1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
11 JH/01 Replace the store_release() internal interface with store_newblock(),
12 which internalises the check required to safely use the old one, plus
13 the allocate and data copy operations duplicated in both (!) of the
16 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
17 modifier. This matches the restriction on the commandline.
19 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
20 Previously only the last row was returned.
22 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
23 we assumed that tags in the header were well-formed, and parsed the
24 element content after inspecting only the first char of the tag.
25 Assumptions at that stage could crash the receive process on malformed
28 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
29 While running the DKIM ACL we operate on the Permanent memory pool so that
30 variables created with "set" persist to the DATA ACL. Also (at any time)
31 DNS lookups that fail create cache records using the Permanent pool. But
32 expansions release any allocations made on the current pool - so a dnsdb
33 lookup expansion done in the DKIM ACL releases the memory used for the
34 DNS negative-cache, and bad things result. Solution is to switch to the
35 Main pool for expansions.
36 While we're in that code, add checks on the DNS cache during store_reset,
37 active in the testsuite.
38 Problem spotted, and debugging aided, by Wolfgang Breyha.
40 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
41 When none of the hosts presented to a transport match an already-open
42 connection, close it and proceed with the list. Previously we would
43 queue the message. Spotted by Lena with Yahoo, probably involving
46 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
47 Previously a spurious "250 OK id=" response was appended to the proper
50 JH/08 The "support for" informational output now, which built with Content
51 Scanning support, has a line for the malware scanner interfaces compiled
52 in. Interface can be individually included or not at build time.
54 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
55 by the template makefile "src/EDITME". The "STREAM" support for an older
56 ClamAV interface method is removed.
58 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
59 rows affected is given instead).
61 JH/11 The runtime Berkeley DB library version is now additionally output by
62 "exim -d -bV". Previously only the compile-time version was shown.
64 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
65 SMTP connection. Previously, when one had more receipients than the
66 first, an abortive onward connection was made. Move to full support for
67 multiple onward connections in sequence, handling cutthrough connection
68 for all multi-message initiating connections.
70 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
71 routers. Previously, a multi-recipient message would fail to match the
72 onward-connection opened for the first recipient, and cause its closure.
74 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
75 a timeout on read on a GnuTLS initiating connection, resulting in the
76 initiating connection being dropped. This mattered most when the callout
77 was marked defer_ok. Fix to keep the two timeout-detection methods
80 JH/15 Relax results from ACL control request to enable cutthrough, in
81 unsupported situations, from error to silently (except under debug)
82 ignoring. This covers use with PRDR, frozen messages, queue-only and
85 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
86 metadata, resulting in a crash in free().
92 JH/01 Rework error string handling in TLS interface so that the caller in
93 more cases is responsible for logging. This permits library-sourced
94 string to be attached to addresses during delivery, and collapses
95 pairs of long lines into single ones.
97 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
98 during configuration. Wildcards are allowed and expanded.
100 JH/02 Rework error string handling in DKIM to pass more info back to callers.
101 This permits better logging.
103 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
104 do not close it down and have the child transport start it up again on
105 the passed-on TCP connection. Instead, proxy the child (and any
106 subsequent ones) for TLS via a unix-domain socket channel. Logging is
107 affected: the continued delivery log lines do not have any DNSSEC, TLS
108 Certificate or OCSP information. TLS cipher information is still logged.
110 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
111 identical IP addresses on different listening ports. Will also affect
114 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
115 add noisy ifdef guards to special-case this sillyness.
116 Patch from Bernd Kuhls.
118 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
119 than 255 are no longer allowed.
121 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
122 Disable the session-cache too, which might reduce our load. Since we
123 currrectly use a new context for every connection, both as server and
124 client, there is no benefit for these.
125 GnuTLS appears to not support tickets server-side by default (we don't
126 call gnutls_session_ticket_enable_server()) but client side is enabled
127 by default on recent versions (3.1.3 +) unless the PFS priority string
130 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
131 <https://reproducible-builds.org/specs/source-date-epoch/>.
133 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
134 the check for any unsuccessful recipients did not notice the limit, and
135 erroneously found still-pending ones.
137 JH/08 Pipeline CHUNKING command and data together, on kernels that support
138 MSG_MORE. Only in-clear (not on TLS connections).
140 JH/09 Avoid using a temporary file during transport using dkim. Unless a
141 transport-filter is involved we can buffer the headers in memory for
142 creating the signature, and read the spool data file once for the
143 signature and again for transmission.
145 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
146 4.77 as the kernel support then wasn't solid, having issues in 64bit
147 mode. Now, it's been long enough. Add support for FreeBSD also.
149 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
150 case where the routing stage had gathered several addresses to send to
151 a host before calling the transport for the first, we previously failed
152 to close down TLS in the old transport process before passing the TCP
153 connection to the new process. The new one sent a STARTTLS command
154 which naturally failed, giving a failed delivery and bloating the retry
155 database. Investigation and fix prototype from Wolfgang Breyha.
157 JH/12 Fix check on SMTP command input synchronisation. Previously there were
158 false-negatives in the check that the sender had not preempted a response
159 or prompt from Exim (running as a server), due to that code's lack of
160 awareness of the SMTP input buffering.
162 PP/04 Add commandline_checks_require_admin option.
163 Exim drops privileges sanely, various checks such as -be aren't a
164 security problem, as long as you trust local users with access to their
165 own account. When invoked by services which pass untrusted data to
166 Exim, this might be an issue. Set this option in main configuration
167 AND make fixes to the calling application, such as using `--` to stop
170 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
171 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
172 responses to those, into a single TLS record each way (this usually means
173 a single packet). As a side issue, smtp_enforce_sync now works on TLS
176 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
177 affects you only if you're dancing at the edge of the param size limits.
178 If you are, and this message makes sense to you, then: raise the
179 configured limit or use OpenSSL 1.1. Nothing we can do for older
182 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
183 cmdline element to get the documented default one. Previously it was
186 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
187 are unsuable for non-message-specific reasons - eg. connection timeout,
188 banner-time rejection.
190 JH/16 Fix logging of delivery remote port, when specified by router, under
193 PP/06 Repair manualroute's ability to take options in any order, even if one
194 is the name of a transport.
197 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
199 JH/17 Change the list-building routines interface to use the expanding-string
200 triplet model, for better allocation and copying behaviour.
202 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
203 Previously it was constructed the first time a possibly-matching string
204 was met in the configuration file input during startup; now it is done
207 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
208 compatible one, to avoid the (poorly documented) possibility of a config
209 file in the working directory redirecting the DB files, possibly correpting
210 some existing file. CVE-2017-10140 assigned for BDB.
212 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
213 cache-hot. Previously, although the result was properly cached, the
214 initial verify call returned a defer.
216 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
217 the main verify for receipient in uncached-mode.
219 JH/22 Retire historical build files to an "unsupported" subdir. These are
220 defined as "ones for which we have no current evidence of testing".
222 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
223 if present. Previously it was ignored.
225 JH/24 Start using specified-initialisers in C structure init coding. This is
226 a C99 feature (it's 2017, so now considered safe).
228 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
229 if was a fixed-sized field and bitmask ops via macros; it is now more
232 PP/07 GitHub PR 56: Apply MariaDB build fix.
233 Patch provided by Jaroslav Škarvada.
235 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
236 during Coverity cleanups [4.87 JH/47]
237 Diagnosis and fix provided by Michael Fischer v. Mollard.
239 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
240 the right size to place the terminating semicolon on its own folded
241 line, the header hash was calculated to an incorrect value thanks to
242 the (relaxed) space the fold became.
244 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
245 and confused the parent.
247 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
248 which could crash as a result. This could lead to undeliverable messages.
250 JH/28 Logging: "next input sent too soon" now shows where input was truncated
253 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
254 matters on fast-turnover and PID-randomising systems, which were getting
255 out-of-order delivery.
257 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
258 a possibly-overlapping copy. The symptom was that "Remote host closed
259 connection in response to HELO" was logged instead of the actual 4xx
262 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
263 Previously only that bufferd was discarded, resulting in SYMTP command
266 JH/32 DKIM: when a message has multiple signatures matching an identity given
267 in dkim_verify_signers, run the dkim acl once for each. Previously only
268 one run was done. Bug 2189.
270 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
271 "panic the current process" to "deliberately defer". The panic log is
272 still written with the problem list name; the mail and reject logs now
273 get a temp-reject line for the message that was being handled, saying
274 something like "domains check lookup or other defer". The SMTP 451
275 message is still "Temporary local problem".
277 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
278 A crafted sequence of BDAT commands could result in in-use memory beeing
279 freed. CVE-2017-16943.
281 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
282 from SMTP input. Previously it was always done; now only done for DATA
283 and not BDAT commands. CVE-2017-16944.
285 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
286 to the message (such as an overlong header line). Previously this was
287 not done and we did not exit BDAT mode. Followon from the previous item
288 though a different problem.
294 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
295 than -2003 did; needs libidn2 in addition to libidn.
297 JH/02 The path option on a pipe transport is now expanded before use.
299 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
300 Patch provided by "Björn", documentation fix added too.
302 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
303 missing a wire-to-host endian conversion.
305 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
306 close after a BDAT command line could be taken as a following command,
307 giving a synch failure. Fix by only checking for synch immediately
308 before acknowledging the chunk.
310 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
311 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
312 macro. Patches provided by Josh Soref.
314 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
315 Previously we did not; the RFC seems ambiguous and VRFY is not listed
316 by IANA as a service extension. However, John Klensin suggests that we
319 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
320 the dkim code may be unix-mode line endings rather than smtp wire-format
321 CRLF, so prepend a CR to any bare LF.
323 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
324 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
326 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
327 the first were themselves being wrongly included in the feed into dkim
328 processing; with most chunk sizes in use this resulted in an incorrect
329 body hash calculated value.
331 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
332 DKIM signature block, for verification. Although advised against by
333 standards it is specifically not ruled illegal.
335 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
337 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
338 missing a body hash (the bh= tag).
340 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
341 It seems that HAProxy sends the Proxy Protocol information in clear and
342 only then does a TLS startup, so do the same.
344 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
345 TCP connections (such as for Spamd) unless the daemon successfully set
346 Fast Open mode on its listening sockets. This fixes breakage seen on
347 too-old kernels or those not configured for Fast Open, at the cost of
348 requiring both directions being enabled for TFO, and TFO never being used
349 by non-daemon-related Exim processes.
351 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
352 endings, at least on the first header line. Try to canonify any that get
353 past that check, despite the cost.
355 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
356 now limited to an arbitrary five deep, while parsing addresses with the
357 strip_excess_angle_brackets option enabled.
359 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
360 instead leave the unprompted TLS handshake in socket buffer for the
361 TLS library to consume.
363 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
365 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
367 JH/16 Drop variables when they go out of scope. Memory management drops a whole
368 region in one operation, for speed, and this leaves assigned pointers
369 dangling. Add checks run only under the testsuite which checks all
370 variables at a store-reset and panics on a dangling pointer; add code
371 explicitly nulling out all the variables discovered. Fixes one known
372 bug: a transport crash, where a dangling pointer for $sending_ip_address
373 originally assigned in a verify callout, is re-used.
375 PP/06 Drop '.' from @INC in various Perl scripts.
377 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
379 PP/08 Reduce a number of compilation warnings under clang; building with
380 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
381 should be warning-free.
383 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
385 HS/01 Fix portability problems introduced by PP/08 for platforms where
386 realloc(NULL) is not equivalent to malloc() [SunOS et al].
388 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
389 chunk. This allows us to accept broken chunked messages. We need a more
390 general solution here.
392 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
393 already-broken messages in the queue.
395 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
397 JH/19 Fix reference counting bug in routing-generated-address tracking.
403 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
404 supports it and a size is available (ie. the sending peer gave us one).
406 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
407 years of being deprecated). The replacements are the ACLs
408 acl_smtp_mime and acl_not_smtp_mime.
410 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
411 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
412 or even an in-clear connection were permitted. Now, if the host lookup
413 was dnssec and dane was requested then the host is only used if the
414 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
415 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
416 if one fails this test.
417 This means that a poorly-configured remote DNS will make it incommunicado;
418 but it protects against a DNS-interception attack on it.
420 JH/04 Bug 1810: make continued-use of an open smtp transport connection
421 non-noisy when a race steals the message being considered.
423 JH/05 If main configuration option tls_certificate is unset, generate a
424 self-signed certificate for inbound TLS connections.
426 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
427 in rewrites and routers.
429 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
430 and logged a warning sing 4.83; now they are a configuration file error.
432 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
433 (lacking @domain). Apply the same qualification processing as RCPT.
435 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
437 JH/10 Support ${sha256:} applied to a string (as well as the previous
440 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
441 a cutthrough deliver is pending, as we always want to make a connection.
442 This also avoids re-routing the message when later placing the cutthrough
443 connection after a verify cache hit.
444 Do not update it with the verify result either.
446 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
447 when routing results in more than one destination address.
449 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
450 signing (which inhibits the cutthrough capability). Previously only
451 the presence of an option was tested; now an expansion evaluating as
452 empty is permissible (obviously it should depend only on data available
453 when the cutthrough connection is made).
455 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
456 the relevant preceding SMTP command did not note the pipelining mode.
458 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
459 Previously they were not counted.
461 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
462 as one having no matching records. Previously we deferred the message
463 that needed the lookup.
465 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
466 distinguished as "(=".
468 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
469 for missing MX records. Previously it only worked for missing A records.
471 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
473 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
474 after the data-go-ahead and data-ack. Patch from Jason Betts.
476 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
477 even for a "none" policy. Patch from Tony Meyer.
479 JH/22 Fix continued use of a connection for further deliveries. If a port was
480 specified by a router, it must also match for the delivery to be
483 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
484 When one of the recipients of a message was unsuitable for the connection
485 (has no matching addresses), we lost track of needing to mark it
486 deferred. As a result mail would be lost.
488 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
490 JH/25 Decoding ACL controls is now done using a binary search; the source code
491 takes up less space and should be simpler to maintain. Merge the ACL
492 condition decode tables also, with similar effect.
494 JH/26 Fix problem with one_time used on a redirect router which returned the
495 parent address unchanged. A retry would see the parent address marked as
496 delivered, so not attempt the (identical) child. As a result mail would
499 JH/27 Fix a possible security hole, wherein a process operating with the Exim
500 UID can gain a root shell. Credit to http://www.halfdog.net/ for
501 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
504 JH/28 Enable {spool,log} filesystem space and inode checks as default.
505 Main config options check_{log,spool}_{inodes,space} are now
506 100 inodes, 10MB unless set otherwise in the configuration.
508 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
509 Previously it only applied to the main-section connection policy
512 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
514 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
515 by me. Added RFC7919 DH primes as an alternative.
517 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
518 are not in the system include path.
520 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
521 GnuTLS, when a session startup failed (eg because the client disconnected)
522 Exim did stdio operations after fclose. This was exposed by a recent
523 change which nulled out the file handle after the fclose.
525 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
526 signed directly by the cert-signing cert, rather than an intermediate
527 OCSP-signing cert. This is the model used by LetsEncrypt.
529 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
531 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
532 an incoming connection.
534 HS/02 Bug 1802: Do not half-close the connection after sending a request
537 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
538 fallback to "prime256v1".
540 JH/34 SECURITY: Use proper copy of DATA command in error message.
541 Could leak key material. Remotely exploitable. CVE-2016-9963.
547 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
548 and 3.4.4 - once the server is enabled to respond to an OCSP request
549 it does even when not requested, resulting in a stapling non-aware
550 client dropping the TLS connection.
552 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
553 support variable-length bit vectors. No functional change.
555 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
556 The I= interface field on outgoing lines is now after the H= remote
557 host field, same as incoming lines. There is a separate
558 outgoing_interface log selector which allows you to disable the
561 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
562 If not running log_selector +smtp_connection the mainlog would be held
563 open indefinitely after a "too many connections" event, including to a
564 deleted file after a log rotate. Leave the per net connection logging
565 leaving it open for efficiency as that will be quickly detected by the
566 check on the next write.
568 HS/01 Bug 1671: Fix post transport crash.
569 Processing the wait-<transport> messages could crash the delivery
570 process if the message IDs didn't exist for some reason. When
571 using 'split_spool_directory=yes' the construction of the spool
572 file name failed already, exposing the same netto behaviour.
574 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
575 mime_regex ACL conditions.
577 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
578 to DSN fail messages (bounces): remote IP, remote greeting, remote response
579 to HELO, local diagnostic string.
581 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
582 log line to debug. Even when configured with a tls authenticator many
583 client connections are expected to not authenticate in this way, so
584 an authenticate fail is not an error.
586 HS/02 Add the Exim version string to the process info. This way exiwhat
587 gives some more detail about the running daemon.
589 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
590 matter for fast-change records such as DNSBLs.
592 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
593 retry record, even if constant. There may be multiple transports with
594 different interface settings and the retry behaviour needs to be kept
597 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
599 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
601 JH/10 Bug 840: fix log_defer_output option of pipe transport
603 JH/11 Bug 830: use same host for all RCPTS of a message, even under
604 hosts_randomize. This matters a lot when combined with mua_wrapper.
606 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
607 ${quote_pgsql:<string>} operator.
609 JH/13 Bug 1708: avoid misaligned access in cached lookup.
611 JH/14 Change header file name for freeradius-client. Relevant if compiling
612 with Radius support; from the Gentoo tree and checked under Fedora.
614 JH/15 Bug 1712: Introduce $prdr_requested flag variable
616 JH/16 Bug 1714: Permit an empty string as expansion result for transport
617 option transport_filter, meaning no filtering.
619 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
621 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
622 defaults to "*" (all hosts). The variable is now available when not built
623 with TLS, default unset, mainly to enable keeping the testsuite sane.
624 If a server certificate is not supplied (via tls_certificate) an error is
625 logged, and clients will find TLS connections fail on startup. Presumably
626 they will retry in-clear.
627 Packagers of Exim are strongly encouraged to create a server certificate
628 at installation time.
630 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
631 with the $config_file variable.
633 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
634 in transport context, after the attempt, and per-recipient. The latter type
635 is per host attempted. The event data is the error message, and the errno
636 information encodes the lookup type (A vs. MX) used for the (first) host,
637 and the trailing two digits of the smtp 4xx response.
639 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
640 to write to mainlog (or rejectlog, paniclog) in the window between file
641 creation and permissions/ownership being changed. Particularly affects
642 installations where exicyclog is run as root, rather than exim user;
643 result is that the running daemon panics and dies.
645 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
647 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
648 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
649 "pri" and "weight". Note that the previous implicit priority given by the
650 list order is no longer honoured.
652 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
655 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
656 by defining SUPPORT_SOCKS.
658 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
659 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
660 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
661 variables are renamed to proxy_{local,external}_{address,port}.
663 JH/27 Move Internationalisation support from Experimental to mainline, enabled
664 for a build by defining SUPPORT_I18N
666 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
667 of the query string, and make ${quote_redis:} do that quoting.
669 JH/29 Move Events support from Experimental to mainline, enabled by default
670 and removable for a build by defining DISABLE_EVENT.
672 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
674 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
675 cached by the daemon.
677 JH/32 Move Redis support from Experimental to mainline, enabled for a build
678 by defining LOOKUP_REDIS. The libhiredis library is required.
680 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
681 keys are given for lookup.
683 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
684 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
685 only supported when built with TLS support. The PolarSSL SHA routines
686 are still used when the TLS library is too old for convenient support.
688 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
689 openssl_options), for security. OpenSSL forces this from version 1.1.0
690 server-side so match that on older versions.
692 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
693 allocation for $value could be released as the expansion processing
694 concluded, but leaving the global pointer active for it.
696 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
697 and to use the domains and local_parts ACL conditions.
699 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
700 incorrectly not doubled on cutthrough transmission, hence seen as a
701 body-termination at the receiving system - resulting in truncated mails.
702 Commonly the sender saw a TCP-level error, and retransmitted the message
703 via the normal store-and-forward channel. This could result in duplicates
704 received - but deduplicating mailstores were liable to retain only the
705 initial truncated version.
707 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
709 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
711 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
712 we're in there, support oversigning also; bug 1309.
714 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
716 HS/04 Add support for keep_environment and add_environment options.
718 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
719 either intentional arithmetic overflow during PRNG, or testing config-
722 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
723 delivery resulted in actual delivery. Cancel cutthrough before DATA
726 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
727 rejects a recipient: pass the reject to the originator.
729 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
730 Many were false-positives and ignorable, but it's worth fixing the
733 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
734 for the new environment-manipulation done at startup. Move the routines
735 from being local to tls.c to being global via the os.c file.
737 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
738 an extract embedded as result-arg for a map, the first arg for extract
739 is unavailable so we cannot tell if this is a numbered or keyed
740 extraction. Accept either.
746 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
749 JH/02 The smtp transport option "multi_domain" is now expanded.
751 JH/03 The smtp transport now requests PRDR by default, if the server offers
754 JH/04 Certificate name checking on server certificates, when exim is a client,
755 is now done by default. The transport option tls_verify_cert_hostnames
756 can be used to disable this per-host. The build option
757 EXPERIMENTAL_CERTNAMES is withdrawn.
759 JH/05 The value of the tls_verify_certificates smtp transport and main options
760 default to the word "system" to access the system default CA bundle.
761 For GnuTLS, only version 3.0.20 or later.
763 JH/06 Verification of the server certificate for a TLS connection is now tried
764 (but not required) by default. The verification status is now logged by
765 default, for both outbound TLS and client-certificate supplying inbound
768 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
771 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
772 Status Notification (bounce) messages are now MIME format per RFC 3464.
773 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
774 under the control of the dsn_advertise_hosts option, and routers may
775 have a dsn_lasthop option.
777 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
778 default, modifiable by a malware= option. The list separator for
779 the options can now be changed in the usual way. Bug 68.
781 JH/10 The smtp_receive_timeout main option is now expanded before use.
783 JH/11 The incoming_interface log option now also enables logging of the
784 local interface on delivery outgoing connections.
786 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
787 if the interface and destination host and port all match.
789 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
792 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
793 Patch from Andrew Lewis.
795 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
796 now supports optional time-restrictions, weighting, and priority
797 modifiers per server. Patch originally by <rommer@active.by>.
799 JH/16 The spamd_address main option now supports a mixed list of local
800 and remote servers. Remote servers can be IPv6 addresses, and
801 specify a port-range.
803 JH/17 Bug 68: The spamd_address main option now supports an optional
804 timeout value per server.
806 JH/18 Bug 1581: Router and transport options headers_add/remove can
807 now have the list separator specified.
809 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
812 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
815 JH/21 Support for the A6 type of dns record is withdrawn.
817 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
818 rather than the verbs used.
820 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
821 from 255 to 1024 chars.
823 JH/24 Verification callouts now attempt to use TLS by default.
825 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
826 are generic router options now. The defaults didn't change.
828 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
829 Original patch from Alexander Shikoff, worked over by JH.
831 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
832 files if ZCAT_COMMAND is not executable.
834 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
836 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
838 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
839 Normally benign, it bites when the pair was led to by a CNAME;
840 modern usage is to not canonicalize the domain to a CNAME target
841 (and we were inconsistent anyway for A-only vs AAAA+A).
843 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
845 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
846 when evaluating $sender_host_dnssec.
848 JH/31 Check the HELO verification lookup for DNSSEC, adding new
849 $sender_helo_dnssec variable.
851 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
853 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
855 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
857 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
858 documented as working, but never had. Support all but $spam_report.
860 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
861 added for tls authenticator.
863 HS/03 Add perl_taintmode main config option
869 TL/01 When running the test suite, the README says that variables such as
870 no_msglog_check are global and can be placed anywhere in a specific
871 test's script, however it was observed that placement needed to be near
872 the beginning for it to behave that way. Changed the runtest perl
873 script to read through the entire script once to detect and set these
874 variables, reset to the beginning of the script, and then run through
875 the script parsing/test process like normal.
877 TL/02 The BSD's have an arc4random API. One of the functions to induce
878 adding randomness was arc4random_stir(), but it has been removed in
879 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
880 function when detected.
882 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
883 cause callback expansion.
885 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
886 syntax errors in an expansion can be treated as a string instead of
887 logging or causing an error, due to the internal use of bool_lax
888 instead of bool when processing it.
890 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
891 server certificates when making smtp deliveries.
893 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
895 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
897 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
899 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
900 Merged patch from Sebastian Wiedenroth.
902 JH/05 Fix results-pipe from transport process. Several recipients, combined
903 with certificate use, exposed issues where response data items split
904 over buffer boundaries were not parsed properly. This eventually
905 resulted in duplicates being sent. This issue only became common enough
906 to notice due to the introduction of connection certificate information,
907 the item size being so much larger. Found and fixed by Wolfgang Breyha.
909 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
910 size buffer was used, resulting in syntax errors when an expansion
913 JH/07 Add support for directories of certificates when compiled with a GnuTLS
914 version 3.3.6 or later.
916 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
917 is EXPERIMENTAL_EVENT, the main-configuration and transport options
918 both become "event_action", the variables become $event_name, $event_data
919 and $event_defer_errno. There is a new variable $verify_mode, usable in
920 routers, transports and related events. The tls:cert event is now also
921 raised for inbound connections, if the main configuration event_action
924 TL/06 In test suite, disable OCSP for old versions of openssl which contained
925 early OCSP support, but no stapling (appears to be less than 1.0.0).
927 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
928 server certificate names available under the smtp transport option
929 "tls_verify_cert_hostname" now do not permit multi-component wildcard
932 JH/10 Time-related extraction expansions from certificates now use the main
933 option "timezone" setting for output formatting, and are consistent
934 between OpenSSL and GnuTLS compilations. Bug 1541.
936 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
937 encoded parameter in the incoming message. Bug 1558.
939 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
940 include certificate info, eximon was claiming there were spoolfile
943 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
945 JH/14 Log delivery-related information more consistently, using the sequence
946 "H=<name> [<ip>]" wherever possible.
948 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
949 are problematic for Debian distribution, omit them from the release
952 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
954 JH/16 Fix string representation of time values on 64bit time_t architectures.
957 JH/17 Fix a null-indirection in certextract expansions when a nondefault
958 output list separator was used.
963 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
964 checkers that were complaining about end of non-void function with no
967 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
968 This was a regression introduced in 4.83 by another bugfix.
970 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
972 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
973 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
979 TF/01 Correctly close the server side of TLS when forking for delivery.
981 When a message was received over SMTP with TLS, Exim failed to clear up
982 the incoming connection properly after forking off the child process to
983 deliver the message. In some situations the subsequent outgoing
984 delivery connection happened to have the same fd number as the incoming
985 connection previously had. Exim would try to use TLS and fail, logging
986 a "Bad file descriptor" error.
