 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4451 changes of various kinds. As a consequence, a big documentation update was
4452 needed. This delayed the release for rather longer than seemed good, especially
4453 in the light of a couple of (minor) security issues. Therefore, the changes
4454 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4455 release. So 4.44 and 4.50 are in effect two different branches that both start
4458 I have left the 4.50 change log unchanged; it contains all the changes since
4459 4.43. The change log for 4.44 is below; many of its items are identical to
4460 those for 4.50. This seems to be the most sensible way to preserve the
4461 historical information.
4467 1. Minor wording change to the doc/README.SIEVE file.
4469 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4470 computation of the current number of files was incorrect.
4472 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4473 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4476 4. Give more explanation in the error message when the command for a transport
4477 filter fails to execute.
4479 5. There are several places where Exim runs a non-Exim command in a
4480 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4481 was being done only for the command run by the queryprogram router. It is
4482 now done for all such subprocesses. The other cases are: ${run, transport
4483 filters, and the commands run by the lmtp and pipe transports.
4485 6. Added CONFIGURE_GROUP build-time option.
4487 7. Some older OS have a limit of 256 on the maximum number of file
4488 descriptors. Exim was using setrlimit() to set 1000 as a large value
4489 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4490 systems. I've change it so that if it can't get 1000, it tries for 256.
4492 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4493 was an oversight, and furthermore, ever since the addition of extra
4494 controls (e.g. 4.43/32), the checks on when to allow different forms of
4495 "control" were broken. There should now be diagnostics for all cases when a
4496 control that does not make sense is encountered.
4498 9. Added the /retain_sender option to "control=submission".
4500 10. $recipients is now available in the predata ACL (oversight).
4502 11. Tidy the search cache before the fork to do a delivery from a message
4503 received from the command line. Otherwise the child will trigger a lookup
4504 failure and thereby defer the delivery if it tries to use (for example) a
4505 cached ldap connection that the parent has called unbind on.
4507 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4508 of $address_data from the recipient verification was clobbered by the
4509 sender verification.
4511 13. The value of address_data from a sender verification is now available in
4512 $sender_address_data in subsequent conditions in the ACL statement.
4514 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4516 15. Added a new option "connect=<time>" to callout options, to set a different
4519 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4520 was its contents. (It was OK if the option was not defined at all.)
4522 17. A "Completed" log line is now written for messages that are removed from
4523 the spool by the -Mrm option.
4525 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4526 information about exactly what failed.
4528 19. Added -dd to debug only the daemon process.
4530 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4531 handles timeouts, both on the server side and network timeouts. Renamed the
4532 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4534 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4535 It is now set to "smtps".
4537 22. $host_address is now set to the target address during the checking of
4538 ignore_target_hosts.
4540 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4541 being passed; this would have caused $sender_host_name to have been used if
4542 matching the list had actually called for a host name (not very likely,
4543 since this list is usually IP addresses). A host name is now passed as
4546 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4547 code that specifies a non-restarting handler (typically sigaction() in
4548 modern systems) in an attempt to fix a rare and obscure crash bug.
4550 25. Narrowed the window for a race in the daemon that could cause it to ignore
4551 SIGCHLD signals. This is not a major problem, because they are used only to
4552 wake it up if nothing else does.
4554 26. A malformed maildirsize file could cause Exim to calculate negative values
4555 for the mailbox size or file count. Odd effects could occur as a result.
4556 The maildirsize information is now recalculated if the size or filecount
4559 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4560 support for a long time. Removed HAVE_SYS_VFS_H.
4562 28. Installed the latest version of exipick from John Jetmore.
4564 29. In an address list, if the pattern was not a regular expression, an empty
4565 subject address (from a bounce message) matched only if the pattern was an
4566 empty string. Non-empty patterns were not even tested. This was the wrong
4567 because it is perfectly reasonable to use an empty address as part of a
4568 database query. An empty address is now tested by patterns that are
4569 lookups. However, all the other forms of pattern expect the subject to
4570 contain a local part and a domain, and therefore, for them, an empty
4571 address still always fails if the pattern is not itself empty.
4573 30. Exim went into a mad DNS loop when attempting to do a callout where the
4574 host was specified on an smtp transport, and looking it up yielded more
4575 than one IP address.
4577 31. Re-factored the code for checking spool and log partition space into a
4578 function that finds that data and another that does the check. The former
4579 is then used to implement four new variables: $spool_space, $log_space,
4580 $spool_inodes, and $log_inodes.
4582 32. The RFC2047 encoding function was originally intended for short strings
4583 such as real names; it was not keeping to the 75-character limit for
4584 encoded words that the RFC imposes. It now respects the limit, and
4585 generates multiple encoded words if necessary. To be on the safe side, I
4586 have increased the buffer size for the ${rfc2047: expansion operator from
4589 33. It is now permitted to omit both strings after an "if" condition; if the
4590 condition is true, the result is "true". As before, when the second string
4591 is omitted, a false condition yields an empty string. This makes it less
4592 cumbersome to write custom ACL and router conditions.
4594 34. Failure to deliver a bounce message always caused it to be frozen, even if
4595 there was an errors_to setting on the router. The errors_to setting is now
4598 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4599 canonical form (fully expanded) before being placed in
4600 $sender_host_address.
4602 36. The table in the code that translates DNS record types into text (T_A to
4603 "A" for instance) was missing entries for NS and CNAME. It is just possible
4604 that this could have caused confusion if both these types were looked up
4605 for the same domain, because the text type is used as part of Exim's
4606 per-process caching. But the chance of anyone hitting this buglet seems
4609 37. The dnsdb lookup has been extended in a number of ways.
4611 (1) There is a new type, "zns", which walks up the domain tree until it
4612 finds some nameserver records. It should be used with care.
4614 (2) There is a new type, "mxh", which is like "mx" except that it returns
4615 just the host names, not the priorities.
4617 (3) It is now possible to give a list of domains (or IP addresses) to be
4618 looked up. The behaviour when one of the lookups defers can be
4619 controlled by a keyword.
4621 (4) It is now possible to specify the separator character for use when
4622 multiple records are returned.
4624 38. The dnslists ACL condition has been extended: it is now possible to supply
4625 a list of IP addresses and/or domains to be looked up in a particular DNS
4628 39. Added log_selector=+queue_time_overall.
4630 40. When running the queue in the test harness, wait just a tad after forking a
4631 delivery process, to get repeatability of debugging output.
4633 41. Include certificate and key file names in error message when GnuTLS fails
4634 to set them up, because the GnuTLS error message doesn't include the name
4635 of the failing file when there is a problem reading it.
4637 42. Allow both -bf and -bF in the same test run.
4639 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4641 44. The "Exiscan patch" is now merged into the mainline Exim source.
4643 45. Sometimes the final signoff response after QUIT could fail to get
4644 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4645 < 0 before doing a fflush(). This bug looks as though it goes back to the
4646 introduction of TLS in release 3.20, but "sometimes" must have been rare
4647 because the tests only now provoked it.
4649 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4650 (this can affect the format of dates).
4652 47. exim_tidydb, when checking for the continued existence of a message for
4653 which it has found a message-specific retry record, was not finding
4654 messages that were in split spool directories. Consequently, it was
4655 deleting retry records that should have stayed in existence.
4657 48. Steve fixed some bugs in eximstats.
4659 49. The SPA authentication driver was not abandoning authentication and moving
4660 on to the next authenticator when an expansion was forced to fail,
4661 contradicting the general specification for all authenticators. Instead it
4662 was generating a temporary error. It now behaves as specified.
4664 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4665 (the order specifies the preference for clients). The order is now AES256,
4666 AES128, 3DES, ARCFOUR128.
4668 51. Small patch to Sieve code - explicitly set From: when generating an
4671 52. Exim crashed if a remote delivery caused a very long error message to be
4672 recorded - for instance if somebody sent an entire SpamAssassin report back
4673 as a large number of 550 error lines. This bug was coincidentally fixed by
4674 increasing the size of one of Exim's internal buffers (big_buffer) that
4675 happened as part of the Exiscan merge. However, to be on the safe side, I
4676 have made the code more robust (and fixed the comments that describe what
4679 53. Now that there can be additional text after "Completed" in log lines (if
4680 the queue_time_overall log selector is set), a one-byte patch to exigrep
4681 was needed to allow it to recognize "Completed" as not the last thing in
4684 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4685 patch that reportedly fixes this has been added. I am not expert enough to
4686 create a test for it. This is what the patch creator wrote:
4688 "I found a little strange behaviour of ldap code when working with
4689 Windows 2003 AD Domain, where users was placed in more than one
4690 Organization Units. When I tried to give exim partial DN, the exit code
4691 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4692 But simultaneously result of request was absolutely normal ldap result,
4693 so I produce this patch..."
4695 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4696 REFERENCE, so I have modified the code to exclude the patch when that macro
4699 55. Some experimental protocols are using DNS PTR records for new purposes. The
4700 keys for these records are domain names, not reversed IP addresses. The
4701 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4702 leaves it alone. Component reversal etc. now happens only for IP addresses.
4705 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4707 57. Double the size of the debug message buffer (to 2048) so that more of very
4708 long debug lines gets shown.
4710 58. The exicyclog utility now does better if the number of log files to keep
4711 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4713 59. Two changes related to the smtp_active_hostname option:
4715 (1) $smtp_active_hostname is now available as a variable.
4716 (2) The default for smtp_banner uses $smtp_active_hostname instead
4717 of $primary_hostname.
4719 60. The host_aton() function is supposed to be passed a string that is known
4720 to be a valid IP address. However, in the case of IPv6 addresses, it was
4721 not checking this. This is a hostage to fortune. Exim now panics and dies
4722 if the condition is not met. A case was found where this could be provoked
4723 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4724 components; fortuitously, this particular loophole had already been fixed
4725 by change 4.50/55 above.
4727 If there are any other similar loopholes, the new check in host_aton()
4728 itself should stop them being exploited. The report I received stated that
4729 data on the command line could provoke the exploit when Exim was running as
4730 exim, but did not say which command line option was involved. All I could
4731 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4732 running as the user.
4735 61. There was a buffer overflow vulnerability in the SPA authentication code
4736 (which came originally from the Samba project). I have added a test to the
4737 spa_base64_to_bits() function which I hope fixes it.
4740 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4741 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4743 63. The daemon start-up calls getloadavg() while still root for those OS that
4744 need the first call to be done as root, but it missed one case: when
4745 deliver_queue_load_max is set with deliver_drop_privilege. This is
4746 necessary for the benefit of the queue runner, because there is no re-exec
4747 when deliver_drop_privilege is set.
4749 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4750 This has been fixed.
4752 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4753 was in use, was not putting the data itself into the right store pool;
4754 consequently, it could be overwritten for a subsequent message in the same
4755 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4758 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4760 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4761 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4762 the function were treating the return as a boolean value, which happened to
4763 work because 0=false and not-0=true, but is not correct code.
4765 68. The host_aton() function was not handling scoped IPv6 addresses (those
4766 with, for example, "%eth0" on the end) correctly.
4768 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4769 negated items (that is, ~something) in unsigned ints. Some compilers
4770 apparently mutter when there is no cast.
4772 70. If an address verification called from an ACL failed, and did not produce a
4773 user-specific message (i.e. there was only a "system" message), nothing was
4774 put in $acl_verify_message. In this situation, it now puts the system
4777 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4778 session; change 4.31/43 added the unwanted input to the log line - except
4779 that it did not do this in the start of session case. It now does.
4781 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4782 This is wrong and can cause the other end to generate a synchronization
4783 error if it is another Exim or anything else that does the synchronization
4784 check. A QUIT command is no longer sent after a timeout.
4786 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4787 during host lookups.
4789 74. The defer_ok option of callout verification was not working if it was used
4790 when verifying addresses in header lines, that is, for this case:
4792 verify = header_sender/callout=defer_ok
4794 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4795 those file descriptors could be used for SMTP connections. If anything
4796 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4797 could be sent to the SMTP client, causing chaos. The daemon now opens
4798 stdin, stdout, and stderr to /dev/null when it puts itself into the
4801 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4802 log by default. The user can override this with suitable Perl magic.
4804 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4805 the log message when discard triggers, was not working for the DATA ACL or
4806 for the non-SMTP ACL.
4808 78. Error message wording change in sieve.c.
4810 79. If smtp_accept_max_per_host was set, the number of connections could be
4811 restricted to fewer than expected, because the daemon was trying to set up
4812 a new connection before checking whether the processes handling previous
4813 connections had finished. The check for completed processes is now done
4814 earlier. On busy systems, this bug wouldn't be noticed because something
4815 else would have woken the daemon, and it would have reaped the completed
4818 80. If a message was submitted locally by a user whose login name contained one
4819 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4820 It caused a spool format error. I have fixed the spool reading code. A
4821 related problem was that the "from" clause in the Received: line became
4822 illegal because of the space(s). It is now covered by ${quote_local_part.
4824 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4827 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4829 83. Patch from Sieve maintainer to fix unterminated string problem in
4830 "vacation" handling.
4832 84. Some minor changes to the Linux configuration files to help with other
4833 OS variants using glibc.
4835 85. One more patch for Sieve to update vacation handling to latest spec.
4838 ----------------------------------------------------
4839 See the note above about the 4.44 and 4.50 releases.
4840 ----------------------------------------------------
4846 1. Change 4.43/35 introduced a bug that caused file counts to be
4847 incorrectly computed when quota_filecount was set in an appendfile
4850 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4851 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4854 3. Give more explanation in the error message when the command for a transport
4855 filter fails to execute.
4857 4. There are several places where Exim runs a non-Exim command in a
4858 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4859 was being done only for the command run by the queryprogram router. It is
4860 now done for all such subprocesses. The other cases are: ${run, transport
4861 filters, and the commands run by the lmtp and pipe transports.
4863 5. Some older OS have a limit of 256 on the maximum number of file
4864 descriptors. Exim was using setrlimit() to set 1000 as a large value
4865 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4866 systems. I've change it so that if it can't get 1000, it tries for 256.
4868 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4869 was an oversight, and furthermore, ever since the addition of extra
4870 controls (e.g. 4.43/32), the checks on when to allow different forms of
4871 "control" were broken. There should now be diagnostics for all cases when a
4872 control that does not make sense is encountered.
4874 7. $recipients is now available in the predata ACL (oversight).
4876 8. Tidy the search cache before the fork to do a delivery from a message
4877 received from the command line. Otherwise the child will trigger a lookup
4878 failure and thereby defer the delivery if it tries to use (for example) a
4879 cached ldap connection that the parent has called unbind on.
4881 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4882 of $address_data from the recipient verification was clobbered by the
4883 sender verification.
4885 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4886 was its contents. (It was OK if the option was not defined at all.)
4888 11. A "Completed" log line is now written for messages that are removed from
4889 the spool by the -Mrm option.
4891 12. $host_address is now set to the target address during the checking of
4892 ignore_target_hosts.
4894 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4895 being passed; this would have caused $sender_host_name to have been used if
4896 matching the list had actually called for a host name (not very likely,
4897 since this list is usually IP addresses). A host name is now passed as
4900 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4901 code that specifies a non-restarting handler (typically sigaction() in
4902 modern systems) in an attempt to fix a rare and obscure crash bug.
4904 15. Narrowed the window for a race in the daemon that could cause it to ignore
4905 SIGCHLD signals. This is not a major problem, because they are used only to
4906 wake it up if nothing else does.
4908 16. A malformed maildirsize file could cause Exim to calculate negative values
4909 for the mailbox size or file count. Odd effects could occur as a result.
4910 The maildirsize information is now recalculated if the size or filecount
4913 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4914 support for a long time. Removed HAVE_SYS_VFS_H.
4916 18. Updated exipick to current release from John Jetmore.
4918 19. Allow an empty sender to be matched against a lookup in an address list.
4919 Previously the only cases considered were a regular expression, or an
4922 20. Exim went into a mad DNS lookup loop when doing a callout where the
4923 host was specified on the transport, if the DNS lookup yielded more than
4926 21. The RFC2047 encoding function was originally intended for short strings
4927 such as real names; it was not keeping to the 75-character limit for
4928 encoded words that the RFC imposes. It now respects the limit, and
4929 generates multiple encoded words if necessary. To be on the safe side, I
4930 have increased the buffer size for the ${rfc2047: expansion operator from
4933 22. Failure to deliver a bounce message always caused it to be frozen, even if
4934 there was an errors_to setting on the router. The errors_to setting is now
4937 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4938 canonical form (fully expanded) before being placed in
4939 $sender_host_address.
4941 24. Updated eximstats to version 1.33
4943 25. Include certificate and key file names in error message when GnuTLS fails
4944 to set them up, because the GnuTLS error message doesn't include the name
4945 of the failing file when there is a problem reading it.
4947 26. Expand error message when OpenSSL has problems setting up cert/key files.
4950 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4951 (this can affect the format of dates).
4953 28. exim_tidydb, when checking for the continued existence of a message for
4954 which it has found a message-specific retry record, was not finding
4955 messages that were in split spool directories. Consequently, it was
4956 deleting retry records that should have stayed in existence.
4958 29. eximstats updated to version 1.35
4959 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4960 1.35 - bugfix such that pie charts by volume are generated correctly
4962 30. The SPA authentication driver was not abandoning authentication and moving
4963 on to the next authenticator when an expansion was forced to fail,
4964 contradicting the general specification for all authenticators. Instead it
4965 was generating a temporary error. It now behaves as specified.
4967 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4968 (the order specifies the preference for clients). The order is now AES256,
4969 AES128, 3DES, ARCFOUR128.
4971 31. Small patch to Sieve code - explicitly set From: when generating an
4974 32. Exim crashed if a remote delivery caused a very long error message to be
4975 recorded - for instance if somebody sent an entire SpamAssassin report back
4976 as a large number of 550 error lines. This bug was coincidentally fixed by
4977 increasing the size of one of Exim's internal buffers (big_buffer) that
4978 happened as part of the Exiscan merge. However, to be on the safe side, I
4979 have made the code more robust (and fixed the comments that describe what
4982 33. Some experimental protocols are using DNS PTR records for new purposes. The
4983 keys for these records are domain names, not reversed IP addresses. The
4984 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4985 leaves it alone. Component reversal etc. now happens only for IP addresses.
4988 34. The host_aton() function is supposed to be passed a string that is known
4989 to be a valid IP address. However, in the case of IPv6 addresses, it was
4990 not checking this. This is a hostage to fortune. Exim now panics and dies
4991 if the condition is not met. A case was found where this could be provoked
4992 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4993 components; fortuitously, this particular loophole had already been fixed
4994 by change 4.50/55 or 4.44/33 above.
4996 If there are any other similar loopholes, the new check in host_aton()
4997 itself should stop them being exploited. The report I received stated that
4998 data on the command line could provoke the exploit when Exim was running as
4999 exim, but did not say which command line option was involved. All I could
5000 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
5001 running as the user.
5004 35. There was a buffer overflow vulnerability in the SPA authentication code
5005 (which came originally from the Samba project). I have added a test to the
5006 spa_base64_to_bits() function which I hope fixes it.
5009 36. The daemon start-up calls getloadavg() while still root for those OS that
5010 need the first call to be done as root, but it missed one case: when
5011 deliver_queue_load_max is set with deliver_drop_privilege. This is
5012 necessary for the benefit of the queue runner, because there is no re-exec
5013 when deliver_drop_privilege is set.
5015 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5016 was in use, was not putting the data itself into the right store pool;
5017 consequently, it could be overwritten for a subsequent message in the same
5018 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5021 38. Sometimes the final signoff response after QUIT could fail to get
5022 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5023 < 0 before doing a fflush(). This bug looks as though it goes back to the
5024 introduction of TLS in release 3.20, but "sometimes" must have been rare
5025 because the tests only now provoked it.
5031 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5032 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5033 Now it returns an integer. A number of calls were still expecting a T/F
5034 return. Fortuitously, in all cases, the tests worked in OK situations,
5035 which is the norm. However, things would have gone wrong on any write
5036 failures on the smtp file descriptor. This function is used when sending
5037 messages over SMTP and also when doing verify callouts.
5039 2. When Exim is called to do synchronous delivery of a locally submitted
5040 message (the -odf or -odi options), it no longer closes stderr before doing
5043 3. Implemented the mua_wrapper option.
5045 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5047 5. Implemented the functions header_remove(), header_testname(),
5048 header_add_at_position(), and receive_remove_recipient(), and exported them
5051 6. If an ACL "warn" statement specified the addition of headers, Exim already
5052 inserted X-ACL-Warn: at the start if there was no header name. However, it
5053 was not making this test for the second and subsequent header lines if
5054 there were newlines in the string. This meant that an invalid header could
5055 be inserted if Exim was badly configured.
5057 7. Allow an ACL "warn" statement to add header lines at the start or after all
5058 the Received: headers, as well as at the end.
5060 8. Added the rcpt_4xx retry error code.
5062 9. Added postmaster_mailfrom=xxx to callout verification option.
5064 10. Added mailfrom=xxxx to the callout verification option, for verify=
5067 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5068 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5070 12. Inserted some casts to stop certain compilers warning when using pointer
5071 differences as field lengths or precisions in printf-type calls (mostly
5072 affecting debugging statements).
5074 13. Added optional readline() support for -be (dynamically loaded).
5076 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5077 same clock tick as a message's arrival, so that its received time was the
5078 same as the "first fail" time on the retry record, and that message
5079 remained on the queue past the ultimate address timeout, every queue runner
5080 would try a delivery (because it was past the ultimate address timeout) but
5081 after another failure, the ultimate address timeout, which should have then
5082 bounced the address, did not kick in. This was a "< instead of <=" error;
5083 in most cases the first failure would have been in the next clock tick
5084 after the received time, and all would be well.
5086 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5087 being recognized when the domain list was tested by the match_domain
5088 condition in an expansion string.
5090 16. Added the ${str2b64: operator.
5092 17. Exim was always calling setrlimit() to set a large limit for the number of
5093 processes, without checking whether the existing limit was already
5094 adequate. (It did check for the limit on file descriptors.) Furthermore,
5095 errors from getrlimit() and setrlimit() were being ignored. Now they are
5096 logged to the main and panic logs, but Exim does carry on, to try to do its
5097 job under whatever limits there are.
5099 18. Imported PCRE 5.0.
5101 19. Trivial typo in log message " temporarily refused connection" (the leading
5104 20. If the log selector return_path_on_delivery was set and an address was
5105 redirected to /dev/null, the delivery process crashed because it assumed
5106 that a return path would always be set for a "successful" delivery. In this
5107 case, the whole delivery is bypassed as an optimization, and therefore no
5110 21. Internal re-arrangement: the function for sending a challenge and reading
5111 a response while authentication was assuming a zero-terminated challenge
5112 string. It's now changed to take a pointer and a length, to allow for
5113 binary data in such strings.
5115 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5117 23. Exim was not respecting finduser_retries when seeking the login of the
5118 uid under which it was called; it was always trying 10 times. (The default
5119 setting of finduser_retries is zero.) Also, it was sleeping after the final
5120 failure, which is pointless.
5122 24. Implemented tls_on_connect_ports.
5124 25. Implemented acl_smtp_predata.
5126 26. If the domain in control=submission is set empty, Exim assumes that the
5127 authenticated id is a complete email address when it generates From: or
5128 Sender: header lines.
5130 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5131 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5132 chown and chgrp in /bin and hostname in /usr/bin.
5134 28. Exim was keeping the "process log" file open after each use, just as it
5135 does for the main log. This opens the possibility of it remaining open for
5136 long periods when the USR1 signal hits a daemon. Occasional processlog
5137 errors were reported, that could have been caused by this. Anyway, it seems
5138 much more sensible not to leave this file open at all, so that is what now
5141 29. The long-running daemon process does not normally write to the log once it
5142 has entered its main loop, and it closes the log before doing so. This is
5143 so that log files can straightforwardly be renamed and moved. However,
5144 there are a couple of unusual error situations where the daemon does write
5145 log entries, and I had neglected to close the log afterwards.
5147 30. The text of an SMTP error response that was received during a remote
5148 delivery was being truncated at 512 bytes. This is too short for some of
5149 the long messages that one sometimes sees. I've increased the limit to
5152 31. It is now possible to make retry rules that apply only when a message has a
5153 specific sender, in particular, an empty sender.
5155 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5156 it possible to be selective about when SMTP synchronization is enforced.
5158 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5160 32. Implemented hosts_connection_nolog.
5162 33. Added an ACL for QUIT.
5164 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5167 35. Added mailbox_size and mailbox_filecount to appendfile.
5169 36. Added control = no_multiline_responses to ACLs.
5171 37. There was a bug in the logic of the code that waits for the clock to tick
5172 in the case where the clock went backwards by a substantial amount such
5173 that the microsecond fraction of "now" was more than the microsecond
5174 fraction of "then" (but the whole seconds number was less).
5176 38. Added support for the libradius Radius client library this is found on
5177 FreeBSD (previously only the radiusclient library was supported).
5183 1. When certain lookups returned multiple values in the form name=value, the
5184 quoting of the values was not always being done properly. Specifically:
5185 (a) If the value started with a double quote, but contained no whitespace,
5187 (b) If the value contained whitespace other than a space character (i.e.
5188 tabs or newlines or carriage returns) it was not quoted.
5189 This fix has been applied to the mysql and pgsql lookups by writing a
5190 separate quoting function and calling it from the lookup code. The fix
5191 should probably also be applied to nisplus, ibase and oracle lookups, but
5192 since I cannot test any of those, I have not disturbed their existing code.
5194 2. A hit in the callout cache for a specific address caused a log line with no
5195 reason for rejecting RCPT. Now it says "Previous (cached) callout
5196 verification failure".
5198 3. There was an off-by-one bug in the queryprogram router. An over-long
5199 return line was truncated at 256 instead of 255 characters, thereby
5200 overflowing its buffer with the terminating zero. As well as fixing this, I
5201 have increased the buffer size to 1024 (and made a note to document this).
5203 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5204 when Exim is waiting for an SMTP response from a remote server, Exim
5205 restarts its select() call on the socket, thereby resetting its timeout.
5206 This is not a problem when such interrupts are rare. Somebody set up a cron
5207 job to run exiwhat every 2 minutes, which is less than the normal select()
5208 timeout (5 or 10 minutes). This meant that the select() timeout never
5209 kicked in because it was always reset. I have fixed this by comparing the
5210 time when an interrupt arrives with the time at the start of the first call
5211 to select(). If more time than the timeout has elapsed, the interrupt is
5212 treated as a timeout.
5214 5. Some internal re-factoring in preparation for the addition of Sieve
5215 extensions (by MH). In particular, the "personal" test is moved to a
5216 separate function, and given an option for scanning Cc: and Bcc: (which is
5217 not set for Exim filters).
5219 6. When Exim created an email address using the login of the caller as the
5220 local part (e.g. when creating a From: or Sender: header line), it was not
5221 quoting the local part when it contained special characters such as @.
5223 7. Installed new OpenBSD configuration files.
5225 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5226 try to make them clearer.
5228 9. Callout options, other than the timeout value, were being ignored when
5229 verifying sender addresses in header lines. For example, when using
5231 verify = header_sender/callout=no_cache
5233 the cache was (incorrectly) being used.
5235 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5236 only the Cygwin environment.
5238 11. When return_path_on_delivery was set as a log selector, if different remote
5239 addresses in the same message used different return paths and parallel
5240 remote delivery occurred, the wrong values would sometimes be logged.
5241 (Whenever a remote delivery process finished, the return path value from
5242 the most recently started remote delivery process was logged.)
5244 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5245 lines when AUTH and/or TLS are in use. This is the "received protocol"
5246 field. Exim used to use "asmtp" for authenticated SMTP, without any
5247 indication (in the protocol name) for TLS use. Now it follows the RFC and
5248 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5249 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5250 names appear in log lines as well as in Received: header lines.
5252 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5253 extensions, and comparison tests, and to fix some bugs.
5255 14. Changes to the "personal" filter test:
5257 (1) The test was buggy in that it was just doing the equivalent of
5258 "contains" tests on header lines. For example, if a user's address was
5259 anne@some.where, the "personal" test would incorrectly be true for
5261 To: susanne@some.where
5263 This test is now done by extracting each address from the header in turn,
5264 and checking the entire address. Other tests that are part of "personal"
5265 are now done using regular expressions (for example, to check local parts
5266 of addresses in From: header lines).
5268 (2) The list of non-personal local parts in From: addresses has been
5269 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5270 taken from the Sieve specification recommendations.
5272 (3) If the message contains any header line starting with "List-" it is
5273 treated as non-personal.
5275 (4) The test for "circular" in the Subject: header line has been removed
5276 because it now seems ill-conceived.
5278 15. Minor typos in src/EDITME comments corrected.
5280 16. Installed latest exipick from John Jetmore.
5282 17. If headers_add on a router specified a text string that was too long for
5283 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5284 of string_sprintf() is now avoided.
5286 18. $message_body_size was not set (it was always zero) when running the DATA
5287 ACL and the local_scan() function.
5289 19. For the "mail" command in an Exim filter, no default was being set for
5290 the once_repeat time, causing a random time value to be used if "once" was
5291 specified. (If the value happened to be <= 0, no repeat happened.) The
5292 default is now 0s, meaning "never repeat". The "vacation" command was OK
5293 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5294 (I found it when inspecting the code).
5296 20. There is now an overall timeout for performing a callout verification. It
5297 defaults to 4 times the callout timeout, which applies to individual SMTP
5298 commands during the callout. The overall timeout applies when there is more
5299 than one host that can be tried. The timeout is checked before trying the
5300 next host. This prevents very long delays if there are a large number of
5301 hosts and all are timing out (e.g. when the network connections are timing
5302 out). The value of the overall timeout can be changed by specifying an
5303 additional sub-option for "callout", called "maxwait". For example:
5305 verify = sender/callout=5s,maxwait=20s
5307 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5308 to the end before writing, but this should make it even safer).
5310 22. Exim was forgetting that it had advertised PIPELINING for the second and
5311 subsequent messages on an SMTP connection. It was also not resetting its
5312 memory on STARTTLS and an internal HELO.
5314 23. When Exim logs an SMTP synchronization error within a session, it now
5315 records whether PIPELINING has been advertised or not.
5317 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5318 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5319 rather than long int.
5321 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5323 26. Added the never_mail option to autoreply.
5329 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5330 crash if the getsockname() call failed; this can happen if a connection is
5331 closed very soon after it is established. The problem was simply in the
5332 order in which certain operations were done, causing Exim to try to write
5333 to the SMTP stream before it had set up the file descriptor. The bug has
5334 been fixed by making things happen in the correct order.
5340 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5341 before the connection was closed, thus losing the rejection response.
5343 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5344 some early Solaris releases, but causes trouble in current releases where
5345 socklen_t is defined.
5347 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5350 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5353 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5354 that releasing the top of it at the end releases what was used for sub-
5355 expansions (unless the block got too big). However, discard this block if
5356 the first thing is a variable or header, so that we can use its block when
5357 it is dynamic (useful for very large $message_headers, for example).
5359 6. Lookups now cache *every* query, not just the most recent. A new, separate
5360 store pool is used for this. It can be recovered when all lookup caches are
5361 flushed. Lookups now release memory at the end of their result strings.
5362 This has involved some general refactoring of the lookup sources.
5364 7. Some code has been added to the store_xxx() functions to reduce the amount
5365 of flapping under certain conditions.
5367 8. log_incoming_interface used to affect only the <= reception log lines. Now
5368 it causes the local interface and port to be added to several more SMTP log
5369 lines, for example "SMTP connection from", and rejection lines.
5371 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5373 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5375 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5376 could be overwritten at the end of the current message (or the start of a
5377 new message if it was set in a HELO ACL). The value is now preserved for
5378 the duration of the SMTP connection.
5380 12. If a transport had a headers_rewrite setting, and a matching header line
5381 contained an unqualified address, that address was qualified, even if it
5382 did not match any rewriting rules. The underlying bug was that the values
5383 of the flags that permit the existence of unqualified sender and recipient
5384 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5385 non-local messages, and by -bnq for local messages) were not being
5386 preserved with the message after it was received.
5388 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5389 "next input=" as part of the text comprising the host identity instead of
5390 the correct text. The code was using the same buffer for two different
5391 strings. However, depending on which order the printing function evaluated
5392 its arguments, the bug did not always show up. Under Linux, for example, my
5393 test suite worked just fine.
5395 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5396 doesn't work with some older versions of Perl. It has been changed to "my",
5397 which in any case is probably the better facility to use.
5399 15. A really picky compiler found some instances of statements for creating
5400 error messages that either had too many or two few arguments for the format
5403 16. The size of the buffer for calls to the DNS resolver has been increased
5404 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5405 for addresses that have a lot of PTR records. This alleviates a problem; it
5406 does not fully solve it.
5408 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5409 buffer now truncates the list and logs the incident, which is the same
5410 action as happens when Exim is looking up a host name and its aliases.
5411 Previously in this situation something unpredictable would happen;
5412 sometimes it was "internal error: store_reset failed".
5414 18. If a server dropped the connection unexpectedly when an Exim client was
5415 using GnuTLS and trying to read a response, the client delivery process
5416 crashed while trying to generate an error log message.
5418 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5419 string, for example:
5421 warn message = H1: something\nH2: something
5423 the text was added as a single header line from Exim's point of view
5424 though it ended up OK in the delivered message. However, searching for the
5425 second and subsequent header lines using $h_h2: did not work. This has been
5426 fixed. Similarly, if a system filter added multiple headers in this way,
5427 the routers could not see them.
5429 20. Expanded the error message when iplsearch is called with an invalid key to
5430 suggest using net-iplsearch in a host list.
5432 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5433 ACLs are no longer actually imposed (and a message to that effect is
5436 22. If a "gecos" field in a passwd entry contained escaped characters, in
5437 particular, if it contained a \" sequence, Exim got it wrong when building
5438 a From: or a Sender: header from that name. A second bug also caused
5439 incorrect handling when an unquoted " was present following a character
5440 that needed quoting.
5442 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5443 was not being matched caselessly.
5445 24. Arranged for all hyphens in the exim.8 source to be escaped with
5448 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5449 properly. Recipient callout cache records were still being keyed to include
5450 the sender, even when use_sender was set false. This led to far more
5451 callouts that were necessary. The sender is no longer included in the key
5452 when use_sender is false.
5454 26. Added "control = submission" modifier to ACLs.
5456 27. Added the ${base62d: operator to decode base 62 numbers.
5458 28. dnsdb lookups can now access SRV records.
5460 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5461 the configuration file.
5463 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5464 (-v) mode. This makes the output for a verbose queue run more intelligible.
5466 31. Added a use_postmaster feature to recipient callouts.
5468 32. Added the $body_zerocount variable, containing the number of binary zero
5469 bytes in the message body.
5471 33. The time of last modification of the "new" subdirectory is now used as the
5472 "mailbox time last read" when there is a quota error for a maildir
5475 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5477 35. Added +ignore_unknown as a special item in host lists.
5479 36. Code for decoding IPv6 addresses in host lists is now included, even if
5480 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5481 address was recognized as an IP address, but was then not correctly decoded
5482 into binary, causing unexpected and incorrect effects when compared with
5489 1. Very minor rewording of debugging text in manualroute to say "list of
5490 hosts" instead of "hostlist".
5492 2. If verify=header_syntax was set, and a header line with an unqualified
5493 address (no domain) and a large number of spaces between the end of the
5494 name and the colon was received, the reception process suffered a buffer
5495 overflow, and (when I tested it) crashed. This was caused by some obsolete
5496 code that should have been removed. The fix is to remove it!
5498 3. When running in the test harness, delay a bit after writing a bounce
5499 message to get a bit more predictability in the log output.
5501 4. Added a call to search_tidyup() just before forking a reception process. In
5502 theory, someone could use a lookup in the expansion of smtp_accept_max_
5503 per_host which, without the tidyup, could leave open a database connection.
5505 5. Added the variables $recipient_data and $sender_data which get set from a
5506 lookup success in an ACL "recipients" or "senders" condition, or a router
5507 "senders" option, similar to $domain_data and $local_part_data.
5509 6. Moved the writing of debug_print from before to after the "senders" test
5512 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5513 problems for message scanning, either using a data ACL, or using
5514 local_scan() because the Received: header was not generated till after they
5515 were called (in order to set the time as the time of reception completion).
5516 I have revised the way this works. The header is now generated after the
5517 body is received, but before the ACL or local_scan() are called. After they
5518 are run, the timestamp in the header is updated.
5524 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5525 before starting a queue runner without re-exec. This happened only when
5526 deliver_drop_privilege was set or when the Exim user was set to root. The
5527 effect of the bug was that timeouts during subsequent deliveries caused
5528 crashes instead of being properly handled. The handler is now left at its
5529 default (and expected) setting.
5531 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5532 message, again when deliver_drop_privilege is set or Exim is run as root.
5533 The bug described in (1) was not present in this case, but the tidying up
5534 of the other signals was missing. I have made the two cases consistent.
5536 3. The ignore_target_hosts setting on a manualroute router was being ignored
5537 for hosts that were looked up using the /MX notation.
5539 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5542 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5543 operated on the sender address. After changing the $sender_address to <>
5544 for the sender address verify, Exim was re-instated it as the original
5545 (before rewriting) address, but remembering that it had rewritten it, so it
5546 wasn't rewriting it again. This bug also had the effect of breaking the
5547 sender address verification caching when the sender address was rewritten.
5549 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5550 This has been changed so that if the ip literal address matches
5551 ignore_target_hosts, the router declines.
5553 7. Added expansion conditions match_domain, match_address, and match_local_
5554 part (NOT match_host).
5556 8. The placeholder for the Received: header didn't have a length field set.
5558 9. Added code to Exim itself and to exim_lock to test for a specific race
5559 condition that could lead to file corruption when using MBX delivery. The
5560 issue is with the lockfile that is created in /tmp. If this file is removed
5561 after a process has opened it but before that process has acquired a lock,
5562 there is the potential for a second process to recreate the file and also
5563 acquire a lock. This could lead to two Exim processes writing to the file
5564 at the same time. The added code performs the same test as UW imapd; it
5565 checks after acquiring the lock that its file descriptor still refers to
5566 the same named file.
5568 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5569 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5572 11. Added the smtp_active_hostname option. If used, this will typically be made
5573 to depend on the incoming interface address. Because $interface_address is
5574 not set up until the daemon has forked a reception process, error responses
5575 that can happen earlier (such as "too many connections") no longer contain
5578 12. If an expansion in a condition on a "warn" statement fails because a lookup
5579 defers, the "warn" statement is abandoned, and the next ACL statement is
5580 processed. Previously this caused the whole ACL to be aborted.
5582 13. Added the iplsearch lookup type.
5584 14. Added ident_timeout as a log selector.
5586 15. Added tls_certificate_verified as a log selector.
5588 16. Added a global option tls_require_ciphers (compare the smtp transport
5589 option of the same name). This controls incoming TLS connections.
5591 17. I finally figured out how to make tls_require_ciphers do a similar thing
5592 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5593 before starting the TLS session.
5595 18. Tabs are now shown as \t in -bP output.
5597 19. If the log selector return_path_on_delivery was set, Exim crashed when
5598 bouncing a message because it had too many Received: header lines.
5600 20. If two routers both had headers_remove settings, and the first one included
5601 a superfluous trailing colon, the final name in the first list and the
5602 first name in the second list were incorrectly joined into one item (with a
5603 colon in the middle).
5609 1. Added -C and -D options to the exinext utility, mainly to make it easier
5610 to include in the automated testing, but these could be helpful when
5611 multiple configurations are in use.
5613 2. The exinext utility was not formatting the output nicely when there was
5614 an alternate port involved in the retry record key, nor when there was a
5615 message id as well (for retries that were specific to a specific message
5616 and a specific host). It was also confused by IPv6 addresses, because of
5617 the additional colons they contain. I have fixed the IPv4 problem, and
5618 patched it up to do a reasonable job for IPv6.
5620 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5621 delivery, the log line now contains "pipelined" if PIPELINING was used.
5623 4. An SMTP transport process used to panic and die if the bind() call to set
5624 an explicit outgoing interface failed. This has been changed; it is now
5625 treated in the same way as a connect() failure.
5627 5. A reference to $sender_host_name in the part of a conditional expansion
5628 that was being skipped was still causing a DNS lookup. This no longer
5631 6. The def: expansion condition was not recognizing references to header lines
5632 that used bh_ and bheader_.
5634 7. Added the _cache feature to named lists.
5636 8. The code for checking quota_filecount in the appendfile transport was
5637 allowing one more file than it should have been.
5639 9. For compatibility with Sendmail, the command line option
5647 and sets the incoming protocol and host name (for trusted callers). The
5648 host name and its colon can be omitted when only the protocol is to be set.
5649 Note the Exim already has two private options, -pd and -ps, that refer to
5650 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5651 "s", but I don't think that's a major issue.
5653 10. A number of refactoring changes to the code, none of which should affect
5656 (a) The number of logging options was getting close to filling up the
5657 32-bit word that was used as a bit map. I have split them into two classes:
5658 those that are passed in the argument to log_write(), and those that are
5659 only ever tested independently outside of that function. These are now in
5660 separate 32-bit words, so there is plenty of room for expansion again.
5661 There is no change in the user interface or the logging behaviour.
5663 (b) When building, for example, log lines, the code previously used a
5664 macro that called string_cat() twice, in order to add two strings. This is
5665 not really sufficiently general. Furthermore, there was one instance where
5666 it was actually wrong because one of the argument was used twice, and in
5667 one call a function was used. (As it happened, calling the function twice
5668 did not affect the overall behaviour.) The macro has been replaced by a
5669 function that can join an arbitrary number of extra strings onto a growing
5672 (c) The code for expansion conditions now uses a table and a binary chop
5673 instead of a serial search (which was left over from when there were very
5674 few conditions). Also, it now recognizes conditions like "pam" even when
5675 the relevant support is not compiled in: a suitably worded error message is
5676 given if an attempt is made to use such a condition.
5678 11. Added ${time_interval:xxxxx}.
5680 12. A bug was causing one of the ddress fields not to be passed back correctly
5681 from remote delivery subprocesses. The field in question was not being
5682 subsequently used, so this caused to problems in practice.
5684 13. Added new log selectors queue_time and deliver_time.
5686 14. Might have fixed a bug in maildirsizefile handling that threw up
5687 "unexpected character" debug warnings, and recalculated the data
5688 unnecessarily. In any case, I expanded the warning message to give more
5691 15. Added the message "Restricted characters in address" to the statements in
5692 the default ACL that block characters like @ and % in local parts.
5694 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5695 Three changes have been made:
5697 (a) There was a serious bug; a negative response to MAIL caused the whole
5698 recipient domain to be cached as invalid, thereby blocking all messages
5699 to all local parts at the same domain, from all senders. This bug has
5700 been fixed. The domain is no longer cached after a negative response to
5701 MAIL if the sender used is not empty.
5703 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5706 (c) A new callout option, "use_sender" has been added for people who want
5707 the modified behaviour.
5713 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5716 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5717 indeed breaks things for older releases.
5719 3. Added additional logging to the case where there is a problem reading data
5720 from a filter that is running in a subprocess using a pipe, in order to
5721 try to track down a specific problem.
5723 4. Testing facility fudge: when running in the test harness and attempting
5724 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5725 getting "No route to host". Convert this to a timeout.
5727 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5730 6. Some OS don't have socklen_t but use size_t instead. This affects the
5731 fifth argument of getsockopt() amongst other things. This is now
5732 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5733 can be set for individual OS. I have set it for SunOS5, OSF1, and
5734 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5735 some earlier ones do not.
5737 7. Change 4.30/15 was not doing the test caselessly.
5739 8. The standard form for an IPv6 address literal was being rejected by address
5740 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5741 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5742 this, as well as the form without the "IPv6" on the front (but only when
5743 address literals are enabled, of course).
5745 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5747 10. Exim crashed if a message with an empty sender address specified by -f
5748 encountered a router with an errors_to setting. This could be provoked only
5749 by a command such as
5753 where an empty string was supplied; "<>" did not hit this bug.
5755 11. Installed PCRE release 4.5.
5757 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5758 remained set. It is now erased.
5760 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5761 times from message ids (which are base 36 rather than the normal 62).
5763 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5764 were being counted as actual protocol errors, and logged if the log
5765 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5766 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5767 and DATA following a set of rejected RCPTs do not count as protocol errors.
5768 In other words, Exim assumes they were pipelined, though this may not
5769 actually be the case. Of course, in all cases the client gets an
5770 appropriate error code.
5772 15. If a lookup fails in an ACL condition, a message about the failure may
5773 be available; it is used if testing the ACL cannot continue, because most
5774 such messages specify what the cause of the deferral is. However, some
5775 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5776 that caused an old message to be retained and used if a later statement
5777 caused a defer, replacing the real cause of the deferral.
5779 16. If an IP address had so many PTR records that the DNS lookup buffer
5780 was not large enough to hold them, Exim could crash while trying to process
5781 the truncated data. It now detects and logs this case.
5783 17. Further to 4.21/58, another change has been made: if (and only if) the
5784 first line of a message (the first header line) ends with CRLF, a bare LF
5785 in a subsequent header line has a space inserted after it, so as not to
5786 terminate the header.
5788 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5789 unnecessarily, used atoi() instead of strtol(), and didn't check the
5790 termination when getting file sizes from file names by regex.
5792 19. Completely re-implemented the support for maildirsize files, in the light
5793 of a number of problems with the previous contributed implementation
5794 (4.30/29). In particular:
5796 . If the quota is zero, the maildirsize file is maintained, but no quota is
5799 . If the maildir directory does not exist, it is created before any attempt
5800 to write a maildirsize file.
5802 . The quota value in the file is just a cache; if the quota is changed in
5803 the transport, the new value overrides.
5805 . A regular expression is available for excluding directories from the
5808 20. The autoreply transport checks the characters in options that define the
5809 message's headers; it allows continued headers, but it was checking with
5810 isspace() after an embedded newline instead of explicitly looking for a
5813 21. If all the "regular" hosts to which an address was routed had passed their
5814 expiry times, and had not reached their retry times, the address was
5815 bounced, even if fallback hosts were defined. Now Exim should go on to try
5818 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5819 equivalent code in the SMTP transport. Some hosts send humungous responses
5820 to HELO/EHLO, more than 1024 it seems.
5822 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5823 gives compiler warnings in some environments. I've now done it "properly",
5826 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5827 (because of problems with the built-in one) was declared to return uschar *
5828 instead of char *, causing compiler failure.
5830 25. Fixed a file descriptor leak when processing alias/forward files.
5832 26. Fixed a minor format string issue in dbfn.c.
5834 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5836 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5837 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5840 29. When a sender address is verified, it is cached, to save repeating the test
5841 when there is more than one recipient in a message. However, when the
5842 verification involves a callout, it is possible for different callout
5843 options to be set for different recipients. It is too complicated to keep
5844 track of this in the cache, so now Exim always runs a verification when a
5845 callout is required, relying on the callout cache for the optimization.
5846 The overhead is duplication of the address routing, but this should not be
5849 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5850 to be verified with callout=postmaster, and the main callout worked but the
5851 postmaster check failed, the verification correctly failed. However, if a
5852 subsequent RCPT command asked for sender verification *without* the
5853 postmaster check, incorrect caching caused this verification also to fail,
5856 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5857 it was not caching the DNS options (qualify_single, search_parents) that
5858 were used when the lookup failed. A subsequent lookup with different
5859 options therefore always gave the same answer, though there were cases
5860 where it should not have. (Example: a "domains = !$mx_any" option on a
5861 dnslookup router: the "domains" option is always processed without any
5862 widening, but the router might have qualify_single set.) Now Exim uses the
5863 cached value only when the same options are set.
5865 32. Added John Jetmore's "exipick" utility to the distribution.
5867 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5868 than a timeout (e.g. a certificate is required, and is not provided), an
5869 Exim server now closes the connection immediately. Previously it waited for
5870 the client to close - but if the client is SSL, it seems that they each
5871 wait for each other, leading to a delay before one of them times out.
5873 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5874 maintained 0.8.x compatibility because I don't think many are using it, and
5875 it is clearly obsolete.
5877 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5880 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5881 client certificate was expired. A simple patch fixes this, though I don't
5882 understand the full logic of why the verify callback is called multiple
5885 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5886 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5887 which causes problems with some clients (such as the Certicom SSL Plus
5888 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5889 disables the coutermeasure allowing Eudora to connect."
5891 38. Exim was not checking that a write() to a log file succeeded. This could
5892 lead to Bad Things if a log got too big, in particular if it hit a file
5893 size limit. Exim now panics and dies if it cannot write to a log file, just
5894 as it does if it cannot open a log file.
5896 39. Modified OS/Makefile-Linux so that it now contains
5898 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5900 The two -D definitions ensure that Exim is compiled with large file
5901 support, which makes it possible to handle log files that are bigger than
5904 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5905 instance) a domain was checked against a named list that involved a lookup,
5906 causing $domain_data to be set, then another domain was checked against the
5907 same list, then the first domain was re-checked, the value of $domain_data
5908 after the final check could be wrong. In particular, if the second check
5909 failed, it could be set empty. This bug probably also applied to
5912 41. The strip_trailing_dot option was not being applied to the address given
5913 with the -f command-line option.
5915 42. The code for reading a message's header from the spool was incrementing
5916 $received_count, but never initializing it. This meant that the value was
5917 incorrect (doubled) while delivering a message in the same process in which
5918 it was received. In the most common configuration of Exim, this never
5919 happens - a fresh exec is done - but it can happen when
5920 deliver_drop_privilege is set.
5922 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5923 it now includes up to 150 characters of the unexpected data in the log
5926 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5927 and building data strings. The size of both of these buffers was 10 000
5928 bytes - far larger than anybody would *ever* want, thought I. Needless to
5929 say, somebody hit the limit. I have increased the maximum line length to
5930 20 000 and the maximum data length of concatenated lines to 100 000. I have
5931 also fixed two bugs, because there was no checking on these buffers. Tsk,
5932 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5933 buffer is too small.
5935 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5936 lsearch lookups. Now it does.
5938 46. When parsing a route_list item in a manualroute router, a fixed-length
5939 buffer was used for the list of hosts. I made this 1024 bytes long,
5940 thinking that nobody would ever have a list of hosts that long. Wrong.
5941 Somebody had a whole pile of complicated expansion conditions, and the
5942 string was silently truncated, leading to an expansion error. It turns out
5943 that it is easier to change to an unlimited length (owing to other changes
5944 that have happened since this code was originally written) than to build
5945 structure for giving a limitation error. The length of the item that
5946 expands into the list of hosts is now unlimited.
5948 47. The lsearch lookup could not handle data where the length of text line was
5949 more than 4095 characters. Such lines were truncated, leading to shortened
5950 data being returned. It should now handle lines of any length.
5952 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5953 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5956 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5957 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5958 no longer happens in comment lines. A list of replacements is now placed
5959 at the head of all of the source files, except those whose only change is
5960 to replace PERL_COMMAND in the very first #! line.
5962 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5963 messages on the queue, with a bottom-up merge sort, using code contributed
5964 by Michael Haardt. This should make operations like -bp somewhat faster on
5965 large queues. It won't affect queue runners, except when queue_run_in_order
5968 51. Installed eximstats 1.31 in the distribution.
5970 52. Added support for SRV lookups to the dnslookup router.
5972 53. If an ACL referred to $message_body or $message_body_end, the value was not
5973 reset for any messages that followed in the same SMTP session.
5975 54. The store-handling optimization for building very long strings was not
5976 differentiating between the different store pools. I don't think this
5977 actually made any difference in practice, but I've tidied it.
5979 55. While running the routers to verify a sender address, $sender_address
5980 was still set to the sender address. This is wrong, because when routing to
5981 send a bounce to the sender, it would be empty. Therefore, I have changed
5982 it so that, while verifying a sender address, $sender_address is set to <>.
5983 (There is no change to what happens when verifying a recipient address.)
5985 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5986 target A or AAAA records (if not already returned) without resetting the
5987 qualify_single or search_parents options of the DNS resolver. These are
5988 inappropriate in this case because the targets of MX and SRV records must
5989 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5990 target that, when qualified, matched something in the local domain. These
5991 two options are now turned off when doing these lookups.
5993 57. It seems that at least some releases of Reiserfs (which does not have the
5994 concept of a fixed number of inodes) returns zero and not -1 for the
5995 number of available inodes. This interacted badly with check_spool_inodes,
5996 which assumed that -1 was the "no such thing" setting. What I have done is
5997 to check that the total number of inodes is greater than zero before doing
5998 the test of how many are available.
6000 58. When a "warn" ACL statement has a log_message modifier, the message is
6001 remembered, and not repeated. This is to avoid a lot of repetition when a
6002 message has many recipients that cause the same warning to be written.
6003 However, Exim was preserving the list of already written lines for an
6004 entire SMTP session, which doesn't seem right. The memory is now reset if a
6005 new message is started.
6007 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6008 debugging output unless log_rewrite was also set.
6010 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6011 of (int)(handle) when we know that handle contains (void *)(-1).
6013 61. The Exim daemon panic-logs an error return when it closes the incoming
6014 connection. However "connection reset by peer" seems to be common, and
6015 isn't really an error worthy of noting specially, so that particular error
6018 62. When Exim is trying to find all the local interfaces, it used to panic and
6019 die if the ioctl to get the interface flags failed. However, it seems that
6020 on at least one OS (Solaris 9) it is possible to have an interface that is
6021 included in the list of interfaces, but for which you get a failure error
6022 for this call. This happens when the interface is not "plumbed" into a
6023 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6024 failure of the "get flags" call assumes that the interface is down.
6026 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6027 makes life easier for people who are doing arithmetic on fields extracted
6028 from dates, where you often get leading zeros that should not be
6029 interpreted as octal.
6031 64. Added qualify_domain to the redirect router, to override the global
6034 65. If a pathologically long header line contained very many addresses (the
6035 report of this problem mentioned 10 000) and each of them was rewritten,
6036 Exim could use up a very large amount of memory. (It kept on making new
6037 copies of the header line as it rewrote, and never released the old ones.)
6038 At the expense of a bit more processing, the header rewriting function has
6039 been changed so that it no longer eats memory in this way.
6041 66. The generation of the Received: header has been moved from the time that a
6042 message starts to be received, to the time that it finishes. The timestamp
6043 in the Received: header should now be very close to that of the <= log
6044 line. There are two side-effects of this change:
6046 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6047 logged header lines no longer include the local Received: line, because
6048 it has not yet been created. The same applies to a copy of the message
6049 that is returned to a non-SMTP sender when a message is rejected.
6051 (b) When a filter file is tested using -bf, no additional Received: header
6052 is added to the test message. After some thought, I decided that this
6055 This change does not affect the value of $received_for. It is still set
6056 after address rewriting, but before local_scan() is called.
6058 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6060 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6061 gave an unhelpful panic error message, and a defer error. I have managed to
6062 change this behaviour so that it now rejects any supplied certificate,
6063 which seems right, as the list of acceptable certificates is empty.
6065 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6066 gave an unhelpful defer error. I have not managed to make this reject any
6067 supplied certificates, but the error message it gives is "no certificate
6068 supplied", which is not helpful.
6070 70. exigrep's output now also includes lines that are not associated with any
6071 message, but which match the given pattern. Implemented by a patch from
6072 Martin Sluka, which also tidied up the Perl a bit.
6074 71. Recipient callout verification, like sender verification, was using <> in
6075 the MAIL FROM command. This isn't really the right thing, since the actual
6076 sender may affect whether the remote host accepts the recipient or not. I
6077 have changed it to use the actual sender in the callout; this means that
6078 the cache record is now keyed on a recipient/sender pair, not just the
6079 recipient address. There doesn't seem to be a real danger of callout loops,
6080 since a callout by the remote host to check the sender would use <>.
6081 [SEE ABOVE: changed after hitting problems.]
6083 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6084 temporary errors. However, in the case of such a code being given after
6085 the end of a data transmission (i.e. after ".") Exim was failing to write
6086 a retry record for the message. (Yes, there was some broken host that was
6087 actually sending 8xx at this point.)
6089 73. An unknown lookup type in a host list could cause Exim to panic-die when
6090 the list was checked. (An example that provoked this was putting <; in the
6091 middle of a list instead of at the start.) If this happened during a DATA
6092 ACL check, a -D file could be left lying around. This kind of configuration
6093 error no longer causes Exim to die; instead it causes a defer error. The
6094 incident is still logged to the main and panic logs.
6096 74. Buglet left over from Exim 3 conversion. The message "too many messages
6097 in one connection" was written to the rejectlog but not the mainlog, except
6098 when address rewriting (yes!) was being logged.
6100 75. Added write_rejectlog option.
6102 76. When a system filter was run not as root (that is, when system_filter_user
6103 was set), the values of the $n variables were not being returned to the
6104 main process; thus, they were not subsequently available in the $sn
6107 77. Added +return_path_on_delivery log selector.
6109 78. A connection timeout was being treated differently from recipients deferred
6110 when testing hosts_max_try with a message that was older than the host's
6111 retry timeout. (The host should not be counted, thus allowing all hosts to
6112 be tried at least once before bouncing.) This may have been the cause of an
6113 occasionally reported bug whereby a message would remain on the queue
6114 longer than the retry timeout, but would be bounced if a delivery was
6115 forced. I say "may" because I never totally pinned down the problem;
6116 setting up timeout/retry tests is difficult. See also the next item.
6118 79. The ultimate address timeout was not being applied to errors that involved
6119 a combination of host plus message (for example, a timeout on a MAIL
6120 command). When an address resolved to a number of possible hosts, and they
6121 were not all tried for each delivery (e.g. because of hosts_max_try), a
6122 message could remain on the queue longer than the retry timeout.
6124 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6127 81. Fixed an obscure SMTP outgoing bug which required at least the following
6128 conditions: (a) there was another message waiting for the same server;
6129 (b) the server returned 5xx to all RCPT commands in the first message so
6130 that the message was not completed; (c) the server dropped the connection
6131 or gave a negative response to the RSET that Exim sends to abort the
6132 transaction. The observed case was a dropped connection after DATA that had
6133 been sent in pipelining mode. That is, the server had advertised PIPELINING
6134 but was not implementing it correctly. The effect of the bug was incorrect
6135 behaviour, such as trying another host, and this could lead to a crash.
6141 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6142 and daemon.c were passed as pointers to ints; they should have been
6143 pointers to socklen_t variables (which are typically unsigned ints).
6145 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6148 3. Fixed a really odd bug that affected only the testing scheme; patching a
6149 certain fixed string in the binary changed the value of another string that
6150 happened to be identical to the end of the original first string.
6152 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6153 name", it returns that address as the IP address. On some operating
6154 systems (e.g. Solaris), it also passes back the IP address string as the
6155 "host name". However, on others (e.g. Linux), it passes back an empty
6156 string. Exim wasn't checking for this, and was changing the host name to an
6157 empty string, assuming it had been canonicalized.
6159 5. Although rare, it is permitted to have more than one PTR record for a given
6160 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6161 all the names associated with an address, because they do in Solaris.
6162 However, it seems that they do not in Linux for data that comes from the
6163 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6164 I found this out when I moved to a new Linux workstation and tried to run
6165 the Exim test suite.
6167 To get round this problem I have changed the code so that it now does its
6168 own call to the DNS to look up PTR records when searching for a host name.
6169 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6170 addresses that are only in /etc/hosts are still found.
6172 This behaviour is, however, controlled by an option called host_lookup_
6173 order, which defaults to "bydns:byaddr". If people want to use the other
6174 order, or indeed, just use one or the other means of lookup, they can
6175 specify it in this variable.
6177 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6178 some operating systems, this comes back from gethostbyaddr() as an empty
6179 string, and this is what Exim used to test for. However, it seems that in
6180 other systems, "." is yielded. Exim now tests for this case too.
6182 7. The values of check_spool_space and check_log_space are now held internally
6183 as a number of kilobytes instead of an absolute number of bytes. If a
6184 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6185 kilobyte. This means that much larger values can be stored.
6187 8. Exim monitor: an attempt to get the action menu when not actually pointing
6188 at a message produces an empty menu entitled "No message selected". This
6189 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6190 no entries in it ("Shell widget menu has zero width and/or height"). So I
6191 have added a single, blank menu entry in this case.
6193 9. Added ${quote_local_part.
6195 10. MIME decoding is now applied to the contents of Subject: header lines when
6198 11. Now that a reference to $sender_host_address automatically causes a reverse
6199 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6200 host lookup before query-style lookups in lists that might use this
6201 variable. This has therefore been abolished, and the "net-" prefix is no
6202 longer necessary for query-style lookups.
6204 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6205 have been a typo for LFLAGS, so it has been changed.
6207 13. The install script calls Exim with "-C /dev/null" in order to find the
6208 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6209 to be output. However, since Exim outputs its version number before the
6210 error, it didn't break the script. It just looked ugly. I fixed this by
6211 always allowing "-C /dev/null" if the caller is root.
6213 14. Ignore overlarge ACL variable number when reading spool file - insurance
6214 against a later release with more variables having written the file.
6216 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6217 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6218 this, as well as the form without the "IPv6" on the front.
6220 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6221 OS/Makefile-Darwin file.
6223 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6224 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6226 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6227 variable. However, some OS use ints for inodes. I've added cast to long int
6228 to get rid of the compiler warning.
6230 19. I had forgotten to lock out "/../" in configuration file names when
6231 ALT_CONFIG_PREFIX was set.
6233 20. Routers used for verification do not need to specify transports. However,
6234 if such a router generated a host list, and callout was configured, Exim
6235 crashed, because it could not find a port number from the (non-existent)
6236 transport. It now assumes port 25 in this circumstance.
6238 21. Added the -t option to exigrep.
6240 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6241 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6242 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6243 them is set, LOOKUP_LSEARCH is forced.
6245 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6247 24. Added sender and host information to the "rejected by local_scan()" log
6248 line; previously there was no indication of these.
6250 25. Added .include_if_exists.
6252 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6253 Linux. It turns out that not all file systems support this. Apparently some
6254 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6255 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6256 not supported on the file descriptor, is now ignored when Exim is trying to
6257 sync a directory. This applies only to Linux.
6259 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6261 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6262 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6265 29. Added support for maildirsize files from supplied patch (modified a bit).
6267 30. The use of :fail: followed by an empty string could lead Exim to respond to
6268 sender verification failures with (e.g.):
6270 550 Verification failed for <xxx>
6271 550 Sender verify failed
6273 where the first response line was missing the '-' that indicates it is not
6274 the final line of the response.
6276 31. The loop for finding the name of the user that called Exim had a hardwired
6277 limit of 10; it now uses the value of finduser_retries, which is used for
6278 all other user lookups.
6280 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6283 33. Exim was neglecting to zero errno before one call of strtol() when
6284 expanding a string and expecting an integer value. On some systems this
6285 resulted in spurious "integer overflow" errors. Also, it was casting the
6286 result into an int without checking.
6288 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6289 did not work. The code looks as if it has *never* worked, though it appears
6290 to have been documented since at least release 1.62. I have made it work.
6292 35. The "timeout_DNS" error in retry rules, also documented since at least
6293 1.62, also never worked. As it isn't clear exactly what this means, and
6294 clearly it isn't a major issue, I have abolished the feature by treating it
6295 as "timeout", and writing a warning to the main and panic logs.
6297 36. The display of retry rules for -brt wasn't always showing the error code
6300 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6301 timeout_connect_A, timeout_connect_MX.
6303 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6304 to the empty sender.
6306 39. The daemon was not analysing the content of -oX till after it had closed
6307 stderr and disconnected from the controlling terminal. This meant that any
6308 syntax errors were only noted on the panic log, and the return code from
6309 the command was 0. By re-arranging the code a little, I've made the
6310 decoding happen first, so such errors now appear on stderr, and the return
6311 code is 1. However, the actual setting up of the sockets still happens in
6312 the disconnected process, so errors there are still only recorded on the
6315 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6316 connections (as happens on some IP stacks) was logged at start up time as
6317 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6318 it from "IPv6 and IPv4", which means that two separate sockets are being
6321 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6322 says whether AF_INET or AF_INET6 was passed as an argument.
6324 42. Exiwhat output was messed up when time zones were included in log
6327 43. Exiwhat now gives more information about the daemon's listening ports,
6328 and whether -tls-on-connect was used.
6330 44. The "port" option of the smtp transport is now expanded.
6332 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6333 silently ignored. Now an error message is written to the main and panic
6336 46. There's a new ACL modifier called "logwrite" which writes to a log file
6337 as soon as it is encountered.
6339 47. Added $local_user_uid and $local_user_gid at routing time.
6341 48. Exim crashed when trying to verify a sender address that was being
6344 49. Exim was recognizing only a space character after ".include". It now also
6345 recognizes a tab character.
6347 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6348 extracting the relevant information from the specification. The man page no
6349 longer contains scrambled data for the -d option, and I've added a section
6350 at the front about calling Exim under different names.
6352 51. Added "extra_headers" argument to the "mail" command in filter files.
6354 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6357 53. Installed eximstats 1.29.
6359 54. Added transport_filter_timeout as a generic transport option.
6361 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6362 Cc: header lines. This was required by RFC 822, but it not required by RFC
6365 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6366 incoming messages that did not have them. Now it does so only if the
6367 message originates locally, that is, if there is no associated remote host
6368 address. When Resent- header lines are present, this applies to the Resent-
6369 lines rather than the non-Resent- lines.
6371 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6372 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6374 58. Messages for configuration errors now include the name of the main
6375 configuration file - useful now that there may be more than one file in a
6376 list (.included file names were always shown).
6378 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6379 for those rare installations that do not start the daemon as root or run it
6380 setuid root. I've cut out the call to initgroups() if the daemon is not
6383 60. The Exim user and group can now be bound into the binary as text strings
6384 that are looked up at the start of Exim's processing.
6386 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6388 62. Added $mailstore_basename variable.
6390 63. Installed patch to sieve.c from Michael Haardt.
6392 64. When Exim failed to open the panic log after failing to open the main log,
6393 the original message it was trying to log was written to stderr and debug
6394 output, but if they were not available (the usual case in production), it
6395 was lost. Now it is written to syslog before the two lines that record the
6396 failures to open the logs.
6398 65. Users' Exim filters run in subprocesses under the user's uid. It is
6399 possible for a "deliver" command or an alias in a "personal" command to
6400 provoke an address rewrite. If logging of address rewriting is configured,
6401 this fails because the process is not running as root or exim. There may be
6402 a better way of dealing with this, but for the moment (because 4.30 needs
6403 to be released), I have disabled address rewrite logging when running a
6404 filter in a non-root, non-exim process.
6410 1. The buildconfig auxiliary program wasn't quoting the value set for
6411 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6412 not defined. This bug was masked in 4.22 by the effect that was fixed in
6415 2. Some messages that were rejected after a message id was allocated were
6416 shown as "incomplete" by exigrep. It no longer does this for messages that
6417 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6419 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6420 have allow_domain_literals set, the ID did not get logged in the <= line.
6421 Domain literals are now always recognized in Message-ID: header lines.
6423 4. The first argument for a ${extract expansion item is the key name or field
6424 number. Leading and trailing spaces in this item were not being ignored,
6425 causing some misleading effects.
6427 5. When deliver_drop_privilege was set, single queue runner processes started
6428 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6429 same command in the process it spins off) were not dropping privilege.
6431 6. When the daemon running as "exim" started a queue runner, it always
6432 re-executed Exim in the spun-off process. This is a waste of effort when
6433 deliver_drop_privilege is set. The new process now just calls the
6434 queue-runner function directly.
6440 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6443 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6444 ignored. Though the use of -oP was forcing the writing of a pid file, it
6445 was always written to the default place.
6447 3. If the message "no IP address found for host xxxx" is generated during
6448 incoming verification, it is now followed by identification of the incoming
6449 connection (so you can more easily find what provoked it).
6451 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6453 5. Added some features to "harden" Exim a bit more against certain attacks:
6455 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6456 be put in Local/Makefile. This is like the never_users runtime option,
6457 but it cannot be overridden. The default setting is "root".
6459 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6460 prefix string with which any file named in a -C command line option
6463 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6464 is retained for -C and -D only if the caller of Exim is root. Without
6465 it, the exim user may also use -C and -D and retain privilege.
6467 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6468 command line option is disabled.
6470 6. Macro names set by the -D option must start with an upper case letter, just
6471 like macro names defined in the configuration file.
6473 7. Added "dereference=" facility to LDAP.
6475 8. Two instances of the typo "uknown" in the source files are fixed.
6477 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6478 the Configure-Makefile script screwed up while processing it.
6480 10. Incorporated PCRE 4.4.
6482 11. The SMTP synchronization check was not operating right at the start of an
6483 SMTP session. For example, it could not catch a HELO sent before the client
6484 waited for the greeting. There is now a check for outstanding input at the
6485 point when the greeting is written. Because of the duplex, asynchronous
6486 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6487 way, but not yet received, when the check is performed.
6489 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6490 on TCP/IP sockets, because this apparently causes some broken clients to
6493 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6494 unchanged) from the Cygwin maintainer.
6496 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6497 was supported instead of testing for mbx. Effectively a typo.
6499 15. The spa authenticator server code was not checking that the input it
6500 received was valid base64.
6502 16. The debug output line for the "set" modifier in ACLs was not showing the
6503 name of the variable that was being set.
6505 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6507 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6508 lookup on its own. Something else was needed to trigger the lookup. For
6509 example, a match in host_lookup or the need for a host name in a host list.
6510 Now, if $sender_host_name is referenced and the host name has not yet been
6511 looked up, a lookup is performed. If the lookup fails, the variable remains
6512 empty, and $host_lookup_failed is set to "1".
6514 19. Added "eqi" as a case-independent comparison operator.
6516 20. The saslauthd authentication condition could segfault if neither service
6517 nor realm was specified.
6519 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6520 error message that was logged was misleading, and incoming SMTP
6521 connections were dropped. The message is now more accurate, and temporary
6522 errors are given to SMTP connections.
6524 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6525 (except RSET) with a 421 error, until QUIT is received. However, it was
6526 failing to send a response to QUIT.
6528 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6529 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6530 ACL is now run after the helo_try_verify_hosts code.
6532 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6533 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6534 case-independent, so other case variants are also recognized). Apparently
6535 some systems use these upper case variants.
6537 25. If more than two messages were waiting for the same host, and a transport
6538 filter was specified for the transport, Exim sent two messages over the
6539 same TCP/IP connection, and then failed with "socket operation on non-
6540 socket" when it tried to send the third.
6542 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6544 27. The extern definition of crypt16() in expand.c was not being excluded when
6545 the OS had its own crypt16() function.
6547 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6548 as a preferred synonym for return_size_limit, both as an option and as an
6551 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6553 30. Changed the default configuration ACL to relax the local part checking rule
6554 for addresses that are not in any local domains. For these addresses,
6555 slashes and pipe symbols are allowed within local parts, but the sequence
6556 /../ is explicitly forbidden.
6558 31. SPA server authentication was not clearing the challenge buffer before
6561 32. log_message in a "warn" ACL statement was writing to the reject log as
6562 well as to the main log, which contradicts the documentation and doesn't
6563 seem right (because no rejection is happening). So I have stopped it.
6565 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6566 However, I am unable to do any testing of this.
6568 34. Fixed an infelicity in the appendfile transport. When checking directories
6569 for a mailbox, to see if any needed to be created, it was accidentally
6570 using path names with one or more superfluous leading slashes; tracing
6571 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6573 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6574 contents are added to the log line that is written for every discarded
6575 recipient. (Previously a log_message setting was ignored.)
6577 36. The ${quote: operator now quotes the string if it is empty.
6579 37. The install script runs exim in order to find its version number. If for
6580 some reason other than non-existence or emptiness, which it checks, it
6581 could not run './exim', it was installing it with an empty version number,
6582 i.e. as "exim-". This error state is now caught, and the installation is
6585 38. An argument was missing from the function that creates an error message
6586 when Exim fails to connect to the socket for saslauthd authentication.
6587 This could cause Exim to crash, or give a corrupted message.
6589 39. Added isip, isip4, and isip6 to ${if conditions.
6591 40. The ACL variables $acl_xx are now saved with the message, and can be
6592 accessed later in routers, transports, and filters.
6594 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6595 strings in the file are not string-expanded.
6597 42. If a MAIL command specified a SIZE value that was too large to fit into an
6598 int variable, the check against message_size_limit failed. Such values are
6599 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6600 day this will have to be increased, but I don't think I want to be around
6601 when emails are that large.
6608 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6609 iconv() is not standard in FreeBSD.
6611 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6612 IPv6 enabled. The observed symptom was a segmentation fault on return from
6613 the function os_common_find_running_interfaces() in src/os.c.
6615 3. In the check_special_case() function in daemon.c I had used "errno" as an
6616 argument name, which causes warnings on some systems. This was basically a
6617 typo, since it was named "eno" in the comments!
6619 4. The code that waits for the clock to tick (at a resolution of some fraction
6620 of a second) so as to ensure message-id uniqueness was always waiting for
6621 at least one whole tick, when it could have waited for less. [This is
6622 almost certainly not relevant at current processor speeds, where it is
6623 unlikely to ever wait at all. But we try to future-proof.]
6625 5. The function that sleeps for a time interval that includes fractions of a
6626 second contained a race. It did not block SIGALRM between setting the
6627 timer, and suspending (a couple of lines later). If the interval was short
6628 and the sigsuspend() was delayed until after it had expired, the suspension
6629 never ended. On busy systems this could lead to processes getting stuck for
6632 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6633 process, before it forks any delivery processes. The open lookup caching
6634 mechanism meant that the open file or database connection was passed into
6635 the delivery process. The problem was that delivery processes always tidy
6636 up cached lookup data. This could cause a problem for the next delivery
6637 process started by the queue runner, because the external queue runner
6638 process does not know about the closure. So the next delivery process
6639 still has data in the lookup cache. In the case of a file lookup, there was
6640 no problem because closing a file descriptor in a subprocess doesn't affect
6641 the parent. However, if the lookup was caching a connection to a database,
6642 the connection was closed, and the second delivery process was likely to
6643 see errors such as "PGSQL: query failed: server closed the connection
6644 unexpectedly". The problem has been fixed by closing all cached lookups
6645 in a queue runner before running a delivery process.
6647 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6648 seem to have the "const" qualifier which it has on other OS. I've
6651 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6652 *of the same type* (client or server) with the same public name that an
6653 error should be diagnosed.
6655 9. When Exim looked up a host name for an IP address, but failed to find the
6656 original IP address when looking up the host name (a safety check), it
6657 output the message "<ip address> does not match any IP for NULL", which was
6658 confusing, to say the least. The bug was that the host name should have
6659 appeared instead of "NULL".
6661 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6662 user that is built into the binary, and the -C or -D options is used, root
6663 privilege is dropped before the configuration file is read. In addition,
6664 logging is switched to stderr instead of the normal log files. If the
6665 configuration then re-defines the Exim user, the unprivileged environment
6666 is probably not what is expected, so Exim logs a panic warning message (but
6669 However, if deliver_drop_privilege is set, the unprivileged state may well
6670 be exactly what is intended, so the warning has been cut out in that case,
6671 and Exim is allowed to try to write to its normal log files.
6677 1. smtp_return_error_details was not giving details for temporary sender
6678 or receiver verification errors.
6680 2. Diagnose a configuration error if two authenticators have the same public
6683 3. Exim used not to create the message log file for a message until the first
6684 delivery attempt. This could be confusing when incoming messages were held
6685 for policy or load reasons. The message log file is now created at the time
6686 the message is received, and an initial "Received" line is written to it.
6688 4. The automatically generated man page for command line options had a minor
6689 bug that caused no ill effects; however, a more serious problem was that
6690 the procedure for building the man page automatically didn't always
6691 operate. Consequently, release 4.20 contains an out-of-date version. This
6692 shouldn't happen again.
6694 5. When building Exim with embedded Perl support, the script that builds the
6695 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6696 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6698 6. The freeze_tell option was not being used for messages that were frozen on
6699 arrival, either by an ACL or by local_scan().
6701 7. Added the smtp_incomplete_transaction log selector.
6703 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6704 was accepting AUTH without a new EHLO.
6706 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6707 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6710 10. Logging of TCP/IP connections (when configured) now happens in the main
6711 daemon process instead of the child process, so that the TCP/IP connection
6712 count is more accurate (but it can never be perfect).
6714 11. The use of "drop" in a nested ACL was not being handled correctly in the
6715 outer ACL. Now, if condition failure induced by the nested "drop" causes
6716 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6717 or "require"), the connection is dropped.
6719 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6720 that yield "discard" can now be used with an "accept" or a "discard" verb,
6721 but an error is generated for any others (because I can't see a useful way
6722 to define what should happen).
6724 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6725 are now processed in the same way as lines in the Exim configuration file.
6726 In particular, continuation lines are supported.
6728 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6730 15. Added -ti meaning -t -i.
6732 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6733 domains, and warn by logging if others are found.
6735 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6736 returned from GIFCONF when Exim is trying to find the list of interfaces on
6737 a host. The code in os.c has been modified to copy each ifreq to an aligned
6738 structure in all cases.
6740 Also, in some cases, the returned ifreq's were being copied to a 'struct
6741 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6742 means the last couple of bytes of an IPv6 address could be chopped if the
6743 ifreq contained only a normal sockaddr (14 bytes storage).
6745 18. Named domain lists were not supported in the hosts_treat_as_local option.
6746 An entry such as +xxxx was not recognized, and was treated as a literal
6749 19. Ensure that header lines added by a DATA ACL are included in the reject log
6750 if the ACL subsequently rejects the message.
6752 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6753 MD5 (which is deprecated).
6755 21. When testing a filter file using -bf, Exim was writing a message when it
6756 took the sender from a "From " line in the message, but it was not doing so
6757 when it took $return_path from a Return-Path: header line. It now does.
6759 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6760 with a valid header line field name (a series of printing characters
6761 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6763 23. Changed "disc" in the source to "disk" to conform to the documentation and
6764 the book and for uniformity.
6766 24. Ignore Sendmail's -Ooption=value command line item.
6768 25. When execve() failed while trying to run a command in a pipe transport,
6769 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6770 could be confused with a return value of 69 from the command itself. This
6771 has been changed to 127, the value the shell returns if it is asked to run
6772 a non-existent command. The wording for the related log line suggests a
6773 non-existent command as the problem.
6775 26. If received_header_text expands to an empty string, do not add a Received:
6776 header line to the message. (Well, it adds a token one on the spool, but
6777 marks it "old" so that it doesn't get used or transmitted.)
6779 27. Installed eximstats 1.28 (addition of -nt option).
6781 28. There was no check for failure on the call to getsockname() in the daemon
6782 code. This can fail if there is a shortage of resources on the system, with
6783 ENOMEM, for example. A temporary error is now given on failure.
6785 29. Contrary to the C standard, it seems that in some environments, the
6786 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6787 program. Exim now does this explicitly; it affects the formatting of
6788 timestamps using strftime().
6790 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6791 complaints. I've now initialized all the variables, to avoid this.
6793 32. Header lines added by a system filter were not being "seen" during
6794 transport-time rewrites.
6796 33. The info_callback() function passed to OpenSSL is set up with type void
6797 (*)(SSL *, int, int), as described somewhere. However, when calling the
6798 function (actually a macro) that sets it up, the type void(*)() is
6799 expected. I've put in a cast to prevent warnings from picky compilers.
6801 34. If a DNS black list lookup found a CNAME record, but there were no A
6802 records associated with the domain it pointed at, Exim crashed.
6804 35. If a DNS black list lookup returned more than one A record, Exim ignored
6805 all but the first. It now scans all returned addresses if a particular IP
6806 value is being sought. In this situation, the contents of the
6807 $dnslist_value variable are a list of all the addresses, separated by a
6810 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6811 to accept a host name and all its aliases if the forward lookup for any of
6812 them yielded the IP address of the incoming connection. Now it accepts only
6813 those names whose forward lookup yields the correct IP address. Any other
6814 names are discarded. This closes a loophole whereby a rogue DNS
6815 administrator could create reverse DNS records to break through a
6816 wildcarded host restriction in an ACL.
6818 37. If a user filter or a system filter that ran in a subprocess used any of
6819 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6820 the wrong values were passed to the pipe command ($thisaddress had the
6821 value of $0, $0 had the value of $1, etc). This bug was introduced by
6822 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6824 38. Improved the line breaking for long SMTP error messages from ACLs.
6825 Previously, if there was no break point between 40 and 75 characters, Exim
6826 left the rest of the message alone. Two changes have been made: (a) I've
6827 reduced the minimum length to 35 characters; (b) if it can't find a break
6828 point between 35 and 75 characters, it looks ahead and uses the first one
6829 that it finds. This may give the occasional overlong line, but at least the
6830 remaining text gets split now.
6832 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6833 file descriptors might be low, and that setting 1000 would always raise it.
6834 It turns out that in some environments, the limit is already over 1000 and
6835 that lowering it causes trouble. So now Exim takes care not to decrease it.
6837 40. When delivering a message, the value of $return_path is set to $sender_
6838 address at the start of routing (routers may change the value). By an
6839 oversight, this default was not being set up when an address was tested by
6840 -bt or -bv, which affected the outcome if any router or filter referred to
6843 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6844 log only when "message" or "log_message" are set. However, if one of the
6845 conditions was an address verification, or a call to a nested ACL, the
6846 messages generated by the underlying test were being passed through. This
6847 no longer happens. The underlying message is available in $acl_verify_
6848 message for both "message" and "log_message" expansions, so it can be
6849 passed through if needed.
6851 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6852 new expansion $bh_ to give the encoded byte string without charset
6853 translation. Translation happens only if iconv() is available; HAVE_ICONV
6854 indicates this at build time. HEADERS_CHARSET gives the charset to
6855 translate to; headers_charset can change it in the configuration, and
6856 "headers charset" can change it in an individual filter file.
6858 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6859 that creates RFC 2047 encoded "words" labels them as that charset instead
6860 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6861 expansion operator; (ii) when Exim creates a From: line for a local
6862 message; (iii) when a header line is rewritten to include a "phrase" part.
6864 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6865 buggy, causing it to skip the first lines of messages whose message ID
6866 ended in 'D'. This would not have bitten before Exim release 4.14, because
6867 message IDs were unlikely to end in 'D' before then. The effect was to have
6868 incorrect size information for certain domains.
6870 45. #include "config.h" was missing at the start of the crypt16.c module. This
6871 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6874 46. If there was a timeout during a "random" callout check, Exim treated it as
6875 a failure of the random address, and carried on sending RSET and the real
6876 address. If the delay was just some slowness somewhere, the response to the
6877 original RCPT would be taken as a response to RSET and so on, causing
6878 mayhem of various kinds.
6880 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6881 when I implemented it. It didn't allow for the fact that some option values
6882 may legitimately be negative (e.g. size_addition), and it didn't even do
6883 the right test for positive values.
6885 48. Domain names in DNS records are case-independent. Exim always looks them up
6886 in lower case. Some resolvers return domain names in exactly the case they
6887 appear in the zone file, that is, they may contain uppercase letters. Not
6888 all resolvers do this - some return always lower case. Exim was treating a
6889 change of case by a resolver as a change of domain, similar to a widening
6890 of a domain abbreviation. This triggered its re-routing code and so it was
6891 trying to route what was effectively the same domain again. This normally
6892 caused routing to fail (because the router wouldn't handle the domain
6893 twice). Now Exim checks for this case specially, and just changes the
6894 casing of the domain that it ultimately uses when it transmits the message
6897 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6900 50. If a filter generated a file delivery with a non-absolute name (possible if
6901 no home directory exists for the router), the forbid_file option was not
6904 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6905 the existing equality matching.
6907 52. Exim was using ints instead of ino_t variables in some places where it was
6908 dealing with inode numbers.
6910 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6911 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6912 called TMPDIR, and if it finds it is different, it changes its value.
6914 54. The smtp_printf() function is now made available to local_scan() so
6915 additional output lines can be written before returning. There is also an
6916 smtp_fflush() function to enable the detection of a dropped connection.
6917 The variables smtp_input and smtp_batched_input are exported to
6920 55. Changed the default runtime configuration: the message "Unknown user"
6921 has been removed from the ACL, and instead placed on the localuser router,
6922 using the cannot_route_message feature. This means that any verification
6923 failures that generate their own messages won't get overridden. Similarly,
6924 the "Unrouteable address" message that was in the ACL for unverifiable
6925 relay addresses has also been removed.
6927 56. Added hosts_avoid_esmtp to the smtp transport.
6929 57. The exicyclog script was not checking for the esoteric option
6930 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6931 will work only if exicyclog is run under the appropriate euid.
6933 58. Following a discussion on the list, the rules by which Exim recognises line
6934 endings on incoming messages have been changed. The -dropcr and drop_cr
6935 options are now no-ops, retained only for backwards compatibility. The
6936 following line terminators are recognized: LF CRLF CR. However, special
6937 processing applies to CR:
6939 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6940 nor a local message in the state where . is a terminator.
6942 (ii) If a bare CR is encountered in a header line, an extra space is added
6943 after the line terminator so as not to end the header. The reasoning
6944 behind this is that bare CRs in header lines are most likely either
6945 to be mistakes, or people trying to play silly games.
6947 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6948 was being incorrectly given as 18 bytes larger than it should have been.
6949 This is a VOB (very old bug).
6951 60. This may never have affected anything current, but just in case it has:
6952 When the local host is found other than at the start of a list of hosts,
6953 the local host, those with the same MX, and any that follow, are discarded.
6954 When the list in question was part of a longer list of hosts, the following
6955 hosts (not currently being processed) were also being discarded. This no
6956 longer happens. I'm not sure if this situation could ever has previously
6959 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6962 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6963 header. This is something that is recommended in a new Internet Draft, and
6964 is something that is documented as being done by Sendmail. There are two
6965 possible values. For messages generated by the autoreply transport, Exim
6968 Auto-Submitted: auto-replied
6970 whereas for all other generated messages (e.g. bounces) it adds
6972 Auto-Submitted: auto-generated
6974 63. The "personal" condition in filters now includes a test for the
6975 Auto-Submitted: header. If it contains the string "auto-" the message it
6976 not considered personal.
6978 64. Added rcpt_include_affixes as a generic transport option.
6980 65. Added queue_only_override (default true).
6982 66. Added the syslog_duplication option.
6984 67. If what should have been the first header line of a message consisted of
6985 a space followed by a colon, Exim was mis-interpreting it as a header line.
6986 It isn't of course - it is syntactically invalid and should therefore be
6987 treated as the start of the message body. The misbehaviour could have
6988 caused a number of strange effects, including loss of data in subsequent
6989 header lines, and spool format errors.
6991 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6992 client host had authenticated. This control can now be exercised by an ACL
6993 for more flexibility.
6995 69. By default, callouts do not happen when testing with -bh. There is now a
6996 variant, -bhc, which does actually run the callout code, including
6997 consulting and updating the callout cache.
6999 70. Added support for saslauthd authentication, courtesy of Alexander
7002 71. If statvfs() failed on the spool or log directories while checking their
7003 size for availability, Exim confusingly gave the error "space shortage".
7004 Furthermore, in debugging mode it crashed with a floating point exception.
7005 These checks are done if check_{spool,log}_{space,inodes} are set, and when
7006 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7007 serious problem, Exim now writes to the main and panic logs when this
7008 happens, with details of the failure. It then refuses to accept the
7009 incoming message, giving the message "spool directory problem" or "log
7010 directory problem" with a 421 code for SMTP messages.
7012 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7013 0, 1, and 2 exist, because some OS complain for execs without them (see
7014 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7015 descriptors. However, the code omitted to check that the open succeeded,
7016 causing mysterious errors if for some reason the permissions on /dev/null
7017 got screwed. Now Exim writes a message to the main and panic logs, and
7018 bombs out if it can't open /dev/null.
7020 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7021 interact so that it is all more flexible. It is supposed to remain
7022 backwards compatible. Also added extra_local_interfaces.
7024 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7025 to bomb out with an assertion failure - to the client this appears as a
7026 connection drop. This problem occurs in the part of the code that was taken
7027 from the Samba project. Fortunately, the assertion is in a very simple
7028 function, so I have fixed this by reproducing the function inline in the
7029 one place where it is called, and arranging for authentication to fail
7030 instead of killing the process with assert().
7032 75. The SPA client code was not working when the server requested OEM rather
7033 than Unicode encoding.
7035 76. Added code to make require_files with a specific uid setting more usable in
7036 the case where statting the file as root fails - usually a non-root-mounted
7037 NFS file system. When this happens and the failure is EACCES, Exim now
7038 forks a subprocess and does the per-uid checking as the relevant uid.
7040 77. Added process_log_path.
7042 78. If log_file_path was not explicitly set, a setting of check_log_space or
7043 check_log_inodes was ignored.
7045 79. If a space check for the spool or log partitions fails, the incident is now
7046 logged. Of course, in the latter case the data may get lost...
7048 80. Added the %p formatting code to string_format() so that it can be used to
7049 print addresses in debug_print(). Adjusted all the address printing in the
7050 debugging in store.c to use %p rather than %d.
7052 81. There was a concern that a line of code in smtp_in.c could overflow a
7053 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7054 initially expressed, the concern was not well-founded, because trailing
7055 spaces are removed early. However, if the trailing spaces were followed by
7056 a NULL, they did not get removed, so the overflow was possible. Two fixes
7059 (a) I re-wrote the offending code in a cleaner fashion.
7060 (b) If an incoming SMTP command contains a NULL character, it is rejected
7063 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7064 runs initgroups(), so that if the Exim user is in any additional groups,
7065 they will be used during message reception.
7071 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob