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
11 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
12 Previously only the last row was returned.
14 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
15 we assumed that tags in the header were well-formed, and parsed the
16 element content after inspecting only the first char of the tag.
17 Assumptions at that stage could crash the receive process on malformed
20 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
21 While running the DKIM ACL we operate on the Permanent memory pool so that
22 variables created with "set" persist to the DATA ACL. Also (at any time)
23 DNS lookups that fail create cache records using the Permanent pool. But
24 expansions release any allocations made on the current pool - so a dnsdb
25 lookup expansion done in the DKIM ACL releases the memory used for the
26 DNS negative-cache, and bad things result. Solution is to switch to the
27 Main pool for expansions.
28 While we're in that code, add checks on the DNS cache during store_reset,
29 active in the testsuite.
30 Problem spotted, and debugging aided, by Wolfgang Breyha.
32 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
33 When none of the hosts presented to a transport match an already-open
34 connection, close it and proceed with the list. Previously we would
35 queue the message. Spotted by Lena with Yahoo, probably involving
42 JH/01 Rework error string handling in TLS interface so that the caller in
43 more cases is responsible for logging. This permits library-sourced
44 string to be attached to addresses during delivery, and collapses
45 pairs of long lines into single ones.
47 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
48 during configuration. Wildcards are allowed and expanded.
50 JH/02 Rework error string handling in DKIM to pass more info back to callers.
51 This permits better logging.
53 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
54 do not close it down and have the child transport start it up again on
55 the passed-on TCP connection. Instead, proxy the child (and any
56 subsequent ones) for TLS via a unix-domain socket channel. Logging is
57 affected: the continued delivery log lines do not have any DNSSEC, TLS
58 Certificate or OCSP information. TLS cipher information is still logged.
60 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
61 identical IP addresses on different listening ports. Will also affect
64 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
65 add noisy ifdef guards to special-case this sillyness.
66 Patch from Bernd Kuhls.
68 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
69 than 255 are no longer allowed.
71 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
72 Disable the session-cache too, which might reduce our load. Since we
73 currrectly use a new context for every connection, both as server and
74 client, there is no benefit for these.
75 GnuTLS appears to not support tickets server-side by default (we don't
76 call gnutls_session_ticket_enable_server()) but client side is enabled
77 by default on recent versions (3.1.3 +) unless the PFS priority string
80 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
81 <https://reproducible-builds.org/specs/source-date-epoch/>.
83 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
84 the check for any unsuccessful recipients did not notice the limit, and
85 erroneously found still-pending ones.
87 JH/08 Pipeline CHUNKING command and data together, on kernels that support
88 MSG_MORE. Only in-clear (not on TLS connections).
90 JH/09 Avoid using a temporary file during transport using dkim. Unless a
91 transport-filter is involved we can buffer the headers in memory for
92 creating the signature, and read the spool data file once for the
93 signature and again for transmission.
95 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
96 4.77 as the kernel support then wasn't solid, having issues in 64bit
97 mode. Now, it's been long enough. Add support for FreeBSD also.
99 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
100 case where the routing stage had gathered several addresses to send to
101 a host before calling the transport for the first, we previously failed
102 to close down TLS in the old transport process before passing the TCP
103 connection to the new process. The new one sent a STARTTLS command
104 which naturally failed, giving a failed delivery and bloating the retry
105 database. Investigation and fix prototype from Wolfgang Breyha.
107 JH/12 Fix check on SMTP command input synchronisation. Previously there were
108 false-negatives in the check that the sender had not preempted a response
109 or prompt from Exim (running as a server), due to that code's lack of
110 awareness of the SMTP input buffering.
112 PP/04 Add commandline_checks_require_admin option.
113 Exim drops privileges sanely, various checks such as -be aren't a
114 security problem, as long as you trust local users with access to their
115 own account. When invoked by services which pass untrusted data to
116 Exim, this might be an issue. Set this option in main configuration
117 AND make fixes to the calling application, such as using `--` to stop
120 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
121 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
122 responses to those, into a single TLS record each way (this usually means
123 a single packet). As a side issue, smtp_enforce_sync now works on TLS
126 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
127 affects you only if you're dancing at the edge of the param size limits.
128 If you are, and this message makes sense to you, then: raise the
129 configured limit or use OpenSSL 1.1. Nothing we can do for older
132 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
133 cmdline element to get the documented default one. Previously it was
136 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
137 are unsuable for non-message-specific reasons - eg. connection timeout,
138 banner-time rejection.
140 JH/16 Fix logging of delivery remote port, when specified by router, under
143 PP/06 Repair manualroute's ability to take options in any order, even if one
144 is the name of a transport.
147 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
149 JH/17 Change the list-building routines interface to use the expanding-string
150 triplet model, for better allocation and copying behaviour.
152 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
153 Previously it was constructed the first time a possibly-matching string
154 was met in the configuration file input during startup; now it is done
157 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
158 compatible one, to avoid the (poorly documented) possibility of a config
159 file in the working directory redirecting the DB files, possibly correpting
160 some existing file. CVE-2017-10140 assigned for BDB.
162 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
163 cache-hot. Previously, although the result was properly cached, the
164 initial verify call returned a defer.
166 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
167 the main verify for receipient in uncached-mode.
169 JH/22 Retire historical build files to an "unsupported" subdir. These are
170 defined as "ones for which we have no current evidence of testing".
172 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
173 if present. Previously it was ignored.
175 JH/24 Start using specified-initialisers in C structure init coding. This is
176 a C99 feature (it's 2017, so now considered safe).
178 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
179 if was a fixed-sized field and bitmask ops via macros; it is now more
182 PP/07 GitHub PR 56: Apply MariaDB build fix.
183 Patch provided by Jaroslav Škarvada.
185 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
186 during Coverity cleanups [4.87 JH/47]
187 Diagnosis and fix provided by Michael Fischer v. Mollard.
189 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
190 the right size to place the terminating semicolon on its own folded
191 line, the header hash was calculated to an incorrect value thanks to
192 the (relaxed) space the fold became.
194 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
195 and confused the parent.
197 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
198 which could crash as a result. This could lead to undeliverable messages.
200 JH/28 Logging: "next input sent too soon" now shows where input was truncated
203 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
204 matters on fast-turnover and PID-randomising systems, which were getting
205 out-of-order delivery.
207 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
208 a possibly-overlapping copy. The symptom was that "Remote host closed
209 connection in response to HELO" was logged instead of the actual 4xx
212 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
213 Previously only that bufferd was discarded, resulting in SYMTP command
216 JH/32 DKIM: when a message has multiple signatures matching an identity given
217 in dkim_verify_signers, run the dkim acl once for each. Previously only
218 one run was done. Bug 2189.
220 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
221 "panic the current process" to "deliberately defer". The panic log is
222 still written with the problem list name; the mail and reject logs now
223 get a temp-reject line for the message that was being handled, saying
224 something like "domains check lookup or other defer". The SMTP 451
225 message is still "Temporary local problem".
227 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
228 A crafted sequence of BDAT commands could result in in-use memory beeing
229 freed. CVE-2017-16943.
231 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
232 from SMTP input. Previously it was always done; now only done for DATA
233 and not BDAT commands. CVE-2017-16944.
235 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
236 to the message (such as an overlong header line). Previously this was
237 not done and we did not exit BDAT mode. Followon from the previous item
238 though a different problem.
244 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
245 than -2003 did; needs libidn2 in addition to libidn.
247 JH/02 The path option on a pipe transport is now expanded before use.
249 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
250 Patch provided by "Björn", documentation fix added too.
252 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
253 missing a wire-to-host endian conversion.
255 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
256 close after a BDAT command line could be taken as a following command,
257 giving a synch failure. Fix by only checking for synch immediately
258 before acknowledging the chunk.
260 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
261 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
262 macro. Patches provided by Josh Soref.
264 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
265 Previously we did not; the RFC seems ambiguous and VRFY is not listed
266 by IANA as a service extension. However, John Klensin suggests that we
269 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
270 the dkim code may be unix-mode line endings rather than smtp wire-format
271 CRLF, so prepend a CR to any bare LF.
273 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
274 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
276 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
277 the first were themselves being wrongly included in the feed into dkim
278 processing; with most chunk sizes in use this resulted in an incorrect
279 body hash calculated value.
281 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
282 DKIM signature block, for verification. Although advised against by
283 standards it is specifically not ruled illegal.
285 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
287 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
288 missing a body hash (the bh= tag).
290 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
291 It seems that HAProxy sends the Proxy Protocol information in clear and
292 only then does a TLS startup, so do the same.
294 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
295 TCP connections (such as for Spamd) unless the daemon successfully set
296 Fast Open mode on its listening sockets. This fixes breakage seen on
297 too-old kernels or those not configured for Fast Open, at the cost of
298 requiring both directions being enabled for TFO, and TFO never being used
299 by non-daemon-related Exim processes.
301 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
302 endings, at least on the first header line. Try to canonify any that get
303 past that check, despite the cost.
305 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
306 now limited to an arbitrary five deep, while parsing addresses with the
307 strip_excess_angle_brackets option enabled.
309 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
310 instead leave the unprompted TLS handshake in socket buffer for the
311 TLS library to consume.
313 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
315 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
317 JH/16 Drop variables when they go out of scope. Memory management drops a whole
318 region in one operation, for speed, and this leaves assigned pointers
319 dangling. Add checks run only under the testsuite which checks all
320 variables at a store-reset and panics on a dangling pointer; add code
321 explicitly nulling out all the variables discovered. Fixes one known
322 bug: a transport crash, where a dangling pointer for $sending_ip_address
323 originally assigned in a verify callout, is re-used.
325 PP/06 Drop '.' from @INC in various Perl scripts.
327 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
329 PP/08 Reduce a number of compilation warnings under clang; building with
330 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
331 should be warning-free.
333 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
335 HS/01 Fix portability problems introduced by PP/08 for platforms where
336 realloc(NULL) is not equivalent to malloc() [SunOS et al].
338 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
339 chunk. This allows us to accept broken chunked messages. We need a more
340 general solution here.
342 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
343 already-broken messages in the queue.
345 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
347 JH/19 Fix reference counting bug in routing-generated-address tracking.
353 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
354 supports it and a size is available (ie. the sending peer gave us one).
356 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
357 years of being deprecated). The replacements are the ACLs
358 acl_smtp_mime and acl_not_smtp_mime.
360 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
361 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
362 or even an in-clear connection were permitted. Now, if the host lookup
363 was dnssec and dane was requested then the host is only used if the
364 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
365 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
366 if one fails this test.
367 This means that a poorly-configured remote DNS will make it incommunicado;
368 but it protects against a DNS-interception attack on it.
370 JH/04 Bug 1810: make continued-use of an open smtp transport connection
371 non-noisy when a race steals the message being considered.
373 JH/05 If main configuration option tls_certificate is unset, generate a
374 self-signed certificate for inbound TLS connections.
376 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
377 in rewrites and routers.
379 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
380 and logged a warning sing 4.83; now they are a configuration file error.
382 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
383 (lacking @domain). Apply the same qualification processing as RCPT.
385 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
387 JH/10 Support ${sha256:} applied to a string (as well as the previous
390 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
391 a cutthrough deliver is pending, as we always want to make a connection.
392 This also avoids re-routing the message when later placing the cutthrough
393 connection after a verify cache hit.
394 Do not update it with the verify result either.
396 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
397 when routing results in more than one destination address.
399 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
400 signing (which inhibits the cutthrough capability). Previously only
401 the presence of an option was tested; now an expansion evaluating as
402 empty is permissible (obviously it should depend only on data available
403 when the cutthrough connection is made).
405 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
406 the relevant preceding SMTP command did not note the pipelining mode.
408 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
409 Previously they were not counted.
411 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
412 as one having no matching records. Previously we deferred the message
413 that needed the lookup.
415 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
416 distinguished as "(=".
418 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
419 for missing MX records. Previously it only worked for missing A records.
421 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
423 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
424 after the data-go-ahead and data-ack. Patch from Jason Betts.
426 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
427 even for a "none" policy. Patch from Tony Meyer.
429 JH/22 Fix continued use of a connection for further deliveries. If a port was
430 specified by a router, it must also match for the delivery to be
433 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
434 When one of the recipients of a message was unsuitable for the connection
435 (has no matching addresses), we lost track of needing to mark it
436 deferred. As a result mail would be lost.
438 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
440 JH/25 Decoding ACL controls is now done using a binary search; the source code
441 takes up less space and should be simpler to maintain. Merge the ACL
442 condition decode tables also, with similar effect.
444 JH/26 Fix problem with one_time used on a redirect router which returned the
445 parent address unchanged. A retry would see the parent address marked as
446 delivered, so not attempt the (identical) child. As a result mail would
449 JH/27 Fix a possible security hole, wherein a process operating with the Exim
450 UID can gain a root shell. Credit to http://www.halfdog.net/ for
451 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
454 JH/28 Enable {spool,log} filesystem space and inode checks as default.
455 Main config options check_{log,spool}_{inodes,space} are now
456 100 inodes, 10MB unless set otherwise in the configuration.
458 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
459 Previously it only applied to the main-section connection policy
462 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
464 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
465 by me. Added RFC7919 DH primes as an alternative.
467 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
468 are not in the system include path.
470 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
471 GnuTLS, when a session startup failed (eg because the client disconnected)
472 Exim did stdio operations after fclose. This was exposed by a recent
473 change which nulled out the file handle after the fclose.
475 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
476 signed directly by the cert-signing cert, rather than an intermediate
477 OCSP-signing cert. This is the model used by LetsEncrypt.
479 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
481 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
482 an incoming connection.
484 HS/02 Bug 1802: Do not half-close the connection after sending a request
487 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
488 fallback to "prime256v1".
490 JH/34 SECURITY: Use proper copy of DATA command in error message.
491 Could leak key material. Remotely exploitable. CVE-2016-9963.
497 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
498 and 3.4.4 - once the server is enabled to respond to an OCSP request
499 it does even when not requested, resulting in a stapling non-aware
500 client dropping the TLS connection.
502 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
503 support variable-length bit vectors. No functional change.
505 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
506 The I= interface field on outgoing lines is now after the H= remote
507 host field, same as incoming lines. There is a separate
508 outgoing_interface log selector which allows you to disable the
511 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
512 If not running log_selector +smtp_connection the mainlog would be held
513 open indefinitely after a "too many connections" event, including to a
514 deleted file after a log rotate. Leave the per net connection logging
515 leaving it open for efficiency as that will be quickly detected by the
516 check on the next write.
518 HS/01 Bug 1671: Fix post transport crash.
519 Processing the wait-<transport> messages could crash the delivery
520 process if the message IDs didn't exist for some reason. When
521 using 'split_spool_directory=yes' the construction of the spool
522 file name failed already, exposing the same netto behaviour.
524 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
525 mime_regex ACL conditions.
527 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
528 to DSN fail messages (bounces): remote IP, remote greeting, remote response
529 to HELO, local diagnostic string.
531 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
532 log line to debug. Even when configured with a tls authenticator many
533 client connections are expected to not authenticate in this way, so
534 an authenticate fail is not an error.
536 HS/02 Add the Exim version string to the process info. This way exiwhat
537 gives some more detail about the running daemon.
539 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
540 matter for fast-change records such as DNSBLs.
542 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
543 retry record, even if constant. There may be multiple transports with
544 different interface settings and the retry behaviour needs to be kept
547 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
549 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
551 JH/10 Bug 840: fix log_defer_output option of pipe transport
553 JH/11 Bug 830: use same host for all RCPTS of a message, even under
554 hosts_randomize. This matters a lot when combined with mua_wrapper.
556 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
557 ${quote_pgsql:<string>} operator.
559 JH/13 Bug 1708: avoid misaligned access in cached lookup.
561 JH/14 Change header file name for freeradius-client. Relevant if compiling
562 with Radius support; from the Gentoo tree and checked under Fedora.
564 JH/15 Bug 1712: Introduce $prdr_requested flag variable
566 JH/16 Bug 1714: Permit an empty string as expansion result for transport
567 option transport_filter, meaning no filtering.
569 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
571 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
572 defaults to "*" (all hosts). The variable is now available when not built
573 with TLS, default unset, mainly to enable keeping the testsuite sane.
574 If a server certificate is not supplied (via tls_certificate) an error is
575 logged, and clients will find TLS connections fail on startup. Presumably
576 they will retry in-clear.
577 Packagers of Exim are strongly encouraged to create a server certificate
578 at installation time.
580 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
581 with the $config_file variable.
583 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
584 in transport context, after the attempt, and per-recipient. The latter type
585 is per host attempted. The event data is the error message, and the errno
586 information encodes the lookup type (A vs. MX) used for the (first) host,
587 and the trailing two digits of the smtp 4xx response.
589 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
590 to write to mainlog (or rejectlog, paniclog) in the window between file
591 creation and permissions/ownership being changed. Particularly affects
592 installations where exicyclog is run as root, rather than exim user;
593 result is that the running daemon panics and dies.
595 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
597 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
598 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
599 "pri" and "weight". Note that the previous implicit priority given by the
600 list order is no longer honoured.
602 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
605 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
606 by defining SUPPORT_SOCKS.
608 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
609 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
610 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
611 variables are renamed to proxy_{local,external}_{address,port}.
613 JH/27 Move Internationalisation support from Experimental to mainline, enabled
614 for a build by defining SUPPORT_I18N
616 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
617 of the query string, and make ${quote_redis:} do that quoting.
619 JH/29 Move Events support from Experimental to mainline, enabled by default
620 and removable for a build by defining DISABLE_EVENT.
622 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
624 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
625 cached by the daemon.
627 JH/32 Move Redis support from Experimental to mainline, enabled for a build
628 by defining LOOKUP_REDIS. The libhiredis library is required.
630 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
631 keys are given for lookup.
633 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
634 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
635 only supported when built with TLS support. The PolarSSL SHA routines
636 are still used when the TLS library is too old for convenient support.
638 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
639 openssl_options), for security. OpenSSL forces this from version 1.1.0
640 server-side so match that on older versions.
642 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
643 allocation for $value could be released as the expansion processing
644 concluded, but leaving the global pointer active for it.
646 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
647 and to use the domains and local_parts ACL conditions.
649 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
650 incorrectly not doubled on cutthrough transmission, hence seen as a
651 body-termination at the receiving system - resulting in truncated mails.
652 Commonly the sender saw a TCP-level error, and retransmitted the message
653 via the normal store-and-forward channel. This could result in duplicates
654 received - but deduplicating mailstores were liable to retain only the
655 initial truncated version.
657 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
659 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
661 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
662 we're in there, support oversigning also; bug 1309.
664 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
666 HS/04 Add support for keep_environment and add_environment options.
668 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
669 either intentional arithmetic overflow during PRNG, or testing config-
672 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
673 delivery resulted in actual delivery. Cancel cutthrough before DATA
676 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
677 rejects a recipient: pass the reject to the originator.
679 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
680 Many were false-positives and ignorable, but it's worth fixing the
683 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
684 for the new environment-manipulation done at startup. Move the routines
685 from being local to tls.c to being global via the os.c file.
687 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
688 an extract embedded as result-arg for a map, the first arg for extract
689 is unavailable so we cannot tell if this is a numbered or keyed
690 extraction. Accept either.
696 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
699 JH/02 The smtp transport option "multi_domain" is now expanded.
701 JH/03 The smtp transport now requests PRDR by default, if the server offers
704 JH/04 Certificate name checking on server certificates, when exim is a client,
705 is now done by default. The transport option tls_verify_cert_hostnames
706 can be used to disable this per-host. The build option
707 EXPERIMENTAL_CERTNAMES is withdrawn.
709 JH/05 The value of the tls_verify_certificates smtp transport and main options
710 default to the word "system" to access the system default CA bundle.
711 For GnuTLS, only version 3.0.20 or later.
713 JH/06 Verification of the server certificate for a TLS connection is now tried
714 (but not required) by default. The verification status is now logged by
715 default, for both outbound TLS and client-certificate supplying inbound
718 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
721 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
722 Status Notification (bounce) messages are now MIME format per RFC 3464.
723 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
724 under the control of the dsn_advertise_hosts option, and routers may
725 have a dsn_lasthop option.
727 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
728 default, modifiable by a malware= option. The list separator for
729 the options can now be changed in the usual way. Bug 68.
731 JH/10 The smtp_receive_timeout main option is now expanded before use.
733 JH/11 The incoming_interface log option now also enables logging of the
734 local interface on delivery outgoing connections.
736 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
737 if the interface and destination host and port all match.
739 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
742 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
743 Patch from Andrew Lewis.
745 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
746 now supports optional time-restrictions, weighting, and priority
747 modifiers per server. Patch originally by <rommer@active.by>.
749 JH/16 The spamd_address main option now supports a mixed list of local
750 and remote servers. Remote servers can be IPv6 addresses, and
751 specify a port-range.
753 JH/17 Bug 68: The spamd_address main option now supports an optional
754 timeout value per server.
756 JH/18 Bug 1581: Router and transport options headers_add/remove can
757 now have the list separator specified.
759 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
762 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
765 JH/21 Support for the A6 type of dns record is withdrawn.
767 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
768 rather than the verbs used.
770 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
771 from 255 to 1024 chars.
773 JH/24 Verification callouts now attempt to use TLS by default.
775 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
776 are generic router options now. The defaults didn't change.
778 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
779 Original patch from Alexander Shikoff, worked over by JH.
781 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
782 files if ZCAT_COMMAND is not executable.
784 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
786 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
788 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
789 Normally benign, it bites when the pair was led to by a CNAME;
790 modern usage is to not canonicalize the domain to a CNAME target
791 (and we were inconsistent anyway for A-only vs AAAA+A).
793 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
795 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
796 when evaluating $sender_host_dnssec.
798 JH/31 Check the HELO verification lookup for DNSSEC, adding new
799 $sender_helo_dnssec variable.
801 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
803 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
805 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
807 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
808 documented as working, but never had. Support all but $spam_report.
810 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
811 added for tls authenticator.
813 HS/03 Add perl_taintmode main config option
819 TL/01 When running the test suite, the README says that variables such as
820 no_msglog_check are global and can be placed anywhere in a specific
821 test's script, however it was observed that placement needed to be near
822 the beginning for it to behave that way. Changed the runtest perl
823 script to read through the entire script once to detect and set these
824 variables, reset to the beginning of the script, and then run through
825 the script parsing/test process like normal.
827 TL/02 The BSD's have an arc4random API. One of the functions to induce
828 adding randomness was arc4random_stir(), but it has been removed in
829 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
830 function when detected.
832 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
833 cause callback expansion.
835 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
836 syntax errors in an expansion can be treated as a string instead of
837 logging or causing an error, due to the internal use of bool_lax
838 instead of bool when processing it.
840 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
841 server certificates when making smtp deliveries.
843 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
845 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
847 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
849 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
850 Merged patch from Sebastian Wiedenroth.
852 JH/05 Fix results-pipe from transport process. Several recipients, combined
853 with certificate use, exposed issues where response data items split
854 over buffer boundaries were not parsed properly. This eventually
855 resulted in duplicates being sent. This issue only became common enough
856 to notice due to the introduction of connection certificate information,
857 the item size being so much larger. Found and fixed by Wolfgang Breyha.
859 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
860 size buffer was used, resulting in syntax errors when an expansion
863 JH/07 Add support for directories of certificates when compiled with a GnuTLS
864 version 3.3.6 or later.
866 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
867 is EXPERIMENTAL_EVENT, the main-configuration and transport options
868 both become "event_action", the variables become $event_name, $event_data
869 and $event_defer_errno. There is a new variable $verify_mode, usable in
870 routers, transports and related events. The tls:cert event is now also
871 raised for inbound connections, if the main configuration event_action
874 TL/06 In test suite, disable OCSP for old versions of openssl which contained
875 early OCSP support, but no stapling (appears to be less than 1.0.0).
877 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
878 server certificate names available under the smtp transport option
879 "tls_verify_cert_hostname" now do not permit multi-component wildcard
882 JH/10 Time-related extraction expansions from certificates now use the main
883 option "timezone" setting for output formatting, and are consistent
884 between OpenSSL and GnuTLS compilations. Bug 1541.
886 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
887 encoded parameter in the incoming message. Bug 1558.
889 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
890 include certificate info, eximon was claiming there were spoolfile
893 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
895 JH/14 Log delivery-related information more consistently, using the sequence
896 "H=<name> [<ip>]" wherever possible.
898 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
899 are problematic for Debian distribution, omit them from the release
902 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
904 JH/16 Fix string representation of time values on 64bit time_t architectures.
907 JH/17 Fix a null-indirection in certextract expansions when a nondefault
908 output list separator was used.
913 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
914 checkers that were complaining about end of non-void function with no
917 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
918 This was a regression introduced in 4.83 by another bugfix.
920 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
922 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
923 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
929 TF/01 Correctly close the server side of TLS when forking for delivery.
931 When a message was received over SMTP with TLS, Exim failed to clear up
932 the incoming connection properly after forking off the child process to
933 deliver the message. In some situations the subsequent outgoing
934 delivery connection happened to have the same fd number as the incoming
935 connection previously had. Exim would try to use TLS and fail, logging
936 a "Bad file descriptor" error.
938 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
939 utilities have not been installed.
941 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
942 temporary space as the ACL may create new global variables.
944 TL/01 LDAP support uses per connection or global context settings, depending
945 upon the detected version of the libraries at build time.
947 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
948 to extract and use the src ip:port in logging and expansions as if it
949 were a direct connection from the outside internet. PPv2 support was
950 updated based on HAProxy spec change in May 2014.
952 JH/02 Add ${listextract {number}{list}{success}{fail}}.
954 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
955 Properly escape header and check for NULL return.
957 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
960 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
962 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
963 characters in header names, implemented as a verify condition.
964 Contributed by Michael Fischer v. Mollard.
966 TL/05 Rename SPF condition results err_perm and err_temp to standardized
967 results permerror and temperror. Previous values are deprecated but
968 still accepted. In a future release, err_perm and err_temp will be
969 completely removed, which will be a backward incompatibility if the
970 ACL tests for either of these two old results. Patch contributed by
971 user bes-internal on the mailing list.
973 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
975 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
976 selectors, in both main and reject logs.
978 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
981 JH/07 Add malware type "sock" for talking to simple daemon.
983 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
985 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
986 routers/transports under cutthrough routing.
988 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
989 numbers. Touch up "bool" conditional to keep the same definition.
991 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
993 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
995 JH/12 Expand items in router/transport headers_add or headers_remove lists
996 individually rather than the list as a whole. Bug 1452.
998 Required for reasonable handling of multiple headers_ options when
999 they may be empty; requires that headers_remove items with embedded
1000 colons must have them doubled (or the list-separator changed).
1002 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1003 view the policy declared in the DMARC record. Currently, $dmarc_status
1004 is a combined value of both the record presence and the result of the
1007 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1009 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1010 dnslookup router and the smtp transport (applying to the forward
1013 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1014 of ldap servers used for a specific lookup. Patch provided by Heiko
1017 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1018 New variable $lookup_dnssec_authenticated for observability.
1020 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1021 Patch submitted by Lars Timman.
1023 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1025 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1026 Requires trusted mode and valid format message id, aborts otherwise.
1027 Patch contributed by Heiko Schlichting.
1029 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1030 certextract with support for various fields. Bug 1358.
1032 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1033 is requested by default, modifiable by smtp transport option
1036 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1037 operate on certificate variables to give certificate fingerprints
1038 Also new ${sha256:cert_variable}.
1040 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1042 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1043 Christian Aistleitner.
1045 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1047 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1048 file. Patch from Wolfgang Breyha.
1050 JH/25 Expand the coverage of the delivery $host and $host_address to
1051 client authenticators run in verify callout. Bug 1476.
1053 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1054 align with daemon_smtp_ports. Bug 72.
1056 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1057 support and error reporting did not work properly.
1059 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1060 and is readable. Patch from Andrew Colin Kissa.
1062 TL/14 Enhance documentation of ${run expansion and how it parses the
1063 commandline after expansion, particularly in the case when an
1064 unquoted variable expansion results in an empty value.
1066 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1068 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1069 by renaming away from it. Addresses GH issue 15, reported by
1072 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1074 TL/15 SECURITY: prevent double expansion in math comparison functions
1075 (can expand unsanitized data). Not remotely exploitable.
1082 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1084 PP/02 Make -n do something, by making it not do something.
1085 When combined with -bP, the name of an option is not output.
1087 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1090 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1091 $sender_host_name and config options to manage this, and basic check
1094 PP/05 DSCP support for outbound connections and control modifier for inbound.
1096 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1097 (Only plugin which currently uses this is kerberos4, which nobody should
1098 be using, but we should make it available and other future plugins might
1099 conceivably use it, even though it would break NAT; stuff *should* be
1100 using channel bindings instead).
1102 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1103 name; added for Sendmail compatibility; requires admin caller.
1104 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1105 just ignore it); requires trusted caller.
1106 Also parse but ignore: -Ac -Am -X<logfile>
1109 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1111 TL/02 Add +smtp_confirmation as a default logging option.
1113 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1114 Patch by Magnus Holmgren from 2007-02-20.
1116 TL/04 Bugzilla 1281 - Spec typo.
1117 Bugzilla 1283 - Spec typo.
1118 Bugzilla 1290 - Spec grammar fixes.
1120 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1122 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1124 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1125 mailing list by by Dmitry Isaikin.
1127 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1129 JH/02 Support "G" suffix to numbers in ${if comparisons.
1131 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1133 NM/01 Bugzilla 1197 - Spec typo
1134 Bugzilla 1196 - Spec examples corrections
1136 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1138 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1139 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1142 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1143 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1145 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1146 "acl {{name}{arg}...}", and optional args on acl condition
1149 JH/05 Permit multiple router/transport headers_add/remove lines.
1151 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1153 JH/07 Avoid using a waiting database for a single-message-only transport.
1154 Performance patch from Paul Fisher. Bugzilla 1262.
1156 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1159 JH/09 Add $headers_added variable, with content from use of ACL modifier
1160 add_header (but not yet added to the message). Bugzilla 199.
1162 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1163 Pulled from Bugzilla 817 by Wolfgang Breyha.
1165 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1167 (nb: this is the same fix as in Exim 4.80.1)
1169 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1172 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1175 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1177 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1178 advertise SMTP AUTH mechanism to us, instead of a generic
1179 protocol violation error. Also, make Exim more robust to bad
1180 data from the Dovecot auth socket.
1182 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1184 When a queue runner is handling a message, Exim first routes the
1185 recipient addresses, during which it prunes them based on the retry
1186 hints database. After that it attempts to deliver the message to
1187 any remaining recipients. It then updates the hints database using
1190 So if a recipient address works intermittently, it can get repeatedly
1191 deferred at routing time. The retry hints record remains fresh so the
1192 address never reaches the final cutoff time.
1194 This is a fairly common occurrence when a user is bumping up against
1195 their storage quota. Exim had some logic in its local delivery code
1196 to deal with this. However it did not apply to per-recipient defers
1197 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1199 This change adds a proper retry rule check during routing so that the
1200 final cutoff time is checked against the message's age. We only do
1201 this check if there is an address retry record and there is not a
1202 domain retry record; this implies that previous attempts to handle
1203 the address had the retry_use_local_parts option turned on. We use
1204 this as an approximation for the destination being like a local
1205 delivery, as in LMTP.
1207 I suspect this new check makes the old local delivery cutoff check
1208 redundant, but I have not verified this so I left the code in place.
1210 TF/02 Correct gecos expansion when From: is a prefix of the username.
1212 Test 0254 submits a message to Exim with the header
1216 When I ran the test suite under the user fanf2, Exim expanded
1217 the header to contain my full name, whereas it should have added
1218 a Resent-Sender: header. It erroneously treats any prefix of the
1219 username as equal to the username.
1221 This change corrects that bug.
1223 GF/01 DCC debug and logging tidyup
1224 Error conditions log to paniclog rather than rejectlog.
1225 Debug lines prefixed by "DCC: " to remove any ambiguity.
1227 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1229 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1230 Bug spotted by Jeremy Harris; was flawed since initial commit.
1231 Would have resulted in OCSP responses post-SNI triggering an Exim
1232 NULL dereference and crash.
1234 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1236 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1237 Bug detection, analysis and fix by Samuel Thibault.
1238 Bugzilla 1331, Debian bug #698092.
1240 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1242 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1243 Server implementation by Todd Lyons, client by JH.
1244 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1245 config variable "prdr_enable" controls whether the server
1246 advertises the facility. If the client requests PRDR a new
1247 acl_data_smtp_prdr ACL is called once for each recipient, after
1248 the body content is received and before the acl_smtp_data ACL.
1249 The client is controlled by both of: a hosts_try_prdr option
1250 on the smtp transport, and the server advertisement.
1251 Default client logging of deliveries and rejections involving
1252 PRDR are flagged with the string "PRDR".
1254 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1255 fclose(). Diagnosis by Todd Lyons.
1257 PP/17 Update configure.default to handle IPv6 localhost better.
1258 Patch by Alain Williams (plus minor tweaks).
1261 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1262 This is now consistent with GnuTLS, and is now documented: the
1263 previous undocumented portable approach to treating the option as
1264 unset was to force an expansion failure. That still works, and
1265 an empty string is now equivalent.
1267 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1268 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1269 not performing validation itself.
1271 PP/20 Added force_command boolean option to pipe transport.
1272 Patch from Nick Koston, of cPanel Inc.
1274 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1277 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1279 PP/21 Fix eximon continuous updating with timestamped log-files.
1280 Broken in a format-string cleanup in 4.80, missed when I repaired the
1281 other false fix of the same issue.
1282 Report and fix from Heiko Schlichting.
1285 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1286 Report from Prashanth Katuri.
1288 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1289 It's SecureTransport, so affects any MacOS clients which use the
1290 system-integrated TLS libraries, including email clients.
1292 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1293 using a MIME ACL for non-SMTP local injection.
1294 Report and assistance in diagnosis by Warren Baker.
1296 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1298 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1300 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1301 last id that failed so it may be referenced in subsequent ACL's.
1303 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1306 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1307 ldap library initialization, allowing self-signed CA's to be
1308 used. Also properly sets require_cert option later in code by
1309 using NULL (global ldap config) instead of ldap handle (per
1310 session). Bug diagnosis and testing by alxgomz.
1312 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1313 the src/util/ subdirectory.
1315 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1316 renamed to Transport Post Delivery Action by Jeremy Harris, as
1319 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1320 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1321 redis_servers = needs to be configured which will be used by the redis
1322 lookup. Patch from Warren Baker, of The Packet Hub.
1324 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1326 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1327 hostname or reverse DNS when processing a host list. Used suggestions
1328 from multiple comments on this bug.
1330 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1332 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1333 Missed a few lines, added it to make the runtest require no keyboard
1336 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1337 contains upper case chars. Make router use caseful_local_part.
1339 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1340 support when GnuTLS has been built with p11-kit.
1346 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1348 This, or similar/improved, will also be change PP/11 of 4.82.
1354 PP/01 Handle short writes when writing local log-files.
1355 In practice, only affects FreeBSD (8 onwards).
1356 Bugzilla 1053, with thanks to Dmitry Isaikin.
1358 NM/01 Bugzilla 949 - Documentation tweak
1360 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1363 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1365 PP/02 Implemented gsasl authenticator.
1367 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1369 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1370 `pkg-config foo` for cflags/libs.
1372 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1373 with rest of GSASL and with heimdal_gssapi.
1375 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1376 `pkg-config foo` for cflags/libs for the TLS implementation.
1378 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1379 properties get this fed in as external SSF. A number of robustness
1380 and debugging improvements to the cyrus_sasl authenticator.
1382 PP/08 cyrus_sasl server now expands the server_realm option.
1384 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1385 Patch by Jeremy Harris.
1387 PP/10 Added dbmjz lookup type.
1389 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1391 PP/12 MAIL args handles TAB as well as SP, for better interop with
1392 non-compliant senders.
1393 Analysis and variant patch by Todd Lyons.
1395 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1396 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1397 Patch from Dirk Mueller <dmueller@suse.com>
1399 PP/13 tls_peerdn now print-escaped for spool files.
1400 Observed some $tls_peerdn in wild which contained \n, which resulted
1401 in spool file corruption.
1403 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1404 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1405 or write after TLS renegotiation, which otherwise led to messages
1408 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1409 as a tracking header (ie: a signed header comes before the signature).
1410 Patch from Wolfgang Breyha.
1412 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1413 comma-sep list; embedded commas doubled.
1415 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1417 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1419 Report and patch from Dmitry Banschikov.
1421 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1422 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1423 failures. We appear to now support TLS1.1+ with Exim.
1425 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1426 lets Exim select keys and certificates based upon TLS SNI from client.
1427 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1428 before an outbound SMTP session. New log_selector, +tls_sni.
1430 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1431 NULL dereference. Report and patch from Alun Jones.
1433 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1434 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1435 Not seeing resolver debug output on NetBSD, but suspect this is a
1436 resolver implementation change.
1438 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1439 Left warnings. Added "eximon gdb" invocation mode.
1441 PP/21 Defaulting "accept_8bitmime" to true, not false.
1443 PP/22 Added -bw for inetd wait mode support.
1445 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1446 locate the relevant includes and libraries. Made this the default.
1448 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1449 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1451 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1452 This may cause build issues on older platforms.
1454 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1455 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1456 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1457 Added SNI support via GnuTLS too.
1458 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1460 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1462 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1463 Applied second patch from Janne, implementing suggestion to default
1464 multiple-strings-in-record handling to match SPF spec.
1466 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1468 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1469 read-only, out of scope).
1470 Patch from Wolfgang Breyha, report from Stuart Northfield.
1472 PP/29 Fix three issues highlighted by clang analyser static analysis.
1473 Only crash-plausible issue would require the Cambridge-specific
1474 iplookup router and a misconfiguration.
1475 Report from Marcin Mirosław.
1477 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1479 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1480 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1481 As part of this, removing so much warning spew let me fix some minor
1482 real issues in debug logging.
1484 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1485 assignment on my part. Fixed.
1487 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1488 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1489 Janne Snabb (who went above and beyond: thank you).
1491 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1492 string otherwise requires a connection and a bunch more work and it's
1493 relatively easy to get wrong. Should also expose TLS library linkage
1496 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1497 64-bit ${eval} (JH/03).
1499 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1500 GNU libc to support some of the 64-bit stuff, should not lead to
1501 conflicts. Defined before os.h is pulled in, so if a given platform
1502 needs to override this, it can.
1504 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1505 protection layer was required, which is not implemented.
1506 Bugzilla 1254, patch from Wolfgang Breyha.
1508 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1509 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1510 tls_dhparam take prime identifiers. Also unbreak combination of
1511 OpenSSL+DH_params+TLSSNI.
1513 PP/39 Disable SSLv2 by default in OpenSSL support.
1519 PP/01 Solaris build fix for Oracle's LDAP libraries.
1520 Bugzilla 1109, patch from Stephen Usher.
1522 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1524 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1527 TF/02 Fix a couple more cases where we did not log the error message
1528 when unlink() failed. See also change 4.74-TF/03.
1530 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1531 lock up or crash if it happened to be inside a call to libc when it
1532 got a SIGUSR1 from exiwhat.
1534 The SIGUSR1 handler appends the current process status to the process
1535 log which is later printed by exiwhat. It used to use the general
1536 purpose logging code to do this, but several functions it calls are
1537 not safe for signals.
1539 The new output code in the SIGUSR1 handler is specific to the process
1540 log, and simple enough that it's easy to inspect for signal safety.
1541 Removing some special cases also simplifies the general logging code.
1542 Removing the spurious timestamps from the process log simplifies
1545 TF/04 Improved ratelimit ACL condition.
1547 The /noupdate option has been deprecated in favour of /readonly which
1548 has clearer semantics. The /leaky, /strict, and /readonly update modes
1549 are mutually exclusive. The update mode is no longer included in the
1550 database key; it just determines when the database is updated. (This
1551 means that when you upgrade Exim will forget old rate measurements.)
1553 Exim now checks that the per_* options are used with an update mode that
1554 makes sense for the current ACL. For example, when Exim is processing a
1555 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1556 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1557 must specify per_mail/readonly. If you omit the update mode it defaults to
1558 /leaky where that makes sense (as before) or /readonly where required.
1560 The /noupdate option is now undocumented but still supported for
1561 backwards compatibility. It is equivalent to /readonly except that in
1562 ACLs where /readonly is required you may specify /leaky/noupdate or
1563 /strict/noupdate which are treated the same as /readonly.
1565 A useful new feature is the /count= option. This is a generalization
1566 of the per_byte option, so that you can measure the throughput of other
1567 aggregate values. For example, the per_byte option is now equivalent
1568 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1570 The per_rcpt option has been generalized using the /count= mechanism
1571 (though it's more complicated than the per_byte equivalence). When it is
1572 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1573 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1574 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1575 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1576 Note that using per_rcpt with a non-readonly update mode in more than
1577 one ACL will cause the recipients to be double-counted. (The per_mail
1578 and per_byte options don't have this problem.)
1580 The handling of very low rates has changed slightly. If the computed rate
1581 is less than the event's count (usually one) then this event is the first
1582 after a long gap. In this case the rate is set to the same as this event's
1583 count, so that the first message of a spam run is counted properly.
1585 The major new feature is a mechanism for counting the rate of unique
1586 events. The new per_addr option counts the number of different
1587 recipients that someone has sent messages to in the last time period. It
1588 behaves like per_rcpt if all the recipient addresses are different, but
1589 duplicate recipient addresses do not increase the measured rate. Like
1590 the /count= option this is a general mechanism, so the per_addr option
1591 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1592 example, measure the rate that a client uses different sender addresses
1593 with the options per_mail/unique=$sender_address. There are further
1594 details in the main documentation.
1596 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1598 TF/06 Removed a few PCRE remnants.
1600 TF/07 Automatically extract Exim's version number from tags in the git
1601 repository when doing development or release builds.
1603 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1604 Bugzilla 879. Patch from Paul Fisher.
1606 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1607 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1610 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1612 PP/05 Variable $av_failed, true if the AV scanner deferred.
1613 Bugzilla 1078. Patch from John Horne.
1615 PP/06 Stop make process more reliably on build failure.
1616 Bugzilla 1087. Patch from Heiko Schlittermann.
1618 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1619 Bugzilla 1089. Patch from Heiko Schlittermann.
1621 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1622 Bugzilla 1131. Patch from Holger Weiß.
1624 PP/09 Handle IPv6 addresses with SPF.
1625 Bugzilla 860. Patch from Wolfgang Breyha.
1627 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1629 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1632 PP/11 match_* no longer expand right-hand-side by default.
1633 New compile-time build option, EXPAND_LISTMATCH_RHS.
1634 New expansion conditions, "inlist", "inlisti".
1636 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1638 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1640 PP/14 fix log_write() format string regression from TF/03.
1641 Bugzilla 1152. Patch from Dmitry Isaikin.
1647 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1649 PP/02 Harmonised TLS library version reporting; only show if debugging.
1650 Layout now matches that introduced for other libraries in 4.74 PP/03.
1652 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1654 PP/04 New "dns_use_edns0" global option.
1656 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1659 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1660 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1662 TK/01 Updated PolarSSL code to 0.14.2.
1663 Bugzilla 1097. Patch from Andreas Metzler.
1665 PP/07 Catch divide-by-zero in ${eval:...}.
1666 Fixes bugzilla 1102.
1668 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1671 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1672 format-string attack -- SECURITY: remote arbitrary code execution.
1674 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1675 time unintentionally subject to list matching rules, letting the header
1676 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1677 arbitrary string expansion). This allowed for information disclosure.
1679 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1680 INT_MIN/-1 -- value coerced to INT_MAX.
1686 NM/01 Workaround for PCRE version dependency in version reporting
1689 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1690 This fixes portability to compilers other than gcc, notably
1691 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1693 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1694 makefiles for portability to HP-UX and POSIX correctness.
1696 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1697 Also via indented variable definition in the Makefile.
1698 (Debugging by Oliver Heesakkers).
1700 PP/02 Restore caching of spamd results with expanded spamd_address.
1701 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1703 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1704 Improves build reliability. Fix from: Frank Elsner
1706 NM/02 Fix wide character breakage in the rfc2047 coding
1707 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1709 NM/03 Allow underscore in dnslist lookups
1710 Fixes bug 1026. Patch from Graeme Fowler
1712 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1713 Code patches from Adam Ciarcinski of NetBSD.
1715 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1718 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1719 is logged, to avoid truncation. Patch from John Horne.
1721 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1722 Patch from Jakob Hirsch.
1724 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1725 SQL string expansion failure details.
1726 Patch from Andrey Oktyabrski.
1728 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1729 Patch from Simon Arlott.
1731 PP/09 New lookups functionality failed to compile on old gcc which rejects
1732 extern declarations in function scope.
1733 Patch from Oliver Fleischmann
1735 PP/10 Use sig_atomic_t for flags set from signal handlers.
1736 Check getgroups() return and improve debugging.
1737 Fixed developed for diagnosis in bug 927 (which turned out to be
1740 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1741 Patch from Mark Zealey.
1743 PP/12 Bugzilla 1056: Improved spamd server selection.
1744 Patch from Mark Zealey.
1746 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1747 Based on patch from Heiko Schlittermann.
1749 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1750 Patch from Uwe Doering, sign-off by Michael Haardt.
1752 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1753 variable declaration deep within a block. Bug and patch from
1756 PP/15 lookups-Makefile IRIX compatibility coercion.
1758 PP/16 Make DISABLE_DKIM build knob functional.
1760 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1761 Patch by Simon Arlott
1763 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1764 variable argument macros. Our copy now differs from upstream.
1770 TF/01 Failure to get a lock on a hints database can have serious
1771 consequences so log it to the panic log.
1773 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1774 controlled using the smtp_confirmation log selector.
1776 TF/03 Include the error message when we fail to unlink a spool file.
1778 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1779 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1780 for maintaining out-of-tree patches for some time.
1782 PP/01 Bugzilla 139: Documentation and portability issues.
1783 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1784 Handle per-OS dynamic-module compilation flags.
1786 PP/02 Let /dev/null have normal permissions.
1787 The 4.73 fixes were a little too stringent and complained about the
1788 permissions on /dev/null. Exempt it from some checks.
1789 Reported by Andreas M. Kirchwitz.
1791 PP/03 Report version information for many libraries, including
1792 Exim version information for dynamically loaded libraries. Created
1793 version.h, now support a version extension string for distributors
1794 who patch heavily. Dynamic module ABI change.
1796 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1797 privilege escalation vulnerability whereby the Exim run-time user
1798 can cause root to append content of the attacker's choosing to
1801 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1804 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1805 If dropping privileges for untrusted macros, we disabled normal logging
1806 on the basis that it would fail; for the Exim run-time user, this is not
1807 the case, and it resulted in successful deliveries going unlogged.
1808 Fixed. Reported by Andreas Metzler.
1814 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1815 only prepend for the Resent-* case. Fixes regression introduced in
1816 Exim 4.70 by NM/22 for Bugzilla 607.
1818 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1819 increasing numbers of administrators be bitten by this.
1821 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1823 PP/03 Bugzilla 994: added openssl_options main configuration option.
1825 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1827 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1829 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1831 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1832 without a peer certificate, leading to a segfault because of an
1833 assumption that peers always have certificates. Be a little more
1834 paranoid. Problem reported by Martin Tscholak.
1836 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1837 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1838 NB: ClamAV planning to remove STREAM in "middle of 2010".
1839 CL also introduces -bmalware, various -d+acl logging additions and
1840 more caution in buffer sizes.
1842 PP/09 Implemented reverse_ip expansion operator.
1844 PP/10 Bugzilla 937: provide a "debug" ACL control.
1846 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1848 PP/12 Bugzilla 973: Implement --version.
1850 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1852 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1854 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1856 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1857 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1858 ignore trailing whitespace.
1860 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1862 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1865 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1866 Notification from Dr Andrew Aitchison.
1868 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1869 ExtendedDetectionInfo response format.
1870 Notification from John Horne.
1872 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1875 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1876 XSL and documented dependency on system catalogs, with examples of how
1879 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1882 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1883 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1884 configuration file which is writeable by the Exim user or group.
1886 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1887 of configuration files to cover files specified with the -C option if
1888 they are going to be used with root privileges, not just the default
1891 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1892 option (effectively making it always true).
1894 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1895 files to be used while preserving root privileges.
1897 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1898 that rogue child processes cannot use them.
1900 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1901 run-time user, instead of root.
1903 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1904 Exim run-time user without dropping privileges.
1906 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1907 result string, instead of calling string_vformat() twice with the same
1910 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1911 for other users. Others should always drop root privileges if they use
1912 -C on the command line, even for a whitelisted configure file.
1914 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1916 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1922 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1923 $data_path, and $header_path variables; fixed documentation bugs and
1926 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1927 exipick to access non-standard spools, including the "frozen" queue
1930 NM/01 Bugzilla 965: Support mysql stored procedures.
1931 Patch from Alain Williams
1933 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1935 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1936 Patch from Andreas Metzler
1938 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1939 Patch from Kirill Miazine
1941 NM/05 Bugzilla 671: Added umask to procmail example.
1943 JJ/03 installed exipick 20100323.0, fixing doc bug
1945 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1946 directory. Notification and patch from Dan Rosenberg.
1948 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1950 TK/02 Improve log output when DKIM signing operation fails.
1952 MH/01 Treat the transport option dkim_domain as a colon separated
1953 list, not as a single string, and sign the message with each element,
1954 omitting multiple occurences of the same signer.
1956 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1957 Bugzilla 985, 986. Patch by Simon Arlott
1959 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1960 Patch by Simon Arlott
1962 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1963 MBX locking. Notification from Dan Rosenberg.
1969 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1971 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1973 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1975 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1977 NM/04 Bugzilla 915: Flag broken perl installation during build.
1983 TK/01 Added patch by Johannes Berg that expands the main option
1984 "spamd_address" if it starts with a dollar sign.
1986 TK/02 Write list of recipients to X-Envelope-Sender header when building
1987 the mbox-format spool file for content scanning (suggested by Jakob
1990 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1991 (http://www.dcc-servers.net/) support via dccifd. Activated by
1992 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1994 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1995 by Mark Daniel Reidel <mr@df.eu>.
1997 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1998 When building exim an external PCRE library is now needed -
1999 PCRE is a system library on the majority of modern systems.
2000 See entry on PCRE_LIBS in EDITME file.
2002 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2003 conversation. Added nologin parameter to request.
2004 Patch contributed by Kirill Miazine.
2006 TF/01 Do not log submission mode rewrites if they do not change the address.
2008 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2010 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2011 log files in place. Contributed by Roberto Lima.
2013 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2015 TF/03 Bugzilla 615: When checking the local_parts router precondition
2016 after a local_part_suffix or local_part_prefix option, Exim now
2017 does not use the address's named list lookup cache, since this
2018 contains cached lookups for the whole local part.
2020 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2021 Robert Millan. Documentation is in experimental-spec.txt.
2023 TF/04 Bugzilla 668: Fix parallel build (make -j).
2025 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2027 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2028 Patch provided by Jan Srzednicki.
2030 TF/05 Leading white space used to be stripped from $spam_report which
2031 wrecked the formatting. Now it is preserved.
2033 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2034 that they are available at delivery time.
2036 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2038 TF/08 TLS error reporting now respects the incoming_interface and
2039 incoming_port log selectors.
2041 TF/09 Produce a more useful error message if an SMTP transport's hosts
2042 setting expands to an empty string.
2044 NM/06 Bugzilla 744: EXPN did not work under TLS.
2045 Patch provided by Phil Pennock.
2047 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2048 Patch provided by Richard Godbee.
2050 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2051 acl_smtp_notquit, added index entry.
2053 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2054 Patch provided by Eugene Bujak.
2056 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2057 accept(). Patch provided by Maxim Dounin.
2059 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2060 Patch provided by Phil Pennock.
2062 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2064 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2065 Patch provided by Brad "anomie" Jorsch.
2067 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2068 Patch provided by Dean Brooks.
2070 TK/05 Add native DKIM support (does not depend on external libraries).
2072 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2073 Patch provided by Graeme Fowler.
2075 NM/16 Bugzilla 851: Documentation example syntax fix.
2077 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2079 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2082 NM/19 Bugzilla 745: TLS version reporting.
2083 Patch provided by Phil Pennock.
2085 NM/20 Bugzilla 167: bool: condition support.
2086 Patch provided by Phil Pennock.
2088 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2089 clients. Patch provided by Phil Pennock.
2091 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2092 Patch provided by Brad "anomie" Jorsch.
2094 NM/23 Bugzilla 687: Fix misparses in eximstats.
2095 Patch provided by Heiko Schlittermann.
2097 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2098 Patch provided by Heiko Schlittermann.
2100 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2101 plus update to original patch.
2103 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2105 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2106 Patch provided by David Brownlee.
2108 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2110 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2112 NM/30 Bugzilla 888: TLS documentation bugfixes.
2114 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2116 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2117 Unlike the original bugzilla I have changed all shell scripts in src tree.
2119 NM/33 Bugzilla 898: Transport filter timeout fix.
2120 Patch by Todd Rinaldo.
2122 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2123 Patch by Serge Demonchaux.
2125 NM/35 Bugzilla 39: Base64 decode bug fixes.
2126 Patch by Jakob Hirsch.
2128 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2130 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2132 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2134 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2140 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2141 ALT-N's libdkim that I have put here:
2142 http://duncanthrax.net/exim-experimental/
2144 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2145 were called 'true' and it seems that C99 defines that as a reserved
2146 keyword to be used with 'bool' variable types. That means you could
2147 not include C99-style headers which use bools without triggering
2148 build errors in sieve.c.
2150 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2151 as mailq or other aliases. Changed the --help handling significantly
2152 to do whats expected. exim_usage() emits usage/help information.
2154 SC/01 Added the -bylocaldomain option to eximstats.
2156 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2158 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2160 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2166 PH/01 Another patch from the Sieve maintainer.
2168 PH/02 When an IPv6 address is converted to a string for single-key lookup
2169 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2170 dots are used instead of colons so that keys in lsearch files need not
2171 contain colons. This was done some time before quoting was made available
2172 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2173 (notated using the quote facility) so as to distinguish them from IPv4
2174 keys. This meant that lookups for IP addresses in host lists did not work
2175 for iplsearch lookups.
2177 This has been fixed by arranging for IPv6 addresses to be expressed with
2178 colons if the lookup type is iplsearch. This is not incompatible, because
2179 previously such lookups could never work.
2181 The situation is now rather anomalous, since one *can* have colons in
2182 ordinary lsearch keys. However, making the change in all cases is
2183 incompatible and would probably break a number of configurations.
2185 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2188 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2189 conversion specification without a maximum field width, thereby enabling
2190 a rogue spamd server to cause a buffer overflow. While nobody in their
2191 right mind would setup Exim to query an untrusted spamd server, an
2192 attacker that gains access to a server running spamd could potentially
2193 exploit this vulnerability to run arbitrary code as the Exim user.
2195 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2196 $primary_hostname instead of what libspf2 thinks the hosts name is.
2198 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2199 a directory entry by the name of the lookup key. Previously, if a
2200 symlink pointed to a non-existing file or a file in a directory that
2201 Exim lacked permissions to read, a lookup for a key matching that
2202 symlink would fail. Now it is enough that a matching directory entry
2203 exists, symlink or not. (Bugzilla 503.)
2205 PH/03 The body_linecount and body_zerocount variables are now exported in the
2208 PH/04 Added the $dnslist_matched variable.
2210 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2211 This means they are set thereafter only if the connection becomes
2214 PH/06 Added the client_condition to authenticators so that some can be skipped
2215 by clients under certain conditions.
2217 PH/07 The error message for a badly-placed control=no_multiline_responses left
2218 "_responses" off the end of the name.
2220 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2222 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2223 (without spaces) instead of just copying the configuration text.
2225 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2227 PH/11 Added $max_received_linelength.
2229 PH/12 Added +ignore_defer and +include_defer to host lists.
2231 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2232 way in which PCRE > 7.0 is built.
2234 PH/14 Implemented queue_only_load_latch.
2236 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2237 MAIL command. The effect was to mangle the value on 64-bit systems.
2239 PH/16 Another patch from the Sieve maintainer.
2241 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2243 PH/18 If a system quota error occurred while trying to create the file for
2244 a maildir delivery, the message "Mailbox is full" was not appended to the
2245 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2246 only to a quota excession during the actual writing of the file.
2248 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2249 characters?) which causes problems in log lines. The DN values are now
2250 passed through string_printing() before being added to log lines.
2252 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2253 and InterBase are left for another time.)
2255 PH/21 Added message_body_newlines option.
2257 PH/22 Guard against possible overflow in moan_check_errorcopy().
2259 PH/23 POSIX allows open() to be a macro; guard against that.
2261 PH/24 If the recipient of an error message contained an @ in the local part
2262 (suitably quoted, of course), incorrect values were put in $domain and
2263 $local_part during the evaluation of errors_copy.
2269 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2270 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2273 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2274 issue a MAIL command.
2276 PH/02 In an ACL statement such as
2278 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2280 if a client was not listed at all, or was listed with a value other than
2281 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2282 the condition was not true (as it should be), so access was not denied.
2283 The bug was that the ! inversion was incorrectly passed on to the second
2284 item. This has been fixed.
2286 PH/03 Added additional dnslists conditions == and =& which are different from
2287 = and & when the dns lookup returns more than one IP address.
2289 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2290 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2292 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2293 FSYNC, which compiles an option called disable_fsync that allows for
2294 bypassing fsync(). The documentation is heavily laced with warnings.
2296 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2298 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2299 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2300 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2301 including adding "make clean"; (3) Added -fPIC when compiling the test
2302 dynamically loaded module, to get rid of a warning.
2304 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2305 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2306 = 0s. The bug is otherwise harmless.
2308 PH/07 There was a bug in the dovecot authenticator such that the value of
2309 $auth1 could be overwritten, and so not correctly preserved, after a
2310 successful authentication. This usually meant that the value preserved by
2311 the server_setid option was incorrect.
2313 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2315 PH/09 Installed PCRE release 7.0.
2317 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2318 run for batched SMTP input. It is now run at the start of every message
2319 in the batch. While fixing this I discovered that the process information
2320 (output by running exiwhat) was not always getting set for -bs and -bS
2321 input. This is fixed, and it now also says "batched" for BSMTP.
2323 PH/11 Added control=no_pipelining.
2325 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2326 patch, slightly modified), and move the expansion of helo_data till after
2327 the connection is made in the smtp transport (so it can use these
2330 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2332 PH/14 Added log_selector = +pid.
2334 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2336 PH/16 Add ${if forany and ${if forall.
2338 PH/17 Added dsn_from option to vary the From: line in DSNs.
2340 PH/18 Flush SMTP output before performing a callout, unless control =
2341 no_callout_flush is set.
2343 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2344 was true (the default) a successful delivery failed to delete the retry
2345 item, thus causing premature timeout of the address. The bug is now
2348 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2350 PH/21 Long custom messages for fakedefer and fakereject are now split up
2351 into multiline responses in the same way that messages for "deny" and
2352 other ACL rejections are.
2354 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2355 with slight modification.
2357 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2358 draft, changing the syntax and factoring some duplicate code".
2360 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2361 for deliveries of the second and subsequent messages over the same SMTP
2364 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2365 ${reduce, with only minor "tidies".
2367 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2369 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2370 expansion side effects.
2372 PH/27 When a message times out after an over-quota error from an Exim-imposed
2373 quota, the bounce message says "mailbox is full". This message was not
2374 being given when it was a system quota that was exceeded. It now should
2377 MH/03 Made $recipients available in local_scan(). local_scan() already has
2378 better access to the recipient list through recipients_list[], but
2379 $recipients can be useful in postmaster-provided expansion strings.
2381 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2382 in the case of a MAIL command with additional options following the
2383 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2384 were accidentally chopped off.
2386 PH/29 SMTP synchronization checks are implemented when a command is read -
2387 there is a check that no more input is waiting when there shouldn't be
2388 any. However, for some commands, a delay in an ACL can mean that it is
2389 some time before the response is written. In this time, more input might
2390 arrive, invalidly. So now there are extra checks after an ACL has run for
2391 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2392 pipelining has not been advertised.
2394 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2396 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2397 returned in the SMTP response when smtp_return_error_details was false.
2398 This has been fixed.
2400 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2401 instead of the C I/O that was originally supplied, because problems were
2402 reported on Solaris.
2404 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2405 Exim which did not show up earlier: it was assuming that a call to
2406 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2407 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2408 was a macro that became an assignment, so it seemed to work. This has
2409 changed to a proper function call with a void return, hence the compile
2410 error. Exim's code has been fixed.
2412 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2415 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2417 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2419 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2420 because it turns out to be too big or there is a timeout) there is no
2421 "Completed" line in the log. When some messages of this type were
2422 selected by exigrep, they were listed as "not completed". Others were
2423 picked up by some special patterns. I have improved the selection
2424 criteria to be more general.
2426 PH/38 The host_find_failed option in the manualroute router can now be set
2427 to "ignore", to completely ignore a host whose IP address cannot be
2428 found. If all hosts are ignored, the behaviour is controlled by the new
2429 host_all_ignored option.
2431 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2432 homing or because of multiple MX records with /mx) generated more than
2433 one IP address, and the following item turned out to be the local host,
2434 all the secondary addresses of the first item were incorrectly removed
2435 from the list, along with the local host and any following hosts (which
2436 is what is supposed to happen).
2438 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2439 whoever called Exim into the -H file. In the case of the daemon it was
2440 behaving confusingly. When first started, it used values for whoever
2441 started the daemon, but after a SIGHUP it used the Exim user (because it
2442 calls itself on a restart). I have changed the code so that it now always
2445 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2446 message are rejected with the same error (e.g. no authentication or bad
2447 sender address), and a DATA command is nevertheless sent (as can happen
2448 with PIPELINING or a stupid MUA), the error message that was given to the
2449 RCPT commands is included in the rejection of the DATA command. This is
2450 intended to be helpful for MUAs that show only the final error to their
2453 PH/42 Another patch from the Sieve maintainer.
2455 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2456 Eximstats - Fixed some broken HTML links and added missing column headers
2458 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2459 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2461 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2463 PH/43 Yet another patch from the Sieve maintainer.
2465 PH/44 I found a way to check for a TCP/IP connection going away before sending
2466 the response to the final '.' that terminates a message, but only in the
2467 case where the client has not sent further data following the '.'
2468 (unfortunately, this is allowed). However, in many cases there won't be
2469 any further data because there won't be any more messages to send. A call
2470 to select() can be used: if it shows that the input is "ready", there is
2471 either input waiting, or the socket has been closed. An attempt to read
2472 the next input character can distinguish the two cases. Previously, Exim
2473 would have sent an OK response which the client would never have see.
2474 This could lead to message repetition. This fix should cure that, at
2475 least in a lot of common cases.
2477 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2478 advertised in response to EHLO.
2484 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2485 fixed by 4.65/MH/01 (is this a record?) are fixed:
2487 (i) An empty string was always treated as zero by the numeric comparison
2488 operators. This behaviour has been restored.
2490 (ii) It is documented that the numeric comparison operators always treat
2491 their arguments as decimal numbers. This was broken in that numbers
2492 starting with 0 were being interpreted as octal.
2494 While fixing these problems I realized that there was another issue that
2495 hadn't been noticed. Values of message_size_limit (both the global option
2496 and the transport option) were treated as octal if they started with 0.
2497 The documentation was vague. These values are now always treated as
2498 decimal, and I will make that clear in the documentation.
2504 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2505 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2508 MH/01 Don't check that the operands of numeric comparison operators are
2509 integers when their expansion is in "skipping" mode (fixes bug
2510 introduced by 4.64-PH/07).
2512 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2513 child addresses, Exim now panics and dies. Previously, because the count
2514 is held in a short int, deliveries were likely to be lost. As such a
2515 large number of recipients for a single message is ridiculous
2516 (performance will be very, very poor), I have chosen to impose a limit
2517 rather than extend the field.
2523 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2524 leftover -K file (the existence of which was triggered by #402).
2525 While we were at it, introduced process PID as part of the -K
2526 filename. This should rule out race conditions when creating
2529 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2530 processing considerably. Previous code took too long for large mails,
2531 triggering a timeout which in turn triggers #401.
2533 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2534 in the DK code in transports.c. sendfile() is not really portable,
2535 hence the _LINUX specificness.
2537 TF/01 In the add_headers option to the mail command in an Exim filter,
2538 there was a bug that Exim would claim a syntax error in any
2539 header after the first one which had an odd number of characters
2542 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2543 callout verification, Exim cached a "reject" for the entire domain. This
2544 is correct for most verifications, but it is not correct for a recipient
2545 verification with use_sender or use_postmaster set, because in that case
2546 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2547 case of MAIL FROM:<> rejection from other early rejections (e.g.
2548 rejection of HELO). When verifying a recipient using a non-null MAIL
2549 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2550 Whatever the result of the callout, the value of the domain cache is
2551 left unchanged (for any other kind of callout, getting as far as trying
2552 RCPT means that the domain itself is ok).
2554 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2557 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2558 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2559 session, the daemon ignores ECONNECT errors and logs others; it now
2560 ignores EPIPE as well.
2562 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2563 (quoted-printable decoding).
2565 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2566 later the small subsequent patch to fix an introduced bug.
2568 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2570 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2572 PH/08 An error is now given if message_size_limit is specified negative.
2574 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2575 to be given (somewhat) arbitrary names.
2577 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2580 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2581 miscellaneous code fixes
2583 PH/10 Added the log_reject_target ACL modifier to specify where to log
2586 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2587 hostname. This is wrong, because it relates to the incoming message (and
2588 probably the interface on which it is arriving) and not to the outgoing
2589 callout (which could be using a different interface). This has been
2590 changed to use the value of the helo_data option from the smtp transport
2591 instead - this is what is used when a message is actually being sent. If
2592 there is no remote transport (possible with a router that sets up host
2593 addresses), $smtp_active_hostname is used.
2595 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2596 tweaks were necessary in order to get it to work (see also 21 below):
2597 (a) The code assumed that strncpy() returns a negative number on buffer
2598 overflow, which isn't the case. Replaced with Exim's string_format()
2600 (b) There were several signed/unsigned issues. I just did the minimum
2601 hacking in of casts. There is scope for a larger refactoring.
2602 (c) The code used strcasecmp() which is not a standard C function.
2603 Replaced with Exim's strcmpic() function.
2604 (d) The code set only $1; it now sets $auth1 as well.
2605 (e) A simple test gave the error "authentication client didn't specify
2606 service in request". It would seem that Dovecot has changed its
2607 interface. Fortunately there's a specification; I followed it and
2608 changed what the client sends and it appears to be working now.
2610 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2613 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2614 address A is aliased to B and C, where B exists and C does not. Without
2615 -v the output is "A verified" because verification stops after a
2616 successful redirection if more than one address is generated. However,
2617 with -v the child addresses are also verified. Exim was outputting "A
2618 failed to verify" and then showing the successful verification for C,
2619 with its parentage. It now outputs "B failed to verify", showing B's
2620 parentage before showing the successful verification of C.
2622 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2623 look up a TXT record in a specific list after matching in a combined
2626 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2627 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2628 they consult the DNS. I had assumed they would set it the way they
2629 wanted; and indeed my experiments on Linux seem to show that in some
2630 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2631 To be on the safe side, however, I have now made the interface to
2632 host_find_byname() similar to host_find_bydns(), with an argument
2633 containing the DNS resolver options. The host_find_byname() function now
2634 sets these options at its start, just as host_find_bydns() does. The smtp
2635 transport options dns_qualify_single and dns_search_parents are passed to
2636 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2637 of host_find_byname() use the default settings of RES_DEFNAMES
2638 (qualify_single) but not RES_DNSRCH (search_parents).
2640 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2641 spool_read_header() do less string testing, by means of a preliminary
2642 switch on the second character of optional "-foo" lines. (This is
2643 overdue, caused by the large number of possibilities that now exist.
2644 Originally there were few.) While I was there, I also converted the
2645 str(n)cmp tests so they don't re-test the leading "-" and the first
2646 character, in the hope this might squeeze out yet more improvement.
2648 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2649 flag allowing group syntax was set by the header_syntax check but not
2650 turned off, possible causing trouble later; (2) The flag was not being
2651 set at all for the header_verify test, causing "group"-style headers to
2652 be rejected. I have now set it in this case, and also caused header_
2653 verify to ignore an empty address taken from a group. While doing this, I
2654 came across some other cases where the code for allowing group syntax
2655 while scanning a header line wasn't quite right (mostly, not resetting
2656 the flag correctly in the right place). These bugs could have caused
2657 trouble for malformed header lines. I hope it is now all correct.
2659 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2660 with the "reply" argument non-NULL. The code, however (which originally
2661 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2662 but it didn't always do it. This confused somebody who was copying the
2663 code for some other use. I have removed all the tests.
2665 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2666 feature that was used to support insecure browsers during the U.S. crypto
2667 embargo. It requires special client support, and Exim is probably the
2668 only MTA that supported it -- and would never use it because real RSA is
2669 always available. This code has been removed, because it had the bad
2670 effect of slowing Exim down by computing (never used) parameters for the
2671 RSA_EXPORT functionality.
2673 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2674 authenticator to fail if there's a tab character in the incoming data
2675 (there should never be unless someone is messing about, as it's supposed
2676 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2677 if the connection is using TLS or if the remote IP is the same as the
2678 local IP, and the "valid-client-cert option" if a client certificate has
2681 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2682 authenticators. This can be used for authorization after authentication
2683 succeeds. (In the case of plaintext, it servers for both authentication
2686 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2687 if any retry times were supplied.
2689 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2690 connection, where there is no client IP address to check. In this
2691 situation, the verify now always succeeds.
2693 PH/25 Applied John Jetmore's -Mset patch.
2695 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2697 PH/27 In a string expansion for a processed (not raw) header when multiple
2698 headers of the same name were present, leading whitespace was being
2699 removed from all of them, but trailing whitespace was being removed only
2700 from the last one. Now trailing whitespace is removed from each header
2701 before concatenation. Completely empty headers in a concatenation (as
2702 before) are ignored.
2704 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2705 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2707 PH/29 [Removed. This was a change that I later backed out, and forgot to
2708 correct the ChangeLog entry (that I had efficiently created) before
2709 committing the later change.]
2711 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2712 address errors (4xx response to RCPT) over the same connection as other
2713 messages routed to the same hosts. Such deliveries are always "forced",
2714 so retry times are not inspected. This resulted in far too many retries
2715 for the affected addresses. The effect occurred only when there were more
2716 hosts than the hosts_max_try setting in the smtp transport when it had
2717 the 4xx errors. Those hosts that it had tried were not added to the list
2718 of hosts for which the message was waiting, so if all were tried, there
2719 was no problem. Two fixes have been applied:
2721 (i) If there are any address or message errors in an SMTP delivery, none
2722 of the hosts (tried or untried) are now added to the list of hosts
2723 for which the message is waiting, so the message should not be a
2724 candidate for sending over the same connection that was used for a
2725 successful delivery of some other message. This seems entirely
2726 reasonable: after all the message is NOT "waiting for some host".
2727 This is so "obvious" that I'm not sure why it wasn't done
2728 previously. Hope I haven't missed anything, but it can't do any
2729 harm, as the worst effect is to miss an optimization.
2731 (ii) If, despite (i), such a delivery is accidentally attempted, the
2732 routing retry time is respected, so at least it doesn't keep
2733 hammering the server.
2735 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2736 in ${readsocket because some servers need this prod.
2738 PH/32 Added some extra debug output when updating a wait-xxx database.
2740 PH/33 The hint "could be header name not terminated by colon", which has been
2741 given for certain expansion errors for a long time, was not being given
2742 for the ${if def:h_colon_omitted{... case.
2744 PH/34 The spec says: "With one important exception, whenever a domain list is
2745 being scanned, $domain contains the subject domain." There was at least
2746 one case where this was not true.
2748 PH/35 The error "getsockname() failed: connection reset by peer" was being
2749 written to the panic log as well as the main log, but it isn't really
2750 panic-worthy as it just means the connection died rather early on. I have
2751 removed the panic log writing for the ECONNRESET error when getsockname()
2754 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2755 runs only) independently of the message's sender address. This meant
2756 that, if the 4xx error was in fact related to the sender, a different
2757 message to the same recipient with a different sender could confuse
2758 things. In particular, this can happen when sending to a greylisting
2759 server, but other circumstances could also provoke similar problems.
2760 I have changed the default so that the retry time for these errors is now
2761 based a combination of the sender and recipient addresses. This change
2762 can be overridden by setting address_retry_include_sender=false in the
2765 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2766 remote server are returned as part of bounce messages. This was not
2767 happening for LMTP over a pipe (the lmtp transport), but now it is the
2768 same for both kinds of LMTP.
2770 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2771 in header lines that were in fact CNAMEs. This is no longer the case.
2773 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2774 and queue runs started by the daemon processed all messages. This has
2775 been fixed so that -R and -S can now usefully be given with -q<time>.
2777 PH/40 Import PCRE release 6.7 (fixes some bugs).
2779 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2781 PH/42 Give an error if -q is specified more than once.
2783 PH/43 Renamed the variables $interface_address and $interface_port as
2784 $received_ip_address and $received_port, to make it clear that these
2785 values apply to message reception, and not to the outgoing interface when
2786 a message is delivered. (The old names remain recognized, of course.)
2788 PH/44 There was no timeout on the connect() call when using a Unix domain
2789 socket in the ${readsocket expansion. There now is.
2791 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2792 be meaningful with "accept".
2794 SC/01 Eximstats V1.43
2795 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2797 SC/02 Eximstats V1.44
2798 Use a glob alias rather than an array ref in the generated
2799 parser. This improves both readability and performance.
2801 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2802 Collect SpamAssassin and rejection statistics.
2803 Don't display local sender or destination tables unless
2804 there is data to show.
2805 Added average volumes into the top table text output.
2807 SC/04 Eximstats V1.46
2808 Collect data on the number of addresses (recipients)
2809 as well as the number of messages.
2811 SC/05 Eximstats V1.47
2812 Added 'Message too big' to the list of mail rejection
2813 reasons (thanks to Marco Gaiarin).
2815 SC/06 Eximstats V1.48
2816 Mainlog lines which have GMT offsets and are too short to
2817 have a flag are now skipped.
2819 SC/07 Eximstats V1.49 (Alain Williams)
2820 Added the -emptyok flag.
2822 SC/08 Eximstats V1.50
2823 Fixes for obtaining the IP address from reject messages.
2825 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2826 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2827 whitespace changes from 4.64-PH/27
2829 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2832 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2833 are found, allow negative numbers in numeric criteria)
2835 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2837 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2838 to match changes made in 4.64-PH/43
2840 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2842 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2843 "domain" is actually an IP address, and force a failure. This locks out
2844 those revolvers/nameservers that support "A-for-A" lookups, in
2845 contravention of the specifications.
2847 PH/48 When a host name was looked up from an IP address, and the subsequent
2848 forward lookup of the name timed out, the host name was left in
2849 $sender_host_name, contrary to the specification.
2851 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2852 restricted to single-key lookups, Exim was not diagnosing an error if
2853 * or *@ was used with a query-style lookup.
2855 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2857 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2858 long ago, but noone interested enough thought of it. Let's just say that
2859 the "1.1" means that there are some new functions that weren't there at
2860 some point in the past.
2862 PH/51 Error processing for expansion failure of helo_data from an smtp
2863 transport during callout processing was broken.
2865 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2866 tested/used via the -bh/-bhc/-bs options.
2868 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2869 bug, fixed in subsequent PCRE releases).
2871 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2872 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2874 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2880 SC/01 Use a glob alias rather than an array ref in eximstats generated
2881 parser. This improves both readability and performance.
2883 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2884 Don't display local sender or destination tables in eximstats unless
2885 there is data to show.
2886 Added average volumes into the eximstats top table text output.
2888 SC/03 Collect data on the number of addresses (recipients) as well
2889 as the number of messages in eximstats.
2891 TF/01 Correct an error in the documentation for the redirect router. Exim
2892 does (usually) call initgroups() when daemonizing.
2894 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2895 with consistent privilege compared to when running as a daemon.
2897 TF/03 Note in the spec that $authenticated_id is not set for local
2898 submissions from trusted users.
2900 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2901 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2903 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2904 by adding some example configuration directives to the default
2905 configuration file. A little bit of work is required to uncomment the
2906 directives and define how usernames and passwords are checked, but
2907 there is now a framework to start from.
2909 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2910 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2911 without this. I don't know how relevant this is to other LDAP libraries.
2913 PH/02 Add the verb name to the "unknown ACL verb" error.
2915 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2917 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2919 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2920 directory not expanded when it should be if an expanded home directory
2921 was set for the address (which is overridden by the transport).
2923 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2926 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2927 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2928 because it is too late at that time, and has no effect.
2930 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2931 security issue with \' (bugzilla #107). I could not use the
2932 PQescapeStringConn() function, because it needs a PGconn value as one of
2935 PH/08 When testing addresses using -bt, indicate those final addresses that
2936 are duplicates that would not cause an additional delivery. At least one
2937 person was confused, thinking that -bt output corresponded to deliveries.
2938 (Suppressing duplicates isn't a good idea as you lose the information
2939 about possibly different redirections that led to the duplicates.)
2941 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2942 systems where poll() doesn't work, in particular OS X.
2944 PH/10 Added more information to debugging output for retry time not reached.
2946 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2947 operations in malware.c.
2949 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2952 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2953 syslog_duplication set false, log lines that would normally be written
2954 both the the main log and to the reject log were not written to syslog at
2957 PH/14 In the default configuration, change the use of "message" in ACL warn
2958 statements to "add_header".
2960 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2961 not followed by a command (e.g. "seen endif").
2963 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2964 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2967 PH/17 Added extra conditions to the default value of delay_warning_condition
2971 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2972 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2973 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2976 The Auto-Submitted: and various List- headers are standardised, whereas I
2977 don't think Precedence: ever was.
2979 PH/18 Refactored debugging code in route_finduser() to show more information,
2980 in particular, the error code if getpwnam() issues one.
2982 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2983 This is apparently needed in addition to the PH/07 change above to avoid
2984 any possible encoding problems.
2986 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2987 but not after initializing Perl.
2989 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2990 output them only if debugging. By default they are written stderr,
2991 apparently, which is not desirable.
2993 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2996 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2999 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3001 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3002 authenticated or an ident call has been made. Suppress the default
3003 values for $authenticated_id and $authenticated_sender (but permit -oMai
3004 and -oMas) when testing with -bh.
3006 PH/24 Re-jigged the order of the tests in the default configuration so that the
3007 tests for valid domains and recipients precede the DNS black list and CSA
3008 tests, on the grounds that those ones are more expensive.
3010 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3011 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3012 This bug exists in every version of Exim that I still have, right back to
3015 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3016 However, an attempt to turn on case-sensitivity in a regex key by
3017 including (?-i) didn't work because the subject string was already
3018 lowercased, and the effects were non-intuitive. It turns out that a
3019 one-line patch can be used to allow (?-i) to work as expected.
3025 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3026 other effects) broke the use of negated acl sub-conditions.
3028 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3031 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3032 "Deny" causes Exim to reject the incoming connection with a 554 error.
3033 Unfortunately, if there is a major crisis, such as a disk failure,
3034 tcp-wrappers gives "deny", whereas what one would like would be some
3035 kind of temporary error. A kludge has been added to help with this.
3036 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3037 554 error is used if errno is still zero or contains ENOENT (which occurs
3038 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3041 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3043 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3044 errors. Otherwise a message that provokes a temporary error (when other
3045 messages do not) can cause a whole host to time out.
3047 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3048 addresses were routed directly to files or pipes from a redirect router.
3049 File deliveries just didn't batch; pipe deliveries might have suffered
3052 PH/06 A failure to get a lock for a hints database would erroneously always say
3053 "Failed to get write lock", even when it was really a read lock.
3055 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3056 of 0600. This has been changed to use the value of the lockfile_mode
3057 option (which defaults to 0600).
3059 PH/08 Applied small patch from the Sieve maintainer.
3061 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3062 folder from quota calculations, a direct delivery into this folder messed
3063 up the contents of the maildirsize file. This was because the regex was
3064 used only to exclude .Trash (or whatever) when the size of the mailbox
3065 was calculated. There was no check that a delivery was happening into an
3066 excluded directory. This bug has been fixed by ignoring all quota
3067 processing for deliveries into excluded directories.
3069 PH/10 Added the maildirfolder_create_regex option to appendfile.
3075 PH/01 The code for finding all the local interface addresses on a FreeBSD
3076 system running IPv6 was broken. This may well have applied to all BSD
3077 systems, as well as to others that have similar system calls. The broken
3078 code found IPv4 interfaces correctly, but gave incorrect values for the
3079 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3080 that it would not match correctly against @[] and not recognize the IPv6
3083 PH/02 The ipliteral router was not recognizing addresses of the form user@
3084 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3086 PH/03 Added disable_ipv6.
3088 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3089 decoded form, because it is most often used to construct To: headers
3090 lines in autoreplies, and the decoded form may well be syntactically
3091 invalid. However, $reply_address has leading white space removed, and all
3092 newlines turned into spaces so that the autoreply transport does not
3095 PH/05 If group was specified without a user on a router, and no group or user
3096 was specified on a transport, the group from the router was ignored.
3098 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3099 for visible compile-time settings that can be used to change these
3100 numbers, for those that want even more. Backwards compatibility with old
3101 spool files has been maintained. However, going back to a previous Exim
3102 release will lost any variables that are in spool files.
3104 PH/07 Two small changes when running in the test harness: increase delay when
3105 passing a TCP/IP connection to a new process, in case the original
3106 process has to generate a bounce, and remove special handling of
3107 127.0.0.2 (sic), which is no longer necessary.
3109 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3110 be the same on different OS.
3112 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3115 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3116 whether --show-vars was specified or not
3118 JJ/02 exipick: Added support for new ACL variable spool format introduced
3121 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3122 syntactically invalid From: or Reply-to: line, and a filter used this to
3123 generate an autoreply, and therefore failed to obtain an address for the
3124 autoreply, Exim could try to deliver to a non-existent relative file
3125 name, causing unrelated and misleading errors. What now happens is that
3126 it logs this as a hard delivery error, but does not attempt to create a
3129 PH/11 The exinext utility has a -C option for testing purposes, but although
3130 the given file was scanned by exinext itself; it wasn't being passed on
3131 when Exim was called.
3133 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3134 an end-of-file indication when reading a command response.
3136 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3137 compiled. In many other places in Exim, IPv6 addresses are always
3138 recognized, so I have changed this. It also means that IPv4 domain
3139 literals of the form [IPV4:n.n.n.n] are now always recognized.
3141 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3142 used if the router is not running as root, for example, when verifying at
3143 ACL time, or when using -bh. The debugging output from this situation was
3144 non-existent - all you got was a failure to exec. I have made two
3147 (a) Failures to set uid/gid, the current directory, or a process leader
3148 in a subprocess such as that created by queryprogram now generate
3149 suitable debugging output when -d is set.
3151 (b) The queryprogram router detects when it is not running as root,
3152 outputs suitable debugging information if -d is set, and then runs
3153 the subprocess without attempting to change uid/gid.
3155 PH/15 Minor change to Makefile for building test_host (undocumented testing
3158 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3159 additional section of a DNS packet that returns MX or SRV records.
3160 Instead, it always explicitly searches for A/AAAA records. This avoids
3161 major problems that occur when a DNS server includes only records of one
3162 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3163 fixed another bug: if SRV records were looked up and the corresponding
3164 address records were *not* found in the additional section, the port
3165 values from the SRV records were lost.
3167 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3168 using the correct key (the original address) when searching the retry
3169 rules in order to find which one to use for generating the retry hint.
3171 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3172 adding the default one. Similarly, if it contains a Reply-To: header, the
3173 errors_reply_to option, if set, is not used.
3175 PH/19 When calculating a retry time, Exim used to measure the "time since
3176 failure" by looking at the "first failed" field in the retry record. Now
3177 it does not use this if it is later than than the arrival time of the
3178 message. Instead it uses the arrival time. This makes for better
3179 behaviour in cases where some deliveries succeed, thus re-setting the
3180 "first failed" field. An example is a quota failure for a huge message
3181 when small messages continue to be delivered. Without this change, the
3182 "time since failure" will always be short, possible causing more frequent
3183 delivery attempts for the huge message than are intended.
3184 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3186 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3187 $1, $2, $3) because the numerical variables can be reset during some
3188 expansion items (e.g. "match"), thereby losing the authentication data.
3190 PH/21 Make -bV show the size of off_t variables so that the test suite can
3191 decide whether to run tests for quotas > 2G.
3193 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3194 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3195 filecount value is greater than 2G or if a quota value is greater than 2G
3196 on a system where the size of off_t is not greater than 4, a panic error
3199 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3200 never match. The debug and -bh output now contains an explicit error
3201 message indicating a malformed IPv4 address or mask.
3203 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3204 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3205 PH/23 above applies.
3207 PH/25 Do not write to syslog when running in the test harness. The only
3208 occasion when this arises is a failure to open the main or panic logs
3209 (for which there is an explicit test).
3211 PH/26 Added the /no_tell option to "control=freeze".
3213 PH/27 If a host name lookup failed very early in a connection, for example, if
3214 the IP address matched host_lookup and the reverse lookup yielded a name
3215 that did not have a forward lookup, an error message of the form "no IP
3216 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3217 could be logged. Now it outputs the IP address instead of "NULL".
3219 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3220 allows the sender and the authenticated sender to be set when
3221 submitting a message from within Exim. Since child_open_exim() is
3222 documented for local_scan(), the new function should be too.
3224 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3225 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3226 results in an empty string is now treated as unset.
3228 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3230 PH/31 Added sender_verify_fail logging option.
3232 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3233 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3234 tidied the source and removed it altogether.
3236 PH/33 When a queue run was abandoned because the load average was too high, a
3237 log line was always written; now it is written only if the queue_run log
3238 selector is set. In addition, the log line for abandonment now contains
3239 information about the queue run such as the pid. This is always present
3240 in "start" and "stop" lines but was omitted from the "abandon" line.
3242 PH/34 Omit spaces between a header name and the colon in the error message that
3243 is given when verify = headers_syntax fails (if there are lots of them,
3244 the message gets confusing).
3246 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3247 names, as there are now some PTR records that contain slashes. This check
3248 is only to protect against broken name servers that fall over on strange
3249 characters, so the fact that it applies to all lookups doesn't matter.
3251 PH/36 Now that the new test suite is complete, we can remove some of the
3252 special code in Exim that was needed for the old test suite. For example,
3253 sorting DNS records because real resolvers return them in an arbitrary
3254 order. The new test suite's fake resolver always returns records in the
3257 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3258 bounces) except when queue_only is set, to avoid logging races between
3259 the different processes.
3261 PH/38 Panic-die if .include specifies a non-absolute path.
3263 PH/39 A tweak to the "H" retry rule from its user.
3265 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3266 a label. They prevented compilation on older perls.
3268 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3269 a warning to be raised on newish perls.
3271 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3272 on queue. Changes to match documented behaviour of showing count of
3273 messages matching specified criteria.
3275 PH/40 Changed the default ident timeout from 30s to 5s.
3277 PH/41 Added support for the use of login_cap features, on those BSD systems
3278 that have them, for controlling the resources used by pipe deliveries.
3280 PH/42 The content-scanning code uses fopen() to create files in which to put
3281 message data. Previously it was not paying any attention to the mode of
3282 the files. Exim runs with umask(0) because the rest of the code creates
3283 files with open(), and sets the required mode explicitly. Thus, these
3284 files were ending up world-writeable. This was not a big issue, because,
3285 being within the spool directory, they were not world-accessible. I have
3286 created a function called modefopen, which takes an additional mode
3287 argument. It sets umask(777), creates the file, chmods it to the required
3288 mode, then resets the umask. All the relevant calls to fopen() in the
3289 content scanning code have been changed to use this function.
3291 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3292 to 24 hours. This avoids potential overflow problems when processing G
3293 and H retry rules. I suspect nobody ever tinkers with this value.
3295 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3297 PH/45 When the plaintext authenticator is running as a client, the server's
3298 challenges are checked to ensure they are valid base64 strings. By
3299 default, the authentication attempt is cancelled if an invalid string is
3300 received. Setting client_ignore_invalid_base64 true ignores these errors.
3301 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3302 they are received. Thus, the responses can be made to depend on the
3303 challenges. If an invalid string is ignored, an empty string is placed in
3306 PH/46 Messages that are created by the autoreply transport now contains a
3307 References: header, in accordance with RFCs 2822 and 3834.
3309 PH/47 Added authenticated_sender_force to the smtp transport.
3311 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3313 PH/49 Installed latest patch from the Sieve maintainer.
3315 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3316 was also set, the appendfile transport was unnecessarily scanning a
3317 directory of message files (e.g. for maildir delivery) to find the count
3318 of files (along with the size), even though it did not need this
3319 information. It now does the scan only if it needs to find either the
3320 size of the count of files.
3322 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3324 PH/52 Two bugs concerned with error handling when the smtp transport is
3327 (i) Exim was not creating retry information for temporary errors given
3328 for individual recipients after the DATA command when the smtp transport
3329 was used in LMTP mode. This meant that they could be retried too
3330 frequently, and not timed out correctly.
3332 (ii) Exim was setting the flag that allows error details to be returned
3333 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3334 recipients that were returned after the DATA command.
3336 PH/53 This is related to PH/52, but is more general: for any failing address,
3337 when detailed error information was permitted to be returned to the
3338 sender, but the error was temporary, then after the final timeout, only
3339 "retry timeout exceeded" was returned. Now it returns the full error as
3340 well as "retry timeout exceeded".
3342 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3343 do this, and (what is worse) MTAs that accept it.
3345 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3346 will now be deprecated.
3348 PH/56 New os.c-cygwin from the Cygwin maintainer.
3350 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3351 formats (previously only available in exim formats via -bpr, -bpru,
3352 and -bpra. Now also available in native and exiqgrep formats)
3354 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3355 with very large, slow to parse queues
3357 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3359 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3361 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3362 responses to authentication challenges, though it was showing the
3363 challenges; (ii) I've removed the CR characters from the debug output for
3366 PH/58 Allow for the insertion of a newline as well as a space when a string
3367 is turned into more than one encoded-word during RFC 2047 encoding. The
3368 Sieve code now uses this.
3370 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3371 data_4xx, lost_connection, tls_required.
3373 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3374 message was being sent as an SMTP response.
3376 PH/61 Add -l and -k options to exicyclog.
3378 PH/62 When verifying, if an address was redirected to one new address, so that
3379 verification continued, and the new address failed or deferred after
3380 having set something in $address_data, the value of $address_data was not
3381 passed back to the ACL. This was different to the case when no
3382 redirection occurred. The value is now passed back in both cases.
3384 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3385 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3386 use login_cap.h, so on its own it isn't the distinguishing feature. The
3387 new name refers directly to the setclassresources() function.
3389 PH/65 Added configuration files for NetBSD3.
3391 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3393 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3394 is preferred over IPv4.
3396 PH/68 The bounce_return_message and bounce_return_body options were not being
3397 honoured for bounces generated during the reception of non-SMTP messages.
3398 In particular, this applied to messages rejected by the ACL. This bug has
3399 been fixed. However, if bounce_return_message is true and bounce_return_
3400 body is false, the headers that are returned for a non-SMTP message
3401 include only those that have been read before the error was detected.
3402 (In the case of an ACL rejection, they have all been read.)
3404 PH/69 The HTML version of the specification is now built in a directory called
3405 spec_html instead of spec.html, because the latter looks like a path with
3406 a MIME-type, and this confuses some software.
3408 PH/70 Catch two compiler warnings in sieve.c.
3410 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3411 function verify_get_ident() calls ip_connect() to connect a socket, but
3412 if the "connect()" function timed out, ip_connect() used to close the
3413 socket. However, verify_get_ident() also closes the socket later, and in
3414 between Exim writes to the log, which may get opened at this point. When
3415 the socket was closed in ip_connect(), the log could get the same file
3416 descriptor number as the socket. This naturally causes chaos. The fix is
3417 not to close the socket in ip_connect(); the socket should be closed by
3418 the function that creates it. There was only one place in the code where
3419 this was missing, in the iplookup router, which I don't think anybody now
3420 uses, but I've fixed it anyway.
3422 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3423 well as to direct DNS lookups. Otherwise the handling of names in host
3424 lists is inconsistent and therefore confusing.
3430 PH/01 Two changes to the default runtime configuration:
3432 (1) Move the checks for relay_from_hosts and authenticated clients from
3433 after to before the (commented out) DNS black list checks.
3435 (2) Add control=submission to the relay_from_hosts and authenticated
3436 clients checks, on the grounds that messages accepted by these
3437 statements are most likely to be submissions.
3439 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3441 (1) Generate an error if the third argument for the ${prvs expansion is
3444 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3447 (3) Reset the variables that are obtained from the first argument of
3448 ${prvscheck and used in the second argument before leaving the code,
3449 because their memory is reclaimed, so using them afterwards may do
3452 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3453 one (it's much easier than Tom thought :-).
3455 (5) Because of (4), we can now allow for the use of $prvscheck_result
3456 inside the third argument.
3458 PH/03 For some reason, the default setting of PATH when running a command from
3459 a pipe transport was just "/usr/bin". I have changed it to
3462 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3463 anything to be listed in the output from -bV.
3465 PH/05 When a filter generated an autoreply, the entire To: header line was
3466 quoted in the delivery log line, like this:
3468 => >A.N.Other <ano@some.domain> <original@ddress> ...
3470 This has been changed so that it extracts the operative address. There
3471 may be more than one such address. If so, they are comma-separated, like
3474 => >ano@some.domain,ona@other.domain <original@ddress> ...
3476 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3477 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3478 not being looked up in the rDNS to get a host name, Exim was showing the
3479 IP address twice in Received: lines, even though the IP addresses were
3480 identical. For example:
3482 Received: from [1.2.3.4] (helo=[1.2.3.4])
3484 However, if the real host name was known, it was omitting the HELO data
3485 if it matched the actual IP address. This has been tidied up so that it
3486 doesn't show the same IP address twice.
3488 PH/07 When both +timestamp and +memory debugging was on, the value given by
3489 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3490 called by the debug printing, thereby overwriting the timestamp buffer.
3491 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3493 PH/08 When the original message was included in an autoreply transport, it
3494 always said "this is a copy of the message, including all the headers",
3495 even if body_only or headers_only was set. It now gives an appropriate
3498 PH/09 Applied a patch from the Sieve maintainer which:
3500 o fixes some comments
3501 o adds the (disabled) notify extension core
3502 o adds some debug output for the result of if/elsif tests
3503 o points to the current vacation draft in the documentation
3504 and documents the missing references header update
3508 o fixes a bug in processing the envelope test (when testing
3509 multiple envelope elements, the last element determined the
3512 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3513 Electronic Mail") by including:
3515 Auto-submitted: auto-generated
3517 in the messages that it generates (bounce messages and others, such as
3518 warnings). In the case of bounce messages for non-SMTP messages, there was
3519 also a typo: it was using "Auto_submitted" (underscore instead of
3520 hyphen). Since every message generated by Exim is necessarily in response
3521 to another message, thes have all been changed to:
3523 Auto-Submitted: auto-replied
3525 in accordance with these statements in the RFC:
3527 The auto-replied keyword:
3529 - SHOULD be used on messages sent in direct response to another
3530 message by an automatic process,
3532 - MUST NOT be used on manually-generated messages,
3534 - MAY be used on Delivery Status Notifications (DSNs) and Message
3535 Disposition Notifications (MDNs),
3537 - MUST NOT be used on messages generated by automatic or periodic
3538 processes, except for messages which are automatic responses to
3541 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3542 to the default Received: header definition.
3544 PH/12 Added log selector acl_warn_skipped (default on).
3546 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3547 variables because (a) they are in the wrong storage pool and (b) even if
3548 they were copied, it wouldn't work properly because of the caching.
3550 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3551 checking when decoding. Apparently there are clients that generate
3552 overlong encoded strings. Why am I not surprised?
3554 PH/15 If the first argument of "${if match_address" was not empty, but did not
3555 contain an "@" character, Exim crashed. Now it writes a panic log message
3556 and treats the condition as false.
3558 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3560 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3561 extension "envelope-auth". The code is finished and in agreement with
3562 other implementations, but there is no documentation so far and in fact,
3563 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3564 not changing the active code.
3566 Print executed "if" and "elsif" statements when debugging is used. This
3567 helps a great deal to understand what a filter does.
3569 Document more things not specified clearly in RFC3028. I had all this
3570 sorted out, when out of a sudden new issues came to my mind. Oops."
3572 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3575 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3576 canonical form (as documented). However, after a host name lookup from
3577 the IP address, check_host() was doing a simple string comparison with
3578 addresses acquired from the DNS when checking that the found name did
3579 have the original IP as one of its addresses. Since any found IPv6
3580 addresses are likely to be in abbreviated form, the comparison could
3581 fail. Luckily, there already exists a function for doing the comparison
3582 by converting both addresses to binary, so now that is used instead of
3583 the text comparison.
3585 PH/20 There was another similar case to PH/19, when a complete host name was
3586 given in a host list; looking up its IP address could give an abbreviated
3587 form, whereas the current host's name might or might not be abbreviated.
3588 The same fix has been applied.
3594 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3595 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3598 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3599 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3601 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3603 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3604 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3605 Also, for "Auto-Submitted", treat anything other than "no" as
3606 non-personal, in accordance with RFC 3834. (Previously it treated
3607 anything starting "auto-" as non-personal.)
3609 TF/01 The control=submission/name=... option had a problem with syntax
3610 errors if the name included a slash character. The /name= option
3611 now slurps the rest of the string, so it can include any characters
3612 but it must come last in the list of options (after /sender_retain
3615 PH/05 Some modifications to the interface to the fake nameserver for the new
3623 TK/01 Added the "success_on_redirect" address verification option. See
3624 NewStuff for rationale and an example.
3626 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3628 PH/02 Patch to exigrep to allow it to work on syslog lines.
3630 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3631 fread() to skip over the body file's header line, because in Cygwin the
3632 header line is locked and is inaccessible.
3634 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3635 co-exist for some time) to make it clear that it is the Exim ID that is
3636 referenced, not the Message-ID: header line.
3638 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3639 string_format() function, because snprintf() does not exist on all
3642 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3643 ${stat: expansion item.
3645 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3646 protocol synchronization error", to keep the pedants happy.
3648 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3649 well as for IRIX systems, when gcc is being used. See the host.c source
3652 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3654 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3657 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3658 transport and to the smtp transport in LMTP mode.
3660 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3662 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3663 run a filter in a subprocess. This could lead to confusion in subsequent
3664 lookups in the parent process. There should also be a search_tidyup() at
3665 the end of the subprocess.
3667 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3668 only if the host matched helo_try_verify_hosts, which caused the
3669 verification to occur when the EHLO/HELO command was issued. The ACL just
3670 tested the remembered result. Now, if a previous verification attempt has
3671 not happened, "verify = helo" does it there and then.
3673 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3675 TK/03 Fix log output including CR from clamd.
3677 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3678 exist provoked a memory error which could cause a segfault.
3680 PH/15 Installed PCRE 6.2
3682 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3684 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3685 of the problem. Specifically, suggested +O2 rather than +O1 for the
3688 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3690 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3691 record, the port was not correctly logged when the outgoing_port log
3692 selector was set (it logged the transort's default port).
3694 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3695 fallback_hosts, and "hosts" in the smtp transport.
3697 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3698 host errors such as "Connection refused".
3700 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3701 authentication with radiusclient 0.4.9:
3703 - Error returned from rc_read_config was caught wrongly
3704 - Username/password not passed on to radius server due to wrong length.
3706 The presumption is that some radiusclient API changes for 4.51/PH/17
3707 were not taken care of correctly. The code is still untested by me (my
3708 Linux distribution still has 0.3.2 of radiusclient), but it was
3709 contributed by a Radius user.
3711 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3712 expanding the "port" option of the smtp transport.
3714 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3715 while reading a MIME header. Thanks to Tom Hughes for a patch.
3717 PH/24 Include config.h inside local_scan.h so that configuration settings are
3720 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3721 that in an ACL for RCPT (for example), you can examine exactly what was
3724 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3725 commands, but it was not correctly comparing the address with the actual
3726 client host address. Thus, it would show the EHLO address in Received:
3727 header lines when this was not necessary.
3729 PH/27 Added the % operator to ${eval:}.
3731 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3732 it should be ignoring failures (because with -C, for example, it has lost
3733 privilege). It wasn't ignoring creation failures other than "already
3736 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3739 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3740 because an input line was too long, either on its own, or by virtue of
3741 too many continuations, the temporary file was not being removed, and the
3742 return code was incorrect.
3744 PH/31 Missing "BOOL" in function definition in filtertest.c.
3746 PH/32 Applied Sieve patches from the maintainer.
3748 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3750 PH/33 Added "verify = not_blind".
3752 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3753 Local/Makefile (with some defaults set). These are used in built scripts
3754 such as exicyclog, but they have never been used in the exim_install
3755 script (though there are many overriding facilities there). I have
3756 arranged that the exim_install script now takes note of these two
3759 PH/35 Installed configuration files for Dragonfly.
3761 PH/36 When a locally submitted message by a trusted user did not contain a
3762 From: header, and the sender address was obtained from -f or from an SMTP
3763 MAIL command, and the trusted user did not use -F to supply a sender
3764 name, $originator_name was incorrectly used when constructing a From:
3765 header. Furthermore, $originator_name was used for submission mode
3766 messages from external hosts without From: headers in a similar way,
3767 which is clearly wrong.
3769 PH/37 Added control=suppress_local_fixups.
3771 PH/38 When log_selector = +received_sender was set, and the addition of the
3772 sender made the log line's construction buffer exactly full, or one byte
3773 less than full, an overflow happened when the terminating "\n" was
3776 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3777 when the result of a list match is failure because a DNS lookup failed.
3779 PH/40 RM_COMMAND is now used in the building process.
3781 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3782 the "build-* directories that it finds.
3784 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3785 address in a domain literal was a prefix of an interface address.
3787 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3788 when verifying a sender address, unless rewrite_headers is false.
3790 PH/44 Wrote a long comment about why errors_to addresses are verified as
3791 recipients, not senders.
3793 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3794 the ratelimit ACL was added.
3796 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3798 PH/46 Added extra information about PostgreSQL errors to the error string.
3800 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3801 avoiding the need to install special zones in a real server. This is
3802 backwards compatible; if it can't find the fake resolver, it drops back.
3803 Thus, both old and new test suites can be run.
3805 TF/02 Added util/ratelimit.pl
3807 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3808 clock is set back in time.
3810 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3811 Candler <B.Candler@pobox.com>.
3813 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3814 OK for addresses that are the result of redirections.
3816 PH/48 A number of further additions for the benefit of the new test suite,
3817 including a fake gethostbyname() that interfaces to the fake DNS resolver
3820 TF/06 The fix for widen_domains has also been applied to qualify_single and
3821 search_parents which are the other dnslookup options that can cause
3824 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3827 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3829 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3830 service" if it encountered an MX record with an empty target hostname.
3831 The message is now "an MX or SRV record indicated no SMTP service".
3833 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3834 if the DNS of the sending site is misconfigured. This is quite a
3835 common situation. This change restores the behaviour of treating a
3836 helo verification defer as a failure.
3838 PH/51 If self=fail was set on a router, the bounce message did not include the
3839 actual error message.
3845 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3847 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3848 command itself ended in error, the underlying message about the transport
3849 filter timeout was being overwritten with the pipe command error. Now the
3850 underlying error message should be appended to the second error message.
3852 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3854 PH/02 Reduce the amount of output that "make" produces by default. Full output
3855 can still be requested.
3857 PH/03 The warning log line about a condition test deferring for a "warn" verb
3858 was being output only once per connection, rather than after each
3859 occurrence (because it was using the same function as for successful
3860 "warn" verbs). This seems wrong, so I have changed it.
3862 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3863 it should not have, which might have caused a crash in the right
3864 circumstances, but probably never did.
3866 PH/04 Installed a modified version of Tony Finch's patch to make submission
3867 mode fix the return path as well as the Sender: header line, and to
3868 add a /name= option so that you can make the user's friendly name appear
3871 TF/03 Added the control = fakedefer ACL modifier.
3873 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3874 Mark Lowes for thorough testing.
3876 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3878 TK/03 Merged latest SRS patch from Miles Wilton.
3880 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3881 with the definition in sysexits.h (which is #included earlier).
3882 Fortunately, Exim does not actually use EX_OK. The code used to try to
3883 preserve the sysexits.h value, by assuming that macro definitions were
3884 scanned for macro replacements. I have been disabused of this notion,
3885 so now the code just undefines EX_OK before #including unistd.h.
3887 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3888 in the smtp transport. When a block could not be written in a single
3889 write() function, the timeout was being re-applied to each part-write.
3890 This seems wrong - if the receiver was accepting one byte at a time it
3891 would take for ever. The timeout is now adjusted when this happens. It
3892 doesn't have to be particularly precise.
3894 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3895 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3897 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3898 without a domain if the check to <postmaster@domain> fails.
3900 SC/01 Eximstats: added -xls and the ability to specify output files
3901 (patch written by Frank Heydlauf).
3903 SC/02 Eximstats: use FileHandles for outputting results.
3905 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3907 SC/04 Eximstats: fixed display of large numbers with -nvr option
3909 SC/05 Eximstats: fixed merging of reports with empty tables.
3911 SC/06 Eximstats: added the -include_original_destination flag
3913 SC/07 Eximstats: removed tabs and trailing whitespace.
3915 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3917 TK/06 MBOX spool code: Add real "From " MBOX separator line
3918 so the .eml file is really in mbox format (even though
3919 most programs do not really care). Patch from Alex Miller.
3921 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3922 The latter is generated from $received_to and is only set if the
3923 message has one envelope recipient. SA can use these headers,
3924 obviously out-of-the-box. Patch from Alex Miller.
3926 PH/08 The ${def test on a variable was returning false if the variable's
3927 value was "0", contrary to what the specification has always said!
3928 The result should be true unless the variable is empty.
3930 PH/09 The syntax error of a character other than { following "${if
3931 def:variable_name" (after optional whitespace) was not being diagnosed.
3932 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3933 accidental colon was present, for example, could give incorrect results.
3935 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3936 result is used (not including appendfile, where other changes are about
3939 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3940 This involved changing a lot of size variables from int to off_t. It
3941 should work with maildirs and everything.
3943 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3944 spamd dying while we are connected to it.
3946 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3949 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3950 function for BDB 4.3.
3952 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3954 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3955 the sources for PCRE have been re-arranged and more files are now
3958 PH/15 The code I had for printing potentially long long variables in PH/11
3959 above was not the best (it lost precision). The length of off_t variables
3960 is now inspected at build time, and an appropriate printing format (%ld
3961 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3962 to be "long long int" or "long int". This is needed for the internal
3963 formatting function string_vformat().
3965 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3966 the configuration file to be ":syslog", then the script "guesses" where
3967 the logs files are, rather than using the compiled in default. In our
3968 case the guess is not the same as the compiled default, so the script
3969 suddenly stopped working when I started to use syslog. The patch checks
3970 to see if log_file_path is "". If so, it attempts to read it from exim
3971 with no configuration file to get the compiled in version, before it
3972 falls back to the previous guessing code."
3974 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3975 implementing BATV in an Exim configuration. See NewStuff for the gory
3978 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3979 Makefile that are specific to HP-UX.
3981 PH/18 If the "use_postmaster" option was set for a recipient callout together
3982 with the "random" option, the postmaster address was used as the MAIL
3983 FROM address for the random test, but not for the subsequent recipient
3984 test. It is now used for both.
3986 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3987 patch removes a few documentation additions to RFC 3028, because the
3988 latest draft now contains them. It adds the new en;ascii-case comparator
3989 and a new error check for 8bit text in MIME parts. Comparator and
3990 require names are now matched exactly. I enabled the subaddress
3991 extension, but it is not well tested yet (read: it works for me)."
3993 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3994 rework some of the code of TK/09 above to avoid the hardwired use of
3995 "%lld" and "long long". Replaced the call to snprintf() with a call to
3998 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3999 records point to non-existent hosts", "retry timeout exceeded", and
4000 "retry time not reached for any host after a long failure period".
4002 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4003 experimental DomainKeys support:
4005 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4006 (2) On an error such as an illegally used "control", the wrong name for
4007 the control was given.
4009 These problems did NOT occur unless DomainKeys support was compiled.
4011 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4013 PH/24 Added ${if match_ip condition.
4015 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4016 databases so that it will be absolutely obvious if a crash occurs in the
4017 DB library. This is a regular occurrence (often caused by mis-matched
4020 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4021 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4022 picked up on a user's system that detects such things. There doesn't seem
4023 to be a gcc warning option for this - only an attribute that has to be
4024 put on the function's prototype. It seems that in Fedora Core 4 they have
4025 set this on a number of new functions. No doubt there will be more in due
4028 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4029 specify a transport. However, if an address that was verified by such a
4030 router was the subject of a callout, Exim crashed because it tried to
4031 read the rcpt_include_affixes from the non-existent transport. Now it
4032 just assumes that the setting of that option is false. This bug was
4033 introduced by 4.51/PH/31.
4035 PH/28 Changed -d+all to exclude +memory, because that information is very
4036 rarely of interest, but it makes the output a lot bigger. People tend to
4037 do -d+all out of habit.
4039 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4040 code in os-type was giving problems when libc.so lives in lib64, like on
4043 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4044 aren't the modern standard, and it seems that some systems' include files
4045 don't always have them. Exim was already checking for some of the newer
4046 ones like T_AAAA, and defining it itself. I've added checks for all the
4047 record types that Exim uses.
4049 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4050 not automatically generating a new one, as it is supposed to. This
4051 prevented TLS from working. If the file did exist, but contained invalid
4052 data, a new version was generated, as expected. It was only the case of a
4053 non-existent file that was broken.
4055 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4056 with a change in libdomainkeys > 0.64.
4058 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4059 from DNS. If the selector record carries the flag, it now has
4060 precedence over the domain-wide flag.
4062 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4064 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4065 the use of a port name that isn't defined in /etc/services) occurred, the
4066 message was deferred as in a normal delivery, and thus remained on the
4067 spool, instead of being failed because of the mua_wrapper setting. This
4068 is now fixed, and I tidied up some of the mua_wrapper messages at the
4071 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4072 the messages in a hash of arrays rather than using individual hashes.
4073 This is a bit cleaner and results in dramatic memory savings, albeit
4074 at a slight CPU cost.
4076 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4077 as requested by Marc Sherman.
4079 SC/10 Eximstats: added histograms for user specified patterns as requested
4082 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4084 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4085 fopen() in the content-scanning modules that did not already have it.
4091 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4092 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4094 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4096 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4098 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4099 to test Sieve filters that use "vacation".
4101 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4102 that changes the way the GnuTLS parameters are stored in the cache file.
4103 The new format can be generated externally. For backward compatibility,
4104 if the data in the cache doesn't make sense, Exim assumes it has read an
4105 old-format file, and it generates new data and writes a new file. This
4106 means that you can't go back to an older release without removing the
4109 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4110 work if there are any delivery delays because "one_time" forces the
4111 parent to be marked "delivered", so its unseen clone is never tried
4112 again. For this reason, Exim now forbids the simultaneous setting of
4115 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4116 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4117 entry if you want to know the details. The fix, however, neglected to
4118 consider the case when local delivery batching is involved. The test for
4119 "previously delivered" was not happening when checking to see if an
4120 address could be batched with a previous (undelivered) one; under
4121 certain circumstances this could lead to multiple deliveries to the same
4124 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4125 in its include files, and this causes problems building Exim.
4127 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4128 header_syntax) but Exim was just ignoring anything given after a slash.
4129 In particular, this caused confusion with an attempt to use "verify =
4130 reverse_host_lookup/defer_ok". An error is now given when options are
4131 supplied for verify items that do not have them. (Maybe reverse_host_
4132 lookup should have a defer_ok option, but that's a different point.)
4134 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4135 defined by RFC 821) to 2048, because there were problems with some AUTH
4136 commands, and RFC 1869 says the size should be increased for extended
4137 SMTP commands that take arguments.
4139 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4142 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4143 "unknown" error; now it says that the functionality isn't in the binary.
4145 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4146 an address' error message when a string expansion fails (syntax or
4147 whatever). Otherwise the password may appear in the log. Following change
4148 PH/42 below, there is no longer a chance of it appearing in a bounce
4151 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4153 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4154 of its addresses was ever tried. (Bugzilla bug #2.)
4156 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4157 the result incorrectly in the debug output. (It correctly added a newline
4158 to what was transported.)
4160 TF/01 Added $received_time.
4162 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4163 commented out examples of how to interface to a virus scanner and to
4164 SpamAssassin. Also added commented examples of av_scanner and
4165 spamd_address settings.
4167 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4168 and controls are allowed in which ACLs. There were a couple of minor
4169 errors. Some of the entries in the conditions table (which is a table of
4170 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4171 negation of where the condition IS allowed.
4173 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4175 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4176 header file does not have a version number, so I've had to invent a new
4177 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4178 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4179 radiusclient), but it was contributed by a Radius user.
4181 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4182 files or directories, for OpenSSL.
4184 PH/19 When an Exim process that is running as root has to create an Exim log
4185 file, it does so in a subprocess that runs as exim:exim so as to get the
4186 ownership right at creation (otherwise, other Exim processes might see
4187 the file with the wrong ownership). There was no test for failure of this
4188 fork() call, which would lead to the process getting stuck as it waited
4189 for a non-existent subprocess. Forks do occasionally fail when resources
4190 run out. I reviewed all the other calls to fork(); they all seem to check
4193 PH/20 When checking for unexpected SMTP input at connect time (before writing
4194 the banner), Exim was not dealing correctly with a non-positive return
4195 from the read() function. If the client had disconnected by this time,
4196 the result was a log entry for a synchronization error with an empty
4197 string after "input=" when read() returned zero. If read() returned -1
4198 (an event I could not check), uninitialized data bytes were printed.
4199 There were reports of junk text (parts of files, etc) appearing after
4202 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4204 PH/22 Added support for macro redefinition, and (re)definition in between
4205 driver and ACL definitions.
4207 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4208 forgetting to use the resulting value; it was using the unexpanded value.
4210 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4211 hadn't been configured. The fix is from Juergen Kreileder, who
4212 understands it better than I do:
4214 "Here's what I see happening with three configured cyrus_sasl
4215 authenticators configured (plain, login, cram-md5):
4217 On startup auth_cyrus_sasl_init() gets called for each of these.
4218 This means three calls to sasl_listmech() without a specified mech_list.
4219 => SASL tests which mechs of all available mechs actually work
4220 => three warnings about OTP not working
4221 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4223 With the patch, sasl_listmech() also gets called three times. But now
4224 SASL's mech_list option is set to the server_mech specified in the the
4225 authenticator. Or in other words, the answer from sasl_listmech()
4226 gets limited to just the mech you're testing for (which is different
4228 => the return list contains just 'plain' or 'login', 'cram-md5' or
4229 nothing depending on the value of ob->server_mech.
4231 I've just tested the patch: Authentication still works fine,
4232 unavailable mechs specified in the exim configuration are still
4233 caught, and the auth.log warnings about OTP are gone."
4235 PH/25 When debugging is enabled, the contents of the command line are added
4236 to the debugging output, even when log_selector=+arguments is not
4239 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4240 answer is "GNU", and only if the return is "GNU/something" is the answer
4243 PH/27 $acl_verify_message is now set immediately after the failure of a
4244 verification in an ACL, and so is available in subsequent modifiers. In
4245 particular, the message can be preserved by coding like this:
4247 warn !verify = sender
4248 set acl_m0 = $acl_verify_message
4250 Previously, $acl_verify_message was set only while expanding "message"
4251 and "log_message" when a very denied access.
4253 PH/28 Modified OS/os.c-Linux with
4255 -#ifndef OS_LOAD_AVERAGE
4256 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4258 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4259 nomenclature these days.)
4261 PH/29 Installed patch from the Sieve maintainer that adds the options
4262 sieve_useraddress and sieve_subaddress to the redirect router.
4264 PH/30 In these circumstances:
4265 . Two addresses routed to the same list of hosts;
4266 . First host does not offer TLS;
4267 . First host accepts first address;
4268 . First host gives temporary error to second address;
4269 . Second host offers TLS and a TLS session is established;
4270 . Second host accepts second address.
4271 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4272 and peerdn, if requested) that were in fact used only for the second
4275 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4276 attention to any local part prefix or suffix that was matched by the
4277 router that accepted the address. It now behaves in the same way as it
4278 does for delivery: the affixes are removed from the local part unless
4279 rcpt_include_affixes is set on the transport.
4281 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4282 timeout during the DATA phase of an incoming message.
4284 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4285 applied a patch sanctioned by the Sieve maintainer.
4287 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4288 the uid or gid is negative. A case of a negative gid caused this to be
4289 noticed. The fix allows for either to be negative.
4291 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4292 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4295 PH/36 The change PH/12 above was broken. Fixed it.
4297 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4298 the grounds that routing the same address twice would always produce the
4299 same answer. This might have been true once, but it is certainly no
4300 longer true now. Routing a child address may depend on the previous
4301 routing that produced that child. Some complicated redirection strategies
4302 went wrong when messages had multiple recipients, and made Exim's
4303 behaviour dependent on the order in which the addresses were given.
4305 I have moved the duplicate checking until after the routing is complete.
4306 Exim scans the addresses that are assigned to local and remote
4307 transports, and removes any duplicates. This means that more work will be
4308 done, as duplicates will always all be routed, but duplicates are
4309 presumably rare, so I don't expect this is of any significance.
4311 For deliveries to pipes, files, and autoreplies, the duplicate checking
4312 still happens during the routing process, since they are not going to be
4315 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4316 It corrects a timeout issue with spamd. This is Ian's comment: "The
4317 background is that sometimes spamd either never reads data from a
4318 connection it has accepted, or it never writes response data. The exiscan
4319 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4320 blindly assumes that writes won't block so it may never time out."
4322 PH/39 Allow G after quota size as well as K and M.
4324 PH/40 The value set for $authenticated_id in an authenticator may not contain
4325 binary zeroes or newlines because the value is written to log lines and
4326 to spool files. There was no check on this. Now the value is run through
4327 the string_printing() function so that such characters are converted to
4328 printable escape sequences.
4330 PH/41 $message_linecount is a new variable that contains the total number of
4331 lines in the message. Compare $body_linecount, which is the count for the
4334 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4335 bounce and delay warning messages, except in certain special cases, which
4338 (a) An SMTP error message from a remote host;
4339 (b) A message specified in a :fail: redirection;
4340 (c) A message specified in a "fail" command in a system filter;
4341 (d) A message specified in a FAIL return from the queryprogram router;
4342 (e) A message specified by the cannot_route_message router option.
4344 In these cases only, Exim does include the error details in bounce and
4345 warning messages. There are also a few cases where bland messages such
4346 as "unrouteable address" or "local delivery error" are given.
4348 PH/43 $value is now also set for the "else" part of a ${run expansion.
4350 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4351 being worked on, but at least Exim now implements the latest version to
4354 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4355 process to complete was treated as a delivery failure, a timeout while
4356 writing the message to the pipe was logged, but erroneously treated as a
4357 successful delivery. Such timeouts include transport filter timeouts. For
4358 consistency with the overall process timeout, these timeouts are now
4359 treated as errors, giving rise to delivery failures by default. However,
4360 there is now a new Boolean option for the pipe transport called
4361 timeout_defer, which, if set TRUE, converts the failures into defers for
4362 both kinds of timeout. A transport filter timeout is now identified in
4365 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4366 systems where "make" and "gmake" are different, calling "gmake" at top
4367 level broke things. I've arranged for the value of $(MAKE) to be passed
4368 from the Makefile to this script so that it can call the same version of
4372 A note about Exim versions 4.44 and 4.50
4373 ----------------------------------------
4375 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4376 changes of various kinds. As a consequence, a big documentation update was
4377 needed. This delayed the release for rather longer than seemed good, especially
4378 in the light of a couple of (minor) security issues. Therefore, the changes
4379 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4380 release. So 4.44 and 4.50 are in effect two different branches that both start
4383 I have left the 4.50 change log unchanged; it contains all the changes since
4384 4.43. The change log for 4.44 is below; many of its items are identical to
4385 those for 4.50. This seems to be the most sensible way to preserve the
4386 historical information.
4392 1. Minor wording change to the doc/README.SIEVE file.
4394 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4395 computation of the current number of files was incorrect.
4397 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4398 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4401 4. Give more explanation in the error message when the command for a transport
4402 filter fails to execute.
4404 5. There are several places where Exim runs a non-Exim command in a
4405 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4406 was being done only for the command run by the queryprogram router. It is
4407 now done for all such subprocesses. The other cases are: ${run, transport
4408 filters, and the commands run by the lmtp and pipe transports.
4410 6. Added CONFIGURE_GROUP build-time option.
4412 7. Some older OS have a limit of 256 on the maximum number of file
4413 descriptors. Exim was using setrlimit() to set 1000 as a large value
4414 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4415 systems. I've change it so that if it can't get 1000, it tries for 256.
4417 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4418 was an oversight, and furthermore, ever since the addition of extra
4419 controls (e.g. 4.43/32), the checks on when to allow different forms of
4420 "control" were broken. There should now be diagnostics for all cases when a
4421 control that does not make sense is encountered.
4423 9. Added the /retain_sender option to "control=submission".
4425 10. $recipients is now available in the predata ACL (oversight).
4427 11. Tidy the search cache before the fork to do a delivery from a message
4428 received from the command line. Otherwise the child will trigger a lookup
4429 failure and thereby defer the delivery if it tries to use (for example) a
4430 cached ldap connection that the parent has called unbind on.
4432 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4433 of $address_data from the recipient verification was clobbered by the
4434 sender verification.
4436 13. The value of address_data from a sender verification is now available in
4437 $sender_address_data in subsequent conditions in the ACL statement.
4439 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4441 15. Added a new option "connect=<time>" to callout options, to set a different
4444 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4445 was its contents. (It was OK if the option was not defined at all.)
4447 17. A "Completed" log line is now written for messages that are removed from
4448 the spool by the -Mrm option.
4450 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4451 information about exactly what failed.
4453 19. Added -dd to debug only the daemon process.
4455 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4456 handles timeouts, both on the server side and network timeouts. Renamed the
4457 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4459 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4460 It is now set to "smtps".
4462 22. $host_address is now set to the target address during the checking of
4463 ignore_target_hosts.
4465 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4466 being passed; this would have caused $sender_host_name to have been used if
4467 matching the list had actually called for a host name (not very likely,
4468 since this list is usually IP addresses). A host name is now passed as
4471 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4472 code that specifies a non-restarting handler (typically sigaction() in
4473 modern systems) in an attempt to fix a rare and obscure crash bug.
4475 25. Narrowed the window for a race in the daemon that could cause it to ignore
4476 SIGCHLD signals. This is not a major problem, because they are used only to
4477 wake it up if nothing else does.
4479 26. A malformed maildirsize file could cause Exim to calculate negative values
4480 for the mailbox size or file count. Odd effects could occur as a result.
4481 The maildirsize information is now recalculated if the size or filecount
4484 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4485 support for a long time. Removed HAVE_SYS_VFS_H.
4487 28. Installed the latest version of exipick from John Jetmore.
4489 29. In an address list, if the pattern was not a regular expression, an empty
4490 subject address (from a bounce message) matched only if the pattern was an
4491 empty string. Non-empty patterns were not even tested. This was the wrong
4492 because it is perfectly reasonable to use an empty address as part of a
4493 database query. An empty address is now tested by patterns that are
4494 lookups. However, all the other forms of pattern expect the subject to
4495 contain a local part and a domain, and therefore, for them, an empty
4496 address still always fails if the pattern is not itself empty.
4498 30. Exim went into a mad DNS loop when attempting to do a callout where the
4499 host was specified on an smtp transport, and looking it up yielded more
4500 than one IP address.
4502 31. Re-factored the code for checking spool and log partition space into a
4503 function that finds that data and another that does the check. The former
4504 is then used to implement four new variables: $spool_space, $log_space,
4505 $spool_inodes, and $log_inodes.
4507 32. The RFC2047 encoding function was originally intended for short strings
4508 such as real names; it was not keeping to the 75-character limit for
4509 encoded words that the RFC imposes. It now respects the limit, and
4510 generates multiple encoded words if necessary. To be on the safe side, I
4511 have increased the buffer size for the ${rfc2047: expansion operator from
4514 33. It is now permitted to omit both strings after an "if" condition; if the
4515 condition is true, the result is "true". As before, when the second string
4516 is omitted, a false condition yields an empty string. This makes it less
4517 cumbersome to write custom ACL and router conditions.
4519 34. Failure to deliver a bounce message always caused it to be frozen, even if
4520 there was an errors_to setting on the router. The errors_to setting is now
4523 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4524 canonical form (fully expanded) before being placed in
4525 $sender_host_address.
4527 36. The table in the code that translates DNS record types into text (T_A to
4528 "A" for instance) was missing entries for NS and CNAME. It is just possible
4529 that this could have caused confusion if both these types were looked up
4530 for the same domain, because the text type is used as part of Exim's
4531 per-process caching. But the chance of anyone hitting this buglet seems
4534 37. The dnsdb lookup has been extended in a number of ways.
4536 (1) There is a new type, "zns", which walks up the domain tree until it
4537 finds some nameserver records. It should be used with care.
4539 (2) There is a new type, "mxh", which is like "mx" except that it returns
4540 just the host names, not the priorities.
4542 (3) It is now possible to give a list of domains (or IP addresses) to be
4543 looked up. The behaviour when one of the lookups defers can be
4544 controlled by a keyword.
4546 (4) It is now possible to specify the separator character for use when
4547 multiple records are returned.
4549 38. The dnslists ACL condition has been extended: it is now possible to supply
4550 a list of IP addresses and/or domains to be looked up in a particular DNS
4553 39. Added log_selector=+queue_time_overall.
4555 40. When running the queue in the test harness, wait just a tad after forking a
4556 delivery process, to get repeatability of debugging output.
4558 41. Include certificate and key file names in error message when GnuTLS fails
4559 to set them up, because the GnuTLS error message doesn't include the name
4560 of the failing file when there is a problem reading it.
4562 42. Allow both -bf and -bF in the same test run.
4564 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4566 44. The "Exiscan patch" is now merged into the mainline Exim source.
4568 45. Sometimes the final signoff response after QUIT could fail to get
4569 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4570 < 0 before doing a fflush(). This bug looks as though it goes back to the
4571 introduction of TLS in release 3.20, but "sometimes" must have been rare
4572 because the tests only now provoked it.
4574 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4575 (this can affect the format of dates).
4577 47. exim_tidydb, when checking for the continued existence of a message for
4578 which it has found a message-specific retry record, was not finding
4579 messages that were in split spool directories. Consequently, it was
4580 deleting retry records that should have stayed in existence.
4582 48. Steve fixed some bugs in eximstats.
4584 49. The SPA authentication driver was not abandoning authentication and moving
4585 on to the next authenticator when an expansion was forced to fail,
4586 contradicting the general specification for all authenticators. Instead it
4587 was generating a temporary error. It now behaves as specified.
4589 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4590 (the order specifies the preference for clients). The order is now AES256,
4591 AES128, 3DES, ARCFOUR128.
4593 51. Small patch to Sieve code - explicitly set From: when generating an
4596 52. Exim crashed if a remote delivery caused a very long error message to be
4597 recorded - for instance if somebody sent an entire SpamAssassin report back
4598 as a large number of 550 error lines. This bug was coincidentally fixed by
4599 increasing the size of one of Exim's internal buffers (big_buffer) that
4600 happened as part of the Exiscan merge. However, to be on the safe side, I
4601 have made the code more robust (and fixed the comments that describe what
4604 53. Now that there can be additional text after "Completed" in log lines (if
4605 the queue_time_overall log selector is set), a one-byte patch to exigrep
4606 was needed to allow it to recognize "Completed" as not the last thing in
4609 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4610 patch that reportedly fixes this has been added. I am not expert enough to
4611 create a test for it. This is what the patch creator wrote:
4613 "I found a little strange behaviour of ldap code when working with
4614 Windows 2003 AD Domain, where users was placed in more than one
4615 Organization Units. When I tried to give exim partial DN, the exit code
4616 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4617 But simultaneously result of request was absolutely normal ldap result,
4618 so I produce this patch..."
4620 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4621 REFERENCE, so I have modified the code to exclude the patch when that macro
4624 55. Some experimental protocols are using DNS PTR records for new purposes. The
4625 keys for these records are domain names, not reversed IP addresses. The
4626 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4627 leaves it alone. Component reversal etc. now happens only for IP addresses.
4630 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4632 57. Double the size of the debug message buffer (to 2048) so that more of very
4633 long debug lines gets shown.
4635 58. The exicyclog utility now does better if the number of log files to keep
4636 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4638 59. Two changes related to the smtp_active_hostname option:
4640 (1) $smtp_active_hostname is now available as a variable.
4641 (2) The default for smtp_banner uses $smtp_active_hostname instead
4642 of $primary_hostname.
4644 60. The host_aton() function is supposed to be passed a string that is known
4645 to be a valid IP address. However, in the case of IPv6 addresses, it was
4646 not checking this. This is a hostage to fortune. Exim now panics and dies
4647 if the condition is not met. A case was found where this could be provoked
4648 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4649 components; fortuitously, this particular loophole had already been fixed
4650 by change 4.50/55 above.
4652 If there are any other similar loopholes, the new check in host_aton()
4653 itself should stop them being exploited. The report I received stated that
4654 data on the command line could provoke the exploit when Exim was running as
4655 exim, but did not say which command line option was involved. All I could
4656 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4657 running as the user.
4660 61. There was a buffer overflow vulnerability in the SPA authentication code
4661 (which came originally from the Samba project). I have added a test to the
4662 spa_base64_to_bits() function which I hope fixes it.
4665 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4666 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4668 63. The daemon start-up calls getloadavg() while still root for those OS that
4669 need the first call to be done as root, but it missed one case: when
4670 deliver_queue_load_max is set with deliver_drop_privilege. This is
4671 necessary for the benefit of the queue runner, because there is no re-exec
4672 when deliver_drop_privilege is set.
4674 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4675 This has been fixed.
4677 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4678 was in use, was not putting the data itself into the right store pool;
4679 consequently, it could be overwritten for a subsequent message in the same
4680 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4683 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4685 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4686 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4687 the function were treating the return as a boolean value, which happened to
4688 work because 0=false and not-0=true, but is not correct code.
4690 68. The host_aton() function was not handling scoped IPv6 addresses (those
4691 with, for example, "%eth0" on the end) correctly.
4693 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4694 negated items (that is, ~something) in unsigned ints. Some compilers
4695 apparently mutter when there is no cast.
4697 70. If an address verification called from an ACL failed, and did not produce a
4698 user-specific message (i.e. there was only a "system" message), nothing was
4699 put in $acl_verify_message. In this situation, it now puts the system
4702 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4703 session; change 4.31/43 added the unwanted input to the log line - except
4704 that it did not do this in the start of session case. It now does.
4706 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4707 This is wrong and can cause the other end to generate a synchronization
4708 error if it is another Exim or anything else that does the synchronization
4709 check. A QUIT command is no longer sent after a timeout.
4711 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4712 during host lookups.
4714 74. The defer_ok option of callout verification was not working if it was used
4715 when verifying addresses in header lines, that is, for this case:
4717 verify = header_sender/callout=defer_ok
4719 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4720 those file descriptors could be used for SMTP connections. If anything
4721 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4722 could be sent to the SMTP client, causing chaos. The daemon now opens
4723 stdin, stdout, and stderr to /dev/null when it puts itself into the
4726 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4727 log by default. The user can override this with suitable Perl magic.
4729 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4730 the log message when discard triggers, was not working for the DATA ACL or
4731 for the non-SMTP ACL.
4733 78. Error message wording change in sieve.c.
4735 79. If smtp_accept_max_per_host was set, the number of connections could be
4736 restricted to fewer than expected, because the daemon was trying to set up
4737 a new connection before checking whether the processes handling previous
4738 connections had finished. The check for completed processes is now done
4739 earlier. On busy systems, this bug wouldn't be noticed because something
4740 else would have woken the daemon, and it would have reaped the completed
4743 80. If a message was submitted locally by a user whose login name contained one
4744 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4745 It caused a spool format error. I have fixed the spool reading code. A
4746 related problem was that the "from" clause in the Received: line became
4747 illegal because of the space(s). It is now covered by ${quote_local_part.
4749 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4752 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4754 83. Patch from Sieve maintainer to fix unterminated string problem in
4755 "vacation" handling.
4757 84. Some minor changes to the Linux configuration files to help with other
4758 OS variants using glibc.
4760 85. One more patch for Sieve to update vacation handling to latest spec.
4763 ----------------------------------------------------
4764 See the note above about the 4.44 and 4.50 releases.
4765 ----------------------------------------------------
4771 1. Change 4.43/35 introduced a bug that caused file counts to be
4772 incorrectly computed when quota_filecount was set in an appendfile
4775 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4776 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4779 3. Give more explanation in the error message when the command for a transport
4780 filter fails to execute.
4782 4. There are several places where Exim runs a non-Exim command in a
4783 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4784 was being done only for the command run by the queryprogram router. It is
4785 now done for all such subprocesses. The other cases are: ${run, transport
4786 filters, and the commands run by the lmtp and pipe transports.
4788 5. Some older OS have a limit of 256 on the maximum number of file
4789 descriptors. Exim was using setrlimit() to set 1000 as a large value
4790 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4791 systems. I've change it so that if it can't get 1000, it tries for 256.
4793 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4794 was an oversight, and furthermore, ever since the addition of extra
4795 controls (e.g. 4.43/32), the checks on when to allow different forms of
4796 "control" were broken. There should now be diagnostics for all cases when a
4797 control that does not make sense is encountered.
4799 7. $recipients is now available in the predata ACL (oversight).
4801 8. Tidy the search cache before the fork to do a delivery from a message
4802 received from the command line. Otherwise the child will trigger a lookup
4803 failure and thereby defer the delivery if it tries to use (for example) a
4804 cached ldap connection that the parent has called unbind on.
4806 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4807 of $address_data from the recipient verification was clobbered by the
4808 sender verification.
4810 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4811 was its contents. (It was OK if the option was not defined at all.)
4813 11. A "Completed" log line is now written for messages that are removed from
4814 the spool by the -Mrm option.
4816 12. $host_address is now set to the target address during the checking of
4817 ignore_target_hosts.
4819 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4820 being passed; this would have caused $sender_host_name to have been used if
4821 matching the list had actually called for a host name (not very likely,
4822 since this list is usually IP addresses). A host name is now passed as
4825 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4826 code that specifies a non-restarting handler (typically sigaction() in
4827 modern systems) in an attempt to fix a rare and obscure crash bug.
4829 15. Narrowed the window for a race in the daemon that could cause it to ignore
4830 SIGCHLD signals. This is not a major problem, because they are used only to
4831 wake it up if nothing else does.
4833 16. A malformed maildirsize file could cause Exim to calculate negative values
4834 for the mailbox size or file count. Odd effects could occur as a result.
4835 The maildirsize information is now recalculated if the size or filecount
4838 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4839 support for a long time. Removed HAVE_SYS_VFS_H.
4841 18. Updated exipick to current release from John Jetmore.
4843 19. Allow an empty sender to be matched against a lookup in an address list.
4844 Previously the only cases considered were a regular expression, or an
4847 20. Exim went into a mad DNS lookup loop when doing a callout where the
4848 host was specified on the transport, if the DNS lookup yielded more than
4851 21. The RFC2047 encoding function was originally intended for short strings
4852 such as real names; it was not keeping to the 75-character limit for
4853 encoded words that the RFC imposes. It now respects the limit, and
4854 generates multiple encoded words if necessary. To be on the safe side, I
4855 have increased the buffer size for the ${rfc2047: expansion operator from
4858 22. Failure to deliver a bounce message always caused it to be frozen, even if
4859 there was an errors_to setting on the router. The errors_to setting is now
4862 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4863 canonical form (fully expanded) before being placed in
4864 $sender_host_address.
4866 24. Updated eximstats to version 1.33
4868 25. Include certificate and key file names in error message when GnuTLS fails
4869 to set them up, because the GnuTLS error message doesn't include the name
4870 of the failing file when there is a problem reading it.
4872 26. Expand error message when OpenSSL has problems setting up cert/key files.
4875 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4876 (this can affect the format of dates).
4878 28. exim_tidydb, when checking for the continued existence of a message for
4879 which it has found a message-specific retry record, was not finding
4880 messages that were in split spool directories. Consequently, it was
4881 deleting retry records that should have stayed in existence.
4883 29. eximstats updated to version 1.35
4884 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4885 1.35 - bugfix such that pie charts by volume are generated correctly
4887 30. The SPA authentication driver was not abandoning authentication and moving
4888 on to the next authenticator when an expansion was forced to fail,
4889 contradicting the general specification for all authenticators. Instead it
4890 was generating a temporary error. It now behaves as specified.
4892 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4893 (the order specifies the preference for clients). The order is now AES256,
4894 AES128, 3DES, ARCFOUR128.
4896 31. Small patch to Sieve code - explicitly set From: when generating an
4899 32. Exim crashed if a remote delivery caused a very long error message to be
4900 recorded - for instance if somebody sent an entire SpamAssassin report back
4901 as a large number of 550 error lines. This bug was coincidentally fixed by
4902 increasing the size of one of Exim's internal buffers (big_buffer) that
4903 happened as part of the Exiscan merge. However, to be on the safe side, I
4904 have made the code more robust (and fixed the comments that describe what
4907 33. Some experimental protocols are using DNS PTR records for new purposes. The
4908 keys for these records are domain names, not reversed IP addresses. The
4909 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4910 leaves it alone. Component reversal etc. now happens only for IP addresses.
4913 34. The host_aton() function is supposed to be passed a string that is known
4914 to be a valid IP address. However, in the case of IPv6 addresses, it was
4915 not checking this. This is a hostage to fortune. Exim now panics and dies
4916 if the condition is not met. A case was found where this could be provoked
4917 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4918 components; fortuitously, this particular loophole had already been fixed
4919 by change 4.50/55 or 4.44/33 above.
4921 If there are any other similar loopholes, the new check in host_aton()
4922 itself should stop them being exploited. The report I received stated that
4923 data on the command line could provoke the exploit when Exim was running as
4924 exim, but did not say which command line option was involved. All I could
4925 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4926 running as the user.
4929 35. There was a buffer overflow vulnerability in the SPA authentication code
4930 (which came originally from the Samba project). I have added a test to the
4931 spa_base64_to_bits() function which I hope fixes it.
4934 36. The daemon start-up calls getloadavg() while still root for those OS that
4935 need the first call to be done as root, but it missed one case: when
4936 deliver_queue_load_max is set with deliver_drop_privilege. This is
4937 necessary for the benefit of the queue runner, because there is no re-exec
4938 when deliver_drop_privilege is set.
4940 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4941 was in use, was not putting the data itself into the right store pool;
4942 consequently, it could be overwritten for a subsequent message in the same
4943 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4946 38. Sometimes the final signoff response after QUIT could fail to get
4947 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4948 < 0 before doing a fflush(). This bug looks as though it goes back to the
4949 introduction of TLS in release 3.20, but "sometimes" must have been rare
4950 because the tests only now provoked it.
4956 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4957 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4958 Now it returns an integer. A number of calls were still expecting a T/F
4959 return. Fortuitously, in all cases, the tests worked in OK situations,
4960 which is the norm. However, things would have gone wrong on any write
4961 failures on the smtp file descriptor. This function is used when sending
4962 messages over SMTP and also when doing verify callouts.
4964 2. When Exim is called to do synchronous delivery of a locally submitted
4965 message (the -odf or -odi options), it no longer closes stderr before doing
4968 3. Implemented the mua_wrapper option.
4970 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4972 5. Implemented the functions header_remove(), header_testname(),
4973 header_add_at_position(), and receive_remove_recipient(), and exported them
4976 6. If an ACL "warn" statement specified the addition of headers, Exim already
4977 inserted X-ACL-Warn: at the start if there was no header name. However, it
4978 was not making this test for the second and subsequent header lines if
4979 there were newlines in the string. This meant that an invalid header could
4980 be inserted if Exim was badly configured.
4982 7. Allow an ACL "warn" statement to add header lines at the start or after all
4983 the Received: headers, as well as at the end.
4985 8. Added the rcpt_4xx retry error code.
4987 9. Added postmaster_mailfrom=xxx to callout verification option.
4989 10. Added mailfrom=xxxx to the callout verification option, for verify=
4992 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4993 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4995 12. Inserted some casts to stop certain compilers warning when using pointer
4996 differences as field lengths or precisions in printf-type calls (mostly
4997 affecting debugging statements).
4999 13. Added optional readline() support for -be (dynamically loaded).
5001 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5002 same clock tick as a message's arrival, so that its received time was the
5003 same as the "first fail" time on the retry record, and that message
5004 remained on the queue past the ultimate address timeout, every queue runner
5005 would try a delivery (because it was past the ultimate address timeout) but
5006 after another failure, the ultimate address timeout, which should have then
5007 bounced the address, did not kick in. This was a "< instead of <=" error;
5008 in most cases the first failure would have been in the next clock tick
5009 after the received time, and all would be well.
5011 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5012 being recognized when the domain list was tested by the match_domain
5013 condition in an expansion string.
5015 16. Added the ${str2b64: operator.
5017 17. Exim was always calling setrlimit() to set a large limit for the number of
5018 processes, without checking whether the existing limit was already
5019 adequate. (It did check for the limit on file descriptors.) Furthermore,
5020 errors from getrlimit() and setrlimit() were being ignored. Now they are
5021 logged to the main and panic logs, but Exim does carry on, to try to do its
5022 job under whatever limits there are.
5024 18. Imported PCRE 5.0.
5026 19. Trivial typo in log message " temporarily refused connection" (the leading
5029 20. If the log selector return_path_on_delivery was set and an address was
5030 redirected to /dev/null, the delivery process crashed because it assumed
5031 that a return path would always be set for a "successful" delivery. In this
5032 case, the whole delivery is bypassed as an optimization, and therefore no
5035 21. Internal re-arrangement: the function for sending a challenge and reading
5036 a response while authentication was assuming a zero-terminated challenge
5037 string. It's now changed to take a pointer and a length, to allow for
5038 binary data in such strings.
5040 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5042 23. Exim was not respecting finduser_retries when seeking the login of the
5043 uid under which it was called; it was always trying 10 times. (The default
5044 setting of finduser_retries is zero.) Also, it was sleeping after the final
5045 failure, which is pointless.
5047 24. Implemented tls_on_connect_ports.
5049 25. Implemented acl_smtp_predata.
5051 26. If the domain in control=submission is set empty, Exim assumes that the
5052 authenticated id is a complete email address when it generates From: or
5053 Sender: header lines.
5055 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5056 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5057 chown and chgrp in /bin and hostname in /usr/bin.
5059 28. Exim was keeping the "process log" file open after each use, just as it
5060 does for the main log. This opens the possibility of it remaining open for
5061 long periods when the USR1 signal hits a daemon. Occasional processlog
5062 errors were reported, that could have been caused by this. Anyway, it seems
5063 much more sensible not to leave this file open at all, so that is what now
5066 29. The long-running daemon process does not normally write to the log once it
5067 has entered its main loop, and it closes the log before doing so. This is
5068 so that log files can straightforwardly be renamed and moved. However,
5069 there are a couple of unusual error situations where the daemon does write
5070 log entries, and I had neglected to close the log afterwards.
5072 30. The text of an SMTP error response that was received during a remote
5073 delivery was being truncated at 512 bytes. This is too short for some of
5074 the long messages that one sometimes sees. I've increased the limit to
5077 31. It is now possible to make retry rules that apply only when a message has a
5078 specific sender, in particular, an empty sender.
5080 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5081 it possible to be selective about when SMTP synchronization is enforced.
5083 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5085 32. Implemented hosts_connection_nolog.
5087 33. Added an ACL for QUIT.
5089 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5092 35. Added mailbox_size and mailbox_filecount to appendfile.
5094 36. Added control = no_multiline_responses to ACLs.
5096 37. There was a bug in the logic of the code that waits for the clock to tick
5097 in the case where the clock went backwards by a substantial amount such
5098 that the microsecond fraction of "now" was more than the microsecond
5099 fraction of "then" (but the whole seconds number was less).
5101 38. Added support for the libradius Radius client library this is found on
5102 FreeBSD (previously only the radiusclient library was supported).
5108 1. When certain lookups returned multiple values in the form name=value, the
5109 quoting of the values was not always being done properly. Specifically:
5110 (a) If the value started with a double quote, but contained no whitespace,
5112 (b) If the value contained whitespace other than a space character (i.e.
5113 tabs or newlines or carriage returns) it was not quoted.
5114 This fix has been applied to the mysql and pgsql lookups by writing a
5115 separate quoting function and calling it from the lookup code. The fix
5116 should probably also be applied to nisplus, ibase and oracle lookups, but
5117 since I cannot test any of those, I have not disturbed their existing code.
5119 2. A hit in the callout cache for a specific address caused a log line with no
5120 reason for rejecting RCPT. Now it says "Previous (cached) callout
5121 verification failure".
5123 3. There was an off-by-one bug in the queryprogram router. An over-long
5124 return line was truncated at 256 instead of 255 characters, thereby
5125 overflowing its buffer with the terminating zero. As well as fixing this, I
5126 have increased the buffer size to 1024 (and made a note to document this).
5128 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5129 when Exim is waiting for an SMTP response from a remote server, Exim
5130 restarts its select() call on the socket, thereby resetting its timeout.
5131 This is not a problem when such interrupts are rare. Somebody set up a cron
5132 job to run exiwhat every 2 minutes, which is less than the normal select()
5133 timeout (5 or 10 minutes). This meant that the select() timeout never
5134 kicked in because it was always reset. I have fixed this by comparing the
5135 time when an interrupt arrives with the time at the start of the first call
5136 to select(). If more time than the timeout has elapsed, the interrupt is
5137 treated as a timeout.
5139 5. Some internal re-factoring in preparation for the addition of Sieve
5140 extensions (by MH). In particular, the "personal" test is moved to a
5141 separate function, and given an option for scanning Cc: and Bcc: (which is
5142 not set for Exim filters).
5144 6. When Exim created an email address using the login of the caller as the
5145 local part (e.g. when creating a From: or Sender: header line), it was not
5146 quoting the local part when it contained special characters such as @.
5148 7. Installed new OpenBSD configuration files.
5150 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5151 try to make them clearer.
5153 9. Callout options, other than the timeout value, were being ignored when
5154 verifying sender addresses in header lines. For example, when using
5156 verify = header_sender/callout=no_cache
5158 the cache was (incorrectly) being used.
5160 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5161 only the Cygwin environment.
5163 11. When return_path_on_delivery was set as a log selector, if different remote
5164 addresses in the same message used different return paths and parallel
5165 remote delivery occurred, the wrong values would sometimes be logged.
5166 (Whenever a remote delivery process finished, the return path value from
5167 the most recently started remote delivery process was logged.)
5169 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5170 lines when AUTH and/or TLS are in use. This is the "received protocol"
5171 field. Exim used to use "asmtp" for authenticated SMTP, without any
5172 indication (in the protocol name) for TLS use. Now it follows the RFC and
5173 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5174 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5175 names appear in log lines as well as in Received: header lines.
5177 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5178 extensions, and comparison tests, and to fix some bugs.
5180 14. Changes to the "personal" filter test:
5182 (1) The test was buggy in that it was just doing the equivalent of
5183 "contains" tests on header lines. For example, if a user's address was
5184 anne@some.where, the "personal" test would incorrectly be true for
5186 To: susanne@some.where
5188 This test is now done by extracting each address from the header in turn,
5189 and checking the entire address. Other tests that are part of "personal"
5190 are now done using regular expressions (for example, to check local parts
5191 of addresses in From: header lines).
5193 (2) The list of non-personal local parts in From: addresses has been
5194 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5195 taken from the Sieve specification recommendations.
5197 (3) If the message contains any header line starting with "List-" it is
5198 treated as non-personal.
5200 (4) The test for "circular" in the Subject: header line has been removed
5201 because it now seems ill-conceived.
5203 15. Minor typos in src/EDITME comments corrected.
5205 16. Installed latest exipick from John Jetmore.
5207 17. If headers_add on a router specified a text string that was too long for
5208 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5209 of string_sprintf() is now avoided.
5211 18. $message_body_size was not set (it was always zero) when running the DATA
5212 ACL and the local_scan() function.
5214 19. For the "mail" command in an Exim filter, no default was being set for
5215 the once_repeat time, causing a random time value to be used if "once" was
5216 specified. (If the value happened to be <= 0, no repeat happened.) The
5217 default is now 0s, meaning "never repeat". The "vacation" command was OK
5218 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5219 (I found it when inspecting the code).
5221 20. There is now an overall timeout for performing a callout verification. It
5222 defaults to 4 times the callout timeout, which applies to individual SMTP
5223 commands during the callout. The overall timeout applies when there is more
5224 than one host that can be tried. The timeout is checked before trying the
5225 next host. This prevents very long delays if there are a large number of
5226 hosts and all are timing out (e.g. when the network connections are timing
5227 out). The value of the overall timeout can be changed by specifying an
5228 additional sub-option for "callout", called "maxwait". For example:
5230 verify = sender/callout=5s,maxwait=20s
5232 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5233 to the end before writing, but this should make it even safer).
5235 22. Exim was forgetting that it had advertised PIPELINING for the second and
5236 subsequent messages on an SMTP connection. It was also not resetting its
5237 memory on STARTTLS and an internal HELO.
5239 23. When Exim logs an SMTP synchronization error within a session, it now
5240 records whether PIPELINING has been advertised or not.
5242 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5243 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5244 rather than long int.
5246 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5248 26. Added the never_mail option to autoreply.
5254 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5255 crash if the getsockname() call failed; this can happen if a connection is
5256 closed very soon after it is established. The problem was simply in the
5257 order in which certain operations were done, causing Exim to try to write
5258 to the SMTP stream before it had set up the file descriptor. The bug has
5259 been fixed by making things happen in the correct order.
5265 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5266 before the connection was closed, thus losing the rejection response.
5268 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5269 some early Solaris releases, but causes trouble in current releases where
5270 socklen_t is defined.
5272 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5275 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5278 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5279 that releasing the top of it at the end releases what was used for sub-
5280 expansions (unless the block got too big). However, discard this block if
5281 the first thing is a variable or header, so that we can use its block when
5282 it is dynamic (useful for very large $message_headers, for example).
5284 6. Lookups now cache *every* query, not just the most recent. A new, separate
5285 store pool is used for this. It can be recovered when all lookup caches are
5286 flushed. Lookups now release memory at the end of their result strings.
5287 This has involved some general refactoring of the lookup sources.
5289 7. Some code has been added to the store_xxx() functions to reduce the amount
5290 of flapping under certain conditions.
5292 8. log_incoming_interface used to affect only the <= reception log lines. Now
5293 it causes the local interface and port to be added to several more SMTP log
5294 lines, for example "SMTP connection from", and rejection lines.
5296 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5298 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5300 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5301 could be overwritten at the end of the current message (or the start of a
5302 new message if it was set in a HELO ACL). The value is now preserved for
5303 the duration of the SMTP connection.
5305 12. If a transport had a headers_rewrite setting, and a matching header line
5306 contained an unqualified address, that address was qualified, even if it
5307 did not match any rewriting rules. The underlying bug was that the values
5308 of the flags that permit the existence of unqualified sender and recipient
5309 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5310 non-local messages, and by -bnq for local messages) were not being
5311 preserved with the message after it was received.
5313 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5314 "next input=" as part of the text comprising the host identity instead of
5315 the correct text. The code was using the same buffer for two different
5316 strings. However, depending on which order the printing function evaluated
5317 its arguments, the bug did not always show up. Under Linux, for example, my
5318 test suite worked just fine.
5320 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5321 doesn't work with some older versions of Perl. It has been changed to "my",
5322 which in any case is probably the better facility to use.
5324 15. A really picky compiler found some instances of statements for creating
5325 error messages that either had too many or two few arguments for the format
5328 16. The size of the buffer for calls to the DNS resolver has been increased
5329 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5330 for addresses that have a lot of PTR records. This alleviates a problem; it
5331 does not fully solve it.
5333 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5334 buffer now truncates the list and logs the incident, which is the same
5335 action as happens when Exim is looking up a host name and its aliases.
5336 Previously in this situation something unpredictable would happen;
5337 sometimes it was "internal error: store_reset failed".
5339 18. If a server dropped the connection unexpectedly when an Exim client was
5340 using GnuTLS and trying to read a response, the client delivery process
5341 crashed while trying to generate an error log message.
5343 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5344 string, for example:
5346 warn message = H1: something\nH2: something
5348 the text was added as a single header line from Exim's point of view
5349 though it ended up OK in the delivered message. However, searching for the
5350 second and subsequent header lines using $h_h2: did not work. This has been
5351 fixed. Similarly, if a system filter added multiple headers in this way,
5352 the routers could not see them.
5354 20. Expanded the error message when iplsearch is called with an invalid key to
5355 suggest using net-iplsearch in a host list.
5357 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5358 ACLs are no longer actually imposed (and a message to that effect is
5361 22. If a "gecos" field in a passwd entry contained escaped characters, in
5362 particular, if it contained a \" sequence, Exim got it wrong when building
5363 a From: or a Sender: header from that name. A second bug also caused
5364 incorrect handling when an unquoted " was present following a character
5365 that needed quoting.
5367 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5368 was not being matched caselessly.
5370 24. Arranged for all hyphens in the exim.8 source to be escaped with
5373 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5374 properly. Recipient callout cache records were still being keyed to include
5375 the sender, even when use_sender was set false. This led to far more
5376 callouts that were necessary. The sender is no longer included in the key
5377 when use_sender is false.
5379 26. Added "control = submission" modifier to ACLs.
5381 27. Added the ${base62d: operator to decode base 62 numbers.
5383 28. dnsdb lookups can now access SRV records.
5385 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5386 the configuration file.
5388 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5389 (-v) mode. This makes the output for a verbose queue run more intelligible.
5391 31. Added a use_postmaster feature to recipient callouts.
5393 32. Added the $body_zerocount variable, containing the number of binary zero
5394 bytes in the message body.
5396 33. The time of last modification of the "new" subdirectory is now used as the
5397 "mailbox time last read" when there is a quota error for a maildir
5400 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5402 35. Added +ignore_unknown as a special item in host lists.
5404 36. Code for decoding IPv6 addresses in host lists is now included, even if
5405 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5406 address was recognized as an IP address, but was then not correctly decoded
5407 into binary, causing unexpected and incorrect effects when compared with
5414 1. Very minor rewording of debugging text in manualroute to say "list of
5415 hosts" instead of "hostlist".
5417 2. If verify=header_syntax was set, and a header line with an unqualified
5418 address (no domain) and a large number of spaces between the end of the
5419 name and the colon was received, the reception process suffered a buffer
5420 overflow, and (when I tested it) crashed. This was caused by some obsolete
5421 code that should have been removed. The fix is to remove it!
5423 3. When running in the test harness, delay a bit after writing a bounce
5424 message to get a bit more predictability in the log output.
5426 4. Added a call to search_tidyup() just before forking a reception process. In
5427 theory, someone could use a lookup in the expansion of smtp_accept_max_
5428 per_host which, without the tidyup, could leave open a database connection.
5430 5. Added the variables $recipient_data and $sender_data which get set from a
5431 lookup success in an ACL "recipients" or "senders" condition, or a router
5432 "senders" option, similar to $domain_data and $local_part_data.
5434 6. Moved the writing of debug_print from before to after the "senders" test
5437 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5438 problems for message scanning, either using a data ACL, or using
5439 local_scan() because the Received: header was not generated till after they
5440 were called (in order to set the time as the time of reception completion).
5441 I have revised the way this works. The header is now generated after the
5442 body is received, but before the ACL or local_scan() are called. After they
5443 are run, the timestamp in the header is updated.
5449 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5450 before starting a queue runner without re-exec. This happened only when
5451 deliver_drop_privilege was set or when the Exim user was set to root. The
5452 effect of the bug was that timeouts during subsequent deliveries caused
5453 crashes instead of being properly handled. The handler is now left at its
5454 default (and expected) setting.
5456 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5457 message, again when deliver_drop_privilege is set or Exim is run as root.
5458 The bug described in (1) was not present in this case, but the tidying up
5459 of the other signals was missing. I have made the two cases consistent.
5461 3. The ignore_target_hosts setting on a manualroute router was being ignored
5462 for hosts that were looked up using the /MX notation.
5464 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5467 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5468 operated on the sender address. After changing the $sender_address to <>
5469 for the sender address verify, Exim was re-instated it as the original
5470 (before rewriting) address, but remembering that it had rewritten it, so it
5471 wasn't rewriting it again. This bug also had the effect of breaking the
5472 sender address verification caching when the sender address was rewritten.
5474 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5475 This has been changed so that if the ip literal address matches
5476 ignore_target_hosts, the router declines.
5478 7. Added expansion conditions match_domain, match_address, and match_local_
5479 part (NOT match_host).
5481 8. The placeholder for the Received: header didn't have a length field set.
5483 9. Added code to Exim itself and to exim_lock to test for a specific race
5484 condition that could lead to file corruption when using MBX delivery. The
5485 issue is with the lockfile that is created in /tmp. If this file is removed
5486 after a process has opened it but before that process has acquired a lock,
5487 there is the potential for a second process to recreate the file and also
5488 acquire a lock. This could lead to two Exim processes writing to the file
5489 at the same time. The added code performs the same test as UW imapd; it
5490 checks after acquiring the lock that its file descriptor still refers to
5491 the same named file.
5493 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5494 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5497 11. Added the smtp_active_hostname option. If used, this will typically be made
5498 to depend on the incoming interface address. Because $interface_address is
5499 not set up until the daemon has forked a reception process, error responses
5500 that can happen earlier (such as "too many connections") no longer contain
5503 12. If an expansion in a condition on a "warn" statement fails because a lookup
5504 defers, the "warn" statement is abandoned, and the next ACL statement is
5505 processed. Previously this caused the whole ACL to be aborted.
5507 13. Added the iplsearch lookup type.
5509 14. Added ident_timeout as a log selector.
5511 15. Added tls_certificate_verified as a log selector.
5513 16. Added a global option tls_require_ciphers (compare the smtp transport
5514 option of the same name). This controls incoming TLS connections.
5516 17. I finally figured out how to make tls_require_ciphers do a similar thing
5517 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5518 before starting the TLS session.
5520 18. Tabs are now shown as \t in -bP output.
5522 19. If the log selector return_path_on_delivery was set, Exim crashed when
5523 bouncing a message because it had too many Received: header lines.
5525 20. If two routers both had headers_remove settings, and the first one included
5526 a superfluous trailing colon, the final name in the first list and the
5527 first name in the second list were incorrectly joined into one item (with a
5528 colon in the middle).
5534 1. Added -C and -D options to the exinext utility, mainly to make it easier
5535 to include in the automated testing, but these could be helpful when
5536 multiple configurations are in use.
5538 2. The exinext utility was not formatting the output nicely when there was
5539 an alternate port involved in the retry record key, nor when there was a
5540 message id as well (for retries that were specific to a specific message
5541 and a specific host). It was also confused by IPv6 addresses, because of
5542 the additional colons they contain. I have fixed the IPv4 problem, and
5543 patched it up to do a reasonable job for IPv6.
5545 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5546 delivery, the log line now contains "pipelined" if PIPELINING was used.
5548 4. An SMTP transport process used to panic and die if the bind() call to set
5549 an explicit outgoing interface failed. This has been changed; it is now
5550 treated in the same way as a connect() failure.
5552 5. A reference to $sender_host_name in the part of a conditional expansion
5553 that was being skipped was still causing a DNS lookup. This no longer
5556 6. The def: expansion condition was not recognizing references to header lines
5557 that used bh_ and bheader_.
5559 7. Added the _cache feature to named lists.
5561 8. The code for checking quota_filecount in the appendfile transport was
5562 allowing one more file than it should have been.
5564 9. For compatibility with Sendmail, the command line option
5572 and sets the incoming protocol and host name (for trusted callers). The
5573 host name and its colon can be omitted when only the protocol is to be set.
5574 Note the Exim already has two private options, -pd and -ps, that refer to
5575 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5576 "s", but I don't think that's a major issue.
5578 10. A number of refactoring changes to the code, none of which should affect
5581 (a) The number of logging options was getting close to filling up the
5582 32-bit word that was used as a bit map. I have split them into two classes:
5583 those that are passed in the argument to log_write(), and those that are
5584 only ever tested independently outside of that function. These are now in
5585 separate 32-bit words, so there is plenty of room for expansion again.
5586 There is no change in the user interface or the logging behaviour.
5588 (b) When building, for example, log lines, the code previously used a
5589 macro that called string_cat() twice, in order to add two strings. This is
5590 not really sufficiently general. Furthermore, there was one instance where
5591 it was actually wrong because one of the argument was used twice, and in
5592 one call a function was used. (As it happened, calling the function twice
5593 did not affect the overall behaviour.) The macro has been replaced by a
5594 function that can join an arbitrary number of extra strings onto a growing
5597 (c) The code for expansion conditions now uses a table and a binary chop
5598 instead of a serial search (which was left over from when there were very
5599 few conditions). Also, it now recognizes conditions like "pam" even when
5600 the relevant support is not compiled in: a suitably worded error message is
5601 given if an attempt is made to use such a condition.
5603 11. Added ${time_interval:xxxxx}.
5605 12. A bug was causing one of the ddress fields not to be passed back correctly
5606 from remote delivery subprocesses. The field in question was not being
5607 subsequently used, so this caused to problems in practice.
5609 13. Added new log selectors queue_time and deliver_time.
5611 14. Might have fixed a bug in maildirsizefile handling that threw up
5612 "unexpected character" debug warnings, and recalculated the data
5613 unnecessarily. In any case, I expanded the warning message to give more
5616 15. Added the message "Restricted characters in address" to the statements in
5617 the default ACL that block characters like @ and % in local parts.
5619 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5620 Three changes have been made:
5622 (a) There was a serious bug; a negative response to MAIL caused the whole
5623 recipient domain to be cached as invalid, thereby blocking all messages
5624 to all local parts at the same domain, from all senders. This bug has
5625 been fixed. The domain is no longer cached after a negative response to
5626 MAIL if the sender used is not empty.
5628 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5631 (c) A new callout option, "use_sender" has been added for people who want
5632 the modified behaviour.
5638 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5641 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5642 indeed breaks things for older releases.
5644 3. Added additional logging to the case where there is a problem reading data
5645 from a filter that is running in a subprocess using a pipe, in order to
5646 try to track down a specific problem.
5648 4. Testing facility fudge: when running in the test harness and attempting
5649 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5650 getting "No route to host". Convert this to a timeout.
5652 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5655 6. Some OS don't have socklen_t but use size_t instead. This affects the
5656 fifth argument of getsockopt() amongst other things. This is now
5657 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5658 can be set for individual OS. I have set it for SunOS5, OSF1, and
5659 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5660 some earlier ones do not.
5662 7. Change 4.30/15 was not doing the test caselessly.
5664 8. The standard form for an IPv6 address literal was being rejected by address
5665 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5666 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5667 this, as well as the form without the "IPv6" on the front (but only when
5668 address literals are enabled, of course).
5670 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5672 10. Exim crashed if a message with an empty sender address specified by -f
5673 encountered a router with an errors_to setting. This could be provoked only
5674 by a command such as
5678 where an empty string was supplied; "<>" did not hit this bug.
5680 11. Installed PCRE release 4.5.
5682 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5683 remained set. It is now erased.
5685 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5686 times from message ids (which are base 36 rather than the normal 62).
5688 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5689 were being counted as actual protocol errors, and logged if the log
5690 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5691 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5692 and DATA following a set of rejected RCPTs do not count as protocol errors.
5693 In other words, Exim assumes they were pipelined, though this may not
5694 actually be the case. Of course, in all cases the client gets an
5695 appropriate error code.
5697 15. If a lookup fails in an ACL condition, a message about the failure may
5698 be available; it is used if testing the ACL cannot continue, because most
5699 such messages specify what the cause of the deferral is. However, some
5700 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5701 that caused an old message to be retained and used if a later statement
5702 caused a defer, replacing the real cause of the deferral.
5704 16. If an IP address had so many PTR records that the DNS lookup buffer
5705 was not large enough to hold them, Exim could crash while trying to process
5706 the truncated data. It now detects and logs this case.
5708 17. Further to 4.21/58, another change has been made: if (and only if) the
5709 first line of a message (the first header line) ends with CRLF, a bare LF
5710 in a subsequent header line has a space inserted after it, so as not to
5711 terminate the header.
5713 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5714 unnecessarily, used atoi() instead of strtol(), and didn't check the
5715 termination when getting file sizes from file names by regex.
5717 19. Completely re-implemented the support for maildirsize files, in the light
5718 of a number of problems with the previous contributed implementation
5719 (4.30/29). In particular:
5721 . If the quota is zero, the maildirsize file is maintained, but no quota is
5724 . If the maildir directory does not exist, it is created before any attempt
5725 to write a maildirsize file.
5727 . The quota value in the file is just a cache; if the quota is changed in
5728 the transport, the new value overrides.
5730 . A regular expression is available for excluding directories from the
5733 20. The autoreply transport checks the characters in options that define the
5734 message's headers; it allows continued headers, but it was checking with
5735 isspace() after an embedded newline instead of explicitly looking for a
5738 21. If all the "regular" hosts to which an address was routed had passed their
5739 expiry times, and had not reached their retry times, the address was
5740 bounced, even if fallback hosts were defined. Now Exim should go on to try
5743 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5744 equivalent code in the SMTP transport. Some hosts send humungous responses
5745 to HELO/EHLO, more than 1024 it seems.
5747 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5748 gives compiler warnings in some environments. I've now done it "properly",
5751 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5752 (because of problems with the built-in one) was declared to return uschar *
5753 instead of char *, causing compiler failure.
5755 25. Fixed a file descriptor leak when processing alias/forward files.
5757 26. Fixed a minor format string issue in dbfn.c.
5759 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5761 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5762 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5765 29. When a sender address is verified, it is cached, to save repeating the test
5766 when there is more than one recipient in a message. However, when the
5767 verification involves a callout, it is possible for different callout
5768 options to be set for different recipients. It is too complicated to keep
5769 track of this in the cache, so now Exim always runs a verification when a
5770 callout is required, relying on the callout cache for the optimization.
5771 The overhead is duplication of the address routing, but this should not be
5774 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5775 to be verified with callout=postmaster, and the main callout worked but the
5776 postmaster check failed, the verification correctly failed. However, if a
5777 subsequent RCPT command asked for sender verification *without* the
5778 postmaster check, incorrect caching caused this verification also to fail,
5781 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5782 it was not caching the DNS options (qualify_single, search_parents) that
5783 were used when the lookup failed. A subsequent lookup with different
5784 options therefore always gave the same answer, though there were cases
5785 where it should not have. (Example: a "domains = !$mx_any" option on a
5786 dnslookup router: the "domains" option is always processed without any
5787 widening, but the router might have qualify_single set.) Now Exim uses the
5788 cached value only when the same options are set.
5790 32. Added John Jetmore's "exipick" utility to the distribution.
5792 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5793 than a timeout (e.g. a certificate is required, and is not provided), an
5794 Exim server now closes the connection immediately. Previously it waited for
5795 the client to close - but if the client is SSL, it seems that they each
5796 wait for each other, leading to a delay before one of them times out.
5798 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5799 maintained 0.8.x compatibility because I don't think many are using it, and
5800 it is clearly obsolete.
5802 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5805 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5806 client certificate was expired. A simple patch fixes this, though I don't
5807 understand the full logic of why the verify callback is called multiple
5810 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5811 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5812 which causes problems with some clients (such as the Certicom SSL Plus
5813 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5814 disables the coutermeasure allowing Eudora to connect."
5816 38. Exim was not checking that a write() to a log file succeeded. This could
5817 lead to Bad Things if a log got too big, in particular if it hit a file
5818 size limit. Exim now panics and dies if it cannot write to a log file, just
5819 as it does if it cannot open a log file.
5821 39. Modified OS/Makefile-Linux so that it now contains
5823 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5825 The two -D definitions ensure that Exim is compiled with large file
5826 support, which makes it possible to handle log files that are bigger than
5829 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5830 instance) a domain was checked against a named list that involved a lookup,
5831 causing $domain_data to be set, then another domain was checked against the
5832 same list, then the first domain was re-checked, the value of $domain_data
5833 after the final check could be wrong. In particular, if the second check
5834 failed, it could be set empty. This bug probably also applied to
5837 41. The strip_trailing_dot option was not being applied to the address given
5838 with the -f command-line option.
5840 42. The code for reading a message's header from the spool was incrementing
5841 $received_count, but never initializing it. This meant that the value was
5842 incorrect (doubled) while delivering a message in the same process in which
5843 it was received. In the most common configuration of Exim, this never
5844 happens - a fresh exec is done - but it can happen when
5845 deliver_drop_privilege is set.
5847 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5848 it now includes up to 150 characters of the unexpected data in the log
5851 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5852 and building data strings. The size of both of these buffers was 10 000
5853 bytes - far larger than anybody would *ever* want, thought I. Needless to
5854 say, somebody hit the limit. I have increased the maximum line length to
5855 20 000 and the maximum data length of concatenated lines to 100 000. I have
5856 also fixed two bugs, because there was no checking on these buffers. Tsk,
5857 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5858 buffer is too small.
5860 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5861 lsearch lookups. Now it does.
5863 46. When parsing a route_list item in a manualroute router, a fixed-length
5864 buffer was used for the list of hosts. I made this 1024 bytes long,
5865 thinking that nobody would ever have a list of hosts that long. Wrong.
5866 Somebody had a whole pile of complicated expansion conditions, and the
5867 string was silently truncated, leading to an expansion error. It turns out
5868 that it is easier to change to an unlimited length (owing to other changes
5869 that have happened since this code was originally written) than to build
5870 structure for giving a limitation error. The length of the item that
5871 expands into the list of hosts is now unlimited.
5873 47. The lsearch lookup could not handle data where the length of text line was
5874 more than 4095 characters. Such lines were truncated, leading to shortened
5875 data being returned. It should now handle lines of any length.
5877 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5878 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5881 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5882 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5883 no longer happens in comment lines. A list of replacements is now placed
5884 at the head of all of the source files, except those whose only change is
5885 to replace PERL_COMMAND in the very first #! line.
5887 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5888 messages on the queue, with a bottom-up merge sort, using code contributed
5889 by Michael Haardt. This should make operations like -bp somewhat faster on
5890 large queues. It won't affect queue runners, except when queue_run_in_order
5893 51. Installed eximstats 1.31 in the distribution.
5895 52. Added support for SRV lookups to the dnslookup router.
5897 53. If an ACL referred to $message_body or $message_body_end, the value was not
5898 reset for any messages that followed in the same SMTP session.
5900 54. The store-handling optimization for building very long strings was not
5901 differentiating between the different store pools. I don't think this
5902 actually made any difference in practice, but I've tidied it.
5904 55. While running the routers to verify a sender address, $sender_address
5905 was still set to the sender address. This is wrong, because when routing to
5906 send a bounce to the sender, it would be empty. Therefore, I have changed
5907 it so that, while verifying a sender address, $sender_address is set to <>.
5908 (There is no change to what happens when verifying a recipient address.)
5910 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5911 target A or AAAA records (if not already returned) without resetting the
5912 qualify_single or search_parents options of the DNS resolver. These are
5913 inappropriate in this case because the targets of MX and SRV records must
5914 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5915 target that, when qualified, matched something in the local domain. These
5916 two options are now turned off when doing these lookups.
5918 57. It seems that at least some releases of Reiserfs (which does not have the
5919 concept of a fixed number of inodes) returns zero and not -1 for the
5920 number of available inodes. This interacted badly with check_spool_inodes,
5921 which assumed that -1 was the "no such thing" setting. What I have done is
5922 to check that the total number of inodes is greater than zero before doing
5923 the test of how many are available.
5925 58. When a "warn" ACL statement has a log_message modifier, the message is
5926 remembered, and not repeated. This is to avoid a lot of repetition when a
5927 message has many recipients that cause the same warning to be written.
5928 However, Exim was preserving the list of already written lines for an
5929 entire SMTP session, which doesn't seem right. The memory is now reset if a
5930 new message is started.
5932 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5933 debugging output unless log_rewrite was also set.
5935 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5936 of (int)(handle) when we know that handle contains (void *)(-1).
5938 61. The Exim daemon panic-logs an error return when it closes the incoming
5939 connection. However "connection reset by peer" seems to be common, and
5940 isn't really an error worthy of noting specially, so that particular error
5943 62. When Exim is trying to find all the local interfaces, it used to panic and
5944 die if the ioctl to get the interface flags failed. However, it seems that
5945 on at least one OS (Solaris 9) it is possible to have an interface that is
5946 included in the list of interfaces, but for which you get a failure error
5947 for this call. This happens when the interface is not "plumbed" into a
5948 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5949 failure of the "get flags" call assumes that the interface is down.
5951 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5952 makes life easier for people who are doing arithmetic on fields extracted
5953 from dates, where you often get leading zeros that should not be
5954 interpreted as octal.
5956 64. Added qualify_domain to the redirect router, to override the global
5959 65. If a pathologically long header line contained very many addresses (the
5960 report of this problem mentioned 10 000) and each of them was rewritten,
5961 Exim could use up a very large amount of memory. (It kept on making new
5962 copies of the header line as it rewrote, and never released the old ones.)
5963 At the expense of a bit more processing, the header rewriting function has
5964 been changed so that it no longer eats memory in this way.
5966 66. The generation of the Received: header has been moved from the time that a
5967 message starts to be received, to the time that it finishes. The timestamp
5968 in the Received: header should now be very close to that of the <= log
5969 line. There are two side-effects of this change:
5971 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5972 logged header lines no longer include the local Received: line, because
5973 it has not yet been created. The same applies to a copy of the message
5974 that is returned to a non-SMTP sender when a message is rejected.
5976 (b) When a filter file is tested using -bf, no additional Received: header
5977 is added to the test message. After some thought, I decided that this
5980 This change does not affect the value of $received_for. It is still set
5981 after address rewriting, but before local_scan() is called.
5983 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5985 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5986 gave an unhelpful panic error message, and a defer error. I have managed to
5987 change this behaviour so that it now rejects any supplied certificate,
5988 which seems right, as the list of acceptable certificates is empty.
5990 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5991 gave an unhelpful defer error. I have not managed to make this reject any
5992 supplied certificates, but the error message it gives is "no certificate
5993 supplied", which is not helpful.
5995 70. exigrep's output now also includes lines that are not associated with any
5996 message, but which match the given pattern. Implemented by a patch from
5997 Martin Sluka, which also tidied up the Perl a bit.
5999 71. Recipient callout verification, like sender verification, was using <> in
6000 the MAIL FROM command. This isn't really the right thing, since the actual
6001 sender may affect whether the remote host accepts the recipient or not. I
6002 have changed it to use the actual sender in the callout; this means that
6003 the cache record is now keyed on a recipient/sender pair, not just the
6004 recipient address. There doesn't seem to be a real danger of callout loops,
6005 since a callout by the remote host to check the sender would use <>.
6006 [SEE ABOVE: changed after hitting problems.]
6008 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6009 temporary errors. However, in the case of such a code being given after
6010 the end of a data transmission (i.e. after ".") Exim was failing to write
6011 a retry record for the message. (Yes, there was some broken host that was
6012 actually sending 8xx at this point.)
6014 73. An unknown lookup type in a host list could cause Exim to panic-die when
6015 the list was checked. (An example that provoked this was putting <; in the
6016 middle of a list instead of at the start.) If this happened during a DATA
6017 ACL check, a -D file could be left lying around. This kind of configuration
6018 error no longer causes Exim to die; instead it causes a defer error. The
6019 incident is still logged to the main and panic logs.
6021 74. Buglet left over from Exim 3 conversion. The message "too many messages
6022 in one connection" was written to the rejectlog but not the mainlog, except
6023 when address rewriting (yes!) was being logged.
6025 75. Added write_rejectlog option.
6027 76. When a system filter was run not as root (that is, when system_filter_user
6028 was set), the values of the $n variables were not being returned to the
6029 main process; thus, they were not subsequently available in the $sn
6032 77. Added +return_path_on_delivery log selector.
6034 78. A connection timeout was being treated differently from recipients deferred
6035 when testing hosts_max_try with a message that was older than the host's
6036 retry timeout. (The host should not be counted, thus allowing all hosts to
6037 be tried at least once before bouncing.) This may have been the cause of an
6038 occasionally reported bug whereby a message would remain on the queue
6039 longer than the retry timeout, but would be bounced if a delivery was
6040 forced. I say "may" because I never totally pinned down the problem;
6041 setting up timeout/retry tests is difficult. See also the next item.
6043 79. The ultimate address timeout was not being applied to errors that involved
6044 a combination of host plus message (for example, a timeout on a MAIL
6045 command). When an address resolved to a number of possible hosts, and they
6046 were not all tried for each delivery (e.g. because of hosts_max_try), a
6047 message could remain on the queue longer than the retry timeout.
6049 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6052 81. Fixed an obscure SMTP outgoing bug which required at least the following
6053 conditions: (a) there was another message waiting for the same server;
6054 (b) the server returned 5xx to all RCPT commands in the first message so
6055 that the message was not completed; (c) the server dropped the connection
6056 or gave a negative response to the RSET that Exim sends to abort the
6057 transaction. The observed case was a dropped connection after DATA that had
6058 been sent in pipelining mode. That is, the server had advertised PIPELINING
6059 but was not implementing it correctly. The effect of the bug was incorrect
6060 behaviour, such as trying another host, and this could lead to a crash.
6066 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6067 and daemon.c were passed as pointers to ints; they should have been
6068 pointers to socklen_t variables (which are typically unsigned ints).
6070 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6073 3. Fixed a really odd bug that affected only the testing scheme; patching a
6074 certain fixed string in the binary changed the value of another string that
6075 happened to be identical to the end of the original first string.
6077 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6078 name", it returns that address as the IP address. On some operating
6079 systems (e.g. Solaris), it also passes back the IP address string as the
6080 "host name". However, on others (e.g. Linux), it passes back an empty
6081 string. Exim wasn't checking for this, and was changing the host name to an
6082 empty string, assuming it had been canonicalized.
6084 5. Although rare, it is permitted to have more than one PTR record for a given
6085 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6086 all the names associated with an address, because they do in Solaris.
6087 However, it seems that they do not in Linux for data that comes from the
6088 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6089 I found this out when I moved to a new Linux workstation and tried to run
6090 the Exim test suite.
6092 To get round this problem I have changed the code so that it now does its
6093 own call to the DNS to look up PTR records when searching for a host name.
6094 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6095 addresses that are only in /etc/hosts are still found.
6097 This behaviour is, however, controlled by an option called host_lookup_
6098 order, which defaults to "bydns:byaddr". If people want to use the other
6099 order, or indeed, just use one or the other means of lookup, they can
6100 specify it in this variable.
6102 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6103 some operating systems, this comes back from gethostbyaddr() as an empty
6104 string, and this is what Exim used to test for. However, it seems that in
6105 other systems, "." is yielded. Exim now tests for this case too.
6107 7. The values of check_spool_space and check_log_space are now held internally
6108 as a number of kilobytes instead of an absolute number of bytes. If a
6109 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6110 kilobyte. This means that much larger values can be stored.
6112 8. Exim monitor: an attempt to get the action menu when not actually pointing
6113 at a message produces an empty menu entitled "No message selected". This
6114 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6115 no entries in it ("Shell widget menu has zero width and/or height"). So I
6116 have added a single, blank menu entry in this case.
6118 9. Added ${quote_local_part.
6120 10. MIME decoding is now applied to the contents of Subject: header lines when
6123 11. Now that a reference to $sender_host_address automatically causes a reverse
6124 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6125 host lookup before query-style lookups in lists that might use this
6126 variable. This has therefore been abolished, and the "net-" prefix is no
6127 longer necessary for query-style lookups.
6129 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6130 have been a typo for LFLAGS, so it has been changed.
6132 13. The install script calls Exim with "-C /dev/null" in order to find the
6133 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6134 to be output. However, since Exim outputs its version number before the
6135 error, it didn't break the script. It just looked ugly. I fixed this by
6136 always allowing "-C /dev/null" if the caller is root.
6138 14. Ignore overlarge ACL variable number when reading spool file - insurance
6139 against a later release with more variables having written the file.
6141 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6142 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6143 this, as well as the form without the "IPv6" on the front.
6145 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6146 OS/Makefile-Darwin file.
6148 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6149 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6151 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6152 variable. However, some OS use ints for inodes. I've added cast to long int
6153 to get rid of the compiler warning.
6155 19. I had forgotten to lock out "/../" in configuration file names when
6156 ALT_CONFIG_PREFIX was set.
6158 20. Routers used for verification do not need to specify transports. However,
6159 if such a router generated a host list, and callout was configured, Exim
6160 crashed, because it could not find a port number from the (non-existent)
6161 transport. It now assumes port 25 in this circumstance.
6163 21. Added the -t option to exigrep.
6165 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6166 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6167 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6168 them is set, LOOKUP_LSEARCH is forced.
6170 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6172 24. Added sender and host information to the "rejected by local_scan()" log
6173 line; previously there was no indication of these.
6175 25. Added .include_if_exists.
6177 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6178 Linux. It turns out that not all file systems support this. Apparently some
6179 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6180 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6181 not supported on the file descriptor, is now ignored when Exim is trying to
6182 sync a directory. This applies only to Linux.
6184 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6186 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6187 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6190 29. Added support for maildirsize files from supplied patch (modified a bit).
6192 30. The use of :fail: followed by an empty string could lead Exim to respond to
6193 sender verification failures with (e.g.):
6195 550 Verification failed for <xxx>
6196 550 Sender verify failed
6198 where the first response line was missing the '-' that indicates it is not
6199 the final line of the response.
6201 31. The loop for finding the name of the user that called Exim had a hardwired
6202 limit of 10; it now uses the value of finduser_retries, which is used for
6203 all other user lookups.
6205 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6208 33. Exim was neglecting to zero errno before one call of strtol() when
6209 expanding a string and expecting an integer value. On some systems this
6210 resulted in spurious "integer overflow" errors. Also, it was casting the
6211 result into an int without checking.
6213 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6214 did not work. The code looks as if it has *never* worked, though it appears
6215 to have been documented since at least release 1.62. I have made it work.
6217 35. The "timeout_DNS" error in retry rules, also documented since at least
6218 1.62, also never worked. As it isn't clear exactly what this means, and
6219 clearly it isn't a major issue, I have abolished the feature by treating it
6220 as "timeout", and writing a warning to the main and panic logs.
6222 36. The display of retry rules for -brt wasn't always showing the error code
6225 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6226 timeout_connect_A, timeout_connect_MX.
6228 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6229 to the empty sender.
6231 39. The daemon was not analysing the content of -oX till after it had closed
6232 stderr and disconnected from the controlling terminal. This meant that any
6233 syntax errors were only noted on the panic log, and the return code from
6234 the command was 0. By re-arranging the code a little, I've made the
6235 decoding happen first, so such errors now appear on stderr, and the return
6236 code is 1. However, the actual setting up of the sockets still happens in
6237 the disconnected process, so errors there are still only recorded on the
6240 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6241 connections (as happens on some IP stacks) was logged at start up time as
6242 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6243 it from "IPv6 and IPv4", which means that two separate sockets are being
6246 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6247 says whether AF_INET or AF_INET6 was passed as an argument.
6249 42. Exiwhat output was messed up when time zones were included in log
6252 43. Exiwhat now gives more information about the daemon's listening ports,
6253 and whether -tls-on-connect was used.
6255 44. The "port" option of the smtp transport is now expanded.
6257 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6258 silently ignored. Now an error message is written to the main and panic
6261 46. There's a new ACL modifier called "logwrite" which writes to a log file
6262 as soon as it is encountered.
6264 47. Added $local_user_uid and $local_user_gid at routing time.
6266 48. Exim crashed when trying to verify a sender address that was being
6269 49. Exim was recognizing only a space character after ".include". It now also
6270 recognizes a tab character.
6272 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6273 extracting the relevant information from the specification. The man page no
6274 longer contains scrambled data for the -d option, and I've added a section
6275 at the front about calling Exim under different names.
6277 51. Added "extra_headers" argument to the "mail" command in filter files.
6279 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6282 53. Installed eximstats 1.29.
6284 54. Added transport_filter_timeout as a generic transport option.
6286 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6287 Cc: header lines. This was required by RFC 822, but it not required by RFC
6290 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6291 incoming messages that did not have them. Now it does so only if the
6292 message originates locally, that is, if there is no associated remote host
6293 address. When Resent- header lines are present, this applies to the Resent-
6294 lines rather than the non-Resent- lines.
6296 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6297 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6299 58. Messages for configuration errors now include the name of the main
6300 configuration file - useful now that there may be more than one file in a
6301 list (.included file names were always shown).
6303 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6304 for those rare installations that do not start the daemon as root or run it
6305 setuid root. I've cut out the call to initgroups() if the daemon is not
6308 60. The Exim user and group can now be bound into the binary as text strings
6309 that are looked up at the start of Exim's processing.
6311 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6313 62. Added $mailstore_basename variable.
6315 63. Installed patch to sieve.c from Michael Haardt.
6317 64. When Exim failed to open the panic log after failing to open the main log,
6318 the original message it was trying to log was written to stderr and debug
6319 output, but if they were not available (the usual case in production), it
6320 was lost. Now it is written to syslog before the two lines that record the
6321 failures to open the logs.
6323 65. Users' Exim filters run in subprocesses under the user's uid. It is
6324 possible for a "deliver" command or an alias in a "personal" command to
6325 provoke an address rewrite. If logging of address rewriting is configured,
6326 this fails because the process is not running as root or exim. There may be
6327 a better way of dealing with this, but for the moment (because 4.30 needs
6328 to be released), I have disabled address rewrite logging when running a
6329 filter in a non-root, non-exim process.
6335 1. The buildconfig auxiliary program wasn't quoting the value set for
6336 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6337 not defined. This bug was masked in 4.22 by the effect that was fixed in
6340 2. Some messages that were rejected after a message id was allocated were
6341 shown as "incomplete" by exigrep. It no longer does this for messages that
6342 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6344 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6345 have allow_domain_literals set, the ID did not get logged in the <= line.
6346 Domain literals are now always recognized in Message-ID: header lines.
6348 4. The first argument for a ${extract expansion item is the key name or field
6349 number. Leading and trailing spaces in this item were not being ignored,
6350 causing some misleading effects.
6352 5. When deliver_drop_privilege was set, single queue runner processes started
6353 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6354 same command in the process it spins off) were not dropping privilege.
6356 6. When the daemon running as "exim" started a queue runner, it always
6357 re-executed Exim in the spun-off process. This is a waste of effort when
6358 deliver_drop_privilege is set. The new process now just calls the
6359 queue-runner function directly.
6365 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6368 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6369 ignored. Though the use of -oP was forcing the writing of a pid file, it
6370 was always written to the default place.
6372 3. If the message "no IP address found for host xxxx" is generated during
6373 incoming verification, it is now followed by identification of the incoming
6374 connection (so you can more easily find what provoked it).
6376 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6378 5. Added some features to "harden" Exim a bit more against certain attacks:
6380 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6381 be put in Local/Makefile. This is like the never_users runtime option,
6382 but it cannot be overridden. The default setting is "root".
6384 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6385 prefix string with which any file named in a -C command line option
6388 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6389 is retained for -C and -D only if the caller of Exim is root. Without
6390 it, the exim user may also use -C and -D and retain privilege.
6392 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6393 command line option is disabled.
6395 6. Macro names set by the -D option must start with an upper case letter, just
6396 like macro names defined in the configuration file.
6398 7. Added "dereference=" facility to LDAP.
6400 8. Two instances of the typo "uknown" in the source files are fixed.
6402 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6403 the Configure-Makefile script screwed up while processing it.
6405 10. Incorporated PCRE 4.4.
6407 11. The SMTP synchronization check was not operating right at the start of an
6408 SMTP session. For example, it could not catch a HELO sent before the client
6409 waited for the greeting. There is now a check for outstanding input at the
6410 point when the greeting is written. Because of the duplex, asynchronous
6411 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6412 way, but not yet received, when the check is performed.
6414 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6415 on TCP/IP sockets, because this apparently causes some broken clients to
6418 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6419 unchanged) from the Cygwin maintainer.
6421 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6422 was supported instead of testing for mbx. Effectively a typo.
6424 15. The spa authenticator server code was not checking that the input it
6425 received was valid base64.
6427 16. The debug output line for the "set" modifier in ACLs was not showing the
6428 name of the variable that was being set.
6430 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6432 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6433 lookup on its own. Something else was needed to trigger the lookup. For
6434 example, a match in host_lookup or the need for a host name in a host list.
6435 Now, if $sender_host_name is referenced and the host name has not yet been
6436 looked up, a lookup is performed. If the lookup fails, the variable remains
6437 empty, and $host_lookup_failed is set to "1".
6439 19. Added "eqi" as a case-independent comparison operator.
6441 20. The saslauthd authentication condition could segfault if neither service
6442 nor realm was specified.
6444 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6445 error message that was logged was misleading, and incoming SMTP
6446 connections were dropped. The message is now more accurate, and temporary
6447 errors are given to SMTP connections.
6449 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6450 (except RSET) with a 421 error, until QUIT is received. However, it was
6451 failing to send a response to QUIT.
6453 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6454 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6455 ACL is now run after the helo_try_verify_hosts code.
6457 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6458 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6459 case-independent, so other case variants are also recognized). Apparently
6460 some systems use these upper case variants.
6462 25. If more than two messages were waiting for the same host, and a transport
6463 filter was specified for the transport, Exim sent two messages over the
6464 same TCP/IP connection, and then failed with "socket operation on non-
6465 socket" when it tried to send the third.
6467 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6469 27. The extern definition of crypt16() in expand.c was not being excluded when
6470 the OS had its own crypt16() function.
6472 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6473 as a preferred synonym for return_size_limit, both as an option and as an
6476 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6478 30. Changed the default configuration ACL to relax the local part checking rule
6479 for addresses that are not in any local domains. For these addresses,
6480 slashes and pipe symbols are allowed within local parts, but the sequence
6481 /../ is explicitly forbidden.
6483 31. SPA server authentication was not clearing the challenge buffer before
6486 32. log_message in a "warn" ACL statement was writing to the reject log as
6487 well as to the main log, which contradicts the documentation and doesn't
6488 seem right (because no rejection is happening). So I have stopped it.
6490 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6491 However, I am unable to do any testing of this.
6493 34. Fixed an infelicity in the appendfile transport. When checking directories
6494 for a mailbox, to see if any needed to be created, it was accidentally
6495 using path names with one or more superfluous leading slashes; tracing
6496 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6498 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6499 contents are added to the log line that is written for every discarded
6500 recipient. (Previously a log_message setting was ignored.)
6502 36. The ${quote: operator now quotes the string if it is empty.
6504 37. The install script runs exim in order to find its version number. If for
6505 some reason other than non-existence or emptiness, which it checks, it
6506 could not run './exim', it was installing it with an empty version number,
6507 i.e. as "exim-". This error state is now caught, and the installation is
6510 38. An argument was missing from the function that creates an error message
6511 when Exim fails to connect to the socket for saslauthd authentication.
6512 This could cause Exim to crash, or give a corrupted message.
6514 39. Added isip, isip4, and isip6 to ${if conditions.
6516 40. The ACL variables $acl_xx are now saved with the message, and can be
6517 accessed later in routers, transports, and filters.
6519 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6520 strings in the file are not string-expanded.
6522 42. If a MAIL command specified a SIZE value that was too large to fit into an
6523 int variable, the check against message_size_limit failed. Such values are
6524 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6525 day this will have to be increased, but I don't think I want to be around
6526 when emails are that large.
6533 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6534 iconv() is not standard in FreeBSD.
6536 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6537 IPv6 enabled. The observed symptom was a segmentation fault on return from
6538 the function os_common_find_running_interfaces() in src/os.c.
6540 3. In the check_special_case() function in daemon.c I had used "errno" as an
6541 argument name, which causes warnings on some systems. This was basically a
6542 typo, since it was named "eno" in the comments!
6544 4. The code that waits for the clock to tick (at a resolution of some fraction
6545 of a second) so as to ensure message-id uniqueness was always waiting for
6546 at least one whole tick, when it could have waited for less. [This is
6547 almost certainly not relevant at current processor speeds, where it is
6548 unlikely to ever wait at all. But we try to future-proof.]
6550 5. The function that sleeps for a time interval that includes fractions of a
6551 second contained a race. It did not block SIGALRM between setting the
6552 timer, and suspending (a couple of lines later). If the interval was short
6553 and the sigsuspend() was delayed until after it had expired, the suspension
6554 never ended. On busy systems this could lead to processes getting stuck for
6557 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6558 process, before it forks any delivery processes. The open lookup caching
6559 mechanism meant that the open file or database connection was passed into
6560 the delivery process. The problem was that delivery processes always tidy
6561 up cached lookup data. This could cause a problem for the next delivery
6562 process started by the queue runner, because the external queue runner
6563 process does not know about the closure. So the next delivery process
6564 still has data in the lookup cache. In the case of a file lookup, there was
6565 no problem because closing a file descriptor in a subprocess doesn't affect
6566 the parent. However, if the lookup was caching a connection to a database,
6567 the connection was closed, and the second delivery process was likely to
6568 see errors such as "PGSQL: query failed: server closed the connection
6569 unexpectedly". The problem has been fixed by closing all cached lookups
6570 in a queue runner before running a delivery process.
6572 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6573 seem to have the "const" qualifier which it has on other OS. I've
6576 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6577 *of the same type* (client or server) with the same public name that an
6578 error should be diagnosed.
6580 9. When Exim looked up a host name for an IP address, but failed to find the
6581 original IP address when looking up the host name (a safety check), it
6582 output the message "<ip address> does not match any IP for NULL", which was
6583 confusing, to say the least. The bug was that the host name should have
6584 appeared instead of "NULL".
6586 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6587 user that is built into the binary, and the -C or -D options is used, root
6588 privilege is dropped before the configuration file is read. In addition,
6589 logging is switched to stderr instead of the normal log files. If the
6590 configuration then re-defines the Exim user, the unprivileged environment
6591 is probably not what is expected, so Exim logs a panic warning message (but
6594 However, if deliver_drop_privilege is set, the unprivileged state may well
6595 be exactly what is intended, so the warning has been cut out in that case,
6596 and Exim is allowed to try to write to its normal log files.
6602 1. smtp_return_error_details was not giving details for temporary sender
6603 or receiver verification errors.
6605 2. Diagnose a configuration error if two authenticators have the same public
6608 3. Exim used not to create the message log file for a message until the first
6609 delivery attempt. This could be confusing when incoming messages were held
6610 for policy or load reasons. The message log file is now created at the time
6611 the message is received, and an initial "Received" line is written to it.
6613 4. The automatically generated man page for command line options had a minor
6614 bug that caused no ill effects; however, a more serious problem was that
6615 the procedure for building the man page automatically didn't always
6616 operate. Consequently, release 4.20 contains an out-of-date version. This
6617 shouldn't happen again.
6619 5. When building Exim with embedded Perl support, the script that builds the
6620 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6621 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6623 6. The freeze_tell option was not being used for messages that were frozen on
6624 arrival, either by an ACL or by local_scan().
6626 7. Added the smtp_incomplete_transaction log selector.
6628 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6629 was accepting AUTH without a new EHLO.
6631 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6632 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6635 10. Logging of TCP/IP connections (when configured) now happens in the main
6636 daemon process instead of the child process, so that the TCP/IP connection
6637 count is more accurate (but it can never be perfect).
6639 11. The use of "drop" in a nested ACL was not being handled correctly in the
6640 outer ACL. Now, if condition failure induced by the nested "drop" causes
6641 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6642 or "require"), the connection is dropped.
6644 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6645 that yield "discard" can now be used with an "accept" or a "discard" verb,
6646 but an error is generated for any others (because I can't see a useful way
6647 to define what should happen).
6649 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6650 are now processed in the same way as lines in the Exim configuration file.
6651 In particular, continuation lines are supported.
6653 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6655 15. Added -ti meaning -t -i.
6657 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6658 domains, and warn by logging if others are found.
6660 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6661 returned from GIFCONF when Exim is trying to find the list of interfaces on
6662 a host. The code in os.c has been modified to copy each ifreq to an aligned
6663 structure in all cases.
6665 Also, in some cases, the returned ifreq's were being copied to a 'struct
6666 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6667 means the last couple of bytes of an IPv6 address could be chopped if the
6668 ifreq contained only a normal sockaddr (14 bytes storage).
6670 18. Named domain lists were not supported in the hosts_treat_as_local option.
6671 An entry such as +xxxx was not recognized, and was treated as a literal
6674 19. Ensure that header lines added by a DATA ACL are included in the reject log
6675 if the ACL subsequently rejects the message.
6677 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6678 MD5 (which is deprecated).
6680 21. When testing a filter file using -bf, Exim was writing a message when it
6681 took the sender from a "From " line in the message, but it was not doing so
6682 when it took $return_path from a Return-Path: header line. It now does.
6684 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6685 with a valid header line field name (a series of printing characters
6686 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6688 23. Changed "disc" in the source to "disk" to conform to the documentation and
6689 the book and for uniformity.
6691 24. Ignore Sendmail's -Ooption=value command line item.
6693 25. When execve() failed while trying to run a command in a pipe transport,
6694 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6695 could be confused with a return value of 69 from the command itself. This
6696 has been changed to 127, the value the shell returns if it is asked to run
6697 a non-existent command. The wording for the related log line suggests a
6698 non-existent command as the problem.
6700 26. If received_header_text expands to an empty string, do not add a Received:
6701 header line to the message. (Well, it adds a token one on the spool, but
6702 marks it "old" so that it doesn't get used or transmitted.)
6704 27. Installed eximstats 1.28 (addition of -nt option).
6706 28. There was no check for failure on the call to getsockname() in the daemon
6707 code. This can fail if there is a shortage of resources on the system, with
6708 ENOMEM, for example. A temporary error is now given on failure.
6710 29. Contrary to the C standard, it seems that in some environments, the
6711 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6712 program. Exim now does this explicitly; it affects the formatting of
6713 timestamps using strftime().
6715 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6716 complaints. I've now initialized all the variables, to avoid this.
6718 32. Header lines added by a system filter were not being "seen" during
6719 transport-time rewrites.
6721 33. The info_callback() function passed to OpenSSL is set up with type void
6722 (*)(SSL *, int, int), as described somewhere. However, when calling the
6723 function (actually a macro) that sets it up, the type void(*)() is
6724 expected. I've put in a cast to prevent warnings from picky compilers.
6726 34. If a DNS black list lookup found a CNAME record, but there were no A
6727 records associated with the domain it pointed at, Exim crashed.
6729 35. If a DNS black list lookup returned more than one A record, Exim ignored
6730 all but the first. It now scans all returned addresses if a particular IP
6731 value is being sought. In this situation, the contents of the
6732 $dnslist_value variable are a list of all the addresses, separated by a
6735 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6736 to accept a host name and all its aliases if the forward lookup for any of
6737 them yielded the IP address of the incoming connection. Now it accepts only
6738 those names whose forward lookup yields the correct IP address. Any other
6739 names are discarded. This closes a loophole whereby a rogue DNS
6740 administrator could create reverse DNS records to break through a
6741 wildcarded host restriction in an ACL.
6743 37. If a user filter or a system filter that ran in a subprocess used any of
6744 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6745 the wrong values were passed to the pipe command ($thisaddress had the
6746 value of $0, $0 had the value of $1, etc). This bug was introduced by
6747 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6749 38. Improved the line breaking for long SMTP error messages from ACLs.
6750 Previously, if there was no break point between 40 and 75 characters, Exim
6751 left the rest of the message alone. Two changes have been made: (a) I've
6752 reduced the minimum length to 35 characters; (b) if it can't find a break
6753 point between 35 and 75 characters, it looks ahead and uses the first one
6754 that it finds. This may give the occasional overlong line, but at least the
6755 remaining text gets split now.
6757 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6758 file descriptors might be low, and that setting 1000 would always raise it.
6759 It turns out that in some environments, the limit is already over 1000 and
6760 that lowering it causes trouble. So now Exim takes care not to decrease it.
6762 40. When delivering a message, the value of $return_path is set to $sender_
6763 address at the start of routing (routers may change the value). By an
6764 oversight, this default was not being set up when an address was tested by
6765 -bt or -bv, which affected the outcome if any router or filter referred to
6768 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6769 log only when "message" or "log_message" are set. However, if one of the
6770 conditions was an address verification, or a call to a nested ACL, the
6771 messages generated by the underlying test were being passed through. This
6772 no longer happens. The underlying message is available in $acl_verify_
6773 message for both "message" and "log_message" expansions, so it can be
6774 passed through if needed.
6776 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6777 new expansion $bh_ to give the encoded byte string without charset
6778 translation. Translation happens only if iconv() is available; HAVE_ICONV
6779 indicates this at build time. HEADERS_CHARSET gives the charset to
6780 translate to; headers_charset can change it in the configuration, and
6781 "headers charset" can change it in an individual filter file.
6783 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6784 that creates RFC 2047 encoded "words" labels them as that charset instead
6785 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6786 expansion operator; (ii) when Exim creates a From: line for a local
6787 message; (iii) when a header line is rewritten to include a "phrase" part.
6789 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6790 buggy, causing it to skip the first lines of messages whose message ID
6791 ended in 'D'. This would not have bitten before Exim release 4.14, because
6792 message IDs were unlikely to end in 'D' before then. The effect was to have
6793 incorrect size information for certain domains.
6795 45. #include "config.h" was missing at the start of the crypt16.c module. This
6796 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6799 46. If there was a timeout during a "random" callout check, Exim treated it as
6800 a failure of the random address, and carried on sending RSET and the real
6801 address. If the delay was just some slowness somewhere, the response to the
6802 original RCPT would be taken as a response to RSET and so on, causing
6803 mayhem of various kinds.
6805 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6806 when I implemented it. It didn't allow for the fact that some option values
6807 may legitimately be negative (e.g. size_addition), and it didn't even do
6808 the right test for positive values.
6810 48. Domain names in DNS records are case-independent. Exim always looks them up
6811 in lower case. Some resolvers return domain names in exactly the case they
6812 appear in the zone file, that is, they may contain uppercase letters. Not
6813 all resolvers do this - some return always lower case. Exim was treating a
6814 change of case by a resolver as a change of domain, similar to a widening
6815 of a domain abbreviation. This triggered its re-routing code and so it was
6816 trying to route what was effectively the same domain again. This normally
6817 caused routing to fail (because the router wouldn't handle the domain
6818 twice). Now Exim checks for this case specially, and just changes the
6819 casing of the domain that it ultimately uses when it transmits the message
6822 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6825 50. If a filter generated a file delivery with a non-absolute name (possible if
6826 no home directory exists for the router), the forbid_file option was not
6829 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6830 the existing equality matching.
6832 52. Exim was using ints instead of ino_t variables in some places where it was
6833 dealing with inode numbers.
6835 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6836 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6837 called TMPDIR, and if it finds it is different, it changes its value.
6839 54. The smtp_printf() function is now made available to local_scan() so
6840 additional output lines can be written before returning. There is also an
6841 smtp_fflush() function to enable the detection of a dropped connection.
6842 The variables smtp_input and smtp_batched_input are exported to
6845 55. Changed the default runtime configuration: the message "Unknown user"
6846 has been removed from the ACL, and instead placed on the localuser router,
6847 using the cannot_route_message feature. This means that any verification
6848 failures that generate their own messages won't get overridden. Similarly,
6849 the "Unrouteable address" message that was in the ACL for unverifiable
6850 relay addresses has also been removed.
6852 56. Added hosts_avoid_esmtp to the smtp transport.
6854 57. The exicyclog script was not checking for the esoteric option
6855 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6856 will work only if exicyclog is run under the appropriate euid.
6858 58. Following a discussion on the list, the rules by which Exim recognises line
6859 endings on incoming messages have been changed. The -dropcr and drop_cr
6860 options are now no-ops, retained only for backwards compatibility. The
6861 following line terminators are recognized: LF CRLF CR. However, special
6862 processing applies to CR:
6864 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6865 nor a local message in the state where . is a terminator.
6867 (ii) If a bare CR is encountered in a header line, an extra space is added
6868 after the line terminator so as not to end the header. The reasoning
6869 behind this is that bare CRs in header lines are most likely either
6870 to be mistakes, or people trying to play silly games.
6872 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6873 was being incorrectly given as 18 bytes larger than it should have been.
6874 This is a VOB (very old bug).
6876 60. This may never have affected anything current, but just in case it has:
6877 When the local host is found other than at the start of a list of hosts,
6878 the local host, those with the same MX, and any that follow, are discarded.
6879 When the list in question was part of a longer list of hosts, the following
6880 hosts (not currently being processed) were also being discarded. This no
6881 longer happens. I'm not sure if this situation could ever has previously
6884 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6887 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6888 header. This is something that is recommended in a new Internet Draft, and
6889 is something that is documented as being done by Sendmail. There are two
6890 possible values. For messages generated by the autoreply transport, Exim
6893 Auto-Submitted: auto-replied
6895 whereas for all other generated messages (e.g. bounces) it adds
6897 Auto-Submitted: auto-generated
6899 63. The "personal" condition in filters now includes a test for the
6900 Auto-Submitted: header. If it contains the string "auto-" the message it
6901 not considered personal.
6903 64. Added rcpt_include_affixes as a generic transport option.
6905 65. Added queue_only_override (default true).
6907 66. Added the syslog_duplication option.
6909 67. If what should have been the first header line of a message consisted of
6910 a space followed by a colon, Exim was mis-interpreting it as a header line.
6911 It isn't of course - it is syntactically invalid and should therefore be
6912 treated as the start of the message body. The misbehaviour could have
6913 caused a number of strange effects, including loss of data in subsequent
6914 header lines, and spool format errors.
6916 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6917 client host had authenticated. This control can now be exercised by an ACL
6918 for more flexibility.
6920 69. By default, callouts do not happen when testing with -bh. There is now a
6921 variant, -bhc, which does actually run the callout code, including
6922 consulting and updating the callout cache.
6924 70. Added support for saslauthd authentication, courtesy of Alexander
6927 71. If statvfs() failed on the spool or log directories while checking their
6928 size for availability, Exim confusingly gave the error "space shortage".
6929 Furthermore, in debugging mode it crashed with a floating point exception.
6930 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6931 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6932 serious problem, Exim now writes to the main and panic logs when this
6933 happens, with details of the failure. It then refuses to accept the
6934 incoming message, giving the message "spool directory problem" or "log
6935 directory problem" with a 421 code for SMTP messages.
6937 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6938 0, 1, and 2 exist, because some OS complain for execs without them (see
6939 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6940 descriptors. However, the code omitted to check that the open succeeded,
6941 causing mysterious errors if for some reason the permissions on /dev/null
6942 got screwed. Now Exim writes a message to the main and panic logs, and
6943 bombs out if it can't open /dev/null.
6945 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6946 interact so that it is all more flexible. It is supposed to remain
6947 backwards compatible. Also added extra_local_interfaces.
6949 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6950 to bomb out with an assertion failure - to the client this appears as a
6951 connection drop. This problem occurs in the part of the code that was taken
6952 from the Samba project. Fortunately, the assertion is in a very simple
6953 function, so I have fixed this by reproducing the function inline in the
6954 one place where it is called, and arranging for authentication to fail
6955 instead of killing the process with assert().
6957 75. The SPA client code was not working when the server requested OEM rather
6958 than Unicode encoding.
6960 76. Added code to make require_files with a specific uid setting more usable in
6961 the case where statting the file as root fails - usually a non-root-mounted
6962 NFS file system. When this happens and the failure is EACCES, Exim now
6963 forks a subprocess and does the per-uid checking as the relevant uid.
6965 77. Added process_log_path.
6967 78. If log_file_path was not explicitly set, a setting of check_log_space or
6968 check_log_inodes was ignored.
6970 79. If a space check for the spool or log partitions fails, the incident is now
6971 logged. Of course, in the latter case the data may get lost...
6973 80. Added the %p formatting code to string_format() so that it can be used to
6974 print addresses in debug_print(). Adjusted all the address printing in the
6975 debugging in store.c to use %p rather than %d.
6977 81. There was a concern that a line of code in smtp_in.c could overflow a
6978 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6979 initially expressed, the concern was not well-founded, because trailing
6980 spaces are removed early. However, if the trailing spaces were followed by
6981 a NULL, they did not get removed, so the overflow was possible. Two fixes
6984 (a) I re-wrote the offending code in a cleaner fashion.
6985 (b) If an incoming SMTP command contains a NULL character, it is rejected
6988 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6989 runs initgroups(), so that if the Exim user is in any additional groups,
6990 they will be used during message reception.
6996 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / heiko / exim.git / blob