988 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
989 utilities have not been installed.
991 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
992 temporary space as the ACL may create new global variables.
994 TL/01 LDAP support uses per connection or global context settings, depending
995 upon the detected version of the libraries at build time.
997 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
998 to extract and use the src ip:port in logging and expansions as if it
999 were a direct connection from the outside internet. PPv2 support was
1000 updated based on HAProxy spec change in May 2014.
1002 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1004 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1005 Properly escape header and check for NULL return.
1007 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1010 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1012 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1013 characters in header names, implemented as a verify condition.
1014 Contributed by Michael Fischer v. Mollard.
1016 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1017 results permerror and temperror. Previous values are deprecated but
1018 still accepted. In a future release, err_perm and err_temp will be
1019 completely removed, which will be a backward incompatibility if the
1020 ACL tests for either of these two old results. Patch contributed by
1021 user bes-internal on the mailing list.
1023 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1025 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1026 selectors, in both main and reject logs.
1028 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1031 JH/07 Add malware type "sock" for talking to simple daemon.
1033 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1035 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1036 routers/transports under cutthrough routing.
1038 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1039 numbers. Touch up "bool" conditional to keep the same definition.
1041 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1043 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1045 JH/12 Expand items in router/transport headers_add or headers_remove lists
1046 individually rather than the list as a whole. Bug 1452.
1048 Required for reasonable handling of multiple headers_ options when
1049 they may be empty; requires that headers_remove items with embedded
1050 colons must have them doubled (or the list-separator changed).
1052 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1053 view the policy declared in the DMARC record. Currently, $dmarc_status
1054 is a combined value of both the record presence and the result of the
1057 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1059 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1060 dnslookup router and the smtp transport (applying to the forward
1063 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1064 of ldap servers used for a specific lookup. Patch provided by Heiko
1067 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1068 New variable $lookup_dnssec_authenticated for observability.
1070 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1071 Patch submitted by Lars Timman.
1073 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1075 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1076 Requires trusted mode and valid format message id, aborts otherwise.
1077 Patch contributed by Heiko Schlichting.
1079 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1080 certextract with support for various fields. Bug 1358.
1082 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1083 is requested by default, modifiable by smtp transport option
1086 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1087 operate on certificate variables to give certificate fingerprints
1088 Also new ${sha256:cert_variable}.
1090 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1092 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1093 Christian Aistleitner.
1095 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1097 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1098 file. Patch from Wolfgang Breyha.
1100 JH/25 Expand the coverage of the delivery $host and $host_address to
1101 client authenticators run in verify callout. Bug 1476.
1103 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1104 align with daemon_smtp_ports. Bug 72.
1106 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1107 support and error reporting did not work properly.
1109 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1110 and is readable. Patch from Andrew Colin Kissa.
1112 TL/14 Enhance documentation of ${run expansion and how it parses the
1113 commandline after expansion, particularly in the case when an
1114 unquoted variable expansion results in an empty value.
1116 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1118 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1119 by renaming away from it. Addresses GH issue 15, reported by
1122 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1124 TL/15 SECURITY: prevent double expansion in math comparison functions
1125 (can expand unsanitized data). Not remotely exploitable.
1132 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1134 PP/02 Make -n do something, by making it not do something.
1135 When combined with -bP, the name of an option is not output.
1137 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1140 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1141 $sender_host_name and config options to manage this, and basic check
1144 PP/05 DSCP support for outbound connections and control modifier for inbound.
1146 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1147 (Only plugin which currently uses this is kerberos4, which nobody should
1148 be using, but we should make it available and other future plugins might
1149 conceivably use it, even though it would break NAT; stuff *should* be
1150 using channel bindings instead).
1152 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1153 name; added for Sendmail compatibility; requires admin caller.
1154 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1155 just ignore it); requires trusted caller.
1156 Also parse but ignore: -Ac -Am -X<logfile>
1159 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1161 TL/02 Add +smtp_confirmation as a default logging option.
1163 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1164 Patch by Magnus Holmgren from 2007-02-20.
1166 TL/04 Bugzilla 1281 - Spec typo.
1167 Bugzilla 1283 - Spec typo.
1168 Bugzilla 1290 - Spec grammar fixes.
1170 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1172 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1174 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1175 mailing list by by Dmitry Isaikin.
1177 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1179 JH/02 Support "G" suffix to numbers in ${if comparisons.
1181 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1183 NM/01 Bugzilla 1197 - Spec typo
1184 Bugzilla 1196 - Spec examples corrections
1186 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1188 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1189 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1192 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1193 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1195 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1196 "acl {{name}{arg}...}", and optional args on acl condition
1199 JH/05 Permit multiple router/transport headers_add/remove lines.
1201 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1203 JH/07 Avoid using a waiting database for a single-message-only transport.
1204 Performance patch from Paul Fisher. Bugzilla 1262.
1206 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1209 JH/09 Add $headers_added variable, with content from use of ACL modifier
1210 add_header (but not yet added to the message). Bugzilla 199.
1212 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1213 Pulled from Bugzilla 817 by Wolfgang Breyha.
1215 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1217 (nb: this is the same fix as in Exim 4.80.1)
1219 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1222 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1225 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1227 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1228 advertise SMTP AUTH mechanism to us, instead of a generic
1229 protocol violation error. Also, make Exim more robust to bad
1230 data from the Dovecot auth socket.
1232 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1234 When a queue runner is handling a message, Exim first routes the
1235 recipient addresses, during which it prunes them based on the retry
1236 hints database. After that it attempts to deliver the message to
1237 any remaining recipients. It then updates the hints database using
1240 So if a recipient address works intermittently, it can get repeatedly
1241 deferred at routing time. The retry hints record remains fresh so the
1242 address never reaches the final cutoff time.
1244 This is a fairly common occurrence when a user is bumping up against
1245 their storage quota. Exim had some logic in its local delivery code
1246 to deal with this. However it did not apply to per-recipient defers
1247 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1249 This change adds a proper retry rule check during routing so that the
1250 final cutoff time is checked against the message's age. We only do
1251 this check if there is an address retry record and there is not a
1252 domain retry record; this implies that previous attempts to handle
1253 the address had the retry_use_local_parts option turned on. We use
1254 this as an approximation for the destination being like a local
1255 delivery, as in LMTP.
1257 I suspect this new check makes the old local delivery cutoff check
1258 redundant, but I have not verified this so I left the code in place.
1260 TF/02 Correct gecos expansion when From: is a prefix of the username.
1262 Test 0254 submits a message to Exim with the header
1266 When I ran the test suite under the user fanf2, Exim expanded
1267 the header to contain my full name, whereas it should have added
1268 a Resent-Sender: header. It erroneously treats any prefix of the
1269 username as equal to the username.
1271 This change corrects that bug.
1273 GF/01 DCC debug and logging tidyup
1274 Error conditions log to paniclog rather than rejectlog.
1275 Debug lines prefixed by "DCC: " to remove any ambiguity.
1277 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1279 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1280 Bug spotted by Jeremy Harris; was flawed since initial commit.
1281 Would have resulted in OCSP responses post-SNI triggering an Exim
1282 NULL dereference and crash.
1284 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1286 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1287 Bug detection, analysis and fix by Samuel Thibault.
1288 Bugzilla 1331, Debian bug #698092.
1290 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1292 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1293 Server implementation by Todd Lyons, client by JH.
1294 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1295 config variable "prdr_enable" controls whether the server
1296 advertises the facility. If the client requests PRDR a new
1297 acl_data_smtp_prdr ACL is called once for each recipient, after
1298 the body content is received and before the acl_smtp_data ACL.
1299 The client is controlled by both of: a hosts_try_prdr option
1300 on the smtp transport, and the server advertisement.
1301 Default client logging of deliveries and rejections involving
1302 PRDR are flagged with the string "PRDR".
1304 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1305 fclose(). Diagnosis by Todd Lyons.
1307 PP/17 Update configure.default to handle IPv6 localhost better.
1308 Patch by Alain Williams (plus minor tweaks).
1311 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1312 This is now consistent with GnuTLS, and is now documented: the
1313 previous undocumented portable approach to treating the option as
1314 unset was to force an expansion failure. That still works, and
1315 an empty string is now equivalent.
1317 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1318 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1319 not performing validation itself.
1321 PP/20 Added force_command boolean option to pipe transport.
1322 Patch from Nick Koston, of cPanel Inc.
1324 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1327 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1329 PP/21 Fix eximon continuous updating with timestamped log-files.
1330 Broken in a format-string cleanup in 4.80, missed when I repaired the
1331 other false fix of the same issue.
1332 Report and fix from Heiko Schlichting.
1335 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1336 Report from Prashanth Katuri.
1338 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1339 It's SecureTransport, so affects any MacOS clients which use the
1340 system-integrated TLS libraries, including email clients.
1342 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1343 using a MIME ACL for non-SMTP local injection.
1344 Report and assistance in diagnosis by Warren Baker.
1346 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1348 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1350 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1351 last id that failed so it may be referenced in subsequent ACL's.
1353 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1356 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1357 ldap library initialization, allowing self-signed CA's to be
1358 used. Also properly sets require_cert option later in code by
1359 using NULL (global ldap config) instead of ldap handle (per
1360 session). Bug diagnosis and testing by alxgomz.
1362 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1363 the src/util/ subdirectory.
1365 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1366 renamed to Transport Post Delivery Action by Jeremy Harris, as
1369 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1370 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1371 redis_servers = needs to be configured which will be used by the redis
1372 lookup. Patch from Warren Baker, of The Packet Hub.
1374 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1376 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1377 hostname or reverse DNS when processing a host list. Used suggestions
1378 from multiple comments on this bug.
1380 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1382 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1383 Missed a few lines, added it to make the runtest require no keyboard
1386 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1387 contains upper case chars. Make router use caseful_local_part.
1389 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1390 support when GnuTLS has been built with p11-kit.
1396 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1398 This, or similar/improved, will also be change PP/11 of 4.82.
1404 PP/01 Handle short writes when writing local log-files.
1405 In practice, only affects FreeBSD (8 onwards).
1406 Bugzilla 1053, with thanks to Dmitry Isaikin.
1408 NM/01 Bugzilla 949 - Documentation tweak
1410 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1413 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1415 PP/02 Implemented gsasl authenticator.
1417 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1419 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1420 `pkg-config foo` for cflags/libs.
1422 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1423 with rest of GSASL and with heimdal_gssapi.
1425 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1426 `pkg-config foo` for cflags/libs for the TLS implementation.
1428 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1429 properties get this fed in as external SSF. A number of robustness
1430 and debugging improvements to the cyrus_sasl authenticator.
1432 PP/08 cyrus_sasl server now expands the server_realm option.
1434 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1435 Patch by Jeremy Harris.
1437 PP/10 Added dbmjz lookup type.
1439 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1441 PP/12 MAIL args handles TAB as well as SP, for better interop with
1442 non-compliant senders.
1443 Analysis and variant patch by Todd Lyons.
1445 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1446 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1447 Patch from Dirk Mueller <dmueller@suse.com>
1449 PP/13 tls_peerdn now print-escaped for spool files.
1450 Observed some $tls_peerdn in wild which contained \n, which resulted
1451 in spool file corruption.
1453 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1454 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1455 or write after TLS renegotiation, which otherwise led to messages
1458 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1459 as a tracking header (ie: a signed header comes before the signature).
1460 Patch from Wolfgang Breyha.
1462 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1463 comma-sep list; embedded commas doubled.
1465 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1467 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1469 Report and patch from Dmitry Banschikov.
1471 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1472 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1473 failures. We appear to now support TLS1.1+ with Exim.
1475 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1476 lets Exim select keys and certificates based upon TLS SNI from client.
1477 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1478 before an outbound SMTP session. New log_selector, +tls_sni.
1480 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1481 NULL dereference. Report and patch from Alun Jones.
1483 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1484 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1485 Not seeing resolver debug output on NetBSD, but suspect this is a
1486 resolver implementation change.
1488 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1489 Left warnings. Added "eximon gdb" invocation mode.
1491 PP/21 Defaulting "accept_8bitmime" to true, not false.
1493 PP/22 Added -bw for inetd wait mode support.
1495 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1496 locate the relevant includes and libraries. Made this the default.
1498 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1499 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1501 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1502 This may cause build issues on older platforms.
1504 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1505 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1506 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1507 Added SNI support via GnuTLS too.
1508 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1510 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1512 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1513 Applied second patch from Janne, implementing suggestion to default
1514 multiple-strings-in-record handling to match SPF spec.
1516 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1518 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1519 read-only, out of scope).
1520 Patch from Wolfgang Breyha, report from Stuart Northfield.
1522 PP/29 Fix three issues highlighted by clang analyser static analysis.
1523 Only crash-plausible issue would require the Cambridge-specific
1524 iplookup router and a misconfiguration.
1525 Report from Marcin Mirosław.
1527 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1529 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1530 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1531 As part of this, removing so much warning spew let me fix some minor
1532 real issues in debug logging.
1534 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1535 assignment on my part. Fixed.
1537 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1538 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1539 Janne Snabb (who went above and beyond: thank you).
1541 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1542 string otherwise requires a connection and a bunch more work and it's
1543 relatively easy to get wrong. Should also expose TLS library linkage
1546 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1547 64-bit ${eval} (JH/03).
1549 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1550 GNU libc to support some of the 64-bit stuff, should not lead to
1551 conflicts. Defined before os.h is pulled in, so if a given platform
1552 needs to override this, it can.
1554 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1555 protection layer was required, which is not implemented.
1556 Bugzilla 1254, patch from Wolfgang Breyha.
1558 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1559 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1560 tls_dhparam take prime identifiers. Also unbreak combination of
1561 OpenSSL+DH_params+TLSSNI.
1563 PP/39 Disable SSLv2 by default in OpenSSL support.
1569 PP/01 Solaris build fix for Oracle's LDAP libraries.
1570 Bugzilla 1109, patch from Stephen Usher.
1572 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1574 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1577 TF/02 Fix a couple more cases where we did not log the error message
1578 when unlink() failed. See also change 4.74-TF/03.
1580 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1581 lock up or crash if it happened to be inside a call to libc when it
1582 got a SIGUSR1 from exiwhat.
1584 The SIGUSR1 handler appends the current process status to the process
1585 log which is later printed by exiwhat. It used to use the general
1586 purpose logging code to do this, but several functions it calls are
1587 not safe for signals.
1589 The new output code in the SIGUSR1 handler is specific to the process
1590 log, and simple enough that it's easy to inspect for signal safety.
1591 Removing some special cases also simplifies the general logging code.
1592 Removing the spurious timestamps from the process log simplifies
1595 TF/04 Improved ratelimit ACL condition.
1597 The /noupdate option has been deprecated in favour of /readonly which
1598 has clearer semantics. The /leaky, /strict, and /readonly update modes
1599 are mutually exclusive. The update mode is no longer included in the
1600 database key; it just determines when the database is updated. (This
1601 means that when you upgrade Exim will forget old rate measurements.)
1603 Exim now checks that the per_* options are used with an update mode that
1604 makes sense for the current ACL. For example, when Exim is processing a
1605 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1606 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1607 must specify per_mail/readonly. If you omit the update mode it defaults to
1608 /leaky where that makes sense (as before) or /readonly where required.
1610 The /noupdate option is now undocumented but still supported for
1611 backwards compatibility. It is equivalent to /readonly except that in
1612 ACLs where /readonly is required you may specify /leaky/noupdate or
1613 /strict/noupdate which are treated the same as /readonly.
1615 A useful new feature is the /count= option. This is a generalization
1616 of the per_byte option, so that you can measure the throughput of other
1617 aggregate values. For example, the per_byte option is now equivalent
1618 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1620 The per_rcpt option has been generalized using the /count= mechanism
1621 (though it's more complicated than the per_byte equivalence). When it is
1622 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1623 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1624 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1625 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1626 Note that using per_rcpt with a non-readonly update mode in more than
1627 one ACL will cause the recipients to be double-counted. (The per_mail
1628 and per_byte options don't have this problem.)
1630 The handling of very low rates has changed slightly. If the computed rate
1631 is less than the event's count (usually one) then this event is the first
1632 after a long gap. In this case the rate is set to the same as this event's
1633 count, so that the first message of a spam run is counted properly.
1635 The major new feature is a mechanism for counting the rate of unique
1636 events. The new per_addr option counts the number of different
1637 recipients that someone has sent messages to in the last time period. It
1638 behaves like per_rcpt if all the recipient addresses are different, but
1639 duplicate recipient addresses do not increase the measured rate. Like
1640 the /count= option this is a general mechanism, so the per_addr option
1641 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1642 example, measure the rate that a client uses different sender addresses
1643 with the options per_mail/unique=$sender_address. There are further
1644 details in the main documentation.
1646 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1648 TF/06 Removed a few PCRE remnants.
1650 TF/07 Automatically extract Exim's version number from tags in the git
1651 repository when doing development or release builds.
1653 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1654 Bugzilla 879. Patch from Paul Fisher.
1656 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1657 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1660 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1662 PP/05 Variable $av_failed, true if the AV scanner deferred.
1663 Bugzilla 1078. Patch from John Horne.
1665 PP/06 Stop make process more reliably on build failure.
1666 Bugzilla 1087. Patch from Heiko Schlittermann.
1668 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1669 Bugzilla 1089. Patch from Heiko Schlittermann.
1671 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1672 Bugzilla 1131. Patch from Holger Weiß.
1674 PP/09 Handle IPv6 addresses with SPF.
1675 Bugzilla 860. Patch from Wolfgang Breyha.
1677 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1679 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1682 PP/11 match_* no longer expand right-hand-side by default.
1683 New compile-time build option, EXPAND_LISTMATCH_RHS.
1684 New expansion conditions, "inlist", "inlisti".
1686 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1688 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1690 PP/14 fix log_write() format string regression from TF/03.
1691 Bugzilla 1152. Patch from Dmitry Isaikin.
1697 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1699 PP/02 Harmonised TLS library version reporting; only show if debugging.
1700 Layout now matches that introduced for other libraries in 4.74 PP/03.
1702 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1704 PP/04 New "dns_use_edns0" global option.
1706 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1709 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1710 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1712 TK/01 Updated PolarSSL code to 0.14.2.
1713 Bugzilla 1097. Patch from Andreas Metzler.
1715 PP/07 Catch divide-by-zero in ${eval:...}.
1716 Fixes bugzilla 1102.
1718 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1721 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1722 format-string attack -- SECURITY: remote arbitrary code execution.
1724 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1725 time unintentionally subject to list matching rules, letting the header
1726 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1727 arbitrary string expansion). This allowed for information disclosure.
1729 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1730 INT_MIN/-1 -- value coerced to INT_MAX.
1736 NM/01 Workaround for PCRE version dependency in version reporting
1739 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1740 This fixes portability to compilers other than gcc, notably
1741 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1743 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1744 makefiles for portability to HP-UX and POSIX correctness.
1746 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1747 Also via indented variable definition in the Makefile.
1748 (Debugging by Oliver Heesakkers).
1750 PP/02 Restore caching of spamd results with expanded spamd_address.
1751 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1753 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1754 Improves build reliability. Fix from: Frank Elsner
1756 NM/02 Fix wide character breakage in the rfc2047 coding
1757 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1759 NM/03 Allow underscore in dnslist lookups
1760 Fixes bug 1026. Patch from Graeme Fowler
1762 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1763 Code patches from Adam Ciarcinski of NetBSD.
1765 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1768 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1769 is logged, to avoid truncation. Patch from John Horne.
1771 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1772 Patch from Jakob Hirsch.
1774 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1775 SQL string expansion failure details.
1776 Patch from Andrey Oktyabrski.
1778 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1779 Patch from Simon Arlott.
1781 PP/09 New lookups functionality failed to compile on old gcc which rejects
1782 extern declarations in function scope.
1783 Patch from Oliver Fleischmann
1785 PP/10 Use sig_atomic_t for flags set from signal handlers.
1786 Check getgroups() return and improve debugging.
1787 Fixed developed for diagnosis in bug 927 (which turned out to be
1790 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1791 Patch from Mark Zealey.
1793 PP/12 Bugzilla 1056: Improved spamd server selection.
1794 Patch from Mark Zealey.
1796 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1797 Based on patch from Heiko Schlittermann.
1799 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1800 Patch from Uwe Doering, sign-off by Michael Haardt.
1802 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1803 variable declaration deep within a block. Bug and patch from
1806 PP/15 lookups-Makefile IRIX compatibility coercion.
1808 PP/16 Make DISABLE_DKIM build knob functional.
1810 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1811 Patch by Simon Arlott
1813 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1814 variable argument macros. Our copy now differs from upstream.
1820 TF/01 Failure to get a lock on a hints database can have serious
1821 consequences so log it to the panic log.
1823 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1824 controlled using the smtp_confirmation log selector.
1826 TF/03 Include the error message when we fail to unlink a spool file.
1828 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1829 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1830 for maintaining out-of-tree patches for some time.
1832 PP/01 Bugzilla 139: Documentation and portability issues.
1833 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1834 Handle per-OS dynamic-module compilation flags.
1836 PP/02 Let /dev/null have normal permissions.
1837 The 4.73 fixes were a little too stringent and complained about the
1838 permissions on /dev/null. Exempt it from some checks.
1839 Reported by Andreas M. Kirchwitz.
1841 PP/03 Report version information for many libraries, including
1842 Exim version information for dynamically loaded libraries. Created
1843 version.h, now support a version extension string for distributors
1844 who patch heavily. Dynamic module ABI change.
1846 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1847 privilege escalation vulnerability whereby the Exim run-time user
1848 can cause root to append content of the attacker's choosing to
1851 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1854 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1855 If dropping privileges for untrusted macros, we disabled normal logging
1856 on the basis that it would fail; for the Exim run-time user, this is not
1857 the case, and it resulted in successful deliveries going unlogged.
1858 Fixed. Reported by Andreas Metzler.
1864 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1865 only prepend for the Resent-* case. Fixes regression introduced in
1866 Exim 4.70 by NM/22 for Bugzilla 607.
1868 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1869 increasing numbers of administrators be bitten by this.
1871 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1873 PP/03 Bugzilla 994: added openssl_options main configuration option.
1875 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1877 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1879 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1881 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1882 without a peer certificate, leading to a segfault because of an
1883 assumption that peers always have certificates. Be a little more
1884 paranoid. Problem reported by Martin Tscholak.
1886 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1887 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1888 NB: ClamAV planning to remove STREAM in "middle of 2010".
1889 CL also introduces -bmalware, various -d+acl logging additions and
1890 more caution in buffer sizes.
1892 PP/09 Implemented reverse_ip expansion operator.
1894 PP/10 Bugzilla 937: provide a "debug" ACL control.
1896 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1898 PP/12 Bugzilla 973: Implement --version.
1900 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1902 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1904 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1906 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1907 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1908 ignore trailing whitespace.
1910 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1912 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1915 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1916 Notification from Dr Andrew Aitchison.
1918 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1919 ExtendedDetectionInfo response format.
1920 Notification from John Horne.
1922 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1925 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1926 XSL and documented dependency on system catalogs, with examples of how
1929 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1932 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1933 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1934 configuration file which is writeable by the Exim user or group.
1936 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1937 of configuration files to cover files specified with the -C option if
1938 they are going to be used with root privileges, not just the default
1941 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1942 option (effectively making it always true).
1944 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1945 files to be used while preserving root privileges.
1947 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1948 that rogue child processes cannot use them.
1950 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1951 run-time user, instead of root.
1953 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1954 Exim run-time user without dropping privileges.
1956 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1957 result string, instead of calling string_vformat() twice with the same
1960 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1961 for other users. Others should always drop root privileges if they use
1962 -C on the command line, even for a whitelisted configure file.
1964 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1966 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1972 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1973 $data_path, and $header_path variables; fixed documentation bugs and
1976 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1977 exipick to access non-standard spools, including the "frozen" queue
1980 NM/01 Bugzilla 965: Support mysql stored procedures.
1981 Patch from Alain Williams
1983 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1985 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1986 Patch from Andreas Metzler
1988 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1989 Patch from Kirill Miazine
1991 NM/05 Bugzilla 671: Added umask to procmail example.
1993 JJ/03 installed exipick 20100323.0, fixing doc bug
1995 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1996 directory. Notification and patch from Dan Rosenberg.
1998 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2000 TK/02 Improve log output when DKIM signing operation fails.
2002 MH/01 Treat the transport option dkim_domain as a colon separated
2003 list, not as a single string, and sign the message with each element,
2004 omitting multiple occurences of the same signer.
2006 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2007 Bugzilla 985, 986. Patch by Simon Arlott
2009 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2010 Patch by Simon Arlott
2012 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2013 MBX locking. Notification from Dan Rosenberg.
2019 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2021 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2023 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2025 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2027 NM/04 Bugzilla 915: Flag broken perl installation during build.
2033 TK/01 Added patch by Johannes Berg that expands the main option
2034 "spamd_address" if it starts with a dollar sign.
2036 TK/02 Write list of recipients to X-Envelope-Sender header when building
2037 the mbox-format spool file for content scanning (suggested by Jakob
2040 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2041 (http://www.dcc-servers.net/) support via dccifd. Activated by
2042 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2044 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2045 by Mark Daniel Reidel <mr@df.eu>.
2047 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2048 When building exim an external PCRE library is now needed -
2049 PCRE is a system library on the majority of modern systems.
2050 See entry on PCRE_LIBS in EDITME file.
2052 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2053 conversation. Added nologin parameter to request.
2054 Patch contributed by Kirill Miazine.
2056 TF/01 Do not log submission mode rewrites if they do not change the address.
2058 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2060 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2061 log files in place. Contributed by Roberto Lima.
2063 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2065 TF/03 Bugzilla 615: When checking the local_parts router precondition
2066 after a local_part_suffix or local_part_prefix option, Exim now
2067 does not use the address's named list lookup cache, since this
2068 contains cached lookups for the whole local part.
2070 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2071 Robert Millan. Documentation is in experimental-spec.txt.
2073 TF/04 Bugzilla 668: Fix parallel build (make -j).
2075 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2077 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2078 Patch provided by Jan Srzednicki.
2080 TF/05 Leading white space used to be stripped from $spam_report which
2081 wrecked the formatting. Now it is preserved.
2083 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2084 that they are available at delivery time.
2086 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2088 TF/08 TLS error reporting now respects the incoming_interface and
2089 incoming_port log selectors.
2091 TF/09 Produce a more useful error message if an SMTP transport's hosts
2092 setting expands to an empty string.
2094 NM/06 Bugzilla 744: EXPN did not work under TLS.
2095 Patch provided by Phil Pennock.
2097 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2098 Patch provided by Richard Godbee.
2100 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2101 acl_smtp_notquit, added index entry.
2103 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2104 Patch provided by Eugene Bujak.
2106 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2107 accept(). Patch provided by Maxim Dounin.
2109 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2110 Patch provided by Phil Pennock.
2112 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2114 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2115 Patch provided by Brad "anomie" Jorsch.
2117 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2118 Patch provided by Dean Brooks.
2120 TK/05 Add native DKIM support (does not depend on external libraries).
2122 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2123 Patch provided by Graeme Fowler.
2125 NM/16 Bugzilla 851: Documentation example syntax fix.
2127 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2129 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2132 NM/19 Bugzilla 745: TLS version reporting.
2133 Patch provided by Phil Pennock.
2135 NM/20 Bugzilla 167: bool: condition support.
2136 Patch provided by Phil Pennock.
2138 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2139 clients. Patch provided by Phil Pennock.
2141 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2142 Patch provided by Brad "anomie" Jorsch.
2144 NM/23 Bugzilla 687: Fix misparses in eximstats.
2145 Patch provided by Heiko Schlittermann.
2147 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2148 Patch provided by Heiko Schlittermann.
2150 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2151 plus update to original patch.
2153 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2155 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2156 Patch provided by David Brownlee.
2158 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2160 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2162 NM/30 Bugzilla 888: TLS documentation bugfixes.
2164 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2166 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2167 Unlike the original bugzilla I have changed all shell scripts in src tree.
2169 NM/33 Bugzilla 898: Transport filter timeout fix.
2170 Patch by Todd Rinaldo.
2172 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2173 Patch by Serge Demonchaux.
2175 NM/35 Bugzilla 39: Base64 decode bug fixes.
2176 Patch by Jakob Hirsch.
2178 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2180 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2182 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2184 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2190 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2191 ALT-N's libdkim that I have put here:
2192 http://duncanthrax.net/exim-experimental/
2194 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2195 were called 'true' and it seems that C99 defines that as a reserved
2196 keyword to be used with 'bool' variable types. That means you could
2197 not include C99-style headers which use bools without triggering
2198 build errors in sieve.c.
2200 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2201 as mailq or other aliases. Changed the --help handling significantly
2202 to do whats expected. exim_usage() emits usage/help information.
2204 SC/01 Added the -bylocaldomain option to eximstats.
2206 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2208 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2210 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2216 PH/01 Another patch from the Sieve maintainer.
2218 PH/02 When an IPv6 address is converted to a string for single-key lookup
2219 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2220 dots are used instead of colons so that keys in lsearch files need not
2221 contain colons. This was done some time before quoting was made available
2222 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2223 (notated using the quote facility) so as to distinguish them from IPv4
2224 keys. This meant that lookups for IP addresses in host lists did not work
2225 for iplsearch lookups.
2227 This has been fixed by arranging for IPv6 addresses to be expressed with
2228 colons if the lookup type is iplsearch. This is not incompatible, because
2229 previously such lookups could never work.
2231 The situation is now rather anomalous, since one *can* have colons in
2232 ordinary lsearch keys. However, making the change in all cases is
2233 incompatible and would probably break a number of configurations.
2235 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2238 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2239 conversion specification without a maximum field width, thereby enabling
2240 a rogue spamd server to cause a buffer overflow. While nobody in their
2241 right mind would setup Exim to query an untrusted spamd server, an
2242 attacker that gains access to a server running spamd could potentially
2243 exploit this vulnerability to run arbitrary code as the Exim user.
2245 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2246 $primary_hostname instead of what libspf2 thinks the hosts name is.
2248 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2249 a directory entry by the name of the lookup key. Previously, if a
2250 symlink pointed to a non-existing file or a file in a directory that
2251 Exim lacked permissions to read, a lookup for a key matching that
2252 symlink would fail. Now it is enough that a matching directory entry
2253 exists, symlink or not. (Bugzilla 503.)
2255 PH/03 The body_linecount and body_zerocount variables are now exported in the
2258 PH/04 Added the $dnslist_matched variable.
2260 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2261 This means they are set thereafter only if the connection becomes
2264 PH/06 Added the client_condition to authenticators so that some can be skipped
2265 by clients under certain conditions.
2267 PH/07 The error message for a badly-placed control=no_multiline_responses left
2268 "_responses" off the end of the name.
2270 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2272 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2273 (without spaces) instead of just copying the configuration text.
2275 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2277 PH/11 Added $max_received_linelength.
2279 PH/12 Added +ignore_defer and +include_defer to host lists.
2281 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2282 way in which PCRE > 7.0 is built.
2284 PH/14 Implemented queue_only_load_latch.
2286 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2287 MAIL command. The effect was to mangle the value on 64-bit systems.
2289 PH/16 Another patch from the Sieve maintainer.
2291 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2293 PH/18 If a system quota error occurred while trying to create the file for
2294 a maildir delivery, the message "Mailbox is full" was not appended to the
2295 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2296 only to a quota excession during the actual writing of the file.
2298 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2299 characters?) which causes problems in log lines. The DN values are now
2300 passed through string_printing() before being added to log lines.
2302 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2303 and InterBase are left for another time.)
2305 PH/21 Added message_body_newlines option.
2307 PH/22 Guard against possible overflow in moan_check_errorcopy().
2309 PH/23 POSIX allows open() to be a macro; guard against that.
2311 PH/24 If the recipient of an error message contained an @ in the local part
2312 (suitably quoted, of course), incorrect values were put in $domain and
2313 $local_part during the evaluation of errors_copy.
2319 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2320 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2323 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2324 issue a MAIL command.
2326 PH/02 In an ACL statement such as
2328 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2330 if a client was not listed at all, or was listed with a value other than
2331 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2332 the condition was not true (as it should be), so access was not denied.
2333 The bug was that the ! inversion was incorrectly passed on to the second
2334 item. This has been fixed.
2336 PH/03 Added additional dnslists conditions == and =& which are different from
2337 = and & when the dns lookup returns more than one IP address.
2339 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2340 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2342 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2343 FSYNC, which compiles an option called disable_fsync that allows for
2344 bypassing fsync(). The documentation is heavily laced with warnings.
2346 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2348 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2349 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2350 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2351 including adding "make clean"; (3) Added -fPIC when compiling the test
2352 dynamically loaded module, to get rid of a warning.
2354 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2355 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2356 = 0s. The bug is otherwise harmless.
2358 PH/07 There was a bug in the dovecot authenticator such that the value of
2359 $auth1 could be overwritten, and so not correctly preserved, after a
2360 successful authentication. This usually meant that the value preserved by
2361 the server_setid option was incorrect.
2363 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2365 PH/09 Installed PCRE release 7.0.
2367 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2368 run for batched SMTP input. It is now run at the start of every message
2369 in the batch. While fixing this I discovered that the process information
2370 (output by running exiwhat) was not always getting set for -bs and -bS
2371 input. This is fixed, and it now also says "batched" for BSMTP.
2373 PH/11 Added control=no_pipelining.
2375 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2376 patch, slightly modified), and move the expansion of helo_data till after
2377 the connection is made in the smtp transport (so it can use these
2380 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2382 PH/14 Added log_selector = +pid.
2384 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2386 PH/16 Add ${if forany and ${if forall.
2388 PH/17 Added dsn_from option to vary the From: line in DSNs.
2390 PH/18 Flush SMTP output before performing a callout, unless control =
2391 no_callout_flush is set.
2393 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2394 was true (the default) a successful delivery failed to delete the retry
2395 item, thus causing premature timeout of the address. The bug is now
2398 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2400 PH/21 Long custom messages for fakedefer and fakereject are now split up
2401 into multiline responses in the same way that messages for "deny" and
2402 other ACL rejections are.
2404 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2405 with slight modification.
2407 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2408 draft, changing the syntax and factoring some duplicate code".
2410 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2411 for deliveries of the second and subsequent messages over the same SMTP
2414 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2415 ${reduce, with only minor "tidies".
2417 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2419 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2420 expansion side effects.
2422 PH/27 When a message times out after an over-quota error from an Exim-imposed
2423 quota, the bounce message says "mailbox is full". This message was not
2424 being given when it was a system quota that was exceeded. It now should
2427 MH/03 Made $recipients available in local_scan(). local_scan() already has
2428 better access to the recipient list through recipients_list[], but
2429 $recipients can be useful in postmaster-provided expansion strings.
2431 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2432 in the case of a MAIL command with additional options following the
2433 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2434 were accidentally chopped off.
2436 PH/29 SMTP synchronization checks are implemented when a command is read -
2437 there is a check that no more input is waiting when there shouldn't be
2438 any. However, for some commands, a delay in an ACL can mean that it is
2439 some time before the response is written. In this time, more input might
2440 arrive, invalidly. So now there are extra checks after an ACL has run for
2441 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2442 pipelining has not been advertised.
2444 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2446 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2447 returned in the SMTP response when smtp_return_error_details was false.
2448 This has been fixed.
2450 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2451 instead of the C I/O that was originally supplied, because problems were
2452 reported on Solaris.
2454 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2455 Exim which did not show up earlier: it was assuming that a call to
2456 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2457 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2458 was a macro that became an assignment, so it seemed to work. This has
2459 changed to a proper function call with a void return, hence the compile
2460 error. Exim's code has been fixed.
2462 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2465 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2467 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2469 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2470 because it turns out to be too big or there is a timeout) there is no
2471 "Completed" line in the log. When some messages of this type were
2472 selected by exigrep, they were listed as "not completed". Others were
2473 picked up by some special patterns. I have improved the selection
2474 criteria to be more general.
2476 PH/38 The host_find_failed option in the manualroute router can now be set
2477 to "ignore", to completely ignore a host whose IP address cannot be
2478 found. If all hosts are ignored, the behaviour is controlled by the new
2479 host_all_ignored option.
2481 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2482 homing or because of multiple MX records with /mx) generated more than
2483 one IP address, and the following item turned out to be the local host,
2484 all the secondary addresses of the first item were incorrectly removed
2485 from the list, along with the local host and any following hosts (which
2486 is what is supposed to happen).
2488 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2489 whoever called Exim into the -H file. In the case of the daemon it was
2490 behaving confusingly. When first started, it used values for whoever
2491 started the daemon, but after a SIGHUP it used the Exim user (because it
2492 calls itself on a restart). I have changed the code so that it now always
2495 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2496 message are rejected with the same error (e.g. no authentication or bad
2497 sender address), and a DATA command is nevertheless sent (as can happen
2498 with PIPELINING or a stupid MUA), the error message that was given to the
2499 RCPT commands is included in the rejection of the DATA command. This is
2500 intended to be helpful for MUAs that show only the final error to their
2503 PH/42 Another patch from the Sieve maintainer.
2505 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2506 Eximstats - Fixed some broken HTML links and added missing column headers
2508 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2509 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2511 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2513 PH/43 Yet another patch from the Sieve maintainer.
2515 PH/44 I found a way to check for a TCP/IP connection going away before sending
2516 the response to the final '.' that terminates a message, but only in the
2517 case where the client has not sent further data following the '.'
2518 (unfortunately, this is allowed). However, in many cases there won't be
2519 any further data because there won't be any more messages to send. A call
2520 to select() can be used: if it shows that the input is "ready", there is
2521 either input waiting, or the socket has been closed. An attempt to read
2522 the next input character can distinguish the two cases. Previously, Exim
2523 would have sent an OK response which the client would never have see.
2524 This could lead to message repetition. This fix should cure that, at
2525 least in a lot of common cases.
2527 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2528 advertised in response to EHLO.
2534 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2535 fixed by 4.65/MH/01 (is this a record?) are fixed:
2537 (i) An empty string was always treated as zero by the numeric comparison
2538 operators. This behaviour has been restored.
2540 (ii) It is documented that the numeric comparison operators always treat
2541 their arguments as decimal numbers. This was broken in that numbers
2542 starting with 0 were being interpreted as octal.
2544 While fixing these problems I realized that there was another issue that
2545 hadn't been noticed. Values of message_size_limit (both the global option
2546 and the transport option) were treated as octal if they started with 0.
2547 The documentation was vague. These values are now always treated as
2548 decimal, and I will make that clear in the documentation.
2554 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2555 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2558 MH/01 Don't check that the operands of numeric comparison operators are
2559 integers when their expansion is in "skipping" mode (fixes bug
2560 introduced by 4.64-PH/07).
2562 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2563 child addresses, Exim now panics and dies. Previously, because the count
2564 is held in a short int, deliveries were likely to be lost. As such a
2565 large number of recipients for a single message is ridiculous
2566 (performance will be very, very poor), I have chosen to impose a limit
2567 rather than extend the field.
2573 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2574 leftover -K file (the existence of which was triggered by #402).
2575 While we were at it, introduced process PID as part of the -K
2576 filename. This should rule out race conditions when creating
2579 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2580 processing considerably. Previous code took too long for large mails,
2581 triggering a timeout which in turn triggers #401.
2583 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2584 in the DK code in transports.c. sendfile() is not really portable,
2585 hence the _LINUX specificness.
2587 TF/01 In the add_headers option to the mail command in an Exim filter,
2588 there was a bug that Exim would claim a syntax error in any
2589 header after the first one which had an odd number of characters
2592 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2593 callout verification, Exim cached a "reject" for the entire domain. This
2594 is correct for most verifications, but it is not correct for a recipient
2595 verification with use_sender or use_postmaster set, because in that case
2596 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2597 case of MAIL FROM:<> rejection from other early rejections (e.g.
2598 rejection of HELO). When verifying a recipient using a non-null MAIL
2599 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2600 Whatever the result of the callout, the value of the domain cache is
2601 left unchanged (for any other kind of callout, getting as far as trying
2602 RCPT means that the domain itself is ok).
2604 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2607 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2608 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2609 session, the daemon ignores ECONNECT errors and logs others; it now
2610 ignores EPIPE as well.
2612 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2613 (quoted-printable decoding).
2615 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2616 later the small subsequent patch to fix an introduced bug.
2618 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2620 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2622 PH/08 An error is now given if message_size_limit is specified negative.
2624 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2625 to be given (somewhat) arbitrary names.
2627 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2630 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2631 miscellaneous code fixes
2633 PH/10 Added the log_reject_target ACL modifier to specify where to log
2636 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2637 hostname. This is wrong, because it relates to the incoming message (and
2638 probably the interface on which it is arriving) and not to the outgoing
2639 callout (which could be using a different interface). This has been
2640 changed to use the value of the helo_data option from the smtp transport
2641 instead - this is what is used when a message is actually being sent. If
2642 there is no remote transport (possible with a router that sets up host
2643 addresses), $smtp_active_hostname is used.
2645 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2646 tweaks were necessary in order to get it to work (see also 21 below):
2647 (a) The code assumed that strncpy() returns a negative number on buffer
2648 overflow, which isn't the case. Replaced with Exim's string_format()
2650 (b) There were several signed/unsigned issues. I just did the minimum
2651 hacking in of casts. There is scope for a larger refactoring.
2652 (c) The code used strcasecmp() which is not a standard C function.
2653 Replaced with Exim's strcmpic() function.
2654 (d) The code set only $1; it now sets $auth1 as well.
2655 (e) A simple test gave the error "authentication client didn't specify
2656 service in request". It would seem that Dovecot has changed its
2657 interface. Fortunately there's a specification; I followed it and
2658 changed what the client sends and it appears to be working now.
2660 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2663 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2664 address A is aliased to B and C, where B exists and C does not. Without
2665 -v the output is "A verified" because verification stops after a
2666 successful redirection if more than one address is generated. However,
2667 with -v the child addresses are also verified. Exim was outputting "A
2668 failed to verify" and then showing the successful verification for C,
2669 with its parentage. It now outputs "B failed to verify", showing B's
2670 parentage before showing the successful verification of C.
2672 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2673 look up a TXT record in a specific list after matching in a combined
2676 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2677 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2678 they consult the DNS. I had assumed they would set it the way they
2679 wanted; and indeed my experiments on Linux seem to show that in some
2680 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2681 To be on the safe side, however, I have now made the interface to
2682 host_find_byname() similar to host_find_bydns(), with an argument
2683 containing the DNS resolver options. The host_find_byname() function now
2684 sets these options at its start, just as host_find_bydns() does. The smtp
2685 transport options dns_qualify_single and dns_search_parents are passed to
2686 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2687 of host_find_byname() use the default settings of RES_DEFNAMES
2688 (qualify_single) but not RES_DNSRCH (search_parents).
2690 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2691 spool_read_header() do less string testing, by means of a preliminary
2692 switch on the second character of optional "-foo" lines. (This is
2693 overdue, caused by the large number of possibilities that now exist.
2694 Originally there were few.) While I was there, I also converted the
2695 str(n)cmp tests so they don't re-test the leading "-" and the first
2696 character, in the hope this might squeeze out yet more improvement.
2698 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2699 flag allowing group syntax was set by the header_syntax check but not
2700 turned off, possible causing trouble later; (2) The flag was not being
2701 set at all for the header_verify test, causing "group"-style headers to
2702 be rejected. I have now set it in this case, and also caused header_
2703 verify to ignore an empty address taken from a group. While doing this, I
2704 came across some other cases where the code for allowing group syntax
2705 while scanning a header line wasn't quite right (mostly, not resetting
2706 the flag correctly in the right place). These bugs could have caused
2707 trouble for malformed header lines. I hope it is now all correct.
2709 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2710 with the "reply" argument non-NULL. The code, however (which originally
2711 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2712 but it didn't always do it. This confused somebody who was copying the
2713 code for some other use. I have removed all the tests.
2715 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2716 feature that was used to support insecure browsers during the U.S. crypto
2717 embargo. It requires special client support, and Exim is probably the
2718 only MTA that supported it -- and would never use it because real RSA is
2719 always available. This code has been removed, because it had the bad
2720 effect of slowing Exim down by computing (never used) parameters for the
2721 RSA_EXPORT functionality.
2723 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2724 authenticator to fail if there's a tab character in the incoming data
2725 (there should never be unless someone is messing about, as it's supposed
2726 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2727 if the connection is using TLS or if the remote IP is the same as the
2728 local IP, and the "valid-client-cert option" if a client certificate has
2731 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2732 authenticators. This can be used for authorization after authentication
2733 succeeds. (In the case of plaintext, it servers for both authentication
2736 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2737 if any retry times were supplied.
2739 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2740 connection, where there is no client IP address to check. In this
2741 situation, the verify now always succeeds.
2743 PH/25 Applied John Jetmore's -Mset patch.
2745 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2747 PH/27 In a string expansion for a processed (not raw) header when multiple
2748 headers of the same name were present, leading whitespace was being
2749 removed from all of them, but trailing whitespace was being removed only
2750 from the last one. Now trailing whitespace is removed from each header
2751 before concatenation. Completely empty headers in a concatenation (as
2752 before) are ignored.
2754 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2755 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2757 PH/29 [Removed. This was a change that I later backed out, and forgot to
2758 correct the ChangeLog entry (that I had efficiently created) before
2759 committing the later change.]
2761 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2762 address errors (4xx response to RCPT) over the same connection as other
2763 messages routed to the same hosts. Such deliveries are always "forced",
2764 so retry times are not inspected. This resulted in far too many retries
2765 for the affected addresses. The effect occurred only when there were more
2766 hosts than the hosts_max_try setting in the smtp transport when it had
2767 the 4xx errors. Those hosts that it had tried were not added to the list
2768 of hosts for which the message was waiting, so if all were tried, there
2769 was no problem. Two fixes have been applied:
2771 (i) If there are any address or message errors in an SMTP delivery, none
2772 of the hosts (tried or untried) are now added to the list of hosts
2773 for which the message is waiting, so the message should not be a
2774 candidate for sending over the same connection that was used for a
2775 successful delivery of some other message. This seems entirely
2776 reasonable: after all the message is NOT "waiting for some host".
2777 This is so "obvious" that I'm not sure why it wasn't done
2778 previously. Hope I haven't missed anything, but it can't do any
2779 harm, as the worst effect is to miss an optimization.
2781 (ii) If, despite (i), such a delivery is accidentally attempted, the
2782 routing retry time is respected, so at least it doesn't keep
2783 hammering the server.
2785 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2786 in ${readsocket because some servers need this prod.
2788 PH/32 Added some extra debug output when updating a wait-xxx database.
2790 PH/33 The hint "could be header name not terminated by colon", which has been
2791 given for certain expansion errors for a long time, was not being given
2792 for the ${if def:h_colon_omitted{... case.
2794 PH/34 The spec says: "With one important exception, whenever a domain list is
2795 being scanned, $domain contains the subject domain." There was at least
2796 one case where this was not true.
2798 PH/35 The error "getsockname() failed: connection reset by peer" was being
2799 written to the panic log as well as the main log, but it isn't really
2800 panic-worthy as it just means the connection died rather early on. I have
2801 removed the panic log writing for the ECONNRESET error when getsockname()
2804 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2805 runs only) independently of the message's sender address. This meant
2806 that, if the 4xx error was in fact related to the sender, a different
2807 message to the same recipient with a different sender could confuse
2808 things. In particular, this can happen when sending to a greylisting
2809 server, but other circumstances could also provoke similar problems.
2810 I have changed the default so that the retry time for these errors is now
2811 based a combination of the sender and recipient addresses. This change
2812 can be overridden by setting address_retry_include_sender=false in the
2815 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2816 remote server are returned as part of bounce messages. This was not
2817 happening for LMTP over a pipe (the lmtp transport), but now it is the
2818 same for both kinds of LMTP.
2820 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2821 in header lines that were in fact CNAMEs. This is no longer the case.
2823 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2824 and queue runs started by the daemon processed all messages. This has
2825 been fixed so that -R and -S can now usefully be given with -q<time>.
2827 PH/40 Import PCRE release 6.7 (fixes some bugs).
2829 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2831 PH/42 Give an error if -q is specified more than once.
2833 PH/43 Renamed the variables $interface_address and $interface_port as
2834 $received_ip_address and $received_port, to make it clear that these
2835 values apply to message reception, and not to the outgoing interface when
2836 a message is delivered. (The old names remain recognized, of course.)
2838 PH/44 There was no timeout on the connect() call when using a Unix domain
2839 socket in the ${readsocket expansion. There now is.
2841 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2842 be meaningful with "accept".
2844 SC/01 Eximstats V1.43
2845 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2847 SC/02 Eximstats V1.44
2848 Use a glob alias rather than an array ref in the generated
2849 parser. This improves both readability and performance.
2851 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2852 Collect SpamAssassin and rejection statistics.
2853 Don't display local sender or destination tables unless
2854 there is data to show.
2855 Added average volumes into the top table text output.
2857 SC/04 Eximstats V1.46
2858 Collect data on the number of addresses (recipients)
2859 as well as the number of messages.
2861 SC/05 Eximstats V1.47
2862 Added 'Message too big' to the list of mail rejection
2863 reasons (thanks to Marco Gaiarin).
2865 SC/06 Eximstats V1.48
2866 Mainlog lines which have GMT offsets and are too short to
2867 have a flag are now skipped.
2869 SC/07 Eximstats V1.49 (Alain Williams)
2870 Added the -emptyok flag.
2872 SC/08 Eximstats V1.50
2873 Fixes for obtaining the IP address from reject messages.
2875 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2876 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2877 whitespace changes from 4.64-PH/27
2879 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2882 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2883 are found, allow negative numbers in numeric criteria)
2885 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2887 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2888 to match changes made in 4.64-PH/43
2890 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2892 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2893 "domain" is actually an IP address, and force a failure. This locks out
2894 those revolvers/nameservers that support "A-for-A" lookups, in
2895 contravention of the specifications.
2897 PH/48 When a host name was looked up from an IP address, and the subsequent
2898 forward lookup of the name timed out, the host name was left in
2899 $sender_host_name, contrary to the specification.
2901 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2902 restricted to single-key lookups, Exim was not diagnosing an error if
2903 * or *@ was used with a query-style lookup.
2905 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2907 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2908 long ago, but noone interested enough thought of it. Let's just say that
2909 the "1.1" means that there are some new functions that weren't there at
2910 some point in the past.
2912 PH/51 Error processing for expansion failure of helo_data from an smtp
2913 transport during callout processing was broken.
2915 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2916 tested/used via the -bh/-bhc/-bs options.
2918 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2919 bug, fixed in subsequent PCRE releases).
2921 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2922 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2924 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2930 SC/01 Use a glob alias rather than an array ref in eximstats generated
2931 parser. This improves both readability and performance.
2933 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2934 Don't display local sender or destination tables in eximstats unless
2935 there is data to show.
2936 Added average volumes into the eximstats top table text output.
2938 SC/03 Collect data on the number of addresses (recipients) as well
2939 as the number of messages in eximstats.
2941 TF/01 Correct an error in the documentation for the redirect router. Exim
2942 does (usually) call initgroups() when daemonizing.
2944 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2945 with consistent privilege compared to when running as a daemon.
2947 TF/03 Note in the spec that $authenticated_id is not set for local
2948 submissions from trusted users.
2950 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2951 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2953 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2954 by adding some example configuration directives to the default
2955 configuration file. A little bit of work is required to uncomment the
2956 directives and define how usernames and passwords are checked, but
2957 there is now a framework to start from.
2959 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2960 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2961 without this. I don't know how relevant this is to other LDAP libraries.
2963 PH/02 Add the verb name to the "unknown ACL verb" error.
2965 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2967 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2969 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2970 directory not expanded when it should be if an expanded home directory
2971 was set for the address (which is overridden by the transport).
2973 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2976 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2977 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2978 because it is too late at that time, and has no effect.
2980 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2981 security issue with \' (bugzilla #107). I could not use the
2982 PQescapeStringConn() function, because it needs a PGconn value as one of
2985 PH/08 When testing addresses using -bt, indicate those final addresses that
2986 are duplicates that would not cause an additional delivery. At least one
2987 person was confused, thinking that -bt output corresponded to deliveries.
2988 (Suppressing duplicates isn't a good idea as you lose the information
2989 about possibly different redirections that led to the duplicates.)
2991 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2992 systems where poll() doesn't work, in particular OS X.
2994 PH/10 Added more information to debugging output for retry time not reached.
2996 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2997 operations in malware.c.
2999 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3002 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3003 syslog_duplication set false, log lines that would normally be written
3004 both the the main log and to the reject log were not written to syslog at
3007 PH/14 In the default configuration, change the use of "message" in ACL warn
3008 statements to "add_header".
3010 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3011 not followed by a command (e.g. "seen endif").
3013 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3014 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3017 PH/17 Added extra conditions to the default value of delay_warning_condition
3021 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3022 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3023 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3026 The Auto-Submitted: and various List- headers are standardised, whereas I
3027 don't think Precedence: ever was.
3029 PH/18 Refactored debugging code in route_finduser() to show more information,
3030 in particular, the error code if getpwnam() issues one.
3032 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3033 This is apparently needed in addition to the PH/07 change above to avoid
3034 any possible encoding problems.
3036 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3037 but not after initializing Perl.
3039 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3040 output them only if debugging. By default they are written stderr,
3041 apparently, which is not desirable.
3043 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3046 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3049 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3051 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3052 authenticated or an ident call has been made. Suppress the default
3053 values for $authenticated_id and $authenticated_sender (but permit -oMai
3054 and -oMas) when testing with -bh.
3056 PH/24 Re-jigged the order of the tests in the default configuration so that the
3057 tests for valid domains and recipients precede the DNS black list and CSA
3058 tests, on the grounds that those ones are more expensive.
3060 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3061 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3062 This bug exists in every version of Exim that I still have, right back to
3065 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3066 However, an attempt to turn on case-sensitivity in a regex key by
3067 including (?-i) didn't work because the subject string was already
3068 lowercased, and the effects were non-intuitive. It turns out that a
3069 one-line patch can be used to allow (?-i) to work as expected.
3075 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3076 other effects) broke the use of negated acl sub-conditions.
3078 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3081 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3082 "Deny" causes Exim to reject the incoming connection with a 554 error.
3083 Unfortunately, if there is a major crisis, such as a disk failure,
3084 tcp-wrappers gives "deny", whereas what one would like would be some
3085 kind of temporary error. A kludge has been added to help with this.
3086 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3087 554 error is used if errno is still zero or contains ENOENT (which occurs
3088 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3091 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3093 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3094 errors. Otherwise a message that provokes a temporary error (when other
3095 messages do not) can cause a whole host to time out.
3097 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3098 addresses were routed directly to files or pipes from a redirect router.
3099 File deliveries just didn't batch; pipe deliveries might have suffered
3102 PH/06 A failure to get a lock for a hints database would erroneously always say
3103 "Failed to get write lock", even when it was really a read lock.
3105 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3106 of 0600. This has been changed to use the value of the lockfile_mode
3107 option (which defaults to 0600).
3109 PH/08 Applied small patch from the Sieve maintainer.
3111 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3112 folder from quota calculations, a direct delivery into this folder messed
3113 up the contents of the maildirsize file. This was because the regex was
3114 used only to exclude .Trash (or whatever) when the size of the mailbox
3115 was calculated. There was no check that a delivery was happening into an
3116 excluded directory. This bug has been fixed by ignoring all quota
3117 processing for deliveries into excluded directories.
3119 PH/10 Added the maildirfolder_create_regex option to appendfile.
3125 PH/01 The code for finding all the local interface addresses on a FreeBSD
3126 system running IPv6 was broken. This may well have applied to all BSD
3127 systems, as well as to others that have similar system calls. The broken
3128 code found IPv4 interfaces correctly, but gave incorrect values for the
3129 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3130 that it would not match correctly against @[] and not recognize the IPv6
3133 PH/02 The ipliteral router was not recognizing addresses of the form user@
3134 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3136 PH/03 Added disable_ipv6.
3138 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3139 decoded form, because it is most often used to construct To: headers
3140 lines in autoreplies, and the decoded form may well be syntactically
3141 invalid. However, $reply_address has leading white space removed, and all
3142 newlines turned into spaces so that the autoreply transport does not
3145 PH/05 If group was specified without a user on a router, and no group or user
3146 was specified on a transport, the group from the router was ignored.
3148 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3149 for visible compile-time settings that can be used to change these
3150 numbers, for those that want even more. Backwards compatibility with old
3151 spool files has been maintained. However, going back to a previous Exim
3152 release will lost any variables that are in spool files.
3154 PH/07 Two small changes when running in the test harness: increase delay when
3155 passing a TCP/IP connection to a new process, in case the original
3156 process has to generate a bounce, and remove special handling of
3157 127.0.0.2 (sic), which is no longer necessary.
3159 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3160 be the same on different OS.
3162 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3165 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3166 whether --show-vars was specified or not
3168 JJ/02 exipick: Added support for new ACL variable spool format introduced
3171 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3172 syntactically invalid From: or Reply-to: line, and a filter used this to
3173 generate an autoreply, and therefore failed to obtain an address for the
3174 autoreply, Exim could try to deliver to a non-existent relative file
3175 name, causing unrelated and misleading errors. What now happens is that
3176 it logs this as a hard delivery error, but does not attempt to create a
3179 PH/11 The exinext utility has a -C option for testing purposes, but although
3180 the given file was scanned by exinext itself; it wasn't being passed on
3181 when Exim was called.
3183 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3184 an end-of-file indication when reading a command response.
3186 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3187 compiled. In many other places in Exim, IPv6 addresses are always
3188 recognized, so I have changed this. It also means that IPv4 domain
3189 literals of the form [IPV4:n.n.n.n] are now always recognized.
3191 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3192 used if the router is not running as root, for example, when verifying at
3193 ACL time, or when using -bh. The debugging output from this situation was
3194 non-existent - all you got was a failure to exec. I have made two
3197 (a) Failures to set uid/gid, the current directory, or a process leader
3198 in a subprocess such as that created by queryprogram now generate
3199 suitable debugging output when -d is set.
3201 (b) The queryprogram router detects when it is not running as root,
3202 outputs suitable debugging information if -d is set, and then runs
3203 the subprocess without attempting to change uid/gid.
3205 PH/15 Minor change to Makefile for building test_host (undocumented testing
3208 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3209 additional section of a DNS packet that returns MX or SRV records.
3210 Instead, it always explicitly searches for A/AAAA records. This avoids
3211 major problems that occur when a DNS server includes only records of one
3212 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3213 fixed another bug: if SRV records were looked up and the corresponding
3214 address records were *not* found in the additional section, the port
3215 values from the SRV records were lost.
3217 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3218 using the correct key (the original address) when searching the retry
3219 rules in order to find which one to use for generating the retry hint.
3221 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3222 adding the default one. Similarly, if it contains a Reply-To: header, the
3223 errors_reply_to option, if set, is not used.
3225 PH/19 When calculating a retry time, Exim used to measure the "time since
3226 failure" by looking at the "first failed" field in the retry record. Now
3227 it does not use this if it is later than than the arrival time of the
3228 message. Instead it uses the arrival time. This makes for better
3229 behaviour in cases where some deliveries succeed, thus re-setting the
3230 "first failed" field. An example is a quota failure for a huge message
3231 when small messages continue to be delivered. Without this change, the
3232 "time since failure" will always be short, possible causing more frequent
3233 delivery attempts for the huge message than are intended.
3234 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3236 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3237 $1, $2, $3) because the numerical variables can be reset during some
3238 expansion items (e.g. "match"), thereby losing the authentication data.
3240 PH/21 Make -bV show the size of off_t variables so that the test suite can
3241 decide whether to run tests for quotas > 2G.
3243 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3244 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3245 filecount value is greater than 2G or if a quota value is greater than 2G
3246 on a system where the size of off_t is not greater than 4, a panic error
3249 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3250 never match. The debug and -bh output now contains an explicit error
3251 message indicating a malformed IPv4 address or mask.
3253 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3254 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3255 PH/23 above applies.
3257 PH/25 Do not write to syslog when running in the test harness. The only
3258 occasion when this arises is a failure to open the main or panic logs
3259 (for which there is an explicit test).
3261 PH/26 Added the /no_tell option to "control=freeze".
3263 PH/27 If a host name lookup failed very early in a connection, for example, if
3264 the IP address matched host_lookup and the reverse lookup yielded a name
3265 that did not have a forward lookup, an error message of the form "no IP
3266 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3267 could be logged. Now it outputs the IP address instead of "NULL".
3269 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3270 allows the sender and the authenticated sender to be set when
3271 submitting a message from within Exim. Since child_open_exim() is
3272 documented for local_scan(), the new function should be too.
3274 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3275 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3276 results in an empty string is now treated as unset.
3278 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3280 PH/31 Added sender_verify_fail logging option.
3282 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3283 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3284 tidied the source and removed it altogether.
3286 PH/33 When a queue run was abandoned because the load average was too high, a
3287 log line was always written; now it is written only if the queue_run log
3288 selector is set. In addition, the log line for abandonment now contains
3289 information about the queue run such as the pid. This is always present
3290 in "start" and "stop" lines but was omitted from the "abandon" line.
3292 PH/34 Omit spaces between a header name and the colon in the error message that
3293 is given when verify = headers_syntax fails (if there are lots of them,
3294 the message gets confusing).
3296 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3297 names, as there are now some PTR records that contain slashes. This check
3298 is only to protect against broken name servers that fall over on strange
3299 characters, so the fact that it applies to all lookups doesn't matter.
3301 PH/36 Now that the new test suite is complete, we can remove some of the
3302 special code in Exim that was needed for the old test suite. For example,
3303 sorting DNS records because real resolvers return them in an arbitrary
3304 order. The new test suite's fake resolver always returns records in the
3307 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3308 bounces) except when queue_only is set, to avoid logging races between
3309 the different processes.
3311 PH/38 Panic-die if .include specifies a non-absolute path.
3313 PH/39 A tweak to the "H" retry rule from its user.
3315 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3316 a label. They prevented compilation on older perls.
3318 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3319 a warning to be raised on newish perls.
3321 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3322 on queue. Changes to match documented behaviour of showing count of
3323 messages matching specified criteria.
3325 PH/40 Changed the default ident timeout from 30s to 5s.
3327 PH/41 Added support for the use of login_cap features, on those BSD systems
3328 that have them, for controlling the resources used by pipe deliveries.
3330 PH/42 The content-scanning code uses fopen() to create files in which to put
3331 message data. Previously it was not paying any attention to the mode of
3332 the files. Exim runs with umask(0) because the rest of the code creates
3333 files with open(), and sets the required mode explicitly. Thus, these
3334 files were ending up world-writeable. This was not a big issue, because,
3335 being within the spool directory, they were not world-accessible. I have
3336 created a function called modefopen, which takes an additional mode
3337 argument. It sets umask(777), creates the file, chmods it to the required
3338 mode, then resets the umask. All the relevant calls to fopen() in the
3339 content scanning code have been changed to use this function.
3341 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3342 to 24 hours. This avoids potential overflow problems when processing G
3343 and H retry rules. I suspect nobody ever tinkers with this value.
3345 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3347 PH/45 When the plaintext authenticator is running as a client, the server's
3348 challenges are checked to ensure they are valid base64 strings. By
3349 default, the authentication attempt is cancelled if an invalid string is
3350 received. Setting client_ignore_invalid_base64 true ignores these errors.
3351 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3352 they are received. Thus, the responses can be made to depend on the
3353 challenges. If an invalid string is ignored, an empty string is placed in
3356 PH/46 Messages that are created by the autoreply transport now contains a
3357 References: header, in accordance with RFCs 2822 and 3834.
3359 PH/47 Added authenticated_sender_force to the smtp transport.
3361 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3363 PH/49 Installed latest patch from the Sieve maintainer.
3365 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3366 was also set, the appendfile transport was unnecessarily scanning a
3367 directory of message files (e.g. for maildir delivery) to find the count
3368 of files (along with the size), even though it did not need this
3369 information. It now does the scan only if it needs to find either the
3370 size of the count of files.
3372 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3374 PH/52 Two bugs concerned with error handling when the smtp transport is
3377 (i) Exim was not creating retry information for temporary errors given
3378 for individual recipients after the DATA command when the smtp transport
3379 was used in LMTP mode. This meant that they could be retried too
3380 frequently, and not timed out correctly.
3382 (ii) Exim was setting the flag that allows error details to be returned
3383 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3384 recipients that were returned after the DATA command.
3386 PH/53 This is related to PH/52, but is more general: for any failing address,
3387 when detailed error information was permitted to be returned to the
3388 sender, but the error was temporary, then after the final timeout, only
3389 "retry timeout exceeded" was returned. Now it returns the full error as
3390 well as "retry timeout exceeded".
3392 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3393 do this, and (what is worse) MTAs that accept it.
3395 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3396 will now be deprecated.
3398 PH/56 New os.c-cygwin from the Cygwin maintainer.
3400 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3401 formats (previously only available in exim formats via -bpr, -bpru,
3402 and -bpra. Now also available in native and exiqgrep formats)
3404 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3405 with very large, slow to parse queues
3407 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3409 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3411 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3412 responses to authentication challenges, though it was showing the
3413 challenges; (ii) I've removed the CR characters from the debug output for
3416 PH/58 Allow for the insertion of a newline as well as a space when a string
3417 is turned into more than one encoded-word during RFC 2047 encoding. The
3418 Sieve code now uses this.
3420 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3421 data_4xx, lost_connection, tls_required.
3423 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3424 message was being sent as an SMTP response.
3426 PH/61 Add -l and -k options to exicyclog.
3428 PH/62 When verifying, if an address was redirected to one new address, so that
3429 verification continued, and the new address failed or deferred after
3430 having set something in $address_data, the value of $address_data was not
3431 passed back to the ACL. This was different to the case when no
3432 redirection occurred. The value is now passed back in both cases.
3434 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3435 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3436 use login_cap.h, so on its own it isn't the distinguishing feature. The
3437 new name refers directly to the setclassresources() function.
3439 PH/65 Added configuration files for NetBSD3.
3441 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3443 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3444 is preferred over IPv4.
3446 PH/68 The bounce_return_message and bounce_return_body options were not being
3447 honoured for bounces generated during the reception of non-SMTP messages.
3448 In particular, this applied to messages rejected by the ACL. This bug has
3449 been fixed. However, if bounce_return_message is true and bounce_return_
3450 body is false, the headers that are returned for a non-SMTP message
3451 include only those that have been read before the error was detected.
3452 (In the case of an ACL rejection, they have all been read.)
3454 PH/69 The HTML version of the specification is now built in a directory called
3455 spec_html instead of spec.html, because the latter looks like a path with
3456 a MIME-type, and this confuses some software.
3458 PH/70 Catch two compiler warnings in sieve.c.
3460 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3461 function verify_get_ident() calls ip_connect() to connect a socket, but
3462 if the "connect()" function timed out, ip_connect() used to close the
3463 socket. However, verify_get_ident() also closes the socket later, and in
3464 between Exim writes to the log, which may get opened at this point. When
3465 the socket was closed in ip_connect(), the log could get the same file
3466 descriptor number as the socket. This naturally causes chaos. The fix is
3467 not to close the socket in ip_connect(); the socket should be closed by
3468 the function that creates it. There was only one place in the code where
3469 this was missing, in the iplookup router, which I don't think anybody now
3470 uses, but I've fixed it anyway.
3472 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3473 well as to direct DNS lookups. Otherwise the handling of names in host
3474 lists is inconsistent and therefore confusing.
3480 PH/01 Two changes to the default runtime configuration:
3482 (1) Move the checks for relay_from_hosts and authenticated clients from
3483 after to before the (commented out) DNS black list checks.
3485 (2) Add control=submission to the relay_from_hosts and authenticated
3486 clients checks, on the grounds that messages accepted by these
3487 statements are most likely to be submissions.
3489 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3491 (1) Generate an error if the third argument for the ${prvs expansion is
3494 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3497 (3) Reset the variables that are obtained from the first argument of
3498 ${prvscheck and used in the second argument before leaving the code,
3499 because their memory is reclaimed, so using them afterwards may do
3502 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3503 one (it's much easier than Tom thought :-).
3505 (5) Because of (4), we can now allow for the use of $prvscheck_result
3506 inside the third argument.
3508 PH/03 For some reason, the default setting of PATH when running a command from
3509 a pipe transport was just "/usr/bin". I have changed it to
3512 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3513 anything to be listed in the output from -bV.
3515 PH/05 When a filter generated an autoreply, the entire To: header line was
3516 quoted in the delivery log line, like this:
3518 => >A.N.Other <ano@some.domain> <original@ddress> ...
3520 This has been changed so that it extracts the operative address. There
3521 may be more than one such address. If so, they are comma-separated, like
3524 => >ano@some.domain,ona@other.domain <original@ddress> ...
3526 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3527 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3528 not being looked up in the rDNS to get a host name, Exim was showing the
3529 IP address twice in Received: lines, even though the IP addresses were
3530 identical. For example:
3532 Received: from [1.2.3.4] (helo=[1.2.3.4])
3534 However, if the real host name was known, it was omitting the HELO data
3535 if it matched the actual IP address. This has been tidied up so that it
3536 doesn't show the same IP address twice.
3538 PH/07 When both +timestamp and +memory debugging was on, the value given by
3539 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3540 called by the debug printing, thereby overwriting the timestamp buffer.
3541 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3543 PH/08 When the original message was included in an autoreply transport, it
3544 always said "this is a copy of the message, including all the headers",
3545 even if body_only or headers_only was set. It now gives an appropriate
3548 PH/09 Applied a patch from the Sieve maintainer which:
3550 o fixes some comments
3551 o adds the (disabled) notify extension core
3552 o adds some debug output for the result of if/elsif tests
3553 o points to the current vacation draft in the documentation
3554 and documents the missing references header update
3558 o fixes a bug in processing the envelope test (when testing
3559 multiple envelope elements, the last element determined the
3562 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3563 Electronic Mail") by including:
3565 Auto-submitted: auto-generated
3567 in the messages that it generates (bounce messages and others, such as
3568 warnings). In the case of bounce messages for non-SMTP messages, there was
3569 also a typo: it was using "Auto_submitted" (underscore instead of
3570 hyphen). Since every message generated by Exim is necessarily in response
3571 to another message, thes have all been changed to:
3573 Auto-Submitted: auto-replied
3575 in accordance with these statements in the RFC:
3577 The auto-replied keyword:
3579 - SHOULD be used on messages sent in direct response to another
3580 message by an automatic process,
3582 - MUST NOT be used on manually-generated messages,
3584 - MAY be used on Delivery Status Notifications (DSNs) and Message
3585 Disposition Notifications (MDNs),
3587 - MUST NOT be used on messages generated by automatic or periodic
3588 processes, except for messages which are automatic responses to
3591 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3592 to the default Received: header definition.
3594 PH/12 Added log selector acl_warn_skipped (default on).
3596 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3597 variables because (a) they are in the wrong storage pool and (b) even if
3598 they were copied, it wouldn't work properly because of the caching.
3600 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3601 checking when decoding. Apparently there are clients that generate
3602 overlong encoded strings. Why am I not surprised?
3604 PH/15 If the first argument of "${if match_address" was not empty, but did not
3605 contain an "@" character, Exim crashed. Now it writes a panic log message
3606 and treats the condition as false.
3608 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3610 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3611 extension "envelope-auth". The code is finished and in agreement with
3612 other implementations, but there is no documentation so far and in fact,
3613 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3614 not changing the active code.
3616 Print executed "if" and "elsif" statements when debugging is used. This
3617 helps a great deal to understand what a filter does.
3619 Document more things not specified clearly in RFC3028. I had all this
3620 sorted out, when out of a sudden new issues came to my mind. Oops."
3622 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3625 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3626 canonical form (as documented). However, after a host name lookup from
3627 the IP address, check_host() was doing a simple string comparison with
3628 addresses acquired from the DNS when checking that the found name did
3629 have the original IP as one of its addresses. Since any found IPv6
3630 addresses are likely to be in abbreviated form, the comparison could
3631 fail. Luckily, there already exists a function for doing the comparison
3632 by converting both addresses to binary, so now that is used instead of
3633 the text comparison.
3635 PH/20 There was another similar case to PH/19, when a complete host name was
3636 given in a host list; looking up its IP address could give an abbreviated
3637 form, whereas the current host's name might or might not be abbreviated.
3638 The same fix has been applied.
3644 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3645 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3648 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3649 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3651 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3653 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3654 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3655 Also, for "Auto-Submitted", treat anything other than "no" as
3656 non-personal, in accordance with RFC 3834. (Previously it treated
3657 anything starting "auto-" as non-personal.)
3659 TF/01 The control=submission/name=... option had a problem with syntax
3660 errors if the name included a slash character. The /name= option
3661 now slurps the rest of the string, so it can include any characters
3662 but it must come last in the list of options (after /sender_retain
3665 PH/05 Some modifications to the interface to the fake nameserver for the new
3673 TK/01 Added the "success_on_redirect" address verification option. See
3674 NewStuff for rationale and an example.
3676 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3678 PH/02 Patch to exigrep to allow it to work on syslog lines.
3680 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3681 fread() to skip over the body file's header line, because in Cygwin the
3682 header line is locked and is inaccessible.
3684 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3685 co-exist for some time) to make it clear that it is the Exim ID that is
3686 referenced, not the Message-ID: header line.
3688 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3689 string_format() function, because snprintf() does not exist on all
3692 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3693 ${stat: expansion item.
3695 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3696 protocol synchronization error", to keep the pedants happy.
3698 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3699 well as for IRIX systems, when gcc is being used. See the host.c source
3702 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3704 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3707 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3708 transport and to the smtp transport in LMTP mode.
3710 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3712 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3713 run a filter in a subprocess. This could lead to confusion in subsequent
3714 lookups in the parent process. There should also be a search_tidyup() at
3715 the end of the subprocess.
3717 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3718 only if the host matched helo_try_verify_hosts, which caused the
3719 verification to occur when the EHLO/HELO command was issued. The ACL just
3720 tested the remembered result. Now, if a previous verification attempt has
3721 not happened, "verify = helo" does it there and then.
3723 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3725 TK/03 Fix log output including CR from clamd.
3727 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3728 exist provoked a memory error which could cause a segfault.
3730 PH/15 Installed PCRE 6.2
3732 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3734 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3735 of the problem. Specifically, suggested +O2 rather than +O1 for the
3738 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3740 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3741 record, the port was not correctly logged when the outgoing_port log
3742 selector was set (it logged the transort's default port).
3744 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3745 fallback_hosts, and "hosts" in the smtp transport.
3747 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3748 host errors such as "Connection refused".
3750 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3751 authentication with radiusclient 0.4.9:
3753 - Error returned from rc_read_config was caught wrongly
3754 - Username/password not passed on to radius server due to wrong length.
3756 The presumption is that some radiusclient API changes for 4.51/PH/17
3757 were not taken care of correctly. The code is still untested by me (my
3758 Linux distribution still has 0.3.2 of radiusclient), but it was
3759 contributed by a Radius user.
3761 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3762 expanding the "port" option of the smtp transport.
3764 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3765 while reading a MIME header. Thanks to Tom Hughes for a patch.
3767 PH/24 Include config.h inside local_scan.h so that configuration settings are
3770 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3771 that in an ACL for RCPT (for example), you can examine exactly what was
3774 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3775 commands, but it was not correctly comparing the address with the actual
3776 client host address. Thus, it would show the EHLO address in Received:
3777 header lines when this was not necessary.
3779 PH/27 Added the % operator to ${eval:}.
3781 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3782 it should be ignoring failures (because with -C, for example, it has lost
3783 privilege). It wasn't ignoring creation failures other than "already
3786 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3789 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3790 because an input line was too long, either on its own, or by virtue of
3791 too many continuations, the temporary file was not being removed, and the
3792 return code was incorrect.
3794 PH/31 Missing "BOOL" in function definition in filtertest.c.
3796 PH/32 Applied Sieve patches from the maintainer.
3798 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3800 PH/33 Added "verify = not_blind".
3802 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3803 Local/Makefile (with some defaults set). These are used in built scripts
3804 such as exicyclog, but they have never been used in the exim_install
3805 script (though there are many overriding facilities there). I have
3806 arranged that the exim_install script now takes note of these two
3809 PH/35 Installed configuration files for Dragonfly.
3811 PH/36 When a locally submitted message by a trusted user did not contain a
3812 From: header, and the sender address was obtained from -f or from an SMTP
3813 MAIL command, and the trusted user did not use -F to supply a sender
3814 name, $originator_name was incorrectly used when constructing a From:
3815 header. Furthermore, $originator_name was used for submission mode
3816 messages from external hosts without From: headers in a similar way,
3817 which is clearly wrong.
3819 PH/37 Added control=suppress_local_fixups.
3821 PH/38 When log_selector = +received_sender was set, and the addition of the
3822 sender made the log line's construction buffer exactly full, or one byte
3823 less than full, an overflow happened when the terminating "\n" was
3826 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3827 when the result of a list match is failure because a DNS lookup failed.
3829 PH/40 RM_COMMAND is now used in the building process.
3831 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3832 the "build-* directories that it finds.
3834 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3835 address in a domain literal was a prefix of an interface address.
3837 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3838 when verifying a sender address, unless rewrite_headers is false.
3840 PH/44 Wrote a long comment about why errors_to addresses are verified as
3841 recipients, not senders.
3843 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3844 the ratelimit ACL was added.
3846 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3848 PH/46 Added extra information about PostgreSQL errors to the error string.
3850 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3851 avoiding the need to install special zones in a real server. This is
3852 backwards compatible; if it can't find the fake resolver, it drops back.
3853 Thus, both old and new test suites can be run.
3855 TF/02 Added util/ratelimit.pl
3857 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3858 clock is set back in time.
3860 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3861 Candler <B.Candler@pobox.com>.
3863 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3864 OK for addresses that are the result of redirections.
3866 PH/48 A number of further additions for the benefit of the new test suite,
3867 including a fake gethostbyname() that interfaces to the fake DNS resolver
3870 TF/06 The fix for widen_domains has also been applied to qualify_single and
3871 search_parents which are the other dnslookup options that can cause
3874 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3877 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3879 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3880 service" if it encountered an MX record with an empty target hostname.
3881 The message is now "an MX or SRV record indicated no SMTP service".
3883 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3884 if the DNS of the sending site is misconfigured. This is quite a
3885 common situation. This change restores the behaviour of treating a
3886 helo verification defer as a failure.
3888 PH/51 If self=fail was set on a router, the bounce message did not include the
3889 actual error message.
3895 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3897 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3898 command itself ended in error, the underlying message about the transport
3899 filter timeout was being overwritten with the pipe command error. Now the
3900 underlying error message should be appended to the second error message.
3902 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3904 PH/02 Reduce the amount of output that "make" produces by default. Full output
3905 can still be requested.
3907 PH/03 The warning log line about a condition test deferring for a "warn" verb
3908 was being output only once per connection, rather than after each
3909 occurrence (because it was using the same function as for successful
3910 "warn" verbs). This seems wrong, so I have changed it.
3912 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3913 it should not have, which might have caused a crash in the right
3914 circumstances, but probably never did.
3916 PH/04 Installed a modified version of Tony Finch's patch to make submission
3917 mode fix the return path as well as the Sender: header line, and to
3918 add a /name= option so that you can make the user's friendly name appear
3921 TF/03 Added the control = fakedefer ACL modifier.
3923 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3924 Mark Lowes for thorough testing.
3926 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3928 TK/03 Merged latest SRS patch from Miles Wilton.
3930 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3931 with the definition in sysexits.h (which is #included earlier).
3932 Fortunately, Exim does not actually use EX_OK. The code used to try to
3933 preserve the sysexits.h value, by assuming that macro definitions were
3934 scanned for macro replacements. I have been disabused of this notion,
3935 so now the code just undefines EX_OK before #including unistd.h.
3937 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3938 in the smtp transport. When a block could not be written in a single
3939 write() function, the timeout was being re-applied to each part-write.
3940 This seems wrong - if the receiver was accepting one byte at a time it
3941 would take for ever. The timeout is now adjusted when this happens. It
3942 doesn't have to be particularly precise.
3944 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3945 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3947 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3948 without a domain if the check to <postmaster@domain> fails.
3950 SC/01 Eximstats: added -xls and the ability to specify output files
3951 (patch written by Frank Heydlauf).
3953 SC/02 Eximstats: use FileHandles for outputting results.
3955 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3957 SC/04 Eximstats: fixed display of large numbers with -nvr option
3959 SC/05 Eximstats: fixed merging of reports with empty tables.
3961 SC/06 Eximstats: added the -include_original_destination flag
3963 SC/07 Eximstats: removed tabs and trailing whitespace.
3965 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3967 TK/06 MBOX spool code: Add real "From " MBOX separator line
3968 so the .eml file is really in mbox format (even though
3969 most programs do not really care). Patch from Alex Miller.
3971 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3972 The latter is generated from $received_to and is only set if the
3973 message has one envelope recipient. SA can use these headers,
3974 obviously out-of-the-box. Patch from Alex Miller.
3976 PH/08 The ${def test on a variable was returning false if the variable's
3977 value was "0", contrary to what the specification has always said!
3978 The result should be true unless the variable is empty.
3980 PH/09 The syntax error of a character other than { following "${if
3981 def:variable_name" (after optional whitespace) was not being diagnosed.
3982 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3983 accidental colon was present, for example, could give incorrect results.
3985 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3986 result is used (not including appendfile, where other changes are about
3989 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3990 This involved changing a lot of size variables from int to off_t. It
3991 should work with maildirs and everything.
3993 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3994 spamd dying while we are connected to it.
3996 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3999 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4000 function for BDB 4.3.
4002 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4004 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4005 the sources for PCRE have been re-arranged and more files are now
4008 PH/15 The code I had for printing potentially long long variables in PH/11
4009 above was not the best (it lost precision). The length of off_t variables
4010 is now inspected at build time, and an appropriate printing format (%ld
4011 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4012 to be "long long int" or "long int". This is needed for the internal
4013 formatting function string_vformat().
4015 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4016 the configuration file to be ":syslog", then the script "guesses" where
4017 the logs files are, rather than using the compiled in default. In our
4018 case the guess is not the same as the compiled default, so the script
4019 suddenly stopped working when I started to use syslog. The patch checks
4020 to see if log_file_path is "". If so, it attempts to read it from exim
4021 with no configuration file to get the compiled in version, before it
4022 falls back to the previous guessing code."
4024 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4025 implementing BATV in an Exim configuration. See NewStuff for the gory
4028 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4029 Makefile that are specific to HP-UX.
4031 PH/18 If the "use_postmaster" option was set for a recipient callout together
4032 with the "random" option, the postmaster address was used as the MAIL
4033 FROM address for the random test, but not for the subsequent recipient
4034 test. It is now used for both.
4036 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4037 patch removes a few documentation additions to RFC 3028, because the
4038 latest draft now contains them. It adds the new en;ascii-case comparator
4039 and a new error check for 8bit text in MIME parts. Comparator and
4040 require names are now matched exactly. I enabled the subaddress
4041 extension, but it is not well tested yet (read: it works for me)."
4043 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4044 rework some of the code of TK/09 above to avoid the hardwired use of
4045 "%lld" and "long long". Replaced the call to snprintf() with a call to
4048 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4049 records point to non-existent hosts", "retry timeout exceeded", and
4050 "retry time not reached for any host after a long failure period".
4052 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4053 experimental DomainKeys support:
4055 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4056 (2) On an error such as an illegally used "control", the wrong name for
4057 the control was given.
4059 These problems did NOT occur unless DomainKeys support was compiled.
4061 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4063 PH/24 Added ${if match_ip condition.
4065 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4066 databases so that it will be absolutely obvious if a crash occurs in the
4067 DB library. This is a regular occurrence (often caused by mis-matched
4070 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4071 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4072 picked up on a user's system that detects such things. There doesn't seem
4073 to be a gcc warning option for this - only an attribute that has to be
4074 put on the function's prototype. It seems that in Fedora Core 4 they have
4075 set this on a number of new functions. No doubt there will be more in due
4078 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4079 specify a transport. However, if an address that was verified by such a
4080 router was the subject of a callout, Exim crashed because it tried to
4081 read the rcpt_include_affixes from the non-existent transport. Now it
4082 just assumes that the setting of that option is false. This bug was
4083 introduced by 4.51/PH/31.
4085 PH/28 Changed -d+all to exclude +memory, because that information is very
4086 rarely of interest, but it makes the output a lot bigger. People tend to
4087 do -d+all out of habit.
4089 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4090 code in os-type was giving problems when libc.so lives in lib64, like on
4093 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4094 aren't the modern standard, and it seems that some systems' include files
4095 don't always have them. Exim was already checking for some of the newer
4096 ones like T_AAAA, and defining it itself. I've added checks for all the
4097 record types that Exim uses.
4099 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4100 not automatically generating a new one, as it is supposed to. This
4101 prevented TLS from working. If the file did exist, but contained invalid
4102 data, a new version was generated, as expected. It was only the case of a
4103 non-existent file that was broken.
4105 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4106 with a change in libdomainkeys > 0.64.
4108 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4109 from DNS. If the selector record carries the flag, it now has
4110 precedence over the domain-wide flag.
4112 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4114 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4115 the use of a port name that isn't defined in /etc/services) occurred, the
4116 message was deferred as in a normal delivery, and thus remained on the
4117 spool, instead of being failed because of the mua_wrapper setting. This
4118 is now fixed, and I tidied up some of the mua_wrapper messages at the
4121 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4122 the messages in a hash of arrays rather than using individual hashes.
4123 This is a bit cleaner and results in dramatic memory savings, albeit
4124 at a slight CPU cost.
4126 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4127 as requested by Marc Sherman.
4129 SC/10 Eximstats: added histograms for user specified patterns as requested
4132 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4134 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4135 fopen() in the content-scanning modules that did not already have it.
4141 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4142 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4144 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4146 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4148 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4149 to test Sieve filters that use "vacation".
4151 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4152 that changes the way the GnuTLS parameters are stored in the cache file.
4153 The new format can be generated externally. For backward compatibility,
4154 if the data in the cache doesn't make sense, Exim assumes it has read an
4155 old-format file, and it generates new data and writes a new file. This
4156 means that you can't go back to an older release without removing the
4159 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4160 work if there are any delivery delays because "one_time" forces the
4161 parent to be marked "delivered", so its unseen clone is never tried
4162 again. For this reason, Exim now forbids the simultaneous setting of
4165 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4166 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4167 entry if you want to know the details. The fix, however, neglected to
4168 consider the case when local delivery batching is involved. The test for
4169 "previously delivered" was not happening when checking to see if an
4170 address could be batched with a previous (undelivered) one; under
4171 certain circumstances this could lead to multiple deliveries to the same
4174 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4175 in its include files, and this causes problems building Exim.
4177 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4178 header_syntax) but Exim was just ignoring anything given after a slash.
4179 In particular, this caused confusion with an attempt to use "verify =
4180 reverse_host_lookup/defer_ok". An error is now given when options are
4181 supplied for verify items that do not have them. (Maybe reverse_host_
4182 lookup should have a defer_ok option, but that's a different point.)
4184 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4185 defined by RFC 821) to 2048, because there were problems with some AUTH
4186 commands, and RFC 1869 says the size should be increased for extended
4187 SMTP commands that take arguments.
4189 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4192 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4193 "unknown" error; now it says that the functionality isn't in the binary.
4195 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4196 an address' error message when a string expansion fails (syntax or
4197 whatever). Otherwise the password may appear in the log. Following change
4198 PH/42 below, there is no longer a chance of it appearing in a bounce
4201 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4203 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4204 of its addresses was ever tried. (Bugzilla bug #2.)
4206 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4207 the result incorrectly in the debug output. (It correctly added a newline
4208 to what was transported.)
4210 TF/01 Added $received_time.
4212 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4213 commented out examples of how to interface to a virus scanner and to
4214 SpamAssassin. Also added commented examples of av_scanner and
4215 spamd_address settings.
4217 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4218 and controls are allowed in which ACLs. There were a couple of minor
4219 errors. Some of the entries in the conditions table (which is a table of
4220 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4221 negation of where the condition IS allowed.
4223 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4225 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4226 header file does not have a version number, so I've had to invent a new
4227 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4228 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4229 radiusclient), but it was contributed by a Radius user.
4231 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4232 files or directories, for OpenSSL.
4234 PH/19 When an Exim process that is running as root has to create an Exim log
4235 file, it does so in a subprocess that runs as exim:exim so as to get the
4236 ownership right at creation (otherwise, other Exim processes might see
4237 the file with the wrong ownership). There was no test for failure of this
4238 fork() call, which would lead to the process getting stuck as it waited
4239 for a non-existent subprocess. Forks do occasionally fail when resources
4240 run out. I reviewed all the other calls to fork(); they all seem to check
4243 PH/20 When checking for unexpected SMTP input at connect time (before writing
4244 the banner), Exim was not dealing correctly with a non-positive return
4245 from the read() function. If the client had disconnected by this time,
4246 the result was a log entry for a synchronization error with an empty
4247 string after "input=" when read() returned zero. If read() returned -1
4248 (an event I could not check), uninitialized data bytes were printed.
4249 There were reports of junk text (parts of files, etc) appearing after
4252 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4254 PH/22 Added support for macro redefinition, and (re)definition in between
4255 driver and ACL definitions.
4257 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4258 forgetting to use the resulting value; it was using the unexpanded value.
4260 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4261 hadn't been configured. The fix is from Juergen Kreileder, who
4262 understands it better than I do:
4264 "Here's what I see happening with three configured cyrus_sasl
4265 authenticators configured (plain, login, cram-md5):
4267 On startup auth_cyrus_sasl_init() gets called for each of these.
4268 This means three calls to sasl_listmech() without a specified mech_list.
4269 => SASL tests which mechs of all available mechs actually work
4270 => three warnings about OTP not working
4271 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4273 With the patch, sasl_listmech() also gets called three times. But now
4274 SASL's mech_list option is set to the server_mech specified in the the
4275 authenticator. Or in other words, the answer from sasl_listmech()
4276 gets limited to just the mech you're testing for (which is different
4278 => the return list contains just 'plain' or 'login', 'cram-md5' or
4279 nothing depending on the value of ob->server_mech.
4281 I've just tested the patch: Authentication still works fine,
4282 unavailable mechs specified in the exim configuration are still
4283 caught, and the auth.log warnings about OTP are gone."
4285 PH/25 When debugging is enabled, the contents of the command line are added
4286 to the debugging output, even when log_selector=+arguments is not
4289 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4290 answer is "GNU", and only if the return is "GNU/something" is the answer
4293 PH/27 $acl_verify_message is now set immediately after the failure of a
4294 verification in an ACL, and so is available in subsequent modifiers. In
4295 particular, the message can be preserved by coding like this:
4297 warn !verify = sender
4298 set acl_m0 = $acl_verify_message
4300 Previously, $acl_verify_message was set only while expanding "message"
4301 and "log_message" when a very denied access.
4303 PH/28 Modified OS/os.c-Linux with
4305 -#ifndef OS_LOAD_AVERAGE
4306 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4308 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4309 nomenclature these days.)
4311 PH/29 Installed patch from the Sieve maintainer that adds the options
4312 sieve_useraddress and sieve_subaddress to the redirect router.
4314 PH/30 In these circumstances:
4315 . Two addresses routed to the same list of hosts;
4316 . First host does not offer TLS;
4317 . First host accepts first address;
4318 . First host gives temporary error to second address;
4319 . Second host offers TLS and a TLS session is established;
4320 . Second host accepts second address.
4321 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4322 and peerdn, if requested) that were in fact used only for the second
4325 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4326 attention to any local part prefix or suffix that was matched by the
4327 router that accepted the address. It now behaves in the same way as it
4328 does for delivery: the affixes are removed from the local part unless
4329 rcpt_include_affixes is set on the transport.
4331 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4332 timeout during the DATA phase of an incoming message.
4334 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4335 applied a patch sanctioned by the Sieve maintainer.
4337 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4338 the uid or gid is negative. A case of a negative gid caused this to be
4339 noticed. The fix allows for either to be negative.
4341 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4342 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4345 PH/36 The change PH/12 above was broken. Fixed it.
4347 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4348 the grounds that routing the same address twice would always produce the
4349 same answer. This might have been true once, but it is certainly no
4350 longer true now. Routing a child address may depend on the previous
4351 routing that produced that child. Some complicated redirection strategies
4352 went wrong when messages had multiple recipients, and made Exim's
4353 behaviour dependent on the order in which the addresses were given.
4355 I have moved the duplicate checking until after the routing is complete.
4356 Exim scans the addresses that are assigned to local and remote
4357 transports, and removes any duplicates. This means that more work will be
4358 done, as duplicates will always all be routed, but duplicates are
4359 presumably rare, so I don't expect this is of any significance.
4361 For deliveries to pipes, files, and autoreplies, the duplicate checking
4362 still happens during the routing process, since they are not going to be
4365 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4366 It corrects a timeout issue with spamd. This is Ian's comment: "The
4367 background is that sometimes spamd either never reads data from a
4368 connection it has accepted, or it never writes response data. The exiscan
4369 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4370 blindly assumes that writes won't block so it may never time out."
4372 PH/39 Allow G after quota size as well as K and M.
4374 PH/40 The value set for $authenticated_id in an authenticator may not contain
4375 binary zeroes or newlines because the value is written to log lines and
4376 to spool files. There was no check on this. Now the value is run through
4377 the string_printing() function so that such characters are converted to
4378 printable escape sequences.
4380 PH/41 $message_linecount is a new variable that contains the total number of
4381 lines in the message. Compare $body_linecount, which is the count for the
4384 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4385 bounce and delay warning messages, except in certain special cases, which
4388 (a) An SMTP error message from a remote host;
4389 (b) A message specified in a :fail: redirection;
4390 (c) A message specified in a "fail" command in a system filter;
4391 (d) A message specified in a FAIL return from the queryprogram router;
4392 (e) A message specified by the cannot_route_message router option.
4394 In these cases only, Exim does include the error details in bounce and
4395 warning messages. There are also a few cases where bland messages such
4396 as "unrouteable address" or "local delivery error" are given.
4398 PH/43 $value is now also set for the "else" part of a ${run expansion.
4400 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4401 being worked on, but at least Exim now implements the latest version to
4404 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4405 process to complete was treated as a delivery failure, a timeout while
4406 writing the message to the pipe was logged, but erroneously treated as a
4407 successful delivery. Such timeouts include transport filter timeouts. For
4408 consistency with the overall process timeout, these timeouts are now
4409 treated as errors, giving rise to delivery failures by default. However,
4410 there is now a new Boolean option for the pipe transport called
4411 timeout_defer, which, if set TRUE, converts the failures into defers for
4412 both kinds of timeout. A transport filter timeout is now identified in
4415 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4416 systems where "make" and "gmake" are different, calling "gmake" at top
4417 level broke things. I've arranged for the value of $(MAKE) to be passed
4418 from the Makefile to this script so that it can call the same version of
4422 A note about Exim versions 4.44 and 4.50
4423 ----------------------------------------
4425 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4426 changes of various kinds. As a consequence, a big documentation update was
4427 needed. This delayed the release for rather longer than seemed good, especially
4428 in the light of a couple of (minor) security issues. Therefore, the changes
4429 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4430 release. So 4.44 and 4.50 are in effect two different branches that both start
4433 I have left the 4.50 change log unchanged; it contains all the changes since
4434 4.43. The change log for 4.44 is below; many of its items are identical to
4435 those for 4.50. This seems to be the most sensible way to preserve the
4436 historical information.
4442 1. Minor wording change to the doc/README.SIEVE file.
4444 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4445 computation of the current number of files was incorrect.
4447 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4448 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4451 4. Give more explanation in the error message when the command for a transport
4452 filter fails to execute.
4454 5. There are several places where Exim runs a non-Exim command in a
4455 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4456 was being done only for the command run by the queryprogram router. It is
4457 now done for all such subprocesses. The other cases are: ${run, transport
4458 filters, and the commands run by the lmtp and pipe transports.
4460 6. Added CONFIGURE_GROUP build-time option.
4462 7. Some older OS have a limit of 256 on the maximum number of file
4463 descriptors. Exim was using setrlimit() to set 1000 as a large value
4464 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4465 systems. I've change it so that if it can't get 1000, it tries for 256.
4467 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4468 was an oversight, and furthermore, ever since the addition of extra
4469 controls (e.g. 4.43/32), the checks on when to allow different forms of
4470 "control" were broken. There should now be diagnostics for all cases when a
4471 control that does not make sense is encountered.
4473 9. Added the /retain_sender option to "control=submission".
4475 10. $recipients is now available in the predata ACL (oversight).
4477 11. Tidy the search cache before the fork to do a delivery from a message
4478 received from the command line. Otherwise the child will trigger a lookup
4479 failure and thereby defer the delivery if it tries to use (for example) a
4480 cached ldap connection that the parent has called unbind on.
4482 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4483 of $address_data from the recipient verification was clobbered by the
4484 sender verification.
4486 13. The value of address_data from a sender verification is now available in
4487 $sender_address_data in subsequent conditions in the ACL statement.
4489 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4491 15. Added a new option "connect=<time>" to callout options, to set a different
4494 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4495 was its contents. (It was OK if the option was not defined at all.)
4497 17. A "Completed" log line is now written for messages that are removed from
4498 the spool by the -Mrm option.
4500 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4501 information about exactly what failed.
4503 19. Added -dd to debug only the daemon process.
4505 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4506 handles timeouts, both on the server side and network timeouts. Renamed the
4507 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4509 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4510 It is now set to "smtps".
4512 22. $host_address is now set to the target address during the checking of
4513 ignore_target_hosts.
4515 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4516 being passed; this would have caused $sender_host_name to have been used if
4517 matching the list had actually called for a host name (not very likely,
4518 since this list is usually IP addresses). A host name is now passed as
4521 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4522 code that specifies a non-restarting handler (typically sigaction() in
4523 modern systems) in an attempt to fix a rare and obscure crash bug.
4525 25. Narrowed the window for a race in the daemon that could cause it to ignore
4526 SIGCHLD signals. This is not a major problem, because they are used only to
4527 wake it up if nothing else does.
4529 26. A malformed maildirsize file could cause Exim to calculate negative values
4530 for the mailbox size or file count. Odd effects could occur as a result.
4531 The maildirsize information is now recalculated if the size or filecount
4534 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4535 support for a long time. Removed HAVE_SYS_VFS_H.
4537 28. Installed the latest version of exipick from John Jetmore.
4539 29. In an address list, if the pattern was not a regular expression, an empty
4540 subject address (from a bounce message) matched only if the pattern was an
4541 empty string. Non-empty patterns were not even tested. This was the wrong
4542 because it is perfectly reasonable to use an empty address as part of a
4543 database query. An empty address is now tested by patterns that are
4544 lookups. However, all the other forms of pattern expect the subject to
4545 contain a local part and a domain, and therefore, for them, an empty
4546 address still always fails if the pattern is not itself empty.
4548 30. Exim went into a mad DNS loop when attempting to do a callout where the
4549 host was specified on an smtp transport, and looking it up yielded more
4550 than one IP address.
4552 31. Re-factored the code for checking spool and log partition space into a
4553 function that finds that data and another that does the check. The former
4554 is then used to implement four new variables: $spool_space, $log_space,
4555 $spool_inodes, and $log_inodes.
4557 32. The RFC2047 encoding function was originally intended for short strings
4558 such as real names; it was not keeping to the 75-character limit for
4559 encoded words that the RFC imposes. It now respects the limit, and
4560 generates multiple encoded words if necessary. To be on the safe side, I
4561 have increased the buffer size for the ${rfc2047: expansion operator from
4564 33. It is now permitted to omit both strings after an "if" condition; if the
4565 condition is true, the result is "true". As before, when the second string
4566 is omitted, a false condition yields an empty string. This makes it less
4567 cumbersome to write custom ACL and router conditions.
4569 34. Failure to deliver a bounce message always caused it to be frozen, even if
4570 there was an errors_to setting on the router. The errors_to setting is now
4573 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4574 canonical form (fully expanded) before being placed in
4575 $sender_host_address.
4577 36. The table in the code that translates DNS record types into text (T_A to
4578 "A" for instance) was missing entries for NS and CNAME. It is just possible
4579 that this could have caused confusion if both these types were looked up
4580 for the same domain, because the text type is used as part of Exim's
4581 per-process caching. But the chance of anyone hitting this buglet seems
4584 37. The dnsdb lookup has been extended in a number of ways.
4586 (1) There is a new type, "zns", which walks up the domain tree until it
4587 finds some nameserver records. It should be used with care.
4589 (2) There is a new type, "mxh", which is like "mx" except that it returns
4590 just the host names, not the priorities.
4592 (3) It is now possible to give a list of domains (or IP addresses) to be
4593 looked up. The behaviour when one of the lookups defers can be
4594 controlled by a keyword.
4596 (4) It is now possible to specify the separator character for use when
4597 multiple records are returned.
4599 38. The dnslists ACL condition has been extended: it is now possible to supply
4600 a list of IP addresses and/or domains to be looked up in a particular DNS
4603 39. Added log_selector=+queue_time_overall.
4605 40. When running the queue in the test harness, wait just a tad after forking a
4606 delivery process, to get repeatability of debugging output.
4608 41. Include certificate and key file names in error message when GnuTLS fails
4609 to set them up, because the GnuTLS error message doesn't include the name
4610 of the failing file when there is a problem reading it.
4612 42. Allow both -bf and -bF in the same test run.
4614 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4616 44. The "Exiscan patch" is now merged into the mainline Exim source.
4618 45. Sometimes the final signoff response after QUIT could fail to get
4619 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4620 < 0 before doing a fflush(). This bug looks as though it goes back to the
4621 introduction of TLS in release 3.20, but "sometimes" must have been rare
4622 because the tests only now provoked it.
4624 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4625 (this can affect the format of dates).
4627 47. exim_tidydb, when checking for the continued existence of a message for
4628 which it has found a message-specific retry record, was not finding
4629 messages that were in split spool directories. Consequently, it was
4630 deleting retry records that should have stayed in existence.
4632 48. Steve fixed some bugs in eximstats.
4634 49. The SPA authentication driver was not abandoning authentication and moving
4635 on to the next authenticator when an expansion was forced to fail,
4636 contradicting the general specification for all authenticators. Instead it
4637 was generating a temporary error. It now behaves as specified.
4639 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4640 (the order specifies the preference for clients). The order is now AES256,
4641 AES128, 3DES, ARCFOUR128.
4643 51. Small patch to Sieve code - explicitly set From: when generating an
4646 52. Exim crashed if a remote delivery caused a very long error message to be
4647 recorded - for instance if somebody sent an entire SpamAssassin report back
4648 as a large number of 550 error lines. This bug was coincidentally fixed by
4649 increasing the size of one of Exim's internal buffers (big_buffer) that
4650 happened as part of the Exiscan merge. However, to be on the safe side, I
4651 have made the code more robust (and fixed the comments that describe what
4654 53. Now that there can be additional text after "Completed" in log lines (if
4655 the queue_time_overall log selector is set), a one-byte patch to exigrep
4656 was needed to allow it to recognize "Completed" as not the last thing in
4659 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4660 patch that reportedly fixes this has been added. I am not expert enough to
4661 create a test for it. This is what the patch creator wrote:
4663 "I found a little strange behaviour of ldap code when working with
4664 Windows 2003 AD Domain, where users was placed in more than one
4665 Organization Units. When I tried to give exim partial DN, the exit code
4666 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4667 But simultaneously result of request was absolutely normal ldap result,
4668 so I produce this patch..."
4670 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4671 REFERENCE, so I have modified the code to exclude the patch when that macro
4674 55. Some experimental protocols are using DNS PTR records for new purposes. The
4675 keys for these records are domain names, not reversed IP addresses. The
4676 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4677 leaves it alone. Component reversal etc. now happens only for IP addresses.
4680 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4682 57. Double the size of the debug message buffer (to 2048) so that more of very
4683 long debug lines gets shown.
4685 58. The exicyclog utility now does better if the number of log files to keep
4686 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4688 59. Two changes related to the smtp_active_hostname option:
4690 (1) $smtp_active_hostname is now available as a variable.
4691 (2) The default for smtp_banner uses $smtp_active_hostname instead
4692 of $primary_hostname.
4694 60. The host_aton() function is supposed to be passed a string that is known
4695 to be a valid IP address. However, in the case of IPv6 addresses, it was
4696 not checking this. This is a hostage to fortune. Exim now panics and dies
4697 if the condition is not met. A case was found where this could be provoked
4698 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4699 components; fortuitously, this particular loophole had already been fixed
4700 by change 4.50/55 above.
4702 If there are any other similar loopholes, the new check in host_aton()
4703 itself should stop them being exploited. The report I received stated that
4704 data on the command line could provoke the exploit when Exim was running as
4705 exim, but did not say which command line option was involved. All I could
4706 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4707 running as the user.
4710 61. There was a buffer overflow vulnerability in the SPA authentication code
4711 (which came originally from the Samba project). I have added a test to the
4712 spa_base64_to_bits() function which I hope fixes it.
4715 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4716 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4718 63. The daemon start-up calls getloadavg() while still root for those OS that
4719 need the first call to be done as root, but it missed one case: when
4720 deliver_queue_load_max is set with deliver_drop_privilege. This is
4721 necessary for the benefit of the queue runner, because there is no re-exec
4722 when deliver_drop_privilege is set.
4724 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4725 This has been fixed.
4727 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4728 was in use, was not putting the data itself into the right store pool;
4729 consequently, it could be overwritten for a subsequent message in the same
4730 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4733 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4735 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4736 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4737 the function were treating the return as a boolean value, which happened to
4738 work because 0=false and not-0=true, but is not correct code.
4740 68. The host_aton() function was not handling scoped IPv6 addresses (those
4741 with, for example, "%eth0" on the end) correctly.
4743 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4744 negated items (that is, ~something) in unsigned ints. Some compilers
4745 apparently mutter when there is no cast.
4747 70. If an address verification called from an ACL failed, and did not produce a
4748 user-specific message (i.e. there was only a "system" message), nothing was
4749 put in $acl_verify_message. In this situation, it now puts the system
4752 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4753 session; change 4.31/43 added the unwanted input to the log line - except
4754 that it did not do this in the start of session case. It now does.
4756 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4757 This is wrong and can cause the other end to generate a synchronization
4758 error if it is another Exim or anything else that does the synchronization
4759 check. A QUIT command is no longer sent after a timeout.
4761 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4762 during host lookups.
4764 74. The defer_ok option of callout verification was not working if it was used
4765 when verifying addresses in header lines, that is, for this case:
4767 verify = header_sender/callout=defer_ok
4769 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4770 those file descriptors could be used for SMTP connections. If anything
4771 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4772 could be sent to the SMTP client, causing chaos. The daemon now opens
4773 stdin, stdout, and stderr to /dev/null when it puts itself into the
4776 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4777 log by default. The user can override this with suitable Perl magic.
4779 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4780 the log message when discard triggers, was not working for the DATA ACL or
4781 for the non-SMTP ACL.
4783 78. Error message wording change in sieve.c.
4785 79. If smtp_accept_max_per_host was set, the number of connections could be
4786 restricted to fewer than expected, because the daemon was trying to set up
4787 a new connection before checking whether the processes handling previous
4788 connections had finished. The check for completed processes is now done
4789 earlier. On busy systems, this bug wouldn't be noticed because something
4790 else would have woken the daemon, and it would have reaped the completed
4793 80. If a message was submitted locally by a user whose login name contained one
4794 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4795 It caused a spool format error. I have fixed the spool reading code. A
4796 related problem was that the "from" clause in the Received: line became
4797 illegal because of the space(s). It is now covered by ${quote_local_part.
4799 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4802 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4804 83. Patch from Sieve maintainer to fix unterminated string problem in
4805 "vacation" handling.
4807 84. Some minor changes to the Linux configuration files to help with other
4808 OS variants using glibc.
4810 85. One more patch for Sieve to update vacation handling to latest spec.
4813 ----------------------------------------------------
4814 See the note above about the 4.44 and 4.50 releases.
4815 ----------------------------------------------------
4821 1. Change 4.43/35 introduced a bug that caused file counts to be
4822 incorrectly computed when quota_filecount was set in an appendfile
4825 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4826 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4829 3. Give more explanation in the error message when the command for a transport
4830 filter fails to execute.
4832 4. There are several places where Exim runs a non-Exim command in a
4833 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4834 was being done only for the command run by the queryprogram router. It is
4835 now done for all such subprocesses. The other cases are: ${run, transport
4836 filters, and the commands run by the lmtp and pipe transports.
4838 5. Some older OS have a limit of 256 on the maximum number of file
4839 descriptors. Exim was using setrlimit() to set 1000 as a large value
4840 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4841 systems. I've change it so that if it can't get 1000, it tries for 256.
4843 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4844 was an oversight, and furthermore, ever since the addition of extra
4845 controls (e.g. 4.43/32), the checks on when to allow different forms of
4846 "control" were broken. There should now be diagnostics for all cases when a
4847 control that does not make sense is encountered.
4849 7. $recipients is now available in the predata ACL (oversight).
4851 8. Tidy the search cache before the fork to do a delivery from a message
4852 received from the command line. Otherwise the child will trigger a lookup
4853 failure and thereby defer the delivery if it tries to use (for example) a
4854 cached ldap connection that the parent has called unbind on.
4856 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4857 of $address_data from the recipient verification was clobbered by the
4858 sender verification.
4860 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4861 was its contents. (It was OK if the option was not defined at all.)
4863 11. A "Completed" log line is now written for messages that are removed from
4864 the spool by the -Mrm option.
4866 12. $host_address is now set to the target address during the checking of
4867 ignore_target_hosts.
4869 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4870 being passed; this would have caused $sender_host_name to have been used if
4871 matching the list had actually called for a host name (not very likely,
4872 since this list is usually IP addresses). A host name is now passed as
4875 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4876 code that specifies a non-restarting handler (typically sigaction() in
4877 modern systems) in an attempt to fix a rare and obscure crash bug.
4879 15. Narrowed the window for a race in the daemon that could cause it to ignore
4880 SIGCHLD signals. This is not a major problem, because they are used only to
4881 wake it up if nothing else does.
4883 16. A malformed maildirsize file could cause Exim to calculate negative values
4884 for the mailbox size or file count. Odd effects could occur as a result.
4885 The maildirsize information is now recalculated if the size or filecount
4888 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4889 support for a long time. Removed HAVE_SYS_VFS_H.
4891 18. Updated exipick to current release from John Jetmore.
4893 19. Allow an empty sender to be matched against a lookup in an address list.
4894 Previously the only cases considered were a regular expression, or an
4897 20. Exim went into a mad DNS lookup loop when doing a callout where the
4898 host was specified on the transport, if the DNS lookup yielded more than
4901 21. The RFC2047 encoding function was originally intended for short strings
4902 such as real names; it was not keeping to the 75-character limit for
4903 encoded words that the RFC imposes. It now respects the limit, and
4904 generates multiple encoded words if necessary. To be on the safe side, I
4905 have increased the buffer size for the ${rfc2047: expansion operator from
4908 22. Failure to deliver a bounce message always caused it to be frozen, even if
4909 there was an errors_to setting on the router. The errors_to setting is now
4912 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4913 canonical form (fully expanded) before being placed in
4914 $sender_host_address.
4916 24. Updated eximstats to version 1.33
4918 25. Include certificate and key file names in error message when GnuTLS fails
4919 to set them up, because the GnuTLS error message doesn't include the name
4920 of the failing file when there is a problem reading it.
4922 26. Expand error message when OpenSSL has problems setting up cert/key files.
4925 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4926 (this can affect the format of dates).
4928 28. exim_tidydb, when checking for the continued existence of a message for
4929 which it has found a message-specific retry record, was not finding
4930 messages that were in split spool directories. Consequently, it was
4931 deleting retry records that should have stayed in existence.
4933 29. eximstats updated to version 1.35
4934 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4935 1.35 - bugfix such that pie charts by volume are generated correctly
4937 30. The SPA authentication driver was not abandoning authentication and moving
4938 on to the next authenticator when an expansion was forced to fail,
4939 contradicting the general specification for all authenticators. Instead it
4940 was generating a temporary error. It now behaves as specified.
4942 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4943 (the order specifies the preference for clients). The order is now AES256,
4944 AES128, 3DES, ARCFOUR128.
4946 31. Small patch to Sieve code - explicitly set From: when generating an
4949 32. Exim crashed if a remote delivery caused a very long error message to be
4950 recorded - for instance if somebody sent an entire SpamAssassin report back
4951 as a large number of 550 error lines. This bug was coincidentally fixed by
4952 increasing the size of one of Exim's internal buffers (big_buffer) that
4953 happened as part of the Exiscan merge. However, to be on the safe side, I
4954 have made the code more robust (and fixed the comments that describe what
4957 33. Some experimental protocols are using DNS PTR records for new purposes. The
4958 keys for these records are domain names, not reversed IP addresses. The
4959 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4960 leaves it alone. Component reversal etc. now happens only for IP addresses.
4963 34. The host_aton() function is supposed to be passed a string that is known
4964 to be a valid IP address. However, in the case of IPv6 addresses, it was
4965 not checking this. This is a hostage to fortune. Exim now panics and dies
4966 if the condition is not met. A case was found where this could be provoked
4967 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4968 components; fortuitously, this particular loophole had already been fixed
4969 by change 4.50/55 or 4.44/33 above.
4971 If there are any other similar loopholes, the new check in host_aton()
4972 itself should stop them being exploited. The report I received stated that
4973 data on the command line could provoke the exploit when Exim was running as
4974 exim, but did not say which command line option was involved. All I could
4975 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4976 running as the user.
4979 35. There was a buffer overflow vulnerability in the SPA authentication code
4980 (which came originally from the Samba project). I have added a test to the
4981 spa_base64_to_bits() function which I hope fixes it.
4984 36. The daemon start-up calls getloadavg() while still root for those OS that
4985 need the first call to be done as root, but it missed one case: when
4986 deliver_queue_load_max is set with deliver_drop_privilege. This is
4987 necessary for the benefit of the queue runner, because there is no re-exec
4988 when deliver_drop_privilege is set.
4990 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4991 was in use, was not putting the data itself into the right store pool;
4992 consequently, it could be overwritten for a subsequent message in the same
4993 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4996 38. Sometimes the final signoff response after QUIT could fail to get
4997 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4998 < 0 before doing a fflush(). This bug looks as though it goes back to the
4999 introduction of TLS in release 3.20, but "sometimes" must have been rare
5000 because the tests only now provoked it.
5006 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5007 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5008 Now it returns an integer. A number of calls were still expecting a T/F
5009 return. Fortuitously, in all cases, the tests worked in OK situations,
5010 which is the norm. However, things would have gone wrong on any write
5011 failures on the smtp file descriptor. This function is used when sending
5012 messages over SMTP and also when doing verify callouts.
5014 2. When Exim is called to do synchronous delivery of a locally submitted
5015 message (the -odf or -odi options), it no longer closes stderr before doing
5018 3. Implemented the mua_wrapper option.
5020 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5022 5. Implemented the functions header_remove(), header_testname(),
5023 header_add_at_position(), and receive_remove_recipient(), and exported them
5026 6. If an ACL "warn" statement specified the addition of headers, Exim already
5027 inserted X-ACL-Warn: at the start if there was no header name. However, it
5028 was not making this test for the second and subsequent header lines if
5029 there were newlines in the string. This meant that an invalid header could
5030 be inserted if Exim was badly configured.
5032 7. Allow an ACL "warn" statement to add header lines at the start or after all
5033 the Received: headers, as well as at the end.
5035 8. Added the rcpt_4xx retry error code.
5037 9. Added postmaster_mailfrom=xxx to callout verification option.
5039 10. Added mailfrom=xxxx to the callout verification option, for verify=
5042 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5043 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5045 12. Inserted some casts to stop certain compilers warning when using pointer
5046 differences as field lengths or precisions in printf-type calls (mostly
5047 affecting debugging statements).
5049 13. Added optional readline() support for -be (dynamically loaded).
5051 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5052 same clock tick as a message's arrival, so that its received time was the
5053 same as the "first fail" time on the retry record, and that message
5054 remained on the queue past the ultimate address timeout, every queue runner
5055 would try a delivery (because it was past the ultimate address timeout) but
5056 after another failure, the ultimate address timeout, which should have then
5057 bounced the address, did not kick in. This was a "< instead of <=" error;
5058 in most cases the first failure would have been in the next clock tick
5059 after the received time, and all would be well.
5061 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5062 being recognized when the domain list was tested by the match_domain
5063 condition in an expansion string.
5065 16. Added the ${str2b64: operator.
5067 17. Exim was always calling setrlimit() to set a large limit for the number of
5068 processes, without checking whether the existing limit was already
5069 adequate. (It did check for the limit on file descriptors.) Furthermore,
5070 errors from getrlimit() and setrlimit() were being ignored. Now they are
5071 logged to the main and panic logs, but Exim does carry on, to try to do its
5072 job under whatever limits there are.
5074 18. Imported PCRE 5.0.
5076 19. Trivial typo in log message " temporarily refused connection" (the leading
5079 20. If the log selector return_path_on_delivery was set and an address was
5080 redirected to /dev/null, the delivery process crashed because it assumed
5081 that a return path would always be set for a "successful" delivery. In this
5082 case, the whole delivery is bypassed as an optimization, and therefore no
5085 21. Internal re-arrangement: the function for sending a challenge and reading
5086 a response while authentication was assuming a zero-terminated challenge
5087 string. It's now changed to take a pointer and a length, to allow for
5088 binary data in such strings.
5090 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5092 23. Exim was not respecting finduser_retries when seeking the login of the
5093 uid under which it was called; it was always trying 10 times. (The default
5094 setting of finduser_retries is zero.) Also, it was sleeping after the final
5095 failure, which is pointless.
5097 24. Implemented tls_on_connect_ports.
5099 25. Implemented acl_smtp_predata.
5101 26. If the domain in control=submission is set empty, Exim assumes that the
5102 authenticated id is a complete email address when it generates From: or
5103 Sender: header lines.
5105 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5106 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5107 chown and chgrp in /bin and hostname in /usr/bin.
5109 28. Exim was keeping the "process log" file open after each use, just as it
5110 does for the main log. This opens the possibility of it remaining open for
5111 long periods when the USR1 signal hits a daemon. Occasional processlog
5112 errors were reported, that could have been caused by this. Anyway, it seems
5113 much more sensible not to leave this file open at all, so that is what now
5116 29. The long-running daemon process does not normally write to the log once it
5117 has entered its main loop, and it closes the log before doing so. This is
5118 so that log files can straightforwardly be renamed and moved. However,
5119 there are a couple of unusual error situations where the daemon does write
5120 log entries, and I had neglected to close the log afterwards.
5122 30. The text of an SMTP error response that was received during a remote
5123 delivery was being truncated at 512 bytes. This is too short for some of
5124 the long messages that one sometimes sees. I've increased the limit to
5127 31. It is now possible to make retry rules that apply only when a message has a
5128 specific sender, in particular, an empty sender.
5130 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5131 it possible to be selective about when SMTP synchronization is enforced.
5133 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5135 32. Implemented hosts_connection_nolog.
5137 33. Added an ACL for QUIT.
5139 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5142 35. Added mailbox_size and mailbox_filecount to appendfile.
5144 36. Added control = no_multiline_responses to ACLs.
5146 37. There was a bug in the logic of the code that waits for the clock to tick
5147 in the case where the clock went backwards by a substantial amount such
5148 that the microsecond fraction of "now" was more than the microsecond
5149 fraction of "then" (but the whole seconds number was less).
5151 38. Added support for the libradius Radius client library this is found on
5152 FreeBSD (previously only the radiusclient library was supported).
5158 1. When certain lookups returned multiple values in the form name=value, the
5159 quoting of the values was not always being done properly. Specifically:
5160 (a) If the value started with a double quote, but contained no whitespace,
5162 (b) If the value contained whitespace other than a space character (i.e.
5163 tabs or newlines or carriage returns) it was not quoted.
5164 This fix has been applied to the mysql and pgsql lookups by writing a
5165 separate quoting function and calling it from the lookup code. The fix
5166 should probably also be applied to nisplus, ibase and oracle lookups, but
5167 since I cannot test any of those, I have not disturbed their existing code.
5169 2. A hit in the callout cache for a specific address caused a log line with no
5170 reason for rejecting RCPT. Now it says "Previous (cached) callout
5171 verification failure".
5173 3. There was an off-by-one bug in the queryprogram router. An over-long
5174 return line was truncated at 256 instead of 255 characters, thereby
5175 overflowing its buffer with the terminating zero. As well as fixing this, I
5176 have increased the buffer size to 1024 (and made a note to document this).
5178 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5179 when Exim is waiting for an SMTP response from a remote server, Exim
5180 restarts its select() call on the socket, thereby resetting its timeout.
5181 This is not a problem when such interrupts are rare. Somebody set up a cron
5182 job to run exiwhat every 2 minutes, which is less than the normal select()
5183 timeout (5 or 10 minutes). This meant that the select() timeout never
5184 kicked in because it was always reset. I have fixed this by comparing the
5185 time when an interrupt arrives with the time at the start of the first call
5186 to select(). If more time than the timeout has elapsed, the interrupt is
5187 treated as a timeout.
5189 5. Some internal re-factoring in preparation for the addition of Sieve
5190 extensions (by MH). In particular, the "personal" test is moved to a
5191 separate function, and given an option for scanning Cc: and Bcc: (which is
5192 not set for Exim filters).
5194 6. When Exim created an email address using the login of the caller as the
5195 local part (e.g. when creating a From: or Sender: header line), it was not
5196 quoting the local part when it contained special characters such as @.
5198 7. Installed new OpenBSD configuration files.
5200 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5201 try to make them clearer.
5203 9. Callout options, other than the timeout value, were being ignored when
5204 verifying sender addresses in header lines. For example, when using
5206 verify = header_sender/callout=no_cache
5208 the cache was (incorrectly) being used.
5210 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5211 only the Cygwin environment.
5213 11. When return_path_on_delivery was set as a log selector, if different remote
5214 addresses in the same message used different return paths and parallel
5215 remote delivery occurred, the wrong values would sometimes be logged.
5216 (Whenever a remote delivery process finished, the return path value from
5217 the most recently started remote delivery process was logged.)
5219 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5220 lines when AUTH and/or TLS are in use. This is the "received protocol"
5221 field. Exim used to use "asmtp" for authenticated SMTP, without any
5222 indication (in the protocol name) for TLS use. Now it follows the RFC and
5223 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5224 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5225 names appear in log lines as well as in Received: header lines.
5227 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5228 extensions, and comparison tests, and to fix some bugs.
5230 14. Changes to the "personal" filter test:
5232 (1) The test was buggy in that it was just doing the equivalent of
5233 "contains" tests on header lines. For example, if a user's address was
5234 anne@some.where, the "personal" test would incorrectly be true for
5236 To: susanne@some.where
5238 This test is now done by extracting each address from the header in turn,
5239 and checking the entire address. Other tests that are part of "personal"
5240 are now done using regular expressions (for example, to check local parts
5241 of addresses in From: header lines).
5243 (2) The list of non-personal local parts in From: addresses has been
5244 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5245 taken from the Sieve specification recommendations.
5247 (3) If the message contains any header line starting with "List-" it is
5248 treated as non-personal.
5250 (4) The test for "circular" in the Subject: header line has been removed
5251 because it now seems ill-conceived.
5253 15. Minor typos in src/EDITME comments corrected.
5255 16. Installed latest exipick from John Jetmore.
5257 17. If headers_add on a router specified a text string that was too long for
5258 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5259 of string_sprintf() is now avoided.
5261 18. $message_body_size was not set (it was always zero) when running the DATA
5262 ACL and the local_scan() function.
5264 19. For the "mail" command in an Exim filter, no default was being set for
5265 the once_repeat time, causing a random time value to be used if "once" was
5266 specified. (If the value happened to be <= 0, no repeat happened.) The
5267 default is now 0s, meaning "never repeat". The "vacation" command was OK
5268 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5269 (I found it when inspecting the code).
5271 20. There is now an overall timeout for performing a callout verification. It
5272 defaults to 4 times the callout timeout, which applies to individual SMTP
5273 commands during the callout. The overall timeout applies when there is more
5274 than one host that can be tried. The timeout is checked before trying the
5275 next host. This prevents very long delays if there are a large number of
5276 hosts and all are timing out (e.g. when the network connections are timing
5277 out). The value of the overall timeout can be changed by specifying an
5278 additional sub-option for "callout", called "maxwait". For example:
5280 verify = sender/callout=5s,maxwait=20s
5282 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5283 to the end before writing, but this should make it even safer).
5285 22. Exim was forgetting that it had advertised PIPELINING for the second and
5286 subsequent messages on an SMTP connection. It was also not resetting its
5287 memory on STARTTLS and an internal HELO.
5289 23. When Exim logs an SMTP synchronization error within a session, it now
5290 records whether PIPELINING has been advertised or not.
5292 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5293 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5294 rather than long int.
5296 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5298 26. Added the never_mail option to autoreply.
5304 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5305 crash if the getsockname() call failed; this can happen if a connection is
5306 closed very soon after it is established. The problem was simply in the
5307 order in which certain operations were done, causing Exim to try to write
5308 to the SMTP stream before it had set up the file descriptor. The bug has
5309 been fixed by making things happen in the correct order.
5315 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5316 before the connection was closed, thus losing the rejection response.
5318 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5319 some early Solaris releases, but causes trouble in current releases where
5320 socklen_t is defined.
5322 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5325 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5328 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5329 that releasing the top of it at the end releases what was used for sub-
5330 expansions (unless the block got too big). However, discard this block if
5331 the first thing is a variable or header, so that we can use its block when
5332 it is dynamic (useful for very large $message_headers, for example).
5334 6. Lookups now cache *every* query, not just the most recent. A new, separate
5335 store pool is used for this. It can be recovered when all lookup caches are
5336 flushed. Lookups now release memory at the end of their result strings.
5337 This has involved some general refactoring of the lookup sources.
5339 7. Some code has been added to the store_xxx() functions to reduce the amount
5340 of flapping under certain conditions.
5342 8. log_incoming_interface used to affect only the <= reception log lines. Now
5343 it causes the local interface and port to be added to several more SMTP log
5344 lines, for example "SMTP connection from", and rejection lines.
5346 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5348 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5350 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5351 could be overwritten at the end of the current message (or the start of a
5352 new message if it was set in a HELO ACL). The value is now preserved for
5353 the duration of the SMTP connection.
5355 12. If a transport had a headers_rewrite setting, and a matching header line
5356 contained an unqualified address, that address was qualified, even if it
5357 did not match any rewriting rules. The underlying bug was that the values
5358 of the flags that permit the existence of unqualified sender and recipient
5359 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5360 non-local messages, and by -bnq for local messages) were not being
5361 preserved with the message after it was received.
5363 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5364 "next input=" as part of the text comprising the host identity instead of
5365 the correct text. The code was using the same buffer for two different
5366 strings. However, depending on which order the printing function evaluated
5367 its arguments, the bug did not always show up. Under Linux, for example, my
5368 test suite worked just fine.
5370 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5371 doesn't work with some older versions of Perl. It has been changed to "my",
5372 which in any case is probably the better facility to use.
5374 15. A really picky compiler found some instances of statements for creating
5375 error messages that either had too many or two few arguments for the format
5378 16. The size of the buffer for calls to the DNS resolver has been increased
5379 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5380 for addresses that have a lot of PTR records. This alleviates a problem; it
5381 does not fully solve it.
5383 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5384 buffer now truncates the list and logs the incident, which is the same
5385 action as happens when Exim is looking up a host name and its aliases.
5386 Previously in this situation something unpredictable would happen;
5387 sometimes it was "internal error: store_reset failed".
5389 18. If a server dropped the connection unexpectedly when an Exim client was
5390 using GnuTLS and trying to read a response, the client delivery process
5391 crashed while trying to generate an error log message.
5393 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5394 string, for example:
5396 warn message = H1: something\nH2: something
5398 the text was added as a single header line from Exim's point of view
5399 though it ended up OK in the delivered message. However, searching for the
5400 second and subsequent header lines using $h_h2: did not work. This has been
5401 fixed. Similarly, if a system filter added multiple headers in this way,
5402 the routers could not see them.
5404 20. Expanded the error message when iplsearch is called with an invalid key to
5405 suggest using net-iplsearch in a host list.
5407 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5408 ACLs are no longer actually imposed (and a message to that effect is
5411 22. If a "gecos" field in a passwd entry contained escaped characters, in
5412 particular, if it contained a \" sequence, Exim got it wrong when building
5413 a From: or a Sender: header from that name. A second bug also caused
5414 incorrect handling when an unquoted " was present following a character
5415 that needed quoting.
5417 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5418 was not being matched caselessly.
5420 24. Arranged for all hyphens in the exim.8 source to be escaped with
5423 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5424 properly. Recipient callout cache records were still being keyed to include
5425 the sender, even when use_sender was set false. This led to far more
5426 callouts that were necessary. The sender is no longer included in the key
5427 when use_sender is false.
5429 26. Added "control = submission" modifier to ACLs.
5431 27. Added the ${base62d: operator to decode base 62 numbers.
5433 28. dnsdb lookups can now access SRV records.
5435 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5436 the configuration file.
5438 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5439 (-v) mode. This makes the output for a verbose queue run more intelligible.
5441 31. Added a use_postmaster feature to recipient callouts.
5443 32. Added the $body_zerocount variable, containing the number of binary zero
5444 bytes in the message body.
5446 33. The time of last modification of the "new" subdirectory is now used as the
5447 "mailbox time last read" when there is a quota error for a maildir
5450 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5452 35. Added +ignore_unknown as a special item in host lists.
5454 36. Code for decoding IPv6 addresses in host lists is now included, even if
5455 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5456 address was recognized as an IP address, but was then not correctly decoded
5457 into binary, causing unexpected and incorrect effects when compared with
5464 1. Very minor rewording of debugging text in manualroute to say "list of
5465 hosts" instead of "hostlist".
5467 2. If verify=header_syntax was set, and a header line with an unqualified
5468 address (no domain) and a large number of spaces between the end of the
5469 name and the colon was received, the reception process suffered a buffer
5470 overflow, and (when I tested it) crashed. This was caused by some obsolete
5471 code that should have been removed. The fix is to remove it!
5473 3. When running in the test harness, delay a bit after writing a bounce
5474 message to get a bit more predictability in the log output.
5476 4. Added a call to search_tidyup() just before forking a reception process. In
5477 theory, someone could use a lookup in the expansion of smtp_accept_max_
5478 per_host which, without the tidyup, could leave open a database connection.
5480 5. Added the variables $recipient_data and $sender_data which get set from a
5481 lookup success in an ACL "recipients" or "senders" condition, or a router
5482 "senders" option, similar to $domain_data and $local_part_data.
5484 6. Moved the writing of debug_print from before to after the "senders" test
5487 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5488 problems for message scanning, either using a data ACL, or using
5489 local_scan() because the Received: header was not generated till after they
5490 were called (in order to set the time as the time of reception completion).
5491 I have revised the way this works. The header is now generated after the
5492 body is received, but before the ACL or local_scan() are called. After they
5493 are run, the timestamp in the header is updated.
5499 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5500 before starting a queue runner without re-exec. This happened only when
5501 deliver_drop_privilege was set or when the Exim user was set to root. The
5502 effect of the bug was that timeouts during subsequent deliveries caused
5503 crashes instead of being properly handled. The handler is now left at its
5504 default (and expected) setting.
5506 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5507 message, again when deliver_drop_privilege is set or Exim is run as root.
5508 The bug described in (1) was not present in this case, but the tidying up
5509 of the other signals was missing. I have made the two cases consistent.
5511 3. The ignore_target_hosts setting on a manualroute router was being ignored
5512 for hosts that were looked up using the /MX notation.
5514 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5517 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5518 operated on the sender address. After changing the $sender_address to <>
5519 for the sender address verify, Exim was re-instated it as the original
5520 (before rewriting) address, but remembering that it had rewritten it, so it
5521 wasn't rewriting it again. This bug also had the effect of breaking the
5522 sender address verification caching when the sender address was rewritten.
5524 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5525 This has been changed so that if the ip literal address matches
5526 ignore_target_hosts, the router declines.
5528 7. Added expansion conditions match_domain, match_address, and match_local_
5529 part (NOT match_host).
5531 8. The placeholder for the Received: header didn't have a length field set.
5533 9. Added code to Exim itself and to exim_lock to test for a specific race
5534 condition that could lead to file corruption when using MBX delivery. The
5535 issue is with the lockfile that is created in /tmp. If this file is removed
5536 after a process has opened it but before that process has acquired a lock,
5537 there is the potential for a second process to recreate the file and also
5538 acquire a lock. This could lead to two Exim processes writing to the file
5539 at the same time. The added code performs the same test as UW imapd; it
5540 checks after acquiring the lock that its file descriptor still refers to
5541 the same named file.
5543 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5544 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5547 11. Added the smtp_active_hostname option. If used, this will typically be made
5548 to depend on the incoming interface address. Because $interface_address is
5549 not set up until the daemon has forked a reception process, error responses
5550 that can happen earlier (such as "too many connections") no longer contain
5553 12. If an expansion in a condition on a "warn" statement fails because a lookup
5554 defers, the "warn" statement is abandoned, and the next ACL statement is
5555 processed. Previously this caused the whole ACL to be aborted.
5557 13. Added the iplsearch lookup type.
5559 14. Added ident_timeout as a log selector.
5561 15. Added tls_certificate_verified as a log selector.
5563 16. Added a global option tls_require_ciphers (compare the smtp transport
5564 option of the same name). This controls incoming TLS connections.
5566 17. I finally figured out how to make tls_require_ciphers do a similar thing
5567 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5568 before starting the TLS session.
5570 18. Tabs are now shown as \t in -bP output.
5572 19. If the log selector return_path_on_delivery was set, Exim crashed when
5573 bouncing a message because it had too many Received: header lines.
5575 20. If two routers both had headers_remove settings, and the first one included
5576 a superfluous trailing colon, the final name in the first list and the
5577 first name in the second list were incorrectly joined into one item (with a
5578 colon in the middle).
5584 1. Added -C and -D options to the exinext utility, mainly to make it easier
5585 to include in the automated testing, but these could be helpful when
5586 multiple configurations are in use.
5588 2. The exinext utility was not formatting the output nicely when there was
5589 an alternate port involved in the retry record key, nor when there was a
5590 message id as well (for retries that were specific to a specific message
5591 and a specific host). It was also confused by IPv6 addresses, because of
5592 the additional colons they contain. I have fixed the IPv4 problem, and
5593 patched it up to do a reasonable job for IPv6.
5595 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5596 delivery, the log line now contains "pipelined" if PIPELINING was used.
5598 4. An SMTP transport process used to panic and die if the bind() call to set
5599 an explicit outgoing interface failed. This has been changed; it is now
5600 treated in the same way as a connect() failure.
5602 5. A reference to $sender_host_name in the part of a conditional expansion
5603 that was being skipped was still causing a DNS lookup. This no longer
5606 6. The def: expansion condition was not recognizing references to header lines
5607 that used bh_ and bheader_.
5609 7. Added the _cache feature to named lists.
5611 8. The code for checking quota_filecount in the appendfile transport was
5612 allowing one more file than it should have been.
5614 9. For compatibility with Sendmail, the command line option
5622 and sets the incoming protocol and host name (for trusted callers). The
5623 host name and its colon can be omitted when only the protocol is to be set.
5624 Note the Exim already has two private options, -pd and -ps, that refer to
5625 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5626 "s", but I don't think that's a major issue.
5628 10. A number of refactoring changes to the code, none of which should affect
5631 (a) The number of logging options was getting close to filling up the
5632 32-bit word that was used as a bit map. I have split them into two classes:
5633 those that are passed in the argument to log_write(), and those that are
5634 only ever tested independently outside of that function. These are now in
5635 separate 32-bit words, so there is plenty of room for expansion again.
5636 There is no change in the user interface or the logging behaviour.
5638 (b) When building, for example, log lines, the code previously used a
5639 macro that called string_cat() twice, in order to add two strings. This is
5640 not really sufficiently general. Furthermore, there was one instance where
5641 it was actually wrong because one of the argument was used twice, and in
5642 one call a function was used. (As it happened, calling the function twice
5643 did not affect the overall behaviour.) The macro has been replaced by a
5644 function that can join an arbitrary number of extra strings onto a growing
5647 (c) The code for expansion conditions now uses a table and a binary chop
5648 instead of a serial search (which was left over from when there were very
5649 few conditions). Also, it now recognizes conditions like "pam" even when
5650 the relevant support is not compiled in: a suitably worded error message is
5651 given if an attempt is made to use such a condition.
5653 11. Added ${time_interval:xxxxx}.
5655 12. A bug was causing one of the ddress fields not to be passed back correctly
5656 from remote delivery subprocesses. The field in question was not being
5657 subsequently used, so this caused to problems in practice.
5659 13. Added new log selectors queue_time and deliver_time.
5661 14. Might have fixed a bug in maildirsizefile handling that threw up
5662 "unexpected character" debug warnings, and recalculated the data
5663 unnecessarily. In any case, I expanded the warning message to give more
5666 15. Added the message "Restricted characters in address" to the statements in
5667 the default ACL that block characters like @ and % in local parts.
5669 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5670 Three changes have been made:
5672 (a) There was a serious bug; a negative response to MAIL caused the whole
5673 recipient domain to be cached as invalid, thereby blocking all messages
5674 to all local parts at the same domain, from all senders. This bug has
5675 been fixed. The domain is no longer cached after a negative response to
5676 MAIL if the sender used is not empty.
5678 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5681 (c) A new callout option, "use_sender" has been added for people who want
5682 the modified behaviour.
5688 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5691 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5692 indeed breaks things for older releases.
5694 3. Added additional logging to the case where there is a problem reading data
5695 from a filter that is running in a subprocess using a pipe, in order to
5696 try to track down a specific problem.
5698 4. Testing facility fudge: when running in the test harness and attempting
5699 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5700 getting "No route to host". Convert this to a timeout.
5702 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5705 6. Some OS don't have socklen_t but use size_t instead. This affects the
5706 fifth argument of getsockopt() amongst other things. This is now
5707 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5708 can be set for individual OS. I have set it for SunOS5, OSF1, and
5709 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5710 some earlier ones do not.
5712 7. Change 4.30/15 was not doing the test caselessly.
5714 8. The standard form for an IPv6 address literal was being rejected by address
5715 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5716 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5717 this, as well as the form without the "IPv6" on the front (but only when
5718 address literals are enabled, of course).
5720 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5722 10. Exim crashed if a message with an empty sender address specified by -f
5723 encountered a router with an errors_to setting. This could be provoked only
5724 by a command such as
5728 where an empty string was supplied; "<>" did not hit this bug.
5730 11. Installed PCRE release 4.5.
5732 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5733 remained set. It is now erased.
5735 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5736 times from message ids (which are base 36 rather than the normal 62).
5738 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5739 were being counted as actual protocol errors, and logged if the log
5740 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5741 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5742 and DATA following a set of rejected RCPTs do not count as protocol errors.
5743 In other words, Exim assumes they were pipelined, though this may not
5744 actually be the case. Of course, in all cases the client gets an
5745 appropriate error code.
5747 15. If a lookup fails in an ACL condition, a message about the failure may
5748 be available; it is used if testing the ACL cannot continue, because most
5749 such messages specify what the cause of the deferral is. However, some
5750 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5751 that caused an old message to be retained and used if a later statement
5752 caused a defer, replacing the real cause of the deferral.
5754 16. If an IP address had so many PTR records that the DNS lookup buffer
5755 was not large enough to hold them, Exim could crash while trying to process
5756 the truncated data. It now detects and logs this case.
5758 17. Further to 4.21/58, another change has been made: if (and only if) the
5759 first line of a message (the first header line) ends with CRLF, a bare LF
5760 in a subsequent header line has a space inserted after it, so as not to
5761 terminate the header.
5763 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5764 unnecessarily, used atoi() instead of strtol(), and didn't check the
5765 termination when getting file sizes from file names by regex.
5767 19. Completely re-implemented the support for maildirsize files, in the light
5768 of a number of problems with the previous contributed implementation
5769 (4.30/29). In particular:
5771 . If the quota is zero, the maildirsize file is maintained, but no quota is
5774 . If the maildir directory does not exist, it is created before any attempt
5775 to write a maildirsize file.
5777 . The quota value in the file is just a cache; if the quota is changed in
5778 the transport, the new value overrides.
5780 . A regular expression is available for excluding directories from the
5783 20. The autoreply transport checks the characters in options that define the
5784 message's headers; it allows continued headers, but it was checking with
5785 isspace() after an embedded newline instead of explicitly looking for a
5788 21. If all the "regular" hosts to which an address was routed had passed their
5789 expiry times, and had not reached their retry times, the address was
5790 bounced, even if fallback hosts were defined. Now Exim should go on to try
5793 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5794 equivalent code in the SMTP transport. Some hosts send humungous responses
5795 to HELO/EHLO, more than 1024 it seems.
5797 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5798 gives compiler warnings in some environments. I've now done it "properly",
5801 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5802 (because of problems with the built-in one) was declared to return uschar *
5803 instead of char *, causing compiler failure.
5805 25. Fixed a file descriptor leak when processing alias/forward files.
5807 26. Fixed a minor format string issue in dbfn.c.
5809 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5811 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5812 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5815 29. When a sender address is verified, it is cached, to save repeating the test
5816 when there is more than one recipient in a message. However, when the
5817 verification involves a callout, it is possible for different callout
5818 options to be set for different recipients. It is too complicated to keep
5819 track of this in the cache, so now Exim always runs a verification when a
5820 callout is required, relying on the callout cache for the optimization.
5821 The overhead is duplication of the address routing, but this should not be
5824 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5825 to be verified with callout=postmaster, and the main callout worked but the
5826 postmaster check failed, the verification correctly failed. However, if a
5827 subsequent RCPT command asked for sender verification *without* the
5828 postmaster check, incorrect caching caused this verification also to fail,
5831 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5832 it was not caching the DNS options (qualify_single, search_parents) that
5833 were used when the lookup failed. A subsequent lookup with different
5834 options therefore always gave the same answer, though there were cases
5835 where it should not have. (Example: a "domains = !$mx_any" option on a
5836 dnslookup router: the "domains" option is always processed without any
5837 widening, but the router might have qualify_single set.) Now Exim uses the
5838 cached value only when the same options are set.
5840 32. Added John Jetmore's "exipick" utility to the distribution.
5842 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5843 than a timeout (e.g. a certificate is required, and is not provided), an
5844 Exim server now closes the connection immediately. Previously it waited for
5845 the client to close - but if the client is SSL, it seems that they each
5846 wait for each other, leading to a delay before one of them times out.
5848 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5849 maintained 0.8.x compatibility because I don't think many are using it, and
5850 it is clearly obsolete.
5852 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5855 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5856 client certificate was expired. A simple patch fixes this, though I don't
5857 understand the full logic of why the verify callback is called multiple
5860 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5861 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5862 which causes problems with some clients (such as the Certicom SSL Plus
5863 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5864 disables the coutermeasure allowing Eudora to connect."
5866 38. Exim was not checking that a write() to a log file succeeded. This could
5867 lead to Bad Things if a log got too big, in particular if it hit a file
5868 size limit. Exim now panics and dies if it cannot write to a log file, just
5869 as it does if it cannot open a log file.
5871 39. Modified OS/Makefile-Linux so that it now contains
5873 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5875 The two -D definitions ensure that Exim is compiled with large file
5876 support, which makes it possible to handle log files that are bigger than
5879 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5880 instance) a domain was checked against a named list that involved a lookup,
5881 causing $domain_data to be set, then another domain was checked against the
5882 same list, then the first domain was re-checked, the value of $domain_data
5883 after the final check could be wrong. In particular, if the second check
5884 failed, it could be set empty. This bug probably also applied to
5887 41. The strip_trailing_dot option was not being applied to the address given
5888 with the -f command-line option.
5890 42. The code for reading a message's header from the spool was incrementing
5891 $received_count, but never initializing it. This meant that the value was
5892 incorrect (doubled) while delivering a message in the same process in which
5893 it was received. In the most common configuration of Exim, this never
5894 happens - a fresh exec is done - but it can happen when
5895 deliver_drop_privilege is set.
5897 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5898 it now includes up to 150 characters of the unexpected data in the log
5901 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5902 and building data strings. The size of both of these buffers was 10 000
5903 bytes - far larger than anybody would *ever* want, thought I. Needless to
5904 say, somebody hit the limit. I have increased the maximum line length to
5905 20 000 and the maximum data length of concatenated lines to 100 000. I have
5906 also fixed two bugs, because there was no checking on these buffers. Tsk,
5907 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5908 buffer is too small.
5910 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5911 lsearch lookups. Now it does.
5913 46. When parsing a route_list item in a manualroute router, a fixed-length
5914 buffer was used for the list of hosts. I made this 1024 bytes long,
5915 thinking that nobody would ever have a list of hosts that long. Wrong.
5916 Somebody had a whole pile of complicated expansion conditions, and the
5917 string was silently truncated, leading to an expansion error. It turns out
5918 that it is easier to change to an unlimited length (owing to other changes
5919 that have happened since this code was originally written) than to build
5920 structure for giving a limitation error. The length of the item that
5921 expands into the list of hosts is now unlimited.
5923 47. The lsearch lookup could not handle data where the length of text line was
5924 more than 4095 characters. Such lines were truncated, leading to shortened
5925 data being returned. It should now handle lines of any length.
5927 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5928 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5931 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5932 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5933 no longer happens in comment lines. A list of replacements is now placed
5934 at the head of all of the source files, except those whose only change is
5935 to replace PERL_COMMAND in the very first #! line.
5937 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5938 messages on the queue, with a bottom-up merge sort, using code contributed
5939 by Michael Haardt. This should make operations like -bp somewhat faster on
5940 large queues. It won't affect queue runners, except when queue_run_in_order
5943 51. Installed eximstats 1.31 in the distribution.
5945 52. Added support for SRV lookups to the dnslookup router.
5947 53. If an ACL referred to $message_body or $message_body_end, the value was not
5948 reset for any messages that followed in the same SMTP session.
5950 54. The store-handling optimization for building very long strings was not
5951 differentiating between the different store pools. I don't think this
5952 actually made any difference in practice, but I've tidied it.
5954 55. While running the routers to verify a sender address, $sender_address
5955 was still set to the sender address. This is wrong, because when routing to
5956 send a bounce to the sender, it would be empty. Therefore, I have changed
5957 it so that, while verifying a sender address, $sender_address is set to <>.
5958 (There is no change to what happens when verifying a recipient address.)
5960 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5961 target A or AAAA records (if not already returned) without resetting the
5962 qualify_single or search_parents options of the DNS resolver. These are
5963 inappropriate in this case because the targets of MX and SRV records must
5964 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5965 target that, when qualified, matched something in the local domain. These
5966 two options are now turned off when doing these lookups.
5968 57. It seems that at least some releases of Reiserfs (which does not have the
5969 concept of a fixed number of inodes) returns zero and not -1 for the
5970 number of available inodes. This interacted badly with check_spool_inodes,
5971 which assumed that -1 was the "no such thing" setting. What I have done is
5972 to check that the total number of inodes is greater than zero before doing
5973 the test of how many are available.
5975 58. When a "warn" ACL statement has a log_message modifier, the message is
5976 remembered, and not repeated. This is to avoid a lot of repetition when a
5977 message has many recipients that cause the same warning to be written.
5978 However, Exim was preserving the list of already written lines for an
5979 entire SMTP session, which doesn't seem right. The memory is now reset if a
5980 new message is started.
5982 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5983 debugging output unless log_rewrite was also set.
5985 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5986 of (int)(handle) when we know that handle contains (void *)(-1).
5988 61. The Exim daemon panic-logs an error return when it closes the incoming
5989 connection. However "connection reset by peer" seems to be common, and
5990 isn't really an error worthy of noting specially, so that particular error
5993 62. When Exim is trying to find all the local interfaces, it used to panic and
5994 die if the ioctl to get the interface flags failed. However, it seems that
5995 on at least one OS (Solaris 9) it is possible to have an interface that is
5996 included in the list of interfaces, but for which you get a failure error
5997 for this call. This happens when the interface is not "plumbed" into a
5998 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5999 failure of the "get flags" call assumes that the interface is down.
6001 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6002 makes life easier for people who are doing arithmetic on fields extracted
6003 from dates, where you often get leading zeros that should not be
6004 interpreted as octal.
6006 64. Added qualify_domain to the redirect router, to override the global
6009 65. If a pathologically long header line contained very many addresses (the
6010 report of this problem mentioned 10 000) and each of them was rewritten,
6011 Exim could use up a very large amount of memory. (It kept on making new
6012 copies of the header line as it rewrote, and never released the old ones.)
6013 At the expense of a bit more processing, the header rewriting function has
6014 been changed so that it no longer eats memory in this way.
6016 66. The generation of the Received: header has been moved from the time that a
6017 message starts to be received, to the time that it finishes. The timestamp
6018 in the Received: header should now be very close to that of the <= log
6019 line. There are two side-effects of this change:
6021 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6022 logged header lines no longer include the local Received: line, because
6023 it has not yet been created. The same applies to a copy of the message
6024 that is returned to a non-SMTP sender when a message is rejected.
6026 (b) When a filter file is tested using -bf, no additional Received: header
6027 is added to the test message. After some thought, I decided that this
6030 This change does not affect the value of $received_for. It is still set
6031 after address rewriting, but before local_scan() is called.
6033 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6035 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6036 gave an unhelpful panic error message, and a defer error. I have managed to
6037 change this behaviour so that it now rejects any supplied certificate,
6038 which seems right, as the list of acceptable certificates is empty.
6040 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6041 gave an unhelpful defer error. I have not managed to make this reject any
6042 supplied certificates, but the error message it gives is "no certificate
6043 supplied", which is not helpful.
6045 70. exigrep's output now also includes lines that are not associated with any
6046 message, but which match the given pattern. Implemented by a patch from
6047 Martin Sluka, which also tidied up the Perl a bit.
6049 71. Recipient callout verification, like sender verification, was using <> in
6050 the MAIL FROM command. This isn't really the right thing, since the actual
6051 sender may affect whether the remote host accepts the recipient or not. I
6052 have changed it to use the actual sender in the callout; this means that
6053 the cache record is now keyed on a recipient/sender pair, not just the
6054 recipient address. There doesn't seem to be a real danger of callout loops,
6055 since a callout by the remote host to check the sender would use <>.
6056 [SEE ABOVE: changed after hitting problems.]
6058 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6059 temporary errors. However, in the case of such a code being given after
6060 the end of a data transmission (i.e. after ".") Exim was failing to write
6061 a retry record for the message. (Yes, there was some broken host that was
6062 actually sending 8xx at this point.)
6064 73. An unknown lookup type in a host list could cause Exim to panic-die when
6065 the list was checked. (An example that provoked this was putting <; in the
6066 middle of a list instead of at the start.) If this happened during a DATA
6067 ACL check, a -D file could be left lying around. This kind of configuration
6068 error no longer causes Exim to die; instead it causes a defer error. The
6069 incident is still logged to the main and panic logs.
6071 74. Buglet left over from Exim 3 conversion. The message "too many messages
6072 in one connection" was written to the rejectlog but not the mainlog, except
6073 when address rewriting (yes!) was being logged.
6075 75. Added write_rejectlog option.
6077 76. When a system filter was run not as root (that is, when system_filter_user
6078 was set), the values of the $n variables were not being returned to the
6079 main process; thus, they were not subsequently available in the $sn
6082 77. Added +return_path_on_delivery log selector.
6084 78. A connection timeout was being treated differently from recipients deferred
6085 when testing hosts_max_try with a message that was older than the host's
6086 retry timeout. (The host should not be counted, thus allowing all hosts to
6087 be tried at least once before bouncing.) This may have been the cause of an
6088 occasionally reported bug whereby a message would remain on the queue
6089 longer than the retry timeout, but would be bounced if a delivery was
6090 forced. I say "may" because I never totally pinned down the problem;
6091 setting up timeout/retry tests is difficult. See also the next item.
6093 79. The ultimate address timeout was not being applied to errors that involved
6094 a combination of host plus message (for example, a timeout on a MAIL
6095 command). When an address resolved to a number of possible hosts, and they
6096 were not all tried for each delivery (e.g. because of hosts_max_try), a
6097 message could remain on the queue longer than the retry timeout.
6099 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6102 81. Fixed an obscure SMTP outgoing bug which required at least the following
6103 conditions: (a) there was another message waiting for the same server;
6104 (b) the server returned 5xx to all RCPT commands in the first message so
6105 that the message was not completed; (c) the server dropped the connection
6106 or gave a negative response to the RSET that Exim sends to abort the
6107 transaction. The observed case was a dropped connection after DATA that had
6108 been sent in pipelining mode. That is, the server had advertised PIPELINING
6109 but was not implementing it correctly. The effect of the bug was incorrect
6110 behaviour, such as trying another host, and this could lead to a crash.
6116 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6117 and daemon.c were passed as pointers to ints; they should have been
6118 pointers to socklen_t variables (which are typically unsigned ints).
6120 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6123 3. Fixed a really odd bug that affected only the testing scheme; patching a
6124 certain fixed string in the binary changed the value of another string that
6125 happened to be identical to the end of the original first string.
6127 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6128 name", it returns that address as the IP address. On some operating
6129 systems (e.g. Solaris), it also passes back the IP address string as the
6130 "host name". However, on others (e.g. Linux), it passes back an empty
6131 string. Exim wasn't checking for this, and was changing the host name to an
6132 empty string, assuming it had been canonicalized.
6134 5. Although rare, it is permitted to have more than one PTR record for a given
6135 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6136 all the names associated with an address, because they do in Solaris.
6137 However, it seems that they do not in Linux for data that comes from the
6138 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6139 I found this out when I moved to a new Linux workstation and tried to run
6140 the Exim test suite.
6142 To get round this problem I have changed the code so that it now does its
6143 own call to the DNS to look up PTR records when searching for a host name.
6144 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6145 addresses that are only in /etc/hosts are still found.
6147 This behaviour is, however, controlled by an option called host_lookup_
6148 order, which defaults to "bydns:byaddr". If people want to use the other
6149 order, or indeed, just use one or the other means of lookup, they can
6150 specify it in this variable.
6152 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6153 some operating systems, this comes back from gethostbyaddr() as an empty
6154 string, and this is what Exim used to test for. However, it seems that in
6155 other systems, "." is yielded. Exim now tests for this case too.
6157 7. The values of check_spool_space and check_log_space are now held internally
6158 as a number of kilobytes instead of an absolute number of bytes. If a
6159 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6160 kilobyte. This means that much larger values can be stored.
6162 8. Exim monitor: an attempt to get the action menu when not actually pointing
6163 at a message produces an empty menu entitled "No message selected". This
6164 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6165 no entries in it ("Shell widget menu has zero width and/or height"). So I
6166 have added a single, blank menu entry in this case.
6168 9. Added ${quote_local_part.
6170 10. MIME decoding is now applied to the contents of Subject: header lines when
6173 11. Now that a reference to $sender_host_address automatically causes a reverse
6174 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6175 host lookup before query-style lookups in lists that might use this
6176 variable. This has therefore been abolished, and the "net-" prefix is no
6177 longer necessary for query-style lookups.
6179 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6180 have been a typo for LFLAGS, so it has been changed.
6182 13. The install script calls Exim with "-C /dev/null" in order to find the
6183 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6184 to be output. However, since Exim outputs its version number before the
6185 error, it didn't break the script. It just looked ugly. I fixed this by
6186 always allowing "-C /dev/null" if the caller is root.
6188 14. Ignore overlarge ACL variable number when reading spool file - insurance
6189 against a later release with more variables having written the file.
6191 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6192 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6193 this, as well as the form without the "IPv6" on the front.
6195 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6196 OS/Makefile-Darwin file.
6198 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6199 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6201 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6202 variable. However, some OS use ints for inodes. I've added cast to long int
6203 to get rid of the compiler warning.
6205 19. I had forgotten to lock out "/../" in configuration file names when
6206 ALT_CONFIG_PREFIX was set.
6208 20. Routers used for verification do not need to specify transports. However,
6209 if such a router generated a host list, and callout was configured, Exim
6210 crashed, because it could not find a port number from the (non-existent)
6211 transport. It now assumes port 25 in this circumstance.
6213 21. Added the -t option to exigrep.
6215 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6216 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6217 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6218 them is set, LOOKUP_LSEARCH is forced.
6220 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6222 24. Added sender and host information to the "rejected by local_scan()" log
6223 line; previously there was no indication of these.
6225 25. Added .include_if_exists.
6227 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6228 Linux. It turns out that not all file systems support this. Apparently some
6229 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6230 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6231 not supported on the file descriptor, is now ignored when Exim is trying to
6232 sync a directory. This applies only to Linux.
6234 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6236 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6237 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6240 29. Added support for maildirsize files from supplied patch (modified a bit).
6242 30. The use of :fail: followed by an empty string could lead Exim to respond to
6243 sender verification failures with (e.g.):
6245 550 Verification failed for <xxx>
6246 550 Sender verify failed
6248 where the first response line was missing the '-' that indicates it is not
6249 the final line of the response.
6251 31. The loop for finding the name of the user that called Exim had a hardwired
6252 limit of 10; it now uses the value of finduser_retries, which is used for
6253 all other user lookups.
6255 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6258 33. Exim was neglecting to zero errno before one call of strtol() when
6259 expanding a string and expecting an integer value. On some systems this
6260 resulted in spurious "integer overflow" errors. Also, it was casting the
6261 result into an int without checking.
6263 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6264 did not work. The code looks as if it has *never* worked, though it appears
6265 to have been documented since at least release 1.62. I have made it work.
6267 35. The "timeout_DNS" error in retry rules, also documented since at least
6268 1.62, also never worked. As it isn't clear exactly what this means, and
6269 clearly it isn't a major issue, I have abolished the feature by treating it
6270 as "timeout", and writing a warning to the main and panic logs.
6272 36. The display of retry rules for -brt wasn't always showing the error code
6275 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6276 timeout_connect_A, timeout_connect_MX.
6278 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6279 to the empty sender.
6281 39. The daemon was not analysing the content of -oX till after it had closed
6282 stderr and disconnected from the controlling terminal. This meant that any
6283 syntax errors were only noted on the panic log, and the return code from
6284 the command was 0. By re-arranging the code a little, I've made the
6285 decoding happen first, so such errors now appear on stderr, and the return
6286 code is 1. However, the actual setting up of the sockets still happens in
6287 the disconnected process, so errors there are still only recorded on the
6290 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6291 connections (as happens on some IP stacks) was logged at start up time as
6292 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6293 it from "IPv6 and IPv4", which means that two separate sockets are being
6296 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6297 says whether AF_INET or AF_INET6 was passed as an argument.
6299 42. Exiwhat output was messed up when time zones were included in log
6302 43. Exiwhat now gives more information about the daemon's listening ports,
6303 and whether -tls-on-connect was used.
6305 44. The "port" option of the smtp transport is now expanded.
6307 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6308 silently ignored. Now an error message is written to the main and panic
6311 46. There's a new ACL modifier called "logwrite" which writes to a log file
6312 as soon as it is encountered.
6314 47. Added $local_user_uid and $local_user_gid at routing time.
6316 48. Exim crashed when trying to verify a sender address that was being
6319 49. Exim was recognizing only a space character after ".include". It now also
6320 recognizes a tab character.
6322 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6323 extracting the relevant information from the specification. The man page no
6324 longer contains scrambled data for the -d option, and I've added a section
6325 at the front about calling Exim under different names.
6327 51. Added "extra_headers" argument to the "mail" command in filter files.
6329 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6332 53. Installed eximstats 1.29.
6334 54. Added transport_filter_timeout as a generic transport option.
6336 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6337 Cc: header lines. This was required by RFC 822, but it not required by RFC
6340 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6341 incoming messages that did not have them. Now it does so only if the
6342 message originates locally, that is, if there is no associated remote host
6343 address. When Resent- header lines are present, this applies to the Resent-
6344 lines rather than the non-Resent- lines.
6346 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6347 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6349 58. Messages for configuration errors now include the name of the main
6350 configuration file - useful now that there may be more than one file in a
6351 list (.included file names were always shown).
6353 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6354 for those rare installations that do not start the daemon as root or run it
6355 setuid root. I've cut out the call to initgroups() if the daemon is not
6358 60. The Exim user and group can now be bound into the binary as text strings
6359 that are looked up at the start of Exim's processing.
6361 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6363 62. Added $mailstore_basename variable.
6365 63. Installed patch to sieve.c from Michael Haardt.
6367 64. When Exim failed to open the panic log after failing to open the main log,
6368 the original message it was trying to log was written to stderr and debug
6369 output, but if they were not available (the usual case in production), it
6370 was lost. Now it is written to syslog before the two lines that record the
6371 failures to open the logs.
6373 65. Users' Exim filters run in subprocesses under the user's uid. It is
6374 possible for a "deliver" command or an alias in a "personal" command to
6375 provoke an address rewrite. If logging of address rewriting is configured,
6376 this fails because the process is not running as root or exim. There may be
6377 a better way of dealing with this, but for the moment (because 4.30 needs
6378 to be released), I have disabled address rewrite logging when running a
6379 filter in a non-root, non-exim process.
6385 1. The buildconfig auxiliary program wasn't quoting the value set for
6386 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6387 not defined. This bug was masked in 4.22 by the effect that was fixed in
6390 2. Some messages that were rejected after a message id was allocated were
6391 shown as "incomplete" by exigrep. It no longer does this for messages that
6392 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6394 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6395 have allow_domain_literals set, the ID did not get logged in the <= line.
6396 Domain literals are now always recognized in Message-ID: header lines.
6398 4. The first argument for a ${extract expansion item is the key name or field
6399 number. Leading and trailing spaces in this item were not being ignored,
6400 causing some misleading effects.
6402 5. When deliver_drop_privilege was set, single queue runner processes started
6403 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6404 same command in the process it spins off) were not dropping privilege.
6406 6. When the daemon running as "exim" started a queue runner, it always
6407 re-executed Exim in the spun-off process. This is a waste of effort when
6408 deliver_drop_privilege is set. The new process now just calls the
6409 queue-runner function directly.
6415 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6418 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6419 ignored. Though the use of -oP was forcing the writing of a pid file, it
6420 was always written to the default place.
6422 3. If the message "no IP address found for host xxxx" is generated during
6423 incoming verification, it is now followed by identification of the incoming
6424 connection (so you can more easily find what provoked it).
6426 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6428 5. Added some features to "harden" Exim a bit more against certain attacks:
6430 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6431 be put in Local/Makefile. This is like the never_users runtime option,
6432 but it cannot be overridden. The default setting is "root".
6434 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6435 prefix string with which any file named in a -C command line option
6438 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6439 is retained for -C and -D only if the caller of Exim is root. Without
6440 it, the exim user may also use -C and -D and retain privilege.
6442 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6443 command line option is disabled.
6445 6. Macro names set by the -D option must start with an upper case letter, just
6446 like macro names defined in the configuration file.
6448 7. Added "dereference=" facility to LDAP.
6450 8. Two instances of the typo "uknown" in the source files are fixed.
6452 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6453 the Configure-Makefile script screwed up while processing it.
6455 10. Incorporated PCRE 4.4.
6457 11. The SMTP synchronization check was not operating right at the start of an
6458 SMTP session. For example, it could not catch a HELO sent before the client
6459 waited for the greeting. There is now a check for outstanding input at the
6460 point when the greeting is written. Because of the duplex, asynchronous
6461 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6462 way, but not yet received, when the check is performed.
6464 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6465 on TCP/IP sockets, because this apparently causes some broken clients to
6468 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6469 unchanged) from the Cygwin maintainer.
6471 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6472 was supported instead of testing for mbx. Effectively a typo.
6474 15. The spa authenticator server code was not checking that the input it
6475 received was valid base64.
6477 16. The debug output line for the "set" modifier in ACLs was not showing the
6478 name of the variable that was being set.
6480 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6482 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6483 lookup on its own. Something else was needed to trigger the lookup. For
6484 example, a match in host_lookup or the need for a host name in a host list.
6485 Now, if $sender_host_name is referenced and the host name has not yet been
6486 looked up, a lookup is performed. If the lookup fails, the variable remains
6487 empty, and $host_lookup_failed is set to "1".
6489 19. Added "eqi" as a case-independent comparison operator.
6491 20. The saslauthd authentication condition could segfault if neither service
6492 nor realm was specified.
6494 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6495 error message that was logged was misleading, and incoming SMTP
6496 connections were dropped. The message is now more accurate, and temporary
6497 errors are given to SMTP connections.
6499 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6500 (except RSET) with a 421 error, until QUIT is received. However, it was
6501 failing to send a response to QUIT.
6503 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6504 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6505 ACL is now run after the helo_try_verify_hosts code.
6507 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6508 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6509 case-independent, so other case variants are also recognized). Apparently
6510 some systems use these upper case variants.
6512 25. If more than two messages were waiting for the same host, and a transport
6513 filter was specified for the transport, Exim sent two messages over the
6514 same TCP/IP connection, and then failed with "socket operation on non-
6515 socket" when it tried to send the third.
6517 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6519 27. The extern definition of crypt16() in expand.c was not being excluded when
6520 the OS had its own crypt16() function.
6522 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6523 as a preferred synonym for return_size_limit, both as an option and as an
6526 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6528 30. Changed the default configuration ACL to relax the local part checking rule
6529 for addresses that are not in any local domains. For these addresses,
6530 slashes and pipe symbols are allowed within local parts, but the sequence
6531 /../ is explicitly forbidden.
6533 31. SPA server authentication was not clearing the challenge buffer before
6536 32. log_message in a "warn" ACL statement was writing to the reject log as
6537 well as to the main log, which contradicts the documentation and doesn't
6538 seem right (because no rejection is happening). So I have stopped it.
6540 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6541 However, I am unable to do any testing of this.
6543 34. Fixed an infelicity in the appendfile transport. When checking directories
6544 for a mailbox, to see if any needed to be created, it was accidentally
6545 using path names with one or more superfluous leading slashes; tracing
6546 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6548 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6549 contents are added to the log line that is written for every discarded
6550 recipient. (Previously a log_message setting was ignored.)
6552 36. The ${quote: operator now quotes the string if it is empty.
6554 37. The install script runs exim in order to find its version number. If for
6555 some reason other than non-existence or emptiness, which it checks, it
6556 could not run './exim', it was installing it with an empty version number,
6557 i.e. as "exim-". This error state is now caught, and the installation is
6560 38. An argument was missing from the function that creates an error message
6561 when Exim fails to connect to the socket for saslauthd authentication.
6562 This could cause Exim to crash, or give a corrupted message.
6564 39. Added isip, isip4, and isip6 to ${if conditions.
6566 40. The ACL variables $acl_xx are now saved with the message, and can be
6567 accessed later in routers, transports, and filters.
6569 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6570 strings in the file are not string-expanded.
6572 42. If a MAIL command specified a SIZE value that was too large to fit into an
6573 int variable, the check against message_size_limit failed. Such values are
6574 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6575 day this will have to be increased, but I don't think I want to be around
6576 when emails are that large.
6583 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6584 iconv() is not standard in FreeBSD.
6586 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6587 IPv6 enabled. The observed symptom was a segmentation fault on return from
6588 the function os_common_find_running_interfaces() in src/os.c.
6590 3. In the check_special_case() function in daemon.c I had used "errno" as an
6591 argument name, which causes warnings on some systems. This was basically a
6592 typo, since it was named "eno" in the comments!
6594 4. The code that waits for the clock to tick (at a resolution of some fraction
6595 of a second) so as to ensure message-id uniqueness was always waiting for
6596 at least one whole tick, when it could have waited for less. [This is
6597 almost certainly not relevant at current processor speeds, where it is
6598 unlikely to ever wait at all. But we try to future-proof.]
6600 5. The function that sleeps for a time interval that includes fractions of a
6601 second contained a race. It did not block SIGALRM between setting the
6602 timer, and suspending (a couple of lines later). If the interval was short
6603 and the sigsuspend() was delayed until after it had expired, the suspension
6604 never ended. On busy systems this could lead to processes getting stuck for
6607 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6608 process, before it forks any delivery processes. The open lookup caching
6609 mechanism meant that the open file or database connection was passed into
6610 the delivery process. The problem was that delivery processes always tidy
6611 up cached lookup data. This could cause a problem for the next delivery
6612 process started by the queue runner, because the external queue runner
6613 process does not know about the closure. So the next delivery process
6614 still has data in the lookup cache. In the case of a file lookup, there was
6615 no problem because closing a file descriptor in a subprocess doesn't affect
6616 the parent. However, if the lookup was caching a connection to a database,
6617 the connection was closed, and the second delivery process was likely to
6618 see errors such as "PGSQL: query failed: server closed the connection
6619 unexpectedly". The problem has been fixed by closing all cached lookups
6620 in a queue runner before running a delivery process.
6622 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6623 seem to have the "const" qualifier which it has on other OS. I've
6626 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6627 *of the same type* (client or server) with the same public name that an
6628 error should be diagnosed.
6630 9. When Exim looked up a host name for an IP address, but failed to find the
6631 original IP address when looking up the host name (a safety check), it
6632 output the message "<ip address> does not match any IP for NULL", which was
6633 confusing, to say the least. The bug was that the host name should have
6634 appeared instead of "NULL".
6636 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6637 user that is built into the binary, and the -C or -D options is used, root
6638 privilege is dropped before the configuration file is read. In addition,
6639 logging is switched to stderr instead of the normal log files. If the
6640 configuration then re-defines the Exim user, the unprivileged environment
6641 is probably not what is expected, so Exim logs a panic warning message (but
6644 However, if deliver_drop_privilege is set, the unprivileged state may well
6645 be exactly what is intended, so the warning has been cut out in that case,
6646 and Exim is allowed to try to write to its normal log files.
6652 1. smtp_return_error_details was not giving details for temporary sender
6653 or receiver verification errors.
6655 2. Diagnose a configuration error if two authenticators have the same public
6658 3. Exim used not to create the message log file for a message until the first
6659 delivery attempt. This could be confusing when incoming messages were held
6660 for policy or load reasons. The message log file is now created at the time
6661 the message is received, and an initial "Received" line is written to it.
6663 4. The automatically generated man page for command line options had a minor
6664 bug that caused no ill effects; however, a more serious problem was that
6665 the procedure for building the man page automatically didn't always
6666 operate. Consequently, release 4.20 contains an out-of-date version. This
6667 shouldn't happen again.
6669 5. When building Exim with embedded Perl support, the script that builds the
6670 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6671 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6673 6. The freeze_tell option was not being used for messages that were frozen on
6674 arrival, either by an ACL or by local_scan().
6676 7. Added the smtp_incomplete_transaction log selector.
6678 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6679 was accepting AUTH without a new EHLO.
6681 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6682 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6685 10. Logging of TCP/IP connections (when configured) now happens in the main
6686 daemon process instead of the child process, so that the TCP/IP connection
6687 count is more accurate (but it can never be perfect).
6689 11. The use of "drop" in a nested ACL was not being handled correctly in the
6690 outer ACL. Now, if condition failure induced by the nested "drop" causes
6691 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6692 or "require"), the connection is dropped.
6694 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6695 that yield "discard" can now be used with an "accept" or a "discard" verb,
6696 but an error is generated for any others (because I can't see a useful way
6697 to define what should happen).
6699 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6700 are now processed in the same way as lines in the Exim configuration file.
6701 In particular, continuation lines are supported.
6703 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6705 15. Added -ti meaning -t -i.
6707 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6708 domains, and warn by logging if others are found.
6710 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6711 returned from GIFCONF when Exim is trying to find the list of interfaces on
6712 a host. The code in os.c has been modified to copy each ifreq to an aligned
6713 structure in all cases.
6715 Also, in some cases, the returned ifreq's were being copied to a 'struct
6716 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6717 means the last couple of bytes of an IPv6 address could be chopped if the
6718 ifreq contained only a normal sockaddr (14 bytes storage).
6720 18. Named domain lists were not supported in the hosts_treat_as_local option.
6721 An entry such as +xxxx was not recognized, and was treated as a literal
6724 19. Ensure that header lines added by a DATA ACL are included in the reject log
6725 if the ACL subsequently rejects the message.
6727 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6728 MD5 (which is deprecated).
6730 21. When testing a filter file using -bf, Exim was writing a message when it
6731 took the sender from a "From " line in the message, but it was not doing so
6732 when it took $return_path from a Return-Path: header line. It now does.
6734 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6735 with a valid header line field name (a series of printing characters
6736 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6738 23. Changed "disc" in the source to "disk" to conform to the documentation and
6739 the book and for uniformity.
6741 24. Ignore Sendmail's -Ooption=value command line item.
6743 25. When execve() failed while trying to run a command in a pipe transport,
6744 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6745 could be confused with a return value of 69 from the command itself. This
6746 has been changed to 127, the value the shell returns if it is asked to run
6747 a non-existent command. The wording for the related log line suggests a
6748 non-existent command as the problem.
6750 26. If received_header_text expands to an empty string, do not add a Received:
6751 header line to the message. (Well, it adds a token one on the spool, but
6752 marks it "old" so that it doesn't get used or transmitted.)
6754 27. Installed eximstats 1.28 (addition of -nt option).
6756 28. There was no check for failure on the call to getsockname() in the daemon
6757 code. This can fail if there is a shortage of resources on the system, with
6758 ENOMEM, for example. A temporary error is now given on failure.
6760 29. Contrary to the C standard, it seems that in some environments, the
6761 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6762 program. Exim now does this explicitly; it affects the formatting of
6763 timestamps using strftime().
6765 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6766 complaints. I've now initialized all the variables, to avoid this.
6768 32. Header lines added by a system filter were not being "seen" during
6769 transport-time rewrites.
6771 33. The info_callback() function passed to OpenSSL is set up with type void
6772 (*)(SSL *, int, int), as described somewhere. However, when calling the
6773 function (actually a macro) that sets it up, the type void(*)() is
6774 expected. I've put in a cast to prevent warnings from picky compilers.
6776 34. If a DNS black list lookup found a CNAME record, but there were no A
6777 records associated with the domain it pointed at, Exim crashed.
6779 35. If a DNS black list lookup returned more than one A record, Exim ignored
6780 all but the first. It now scans all returned addresses if a particular IP
6781 value is being sought. In this situation, the contents of the
6782 $dnslist_value variable are a list of all the addresses, separated by a
6785 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6786 to accept a host name and all its aliases if the forward lookup for any of
6787 them yielded the IP address of the incoming connection. Now it accepts only
6788 those names whose forward lookup yields the correct IP address. Any other
6789 names are discarded. This closes a loophole whereby a rogue DNS
6790 administrator could create reverse DNS records to break through a
6791 wildcarded host restriction in an ACL.
6793 37. If a user filter or a system filter that ran in a subprocess used any of
6794 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6795 the wrong values were passed to the pipe command ($thisaddress had the
6796 value of $0, $0 had the value of $1, etc). This bug was introduced by
6797 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6799 38. Improved the line breaking for long SMTP error messages from ACLs.
6800 Previously, if there was no break point between 40 and 75 characters, Exim
6801 left the rest of the message alone. Two changes have been made: (a) I've
6802 reduced the minimum length to 35 characters; (b) if it can't find a break
6803 point between 35 and 75 characters, it looks ahead and uses the first one
6804 that it finds. This may give the occasional overlong line, but at least the
6805 remaining text gets split now.
6807 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6808 file descriptors might be low, and that setting 1000 would always raise it.
6809 It turns out that in some environments, the limit is already over 1000 and
6810 that lowering it causes trouble. So now Exim takes care not to decrease it.
6812 40. When delivering a message, the value of $return_path is set to $sender_
6813 address at the start of routing (routers may change the value). By an
6814 oversight, this default was not being set up when an address was tested by
6815 -bt or -bv, which affected the outcome if any router or filter referred to
6818 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6819 log only when "message" or "log_message" are set. However, if one of the
6820 conditions was an address verification, or a call to a nested ACL, the
6821 messages generated by the underlying test were being passed through. This
6822 no longer happens. The underlying message is available in $acl_verify_
6823 message for both "message" and "log_message" expansions, so it can be
6824 passed through if needed.
6826 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6827 new expansion $bh_ to give the encoded byte string without charset
6828 translation. Translation happens only if iconv() is available; HAVE_ICONV
6829 indicates this at build time. HEADERS_CHARSET gives the charset to
6830 translate to; headers_charset can change it in the configuration, and
6831 "headers charset" can change it in an individual filter file.
6833 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6834 that creates RFC 2047 encoded "words" labels them as that charset instead
6835 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6836 expansion operator; (ii) when Exim creates a From: line for a local
6837 message; (iii) when a header line is rewritten to include a "phrase" part.
6839 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6840 buggy, causing it to skip the first lines of messages whose message ID
6841 ended in 'D'. This would not have bitten before Exim release 4.14, because
6842 message IDs were unlikely to end in 'D' before then. The effect was to have
6843 incorrect size information for certain domains.
6845 45. #include "config.h" was missing at the start of the crypt16.c module. This
6846 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6849 46. If there was a timeout during a "random" callout check, Exim treated it as
6850 a failure of the random address, and carried on sending RSET and the real
6851 address. If the delay was just some slowness somewhere, the response to the
6852 original RCPT would be taken as a response to RSET and so on, causing
6853 mayhem of various kinds.
6855 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6856 when I implemented it. It didn't allow for the fact that some option values
6857 may legitimately be negative (e.g. size_addition), and it didn't even do
6858 the right test for positive values.
6860 48. Domain names in DNS records are case-independent. Exim always looks them up
6861 in lower case. Some resolvers return domain names in exactly the case they
6862 appear in the zone file, that is, they may contain uppercase letters. Not
6863 all resolvers do this - some return always lower case. Exim was treating a
6864 change of case by a resolver as a change of domain, similar to a widening
6865 of a domain abbreviation. This triggered its re-routing code and so it was
6866 trying to route what was effectively the same domain again. This normally
6867 caused routing to fail (because the router wouldn't handle the domain
6868 twice). Now Exim checks for this case specially, and just changes the
6869 casing of the domain that it ultimately uses when it transmits the message
6872 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6875 50. If a filter generated a file delivery with a non-absolute name (possible if
6876 no home directory exists for the router), the forbid_file option was not
6879 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6880 the existing equality matching.
6882 52. Exim was using ints instead of ino_t variables in some places where it was
6883 dealing with inode numbers.
6885 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6886 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6887 called TMPDIR, and if it finds it is different, it changes its value.
6889 54. The smtp_printf() function is now made available to local_scan() so
6890 additional output lines can be written before returning. There is also an
6891 smtp_fflush() function to enable the detection of a dropped connection.
6892 The variables smtp_input and smtp_batched_input are exported to
6895 55. Changed the default runtime configuration: the message "Unknown user"
6896 has been removed from the ACL, and instead placed on the localuser router,
6897 using the cannot_route_message feature. This means that any verification
6898 failures that generate their own messages won't get overridden. Similarly,
6899 the "Unrouteable address" message that was in the ACL for unverifiable
6900 relay addresses has also been removed.
6902 56. Added hosts_avoid_esmtp to the smtp transport.
6904 57. The exicyclog script was not checking for the esoteric option
6905 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6906 will work only if exicyclog is run under the appropriate euid.
6908 58. Following a discussion on the list, the rules by which Exim recognises line
6909 endings on incoming messages have been changed. The -dropcr and drop_cr
6910 options are now no-ops, retained only for backwards compatibility. The
6911 following line terminators are recognized: LF CRLF CR. However, special
6912 processing applies to CR:
6914 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6915 nor a local message in the state where . is a terminator.
6917 (ii) If a bare CR is encountered in a header line, an extra space is added
6918 after the line terminator so as not to end the header. The reasoning
6919 behind this is that bare CRs in header lines are most likely either
6920 to be mistakes, or people trying to play silly games.
6922 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6923 was being incorrectly given as 18 bytes larger than it should have been.
6924 This is a VOB (very old bug).
6926 60. This may never have affected anything current, but just in case it has:
6927 When the local host is found other than at the start of a list of hosts,
6928 the local host, those with the same MX, and any that follow, are discarded.
6929 When the list in question was part of a longer list of hosts, the following
6930 hosts (not currently being processed) were also being discarded. This no
6931 longer happens. I'm not sure if this situation could ever has previously
6934 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6937 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6938 header. This is something that is recommended in a new Internet Draft, and
6939 is something that is documented as being done by Sendmail. There are two
6940 possible values. For messages generated by the autoreply transport, Exim
6943 Auto-Submitted: auto-replied
6945 whereas for all other generated messages (e.g. bounces) it adds
6947 Auto-Submitted: auto-generated
6949 63. The "personal" condition in filters now includes a test for the
6950 Auto-Submitted: header. If it contains the string "auto-" the message it
6951 not considered personal.
6953 64. Added rcpt_include_affixes as a generic transport option.
6955 65. Added queue_only_override (default true).
6957 66. Added the syslog_duplication option.
6959 67. If what should have been the first header line of a message consisted of
6960 a space followed by a colon, Exim was mis-interpreting it as a header line.
6961 It isn't of course - it is syntactically invalid and should therefore be
6962 treated as the start of the message body. The misbehaviour could have
6963 caused a number of strange effects, including loss of data in subsequent
6964 header lines, and spool format errors.
6966 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6967 client host had authenticated. This control can now be exercised by an ACL
6968 for more flexibility.
6970 69. By default, callouts do not happen when testing with -bh. There is now a
6971 variant, -bhc, which does actually run the callout code, including
6972 consulting and updating the callout cache.
6974 70. Added support for saslauthd authentication, courtesy of Alexander
6977 71. If statvfs() failed on the spool or log directories while checking their
6978 size for availability, Exim confusingly gave the error "space shortage".
6979 Furthermore, in debugging mode it crashed with a floating point exception.
6980 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6981 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6982 serious problem, Exim now writes to the main and panic logs when this
6983 happens, with details of the failure. It then refuses to accept the
6984 incoming message, giving the message "spool directory problem" or "log
6985 directory problem" with a 421 code for SMTP messages.
6987 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6988 0, 1, and 2 exist, because some OS complain for execs without them (see
6989 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6990 descriptors. However, the code omitted to check that the open succeeded,
6991 causing mysterious errors if for some reason the permissions on /dev/null
6992 got screwed. Now Exim writes a message to the main and panic logs, and
6993 bombs out if it can't open /dev/null.
6995 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6996 interact so that it is all more flexible. It is supposed to remain
6997 backwards compatible. Also added extra_local_interfaces.
6999 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7000 to bomb out with an assertion failure - to the client this appears as a
7001 connection drop. This problem occurs in the part of the code that was taken
7002 from the Samba project. Fortunately, the assertion is in a very simple
7003 function, so I have fixed this by reproducing the function inline in the
7004 one place where it is called, and arranging for authentication to fail
7005 instead of killing the process with assert().
7007 75. The SPA client code was not working when the server requested OEM rather
7008 than Unicode encoding.
7010 76. Added code to make require_files with a specific uid setting more usable in
7011 the case where statting the file as root fails - usually a non-root-mounted
7012 NFS file system. When this happens and the failure is EACCES, Exim now
7013 forks a subprocess and does the per-uid checking as the relevant uid.
7015 77. Added process_log_path.
7017 78. If log_file_path was not explicitly set, a setting of check_log_space or
7018 check_log_inodes was ignored.
7020 79. If a space check for the spool or log partitions fails, the incident is now
7021 logged. Of course, in the latter case the data may get lost...
7023 80. Added the %p formatting code to string_format() so that it can be used to
7024 print addresses in debug_print(). Adjusted all the address printing in the
7025 debugging in store.c to use %p rather than %d.
7027 81. There was a concern that a line of code in smtp_in.c could overflow a
7028 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7029 initially expressed, the concern was not well-founded, because trailing
7030 spaces are removed early. However, if the trailing spaces were followed by
7031 a NULL, they did not get removed, so the overflow was possible. Two fixes
7034 (a) I re-wrote the offending code in a cleaner fashion.
7035 (b) If an incoming SMTP command contains a NULL character, it is rejected
7038 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7039 runs initgroups(), so that if the Exim user is in any additional groups,
7040 they will be used during message reception.
7046 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob