1 Change log file for Exim from version 4.21
2 ------------------------------------------
3 This document describes *changes* to previous versions, that might
4 affect Exim's operation, with an unchanged configuration file. For new
5 options, and new features, see the NewStuff file next to this ChangeLog.
8 Since Exim version 4.90
9 -----------------------
11 JH/01 Replace the store_release() internal interface with store_newblock(),
12 which internalises the check required to safely use the old one, plus
13 the allocate and data copy operations duplicated in both (!) of the
16 JH/02 Disallow '/' characters in queue names specified for the "queue=" ACL
17 modifier. This matches the restriction on the commandline.
19 JH/03 Fix pgsql lookup for multiple result-tuples with a single column.
20 Previously only the last row was returned.
22 JH/04 Bug 2217: Tighten up the parsing of DKIM signature headers. Previously
23 we assumed that tags in the header were well-formed, and parsed the
24 element content after inspecting only the first char of the tag.
25 Assumptions at that stage could crash the receive process on malformed
28 JH/05 Bug 2215: Fix crash associated with dnsdb lookup done from DKIM ACL.
29 While running the DKIM ACL we operate on the Permanent memory pool so that
30 variables created with "set" persist to the DATA ACL. Also (at any time)
31 DNS lookups that fail create cache records using the Permanent pool. But
32 expansions release any allocations made on the current pool - so a dnsdb
33 lookup expansion done in the DKIM ACL releases the memory used for the
34 DNS negative-cache, and bad things result. Solution is to switch to the
35 Main pool for expansions.
36 While we're in that code, add checks on the DNS cache during store_reset,
37 active in the testsuite.
38 Problem spotted, and debugging aided, by Wolfgang Breyha.
40 JH/06 Fix issue with continued-connections when the DNS shifts unreliably.
41 When none of the hosts presented to a transport match an already-open
42 connection, close it and proceed with the list. Previously we would
43 queue the message. Spotted by Lena with Yahoo, probably involving
46 JH/07 Bug 2214: Fix SMTP responses resulting from non-accept result of MIME ACL.
47 Previously a spurious "250 OK id=" response was appended to the proper
50 JH/08 The "support for" informational output now, which built with Content
51 Scanning support, has a line for the malware scanner interfaces compiled
52 in. Interface can be individually included or not at build time.
54 JH/09 The "aveserver", "kavdaemon" and "mksd" interfaces are now not included
55 by the template makefile "src/EDITME". The "STREAM" support for an older
56 ClamAV interface method is removed.
58 JH/10 Bug 2223: Fix mysql lookup returns for the no-data case (when the number of
59 rows affected is given instead).
61 JH/11 The runtime Berkeley DB library version is now additionally output by
62 "exim -d -bV". Previously only the compile-time version was shown.
64 JH/12 Bug 2230: Fix cutthrough routing for nonfirst messages in an initiating
65 SMTP connection. Previously, when one had more receipients than the
66 first, an abortive onward connection was made. Move to full support for
67 multiple onward connections in sequence, handling cutthrough connection
68 for all multi-message initiating connections.
70 JH/13 Bug 2229: Fix cutthrough routing for nonstandard port numbers defined by
71 routers. Previously, a multi-recipient message would fail to match the
72 onward-connection opened for the first recipient, and cause its closure.
74 JH/14 Bug 2174: A timeout on connect for a callout was also erroneously seen as
75 a timeout on read on a GnuTLS initiating connection, resulting in the
76 initiating connection being dropped. This mattered most when the callout
77 was marked defer_ok. Fix to keep the two timeout-detection methods
80 JH/15 Relax results from ACL control request to enable cutthrough, in
81 unsupported situations, from error to silently (except under debug)
82 ignoring. This covers use with PRDR, frozen messages, queue-only and
85 HS/01 Fix Buffer overflow in base64d() (CVE-2018-6789)
87 JH/16 Fix bug in DKIM verify: a buffer overflow could corrupt the malloc
88 metadata, resulting in a crash in free().
90 PP/01 Fix broken Heimdal GSSAPI authenticator integration.
91 Broken in f2ed27cf5, missing an equals sign for specified-initialisers.
92 Broken also in d185889f4, with init system revamp.
94 JH/17 Bug 2113: Fix conversation closedown with the Avast malware scanner.
95 Previously we abruptly closed the connection after reading a malware-
96 found indication; now we go on to read the "scan ok" response line,
99 JH/18 Bug 2239: Enforce non-usability of control=utf8_downconvert in the mail
100 ACL. Previously, a crash would result.
102 JH/19 Speed up macro lookups during configuration file read, by skipping non-
103 macro text after a replacement (previously it was only once per line) and
104 by skipping builtin macros when searching for an uppercase lead character.
110 JH/01 Rework error string handling in TLS interface so that the caller in
111 more cases is responsible for logging. This permits library-sourced
112 string to be attached to addresses during delivery, and collapses
113 pairs of long lines into single ones.
115 PP/01 Allow PKG_CONFIG_PATH to be set in Local/Makefile and use it correctly
116 during configuration. Wildcards are allowed and expanded.
118 JH/02 Rework error string handling in DKIM to pass more info back to callers.
119 This permits better logging.
121 JH/03 Rework the transport continued-connection mechanism: when TLS is active,
122 do not close it down and have the child transport start it up again on
123 the passed-on TCP connection. Instead, proxy the child (and any
124 subsequent ones) for TLS via a unix-domain socket channel. Logging is
125 affected: the continued delivery log lines do not have any DNSSEC, TLS
126 Certificate or OCSP information. TLS cipher information is still logged.
128 JH/04 Shorten the log line for daemon startup by collapsing adjacent sets of
129 identical IP addresses on different listening ports. Will also affect
132 PP/02 Bug 2070: uClibc defines __GLIBC__ without providing glibc headers;
133 add noisy ifdef guards to special-case this sillyness.
134 Patch from Bernd Kuhls.
136 JH/05 Tighten up the checking in isip4 (et al): dotted-quad components larger
137 than 255 are no longer allowed.
139 JH/06 Default openssl_options to include +no_ticket, to reduce load on peers.
140 Disable the session-cache too, which might reduce our load. Since we
141 currrectly use a new context for every connection, both as server and
142 client, there is no benefit for these.
143 GnuTLS appears to not support tickets server-side by default (we don't
144 call gnutls_session_ticket_enable_server()) but client side is enabled
145 by default on recent versions (3.1.3 +) unless the PFS priority string
148 PP/03 Add $SOURCE_DATE_EPOCH support for reproducible builds, per spec at
149 <https://reproducible-builds.org/specs/source-date-epoch/>.
151 JH/07 Fix smtp transport use of limited max_rcpt under mua_wrapper. Previously
152 the check for any unsuccessful recipients did not notice the limit, and
153 erroneously found still-pending ones.
155 JH/08 Pipeline CHUNKING command and data together, on kernels that support
156 MSG_MORE. Only in-clear (not on TLS connections).
158 JH/09 Avoid using a temporary file during transport using dkim. Unless a
159 transport-filter is involved we can buffer the headers in memory for
160 creating the signature, and read the spool data file once for the
161 signature and again for transmission.
163 JH/10 Enable use of sendfile in Linux builds as default. It was disabled in
164 4.77 as the kernel support then wasn't solid, having issues in 64bit
165 mode. Now, it's been long enough. Add support for FreeBSD also.
167 JH/11 Bug 2104: Fix continued use of a transport connection with TLS. In the
168 case where the routing stage had gathered several addresses to send to
169 a host before calling the transport for the first, we previously failed
170 to close down TLS in the old transport process before passing the TCP
171 connection to the new process. The new one sent a STARTTLS command
172 which naturally failed, giving a failed delivery and bloating the retry
173 database. Investigation and fix prototype from Wolfgang Breyha.
175 JH/12 Fix check on SMTP command input synchronisation. Previously there were
176 false-negatives in the check that the sender had not preempted a response
177 or prompt from Exim (running as a server), due to that code's lack of
178 awareness of the SMTP input buffering.
180 PP/04 Add commandline_checks_require_admin option.
181 Exim drops privileges sanely, various checks such as -be aren't a
182 security problem, as long as you trust local users with access to their
183 own account. When invoked by services which pass untrusted data to
184 Exim, this might be an issue. Set this option in main configuration
185 AND make fixes to the calling application, such as using `--` to stop
188 JH/13 Do pipelining under TLS. Previously, although safe, no advantage was
189 taken. Now take care to pack both (client) MAIL,RCPT,DATA, and (server)
190 responses to those, into a single TLS record each way (this usually means
191 a single packet). As a side issue, smtp_enforce_sync now works on TLS
194 PP/05 OpenSSL/1.1: use DH_bits() for more accurate DH param sizes. This
195 affects you only if you're dancing at the edge of the param size limits.
196 If you are, and this message makes sense to you, then: raise the
197 configured limit or use OpenSSL 1.1. Nothing we can do for older
200 JH/14 For the "sock" variant of the malware scanner interface, accept an empty
201 cmdline element to get the documented default one. Previously it was
204 JH/15 Fix a crash in the smtp transport caused when two hosts in succession
205 are unsuable for non-message-specific reasons - eg. connection timeout,
206 banner-time rejection.
208 JH/16 Fix logging of delivery remote port, when specified by router, under
211 PP/06 Repair manualroute's ability to take options in any order, even if one
212 is the name of a transport.
215 HS/01 Cleanup, prevent repeated use of -p/-oMr (CVE-2017-1000369)
217 JH/17 Change the list-building routines interface to use the expanding-string
218 triplet model, for better allocation and copying behaviour.
220 JH/18 Prebuild the data-structure for "builtin" macros, for faster startup.
221 Previously it was constructed the first time a possibly-matching string
222 was met in the configuration file input during startup; now it is done
225 JH/19 Bug 2141: Use the full-complex API for Berkeley DB rather than the legacy-
226 compatible one, to avoid the (poorly documented) possibility of a config
227 file in the working directory redirecting the DB files, possibly correpting
228 some existing file. CVE-2017-10140 assigned for BDB.
230 JH/20 Bug 2147: Do not defer for a verify-with-callout-and-random which is not
231 cache-hot. Previously, although the result was properly cached, the
232 initial verify call returned a defer.
234 JH/21 Bug 2151: Avoid using SIZE on the MAIL for a callout verify, on any but
235 the main verify for receipient in uncached-mode.
237 JH/22 Retire historical build files to an "unsupported" subdir. These are
238 defined as "ones for which we have no current evidence of testing".
240 JH/23 DKIM: enforce the DNS pubkey record "h" permitted-hashes optional field,
241 if present. Previously it was ignored.
243 JH/24 Start using specified-initialisers in C structure init coding. This is
244 a C99 feature (it's 2017, so now considered safe).
246 JH/25 Use one-bit bitfields for flags in the "addr" data structure. Previously
247 if was a fixed-sized field and bitmask ops via macros; it is now more
250 PP/07 GitHub PR 56: Apply MariaDB build fix.
251 Patch provided by Jaroslav Å karvada.
253 PP/08 Bug 2161: Fix regression in sieve quoted-printable handling introduced
254 during Coverity cleanups [4.87 JH/47]
255 Diagnosis and fix provided by Michael Fischer v. Mollard.
257 JH/26 Fix DKIM bug: when the pseudoheader generated for signing was exactly
258 the right size to place the terminating semicolon on its own folded
259 line, the header hash was calculated to an incorrect value thanks to
260 the (relaxed) space the fold became.
262 HS/02 Fix Bug 2130: large writes from the transport subprocess where chunked
263 and confused the parent.
265 JH/27 Fix SOCKS bug: an unitialized pointer was deref'd by the transport process
266 which could crash as a result. This could lead to undeliverable messages.
268 JH/28 Logging: "next input sent too soon" now shows where input was truncated
271 JH/29 Fix queue_run_in_order to ignore the PID portion of the message ID. This
272 matters on fast-turnover and PID-randomising systems, which were getting
273 out-of-order delivery.
275 JH/30 Fix a logging bug on aarch64: an unsafe routine was previously used for
276 a possibly-overlapping copy. The symptom was that "Remote host closed
277 connection in response to HELO" was logged instead of the actual 4xx
280 JH/31 Fix CHUNKING code to properly flush the unwanted chunk after an error.
281 Previously only that bufferd was discarded, resulting in SYMTP command
284 JH/32 DKIM: when a message has multiple signatures matching an identity given
285 in dkim_verify_signers, run the dkim acl once for each. Previously only
286 one run was done. Bug 2189.
288 JH/33 Downgrade an unfound-list name (usually a typo in the config file) from
289 "panic the current process" to "deliberately defer". The panic log is
290 still written with the problem list name; the mail and reject logs now
291 get a temp-reject line for the message that was being handled, saying
292 something like "domains check lookup or other defer". The SMTP 451
293 message is still "Temporary local problem".
295 JH/34 Bug 2199: Fix a use-after-free while reading smtp input for header lines.
296 A crafted sequence of BDAT commands could result in in-use memory beeing
297 freed. CVE-2017-16943.
299 HS/03 Bug 2201: Fix checking for leading-dot on a line during headers reading
300 from SMTP input. Previously it was always done; now only done for DATA
301 and not BDAT commands. CVE-2017-16944.
303 JH/35 Bug 2201: Flush received data in BDAT mode after detecting an error fatal
304 to the message (such as an overlong header line). Previously this was
305 not done and we did not exit BDAT mode. Followon from the previous item
306 though a different problem.
312 JH/01 Bug 1922: Support IDNA2008. This has slightly different conversion rules
313 than -2003 did; needs libidn2 in addition to libidn.
315 JH/02 The path option on a pipe transport is now expanded before use.
317 PP/01 GitHub PR 50: Do not call ldap_start_tls_s on ldapi:// connections.
318 Patch provided by "Björn", documentation fix added too.
320 JH/03 Bug 2003: fix Proxy Protocol v2 handling: the address size field was
321 missing a wire-to-host endian conversion.
323 JH/04 Bug 2004: fix CHUNKING in non-PIPELINEING mode. Chunk data following
324 close after a BDAT command line could be taken as a following command,
325 giving a synch failure. Fix by only checking for synch immediately
326 before acknowledging the chunk.
328 PP/02 GitHub PR 52: many spelling fixes, which include fixing parsing of
329 no_require_dnssec option and creation of _HAVE_TRANSPORT_APPEND_MAILDIR
330 macro. Patches provided by Josh Soref.
332 JH/05 Have the EHLO response advertise VRFY, if there is a vrfy ACL defined.
333 Previously we did not; the RFC seems ambiguous and VRFY is not listed
334 by IANA as a service extension. However, John Klensin suggests that we
337 JH/06 Bug 2017: Fix DKIM verification in -bh test mode. The data feed into
338 the dkim code may be unix-mode line endings rather than smtp wire-format
339 CRLF, so prepend a CR to any bare LF.
341 JH/07 Rationalise the coding for callout smtp conversations and transport ones.
342 As a side-benfit, callouts can now use PIPELINING hence fewer round-trips.
344 JH/08 Bug 2016: Fix DKIM verification vs. CHUNKING. Any BDAT commands after
345 the first were themselves being wrongly included in the feed into dkim
346 processing; with most chunk sizes in use this resulted in an incorrect
347 body hash calculated value.
349 JH/09 Bug 2014: permit inclusion of a DKIM-Signature header in a received
350 DKIM signature block, for verification. Although advised against by
351 standards it is specifically not ruled illegal.
353 JH/10 Bug 2025: Fix reception of (quoted) local-parts with embedded spaces.
355 JH/11 Bug 2029: Fix crash in DKIM verification when a message signature block is
356 missing a body hash (the bh= tag).
358 JH/12 Bug 2018: Re-order Proxy Protocol startup versus TLS-on-connect startup.
359 It seems that HAProxy sends the Proxy Protocol information in clear and
360 only then does a TLS startup, so do the same.
362 JH/13 Bug 2027: Avoid attempting to use TCP Fast Open for non-transport client
363 TCP connections (such as for Spamd) unless the daemon successfully set
364 Fast Open mode on its listening sockets. This fixes breakage seen on
365 too-old kernels or those not configured for Fast Open, at the cost of
366 requiring both directions being enabled for TFO, and TFO never being used
367 by non-daemon-related Exim processes.
369 JH/14 Bug 2000: Reject messages recieved with CHUNKING but with malformed line
370 endings, at least on the first header line. Try to canonify any that get
371 past that check, despite the cost.
373 JH/15 Angle-bracket nesting (an error inserted by broken sendmails) levels are
374 now limited to an arbitrary five deep, while parsing addresses with the
375 strip_excess_angle_brackets option enabled.
377 PP/03 Bug 2018: For Proxy Protocol and TLS-on-connect, do not over-read and
378 instead leave the unprompted TLS handshake in socket buffer for the
379 TLS library to consume.
381 PP/04 Bug 2018: Also handle Proxy Protocol v2 safely.
383 PP/05 FreeBSD compat: handle that Ports no longer create /usr/bin/perl
385 JH/16 Drop variables when they go out of scope. Memory management drops a whole
386 region in one operation, for speed, and this leaves assigned pointers
387 dangling. Add checks run only under the testsuite which checks all
388 variables at a store-reset and panics on a dangling pointer; add code
389 explicitly nulling out all the variables discovered. Fixes one known
390 bug: a transport crash, where a dangling pointer for $sending_ip_address
391 originally assigned in a verify callout, is re-used.
393 PP/06 Drop '.' from @INC in various Perl scripts.
395 PP/07 Switch FreeBSD iconv to always use the base-system libc functions.
397 PP/08 Reduce a number of compilation warnings under clang; building with
398 CC=clang CFLAGS+=-Wno-dangling-else -Wno-logical-op-parentheses
399 should be warning-free.
401 JH/17 Fix inbound CHUNKING when DKIM disabled at runtime.
403 HS/01 Fix portability problems introduced by PP/08 for platforms where
404 realloc(NULL) is not equivalent to malloc() [SunOS et al].
406 HS/02 Bug 1974: Fix missing line terminator on the last received BDAT
407 chunk. This allows us to accept broken chunked messages. We need a more
408 general solution here.
410 PP/09 Wrote util/chunking_fixqueue_finalnewlines.pl to help recover
411 already-broken messages in the queue.
413 JH/18 Bug 2061: Fix ${extract } corrupting an enclosing ${reduce } $value.
415 JH/19 Fix reference counting bug in routing-generated-address tracking.
421 JH/01 Use SIZE on MAIL FROM in a cutthrough connection, if the destination
422 supports it and a size is available (ie. the sending peer gave us one).
424 JH/02 The obsolete acl condition "demime" is removed (finally, after ten
425 years of being deprecated). The replacements are the ACLs
426 acl_smtp_mime and acl_not_smtp_mime.
428 JH/03 Upgrade security requirements imposed for hosts_try_dane: previously
429 a downgraded non-dane trust-anchor for the TLS connection (CA-style)
430 or even an in-clear connection were permitted. Now, if the host lookup
431 was dnssec and dane was requested then the host is only used if the
432 TLSA lookup succeeds and is dnssec. Further hosts (eg. lower priority
433 MXs) will be tried (for hosts_try_dane though not for hosts_require_dane)
434 if one fails this test.
435 This means that a poorly-configured remote DNS will make it incommunicado;
436 but it protects against a DNS-interception attack on it.
438 JH/04 Bug 1810: make continued-use of an open smtp transport connection
439 non-noisy when a race steals the message being considered.
441 JH/05 If main configuration option tls_certificate is unset, generate a
442 self-signed certificate for inbound TLS connections.
444 JH/06 Bug 165: hide more cases of password exposure - this time in expansions
445 in rewrites and routers.
447 JH/07 Retire gnutls_require_mac et.al. These were nonfunctional since 4.80
448 and logged a warning sing 4.83; now they are a configuration file error.
450 JH/08 Bug 1836: Fix crash in VRFY handling when handed an unqualified name
451 (lacking @domain). Apply the same qualification processing as RCPT.
453 JH/09 Bug 1804: Avoid writing msglog files when in -bh or -bhc mode.
455 JH/10 Support ${sha256:} applied to a string (as well as the previous
458 JH/11 Cutthrough: avoid using the callout hints db on a verify callout when
459 a cutthrough deliver is pending, as we always want to make a connection.
460 This also avoids re-routing the message when later placing the cutthrough
461 connection after a verify cache hit.
462 Do not update it with the verify result either.
464 JH/12 Cutthrough: disable when verify option success_on_redirect is used, and
465 when routing results in more than one destination address.
467 JH/13 Cutthrough: expand transport dkim_domain option when testing for dkim
468 signing (which inhibits the cutthrough capability). Previously only
469 the presence of an option was tested; now an expansion evaluating as
470 empty is permissible (obviously it should depend only on data available
471 when the cutthrough connection is made).
473 JH/14 Fix logging of errors under PIPELINING. Previously the log line giving
474 the relevant preceding SMTP command did not note the pipelining mode.
476 JH/15 Fix counting of empty lines in $body_linecount and $message_linecount.
477 Previously they were not counted.
479 JH/16 DANE: treat a TLSA lookup response having all non-TLSA RRs, the same
480 as one having no matching records. Previously we deferred the message
481 that needed the lookup.
483 JH/17 Fakereject: previously logged as a normal message arrival "<="; now
484 distinguished as "(=".
486 JH/18 Bug 1867: make the fail_defer_domains option on a dnslookup router work
487 for missing MX records. Previously it only worked for missing A records.
489 JH/19 Bug 1850: support Radius libraries that return REJECT_RC.
491 JH/20 Bug 1872: Ensure that acl_smtp_notquit is run when the connection drops
492 after the data-go-ahead and data-ack. Patch from Jason Betts.
494 JH/21 Bug 1846: Send DMARC forensic reports for reject and quarantine results,
495 even for a "none" policy. Patch from Tony Meyer.
497 JH/22 Fix continued use of a connection for further deliveries. If a port was
498 specified by a router, it must also match for the delivery to be
501 JH/23 Bug 1874: fix continued use of a connection for further deliveries.
502 When one of the recipients of a message was unsuitable for the connection
503 (has no matching addresses), we lost track of needing to mark it
504 deferred. As a result mail would be lost.
506 JH/24 Bug 1832: Log EHLO response on getting conn-close response for HELO.
508 JH/25 Decoding ACL controls is now done using a binary search; the source code
509 takes up less space and should be simpler to maintain. Merge the ACL
510 condition decode tables also, with similar effect.
512 JH/26 Fix problem with one_time used on a redirect router which returned the
513 parent address unchanged. A retry would see the parent address marked as
514 delivered, so not attempt the (identical) child. As a result mail would
517 JH/27 Fix a possible security hole, wherein a process operating with the Exim
518 UID can gain a root shell. Credit to http://www.halfdog.net/ for
519 discovery and writeup. Ubuntu bug 1580454; no bug raised against Exim
522 JH/28 Enable {spool,log} filesystem space and inode checks as default.
523 Main config options check_{log,spool}_{inodes,space} are now
524 100 inodes, 10MB unless set otherwise in the configuration.
526 JH/29 Fix the connection_reject log selector to apply to the connect ACL.
527 Previously it only applied to the main-section connection policy
530 JH/30 Bug 1897: fix callouts connection fallback from TLS to cleartext.
532 PP/01 Changed default Diffie-Hellman parameters to be Exim-specific, created
533 by me. Added RFC7919 DH primes as an alternative.
535 PP/02 Unbreak build via pkg-config with new hash support when crypto headers
536 are not in the system include path.
538 JH/31 Fix longstanding bug with aborted TLS server connection handling. Under
539 GnuTLS, when a session startup failed (eg because the client disconnected)
540 Exim did stdio operations after fclose. This was exposed by a recent
541 change which nulled out the file handle after the fclose.
543 JH/32 Bug 1909: Fix OCSP proof verification for cases where the proof is
544 signed directly by the cert-signing cert, rather than an intermediate
545 OCSP-signing cert. This is the model used by LetsEncrypt.
547 JH/33 Bug 1914: Ensure socket is nonblocking before draining after SMTP QUIT.
549 HS/01 Fix leak in verify callout under GnuTLS, about 3MB per recipient on
550 an incoming connection.
552 HS/02 Bug 1802: Do not half-close the connection after sending a request
555 HS/03 Use "auto" as the default EC curve parameter. For OpenSSL < 1.0.2
556 fallback to "prime256v1".
558 JH/34 SECURITY: Use proper copy of DATA command in error message.
559 Could leak key material. Remotely exploitable. CVE-2016-9963.
565 JH/01 Bug 1664: Disable OCSP for GnuTLS library versions at/before 3.3.16
566 and 3.4.4 - once the server is enabled to respond to an OCSP request
567 it does even when not requested, resulting in a stapling non-aware
568 client dropping the TLS connection.
570 TF/01 Code cleanup: Overhaul the debug_selector and log_selector machinery to
571 support variable-length bit vectors. No functional change.
573 TF/02 Improve the consistency of logging incoming and outgoing interfaces.
574 The I= interface field on outgoing lines is now after the H= remote
575 host field, same as incoming lines. There is a separate
576 outgoing_interface log selector which allows you to disable the
579 JH/02 Bug 728: Close logfiles after a daemon-process "exceptional" log write.
580 If not running log_selector +smtp_connection the mainlog would be held
581 open indefinitely after a "too many connections" event, including to a
582 deleted file after a log rotate. Leave the per net connection logging
583 leaving it open for efficiency as that will be quickly detected by the
584 check on the next write.
586 HS/01 Bug 1671: Fix post transport crash.
587 Processing the wait-<transport> messages could crash the delivery
588 process if the message IDs didn't exist for some reason. When
589 using 'split_spool_directory=yes' the construction of the spool
590 file name failed already, exposing the same netto behaviour.
592 JH/03 Bug 425: Capture substrings in $regex1, $regex2 etc from regex &
593 mime_regex ACL conditions.
595 JH/04 Bug 1686: When compiled with EXPERIMENTAL_DSN_INFO: Add extra information
596 to DSN fail messages (bounces): remote IP, remote greeting, remote response
597 to HELO, local diagnostic string.
599 JH/05 Downgrade message for a TLS-certificate-based authentication fail from
600 log line to debug. Even when configured with a tls authenticator many
601 client connections are expected to not authenticate in this way, so
602 an authenticate fail is not an error.
604 HS/02 Add the Exim version string to the process info. This way exiwhat
605 gives some more detail about the running daemon.
607 JH/06 Bug 1395: time-limit caching of DNS lookups, to the TTL value. This may
608 matter for fast-change records such as DNSBLs.
610 JH/07 Bug 1678: Always record an interface option value, if set, as part of a
611 retry record, even if constant. There may be multiple transports with
612 different interface settings and the retry behaviour needs to be kept
615 JH/08 Bug 1586: exiqgrep now refuses to run if there are unexpected arguments.
617 JH/09 Bug 1700: ignore space & tab embedded in base64 during decode.
619 JH/10 Bug 840: fix log_defer_output option of pipe transport
621 JH/11 Bug 830: use same host for all RCPTS of a message, even under
622 hosts_randomize. This matters a lot when combined with mua_wrapper.
624 JH/12 Bug 1706: percent and underbar characters are no longer escaped by the
625 ${quote_pgsql:<string>} operator.
627 JH/13 Bug 1708: avoid misaligned access in cached lookup.
629 JH/14 Change header file name for freeradius-client. Relevant if compiling
630 with Radius support; from the Gentoo tree and checked under Fedora.
632 JH/15 Bug 1712: Introduce $prdr_requested flag variable
634 JH/16 Bug 1714: Permit an empty string as expansion result for transport
635 option transport_filter, meaning no filtering.
637 JH/17 Bug 1713: Fix non-PDKIM_DEBUG build. Patch from Jasen Betts.
639 JH/18 Bug 1709: When built with TLS support, the tls_advertise_hosts option now
640 defaults to "*" (all hosts). The variable is now available when not built
641 with TLS, default unset, mainly to enable keeping the testsuite sane.
642 If a server certificate is not supplied (via tls_certificate) an error is
643 logged, and clients will find TLS connections fail on startup. Presumably
644 they will retry in-clear.
645 Packagers of Exim are strongly encouraged to create a server certificate
646 at installation time.
648 HS/03 Add -bP config_file as a synonym for -bP configure_file, for consistency
649 with the $config_file variable.
651 JH/19 Two additional event types: msg:rcpt:defer and msg:rcpt:host:defer. Both
652 in transport context, after the attempt, and per-recipient. The latter type
653 is per host attempted. The event data is the error message, and the errno
654 information encodes the lookup type (A vs. MX) used for the (first) host,
655 and the trailing two digits of the smtp 4xx response.
657 GF/01 Bug 1715: Fix for race condition in exicyclog, where exim could attempt
658 to write to mainlog (or rejectlog, paniclog) in the window between file
659 creation and permissions/ownership being changed. Particularly affects
660 installations where exicyclog is run as root, rather than exim user;
661 result is that the running daemon panics and dies.
663 JH/20 Bug 1701: For MySQL lookups, support MySQL config file option group names.
665 JH/21 Bug 1720: Add support for priority groups and weighted-random proxy
666 selection for the EXPERIMENTAL_SOCKS feature, via new per-proxy options
667 "pri" and "weight". Note that the previous implicit priority given by the
668 list order is no longer honoured.
670 JH/22 Bugs 963, 1721: Fix some corner cases in message body canonicalization
673 JH/23 Move SOCKS5 support from Experimental to mainline, enabled for a build
674 by defining SUPPORT_SOCKS.
676 JH/26 Move PROXY support from Experimental to mainline, enabled for a build
677 by defining SUPPORT_PROXY. Note that the proxy_required_hosts option
678 is renamed to hosts_proxy, and the proxy_{host,target}_{address,port}.
679 variables are renamed to proxy_{local,external}_{address,port}.
681 JH/27 Move Internationalisation support from Experimental to mainline, enabled
682 for a build by defining SUPPORT_I18N
684 JH/28 Bug 1745: Fix redis lookups to handle (quoted) spaces embedded in parts
685 of the query string, and make ${quote_redis:} do that quoting.
687 JH/29 Move Events support from Experimental to mainline, enabled by default
688 and removable for a build by defining DISABLE_EVENT.
690 JH/30 Updated DANE implementation code to current from Viktor Dukhovni.
692 JH/31 Fix bug with hosts_connection_nolog and named-lists which were wrongly
693 cached by the daemon.
695 JH/32 Move Redis support from Experimental to mainline, enabled for a build
696 by defining LOOKUP_REDIS. The libhiredis library is required.
698 JH/33 Bug 1748: Permit ACL dnslists= condition in non-smtp ACLs if explicit
699 keys are given for lookup.
701 JH/34 Bug 1192: replace the embedded copy of PolarSSL RSA routines in the DKIM
702 support, by using OpenSSL or GnuTLS library ones. This means DKIM is
703 only supported when built with TLS support. The PolarSSL SHA routines
704 are still used when the TLS library is too old for convenient support.
706 JH/35 Require SINGLE_DH_USE by default in OpenSSL (main config option
707 openssl_options), for security. OpenSSL forces this from version 1.1.0
708 server-side so match that on older versions.
710 JH/36 Bug 1778: longstanding bug in memory use by the ${run } expansion: A fresh
711 allocation for $value could be released as the expansion processing
712 concluded, but leaving the global pointer active for it.
714 JH/37 Bug 1769: Permit a VRFY ACL to override the default 252 response,
715 and to use the domains and local_parts ACL conditions.
717 JH/38 Fix cutthrough bug with body lines having a single dot. The dot was
718 incorrectly not doubled on cutthrough transmission, hence seen as a
719 body-termination at the receiving system - resulting in truncated mails.
720 Commonly the sender saw a TCP-level error, and retransmitted the message
721 via the normal store-and-forward channel. This could result in duplicates
722 received - but deduplicating mailstores were liable to retain only the
723 initial truncated version.
725 JH/39 Bug 1781: Fix use of DKIM private-keys having trailing '=' in the base-64.
727 JH/40 Fix crash in queryprogram router when compiled with EXPERIMENTAL_SRS.
729 JH/41 Bug 1792: Fix selection of headers to sign for DKIM: bottom-up. While
730 we're in there, support oversigning also; bug 1309.
732 JH/42 Bug 1796: Fix error logged on a malware scanner connection failure.
734 HS/04 Add support for keep_environment and add_environment options.
736 JH/43 Tidy coding issues detected by gcc --fsanitize=undefined. Some remain;
737 either intentional arithmetic overflow during PRNG, or testing config-
740 JH/44 Bug 1800: The combination of a -bhc commandline option and cutthrough
741 delivery resulted in actual delivery. Cancel cutthrough before DATA
744 JH/45 Fix cutthrough, when connection not opened by verify and target hard-
745 rejects a recipient: pass the reject to the originator.
747 JH/46 Multiple issues raised by Coverity. Some were obvious or plausible bugs.
748 Many were false-positives and ignorable, but it's worth fixing the
751 JH/47 Fix build on HP-UX and older Solaris, which need (un)setenv now also
752 for the new environment-manipulation done at startup. Move the routines
753 from being local to tls.c to being global via the os.c file.
755 JH/48 Bug 1807: Fix ${extract } for the numeric/3-string case. While preparsing
756 an extract embedded as result-arg for a map, the first arg for extract
757 is unavailable so we cannot tell if this is a numbered or keyed
758 extraction. Accept either.
764 JH/01 Bug 1545: The smtp transport option "retry_include_ip_address" is now
767 JH/02 The smtp transport option "multi_domain" is now expanded.
769 JH/03 The smtp transport now requests PRDR by default, if the server offers
772 JH/04 Certificate name checking on server certificates, when exim is a client,
773 is now done by default. The transport option tls_verify_cert_hostnames
774 can be used to disable this per-host. The build option
775 EXPERIMENTAL_CERTNAMES is withdrawn.
777 JH/05 The value of the tls_verify_certificates smtp transport and main options
778 default to the word "system" to access the system default CA bundle.
779 For GnuTLS, only version 3.0.20 or later.
781 JH/06 Verification of the server certificate for a TLS connection is now tried
782 (but not required) by default. The verification status is now logged by
783 default, for both outbound TLS and client-certificate supplying inbound
786 JH/07 Changed the default rfc1413 lookup settings to disable calls. Few
789 JH/08 The EXPERIMENTAL_DSN compile option is no longer needed; all Delivery
790 Status Notification (bounce) messages are now MIME format per RFC 3464.
791 Support for RFC 3461 DSN options NOTIFY,ENVID,RET,ORCPT can be advertised
792 under the control of the dsn_advertise_hosts option, and routers may
793 have a dsn_lasthop option.
795 JH/09 A timeout of 2 minutes is now applied to all malware scanner types by
796 default, modifiable by a malware= option. The list separator for
797 the options can now be changed in the usual way. Bug 68.
799 JH/10 The smtp_receive_timeout main option is now expanded before use.
801 JH/11 The incoming_interface log option now also enables logging of the
802 local interface on delivery outgoing connections.
804 JH/12 The cutthrough-routing facility now supports multi-recipient mails,
805 if the interface and destination host and port all match.
807 JH/13 Bug 344: The verify = reverse_host_lookup ACL condition now accepts a
810 JH/14 Bug 1573: The spam= ACL condition now additionally supports Rspamd.
811 Patch from Andrew Lewis.
813 JH/15 Bug 670: The spamd_address main option (for the spam= ACL condition)
814 now supports optional time-restrictions, weighting, and priority
815 modifiers per server. Patch originally by <rommer@active.by>.
817 JH/16 The spamd_address main option now supports a mixed list of local
818 and remote servers. Remote servers can be IPv6 addresses, and
819 specify a port-range.
821 JH/17 Bug 68: The spamd_address main option now supports an optional
822 timeout value per server.
824 JH/18 Bug 1581: Router and transport options headers_add/remove can
825 now have the list separator specified.
827 JH/19 Bug 392: spamd_address, and clamd av_scanner, now support retry
830 JH/20 Bug 1571: Ensure that $tls_in_peerdn is set, when verification fails
833 JH/21 Support for the A6 type of dns record is withdrawn.
835 JH/22 Bug 608: The result of a QUIT or not-QUIT toplevel ACL now matters
836 rather than the verbs used.
838 JH/23 Bug 1572: Increase limit on SMTP confirmation message copy size
839 from 255 to 1024 chars.
841 JH/24 Verification callouts now attempt to use TLS by default.
843 HS/01 DNSSEC options (dnssec_require_domains, dnssec_request_domains)
844 are generic router options now. The defaults didn't change.
846 JH/25 Bug 466: Add RFC2322 support for MIME attachment filenames.
847 Original patch from Alexander Shikoff, worked over by JH.
849 HS/02 Bug 1575: exigrep falls back to autodetection of compressed
850 files if ZCAT_COMMAND is not executable.
852 JH/26 Bug 1539: Add timeout/retry options on dnsdb lookups.
854 JH/27 Bug 286: Support SOA lookup in dnsdb lookups.
856 JH/28 Bug 1588: Do not use the A lookup following an AAAA for setting the FQDN.
857 Normally benign, it bites when the pair was led to by a CNAME;
858 modern usage is to not canonicalize the domain to a CNAME target
859 (and we were inconsistent anyway for A-only vs AAAA+A).
861 JH/29 Bug 1632: Removed the word "rejected" from line logged for ACL discards.
863 JH/30 Check the forward DNS lookup for DNSSEC, in addition to the reverse,
864 when evaluating $sender_host_dnssec.
866 JH/31 Check the HELO verification lookup for DNSSEC, adding new
867 $sender_helo_dnssec variable.
869 JH/32 Bug 1397: Enable ECDHE on OpenSSL, just the NIST P-256 curve.
871 JH/33 Bug 1346: Note MAIL cmd seen in -bS batch, to avoid smtp_no_mail log.
873 JH/34 Bug 1648: Fix a memory leak seen with "mailq" and large queues.
875 JH/35 Bug 1642: Fix support of $spam_ variables at delivery time. Was
876 documented as working, but never had. Support all but $spam_report.
878 JH/36 Bug 1659: Guard checking of input smtp commands again pseudo-command
879 added for tls authenticator.
881 HS/03 Add perl_taintmode main config option
887 TL/01 When running the test suite, the README says that variables such as
888 no_msglog_check are global and can be placed anywhere in a specific
889 test's script, however it was observed that placement needed to be near
890 the beginning for it to behave that way. Changed the runtest perl
891 script to read through the entire script once to detect and set these
892 variables, reset to the beginning of the script, and then run through
893 the script parsing/test process like normal.
895 TL/02 The BSD's have an arc4random API. One of the functions to induce
896 adding randomness was arc4random_stir(), but it has been removed in
897 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
898 function when detected.
900 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
901 cause callback expansion.
903 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
904 syntax errors in an expansion can be treated as a string instead of
905 logging or causing an error, due to the internal use of bool_lax
906 instead of bool when processing it.
908 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
909 server certificates when making smtp deliveries.
911 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
913 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
915 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
917 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
918 Merged patch from Sebastian Wiedenroth.
920 JH/05 Fix results-pipe from transport process. Several recipients, combined
921 with certificate use, exposed issues where response data items split
922 over buffer boundaries were not parsed properly. This eventually
923 resulted in duplicates being sent. This issue only became common enough
924 to notice due to the introduction of connection certificate information,
925 the item size being so much larger. Found and fixed by Wolfgang Breyha.
927 JH/06 Bug 1533: Fix truncation of items in headers_remove lists. A fixed
928 size buffer was used, resulting in syntax errors when an expansion
931 JH/07 Add support for directories of certificates when compiled with a GnuTLS
932 version 3.3.6 or later.
934 JH/08 Rename the TPDA experimental facility to Event Actions. The #ifdef
935 is EXPERIMENTAL_EVENT, the main-configuration and transport options
936 both become "event_action", the variables become $event_name, $event_data
937 and $event_defer_errno. There is a new variable $verify_mode, usable in
938 routers, transports and related events. The tls:cert event is now also
939 raised for inbound connections, if the main configuration event_action
942 TL/06 In test suite, disable OCSP for old versions of openssl which contained
943 early OCSP support, but no stapling (appears to be less than 1.0.0).
945 JH/09 When compiled with OpenSSL and EXPERIMENTAL_CERTNAMES, the checks on
946 server certificate names available under the smtp transport option
947 "tls_verify_cert_hostname" now do not permit multi-component wildcard
950 JH/10 Time-related extraction expansions from certificates now use the main
951 option "timezone" setting for output formatting, and are consistent
952 between OpenSSL and GnuTLS compilations. Bug 1541.
954 JH/11 Fix a crash in mime ACL when meeting a zero-length, quoted or RFC2047-
955 encoded parameter in the incoming message. Bug 1558.
957 JH/12 Bug 1527: Autogrow buffer used in reading spool files. Since they now
958 include certificate info, eximon was claiming there were spoolfile
961 JH/13 Bug 1521: Fix ldap lookup for single-attr request, multiple-attr return.
963 JH/14 Log delivery-related information more consistently, using the sequence
964 "H=<name> [<ip>]" wherever possible.
966 TL/07 Bug 1547: Omit RFCs from release. Draft and RFCs have licenses which
967 are problematic for Debian distribution, omit them from the release
970 JH/15 Updates and fixes to the EXPERIMENTAL_DSN feature.
972 JH/16 Fix string representation of time values on 64bit time_t architectures.
975 JH/17 Fix a null-indirection in certextract expansions when a nondefault
976 output list separator was used.
981 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
982 checkers that were complaining about end of non-void function with no
985 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
986 This was a regression introduced in 4.83 by another bugfix.
988 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
990 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
991 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
997 TF/01 Correctly close the server side of TLS when forking for delivery.
999 When a message was received over SMTP with TLS, Exim failed to clear up
1000 the incoming connection properly after forking off the child process to
1001 deliver the message. In some situations the subsequent outgoing
1002 delivery connection happened to have the same fd number as the incoming
1003 connection previously had. Exim would try to use TLS and fail, logging
1004 a "Bad file descriptor" error.
1006 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
1007 utilities have not been installed.
1009 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
1010 temporary space as the ACL may create new global variables.
1012 TL/01 LDAP support uses per connection or global context settings, depending
1013 upon the detected version of the libraries at build time.
1015 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
1016 to extract and use the src ip:port in logging and expansions as if it
1017 were a direct connection from the outside internet. PPv2 support was
1018 updated based on HAProxy spec change in May 2014.
1020 JH/02 Add ${listextract {number}{list}{success}{fail}}.
1022 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
1023 Properly escape header and check for NULL return.
1025 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
1028 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
1030 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
1031 characters in header names, implemented as a verify condition.
1032 Contributed by Michael Fischer v. Mollard.
1034 TL/05 Rename SPF condition results err_perm and err_temp to standardized
1035 results permerror and temperror. Previous values are deprecated but
1036 still accepted. In a future release, err_perm and err_temp will be
1037 completely removed, which will be a backward incompatibility if the
1038 ACL tests for either of these two old results. Patch contributed by
1039 user bes-internal on the mailing list.
1041 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
1043 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
1044 selectors, in both main and reject logs.
1046 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
1049 JH/07 Add malware type "sock" for talking to simple daemon.
1051 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
1053 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
1054 routers/transports under cutthrough routing.
1056 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
1057 numbers. Touch up "bool" conditional to keep the same definition.
1059 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
1061 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
1063 JH/12 Expand items in router/transport headers_add or headers_remove lists
1064 individually rather than the list as a whole. Bug 1452.
1066 Required for reasonable handling of multiple headers_ options when
1067 they may be empty; requires that headers_remove items with embedded
1068 colons must have them doubled (or the list-separator changed).
1070 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
1071 view the policy declared in the DMARC record. Currently, $dmarc_status
1072 is a combined value of both the record presence and the result of the
1075 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
1077 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
1078 dnslookup router and the smtp transport (applying to the forward
1081 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
1082 of ldap servers used for a specific lookup. Patch provided by Heiko
1085 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
1086 New variable $lookup_dnssec_authenticated for observability.
1088 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
1089 Patch submitted by Lars Timman.
1091 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
1093 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
1094 Requires trusted mode and valid format message id, aborts otherwise.
1095 Patch contributed by Heiko Schlichting.
1097 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
1098 certextract with support for various fields. Bug 1358.
1100 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
1101 is requested by default, modifiable by smtp transport option
1104 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
1105 operate on certificate variables to give certificate fingerprints
1106 Also new ${sha256:cert_variable}.
1108 JH/23 The PRDR feature is moved from being Experimental into the mainline.
1110 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
1111 Christian Aistleitner.
1113 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
1115 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
1116 file. Patch from Wolfgang Breyha.
1118 JH/25 Expand the coverage of the delivery $host and $host_address to
1119 client authenticators run in verify callout. Bug 1476.
1121 JH/26 Port service names are now accepted for tls_on_connect_ports, to
1122 align with daemon_smtp_ports. Bug 72.
1124 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
1125 support and error reporting did not work properly.
1127 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
1128 and is readable. Patch from Andrew Colin Kissa.
1130 TL/14 Enhance documentation of ${run expansion and how it parses the
1131 commandline after expansion, particularly in the case when an
1132 unquoted variable expansion results in an empty value.
1134 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
1136 PP/02 Fix internal collision of T_APL on systems which support RFC3123
1137 by renaming away from it. Addresses GH issue 15, reported by
1140 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
1142 TL/15 SECURITY: prevent double expansion in math comparison functions
1143 (can expand unsanitized data). Not remotely exploitable.
1150 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
1152 PP/02 Make -n do something, by making it not do something.
1153 When combined with -bP, the name of an option is not output.
1155 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
1158 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
1159 $sender_host_name and config options to manage this, and basic check
1162 PP/05 DSCP support for outbound connections and control modifier for inbound.
1164 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
1165 (Only plugin which currently uses this is kerberos4, which nobody should
1166 be using, but we should make it available and other future plugins might
1167 conceivably use it, even though it would break NAT; stuff *should* be
1168 using channel bindings instead).
1170 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
1171 name; added for Sendmail compatibility; requires admin caller.
1172 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
1173 just ignore it); requires trusted caller.
1174 Also parse but ignore: -Ac -Am -X<logfile>
1177 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
1179 TL/02 Add +smtp_confirmation as a default logging option.
1181 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
1182 Patch by Magnus Holmgren from 2007-02-20.
1184 TL/04 Bugzilla 1281 - Spec typo.
1185 Bugzilla 1283 - Spec typo.
1186 Bugzilla 1290 - Spec grammar fixes.
1188 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
1190 TL/06 Add Experimental DMARC support using libopendmarc libraries.
1192 TL/07 Fix an out of order global option causing a segfault. Reported to dev
1193 mailing list by by Dmitry Isaikin.
1195 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
1197 JH/02 Support "G" suffix to numbers in ${if comparisons.
1199 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
1201 NM/01 Bugzilla 1197 - Spec typo
1202 Bugzilla 1196 - Spec examples corrections
1204 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
1206 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
1207 gnutls_enable_pkcs11, but renamed to more accurately indicate its
1210 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
1211 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
1213 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
1214 "acl {{name}{arg}...}", and optional args on acl condition
1217 JH/05 Permit multiple router/transport headers_add/remove lines.
1219 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
1221 JH/07 Avoid using a waiting database for a single-message-only transport.
1222 Performance patch from Paul Fisher. Bugzilla 1262.
1224 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
1227 JH/09 Add $headers_added variable, with content from use of ACL modifier
1228 add_header (but not yet added to the message). Bugzilla 199.
1230 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
1231 Pulled from Bugzilla 817 by Wolfgang Breyha.
1233 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
1235 (nb: this is the same fix as in Exim 4.80.1)
1237 JH/11 Add A= logging on delivery lines, and a client_set_id option on
1240 JH/12 Add optional authenticated_sender logging to A= and a log_selector
1243 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
1245 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
1246 advertise SMTP AUTH mechanism to us, instead of a generic
1247 protocol violation error. Also, make Exim more robust to bad
1248 data from the Dovecot auth socket.
1250 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
1252 When a queue runner is handling a message, Exim first routes the
1253 recipient addresses, during which it prunes them based on the retry
1254 hints database. After that it attempts to deliver the message to
1255 any remaining recipients. It then updates the hints database using
1258 So if a recipient address works intermittently, it can get repeatedly
1259 deferred at routing time. The retry hints record remains fresh so the
1260 address never reaches the final cutoff time.
1262 This is a fairly common occurrence when a user is bumping up against
1263 their storage quota. Exim had some logic in its local delivery code
1264 to deal with this. However it did not apply to per-recipient defers
1265 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
1267 This change adds a proper retry rule check during routing so that the
1268 final cutoff time is checked against the message's age. We only do
1269 this check if there is an address retry record and there is not a
1270 domain retry record; this implies that previous attempts to handle
1271 the address had the retry_use_local_parts option turned on. We use
1272 this as an approximation for the destination being like a local
1273 delivery, as in LMTP.
1275 I suspect this new check makes the old local delivery cutoff check
1276 redundant, but I have not verified this so I left the code in place.
1278 TF/02 Correct gecos expansion when From: is a prefix of the username.
1280 Test 0254 submits a message to Exim with the header
1284 When I ran the test suite under the user fanf2, Exim expanded
1285 the header to contain my full name, whereas it should have added
1286 a Resent-Sender: header. It erroneously treats any prefix of the
1287 username as equal to the username.
1289 This change corrects that bug.
1291 GF/01 DCC debug and logging tidyup
1292 Error conditions log to paniclog rather than rejectlog.
1293 Debug lines prefixed by "DCC: " to remove any ambiguity.
1295 TF/03 Avoid unnecessary rebuilds of lookup-related code.
1297 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
1298 Bug spotted by Jeremy Harris; was flawed since initial commit.
1299 Would have resulted in OCSP responses post-SNI triggering an Exim
1300 NULL dereference and crash.
1302 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
1304 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
1305 Bug detection, analysis and fix by Samuel Thibault.
1306 Bugzilla 1331, Debian bug #698092.
1308 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
1310 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
1311 Server implementation by Todd Lyons, client by JH.
1312 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
1313 config variable "prdr_enable" controls whether the server
1314 advertises the facility. If the client requests PRDR a new
1315 acl_data_smtp_prdr ACL is called once for each recipient, after
1316 the body content is received and before the acl_smtp_data ACL.
1317 The client is controlled by both of: a hosts_try_prdr option
1318 on the smtp transport, and the server advertisement.
1319 Default client logging of deliveries and rejections involving
1320 PRDR are flagged with the string "PRDR".
1322 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
1323 fclose(). Diagnosis by Todd Lyons.
1325 PP/17 Update configure.default to handle IPv6 localhost better.
1326 Patch by Alain Williams (plus minor tweaks).
1329 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
1330 This is now consistent with GnuTLS, and is now documented: the
1331 previous undocumented portable approach to treating the option as
1332 unset was to force an expansion failure. That still works, and
1333 an empty string is now equivalent.
1335 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
1336 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
1337 not performing validation itself.
1339 PP/20 Added force_command boolean option to pipe transport.
1340 Patch from Nick Koston, of cPanel Inc.
1342 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
1345 TF/04 Added udpsend ACL modifier and hexquote expansion operator
1347 PP/21 Fix eximon continuous updating with timestamped log-files.
1348 Broken in a format-string cleanup in 4.80, missed when I repaired the
1349 other false fix of the same issue.
1350 Report and fix from Heiko Schlichting.
1353 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
1354 Report from Prashanth Katuri.
1356 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
1357 It's SecureTransport, so affects any MacOS clients which use the
1358 system-integrated TLS libraries, including email clients.
1360 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
1361 using a MIME ACL for non-SMTP local injection.
1362 Report and assistance in diagnosis by Warren Baker.
1364 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
1366 JH/16 Fix comparisons for 64b. Bugzilla 1385.
1368 TL/09 Add expansion variable $authenticated_fail_id to keep track of
1369 last id that failed so it may be referenced in subsequent ACL's.
1371 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
1374 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
1375 ldap library initialization, allowing self-signed CA's to be
1376 used. Also properly sets require_cert option later in code by
1377 using NULL (global ldap config) instead of ldap handle (per
1378 session). Bug diagnosis and testing by alxgomz.
1380 TL/12 Enhanced documentation in the ratelimit.pl script provided in
1381 the src/util/ subdirectory.
1383 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
1384 renamed to Transport Post Delivery Action by Jeremy Harris, as
1387 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
1388 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
1389 redis_servers = needs to be configured which will be used by the redis
1390 lookup. Patch from Warren Baker, of The Packet Hub.
1392 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
1394 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
1395 hostname or reverse DNS when processing a host list. Used suggestions
1396 from multiple comments on this bug.
1398 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
1400 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
1401 Missed a few lines, added it to make the runtest require no keyboard
1404 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
1405 contains upper case chars. Make router use caseful_local_part.
1407 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
1408 support when GnuTLS has been built with p11-kit.
1414 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
1416 This, or similar/improved, will also be change PP/11 of 4.82.
1422 PP/01 Handle short writes when writing local log-files.
1423 In practice, only affects FreeBSD (8 onwards).
1424 Bugzilla 1053, with thanks to Dmitry Isaikin.
1426 NM/01 Bugzilla 949 - Documentation tweak
1428 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
1431 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
1433 PP/02 Implemented gsasl authenticator.
1435 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
1437 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
1438 `pkg-config foo` for cflags/libs.
1440 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
1441 with rest of GSASL and with heimdal_gssapi.
1443 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
1444 `pkg-config foo` for cflags/libs for the TLS implementation.
1446 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
1447 properties get this fed in as external SSF. A number of robustness
1448 and debugging improvements to the cyrus_sasl authenticator.
1450 PP/08 cyrus_sasl server now expands the server_realm option.
1452 PP/09 Bugzilla 1214 - Log authentication information in reject log.
1453 Patch by Jeremy Harris.
1455 PP/10 Added dbmjz lookup type.
1457 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
1459 PP/12 MAIL args handles TAB as well as SP, for better interop with
1460 non-compliant senders.
1461 Analysis and variant patch by Todd Lyons.
1463 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
1464 Bug report from Lars Müller <lars@samba.org> (via SUSE),
1465 Patch from Dirk Mueller <dmueller@suse.com>
1467 PP/13 tls_peerdn now print-escaped for spool files.
1468 Observed some $tls_peerdn in wild which contained \n, which resulted
1469 in spool file corruption.
1471 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
1472 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
1473 or write after TLS renegotiation, which otherwise led to messages
1476 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
1477 as a tracking header (ie: a signed header comes before the signature).
1478 Patch from Wolfgang Breyha.
1480 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
1481 comma-sep list; embedded commas doubled.
1483 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
1485 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
1487 Report and patch from Dmitry Banschikov.
1489 PP/16 Removed "dont_insert_empty_fragments" from "openssl_options".
1490 Removed SSL_clear() after SSL_new() which led to protocol negotiation
1491 failures. We appear to now support TLS1.1+ with Exim.
1493 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
1494 lets Exim select keys and certificates based upon TLS SNI from client.
1495 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
1496 before an outbound SMTP session. New log_selector, +tls_sni.
1498 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
1499 NULL dereference. Report and patch from Alun Jones.
1501 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
1502 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
1503 Not seeing resolver debug output on NetBSD, but suspect this is a
1504 resolver implementation change.
1506 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
1507 Left warnings. Added "eximon gdb" invocation mode.
1509 PP/21 Defaulting "accept_8bitmime" to true, not false.
1511 PP/22 Added -bw for inetd wait mode support.
1513 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
1514 locate the relevant includes and libraries. Made this the default.
1516 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
1517 Bugzilla 1246, report and most of solution from Tomasz Kusy.
1519 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
1520 This may cause build issues on older platforms.
1522 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
1523 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
1524 gnutls_require_mac & gnutls_require_protocols (no longer supported).
1525 Added SNI support via GnuTLS too.
1526 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
1528 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
1530 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
1531 Applied second patch from Janne, implementing suggestion to default
1532 multiple-strings-in-record handling to match SPF spec.
1534 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
1536 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
1537 read-only, out of scope).
1538 Patch from Wolfgang Breyha, report from Stuart Northfield.
1540 PP/29 Fix three issues highlighted by clang analyser static analysis.
1541 Only crash-plausible issue would require the Cambridge-specific
1542 iplookup router and a misconfiguration.
1543 Report from Marcin Mirosław.
1545 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
1547 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
1548 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
1549 As part of this, removing so much warning spew let me fix some minor
1550 real issues in debug logging.
1552 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
1553 assignment on my part. Fixed.
1555 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
1556 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
1557 Janne Snabb (who went above and beyond: thank you).
1559 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
1560 string otherwise requires a connection and a bunch more work and it's
1561 relatively easy to get wrong. Should also expose TLS library linkage
1564 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
1565 64-bit ${eval} (JH/03).
1567 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
1568 GNU libc to support some of the 64-bit stuff, should not lead to
1569 conflicts. Defined before os.h is pulled in, so if a given platform
1570 needs to override this, it can.
1572 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
1573 protection layer was required, which is not implemented.
1574 Bugzilla 1254, patch from Wolfgang Breyha.
1576 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
1577 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
1578 tls_dhparam take prime identifiers. Also unbreak combination of
1579 OpenSSL+DH_params+TLSSNI.
1581 PP/39 Disable SSLv2 by default in OpenSSL support.
1587 PP/01 Solaris build fix for Oracle's LDAP libraries.
1588 Bugzilla 1109, patch from Stephen Usher.
1590 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
1592 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
1595 TF/02 Fix a couple more cases where we did not log the error message
1596 when unlink() failed. See also change 4.74-TF/03.
1598 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
1599 lock up or crash if it happened to be inside a call to libc when it
1600 got a SIGUSR1 from exiwhat.
1602 The SIGUSR1 handler appends the current process status to the process
1603 log which is later printed by exiwhat. It used to use the general
1604 purpose logging code to do this, but several functions it calls are
1605 not safe for signals.
1607 The new output code in the SIGUSR1 handler is specific to the process
1608 log, and simple enough that it's easy to inspect for signal safety.
1609 Removing some special cases also simplifies the general logging code.
1610 Removing the spurious timestamps from the process log simplifies
1613 TF/04 Improved ratelimit ACL condition.
1615 The /noupdate option has been deprecated in favour of /readonly which
1616 has clearer semantics. The /leaky, /strict, and /readonly update modes
1617 are mutually exclusive. The update mode is no longer included in the
1618 database key; it just determines when the database is updated. (This
1619 means that when you upgrade Exim will forget old rate measurements.)
1621 Exim now checks that the per_* options are used with an update mode that
1622 makes sense for the current ACL. For example, when Exim is processing a
1623 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
1624 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
1625 must specify per_mail/readonly. If you omit the update mode it defaults to
1626 /leaky where that makes sense (as before) or /readonly where required.
1628 The /noupdate option is now undocumented but still supported for
1629 backwards compatibility. It is equivalent to /readonly except that in
1630 ACLs where /readonly is required you may specify /leaky/noupdate or
1631 /strict/noupdate which are treated the same as /readonly.
1633 A useful new feature is the /count= option. This is a generalization
1634 of the per_byte option, so that you can measure the throughput of other
1635 aggregate values. For example, the per_byte option is now equivalent
1636 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
1638 The per_rcpt option has been generalized using the /count= mechanism
1639 (though it's more complicated than the per_byte equivalence). When it is
1640 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
1641 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
1642 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
1643 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
1644 Note that using per_rcpt with a non-readonly update mode in more than
1645 one ACL will cause the recipients to be double-counted. (The per_mail
1646 and per_byte options don't have this problem.)
1648 The handling of very low rates has changed slightly. If the computed rate
1649 is less than the event's count (usually one) then this event is the first
1650 after a long gap. In this case the rate is set to the same as this event's
1651 count, so that the first message of a spam run is counted properly.
1653 The major new feature is a mechanism for counting the rate of unique
1654 events. The new per_addr option counts the number of different
1655 recipients that someone has sent messages to in the last time period. It
1656 behaves like per_rcpt if all the recipient addresses are different, but
1657 duplicate recipient addresses do not increase the measured rate. Like
1658 the /count= option this is a general mechanism, so the per_addr option
1659 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
1660 example, measure the rate that a client uses different sender addresses
1661 with the options per_mail/unique=$sender_address. There are further
1662 details in the main documentation.
1664 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
1666 TF/06 Removed a few PCRE remnants.
1668 TF/07 Automatically extract Exim's version number from tags in the git
1669 repository when doing development or release builds.
1671 PP/02 Raise smtp_cmd_buffer_size to 16kB.
1672 Bugzilla 879. Patch from Paul Fisher.
1674 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
1675 Heavily based on revision 40f9a89a from Simon Arlott's tree.
1678 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
1680 PP/05 Variable $av_failed, true if the AV scanner deferred.
1681 Bugzilla 1078. Patch from John Horne.
1683 PP/06 Stop make process more reliably on build failure.
1684 Bugzilla 1087. Patch from Heiko Schlittermann.
1686 PP/07 Make maildir_use_size_file an _expandable_ boolean.
1687 Bugzilla 1089. Patch from Heiko Schlittermann.
1689 PP/08 Handle ${run} returning more data than OS pipe buffer size.
1690 Bugzilla 1131. Patch from Holger Weiß.
1692 PP/09 Handle IPv6 addresses with SPF.
1693 Bugzilla 860. Patch from Wolfgang Breyha.
1695 PP/10 GnuTLS: support TLS 1.2 & 1.1.
1697 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
1700 PP/11 match_* no longer expand right-hand-side by default.
1701 New compile-time build option, EXPAND_LISTMATCH_RHS.
1702 New expansion conditions, "inlist", "inlisti".
1704 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
1706 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
1708 PP/14 fix log_write() format string regression from TF/03.
1709 Bugzilla 1152. Patch from Dmitry Isaikin.
1715 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
1717 PP/02 Harmonised TLS library version reporting; only show if debugging.
1718 Layout now matches that introduced for other libraries in 4.74 PP/03.
1720 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
1722 PP/04 New "dns_use_edns0" global option.
1724 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
1727 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
1728 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
1730 TK/01 Updated PolarSSL code to 0.14.2.
1731 Bugzilla 1097. Patch from Andreas Metzler.
1733 PP/07 Catch divide-by-zero in ${eval:...}.
1734 Fixes bugzilla 1102.
1736 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
1739 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
1740 format-string attack -- SECURITY: remote arbitrary code execution.
1742 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
1743 time unintentionally subject to list matching rules, letting the header
1744 cause arbitrary Exim lookups (of items which can occur in lists, *not*
1745 arbitrary string expansion). This allowed for information disclosure.
1747 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
1748 INT_MIN/-1 -- value coerced to INT_MAX.
1754 NM/01 Workaround for PCRE version dependency in version reporting
1757 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
1758 This fixes portability to compilers other than gcc, notably
1759 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
1761 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
1762 makefiles for portability to HP-UX and POSIX correctness.
1764 PP/01 Permit LOOKUP_foo enabling on the make command-line.
1765 Also via indented variable definition in the Makefile.
1766 (Debugging by Oliver Heesakkers).
1768 PP/02 Restore caching of spamd results with expanded spamd_address.
1769 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
1771 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
1772 Improves build reliability. Fix from: Frank Elsner
1774 NM/02 Fix wide character breakage in the rfc2047 coding
1775 Fixes bug 1064. Patch from Andrey N. Oktyabrski
1777 NM/03 Allow underscore in dnslist lookups
1778 Fixes bug 1026. Patch from Graeme Fowler
1780 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
1781 Code patches from Adam Ciarcinski of NetBSD.
1783 NM/04 Fixed exiqgrep to cope with mailq missing size issue
1786 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
1787 is logged, to avoid truncation. Patch from John Horne.
1789 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
1790 Patch from Jakob Hirsch.
1792 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
1793 SQL string expansion failure details.
1794 Patch from Andrey Oktyabrski.
1796 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
1797 Patch from Simon Arlott.
1799 PP/09 New lookups functionality failed to compile on old gcc which rejects
1800 extern declarations in function scope.
1801 Patch from Oliver Fleischmann
1803 PP/10 Use sig_atomic_t for flags set from signal handlers.
1804 Check getgroups() return and improve debugging.
1805 Fixed developed for diagnosis in bug 927 (which turned out to be
1808 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
1809 Patch from Mark Zealey.
1811 PP/12 Bugzilla 1056: Improved spamd server selection.
1812 Patch from Mark Zealey.
1814 PP/13 Bugzilla 1086: Deal with maildir quota file races.
1815 Based on patch from Heiko Schlittermann.
1817 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
1818 Patch from Uwe Doering, sign-off by Michael Haardt.
1820 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
1821 variable declaration deep within a block. Bug and patch from
1824 PP/15 lookups-Makefile IRIX compatibility coercion.
1826 PP/16 Make DISABLE_DKIM build knob functional.
1828 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
1829 Patch by Simon Arlott
1831 TF/03 Fix valgrind.h portability to C89 compilers that do not support
1832 variable argument macros. Our copy now differs from upstream.
1838 TF/01 Failure to get a lock on a hints database can have serious
1839 consequences so log it to the panic log.
1841 TF/02 Log LMTP confirmation messages in the same way as SMTP,
1842 controlled using the smtp_confirmation log selector.
1844 TF/03 Include the error message when we fail to unlink a spool file.
1846 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
1847 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
1848 for maintaining out-of-tree patches for some time.
1850 PP/01 Bugzilla 139: Documentation and portability issues.
1851 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
1852 Handle per-OS dynamic-module compilation flags.
1854 PP/02 Let /dev/null have normal permissions.
1855 The 4.73 fixes were a little too stringent and complained about the
1856 permissions on /dev/null. Exempt it from some checks.
1857 Reported by Andreas M. Kirchwitz.
1859 PP/03 Report version information for many libraries, including
1860 Exim version information for dynamically loaded libraries. Created
1861 version.h, now support a version extension string for distributors
1862 who patch heavily. Dynamic module ABI change.
1864 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
1865 privilege escalation vulnerability whereby the Exim run-time user
1866 can cause root to append content of the attacker's choosing to
1869 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
1872 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
1873 If dropping privileges for untrusted macros, we disabled normal logging
1874 on the basis that it would fail; for the Exim run-time user, this is not
1875 the case, and it resulted in successful deliveries going unlogged.
1876 Fixed. Reported by Andreas Metzler.
1882 PP/01 Date: & Message-Id: revert to normally being appended to a message,
1883 only prepend for the Resent-* case. Fixes regression introduced in
1884 Exim 4.70 by NM/22 for Bugzilla 607.
1886 PP/02 Include check_rfc2047_length in configure.default because we're seeing
1887 increasing numbers of administrators be bitten by this.
1889 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
1891 PP/03 Bugzilla 994: added openssl_options main configuration option.
1893 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
1895 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
1897 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
1899 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
1900 without a peer certificate, leading to a segfault because of an
1901 assumption that peers always have certificates. Be a little more
1902 paranoid. Problem reported by Martin Tscholak.
1904 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
1905 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
1906 NB: ClamAV planning to remove STREAM in "middle of 2010".
1907 CL also introduces -bmalware, various -d+acl logging additions and
1908 more caution in buffer sizes.
1910 PP/09 Implemented reverse_ip expansion operator.
1912 PP/10 Bugzilla 937: provide a "debug" ACL control.
1914 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
1916 PP/12 Bugzilla 973: Implement --version.
1918 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
1920 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
1922 PP/15 Bugzilla 816: support multiple condition rules on Routers.
1924 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
1925 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
1926 ignore trailing whitespace.
1928 JJ/02 prevent non-panic DKIM error from being sent to paniclog
1930 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
1933 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
1934 Notification from Dr Andrew Aitchison.
1936 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
1937 ExtendedDetectionInfo response format.
1938 Notification from John Horne.
1940 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1943 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1944 XSL and documented dependency on system catalogs, with examples of how
1947 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1950 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1951 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1952 configuration file which is writeable by the Exim user or group.
1954 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1955 of configuration files to cover files specified with the -C option if
1956 they are going to be used with root privileges, not just the default
1959 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1960 option (effectively making it always true).
1962 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1963 files to be used while preserving root privileges.
1965 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1966 that rogue child processes cannot use them.
1968 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1969 run-time user, instead of root.
1971 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1972 Exim run-time user without dropping privileges.
1974 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1975 result string, instead of calling string_vformat() twice with the same
1978 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1979 for other users. Others should always drop root privileges if they use
1980 -C on the command line, even for a whitelisted configure file.
1982 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1984 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1990 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1991 $data_path, and $header_path variables; fixed documentation bugs and
1994 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1995 exipick to access non-standard spools, including the "frozen" queue
1998 NM/01 Bugzilla 965: Support mysql stored procedures.
1999 Patch from Alain Williams
2001 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
2003 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
2004 Patch from Andreas Metzler
2006 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
2007 Patch from Kirill Miazine
2009 NM/05 Bugzilla 671: Added umask to procmail example.
2011 JJ/03 installed exipick 20100323.0, fixing doc bug
2013 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
2014 directory. Notification and patch from Dan Rosenberg.
2016 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
2018 TK/02 Improve log output when DKIM signing operation fails.
2020 MH/01 Treat the transport option dkim_domain as a colon separated
2021 list, not as a single string, and sign the message with each element,
2022 omitting multiple occurences of the same signer.
2024 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
2025 Bugzilla 985, 986. Patch by Simon Arlott
2027 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
2028 Patch by Simon Arlott
2030 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
2031 MBX locking. Notification from Dan Rosenberg.
2037 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
2039 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
2041 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
2043 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
2045 NM/04 Bugzilla 915: Flag broken perl installation during build.
2051 TK/01 Added patch by Johannes Berg that expands the main option
2052 "spamd_address" if it starts with a dollar sign.
2054 TK/02 Write list of recipients to X-Envelope-Sender header when building
2055 the mbox-format spool file for content scanning (suggested by Jakob
2058 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
2059 (http://www.dcc-servers.net/) support via dccifd. Activated by
2060 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
2062 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
2063 by Mark Daniel Reidel <mr@df.eu>.
2065 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
2066 When building exim an external PCRE library is now needed -
2067 PCRE is a system library on the majority of modern systems.
2068 See entry on PCRE_LIBS in EDITME file.
2070 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
2071 conversation. Added nologin parameter to request.
2072 Patch contributed by Kirill Miazine.
2074 TF/01 Do not log submission mode rewrites if they do not change the address.
2076 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
2078 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
2079 log files in place. Contributed by Roberto Lima.
2081 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
2083 TF/03 Bugzilla 615: When checking the local_parts router precondition
2084 after a local_part_suffix or local_part_prefix option, Exim now
2085 does not use the address's named list lookup cache, since this
2086 contains cached lookups for the whole local part.
2088 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
2089 Robert Millan. Documentation is in experimental-spec.txt.
2091 TF/04 Bugzilla 668: Fix parallel build (make -j).
2093 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
2095 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
2096 Patch provided by Jan Srzednicki.
2098 TF/05 Leading white space used to be stripped from $spam_report which
2099 wrecked the formatting. Now it is preserved.
2101 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
2102 that they are available at delivery time.
2104 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
2106 TF/08 TLS error reporting now respects the incoming_interface and
2107 incoming_port log selectors.
2109 TF/09 Produce a more useful error message if an SMTP transport's hosts
2110 setting expands to an empty string.
2112 NM/06 Bugzilla 744: EXPN did not work under TLS.
2113 Patch provided by Phil Pennock.
2115 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
2116 Patch provided by Richard Godbee.
2118 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
2119 acl_smtp_notquit, added index entry.
2121 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
2122 Patch provided by Eugene Bujak.
2124 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
2125 accept(). Patch provided by Maxim Dounin.
2127 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
2128 Patch provided by Phil Pennock.
2130 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
2132 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
2133 Patch provided by Brad "anomie" Jorsch.
2135 NM/14 Bugzilla 622: Added timeout setting to transport filter.
2136 Patch provided by Dean Brooks.
2138 TK/05 Add native DKIM support (does not depend on external libraries).
2140 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
2141 Patch provided by Graeme Fowler.
2143 NM/16 Bugzilla 851: Documentation example syntax fix.
2145 NM/17 Changed NOTICE file to remove references to embedded PCRE.
2147 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
2150 NM/19 Bugzilla 745: TLS version reporting.
2151 Patch provided by Phil Pennock.
2153 NM/20 Bugzilla 167: bool: condition support.
2154 Patch provided by Phil Pennock.
2156 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
2157 clients. Patch provided by Phil Pennock.
2159 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
2160 Patch provided by Brad "anomie" Jorsch.
2162 NM/23 Bugzilla 687: Fix misparses in eximstats.
2163 Patch provided by Heiko Schlittermann.
2165 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
2166 Patch provided by Heiko Schlittermann.
2168 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
2169 plus update to original patch.
2171 NM/26 Bugzilla 799: Documentation correction for ratelimit.
2173 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
2174 Patch provided by David Brownlee.
2176 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
2178 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
2180 NM/30 Bugzilla 888: TLS documentation bugfixes.
2182 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
2184 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
2185 Unlike the original bugzilla I have changed all shell scripts in src tree.
2187 NM/33 Bugzilla 898: Transport filter timeout fix.
2188 Patch by Todd Rinaldo.
2190 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
2191 Patch by Serge Demonchaux.
2193 NM/35 Bugzilla 39: Base64 decode bug fixes.
2194 Patch by Jakob Hirsch.
2196 NM/36 Bugzilla 909: Correct connect() call in dcc code.
2198 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
2200 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
2202 NM/39 Bugzilla 911: Fixed MakeLinks build script.
2208 TK/01 Add preliminary DKIM support. Currently requires a forked version of
2209 ALT-N's libdkim that I have put here:
2210 http://duncanthrax.net/exim-experimental/
2212 Note to Michael Haardt: I had to rename some vars in sieve.c. They
2213 were called 'true' and it seems that C99 defines that as a reserved
2214 keyword to be used with 'bool' variable types. That means you could
2215 not include C99-style headers which use bools without triggering
2216 build errors in sieve.c.
2218 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
2219 as mailq or other aliases. Changed the --help handling significantly
2220 to do whats expected. exim_usage() emits usage/help information.
2222 SC/01 Added the -bylocaldomain option to eximstats.
2224 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
2226 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
2228 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
2234 PH/01 Another patch from the Sieve maintainer.
2236 PH/02 When an IPv6 address is converted to a string for single-key lookup
2237 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
2238 dots are used instead of colons so that keys in lsearch files need not
2239 contain colons. This was done some time before quoting was made available
2240 in lsearch files. However, iplsearch files do require colons in IPv6 keys
2241 (notated using the quote facility) so as to distinguish them from IPv4
2242 keys. This meant that lookups for IP addresses in host lists did not work
2243 for iplsearch lookups.
2245 This has been fixed by arranging for IPv6 addresses to be expressed with
2246 colons if the lookup type is iplsearch. This is not incompatible, because
2247 previously such lookups could never work.
2249 The situation is now rather anomalous, since one *can* have colons in
2250 ordinary lsearch keys. However, making the change in all cases is
2251 incompatible and would probably break a number of configurations.
2253 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
2256 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
2257 conversion specification without a maximum field width, thereby enabling
2258 a rogue spamd server to cause a buffer overflow. While nobody in their
2259 right mind would setup Exim to query an untrusted spamd server, an
2260 attacker that gains access to a server running spamd could potentially
2261 exploit this vulnerability to run arbitrary code as the Exim user.
2263 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
2264 $primary_hostname instead of what libspf2 thinks the hosts name is.
2266 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
2267 a directory entry by the name of the lookup key. Previously, if a
2268 symlink pointed to a non-existing file or a file in a directory that
2269 Exim lacked permissions to read, a lookup for a key matching that
2270 symlink would fail. Now it is enough that a matching directory entry
2271 exists, symlink or not. (Bugzilla 503.)
2273 PH/03 The body_linecount and body_zerocount variables are now exported in the
2276 PH/04 Added the $dnslist_matched variable.
2278 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
2279 This means they are set thereafter only if the connection becomes
2282 PH/06 Added the client_condition to authenticators so that some can be skipped
2283 by clients under certain conditions.
2285 PH/07 The error message for a badly-placed control=no_multiline_responses left
2286 "_responses" off the end of the name.
2288 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
2290 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
2291 (without spaces) instead of just copying the configuration text.
2293 PH/10 Added the /noupdate option to the ratelimit ACL condition.
2295 PH/11 Added $max_received_linelength.
2297 PH/12 Added +ignore_defer and +include_defer to host lists.
2299 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
2300 way in which PCRE > 7.0 is built.
2302 PH/14 Implemented queue_only_load_latch.
2304 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
2305 MAIL command. The effect was to mangle the value on 64-bit systems.
2307 PH/16 Another patch from the Sieve maintainer.
2309 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
2311 PH/18 If a system quota error occurred while trying to create the file for
2312 a maildir delivery, the message "Mailbox is full" was not appended to the
2313 bounce if the delivery eventually timed out. Change 4.67/27 below applied
2314 only to a quota excession during the actual writing of the file.
2316 PH/19 It seems that peer DN values may contain newlines (and other non-printing
2317 characters?) which causes problems in log lines. The DN values are now
2318 passed through string_printing() before being added to log lines.
2320 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
2321 and InterBase are left for another time.)
2323 PH/21 Added message_body_newlines option.
2325 PH/22 Guard against possible overflow in moan_check_errorcopy().
2327 PH/23 POSIX allows open() to be a macro; guard against that.
2329 PH/24 If the recipient of an error message contained an @ in the local part
2330 (suitably quoted, of course), incorrect values were put in $domain and
2331 $local_part during the evaluation of errors_copy.
2337 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
2338 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
2341 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
2342 issue a MAIL command.
2344 PH/02 In an ACL statement such as
2346 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
2348 if a client was not listed at all, or was listed with a value other than
2349 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
2350 the condition was not true (as it should be), so access was not denied.
2351 The bug was that the ! inversion was incorrectly passed on to the second
2352 item. This has been fixed.
2354 PH/03 Added additional dnslists conditions == and =& which are different from
2355 = and & when the dns lookup returns more than one IP address.
2357 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
2358 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
2360 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
2361 FSYNC, which compiles an option called disable_fsync that allows for
2362 bypassing fsync(). The documentation is heavily laced with warnings.
2364 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
2366 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
2367 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
2368 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
2369 including adding "make clean"; (3) Added -fPIC when compiling the test
2370 dynamically loaded module, to get rid of a warning.
2372 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
2373 message fails, move_frozen_messages = true and ignore_bounce_errors_after
2374 = 0s. The bug is otherwise harmless.
2376 PH/07 There was a bug in the dovecot authenticator such that the value of
2377 $auth1 could be overwritten, and so not correctly preserved, after a
2378 successful authentication. This usually meant that the value preserved by
2379 the server_setid option was incorrect.
2381 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
2383 PH/09 Installed PCRE release 7.0.
2385 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
2386 run for batched SMTP input. It is now run at the start of every message
2387 in the batch. While fixing this I discovered that the process information
2388 (output by running exiwhat) was not always getting set for -bs and -bS
2389 input. This is fixed, and it now also says "batched" for BSMTP.
2391 PH/11 Added control=no_pipelining.
2393 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
2394 patch, slightly modified), and move the expansion of helo_data till after
2395 the connection is made in the smtp transport (so it can use these
2398 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
2400 PH/14 Added log_selector = +pid.
2402 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
2404 PH/16 Add ${if forany and ${if forall.
2406 PH/17 Added dsn_from option to vary the From: line in DSNs.
2408 PH/18 Flush SMTP output before performing a callout, unless control =
2409 no_callout_flush is set.
2411 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
2412 was true (the default) a successful delivery failed to delete the retry
2413 item, thus causing premature timeout of the address. The bug is now
2416 PH/20 Added hosts_avoid_pipelining to the smtp transport.
2418 PH/21 Long custom messages for fakedefer and fakereject are now split up
2419 into multiline responses in the same way that messages for "deny" and
2420 other ACL rejections are.
2422 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
2423 with slight modification.
2425 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
2426 draft, changing the syntax and factoring some duplicate code".
2428 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
2429 for deliveries of the second and subsequent messages over the same SMTP
2432 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
2433 ${reduce, with only minor "tidies".
2435 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
2437 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
2438 expansion side effects.
2440 PH/27 When a message times out after an over-quota error from an Exim-imposed
2441 quota, the bounce message says "mailbox is full". This message was not
2442 being given when it was a system quota that was exceeded. It now should
2445 MH/03 Made $recipients available in local_scan(). local_scan() already has
2446 better access to the recipient list through recipients_list[], but
2447 $recipients can be useful in postmaster-provided expansion strings.
2449 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
2450 in the case of a MAIL command with additional options following the
2451 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
2452 were accidentally chopped off.
2454 PH/29 SMTP synchronization checks are implemented when a command is read -
2455 there is a check that no more input is waiting when there shouldn't be
2456 any. However, for some commands, a delay in an ACL can mean that it is
2457 some time before the response is written. In this time, more input might
2458 arrive, invalidly. So now there are extra checks after an ACL has run for
2459 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
2460 pipelining has not been advertised.
2462 PH/30 MH's patch to allow iscntrl() characters to be list separators.
2464 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
2465 returned in the SMTP response when smtp_return_error_details was false.
2466 This has been fixed.
2468 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
2469 instead of the C I/O that was originally supplied, because problems were
2470 reported on Solaris.
2472 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
2473 Exim which did not show up earlier: it was assuming that a call to
2474 SSL_CTX_set_info_callback() might give an error value. In fact, there is
2475 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
2476 was a macro that became an assignment, so it seemed to work. This has
2477 changed to a proper function call with a void return, hence the compile
2478 error. Exim's code has been fixed.
2480 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
2483 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
2485 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
2487 PH/37 If a message is not accepted after it has had an id assigned (e.g.
2488 because it turns out to be too big or there is a timeout) there is no
2489 "Completed" line in the log. When some messages of this type were
2490 selected by exigrep, they were listed as "not completed". Others were
2491 picked up by some special patterns. I have improved the selection
2492 criteria to be more general.
2494 PH/38 The host_find_failed option in the manualroute router can now be set
2495 to "ignore", to completely ignore a host whose IP address cannot be
2496 found. If all hosts are ignored, the behaviour is controlled by the new
2497 host_all_ignored option.
2499 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
2500 homing or because of multiple MX records with /mx) generated more than
2501 one IP address, and the following item turned out to be the local host,
2502 all the secondary addresses of the first item were incorrectly removed
2503 from the list, along with the local host and any following hosts (which
2504 is what is supposed to happen).
2506 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
2507 whoever called Exim into the -H file. In the case of the daemon it was
2508 behaving confusingly. When first started, it used values for whoever
2509 started the daemon, but after a SIGHUP it used the Exim user (because it
2510 calls itself on a restart). I have changed the code so that it now always
2513 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
2514 message are rejected with the same error (e.g. no authentication or bad
2515 sender address), and a DATA command is nevertheless sent (as can happen
2516 with PIPELINING or a stupid MUA), the error message that was given to the
2517 RCPT commands is included in the rejection of the DATA command. This is
2518 intended to be helpful for MUAs that show only the final error to their
2521 PH/42 Another patch from the Sieve maintainer.
2523 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
2524 Eximstats - Fixed some broken HTML links and added missing column headers
2526 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
2527 columns for Rejects, Temp Rejects, Ham, and Spam rows.
2529 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
2531 PH/43 Yet another patch from the Sieve maintainer.
2533 PH/44 I found a way to check for a TCP/IP connection going away before sending
2534 the response to the final '.' that terminates a message, but only in the
2535 case where the client has not sent further data following the '.'
2536 (unfortunately, this is allowed). However, in many cases there won't be
2537 any further data because there won't be any more messages to send. A call
2538 to select() can be used: if it shows that the input is "ready", there is
2539 either input waiting, or the socket has been closed. An attempt to read
2540 the next input character can distinguish the two cases. Previously, Exim
2541 would have sent an OK response which the client would never have see.
2542 This could lead to message repetition. This fix should cure that, at
2543 least in a lot of common cases.
2545 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
2546 advertised in response to EHLO.
2552 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
2553 fixed by 4.65/MH/01 (is this a record?) are fixed:
2555 (i) An empty string was always treated as zero by the numeric comparison
2556 operators. This behaviour has been restored.
2558 (ii) It is documented that the numeric comparison operators always treat
2559 their arguments as decimal numbers. This was broken in that numbers
2560 starting with 0 were being interpreted as octal.
2562 While fixing these problems I realized that there was another issue that
2563 hadn't been noticed. Values of message_size_limit (both the global option
2564 and the transport option) were treated as octal if they started with 0.
2565 The documentation was vague. These values are now always treated as
2566 decimal, and I will make that clear in the documentation.
2572 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
2573 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
2576 MH/01 Don't check that the operands of numeric comparison operators are
2577 integers when their expansion is in "skipping" mode (fixes bug
2578 introduced by 4.64-PH/07).
2580 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
2581 child addresses, Exim now panics and dies. Previously, because the count
2582 is held in a short int, deliveries were likely to be lost. As such a
2583 large number of recipients for a single message is ridiculous
2584 (performance will be very, very poor), I have chosen to impose a limit
2585 rather than extend the field.
2591 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
2592 leftover -K file (the existence of which was triggered by #402).
2593 While we were at it, introduced process PID as part of the -K
2594 filename. This should rule out race conditions when creating
2597 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
2598 processing considerably. Previous code took too long for large mails,
2599 triggering a timeout which in turn triggers #401.
2601 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
2602 in the DK code in transports.c. sendfile() is not really portable,
2603 hence the _LINUX specificness.
2605 TF/01 In the add_headers option to the mail command in an Exim filter,
2606 there was a bug that Exim would claim a syntax error in any
2607 header after the first one which had an odd number of characters
2610 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
2611 callout verification, Exim cached a "reject" for the entire domain. This
2612 is correct for most verifications, but it is not correct for a recipient
2613 verification with use_sender or use_postmaster set, because in that case
2614 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
2615 case of MAIL FROM:<> rejection from other early rejections (e.g.
2616 rejection of HELO). When verifying a recipient using a non-null MAIL
2617 address, the cache is ignored if it shows MAIL FROM:<> rejection.
2618 Whatever the result of the callout, the value of the domain cache is
2619 left unchanged (for any other kind of callout, getting as far as trying
2620 RCPT means that the domain itself is ok).
2622 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
2625 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
2626 manifest itself as EPIPE rather than ECONNECT. When tidying away a
2627 session, the daemon ignores ECONNECT errors and logs others; it now
2628 ignores EPIPE as well.
2630 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
2631 (quoted-printable decoding).
2633 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
2634 later the small subsequent patch to fix an introduced bug.
2636 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
2638 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
2640 PH/08 An error is now given if message_size_limit is specified negative.
2642 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
2643 to be given (somewhat) arbitrary names.
2645 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
2648 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
2649 miscellaneous code fixes
2651 PH/10 Added the log_reject_target ACL modifier to specify where to log
2654 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
2655 hostname. This is wrong, because it relates to the incoming message (and
2656 probably the interface on which it is arriving) and not to the outgoing
2657 callout (which could be using a different interface). This has been
2658 changed to use the value of the helo_data option from the smtp transport
2659 instead - this is what is used when a message is actually being sent. If
2660 there is no remote transport (possible with a router that sets up host
2661 addresses), $smtp_active_hostname is used.
2663 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
2664 tweaks were necessary in order to get it to work (see also 21 below):
2665 (a) The code assumed that strncpy() returns a negative number on buffer
2666 overflow, which isn't the case. Replaced with Exim's string_format()
2668 (b) There were several signed/unsigned issues. I just did the minimum
2669 hacking in of casts. There is scope for a larger refactoring.
2670 (c) The code used strcasecmp() which is not a standard C function.
2671 Replaced with Exim's strcmpic() function.
2672 (d) The code set only $1; it now sets $auth1 as well.
2673 (e) A simple test gave the error "authentication client didn't specify
2674 service in request". It would seem that Dovecot has changed its
2675 interface. Fortunately there's a specification; I followed it and
2676 changed what the client sends and it appears to be working now.
2678 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
2681 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
2682 address A is aliased to B and C, where B exists and C does not. Without
2683 -v the output is "A verified" because verification stops after a
2684 successful redirection if more than one address is generated. However,
2685 with -v the child addresses are also verified. Exim was outputting "A
2686 failed to verify" and then showing the successful verification for C,
2687 with its parentage. It now outputs "B failed to verify", showing B's
2688 parentage before showing the successful verification of C.
2690 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
2691 look up a TXT record in a specific list after matching in a combined
2694 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
2695 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
2696 they consult the DNS. I had assumed they would set it the way they
2697 wanted; and indeed my experiments on Linux seem to show that in some
2698 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
2699 To be on the safe side, however, I have now made the interface to
2700 host_find_byname() similar to host_find_bydns(), with an argument
2701 containing the DNS resolver options. The host_find_byname() function now
2702 sets these options at its start, just as host_find_bydns() does. The smtp
2703 transport options dns_qualify_single and dns_search_parents are passed to
2704 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
2705 of host_find_byname() use the default settings of RES_DEFNAMES
2706 (qualify_single) but not RES_DNSRCH (search_parents).
2708 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
2709 spool_read_header() do less string testing, by means of a preliminary
2710 switch on the second character of optional "-foo" lines. (This is
2711 overdue, caused by the large number of possibilities that now exist.
2712 Originally there were few.) While I was there, I also converted the
2713 str(n)cmp tests so they don't re-test the leading "-" and the first
2714 character, in the hope this might squeeze out yet more improvement.
2716 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
2717 flag allowing group syntax was set by the header_syntax check but not
2718 turned off, possible causing trouble later; (2) The flag was not being
2719 set at all for the header_verify test, causing "group"-style headers to
2720 be rejected. I have now set it in this case, and also caused header_
2721 verify to ignore an empty address taken from a group. While doing this, I
2722 came across some other cases where the code for allowing group syntax
2723 while scanning a header line wasn't quite right (mostly, not resetting
2724 the flag correctly in the right place). These bugs could have caused
2725 trouble for malformed header lines. I hope it is now all correct.
2727 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
2728 with the "reply" argument non-NULL. The code, however (which originally
2729 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
2730 but it didn't always do it. This confused somebody who was copying the
2731 code for some other use. I have removed all the tests.
2733 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
2734 feature that was used to support insecure browsers during the U.S. crypto
2735 embargo. It requires special client support, and Exim is probably the
2736 only MTA that supported it -- and would never use it because real RSA is
2737 always available. This code has been removed, because it had the bad
2738 effect of slowing Exim down by computing (never used) parameters for the
2739 RSA_EXPORT functionality.
2741 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
2742 authenticator to fail if there's a tab character in the incoming data
2743 (there should never be unless someone is messing about, as it's supposed
2744 to be base64-encoded). Also added, on Timo's advice, the "secured" option
2745 if the connection is using TLS or if the remote IP is the same as the
2746 local IP, and the "valid-client-cert option" if a client certificate has
2749 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
2750 authenticators. This can be used for authorization after authentication
2751 succeeds. (In the case of plaintext, it servers for both authentication
2754 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
2755 if any retry times were supplied.
2757 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
2758 connection, where there is no client IP address to check. In this
2759 situation, the verify now always succeeds.
2761 PH/25 Applied John Jetmore's -Mset patch.
2763 PH/26 Added -bem to be like -Mset, but loading a message from a file.
2765 PH/27 In a string expansion for a processed (not raw) header when multiple
2766 headers of the same name were present, leading whitespace was being
2767 removed from all of them, but trailing whitespace was being removed only
2768 from the last one. Now trailing whitespace is removed from each header
2769 before concatenation. Completely empty headers in a concatenation (as
2770 before) are ignored.
2772 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
2773 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
2775 PH/29 [Removed. This was a change that I later backed out, and forgot to
2776 correct the ChangeLog entry (that I had efficiently created) before
2777 committing the later change.]
2779 PH/30 Exim was sometimes attempting to deliver messages that had suffered
2780 address errors (4xx response to RCPT) over the same connection as other
2781 messages routed to the same hosts. Such deliveries are always "forced",
2782 so retry times are not inspected. This resulted in far too many retries
2783 for the affected addresses. The effect occurred only when there were more
2784 hosts than the hosts_max_try setting in the smtp transport when it had
2785 the 4xx errors. Those hosts that it had tried were not added to the list
2786 of hosts for which the message was waiting, so if all were tried, there
2787 was no problem. Two fixes have been applied:
2789 (i) If there are any address or message errors in an SMTP delivery, none
2790 of the hosts (tried or untried) are now added to the list of hosts
2791 for which the message is waiting, so the message should not be a
2792 candidate for sending over the same connection that was used for a
2793 successful delivery of some other message. This seems entirely
2794 reasonable: after all the message is NOT "waiting for some host".
2795 This is so "obvious" that I'm not sure why it wasn't done
2796 previously. Hope I haven't missed anything, but it can't do any
2797 harm, as the worst effect is to miss an optimization.
2799 (ii) If, despite (i), such a delivery is accidentally attempted, the
2800 routing retry time is respected, so at least it doesn't keep
2801 hammering the server.
2803 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
2804 in ${readsocket because some servers need this prod.
2806 PH/32 Added some extra debug output when updating a wait-xxx database.
2808 PH/33 The hint "could be header name not terminated by colon", which has been
2809 given for certain expansion errors for a long time, was not being given
2810 for the ${if def:h_colon_omitted{... case.
2812 PH/34 The spec says: "With one important exception, whenever a domain list is
2813 being scanned, $domain contains the subject domain." There was at least
2814 one case where this was not true.
2816 PH/35 The error "getsockname() failed: connection reset by peer" was being
2817 written to the panic log as well as the main log, but it isn't really
2818 panic-worthy as it just means the connection died rather early on. I have
2819 removed the panic log writing for the ECONNRESET error when getsockname()
2822 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
2823 runs only) independently of the message's sender address. This meant
2824 that, if the 4xx error was in fact related to the sender, a different
2825 message to the same recipient with a different sender could confuse
2826 things. In particular, this can happen when sending to a greylisting
2827 server, but other circumstances could also provoke similar problems.
2828 I have changed the default so that the retry time for these errors is now
2829 based a combination of the sender and recipient addresses. This change
2830 can be overridden by setting address_retry_include_sender=false in the
2833 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
2834 remote server are returned as part of bounce messages. This was not
2835 happening for LMTP over a pipe (the lmtp transport), but now it is the
2836 same for both kinds of LMTP.
2838 PH/38 Despite being documented as not happening, Exim was rewriting addresses
2839 in header lines that were in fact CNAMEs. This is no longer the case.
2841 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
2842 and queue runs started by the daemon processed all messages. This has
2843 been fixed so that -R and -S can now usefully be given with -q<time>.
2845 PH/40 Import PCRE release 6.7 (fixes some bugs).
2847 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
2849 PH/42 Give an error if -q is specified more than once.
2851 PH/43 Renamed the variables $interface_address and $interface_port as
2852 $received_ip_address and $received_port, to make it clear that these
2853 values apply to message reception, and not to the outgoing interface when
2854 a message is delivered. (The old names remain recognized, of course.)
2856 PH/44 There was no timeout on the connect() call when using a Unix domain
2857 socket in the ${readsocket expansion. There now is.
2859 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
2860 be meaningful with "accept".
2862 SC/01 Eximstats V1.43
2863 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
2865 SC/02 Eximstats V1.44
2866 Use a glob alias rather than an array ref in the generated
2867 parser. This improves both readability and performance.
2869 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
2870 Collect SpamAssassin and rejection statistics.
2871 Don't display local sender or destination tables unless
2872 there is data to show.
2873 Added average volumes into the top table text output.
2875 SC/04 Eximstats V1.46
2876 Collect data on the number of addresses (recipients)
2877 as well as the number of messages.
2879 SC/05 Eximstats V1.47
2880 Added 'Message too big' to the list of mail rejection
2881 reasons (thanks to Marco Gaiarin).
2883 SC/06 Eximstats V1.48
2884 Mainlog lines which have GMT offsets and are too short to
2885 have a flag are now skipped.
2887 SC/07 Eximstats V1.49 (Alain Williams)
2888 Added the -emptyok flag.
2890 SC/08 Eximstats V1.50
2891 Fixes for obtaining the IP address from reject messages.
2893 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
2894 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
2895 whitespace changes from 4.64-PH/27
2897 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
2900 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
2901 are found, allow negative numbers in numeric criteria)
2903 JJ/06 exipick.20061117.2, added new $message_body_missing variable
2905 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
2906 to match changes made in 4.64-PH/43
2908 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
2910 PH/47 Put in an explicit test for a DNS lookup of an address record where the
2911 "domain" is actually an IP address, and force a failure. This locks out
2912 those revolvers/nameservers that support "A-for-A" lookups, in
2913 contravention of the specifications.
2915 PH/48 When a host name was looked up from an IP address, and the subsequent
2916 forward lookup of the name timed out, the host name was left in
2917 $sender_host_name, contrary to the specification.
2919 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
2920 restricted to single-key lookups, Exim was not diagnosing an error if
2921 * or *@ was used with a query-style lookup.
2923 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
2925 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
2926 long ago, but noone interested enough thought of it. Let's just say that
2927 the "1.1" means that there are some new functions that weren't there at
2928 some point in the past.
2930 PH/51 Error processing for expansion failure of helo_data from an smtp
2931 transport during callout processing was broken.
2933 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
2934 tested/used via the -bh/-bhc/-bs options.
2936 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
2937 bug, fixed in subsequent PCRE releases).
2939 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2940 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2942 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2948 SC/01 Use a glob alias rather than an array ref in eximstats generated
2949 parser. This improves both readability and performance.
2951 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2952 Don't display local sender or destination tables in eximstats unless
2953 there is data to show.
2954 Added average volumes into the eximstats top table text output.
2956 SC/03 Collect data on the number of addresses (recipients) as well
2957 as the number of messages in eximstats.
2959 TF/01 Correct an error in the documentation for the redirect router. Exim
2960 does (usually) call initgroups() when daemonizing.
2962 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2963 with consistent privilege compared to when running as a daemon.
2965 TF/03 Note in the spec that $authenticated_id is not set for local
2966 submissions from trusted users.
2968 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2969 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2971 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2972 by adding some example configuration directives to the default
2973 configuration file. A little bit of work is required to uncomment the
2974 directives and define how usernames and passwords are checked, but
2975 there is now a framework to start from.
2977 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2978 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2979 without this. I don't know how relevant this is to other LDAP libraries.
2981 PH/02 Add the verb name to the "unknown ACL verb" error.
2983 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2985 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2987 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2988 directory not expanded when it should be if an expanded home directory
2989 was set for the address (which is overridden by the transport).
2991 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2994 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2995 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2996 because it is too late at that time, and has no effect.
2998 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2999 security issue with \' (bugzilla #107). I could not use the
3000 PQescapeStringConn() function, because it needs a PGconn value as one of
3003 PH/08 When testing addresses using -bt, indicate those final addresses that
3004 are duplicates that would not cause an additional delivery. At least one
3005 person was confused, thinking that -bt output corresponded to deliveries.
3006 (Suppressing duplicates isn't a good idea as you lose the information
3007 about possibly different redirections that led to the duplicates.)
3009 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
3010 systems where poll() doesn't work, in particular OS X.
3012 PH/10 Added more information to debugging output for retry time not reached.
3014 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
3015 operations in malware.c.
3017 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
3020 PH/13 If write_rejectlog was set false when logging was sent to syslog with
3021 syslog_duplication set false, log lines that would normally be written
3022 both the the main log and to the reject log were not written to syslog at
3025 PH/14 In the default configuration, change the use of "message" in ACL warn
3026 statements to "add_header".
3028 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
3029 not followed by a command (e.g. "seen endif").
3031 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
3032 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
3035 PH/17 Added extra conditions to the default value of delay_warning_condition
3039 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
3040 { match{$h_precedence:}{(?i)bulk|list|junk} } \
3041 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
3044 The Auto-Submitted: and various List- headers are standardised, whereas I
3045 don't think Precedence: ever was.
3047 PH/18 Refactored debugging code in route_finduser() to show more information,
3048 in particular, the error code if getpwnam() issues one.
3050 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
3051 This is apparently needed in addition to the PH/07 change above to avoid
3052 any possible encoding problems.
3054 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
3055 but not after initializing Perl.
3057 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
3058 output them only if debugging. By default they are written stderr,
3059 apparently, which is not desirable.
3061 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
3064 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
3067 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
3069 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
3070 authenticated or an ident call has been made. Suppress the default
3071 values for $authenticated_id and $authenticated_sender (but permit -oMai
3072 and -oMas) when testing with -bh.
3074 PH/24 Re-jigged the order of the tests in the default configuration so that the
3075 tests for valid domains and recipients precede the DNS black list and CSA
3076 tests, on the grounds that those ones are more expensive.
3078 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
3079 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
3080 This bug exists in every version of Exim that I still have, right back to
3083 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
3084 However, an attempt to turn on case-sensitivity in a regex key by
3085 including (?-i) didn't work because the subject string was already
3086 lowercased, and the effects were non-intuitive. It turns out that a
3087 one-line patch can be used to allow (?-i) to work as expected.
3093 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
3094 other effects) broke the use of negated acl sub-conditions.
3096 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
3099 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
3100 "Deny" causes Exim to reject the incoming connection with a 554 error.
3101 Unfortunately, if there is a major crisis, such as a disk failure,
3102 tcp-wrappers gives "deny", whereas what one would like would be some
3103 kind of temporary error. A kludge has been added to help with this.
3104 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
3105 554 error is used if errno is still zero or contains ENOENT (which occurs
3106 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
3109 PH/03 Add -lutil to the default FreeBSD LIBS setting.
3111 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
3112 errors. Otherwise a message that provokes a temporary error (when other
3113 messages do not) can cause a whole host to time out.
3115 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
3116 addresses were routed directly to files or pipes from a redirect router.
3117 File deliveries just didn't batch; pipe deliveries might have suffered
3120 PH/06 A failure to get a lock for a hints database would erroneously always say
3121 "Failed to get write lock", even when it was really a read lock.
3123 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
3124 of 0600. This has been changed to use the value of the lockfile_mode
3125 option (which defaults to 0600).
3127 PH/08 Applied small patch from the Sieve maintainer.
3129 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
3130 folder from quota calculations, a direct delivery into this folder messed
3131 up the contents of the maildirsize file. This was because the regex was
3132 used only to exclude .Trash (or whatever) when the size of the mailbox
3133 was calculated. There was no check that a delivery was happening into an
3134 excluded directory. This bug has been fixed by ignoring all quota
3135 processing for deliveries into excluded directories.
3137 PH/10 Added the maildirfolder_create_regex option to appendfile.
3143 PH/01 The code for finding all the local interface addresses on a FreeBSD
3144 system running IPv6 was broken. This may well have applied to all BSD
3145 systems, as well as to others that have similar system calls. The broken
3146 code found IPv4 interfaces correctly, but gave incorrect values for the
3147 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
3148 that it would not match correctly against @[] and not recognize the IPv6
3151 PH/02 The ipliteral router was not recognizing addresses of the form user@
3152 [ipv6:....] because it didn't know about the "ipv6:" prefix.
3154 PH/03 Added disable_ipv6.
3156 PH/04 Changed $reply_address to use the raw form of the headers instead of the
3157 decoded form, because it is most often used to construct To: headers
3158 lines in autoreplies, and the decoded form may well be syntactically
3159 invalid. However, $reply_address has leading white space removed, and all
3160 newlines turned into spaces so that the autoreply transport does not
3163 PH/05 If group was specified without a user on a router, and no group or user
3164 was specified on a transport, the group from the router was ignored.
3166 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
3167 for visible compile-time settings that can be used to change these
3168 numbers, for those that want even more. Backwards compatibility with old
3169 spool files has been maintained. However, going back to a previous Exim
3170 release will lost any variables that are in spool files.
3172 PH/07 Two small changes when running in the test harness: increase delay when
3173 passing a TCP/IP connection to a new process, in case the original
3174 process has to generate a bounce, and remove special handling of
3175 127.0.0.2 (sic), which is no longer necessary.
3177 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
3178 be the same on different OS.
3180 PH/09 Moved a debug statement in filter processing to avoid a race problem when
3183 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
3184 whether --show-vars was specified or not
3186 JJ/02 exipick: Added support for new ACL variable spool format introduced
3189 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
3190 syntactically invalid From: or Reply-to: line, and a filter used this to
3191 generate an autoreply, and therefore failed to obtain an address for the
3192 autoreply, Exim could try to deliver to a non-existent relative file
3193 name, causing unrelated and misleading errors. What now happens is that
3194 it logs this as a hard delivery error, but does not attempt to create a
3197 PH/11 The exinext utility has a -C option for testing purposes, but although
3198 the given file was scanned by exinext itself; it wasn't being passed on
3199 when Exim was called.
3201 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
3202 an end-of-file indication when reading a command response.
3204 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
3205 compiled. In many other places in Exim, IPv6 addresses are always
3206 recognized, so I have changed this. It also means that IPv4 domain
3207 literals of the form [IPV4:n.n.n.n] are now always recognized.
3209 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
3210 used if the router is not running as root, for example, when verifying at
3211 ACL time, or when using -bh. The debugging output from this situation was
3212 non-existent - all you got was a failure to exec. I have made two
3215 (a) Failures to set uid/gid, the current directory, or a process leader
3216 in a subprocess such as that created by queryprogram now generate
3217 suitable debugging output when -d is set.
3219 (b) The queryprogram router detects when it is not running as root,
3220 outputs suitable debugging information if -d is set, and then runs
3221 the subprocess without attempting to change uid/gid.
3223 PH/15 Minor change to Makefile for building test_host (undocumented testing
3226 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
3227 additional section of a DNS packet that returns MX or SRV records.
3228 Instead, it always explicitly searches for A/AAAA records. This avoids
3229 major problems that occur when a DNS server includes only records of one
3230 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
3231 fixed another bug: if SRV records were looked up and the corresponding
3232 address records were *not* found in the additional section, the port
3233 values from the SRV records were lost.
3235 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
3236 using the correct key (the original address) when searching the retry
3237 rules in order to find which one to use for generating the retry hint.
3239 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
3240 adding the default one. Similarly, if it contains a Reply-To: header, the
3241 errors_reply_to option, if set, is not used.
3243 PH/19 When calculating a retry time, Exim used to measure the "time since
3244 failure" by looking at the "first failed" field in the retry record. Now
3245 it does not use this if it is later than than the arrival time of the
3246 message. Instead it uses the arrival time. This makes for better
3247 behaviour in cases where some deliveries succeed, thus re-setting the
3248 "first failed" field. An example is a quota failure for a huge message
3249 when small messages continue to be delivered. Without this change, the
3250 "time since failure" will always be short, possible causing more frequent
3251 delivery attempts for the huge message than are intended.
3252 [Note: This change was subsequently modified - see PH/04 for 4.62.]
3254 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
3255 $1, $2, $3) because the numerical variables can be reset during some
3256 expansion items (e.g. "match"), thereby losing the authentication data.
3258 PH/21 Make -bV show the size of off_t variables so that the test suite can
3259 decide whether to run tests for quotas > 2G.
3261 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
3262 mailbox_size, and mailbox_filecount in the appendfile transport. If a
3263 filecount value is greater than 2G or if a quota value is greater than 2G
3264 on a system where the size of off_t is not greater than 4, a panic error
3267 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
3268 never match. The debug and -bh output now contains an explicit error
3269 message indicating a malformed IPv4 address or mask.
3271 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
3272 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
3273 PH/23 above applies.
3275 PH/25 Do not write to syslog when running in the test harness. The only
3276 occasion when this arises is a failure to open the main or panic logs
3277 (for which there is an explicit test).
3279 PH/26 Added the /no_tell option to "control=freeze".
3281 PH/27 If a host name lookup failed very early in a connection, for example, if
3282 the IP address matched host_lookup and the reverse lookup yielded a name
3283 that did not have a forward lookup, an error message of the form "no IP
3284 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
3285 could be logged. Now it outputs the IP address instead of "NULL".
3287 PH/28 An enabling patch from MH: add new function child_open_exim2() which
3288 allows the sender and the authenticated sender to be set when
3289 submitting a message from within Exim. Since child_open_exim() is
3290 documented for local_scan(), the new function should be too.
3292 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
3293 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
3294 results in an empty string is now treated as unset.
3296 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
3298 PH/31 Added sender_verify_fail logging option.
3300 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
3301 needed by RFC 822 but not by RFC 2822 was commented out. I have now
3302 tidied the source and removed it altogether.
3304 PH/33 When a queue run was abandoned because the load average was too high, a
3305 log line was always written; now it is written only if the queue_run log
3306 selector is set. In addition, the log line for abandonment now contains
3307 information about the queue run such as the pid. This is always present
3308 in "start" and "stop" lines but was omitted from the "abandon" line.
3310 PH/34 Omit spaces between a header name and the colon in the error message that
3311 is given when verify = headers_syntax fails (if there are lots of them,
3312 the message gets confusing).
3314 PH/35 Change the default for dns_check_names_pattern to allow slashes within
3315 names, as there are now some PTR records that contain slashes. This check
3316 is only to protect against broken name servers that fall over on strange
3317 characters, so the fact that it applies to all lookups doesn't matter.
3319 PH/36 Now that the new test suite is complete, we can remove some of the
3320 special code in Exim that was needed for the old test suite. For example,
3321 sorting DNS records because real resolvers return them in an arbitrary
3322 order. The new test suite's fake resolver always returns records in the
3325 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
3326 bounces) except when queue_only is set, to avoid logging races between
3327 the different processes.
3329 PH/38 Panic-die if .include specifies a non-absolute path.
3331 PH/39 A tweak to the "H" retry rule from its user.
3333 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
3334 a label. They prevented compilation on older perls.
3336 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
3337 a warning to be raised on newish perls.
3339 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
3340 on queue. Changes to match documented behaviour of showing count of
3341 messages matching specified criteria.
3343 PH/40 Changed the default ident timeout from 30s to 5s.
3345 PH/41 Added support for the use of login_cap features, on those BSD systems
3346 that have them, for controlling the resources used by pipe deliveries.
3348 PH/42 The content-scanning code uses fopen() to create files in which to put
3349 message data. Previously it was not paying any attention to the mode of
3350 the files. Exim runs with umask(0) because the rest of the code creates
3351 files with open(), and sets the required mode explicitly. Thus, these
3352 files were ending up world-writeable. This was not a big issue, because,
3353 being within the spool directory, they were not world-accessible. I have
3354 created a function called modefopen, which takes an additional mode
3355 argument. It sets umask(777), creates the file, chmods it to the required
3356 mode, then resets the umask. All the relevant calls to fopen() in the
3357 content scanning code have been changed to use this function.
3359 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
3360 to 24 hours. This avoids potential overflow problems when processing G
3361 and H retry rules. I suspect nobody ever tinkers with this value.
3363 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
3365 PH/45 When the plaintext authenticator is running as a client, the server's
3366 challenges are checked to ensure they are valid base64 strings. By
3367 default, the authentication attempt is cancelled if an invalid string is
3368 received. Setting client_ignore_invalid_base64 true ignores these errors.
3369 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
3370 they are received. Thus, the responses can be made to depend on the
3371 challenges. If an invalid string is ignored, an empty string is placed in
3374 PH/46 Messages that are created by the autoreply transport now contains a
3375 References: header, in accordance with RFCs 2822 and 3834.
3377 PH/47 Added authenticated_sender_force to the smtp transport.
3379 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
3381 PH/49 Installed latest patch from the Sieve maintainer.
3383 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
3384 was also set, the appendfile transport was unnecessarily scanning a
3385 directory of message files (e.g. for maildir delivery) to find the count
3386 of files (along with the size), even though it did not need this
3387 information. It now does the scan only if it needs to find either the
3388 size of the count of files.
3390 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
3392 PH/52 Two bugs concerned with error handling when the smtp transport is
3395 (i) Exim was not creating retry information for temporary errors given
3396 for individual recipients after the DATA command when the smtp transport
3397 was used in LMTP mode. This meant that they could be retried too
3398 frequently, and not timed out correctly.
3400 (ii) Exim was setting the flag that allows error details to be returned
3401 for LMTP errors on RCPT commands, but not for LMTP errors for individual
3402 recipients that were returned after the DATA command.
3404 PH/53 This is related to PH/52, but is more general: for any failing address,
3405 when detailed error information was permitted to be returned to the
3406 sender, but the error was temporary, then after the final timeout, only
3407 "retry timeout exceeded" was returned. Now it returns the full error as
3408 well as "retry timeout exceeded".
3410 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
3411 do this, and (what is worse) MTAs that accept it.
3413 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
3414 will now be deprecated.
3416 PH/56 New os.c-cygwin from the Cygwin maintainer.
3418 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
3419 formats (previously only available in exim formats via -bpr, -bpru,
3420 and -bpra. Now also available in native and exiqgrep formats)
3422 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
3423 with very large, slow to parse queues
3425 JJ/08 exipick: added ! as generic prefix to negate any criteria format
3427 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
3429 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
3430 responses to authentication challenges, though it was showing the
3431 challenges; (ii) I've removed the CR characters from the debug output for
3434 PH/58 Allow for the insertion of a newline as well as a space when a string
3435 is turned into more than one encoded-word during RFC 2047 encoding. The
3436 Sieve code now uses this.
3438 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
3439 data_4xx, lost_connection, tls_required.
3441 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
3442 message was being sent as an SMTP response.
3444 PH/61 Add -l and -k options to exicyclog.
3446 PH/62 When verifying, if an address was redirected to one new address, so that
3447 verification continued, and the new address failed or deferred after
3448 having set something in $address_data, the value of $address_data was not
3449 passed back to the ACL. This was different to the case when no
3450 redirection occurred. The value is now passed back in both cases.
3452 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
3453 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
3454 use login_cap.h, so on its own it isn't the distinguishing feature. The
3455 new name refers directly to the setclassresources() function.
3457 PH/65 Added configuration files for NetBSD3.
3459 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
3461 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
3462 is preferred over IPv4.
3464 PH/68 The bounce_return_message and bounce_return_body options were not being
3465 honoured for bounces generated during the reception of non-SMTP messages.
3466 In particular, this applied to messages rejected by the ACL. This bug has
3467 been fixed. However, if bounce_return_message is true and bounce_return_
3468 body is false, the headers that are returned for a non-SMTP message
3469 include only those that have been read before the error was detected.
3470 (In the case of an ACL rejection, they have all been read.)
3472 PH/69 The HTML version of the specification is now built in a directory called
3473 spec_html instead of spec.html, because the latter looks like a path with
3474 a MIME-type, and this confuses some software.
3476 PH/70 Catch two compiler warnings in sieve.c.
3478 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
3479 function verify_get_ident() calls ip_connect() to connect a socket, but
3480 if the "connect()" function timed out, ip_connect() used to close the
3481 socket. However, verify_get_ident() also closes the socket later, and in
3482 between Exim writes to the log, which may get opened at this point. When
3483 the socket was closed in ip_connect(), the log could get the same file
3484 descriptor number as the socket. This naturally causes chaos. The fix is
3485 not to close the socket in ip_connect(); the socket should be closed by
3486 the function that creates it. There was only one place in the code where
3487 this was missing, in the iplookup router, which I don't think anybody now
3488 uses, but I've fixed it anyway.
3490 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
3491 well as to direct DNS lookups. Otherwise the handling of names in host
3492 lists is inconsistent and therefore confusing.
3498 PH/01 Two changes to the default runtime configuration:
3500 (1) Move the checks for relay_from_hosts and authenticated clients from
3501 after to before the (commented out) DNS black list checks.
3503 (2) Add control=submission to the relay_from_hosts and authenticated
3504 clients checks, on the grounds that messages accepted by these
3505 statements are most likely to be submissions.
3507 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
3509 (1) Generate an error if the third argument for the ${prvs expansion is
3512 (2) Treat a missing third argument of ${prvscheck as if it were an empty
3515 (3) Reset the variables that are obtained from the first argument of
3516 ${prvscheck and used in the second argument before leaving the code,
3517 because their memory is reclaimed, so using them afterwards may do
3520 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
3521 one (it's much easier than Tom thought :-).
3523 (5) Because of (4), we can now allow for the use of $prvscheck_result
3524 inside the third argument.
3526 PH/03 For some reason, the default setting of PATH when running a command from
3527 a pipe transport was just "/usr/bin". I have changed it to
3530 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
3531 anything to be listed in the output from -bV.
3533 PH/05 When a filter generated an autoreply, the entire To: header line was
3534 quoted in the delivery log line, like this:
3536 => >A.N.Other <ano@some.domain> <original@ddress> ...
3538 This has been changed so that it extracts the operative address. There
3539 may be more than one such address. If so, they are comma-separated, like
3542 => >ano@some.domain,ona@other.domain <original@ddress> ...
3544 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
3545 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
3546 not being looked up in the rDNS to get a host name, Exim was showing the
3547 IP address twice in Received: lines, even though the IP addresses were
3548 identical. For example:
3550 Received: from [1.2.3.4] (helo=[1.2.3.4])
3552 However, if the real host name was known, it was omitting the HELO data
3553 if it matched the actual IP address. This has been tidied up so that it
3554 doesn't show the same IP address twice.
3556 PH/07 When both +timestamp and +memory debugging was on, the value given by
3557 $tod_xxx expansions could be wrong, because the tod_stamp() function was
3558 called by the debug printing, thereby overwriting the timestamp buffer.
3559 Debugging no longer uses the tod_stamp() function when +timestamp is set.
3561 PH/08 When the original message was included in an autoreply transport, it
3562 always said "this is a copy of the message, including all the headers",
3563 even if body_only or headers_only was set. It now gives an appropriate
3566 PH/09 Applied a patch from the Sieve maintainer which:
3568 o fixes some comments
3569 o adds the (disabled) notify extension core
3570 o adds some debug output for the result of if/elsif tests
3571 o points to the current vacation draft in the documentation
3572 and documents the missing references header update
3576 o fixes a bug in processing the envelope test (when testing
3577 multiple envelope elements, the last element determined the
3580 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
3581 Electronic Mail") by including:
3583 Auto-submitted: auto-generated
3585 in the messages that it generates (bounce messages and others, such as
3586 warnings). In the case of bounce messages for non-SMTP messages, there was
3587 also a typo: it was using "Auto_submitted" (underscore instead of
3588 hyphen). Since every message generated by Exim is necessarily in response
3589 to another message, thes have all been changed to:
3591 Auto-Submitted: auto-replied
3593 in accordance with these statements in the RFC:
3595 The auto-replied keyword:
3597 - SHOULD be used on messages sent in direct response to another
3598 message by an automatic process,
3600 - MUST NOT be used on manually-generated messages,
3602 - MAY be used on Delivery Status Notifications (DSNs) and Message
3603 Disposition Notifications (MDNs),
3605 - MUST NOT be used on messages generated by automatic or periodic
3606 processes, except for messages which are automatic responses to
3609 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
3610 to the default Received: header definition.
3612 PH/12 Added log selector acl_warn_skipped (default on).
3614 PH/13 After a successful wildlsearch lookup, discard the values of numeric
3615 variables because (a) they are in the wrong storage pool and (b) even if
3616 they were copied, it wouldn't work properly because of the caching.
3618 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
3619 checking when decoding. Apparently there are clients that generate
3620 overlong encoded strings. Why am I not surprised?
3622 PH/15 If the first argument of "${if match_address" was not empty, but did not
3623 contain an "@" character, Exim crashed. Now it writes a panic log message
3624 and treats the condition as false.
3626 PH/16 In autoreply, treat an empty string for "once" the same as unset.
3628 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
3629 extension "envelope-auth". The code is finished and in agreement with
3630 other implementations, but there is no documentation so far and in fact,
3631 nobody wrote the draft yet. This extension is currently #undef'ed, thus
3632 not changing the active code.
3634 Print executed "if" and "elsif" statements when debugging is used. This
3635 helps a great deal to understand what a filter does.
3637 Document more things not specified clearly in RFC3028. I had all this
3638 sorted out, when out of a sudden new issues came to my mind. Oops."
3640 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
3643 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
3644 canonical form (as documented). However, after a host name lookup from
3645 the IP address, check_host() was doing a simple string comparison with
3646 addresses acquired from the DNS when checking that the found name did
3647 have the original IP as one of its addresses. Since any found IPv6
3648 addresses are likely to be in abbreviated form, the comparison could
3649 fail. Luckily, there already exists a function for doing the comparison
3650 by converting both addresses to binary, so now that is used instead of
3651 the text comparison.
3653 PH/20 There was another similar case to PH/19, when a complete host name was
3654 given in a host list; looking up its IP address could give an abbreviated
3655 form, whereas the current host's name might or might not be abbreviated.
3656 The same fix has been applied.
3662 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
3663 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
3666 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
3667 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
3669 PH/03 Typo: missing ".o" in src/pcre/Makefile.
3671 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
3672 header line, restrict the check to what is listed in RFCs 2369 and 2929.
3673 Also, for "Auto-Submitted", treat anything other than "no" as
3674 non-personal, in accordance with RFC 3834. (Previously it treated
3675 anything starting "auto-" as non-personal.)
3677 TF/01 The control=submission/name=... option had a problem with syntax
3678 errors if the name included a slash character. The /name= option
3679 now slurps the rest of the string, so it can include any characters
3680 but it must come last in the list of options (after /sender_retain
3683 PH/05 Some modifications to the interface to the fake nameserver for the new
3691 TK/01 Added the "success_on_redirect" address verification option. See
3692 NewStuff for rationale and an example.
3694 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
3696 PH/02 Patch to exigrep to allow it to work on syslog lines.
3698 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
3699 fread() to skip over the body file's header line, because in Cygwin the
3700 header line is locked and is inaccessible.
3702 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
3703 co-exist for some time) to make it clear that it is the Exim ID that is
3704 referenced, not the Message-ID: header line.
3706 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
3707 string_format() function, because snprintf() does not exist on all
3710 PH/06 The use of forbid_filter_existstest now also locks out the use of the
3711 ${stat: expansion item.
3713 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
3714 protocol synchronization error", to keep the pedants happy.
3716 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
3717 well as for IRIX systems, when gcc is being used. See the host.c source
3720 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
3722 PH/10 Named domain lists were not working if used in a queue_smtp_domains
3725 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
3726 transport and to the smtp transport in LMTP mode.
3728 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
3730 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
3731 run a filter in a subprocess. This could lead to confusion in subsequent
3732 lookups in the parent process. There should also be a search_tidyup() at
3733 the end of the subprocess.
3735 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
3736 only if the host matched helo_try_verify_hosts, which caused the
3737 verification to occur when the EHLO/HELO command was issued. The ACL just
3738 tested the remembered result. Now, if a previous verification attempt has
3739 not happened, "verify = helo" does it there and then.
3741 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
3743 TK/03 Fix log output including CR from clamd.
3745 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
3746 exist provoked a memory error which could cause a segfault.
3748 PH/15 Installed PCRE 6.2
3750 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
3752 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
3753 of the problem. Specifically, suggested +O2 rather than +O1 for the
3756 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
3758 PH/20 If a delivery was routed to a non-standard port by means of an SRV
3759 record, the port was not correctly logged when the outgoing_port log
3760 selector was set (it logged the transort's default port).
3762 PH/21 Added support for host-specific ports to manualroute, queryprogram,
3763 fallback_hosts, and "hosts" in the smtp transport.
3765 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
3766 host errors such as "Connection refused".
3768 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
3769 authentication with radiusclient 0.4.9:
3771 - Error returned from rc_read_config was caught wrongly
3772 - Username/password not passed on to radius server due to wrong length.
3774 The presumption is that some radiusclient API changes for 4.51/PH/17
3775 were not taken care of correctly. The code is still untested by me (my
3776 Linux distribution still has 0.3.2 of radiusclient), but it was
3777 contributed by a Radius user.
3779 PH/24 When doing a callout, the value of $domain wasn't set correctly when
3780 expanding the "port" option of the smtp transport.
3782 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
3783 while reading a MIME header. Thanks to Tom Hughes for a patch.
3785 PH/24 Include config.h inside local_scan.h so that configuration settings are
3788 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
3789 that in an ACL for RCPT (for example), you can examine exactly what was
3792 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
3793 commands, but it was not correctly comparing the address with the actual
3794 client host address. Thus, it would show the EHLO address in Received:
3795 header lines when this was not necessary.
3797 PH/27 Added the % operator to ${eval:}.
3799 PH/28 Exim tries to create and chdir to its spool directory when it starts;
3800 it should be ignoring failures (because with -C, for example, it has lost
3801 privilege). It wasn't ignoring creation failures other than "already
3804 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
3807 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
3808 because an input line was too long, either on its own, or by virtue of
3809 too many continuations, the temporary file was not being removed, and the
3810 return code was incorrect.
3812 PH/31 Missing "BOOL" in function definition in filtertest.c.
3814 PH/32 Applied Sieve patches from the maintainer.
3816 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
3818 PH/33 Added "verify = not_blind".
3820 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
3821 Local/Makefile (with some defaults set). These are used in built scripts
3822 such as exicyclog, but they have never been used in the exim_install
3823 script (though there are many overriding facilities there). I have
3824 arranged that the exim_install script now takes note of these two
3827 PH/35 Installed configuration files for Dragonfly.
3829 PH/36 When a locally submitted message by a trusted user did not contain a
3830 From: header, and the sender address was obtained from -f or from an SMTP
3831 MAIL command, and the trusted user did not use -F to supply a sender
3832 name, $originator_name was incorrectly used when constructing a From:
3833 header. Furthermore, $originator_name was used for submission mode
3834 messages from external hosts without From: headers in a similar way,
3835 which is clearly wrong.
3837 PH/37 Added control=suppress_local_fixups.
3839 PH/38 When log_selector = +received_sender was set, and the addition of the
3840 sender made the log line's construction buffer exactly full, or one byte
3841 less than full, an overflow happened when the terminating "\n" was
3844 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
3845 when the result of a list match is failure because a DNS lookup failed.
3847 PH/40 RM_COMMAND is now used in the building process.
3849 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
3850 the "build-* directories that it finds.
3852 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
3853 address in a domain literal was a prefix of an interface address.
3855 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
3856 when verifying a sender address, unless rewrite_headers is false.
3858 PH/44 Wrote a long comment about why errors_to addresses are verified as
3859 recipients, not senders.
3861 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
3862 the ratelimit ACL was added.
3864 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
3866 PH/46 Added extra information about PostgreSQL errors to the error string.
3868 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
3869 avoiding the need to install special zones in a real server. This is
3870 backwards compatible; if it can't find the fake resolver, it drops back.
3871 Thus, both old and new test suites can be run.
3873 TF/02 Added util/ratelimit.pl
3875 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
3876 clock is set back in time.
3878 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
3879 Candler <B.Candler@pobox.com>.
3881 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
3882 OK for addresses that are the result of redirections.
3884 PH/48 A number of further additions for the benefit of the new test suite,
3885 including a fake gethostbyname() that interfaces to the fake DNS resolver
3888 TF/06 The fix for widen_domains has also been applied to qualify_single and
3889 search_parents which are the other dnslookup options that can cause
3892 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
3895 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
3897 TF/07 Exim produced the error message "an SRV record indicated no SMTP
3898 service" if it encountered an MX record with an empty target hostname.
3899 The message is now "an MX or SRV record indicated no SMTP service".
3901 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
3902 if the DNS of the sending site is misconfigured. This is quite a
3903 common situation. This change restores the behaviour of treating a
3904 helo verification defer as a failure.
3906 PH/51 If self=fail was set on a router, the bounce message did not include the
3907 actual error message.
3913 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
3915 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
3916 command itself ended in error, the underlying message about the transport
3917 filter timeout was being overwritten with the pipe command error. Now the
3918 underlying error message should be appended to the second error message.
3920 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
3922 PH/02 Reduce the amount of output that "make" produces by default. Full output
3923 can still be requested.
3925 PH/03 The warning log line about a condition test deferring for a "warn" verb
3926 was being output only once per connection, rather than after each
3927 occurrence (because it was using the same function as for successful
3928 "warn" verbs). This seems wrong, so I have changed it.
3930 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
3931 it should not have, which might have caused a crash in the right
3932 circumstances, but probably never did.
3934 PH/04 Installed a modified version of Tony Finch's patch to make submission
3935 mode fix the return path as well as the Sender: header line, and to
3936 add a /name= option so that you can make the user's friendly name appear
3939 TF/03 Added the control = fakedefer ACL modifier.
3941 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3942 Mark Lowes for thorough testing.
3944 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3946 TK/03 Merged latest SRS patch from Miles Wilton.
3948 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3949 with the definition in sysexits.h (which is #included earlier).
3950 Fortunately, Exim does not actually use EX_OK. The code used to try to
3951 preserve the sysexits.h value, by assuming that macro definitions were
3952 scanned for macro replacements. I have been disabused of this notion,
3953 so now the code just undefines EX_OK before #including unistd.h.
3955 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3956 in the smtp transport. When a block could not be written in a single
3957 write() function, the timeout was being re-applied to each part-write.
3958 This seems wrong - if the receiver was accepting one byte at a time it
3959 would take for ever. The timeout is now adjusted when this happens. It
3960 doesn't have to be particularly precise.
3962 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3963 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3965 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3966 without a domain if the check to <postmaster@domain> fails.
3968 SC/01 Eximstats: added -xls and the ability to specify output files
3969 (patch written by Frank Heydlauf).
3971 SC/02 Eximstats: use FileHandles for outputting results.
3973 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3975 SC/04 Eximstats: fixed display of large numbers with -nvr option
3977 SC/05 Eximstats: fixed merging of reports with empty tables.
3979 SC/06 Eximstats: added the -include_original_destination flag
3981 SC/07 Eximstats: removed tabs and trailing whitespace.
3983 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3985 TK/06 MBOX spool code: Add real "From " MBOX separator line
3986 so the .eml file is really in mbox format (even though
3987 most programs do not really care). Patch from Alex Miller.
3989 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3990 The latter is generated from $received_to and is only set if the
3991 message has one envelope recipient. SA can use these headers,
3992 obviously out-of-the-box. Patch from Alex Miller.
3994 PH/08 The ${def test on a variable was returning false if the variable's
3995 value was "0", contrary to what the specification has always said!
3996 The result should be true unless the variable is empty.
3998 PH/09 The syntax error of a character other than { following "${if
3999 def:variable_name" (after optional whitespace) was not being diagnosed.
4000 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
4001 accidental colon was present, for example, could give incorrect results.
4003 PH/10 Tidied the code in a number of places where the st_size field of a stat()
4004 result is used (not including appendfile, where other changes are about
4007 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
4008 This involved changing a lot of size variables from int to off_t. It
4009 should work with maildirs and everything.
4011 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
4012 spamd dying while we are connected to it.
4014 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
4017 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
4018 function for BDB 4.3.
4020 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
4022 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
4023 the sources for PCRE have been re-arranged and more files are now
4026 PH/15 The code I had for printing potentially long long variables in PH/11
4027 above was not the best (it lost precision). The length of off_t variables
4028 is now inspected at build time, and an appropriate printing format (%ld
4029 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
4030 to be "long long int" or "long int". This is needed for the internal
4031 formatting function string_vformat().
4033 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
4034 the configuration file to be ":syslog", then the script "guesses" where
4035 the logs files are, rather than using the compiled in default. In our
4036 case the guess is not the same as the compiled default, so the script
4037 suddenly stopped working when I started to use syslog. The patch checks
4038 to see if log_file_path is "". If so, it attempts to read it from exim
4039 with no configuration file to get the compiled in version, before it
4040 falls back to the previous guessing code."
4042 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
4043 implementing BATV in an Exim configuration. See NewStuff for the gory
4046 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
4047 Makefile that are specific to HP-UX.
4049 PH/18 If the "use_postmaster" option was set for a recipient callout together
4050 with the "random" option, the postmaster address was used as the MAIL
4051 FROM address for the random test, but not for the subsequent recipient
4052 test. It is now used for both.
4054 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
4055 patch removes a few documentation additions to RFC 3028, because the
4056 latest draft now contains them. It adds the new en;ascii-case comparator
4057 and a new error check for 8bit text in MIME parts. Comparator and
4058 require names are now matched exactly. I enabled the subaddress
4059 extension, but it is not well tested yet (read: it works for me)."
4061 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
4062 rework some of the code of TK/09 above to avoid the hardwired use of
4063 "%lld" and "long long". Replaced the call to snprintf() with a call to
4066 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
4067 records point to non-existent hosts", "retry timeout exceeded", and
4068 "retry time not reached for any host after a long failure period".
4070 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
4071 experimental DomainKeys support:
4073 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
4074 (2) On an error such as an illegally used "control", the wrong name for
4075 the control was given.
4077 These problems did NOT occur unless DomainKeys support was compiled.
4079 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
4081 PH/24 Added ${if match_ip condition.
4083 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
4084 databases so that it will be absolutely obvious if a crash occurs in the
4085 DB library. This is a regular occurrence (often caused by mis-matched
4088 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
4089 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
4090 picked up on a user's system that detects such things. There doesn't seem
4091 to be a gcc warning option for this - only an attribute that has to be
4092 put on the function's prototype. It seems that in Fedora Core 4 they have
4093 set this on a number of new functions. No doubt there will be more in due
4096 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
4097 specify a transport. However, if an address that was verified by such a
4098 router was the subject of a callout, Exim crashed because it tried to
4099 read the rcpt_include_affixes from the non-existent transport. Now it
4100 just assumes that the setting of that option is false. This bug was
4101 introduced by 4.51/PH/31.
4103 PH/28 Changed -d+all to exclude +memory, because that information is very
4104 rarely of interest, but it makes the output a lot bigger. People tend to
4105 do -d+all out of habit.
4107 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
4108 code in os-type was giving problems when libc.so lives in lib64, like on
4111 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
4112 aren't the modern standard, and it seems that some systems' include files
4113 don't always have them. Exim was already checking for some of the newer
4114 ones like T_AAAA, and defining it itself. I've added checks for all the
4115 record types that Exim uses.
4117 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
4118 not automatically generating a new one, as it is supposed to. This
4119 prevented TLS from working. If the file did exist, but contained invalid
4120 data, a new version was generated, as expected. It was only the case of a
4121 non-existent file that was broken.
4123 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
4124 with a change in libdomainkeys > 0.64.
4126 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
4127 from DNS. If the selector record carries the flag, it now has
4128 precedence over the domain-wide flag.
4130 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
4132 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
4133 the use of a port name that isn't defined in /etc/services) occurred, the
4134 message was deferred as in a normal delivery, and thus remained on the
4135 spool, instead of being failed because of the mua_wrapper setting. This
4136 is now fixed, and I tidied up some of the mua_wrapper messages at the
4139 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
4140 the messages in a hash of arrays rather than using individual hashes.
4141 This is a bit cleaner and results in dramatic memory savings, albeit
4142 at a slight CPU cost.
4144 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
4145 as requested by Marc Sherman.
4147 SC/10 Eximstats: added histograms for user specified patterns as requested
4150 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
4152 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
4153 fopen() in the content-scanning modules that did not already have it.
4159 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
4160 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
4162 TK/02 Fix ACL "control" statement not being available in MIME ACL.
4164 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
4166 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
4167 to test Sieve filters that use "vacation".
4169 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
4170 that changes the way the GnuTLS parameters are stored in the cache file.
4171 The new format can be generated externally. For backward compatibility,
4172 if the data in the cache doesn't make sense, Exim assumes it has read an
4173 old-format file, and it generates new data and writes a new file. This
4174 means that you can't go back to an older release without removing the
4177 PH/03 A redirect router that has both "unseen" and "one_time" set does not
4178 work if there are any delivery delays because "one_time" forces the
4179 parent to be marked "delivered", so its unseen clone is never tried
4180 again. For this reason, Exim now forbids the simultaneous setting of
4183 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
4184 redirected to themselves ("homonym" addresses). Read the long ChangeLog
4185 entry if you want to know the details. The fix, however, neglected to
4186 consider the case when local delivery batching is involved. The test for
4187 "previously delivered" was not happening when checking to see if an
4188 address could be batched with a previous (undelivered) one; under
4189 certain circumstances this could lead to multiple deliveries to the same
4192 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
4193 in its include files, and this causes problems building Exim.
4195 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
4196 header_syntax) but Exim was just ignoring anything given after a slash.
4197 In particular, this caused confusion with an attempt to use "verify =
4198 reverse_host_lookup/defer_ok". An error is now given when options are
4199 supplied for verify items that do not have them. (Maybe reverse_host_
4200 lookup should have a defer_ok option, but that's a different point.)
4202 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
4203 defined by RFC 821) to 2048, because there were problems with some AUTH
4204 commands, and RFC 1869 says the size should be increased for extended
4205 SMTP commands that take arguments.
4207 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
4210 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
4211 "unknown" error; now it says that the functionality isn't in the binary.
4213 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
4214 an address' error message when a string expansion fails (syntax or
4215 whatever). Otherwise the password may appear in the log. Following change
4216 PH/42 below, there is no longer a chance of it appearing in a bounce
4219 PH/11 Installed exipick version 20050225.0 from John Jetmore.
4221 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
4222 of its addresses was ever tried. (Bugzilla bug #2.)
4224 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
4225 the result incorrectly in the debug output. (It correctly added a newline
4226 to what was transported.)
4228 TF/01 Added $received_time.
4230 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
4231 commented out examples of how to interface to a virus scanner and to
4232 SpamAssassin. Also added commented examples of av_scanner and
4233 spamd_address settings.
4235 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
4236 and controls are allowed in which ACLs. There were a couple of minor
4237 errors. Some of the entries in the conditions table (which is a table of
4238 where they are NOT allowed) were getting very unwieldy; rewrote them as a
4239 negation of where the condition IS allowed.
4241 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
4243 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
4244 header file does not have a version number, so I've had to invent a new
4245 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
4246 API. The code is untested by me (my Linux distribution still has 0.3.2 of
4247 radiusclient), but it was contributed by a Radius user.
4249 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
4250 files or directories, for OpenSSL.
4252 PH/19 When an Exim process that is running as root has to create an Exim log
4253 file, it does so in a subprocess that runs as exim:exim so as to get the
4254 ownership right at creation (otherwise, other Exim processes might see
4255 the file with the wrong ownership). There was no test for failure of this
4256 fork() call, which would lead to the process getting stuck as it waited
4257 for a non-existent subprocess. Forks do occasionally fail when resources
4258 run out. I reviewed all the other calls to fork(); they all seem to check
4261 PH/20 When checking for unexpected SMTP input at connect time (before writing
4262 the banner), Exim was not dealing correctly with a non-positive return
4263 from the read() function. If the client had disconnected by this time,
4264 the result was a log entry for a synchronization error with an empty
4265 string after "input=" when read() returned zero. If read() returned -1
4266 (an event I could not check), uninitialized data bytes were printed.
4267 There were reports of junk text (parts of files, etc) appearing after
4270 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
4272 PH/22 Added support for macro redefinition, and (re)definition in between
4273 driver and ACL definitions.
4275 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
4276 forgetting to use the resulting value; it was using the unexpanded value.
4278 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
4279 hadn't been configured. The fix is from Juergen Kreileder, who
4280 understands it better than I do:
4282 "Here's what I see happening with three configured cyrus_sasl
4283 authenticators configured (plain, login, cram-md5):
4285 On startup auth_cyrus_sasl_init() gets called for each of these.
4286 This means three calls to sasl_listmech() without a specified mech_list.
4287 => SASL tests which mechs of all available mechs actually work
4288 => three warnings about OTP not working
4289 => the returned list contains: plain, login, cram-md5, digest-md5, ...
4291 With the patch, sasl_listmech() also gets called three times. But now
4292 SASL's mech_list option is set to the server_mech specified in the the
4293 authenticator. Or in other words, the answer from sasl_listmech()
4294 gets limited to just the mech you're testing for (which is different
4296 => the return list contains just 'plain' or 'login', 'cram-md5' or
4297 nothing depending on the value of ob->server_mech.
4299 I've just tested the patch: Authentication still works fine,
4300 unavailable mechs specified in the exim configuration are still
4301 caught, and the auth.log warnings about OTP are gone."
4303 PH/25 When debugging is enabled, the contents of the command line are added
4304 to the debugging output, even when log_selector=+arguments is not
4307 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
4308 answer is "GNU", and only if the return is "GNU/something" is the answer
4311 PH/27 $acl_verify_message is now set immediately after the failure of a
4312 verification in an ACL, and so is available in subsequent modifiers. In
4313 particular, the message can be preserved by coding like this:
4315 warn !verify = sender
4316 set acl_m0 = $acl_verify_message
4318 Previously, $acl_verify_message was set only while expanding "message"
4319 and "log_message" when a very denied access.
4321 PH/28 Modified OS/os.c-Linux with
4323 -#ifndef OS_LOAD_AVERAGE
4324 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
4326 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
4327 nomenclature these days.)
4329 PH/29 Installed patch from the Sieve maintainer that adds the options
4330 sieve_useraddress and sieve_subaddress to the redirect router.
4332 PH/30 In these circumstances:
4333 . Two addresses routed to the same list of hosts;
4334 . First host does not offer TLS;
4335 . First host accepts first address;
4336 . First host gives temporary error to second address;
4337 . Second host offers TLS and a TLS session is established;
4338 . Second host accepts second address.
4339 Exim incorrectly logged both deliveries with the TLS parameters (cipher
4340 and peerdn, if requested) that were in fact used only for the second
4343 PH/31 When doing a callout as part of verifying an address, Exim was not paying
4344 attention to any local part prefix or suffix that was matched by the
4345 router that accepted the address. It now behaves in the same way as it
4346 does for delivery: the affixes are removed from the local part unless
4347 rcpt_include_affixes is set on the transport.
4349 PH/32 Add the sender address, as F=<...>, to the log line when logging a
4350 timeout during the DATA phase of an incoming message.
4352 PH/33 Sieve envelope tests were broken for match types other than :is. I have
4353 applied a patch sanctioned by the Sieve maintainer.
4355 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
4356 the uid or gid is negative. A case of a negative gid caused this to be
4357 noticed. The fix allows for either to be negative.
4359 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
4360 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
4363 PH/36 The change PH/12 above was broken. Fixed it.
4365 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
4366 the grounds that routing the same address twice would always produce the
4367 same answer. This might have been true once, but it is certainly no
4368 longer true now. Routing a child address may depend on the previous
4369 routing that produced that child. Some complicated redirection strategies
4370 went wrong when messages had multiple recipients, and made Exim's
4371 behaviour dependent on the order in which the addresses were given.
4373 I have moved the duplicate checking until after the routing is complete.
4374 Exim scans the addresses that are assigned to local and remote
4375 transports, and removes any duplicates. This means that more work will be
4376 done, as duplicates will always all be routed, but duplicates are
4377 presumably rare, so I don't expect this is of any significance.
4379 For deliveries to pipes, files, and autoreplies, the duplicate checking
4380 still happens during the routing process, since they are not going to be
4383 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
4384 It corrects a timeout issue with spamd. This is Ian's comment: "The
4385 background is that sometimes spamd either never reads data from a
4386 connection it has accepted, or it never writes response data. The exiscan
4387 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
4388 blindly assumes that writes won't block so it may never time out."
4390 PH/39 Allow G after quota size as well as K and M.
4392 PH/40 The value set for $authenticated_id in an authenticator may not contain
4393 binary zeroes or newlines because the value is written to log lines and
4394 to spool files. There was no check on this. Now the value is run through
4395 the string_printing() function so that such characters are converted to
4396 printable escape sequences.
4398 PH/41 $message_linecount is a new variable that contains the total number of
4399 lines in the message. Compare $body_linecount, which is the count for the
4402 PH/42 Exim no longer gives details of delivery errors for specific addresses in
4403 bounce and delay warning messages, except in certain special cases, which
4406 (a) An SMTP error message from a remote host;
4407 (b) A message specified in a :fail: redirection;
4408 (c) A message specified in a "fail" command in a system filter;
4409 (d) A message specified in a FAIL return from the queryprogram router;
4410 (e) A message specified by the cannot_route_message router option.
4412 In these cases only, Exim does include the error details in bounce and
4413 warning messages. There are also a few cases where bland messages such
4414 as "unrouteable address" or "local delivery error" are given.
4416 PH/43 $value is now also set for the "else" part of a ${run expansion.
4418 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
4419 being worked on, but at least Exim now implements the latest version to
4422 PH/45 In a pipe transport, although a timeout while waiting for the pipe
4423 process to complete was treated as a delivery failure, a timeout while
4424 writing the message to the pipe was logged, but erroneously treated as a
4425 successful delivery. Such timeouts include transport filter timeouts. For
4426 consistency with the overall process timeout, these timeouts are now
4427 treated as errors, giving rise to delivery failures by default. However,
4428 there is now a new Boolean option for the pipe transport called
4429 timeout_defer, which, if set TRUE, converts the failures into defers for
4430 both kinds of timeout. A transport filter timeout is now identified in
4433 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
4434 systems where "make" and "gmake" are different, calling "gmake" at top
4435 level broke things. I've arranged for the value of $(MAKE) to be passed
4436 from the Makefile to this script so that it can call the same version of
4440 A note about Exim versions 4.44 and 4.50
4441 ----------------------------------------
4443 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
4444 changes of various kinds. As a consequence, a big documentation update was
4445 needed. This delayed the release for rather longer than seemed good, especially
4446 in the light of a couple of (minor) security issues. Therefore, the changes
4447 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
4448 release. So 4.44 and 4.50 are in effect two different branches that both start
4451 I have left the 4.50 change log unchanged; it contains all the changes since
4452 4.43. The change log for 4.44 is below; many of its items are identical to
4453 those for 4.50. This seems to be the most sensible way to preserve the
4454 historical information.
4460 1. Minor wording change to the doc/README.SIEVE file.
4462 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
4463 computation of the current number of files was incorrect.
4465 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4466 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4469 4. Give more explanation in the error message when the command for a transport
4470 filter fails to execute.
4472 5. There are several places where Exim runs a non-Exim command in a
4473 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4474 was being done only for the command run by the queryprogram router. It is
4475 now done for all such subprocesses. The other cases are: ${run, transport
4476 filters, and the commands run by the lmtp and pipe transports.
4478 6. Added CONFIGURE_GROUP build-time option.
4480 7. Some older OS have a limit of 256 on the maximum number of file
4481 descriptors. Exim was using setrlimit() to set 1000 as a large value
4482 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4483 systems. I've change it so that if it can't get 1000, it tries for 256.
4485 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4486 was an oversight, and furthermore, ever since the addition of extra
4487 controls (e.g. 4.43/32), the checks on when to allow different forms of
4488 "control" were broken. There should now be diagnostics for all cases when a
4489 control that does not make sense is encountered.
4491 9. Added the /retain_sender option to "control=submission".
4493 10. $recipients is now available in the predata ACL (oversight).
4495 11. Tidy the search cache before the fork to do a delivery from a message
4496 received from the command line. Otherwise the child will trigger a lookup
4497 failure and thereby defer the delivery if it tries to use (for example) a
4498 cached ldap connection that the parent has called unbind on.
4500 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4501 of $address_data from the recipient verification was clobbered by the
4502 sender verification.
4504 13. The value of address_data from a sender verification is now available in
4505 $sender_address_data in subsequent conditions in the ACL statement.
4507 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
4509 15. Added a new option "connect=<time>" to callout options, to set a different
4512 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4513 was its contents. (It was OK if the option was not defined at all.)
4515 17. A "Completed" log line is now written for messages that are removed from
4516 the spool by the -Mrm option.
4518 18. New variables $sender_verify_failure and $recipient_verify_failure contain
4519 information about exactly what failed.
4521 19. Added -dd to debug only the daemon process.
4523 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
4524 handles timeouts, both on the server side and network timeouts. Renamed the
4525 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
4527 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
4528 It is now set to "smtps".
4530 22. $host_address is now set to the target address during the checking of
4531 ignore_target_hosts.
4533 23. When checking ignore_target_hosts for an ipliteral router, no host name was
4534 being passed; this would have caused $sender_host_name to have been used if
4535 matching the list had actually called for a host name (not very likely,
4536 since this list is usually IP addresses). A host name is now passed as
4539 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4540 code that specifies a non-restarting handler (typically sigaction() in
4541 modern systems) in an attempt to fix a rare and obscure crash bug.
4543 25. Narrowed the window for a race in the daemon that could cause it to ignore
4544 SIGCHLD signals. This is not a major problem, because they are used only to
4545 wake it up if nothing else does.
4547 26. A malformed maildirsize file could cause Exim to calculate negative values
4548 for the mailbox size or file count. Odd effects could occur as a result.
4549 The maildirsize information is now recalculated if the size or filecount
4552 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4553 support for a long time. Removed HAVE_SYS_VFS_H.
4555 28. Installed the latest version of exipick from John Jetmore.
4557 29. In an address list, if the pattern was not a regular expression, an empty
4558 subject address (from a bounce message) matched only if the pattern was an
4559 empty string. Non-empty patterns were not even tested. This was the wrong
4560 because it is perfectly reasonable to use an empty address as part of a
4561 database query. An empty address is now tested by patterns that are
4562 lookups. However, all the other forms of pattern expect the subject to
4563 contain a local part and a domain, and therefore, for them, an empty
4564 address still always fails if the pattern is not itself empty.
4566 30. Exim went into a mad DNS loop when attempting to do a callout where the
4567 host was specified on an smtp transport, and looking it up yielded more
4568 than one IP address.
4570 31. Re-factored the code for checking spool and log partition space into a
4571 function that finds that data and another that does the check. The former
4572 is then used to implement four new variables: $spool_space, $log_space,
4573 $spool_inodes, and $log_inodes.
4575 32. The RFC2047 encoding function was originally intended for short strings
4576 such as real names; it was not keeping to the 75-character limit for
4577 encoded words that the RFC imposes. It now respects the limit, and
4578 generates multiple encoded words if necessary. To be on the safe side, I
4579 have increased the buffer size for the ${rfc2047: expansion operator from
4582 33. It is now permitted to omit both strings after an "if" condition; if the
4583 condition is true, the result is "true". As before, when the second string
4584 is omitted, a false condition yields an empty string. This makes it less
4585 cumbersome to write custom ACL and router conditions.
4587 34. Failure to deliver a bounce message always caused it to be frozen, even if
4588 there was an errors_to setting on the router. The errors_to setting is now
4591 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4592 canonical form (fully expanded) before being placed in
4593 $sender_host_address.
4595 36. The table in the code that translates DNS record types into text (T_A to
4596 "A" for instance) was missing entries for NS and CNAME. It is just possible
4597 that this could have caused confusion if both these types were looked up
4598 for the same domain, because the text type is used as part of Exim's
4599 per-process caching. But the chance of anyone hitting this buglet seems
4602 37. The dnsdb lookup has been extended in a number of ways.
4604 (1) There is a new type, "zns", which walks up the domain tree until it
4605 finds some nameserver records. It should be used with care.
4607 (2) There is a new type, "mxh", which is like "mx" except that it returns
4608 just the host names, not the priorities.
4610 (3) It is now possible to give a list of domains (or IP addresses) to be
4611 looked up. The behaviour when one of the lookups defers can be
4612 controlled by a keyword.
4614 (4) It is now possible to specify the separator character for use when
4615 multiple records are returned.
4617 38. The dnslists ACL condition has been extended: it is now possible to supply
4618 a list of IP addresses and/or domains to be looked up in a particular DNS
4621 39. Added log_selector=+queue_time_overall.
4623 40. When running the queue in the test harness, wait just a tad after forking a
4624 delivery process, to get repeatability of debugging output.
4626 41. Include certificate and key file names in error message when GnuTLS fails
4627 to set them up, because the GnuTLS error message doesn't include the name
4628 of the failing file when there is a problem reading it.
4630 42. Allow both -bf and -bF in the same test run.
4632 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
4634 44. The "Exiscan patch" is now merged into the mainline Exim source.
4636 45. Sometimes the final signoff response after QUIT could fail to get
4637 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4638 < 0 before doing a fflush(). This bug looks as though it goes back to the
4639 introduction of TLS in release 3.20, but "sometimes" must have been rare
4640 because the tests only now provoked it.
4642 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
4643 (this can affect the format of dates).
4645 47. exim_tidydb, when checking for the continued existence of a message for
4646 which it has found a message-specific retry record, was not finding
4647 messages that were in split spool directories. Consequently, it was
4648 deleting retry records that should have stayed in existence.
4650 48. Steve fixed some bugs in eximstats.
4652 49. The SPA authentication driver was not abandoning authentication and moving
4653 on to the next authenticator when an expansion was forced to fail,
4654 contradicting the general specification for all authenticators. Instead it
4655 was generating a temporary error. It now behaves as specified.
4657 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
4658 (the order specifies the preference for clients). The order is now AES256,
4659 AES128, 3DES, ARCFOUR128.
4661 51. Small patch to Sieve code - explicitly set From: when generating an
4664 52. Exim crashed if a remote delivery caused a very long error message to be
4665 recorded - for instance if somebody sent an entire SpamAssassin report back
4666 as a large number of 550 error lines. This bug was coincidentally fixed by
4667 increasing the size of one of Exim's internal buffers (big_buffer) that
4668 happened as part of the Exiscan merge. However, to be on the safe side, I
4669 have made the code more robust (and fixed the comments that describe what
4672 53. Now that there can be additional text after "Completed" in log lines (if
4673 the queue_time_overall log selector is set), a one-byte patch to exigrep
4674 was needed to allow it to recognize "Completed" as not the last thing in
4677 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
4678 patch that reportedly fixes this has been added. I am not expert enough to
4679 create a test for it. This is what the patch creator wrote:
4681 "I found a little strange behaviour of ldap code when working with
4682 Windows 2003 AD Domain, where users was placed in more than one
4683 Organization Units. When I tried to give exim partial DN, the exit code
4684 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
4685 But simultaneously result of request was absolutely normal ldap result,
4686 so I produce this patch..."
4688 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
4689 REFERENCE, so I have modified the code to exclude the patch when that macro
4692 55. Some experimental protocols are using DNS PTR records for new purposes. The
4693 keys for these records are domain names, not reversed IP addresses. The
4694 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4695 leaves it alone. Component reversal etc. now happens only for IP addresses.
4698 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
4700 57. Double the size of the debug message buffer (to 2048) so that more of very
4701 long debug lines gets shown.
4703 58. The exicyclog utility now does better if the number of log files to keep
4704 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
4706 59. Two changes related to the smtp_active_hostname option:
4708 (1) $smtp_active_hostname is now available as a variable.
4709 (2) The default for smtp_banner uses $smtp_active_hostname instead
4710 of $primary_hostname.
4712 60. The host_aton() function is supposed to be passed a string that is known
4713 to be a valid IP address. However, in the case of IPv6 addresses, it was
4714 not checking this. This is a hostage to fortune. Exim now panics and dies
4715 if the condition is not met. A case was found where this could be provoked
4716 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4717 components; fortuitously, this particular loophole had already been fixed
4718 by change 4.50/55 above.
4720 If there are any other similar loopholes, the new check in host_aton()
4721 itself should stop them being exploited. The report I received stated that
4722 data on the command line could provoke the exploit when Exim was running as
4723 exim, but did not say which command line option was involved. All I could
4724 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4725 running as the user.
4728 61. There was a buffer overflow vulnerability in the SPA authentication code
4729 (which came originally from the Samba project). I have added a test to the
4730 spa_base64_to_bits() function which I hope fixes it.
4733 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
4734 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
4736 63. The daemon start-up calls getloadavg() while still root for those OS that
4737 need the first call to be done as root, but it missed one case: when
4738 deliver_queue_load_max is set with deliver_drop_privilege. This is
4739 necessary for the benefit of the queue runner, because there is no re-exec
4740 when deliver_drop_privilege is set.
4742 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
4743 This has been fixed.
4745 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4746 was in use, was not putting the data itself into the right store pool;
4747 consequently, it could be overwritten for a subsequent message in the same
4748 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4751 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
4753 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
4754 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
4755 the function were treating the return as a boolean value, which happened to
4756 work because 0=false and not-0=true, but is not correct code.
4758 68. The host_aton() function was not handling scoped IPv6 addresses (those
4759 with, for example, "%eth0" on the end) correctly.
4761 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
4762 negated items (that is, ~something) in unsigned ints. Some compilers
4763 apparently mutter when there is no cast.
4765 70. If an address verification called from an ACL failed, and did not produce a
4766 user-specific message (i.e. there was only a "system" message), nothing was
4767 put in $acl_verify_message. In this situation, it now puts the system
4770 71. Change 4.23/11 added synchronization checking at the start of an SMTP
4771 session; change 4.31/43 added the unwanted input to the log line - except
4772 that it did not do this in the start of session case. It now does.
4774 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
4775 This is wrong and can cause the other end to generate a synchronization
4776 error if it is another Exim or anything else that does the synchronization
4777 check. A QUIT command is no longer sent after a timeout.
4779 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
4780 during host lookups.
4782 74. The defer_ok option of callout verification was not working if it was used
4783 when verifying addresses in header lines, that is, for this case:
4785 verify = header_sender/callout=defer_ok
4787 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
4788 those file descriptors could be used for SMTP connections. If anything
4789 wrote to stderr (the example that came up was "warn" in embedded Perl), it
4790 could be sent to the SMTP client, causing chaos. The daemon now opens
4791 stdin, stdout, and stderr to /dev/null when it puts itself into the
4794 76. Arrange for output from Perl's "warn" command to be written to Exim's main
4795 log by default. The user can override this with suitable Perl magic.
4797 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
4798 the log message when discard triggers, was not working for the DATA ACL or
4799 for the non-SMTP ACL.
4801 78. Error message wording change in sieve.c.
4803 79. If smtp_accept_max_per_host was set, the number of connections could be
4804 restricted to fewer than expected, because the daemon was trying to set up
4805 a new connection before checking whether the processes handling previous
4806 connections had finished. The check for completed processes is now done
4807 earlier. On busy systems, this bug wouldn't be noticed because something
4808 else would have woken the daemon, and it would have reaped the completed
4811 80. If a message was submitted locally by a user whose login name contained one
4812 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
4813 It caused a spool format error. I have fixed the spool reading code. A
4814 related problem was that the "from" clause in the Received: line became
4815 illegal because of the space(s). It is now covered by ${quote_local_part.
4817 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
4820 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
4822 83. Patch from Sieve maintainer to fix unterminated string problem in
4823 "vacation" handling.
4825 84. Some minor changes to the Linux configuration files to help with other
4826 OS variants using glibc.
4828 85. One more patch for Sieve to update vacation handling to latest spec.
4831 ----------------------------------------------------
4832 See the note above about the 4.44 and 4.50 releases.
4833 ----------------------------------------------------
4839 1. Change 4.43/35 introduced a bug that caused file counts to be
4840 incorrectly computed when quota_filecount was set in an appendfile
4843 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
4844 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
4847 3. Give more explanation in the error message when the command for a transport
4848 filter fails to execute.
4850 4. There are several places where Exim runs a non-Exim command in a
4851 subprocess. The SIGUSR1 signal should be disabled for these processes. This
4852 was being done only for the command run by the queryprogram router. It is
4853 now done for all such subprocesses. The other cases are: ${run, transport
4854 filters, and the commands run by the lmtp and pipe transports.
4856 5. Some older OS have a limit of 256 on the maximum number of file
4857 descriptors. Exim was using setrlimit() to set 1000 as a large value
4858 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
4859 systems. I've change it so that if it can't get 1000, it tries for 256.
4861 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
4862 was an oversight, and furthermore, ever since the addition of extra
4863 controls (e.g. 4.43/32), the checks on when to allow different forms of
4864 "control" were broken. There should now be diagnostics for all cases when a
4865 control that does not make sense is encountered.
4867 7. $recipients is now available in the predata ACL (oversight).
4869 8. Tidy the search cache before the fork to do a delivery from a message
4870 received from the command line. Otherwise the child will trigger a lookup
4871 failure and thereby defer the delivery if it tries to use (for example) a
4872 cached ldap connection that the parent has called unbind on.
4874 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
4875 of $address_data from the recipient verification was clobbered by the
4876 sender verification.
4878 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
4879 was its contents. (It was OK if the option was not defined at all.)
4881 11. A "Completed" log line is now written for messages that are removed from
4882 the spool by the -Mrm option.
4884 12. $host_address is now set to the target address during the checking of
4885 ignore_target_hosts.
4887 13. When checking ignore_target_hosts for an ipliteral router, no host name was
4888 being passed; this would have caused $sender_host_name to have been used if
4889 matching the list had actually called for a host name (not very likely,
4890 since this list is usually IP addresses). A host name is now passed as
4893 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
4894 code that specifies a non-restarting handler (typically sigaction() in
4895 modern systems) in an attempt to fix a rare and obscure crash bug.
4897 15. Narrowed the window for a race in the daemon that could cause it to ignore
4898 SIGCHLD signals. This is not a major problem, because they are used only to
4899 wake it up if nothing else does.
4901 16. A malformed maildirsize file could cause Exim to calculate negative values
4902 for the mailbox size or file count. Odd effects could occur as a result.
4903 The maildirsize information is now recalculated if the size or filecount
4906 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
4907 support for a long time. Removed HAVE_SYS_VFS_H.
4909 18. Updated exipick to current release from John Jetmore.
4911 19. Allow an empty sender to be matched against a lookup in an address list.
4912 Previously the only cases considered were a regular expression, or an
4915 20. Exim went into a mad DNS lookup loop when doing a callout where the
4916 host was specified on the transport, if the DNS lookup yielded more than
4919 21. The RFC2047 encoding function was originally intended for short strings
4920 such as real names; it was not keeping to the 75-character limit for
4921 encoded words that the RFC imposes. It now respects the limit, and
4922 generates multiple encoded words if necessary. To be on the safe side, I
4923 have increased the buffer size for the ${rfc2047: expansion operator from
4926 22. Failure to deliver a bounce message always caused it to be frozen, even if
4927 there was an errors_to setting on the router. The errors_to setting is now
4930 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
4931 canonical form (fully expanded) before being placed in
4932 $sender_host_address.
4934 24. Updated eximstats to version 1.33
4936 25. Include certificate and key file names in error message when GnuTLS fails
4937 to set them up, because the GnuTLS error message doesn't include the name
4938 of the failing file when there is a problem reading it.
4940 26. Expand error message when OpenSSL has problems setting up cert/key files.
4943 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4944 (this can affect the format of dates).
4946 28. exim_tidydb, when checking for the continued existence of a message for
4947 which it has found a message-specific retry record, was not finding
4948 messages that were in split spool directories. Consequently, it was
4949 deleting retry records that should have stayed in existence.
4951 29. eximstats updated to version 1.35
4952 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4953 1.35 - bugfix such that pie charts by volume are generated correctly
4955 30. The SPA authentication driver was not abandoning authentication and moving
4956 on to the next authenticator when an expansion was forced to fail,
4957 contradicting the general specification for all authenticators. Instead it
4958 was generating a temporary error. It now behaves as specified.
4960 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4961 (the order specifies the preference for clients). The order is now AES256,
4962 AES128, 3DES, ARCFOUR128.
4964 31. Small patch to Sieve code - explicitly set From: when generating an
4967 32. Exim crashed if a remote delivery caused a very long error message to be
4968 recorded - for instance if somebody sent an entire SpamAssassin report back
4969 as a large number of 550 error lines. This bug was coincidentally fixed by
4970 increasing the size of one of Exim's internal buffers (big_buffer) that
4971 happened as part of the Exiscan merge. However, to be on the safe side, I
4972 have made the code more robust (and fixed the comments that describe what
4975 33. Some experimental protocols are using DNS PTR records for new purposes. The
4976 keys for these records are domain names, not reversed IP addresses. The
4977 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4978 leaves it alone. Component reversal etc. now happens only for IP addresses.
4981 34. The host_aton() function is supposed to be passed a string that is known
4982 to be a valid IP address. However, in the case of IPv6 addresses, it was
4983 not checking this. This is a hostage to fortune. Exim now panics and dies
4984 if the condition is not met. A case was found where this could be provoked
4985 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4986 components; fortuitously, this particular loophole had already been fixed
4987 by change 4.50/55 or 4.44/33 above.
4989 If there are any other similar loopholes, the new check in host_aton()
4990 itself should stop them being exploited. The report I received stated that
4991 data on the command line could provoke the exploit when Exim was running as
4992 exim, but did not say which command line option was involved. All I could
4993 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4994 running as the user.
4997 35. There was a buffer overflow vulnerability in the SPA authentication code
4998 (which came originally from the Samba project). I have added a test to the
4999 spa_base64_to_bits() function which I hope fixes it.
5002 36. The daemon start-up calls getloadavg() while still root for those OS that
5003 need the first call to be done as root, but it missed one case: when
5004 deliver_queue_load_max is set with deliver_drop_privilege. This is
5005 necessary for the benefit of the queue runner, because there is no re-exec
5006 when deliver_drop_privilege is set.
5008 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
5009 was in use, was not putting the data itself into the right store pool;
5010 consequently, it could be overwritten for a subsequent message in the same
5011 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
5014 38. Sometimes the final signoff response after QUIT could fail to get
5015 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
5016 < 0 before doing a fflush(). This bug looks as though it goes back to the
5017 introduction of TLS in release 3.20, but "sometimes" must have been rare
5018 because the tests only now provoked it.
5024 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
5025 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
5026 Now it returns an integer. A number of calls were still expecting a T/F
5027 return. Fortuitously, in all cases, the tests worked in OK situations,
5028 which is the norm. However, things would have gone wrong on any write
5029 failures on the smtp file descriptor. This function is used when sending
5030 messages over SMTP and also when doing verify callouts.
5032 2. When Exim is called to do synchronous delivery of a locally submitted
5033 message (the -odf or -odi options), it no longer closes stderr before doing
5036 3. Implemented the mua_wrapper option.
5038 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
5040 5. Implemented the functions header_remove(), header_testname(),
5041 header_add_at_position(), and receive_remove_recipient(), and exported them
5044 6. If an ACL "warn" statement specified the addition of headers, Exim already
5045 inserted X-ACL-Warn: at the start if there was no header name. However, it
5046 was not making this test for the second and subsequent header lines if
5047 there were newlines in the string. This meant that an invalid header could
5048 be inserted if Exim was badly configured.
5050 7. Allow an ACL "warn" statement to add header lines at the start or after all
5051 the Received: headers, as well as at the end.
5053 8. Added the rcpt_4xx retry error code.
5055 9. Added postmaster_mailfrom=xxx to callout verification option.
5057 10. Added mailfrom=xxxx to the callout verification option, for verify=
5060 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
5061 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
5063 12. Inserted some casts to stop certain compilers warning when using pointer
5064 differences as field lengths or precisions in printf-type calls (mostly
5065 affecting debugging statements).
5067 13. Added optional readline() support for -be (dynamically loaded).
5069 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
5070 same clock tick as a message's arrival, so that its received time was the
5071 same as the "first fail" time on the retry record, and that message
5072 remained on the queue past the ultimate address timeout, every queue runner
5073 would try a delivery (because it was past the ultimate address timeout) but
5074 after another failure, the ultimate address timeout, which should have then
5075 bounced the address, did not kick in. This was a "< instead of <=" error;
5076 in most cases the first failure would have been in the next clock tick
5077 after the received time, and all would be well.
5079 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
5080 being recognized when the domain list was tested by the match_domain
5081 condition in an expansion string.
5083 16. Added the ${str2b64: operator.
5085 17. Exim was always calling setrlimit() to set a large limit for the number of
5086 processes, without checking whether the existing limit was already
5087 adequate. (It did check for the limit on file descriptors.) Furthermore,
5088 errors from getrlimit() and setrlimit() were being ignored. Now they are
5089 logged to the main and panic logs, but Exim does carry on, to try to do its
5090 job under whatever limits there are.
5092 18. Imported PCRE 5.0.
5094 19. Trivial typo in log message " temporarily refused connection" (the leading
5097 20. If the log selector return_path_on_delivery was set and an address was
5098 redirected to /dev/null, the delivery process crashed because it assumed
5099 that a return path would always be set for a "successful" delivery. In this
5100 case, the whole delivery is bypassed as an optimization, and therefore no
5103 21. Internal re-arrangement: the function for sending a challenge and reading
5104 a response while authentication was assuming a zero-terminated challenge
5105 string. It's now changed to take a pointer and a length, to allow for
5106 binary data in such strings.
5108 22. Added the cyrus_sasl authenticator (code supplied by MBM).
5110 23. Exim was not respecting finduser_retries when seeking the login of the
5111 uid under which it was called; it was always trying 10 times. (The default
5112 setting of finduser_retries is zero.) Also, it was sleeping after the final
5113 failure, which is pointless.
5115 24. Implemented tls_on_connect_ports.
5117 25. Implemented acl_smtp_predata.
5119 26. If the domain in control=submission is set empty, Exim assumes that the
5120 authenticated id is a complete email address when it generates From: or
5121 Sender: header lines.
5123 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
5124 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
5125 chown and chgrp in /bin and hostname in /usr/bin.
5127 28. Exim was keeping the "process log" file open after each use, just as it
5128 does for the main log. This opens the possibility of it remaining open for
5129 long periods when the USR1 signal hits a daemon. Occasional processlog
5130 errors were reported, that could have been caused by this. Anyway, it seems
5131 much more sensible not to leave this file open at all, so that is what now
5134 29. The long-running daemon process does not normally write to the log once it
5135 has entered its main loop, and it closes the log before doing so. This is
5136 so that log files can straightforwardly be renamed and moved. However,
5137 there are a couple of unusual error situations where the daemon does write
5138 log entries, and I had neglected to close the log afterwards.
5140 30. The text of an SMTP error response that was received during a remote
5141 delivery was being truncated at 512 bytes. This is too short for some of
5142 the long messages that one sometimes sees. I've increased the limit to
5145 31. It is now possible to make retry rules that apply only when a message has a
5146 specific sender, in particular, an empty sender.
5148 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
5149 it possible to be selective about when SMTP synchronization is enforced.
5151 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
5153 32. Implemented hosts_connection_nolog.
5155 33. Added an ACL for QUIT.
5157 34. Setting "delay_warning=" to disable warnings was not working; it gave a
5160 35. Added mailbox_size and mailbox_filecount to appendfile.
5162 36. Added control = no_multiline_responses to ACLs.
5164 37. There was a bug in the logic of the code that waits for the clock to tick
5165 in the case where the clock went backwards by a substantial amount such
5166 that the microsecond fraction of "now" was more than the microsecond
5167 fraction of "then" (but the whole seconds number was less).
5169 38. Added support for the libradius Radius client library this is found on
5170 FreeBSD (previously only the radiusclient library was supported).
5176 1. When certain lookups returned multiple values in the form name=value, the
5177 quoting of the values was not always being done properly. Specifically:
5178 (a) If the value started with a double quote, but contained no whitespace,
5180 (b) If the value contained whitespace other than a space character (i.e.
5181 tabs or newlines or carriage returns) it was not quoted.
5182 This fix has been applied to the mysql and pgsql lookups by writing a
5183 separate quoting function and calling it from the lookup code. The fix
5184 should probably also be applied to nisplus, ibase and oracle lookups, but
5185 since I cannot test any of those, I have not disturbed their existing code.
5187 2. A hit in the callout cache for a specific address caused a log line with no
5188 reason for rejecting RCPT. Now it says "Previous (cached) callout
5189 verification failure".
5191 3. There was an off-by-one bug in the queryprogram router. An over-long
5192 return line was truncated at 256 instead of 255 characters, thereby
5193 overflowing its buffer with the terminating zero. As well as fixing this, I
5194 have increased the buffer size to 1024 (and made a note to document this).
5196 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
5197 when Exim is waiting for an SMTP response from a remote server, Exim
5198 restarts its select() call on the socket, thereby resetting its timeout.
5199 This is not a problem when such interrupts are rare. Somebody set up a cron
5200 job to run exiwhat every 2 minutes, which is less than the normal select()
5201 timeout (5 or 10 minutes). This meant that the select() timeout never
5202 kicked in because it was always reset. I have fixed this by comparing the
5203 time when an interrupt arrives with the time at the start of the first call
5204 to select(). If more time than the timeout has elapsed, the interrupt is
5205 treated as a timeout.
5207 5. Some internal re-factoring in preparation for the addition of Sieve
5208 extensions (by MH). In particular, the "personal" test is moved to a
5209 separate function, and given an option for scanning Cc: and Bcc: (which is
5210 not set for Exim filters).
5212 6. When Exim created an email address using the login of the caller as the
5213 local part (e.g. when creating a From: or Sender: header line), it was not
5214 quoting the local part when it contained special characters such as @.
5216 7. Installed new OpenBSD configuration files.
5218 8. Reworded some messages for syntax errors in "and" and "or" conditions to
5219 try to make them clearer.
5221 9. Callout options, other than the timeout value, were being ignored when
5222 verifying sender addresses in header lines. For example, when using
5224 verify = header_sender/callout=no_cache
5226 the cache was (incorrectly) being used.
5228 10. Added a missing instance of ${EXE} to the exim_install script; this affects
5229 only the Cygwin environment.
5231 11. When return_path_on_delivery was set as a log selector, if different remote
5232 addresses in the same message used different return paths and parallel
5233 remote delivery occurred, the wrong values would sometimes be logged.
5234 (Whenever a remote delivery process finished, the return path value from
5235 the most recently started remote delivery process was logged.)
5237 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
5238 lines when AUTH and/or TLS are in use. This is the "received protocol"
5239 field. Exim used to use "asmtp" for authenticated SMTP, without any
5240 indication (in the protocol name) for TLS use. Now it follows the RFC and
5241 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
5242 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
5243 names appear in log lines as well as in Received: header lines.
5245 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
5246 extensions, and comparison tests, and to fix some bugs.
5248 14. Changes to the "personal" filter test:
5250 (1) The test was buggy in that it was just doing the equivalent of
5251 "contains" tests on header lines. For example, if a user's address was
5252 anne@some.where, the "personal" test would incorrectly be true for
5254 To: susanne@some.where
5256 This test is now done by extracting each address from the header in turn,
5257 and checking the entire address. Other tests that are part of "personal"
5258 are now done using regular expressions (for example, to check local parts
5259 of addresses in From: header lines).
5261 (2) The list of non-personal local parts in From: addresses has been
5262 extended to include "listserv", "majordomo", "*-request", and "owner-*",
5263 taken from the Sieve specification recommendations.
5265 (3) If the message contains any header line starting with "List-" it is
5266 treated as non-personal.
5268 (4) The test for "circular" in the Subject: header line has been removed
5269 because it now seems ill-conceived.
5271 15. Minor typos in src/EDITME comments corrected.
5273 16. Installed latest exipick from John Jetmore.
5275 17. If headers_add on a router specified a text string that was too long for
5276 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
5277 of string_sprintf() is now avoided.
5279 18. $message_body_size was not set (it was always zero) when running the DATA
5280 ACL and the local_scan() function.
5282 19. For the "mail" command in an Exim filter, no default was being set for
5283 the once_repeat time, causing a random time value to be used if "once" was
5284 specified. (If the value happened to be <= 0, no repeat happened.) The
5285 default is now 0s, meaning "never repeat". The "vacation" command was OK
5286 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
5287 (I found it when inspecting the code).
5289 20. There is now an overall timeout for performing a callout verification. It
5290 defaults to 4 times the callout timeout, which applies to individual SMTP
5291 commands during the callout. The overall timeout applies when there is more
5292 than one host that can be tried. The timeout is checked before trying the
5293 next host. This prevents very long delays if there are a large number of
5294 hosts and all are timing out (e.g. when the network connections are timing
5295 out). The value of the overall timeout can be changed by specifying an
5296 additional sub-option for "callout", called "maxwait". For example:
5298 verify = sender/callout=5s,maxwait=20s
5300 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
5301 to the end before writing, but this should make it even safer).
5303 22. Exim was forgetting that it had advertised PIPELINING for the second and
5304 subsequent messages on an SMTP connection. It was also not resetting its
5305 memory on STARTTLS and an internal HELO.
5307 23. When Exim logs an SMTP synchronization error within a session, it now
5308 records whether PIPELINING has been advertised or not.
5310 24. Added 3 instances of "(long int)" casts to time_t variables that were being
5311 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
5312 rather than long int.
5314 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
5316 26. Added the never_mail option to autoreply.
5322 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
5323 crash if the getsockname() call failed; this can happen if a connection is
5324 closed very soon after it is established. The problem was simply in the
5325 order in which certain operations were done, causing Exim to try to write
5326 to the SMTP stream before it had set up the file descriptor. The bug has
5327 been fixed by making things happen in the correct order.
5333 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
5334 before the connection was closed, thus losing the rejection response.
5336 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
5337 some early Solaris releases, but causes trouble in current releases where
5338 socklen_t is defined.
5340 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
5343 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
5346 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
5347 that releasing the top of it at the end releases what was used for sub-
5348 expansions (unless the block got too big). However, discard this block if
5349 the first thing is a variable or header, so that we can use its block when
5350 it is dynamic (useful for very large $message_headers, for example).
5352 6. Lookups now cache *every* query, not just the most recent. A new, separate
5353 store pool is used for this. It can be recovered when all lookup caches are
5354 flushed. Lookups now release memory at the end of their result strings.
5355 This has involved some general refactoring of the lookup sources.
5357 7. Some code has been added to the store_xxx() functions to reduce the amount
5358 of flapping under certain conditions.
5360 8. log_incoming_interface used to affect only the <= reception log lines. Now
5361 it causes the local interface and port to be added to several more SMTP log
5362 lines, for example "SMTP connection from", and rejection lines.
5364 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
5366 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
5368 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
5369 could be overwritten at the end of the current message (or the start of a
5370 new message if it was set in a HELO ACL). The value is now preserved for
5371 the duration of the SMTP connection.
5373 12. If a transport had a headers_rewrite setting, and a matching header line
5374 contained an unqualified address, that address was qualified, even if it
5375 did not match any rewriting rules. The underlying bug was that the values
5376 of the flags that permit the existence of unqualified sender and recipient
5377 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
5378 non-local messages, and by -bnq for local messages) were not being
5379 preserved with the message after it was received.
5381 13. When Exim was logging an SMTP synchronization error, it could sometimes log
5382 "next input=" as part of the text comprising the host identity instead of
5383 the correct text. The code was using the same buffer for two different
5384 strings. However, depending on which order the printing function evaluated
5385 its arguments, the bug did not always show up. Under Linux, for example, my
5386 test suite worked just fine.
5388 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
5389 doesn't work with some older versions of Perl. It has been changed to "my",
5390 which in any case is probably the better facility to use.
5392 15. A really picky compiler found some instances of statements for creating
5393 error messages that either had too many or two few arguments for the format
5396 16. The size of the buffer for calls to the DNS resolver has been increased
5397 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
5398 for addresses that have a lot of PTR records. This alleviates a problem; it
5399 does not fully solve it.
5401 17. A dnsdb lookup for PTR records that receives more data than will fit in the
5402 buffer now truncates the list and logs the incident, which is the same
5403 action as happens when Exim is looking up a host name and its aliases.
5404 Previously in this situation something unpredictable would happen;
5405 sometimes it was "internal error: store_reset failed".
5407 18. If a server dropped the connection unexpectedly when an Exim client was
5408 using GnuTLS and trying to read a response, the client delivery process
5409 crashed while trying to generate an error log message.
5411 19. If a "warn" verb in an ACL added multiple headers to a message in a single
5412 string, for example:
5414 warn message = H1: something\nH2: something
5416 the text was added as a single header line from Exim's point of view
5417 though it ended up OK in the delivered message. However, searching for the
5418 second and subsequent header lines using $h_h2: did not work. This has been
5419 fixed. Similarly, if a system filter added multiple headers in this way,
5420 the routers could not see them.
5422 20. Expanded the error message when iplsearch is called with an invalid key to
5423 suggest using net-iplsearch in a host list.
5425 21. When running tests using -bh, any delays imposed by "delay" modifiers in
5426 ACLs are no longer actually imposed (and a message to that effect is
5429 22. If a "gecos" field in a passwd entry contained escaped characters, in
5430 particular, if it contained a \" sequence, Exim got it wrong when building
5431 a From: or a Sender: header from that name. A second bug also caused
5432 incorrect handling when an unquoted " was present following a character
5433 that needed quoting.
5435 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
5436 was not being matched caselessly.
5438 24. Arranged for all hyphens in the exim.8 source to be escaped with
5441 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
5442 properly. Recipient callout cache records were still being keyed to include
5443 the sender, even when use_sender was set false. This led to far more
5444 callouts that were necessary. The sender is no longer included in the key
5445 when use_sender is false.
5447 26. Added "control = submission" modifier to ACLs.
5449 27. Added the ${base62d: operator to decode base 62 numbers.
5451 28. dnsdb lookups can now access SRV records.
5453 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
5454 the configuration file.
5456 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
5457 (-v) mode. This makes the output for a verbose queue run more intelligible.
5459 31. Added a use_postmaster feature to recipient callouts.
5461 32. Added the $body_zerocount variable, containing the number of binary zero
5462 bytes in the message body.
5464 33. The time of last modification of the "new" subdirectory is now used as the
5465 "mailbox time last read" when there is a quota error for a maildir
5468 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
5470 35. Added +ignore_unknown as a special item in host lists.
5472 36. Code for decoding IPv6 addresses in host lists is now included, even if
5473 IPv6 support is not being compiled. This fixes a bug in which an IPv6
5474 address was recognized as an IP address, but was then not correctly decoded
5475 into binary, causing unexpected and incorrect effects when compared with
5482 1. Very minor rewording of debugging text in manualroute to say "list of
5483 hosts" instead of "hostlist".
5485 2. If verify=header_syntax was set, and a header line with an unqualified
5486 address (no domain) and a large number of spaces between the end of the
5487 name and the colon was received, the reception process suffered a buffer
5488 overflow, and (when I tested it) crashed. This was caused by some obsolete
5489 code that should have been removed. The fix is to remove it!
5491 3. When running in the test harness, delay a bit after writing a bounce
5492 message to get a bit more predictability in the log output.
5494 4. Added a call to search_tidyup() just before forking a reception process. In
5495 theory, someone could use a lookup in the expansion of smtp_accept_max_
5496 per_host which, without the tidyup, could leave open a database connection.
5498 5. Added the variables $recipient_data and $sender_data which get set from a
5499 lookup success in an ACL "recipients" or "senders" condition, or a router
5500 "senders" option, similar to $domain_data and $local_part_data.
5502 6. Moved the writing of debug_print from before to after the "senders" test
5505 7. Change 4.31/66 (moving the time when the Received: is generated) caused
5506 problems for message scanning, either using a data ACL, or using
5507 local_scan() because the Received: header was not generated till after they
5508 were called (in order to set the time as the time of reception completion).
5509 I have revised the way this works. The header is now generated after the
5510 body is received, but before the ACL or local_scan() are called. After they
5511 are run, the timestamp in the header is updated.
5517 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
5518 before starting a queue runner without re-exec. This happened only when
5519 deliver_drop_privilege was set or when the Exim user was set to root. The
5520 effect of the bug was that timeouts during subsequent deliveries caused
5521 crashes instead of being properly handled. The handler is now left at its
5522 default (and expected) setting.
5524 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
5525 message, again when deliver_drop_privilege is set or Exim is run as root.
5526 The bug described in (1) was not present in this case, but the tidying up
5527 of the other signals was missing. I have made the two cases consistent.
5529 3. The ignore_target_hosts setting on a manualroute router was being ignored
5530 for hosts that were looked up using the /MX notation.
5532 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
5535 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
5536 operated on the sender address. After changing the $sender_address to <>
5537 for the sender address verify, Exim was re-instated it as the original
5538 (before rewriting) address, but remembering that it had rewritten it, so it
5539 wasn't rewriting it again. This bug also had the effect of breaking the
5540 sender address verification caching when the sender address was rewritten.
5542 6. The ignore_target_hosts option was being ignored by the ipliteral router.
5543 This has been changed so that if the ip literal address matches
5544 ignore_target_hosts, the router declines.
5546 7. Added expansion conditions match_domain, match_address, and match_local_
5547 part (NOT match_host).
5549 8. The placeholder for the Received: header didn't have a length field set.
5551 9. Added code to Exim itself and to exim_lock to test for a specific race
5552 condition that could lead to file corruption when using MBX delivery. The
5553 issue is with the lockfile that is created in /tmp. If this file is removed
5554 after a process has opened it but before that process has acquired a lock,
5555 there is the potential for a second process to recreate the file and also
5556 acquire a lock. This could lead to two Exim processes writing to the file
5557 at the same time. The added code performs the same test as UW imapd; it
5558 checks after acquiring the lock that its file descriptor still refers to
5559 the same named file.
5561 10. The buffer for building added header lines was of fixed size, 8192 bytes.
5562 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
5565 11. Added the smtp_active_hostname option. If used, this will typically be made
5566 to depend on the incoming interface address. Because $interface_address is
5567 not set up until the daemon has forked a reception process, error responses
5568 that can happen earlier (such as "too many connections") no longer contain
5571 12. If an expansion in a condition on a "warn" statement fails because a lookup
5572 defers, the "warn" statement is abandoned, and the next ACL statement is
5573 processed. Previously this caused the whole ACL to be aborted.
5575 13. Added the iplsearch lookup type.
5577 14. Added ident_timeout as a log selector.
5579 15. Added tls_certificate_verified as a log selector.
5581 16. Added a global option tls_require_ciphers (compare the smtp transport
5582 option of the same name). This controls incoming TLS connections.
5584 17. I finally figured out how to make tls_require_ciphers do a similar thing
5585 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
5586 before starting the TLS session.
5588 18. Tabs are now shown as \t in -bP output.
5590 19. If the log selector return_path_on_delivery was set, Exim crashed when
5591 bouncing a message because it had too many Received: header lines.
5593 20. If two routers both had headers_remove settings, and the first one included
5594 a superfluous trailing colon, the final name in the first list and the
5595 first name in the second list were incorrectly joined into one item (with a
5596 colon in the middle).
5602 1. Added -C and -D options to the exinext utility, mainly to make it easier
5603 to include in the automated testing, but these could be helpful when
5604 multiple configurations are in use.
5606 2. The exinext utility was not formatting the output nicely when there was
5607 an alternate port involved in the retry record key, nor when there was a
5608 message id as well (for retries that were specific to a specific message
5609 and a specific host). It was also confused by IPv6 addresses, because of
5610 the additional colons they contain. I have fixed the IPv4 problem, and
5611 patched it up to do a reasonable job for IPv6.
5613 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
5614 delivery, the log line now contains "pipelined" if PIPELINING was used.
5616 4. An SMTP transport process used to panic and die if the bind() call to set
5617 an explicit outgoing interface failed. This has been changed; it is now
5618 treated in the same way as a connect() failure.
5620 5. A reference to $sender_host_name in the part of a conditional expansion
5621 that was being skipped was still causing a DNS lookup. This no longer
5624 6. The def: expansion condition was not recognizing references to header lines
5625 that used bh_ and bheader_.
5627 7. Added the _cache feature to named lists.
5629 8. The code for checking quota_filecount in the appendfile transport was
5630 allowing one more file than it should have been.
5632 9. For compatibility with Sendmail, the command line option
5640 and sets the incoming protocol and host name (for trusted callers). The
5641 host name and its colon can be omitted when only the protocol is to be set.
5642 Note the Exim already has two private options, -pd and -ps, that refer to
5643 embedded Perl. It is therefore impossible to set a protocol value of "d" or
5644 "s", but I don't think that's a major issue.
5646 10. A number of refactoring changes to the code, none of which should affect
5649 (a) The number of logging options was getting close to filling up the
5650 32-bit word that was used as a bit map. I have split them into two classes:
5651 those that are passed in the argument to log_write(), and those that are
5652 only ever tested independently outside of that function. These are now in
5653 separate 32-bit words, so there is plenty of room for expansion again.
5654 There is no change in the user interface or the logging behaviour.
5656 (b) When building, for example, log lines, the code previously used a
5657 macro that called string_cat() twice, in order to add two strings. This is
5658 not really sufficiently general. Furthermore, there was one instance where
5659 it was actually wrong because one of the argument was used twice, and in
5660 one call a function was used. (As it happened, calling the function twice
5661 did not affect the overall behaviour.) The macro has been replaced by a
5662 function that can join an arbitrary number of extra strings onto a growing
5665 (c) The code for expansion conditions now uses a table and a binary chop
5666 instead of a serial search (which was left over from when there were very
5667 few conditions). Also, it now recognizes conditions like "pam" even when
5668 the relevant support is not compiled in: a suitably worded error message is
5669 given if an attempt is made to use such a condition.
5671 11. Added ${time_interval:xxxxx}.
5673 12. A bug was causing one of the ddress fields not to be passed back correctly
5674 from remote delivery subprocesses. The field in question was not being
5675 subsequently used, so this caused to problems in practice.
5677 13. Added new log selectors queue_time and deliver_time.
5679 14. Might have fixed a bug in maildirsizefile handling that threw up
5680 "unexpected character" debug warnings, and recalculated the data
5681 unnecessarily. In any case, I expanded the warning message to give more
5684 15. Added the message "Restricted characters in address" to the statements in
5685 the default ACL that block characters like @ and % in local parts.
5687 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
5688 Three changes have been made:
5690 (a) There was a serious bug; a negative response to MAIL caused the whole
5691 recipient domain to be cached as invalid, thereby blocking all messages
5692 to all local parts at the same domain, from all senders. This bug has
5693 been fixed. The domain is no longer cached after a negative response to
5694 MAIL if the sender used is not empty.
5696 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
5699 (c) A new callout option, "use_sender" has been added for people who want
5700 the modified behaviour.
5706 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
5709 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
5710 indeed breaks things for older releases.
5712 3. Added additional logging to the case where there is a problem reading data
5713 from a filter that is running in a subprocess using a pipe, in order to
5714 try to track down a specific problem.
5716 4. Testing facility fudge: when running in the test harness and attempting
5717 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
5718 getting "No route to host". Convert this to a timeout.
5720 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
5723 6. Some OS don't have socklen_t but use size_t instead. This affects the
5724 fifth argument of getsockopt() amongst other things. This is now
5725 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
5726 can be set for individual OS. I have set it for SunOS5, OSF1, and
5727 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
5728 some earlier ones do not.
5730 7. Change 4.30/15 was not doing the test caselessly.
5732 8. The standard form for an IPv6 address literal was being rejected by address
5733 parsing in, for example, MAIL and RCPT commands. An example of this kind of
5734 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5735 this, as well as the form without the "IPv6" on the front (but only when
5736 address literals are enabled, of course).
5738 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
5740 10. Exim crashed if a message with an empty sender address specified by -f
5741 encountered a router with an errors_to setting. This could be provoked only
5742 by a command such as
5746 where an empty string was supplied; "<>" did not hit this bug.
5748 11. Installed PCRE release 4.5.
5750 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
5751 remained set. It is now erased.
5753 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
5754 times from message ids (which are base 36 rather than the normal 62).
5756 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
5757 were being counted as actual protocol errors, and logged if the log
5758 selector +smtp_protocol_error was set. One cannot be perfect in this test,
5759 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
5760 and DATA following a set of rejected RCPTs do not count as protocol errors.
5761 In other words, Exim assumes they were pipelined, though this may not
5762 actually be the case. Of course, in all cases the client gets an
5763 appropriate error code.
5765 15. If a lookup fails in an ACL condition, a message about the failure may
5766 be available; it is used if testing the ACL cannot continue, because most
5767 such messages specify what the cause of the deferral is. However, some
5768 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
5769 that caused an old message to be retained and used if a later statement
5770 caused a defer, replacing the real cause of the deferral.
5772 16. If an IP address had so many PTR records that the DNS lookup buffer
5773 was not large enough to hold them, Exim could crash while trying to process
5774 the truncated data. It now detects and logs this case.
5776 17. Further to 4.21/58, another change has been made: if (and only if) the
5777 first line of a message (the first header line) ends with CRLF, a bare LF
5778 in a subsequent header line has a space inserted after it, so as not to
5779 terminate the header.
5781 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
5782 unnecessarily, used atoi() instead of strtol(), and didn't check the
5783 termination when getting file sizes from file names by regex.
5785 19. Completely re-implemented the support for maildirsize files, in the light
5786 of a number of problems with the previous contributed implementation
5787 (4.30/29). In particular:
5789 . If the quota is zero, the maildirsize file is maintained, but no quota is
5792 . If the maildir directory does not exist, it is created before any attempt
5793 to write a maildirsize file.
5795 . The quota value in the file is just a cache; if the quota is changed in
5796 the transport, the new value overrides.
5798 . A regular expression is available for excluding directories from the
5801 20. The autoreply transport checks the characters in options that define the
5802 message's headers; it allows continued headers, but it was checking with
5803 isspace() after an embedded newline instead of explicitly looking for a
5806 21. If all the "regular" hosts to which an address was routed had passed their
5807 expiry times, and had not reached their retry times, the address was
5808 bounced, even if fallback hosts were defined. Now Exim should go on to try
5811 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
5812 equivalent code in the SMTP transport. Some hosts send humungous responses
5813 to HELO/EHLO, more than 1024 it seems.
5815 23. Refactoring: code in filter.c used (void *) for "any old type" but this
5816 gives compiler warnings in some environments. I've now done it "properly",
5819 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
5820 (because of problems with the built-in one) was declared to return uschar *
5821 instead of char *, causing compiler failure.
5823 25. Fixed a file descriptor leak when processing alias/forward files.
5825 26. Fixed a minor format string issue in dbfn.c.
5827 27. Typo in exim.c: ("dmbnz" for "dbmnz").
5829 28. If a filter file refered to $h_xxx or $message_headers, and the headers
5830 contained RFC 2047 "words", Exim's memory could, under certain conditions,
5833 29. When a sender address is verified, it is cached, to save repeating the test
5834 when there is more than one recipient in a message. However, when the
5835 verification involves a callout, it is possible for different callout
5836 options to be set for different recipients. It is too complicated to keep
5837 track of this in the cache, so now Exim always runs a verification when a
5838 callout is required, relying on the callout cache for the optimization.
5839 The overhead is duplication of the address routing, but this should not be
5842 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
5843 to be verified with callout=postmaster, and the main callout worked but the
5844 postmaster check failed, the verification correctly failed. However, if a
5845 subsequent RCPT command asked for sender verification *without* the
5846 postmaster check, incorrect caching caused this verification also to fail,
5849 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
5850 it was not caching the DNS options (qualify_single, search_parents) that
5851 were used when the lookup failed. A subsequent lookup with different
5852 options therefore always gave the same answer, though there were cases
5853 where it should not have. (Example: a "domains = !$mx_any" option on a
5854 dnslookup router: the "domains" option is always processed without any
5855 widening, but the router might have qualify_single set.) Now Exim uses the
5856 cached value only when the same options are set.
5858 32. Added John Jetmore's "exipick" utility to the distribution.
5860 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
5861 than a timeout (e.g. a certificate is required, and is not provided), an
5862 Exim server now closes the connection immediately. Previously it waited for
5863 the client to close - but if the client is SSL, it seems that they each
5864 wait for each other, leading to a delay before one of them times out.
5866 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
5867 maintained 0.8.x compatibility because I don't think many are using it, and
5868 it is clearly obsolete.
5870 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
5873 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
5874 client certificate was expired. A simple patch fixes this, though I don't
5875 understand the full logic of why the verify callback is called multiple
5878 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
5879 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
5880 which causes problems with some clients (such as the Certicom SSL Plus
5881 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
5882 disables the coutermeasure allowing Eudora to connect."
5884 38. Exim was not checking that a write() to a log file succeeded. This could
5885 lead to Bad Things if a log got too big, in particular if it hit a file
5886 size limit. Exim now panics and dies if it cannot write to a log file, just
5887 as it does if it cannot open a log file.
5889 39. Modified OS/Makefile-Linux so that it now contains
5891 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
5893 The two -D definitions ensure that Exim is compiled with large file
5894 support, which makes it possible to handle log files that are bigger than
5897 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
5898 instance) a domain was checked against a named list that involved a lookup,
5899 causing $domain_data to be set, then another domain was checked against the
5900 same list, then the first domain was re-checked, the value of $domain_data
5901 after the final check could be wrong. In particular, if the second check
5902 failed, it could be set empty. This bug probably also applied to
5905 41. The strip_trailing_dot option was not being applied to the address given
5906 with the -f command-line option.
5908 42. The code for reading a message's header from the spool was incrementing
5909 $received_count, but never initializing it. This meant that the value was
5910 incorrect (doubled) while delivering a message in the same process in which
5911 it was received. In the most common configuration of Exim, this never
5912 happens - a fresh exec is done - but it can happen when
5913 deliver_drop_privilege is set.
5915 43. When Exim logs an SMTP synchronization error - client data sent too soon -
5916 it now includes up to 150 characters of the unexpected data in the log
5919 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
5920 and building data strings. The size of both of these buffers was 10 000
5921 bytes - far larger than anybody would *ever* want, thought I. Needless to
5922 say, somebody hit the limit. I have increased the maximum line length to
5923 20 000 and the maximum data length of concatenated lines to 100 000. I have
5924 also fixed two bugs, because there was no checking on these buffers. Tsk,
5925 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
5926 buffer is too small.
5928 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
5929 lsearch lookups. Now it does.
5931 46. When parsing a route_list item in a manualroute router, a fixed-length
5932 buffer was used for the list of hosts. I made this 1024 bytes long,
5933 thinking that nobody would ever have a list of hosts that long. Wrong.
5934 Somebody had a whole pile of complicated expansion conditions, and the
5935 string was silently truncated, leading to an expansion error. It turns out
5936 that it is easier to change to an unlimited length (owing to other changes
5937 that have happened since this code was originally written) than to build
5938 structure for giving a limitation error. The length of the item that
5939 expands into the list of hosts is now unlimited.
5941 47. The lsearch lookup could not handle data where the length of text line was
5942 more than 4095 characters. Such lines were truncated, leading to shortened
5943 data being returned. It should now handle lines of any length.
5945 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5946 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5949 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5950 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5951 no longer happens in comment lines. A list of replacements is now placed
5952 at the head of all of the source files, except those whose only change is
5953 to replace PERL_COMMAND in the very first #! line.
5955 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5956 messages on the queue, with a bottom-up merge sort, using code contributed
5957 by Michael Haardt. This should make operations like -bp somewhat faster on
5958 large queues. It won't affect queue runners, except when queue_run_in_order
5961 51. Installed eximstats 1.31 in the distribution.
5963 52. Added support for SRV lookups to the dnslookup router.
5965 53. If an ACL referred to $message_body or $message_body_end, the value was not
5966 reset for any messages that followed in the same SMTP session.
5968 54. The store-handling optimization for building very long strings was not
5969 differentiating between the different store pools. I don't think this
5970 actually made any difference in practice, but I've tidied it.
5972 55. While running the routers to verify a sender address, $sender_address
5973 was still set to the sender address. This is wrong, because when routing to
5974 send a bounce to the sender, it would be empty. Therefore, I have changed
5975 it so that, while verifying a sender address, $sender_address is set to <>.
5976 (There is no change to what happens when verifying a recipient address.)
5978 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5979 target A or AAAA records (if not already returned) without resetting the
5980 qualify_single or search_parents options of the DNS resolver. These are
5981 inappropriate in this case because the targets of MX and SRV records must
5982 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5983 target that, when qualified, matched something in the local domain. These
5984 two options are now turned off when doing these lookups.
5986 57. It seems that at least some releases of Reiserfs (which does not have the
5987 concept of a fixed number of inodes) returns zero and not -1 for the
5988 number of available inodes. This interacted badly with check_spool_inodes,
5989 which assumed that -1 was the "no such thing" setting. What I have done is
5990 to check that the total number of inodes is greater than zero before doing
5991 the test of how many are available.
5993 58. When a "warn" ACL statement has a log_message modifier, the message is
5994 remembered, and not repeated. This is to avoid a lot of repetition when a
5995 message has many recipients that cause the same warning to be written.
5996 However, Exim was preserving the list of already written lines for an
5997 entire SMTP session, which doesn't seem right. The memory is now reset if a
5998 new message is started.
6000 59. The "rewrite" debugging flag was not showing the result of rewriting in the
6001 debugging output unless log_rewrite was also set.
6003 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
6004 of (int)(handle) when we know that handle contains (void *)(-1).
6006 61. The Exim daemon panic-logs an error return when it closes the incoming
6007 connection. However "connection reset by peer" seems to be common, and
6008 isn't really an error worthy of noting specially, so that particular error
6011 62. When Exim is trying to find all the local interfaces, it used to panic and
6012 die if the ioctl to get the interface flags failed. However, it seems that
6013 on at least one OS (Solaris 9) it is possible to have an interface that is
6014 included in the list of interfaces, but for which you get a failure error
6015 for this call. This happens when the interface is not "plumbed" into a
6016 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
6017 failure of the "get flags" call assumes that the interface is down.
6019 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
6020 makes life easier for people who are doing arithmetic on fields extracted
6021 from dates, where you often get leading zeros that should not be
6022 interpreted as octal.
6024 64. Added qualify_domain to the redirect router, to override the global
6027 65. If a pathologically long header line contained very many addresses (the
6028 report of this problem mentioned 10 000) and each of them was rewritten,
6029 Exim could use up a very large amount of memory. (It kept on making new
6030 copies of the header line as it rewrote, and never released the old ones.)
6031 At the expense of a bit more processing, the header rewriting function has
6032 been changed so that it no longer eats memory in this way.
6034 66. The generation of the Received: header has been moved from the time that a
6035 message starts to be received, to the time that it finishes. The timestamp
6036 in the Received: header should now be very close to that of the <= log
6037 line. There are two side-effects of this change:
6039 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
6040 logged header lines no longer include the local Received: line, because
6041 it has not yet been created. The same applies to a copy of the message
6042 that is returned to a non-SMTP sender when a message is rejected.
6044 (b) When a filter file is tested using -bf, no additional Received: header
6045 is added to the test message. After some thought, I decided that this
6048 This change does not affect the value of $received_for. It is still set
6049 after address rewriting, but before local_scan() is called.
6051 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
6053 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
6054 gave an unhelpful panic error message, and a defer error. I have managed to
6055 change this behaviour so that it now rejects any supplied certificate,
6056 which seems right, as the list of acceptable certificates is empty.
6058 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
6059 gave an unhelpful defer error. I have not managed to make this reject any
6060 supplied certificates, but the error message it gives is "no certificate
6061 supplied", which is not helpful.
6063 70. exigrep's output now also includes lines that are not associated with any
6064 message, but which match the given pattern. Implemented by a patch from
6065 Martin Sluka, which also tidied up the Perl a bit.
6067 71. Recipient callout verification, like sender verification, was using <> in
6068 the MAIL FROM command. This isn't really the right thing, since the actual
6069 sender may affect whether the remote host accepts the recipient or not. I
6070 have changed it to use the actual sender in the callout; this means that
6071 the cache record is now keyed on a recipient/sender pair, not just the
6072 recipient address. There doesn't seem to be a real danger of callout loops,
6073 since a callout by the remote host to check the sender would use <>.
6074 [SEE ABOVE: changed after hitting problems.]
6076 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
6077 temporary errors. However, in the case of such a code being given after
6078 the end of a data transmission (i.e. after ".") Exim was failing to write
6079 a retry record for the message. (Yes, there was some broken host that was
6080 actually sending 8xx at this point.)
6082 73. An unknown lookup type in a host list could cause Exim to panic-die when
6083 the list was checked. (An example that provoked this was putting <; in the
6084 middle of a list instead of at the start.) If this happened during a DATA
6085 ACL check, a -D file could be left lying around. This kind of configuration
6086 error no longer causes Exim to die; instead it causes a defer error. The
6087 incident is still logged to the main and panic logs.
6089 74. Buglet left over from Exim 3 conversion. The message "too many messages
6090 in one connection" was written to the rejectlog but not the mainlog, except
6091 when address rewriting (yes!) was being logged.
6093 75. Added write_rejectlog option.
6095 76. When a system filter was run not as root (that is, when system_filter_user
6096 was set), the values of the $n variables were not being returned to the
6097 main process; thus, they were not subsequently available in the $sn
6100 77. Added +return_path_on_delivery log selector.
6102 78. A connection timeout was being treated differently from recipients deferred
6103 when testing hosts_max_try with a message that was older than the host's
6104 retry timeout. (The host should not be counted, thus allowing all hosts to
6105 be tried at least once before bouncing.) This may have been the cause of an
6106 occasionally reported bug whereby a message would remain on the queue
6107 longer than the retry timeout, but would be bounced if a delivery was
6108 forced. I say "may" because I never totally pinned down the problem;
6109 setting up timeout/retry tests is difficult. See also the next item.
6111 79. The ultimate address timeout was not being applied to errors that involved
6112 a combination of host plus message (for example, a timeout on a MAIL
6113 command). When an address resolved to a number of possible hosts, and they
6114 were not all tried for each delivery (e.g. because of hosts_max_try), a
6115 message could remain on the queue longer than the retry timeout.
6117 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
6120 81. Fixed an obscure SMTP outgoing bug which required at least the following
6121 conditions: (a) there was another message waiting for the same server;
6122 (b) the server returned 5xx to all RCPT commands in the first message so
6123 that the message was not completed; (c) the server dropped the connection
6124 or gave a negative response to the RSET that Exim sends to abort the
6125 transaction. The observed case was a dropped connection after DATA that had
6126 been sent in pipelining mode. That is, the server had advertised PIPELINING
6127 but was not implementing it correctly. The effect of the bug was incorrect
6128 behaviour, such as trying another host, and this could lead to a crash.
6134 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
6135 and daemon.c were passed as pointers to ints; they should have been
6136 pointers to socklen_t variables (which are typically unsigned ints).
6138 2. Some signed/unsigned type warnings in the os.c file for Linux have been
6141 3. Fixed a really odd bug that affected only the testing scheme; patching a
6142 certain fixed string in the binary changed the value of another string that
6143 happened to be identical to the end of the original first string.
6145 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
6146 name", it returns that address as the IP address. On some operating
6147 systems (e.g. Solaris), it also passes back the IP address string as the
6148 "host name". However, on others (e.g. Linux), it passes back an empty
6149 string. Exim wasn't checking for this, and was changing the host name to an
6150 empty string, assuming it had been canonicalized.
6152 5. Although rare, it is permitted to have more than one PTR record for a given
6153 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
6154 all the names associated with an address, because they do in Solaris.
6155 However, it seems that they do not in Linux for data that comes from the
6156 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
6157 I found this out when I moved to a new Linux workstation and tried to run
6158 the Exim test suite.
6160 To get round this problem I have changed the code so that it now does its
6161 own call to the DNS to look up PTR records when searching for a host name.
6162 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
6163 addresses that are only in /etc/hosts are still found.
6165 This behaviour is, however, controlled by an option called host_lookup_
6166 order, which defaults to "bydns:byaddr". If people want to use the other
6167 order, or indeed, just use one or the other means of lookup, they can
6168 specify it in this variable.
6170 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
6171 some operating systems, this comes back from gethostbyaddr() as an empty
6172 string, and this is what Exim used to test for. However, it seems that in
6173 other systems, "." is yielded. Exim now tests for this case too.
6175 7. The values of check_spool_space and check_log_space are now held internally
6176 as a number of kilobytes instead of an absolute number of bytes. If a
6177 numbers is specified without 'K' or 'M', it is rounded up to the nearest
6178 kilobyte. This means that much larger values can be stored.
6180 8. Exim monitor: an attempt to get the action menu when not actually pointing
6181 at a message produces an empty menu entitled "No message selected". This
6182 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
6183 no entries in it ("Shell widget menu has zero width and/or height"). So I
6184 have added a single, blank menu entry in this case.
6186 9. Added ${quote_local_part.
6188 10. MIME decoding is now applied to the contents of Subject: header lines when
6191 11. Now that a reference to $sender_host_address automatically causes a reverse
6192 lookup to occur if necessary (4.13/18), there is no need to arrange for a
6193 host lookup before query-style lookups in lists that might use this
6194 variable. This has therefore been abolished, and the "net-" prefix is no
6195 longer necessary for query-style lookups.
6197 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
6198 have been a typo for LFLAGS, so it has been changed.
6200 13. The install script calls Exim with "-C /dev/null" in order to find the
6201 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
6202 to be output. However, since Exim outputs its version number before the
6203 error, it didn't break the script. It just looked ugly. I fixed this by
6204 always allowing "-C /dev/null" if the caller is root.
6206 14. Ignore overlarge ACL variable number when reading spool file - insurance
6207 against a later release with more variables having written the file.
6209 15. The standard form for an IPv6 address literal was being rejected by EHLO.
6210 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
6211 this, as well as the form without the "IPv6" on the front.
6213 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
6214 OS/Makefile-Darwin file.
6216 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
6217 only to LDAP libraries that do not have LDAP_OPT_DEREF.
6219 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
6220 variable. However, some OS use ints for inodes. I've added cast to long int
6221 to get rid of the compiler warning.
6223 19. I had forgotten to lock out "/../" in configuration file names when
6224 ALT_CONFIG_PREFIX was set.
6226 20. Routers used for verification do not need to specify transports. However,
6227 if such a router generated a host list, and callout was configured, Exim
6228 crashed, because it could not find a port number from the (non-existent)
6229 transport. It now assumes port 25 in this circumstance.
6231 21. Added the -t option to exigrep.
6233 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
6234 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
6235 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
6236 them is set, LOOKUP_LSEARCH is forced.
6238 23. "exim -bV" now outputs a list of lookups that are included in the binary.
6240 24. Added sender and host information to the "rejected by local_scan()" log
6241 line; previously there was no indication of these.
6243 25. Added .include_if_exists.
6245 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
6246 Linux. It turns out that not all file systems support this. Apparently some
6247 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
6248 do it.) To cope with this, the error EINVAL, which means that sync-ing is
6249 not supported on the file descriptor, is now ignored when Exim is trying to
6250 sync a directory. This applies only to Linux.
6252 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
6254 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
6255 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
6258 29. Added support for maildirsize files from supplied patch (modified a bit).
6260 30. The use of :fail: followed by an empty string could lead Exim to respond to
6261 sender verification failures with (e.g.):
6263 550 Verification failed for <xxx>
6264 550 Sender verify failed
6266 where the first response line was missing the '-' that indicates it is not
6267 the final line of the response.
6269 31. The loop for finding the name of the user that called Exim had a hardwired
6270 limit of 10; it now uses the value of finduser_retries, which is used for
6271 all other user lookups.
6273 32. Added $received_count variable, available in data and not_smtp ACLs, and at
6276 33. Exim was neglecting to zero errno before one call of strtol() when
6277 expanding a string and expecting an integer value. On some systems this
6278 resulted in spurious "integer overflow" errors. Also, it was casting the
6279 result into an int without checking.
6281 34. Testing for a connection timeout using "timeout_connect" in the retry rules
6282 did not work. The code looks as if it has *never* worked, though it appears
6283 to have been documented since at least release 1.62. I have made it work.
6285 35. The "timeout_DNS" error in retry rules, also documented since at least
6286 1.62, also never worked. As it isn't clear exactly what this means, and
6287 clearly it isn't a major issue, I have abolished the feature by treating it
6288 as "timeout", and writing a warning to the main and panic logs.
6290 36. The display of retry rules for -brt wasn't always showing the error code
6293 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
6294 timeout_connect_A, timeout_connect_MX.
6296 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
6297 to the empty sender.
6299 39. The daemon was not analysing the content of -oX till after it had closed
6300 stderr and disconnected from the controlling terminal. This meant that any
6301 syntax errors were only noted on the panic log, and the return code from
6302 the command was 0. By re-arranging the code a little, I've made the
6303 decoding happen first, so such errors now appear on stderr, and the return
6304 code is 1. However, the actual setting up of the sockets still happens in
6305 the disconnected process, so errors there are still only recorded on the
6308 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
6309 connections (as happens on some IP stacks) was logged at start up time as
6310 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
6311 it from "IPv6 and IPv4", which means that two separate sockets are being
6314 41. The debug output for gethostbyname2() or getipnodebyname() failures now
6315 says whether AF_INET or AF_INET6 was passed as an argument.
6317 42. Exiwhat output was messed up when time zones were included in log
6320 43. Exiwhat now gives more information about the daemon's listening ports,
6321 and whether -tls-on-connect was used.
6323 44. The "port" option of the smtp transport is now expanded.
6325 45. A "message" modifier in a "warn" statement in a non-message ACL was being
6326 silently ignored. Now an error message is written to the main and panic
6329 46. There's a new ACL modifier called "logwrite" which writes to a log file
6330 as soon as it is encountered.
6332 47. Added $local_user_uid and $local_user_gid at routing time.
6334 48. Exim crashed when trying to verify a sender address that was being
6337 49. Exim was recognizing only a space character after ".include". It now also
6338 recognizes a tab character.
6340 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
6341 extracting the relevant information from the specification. The man page no
6342 longer contains scrambled data for the -d option, and I've added a section
6343 at the front about calling Exim under different names.
6345 51. Added "extra_headers" argument to the "mail" command in filter files.
6347 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
6350 53. Installed eximstats 1.29.
6352 54. Added transport_filter_timeout as a generic transport option.
6354 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
6355 Cc: header lines. This was required by RFC 822, but it not required by RFC
6358 56. Exim used to add From:, Date:, and Message-Id: header lines to any
6359 incoming messages that did not have them. Now it does so only if the
6360 message originates locally, that is, if there is no associated remote host
6361 address. When Resent- header lines are present, this applies to the Resent-
6362 lines rather than the non-Resent- lines.
6364 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
6365 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
6367 58. Messages for configuration errors now include the name of the main
6368 configuration file - useful now that there may be more than one file in a
6369 list (.included file names were always shown).
6371 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
6372 for those rare installations that do not start the daemon as root or run it
6373 setuid root. I've cut out the call to initgroups() if the daemon is not
6376 60. The Exim user and group can now be bound into the binary as text strings
6377 that are looked up at the start of Exim's processing.
6379 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
6381 62. Added $mailstore_basename variable.
6383 63. Installed patch to sieve.c from Michael Haardt.
6385 64. When Exim failed to open the panic log after failing to open the main log,
6386 the original message it was trying to log was written to stderr and debug
6387 output, but if they were not available (the usual case in production), it
6388 was lost. Now it is written to syslog before the two lines that record the
6389 failures to open the logs.
6391 65. Users' Exim filters run in subprocesses under the user's uid. It is
6392 possible for a "deliver" command or an alias in a "personal" command to
6393 provoke an address rewrite. If logging of address rewriting is configured,
6394 this fails because the process is not running as root or exim. There may be
6395 a better way of dealing with this, but for the moment (because 4.30 needs
6396 to be released), I have disabled address rewrite logging when running a
6397 filter in a non-root, non-exim process.
6403 1. The buildconfig auxiliary program wasn't quoting the value set for
6404 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
6405 not defined. This bug was masked in 4.22 by the effect that was fixed in
6408 2. Some messages that were rejected after a message id was allocated were
6409 shown as "incomplete" by exigrep. It no longer does this for messages that
6410 are rejected by local_scan() or the DATA or non-SMTP ACLs.
6412 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
6413 have allow_domain_literals set, the ID did not get logged in the <= line.
6414 Domain literals are now always recognized in Message-ID: header lines.
6416 4. The first argument for a ${extract expansion item is the key name or field
6417 number. Leading and trailing spaces in this item were not being ignored,
6418 causing some misleading effects.
6420 5. When deliver_drop_privilege was set, single queue runner processes started
6421 manually (i.e. by the command "exim -q") or by the daemon (which uses the
6422 same command in the process it spins off) were not dropping privilege.
6424 6. When the daemon running as "exim" started a queue runner, it always
6425 re-executed Exim in the spun-off process. This is a waste of effort when
6426 deliver_drop_privilege is set. The new process now just calls the
6427 queue-runner function directly.
6433 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
6436 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
6437 ignored. Though the use of -oP was forcing the writing of a pid file, it
6438 was always written to the default place.
6440 3. If the message "no IP address found for host xxxx" is generated during
6441 incoming verification, it is now followed by identification of the incoming
6442 connection (so you can more easily find what provoked it).
6444 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
6446 5. Added some features to "harden" Exim a bit more against certain attacks:
6448 (a) There is now a build-time option called FIXED_NEVER_USERS that can
6449 be put in Local/Makefile. This is like the never_users runtime option,
6450 but it cannot be overridden. The default setting is "root".
6452 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
6453 prefix string with which any file named in a -C command line option
6456 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
6457 is retained for -C and -D only if the caller of Exim is root. Without
6458 it, the exim user may also use -C and -D and retain privilege.
6460 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
6461 command line option is disabled.
6463 6. Macro names set by the -D option must start with an upper case letter, just
6464 like macro names defined in the configuration file.
6466 7. Added "dereference=" facility to LDAP.
6468 8. Two instances of the typo "uknown" in the source files are fixed.
6470 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
6471 the Configure-Makefile script screwed up while processing it.
6473 10. Incorporated PCRE 4.4.
6475 11. The SMTP synchronization check was not operating right at the start of an
6476 SMTP session. For example, it could not catch a HELO sent before the client
6477 waited for the greeting. There is now a check for outstanding input at the
6478 point when the greeting is written. Because of the duplex, asynchronous
6479 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
6480 way, but not yet received, when the check is performed.
6482 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
6483 on TCP/IP sockets, because this apparently causes some broken clients to
6486 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
6487 unchanged) from the Cygwin maintainer.
6489 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
6490 was supported instead of testing for mbx. Effectively a typo.
6492 15. The spa authenticator server code was not checking that the input it
6493 received was valid base64.
6495 16. The debug output line for the "set" modifier in ACLs was not showing the
6496 name of the variable that was being set.
6498 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
6500 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
6501 lookup on its own. Something else was needed to trigger the lookup. For
6502 example, a match in host_lookup or the need for a host name in a host list.
6503 Now, if $sender_host_name is referenced and the host name has not yet been
6504 looked up, a lookup is performed. If the lookup fails, the variable remains
6505 empty, and $host_lookup_failed is set to "1".
6507 19. Added "eqi" as a case-independent comparison operator.
6509 20. The saslauthd authentication condition could segfault if neither service
6510 nor realm was specified.
6512 21. If an overflowing value such as "2048M" was set for message_size_limit, the
6513 error message that was logged was misleading, and incoming SMTP
6514 connections were dropped. The message is now more accurate, and temporary
6515 errors are given to SMTP connections.
6517 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
6518 (except RSET) with a 421 error, until QUIT is received. However, it was
6519 failing to send a response to QUIT.
6521 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
6522 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
6523 ACL is now run after the helo_try_verify_hosts code.
6525 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
6526 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
6527 case-independent, so other case variants are also recognized). Apparently
6528 some systems use these upper case variants.
6530 25. If more than two messages were waiting for the same host, and a transport
6531 filter was specified for the transport, Exim sent two messages over the
6532 same TCP/IP connection, and then failed with "socket operation on non-
6533 socket" when it tried to send the third.
6535 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
6537 27. The extern definition of crypt16() in expand.c was not being excluded when
6538 the OS had its own crypt16() function.
6540 28. Added bounce_return_body as a new option, and bounce_return_size_limit
6541 as a preferred synonym for return_size_limit, both as an option and as an
6544 29. Added LIBS=-liconv to OS/Makefile-OSF1.
6546 30. Changed the default configuration ACL to relax the local part checking rule
6547 for addresses that are not in any local domains. For these addresses,
6548 slashes and pipe symbols are allowed within local parts, but the sequence
6549 /../ is explicitly forbidden.
6551 31. SPA server authentication was not clearing the challenge buffer before
6554 32. log_message in a "warn" ACL statement was writing to the reject log as
6555 well as to the main log, which contradicts the documentation and doesn't
6556 seem right (because no rejection is happening). So I have stopped it.
6558 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
6559 However, I am unable to do any testing of this.
6561 34. Fixed an infelicity in the appendfile transport. When checking directories
6562 for a mailbox, to see if any needed to be created, it was accidentally
6563 using path names with one or more superfluous leading slashes; tracing
6564 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
6566 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
6567 contents are added to the log line that is written for every discarded
6568 recipient. (Previously a log_message setting was ignored.)
6570 36. The ${quote: operator now quotes the string if it is empty.
6572 37. The install script runs exim in order to find its version number. If for
6573 some reason other than non-existence or emptiness, which it checks, it
6574 could not run './exim', it was installing it with an empty version number,
6575 i.e. as "exim-". This error state is now caught, and the installation is
6578 38. An argument was missing from the function that creates an error message
6579 when Exim fails to connect to the socket for saslauthd authentication.
6580 This could cause Exim to crash, or give a corrupted message.
6582 39. Added isip, isip4, and isip6 to ${if conditions.
6584 40. The ACL variables $acl_xx are now saved with the message, and can be
6585 accessed later in routers, transports, and filters.
6587 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
6588 strings in the file are not string-expanded.
6590 42. If a MAIL command specified a SIZE value that was too large to fit into an
6591 int variable, the check against message_size_limit failed. Such values are
6592 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
6593 day this will have to be increased, but I don't think I want to be around
6594 when emails are that large.
6601 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
6602 iconv() is not standard in FreeBSD.
6604 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
6605 IPv6 enabled. The observed symptom was a segmentation fault on return from
6606 the function os_common_find_running_interfaces() in src/os.c.
6608 3. In the check_special_case() function in daemon.c I had used "errno" as an
6609 argument name, which causes warnings on some systems. This was basically a
6610 typo, since it was named "eno" in the comments!
6612 4. The code that waits for the clock to tick (at a resolution of some fraction
6613 of a second) so as to ensure message-id uniqueness was always waiting for
6614 at least one whole tick, when it could have waited for less. [This is
6615 almost certainly not relevant at current processor speeds, where it is
6616 unlikely to ever wait at all. But we try to future-proof.]
6618 5. The function that sleeps for a time interval that includes fractions of a
6619 second contained a race. It did not block SIGALRM between setting the
6620 timer, and suspending (a couple of lines later). If the interval was short
6621 and the sigsuspend() was delayed until after it had expired, the suspension
6622 never ended. On busy systems this could lead to processes getting stuck for
6625 6. Some uncommon configurations may cause a lookup to happen in a queue runner
6626 process, before it forks any delivery processes. The open lookup caching
6627 mechanism meant that the open file or database connection was passed into
6628 the delivery process. The problem was that delivery processes always tidy
6629 up cached lookup data. This could cause a problem for the next delivery
6630 process started by the queue runner, because the external queue runner
6631 process does not know about the closure. So the next delivery process
6632 still has data in the lookup cache. In the case of a file lookup, there was
6633 no problem because closing a file descriptor in a subprocess doesn't affect
6634 the parent. However, if the lookup was caching a connection to a database,
6635 the connection was closed, and the second delivery process was likely to
6636 see errors such as "PGSQL: query failed: server closed the connection
6637 unexpectedly". The problem has been fixed by closing all cached lookups
6638 in a queue runner before running a delivery process.
6640 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
6641 seem to have the "const" qualifier which it has on other OS. I've
6644 8. Change 4.21/2 was too strict. It is only if there are two authenticators
6645 *of the same type* (client or server) with the same public name that an
6646 error should be diagnosed.
6648 9. When Exim looked up a host name for an IP address, but failed to find the
6649 original IP address when looking up the host name (a safety check), it
6650 output the message "<ip address> does not match any IP for NULL", which was
6651 confusing, to say the least. The bug was that the host name should have
6652 appeared instead of "NULL".
6654 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
6655 user that is built into the binary, and the -C or -D options is used, root
6656 privilege is dropped before the configuration file is read. In addition,
6657 logging is switched to stderr instead of the normal log files. If the
6658 configuration then re-defines the Exim user, the unprivileged environment
6659 is probably not what is expected, so Exim logs a panic warning message (but
6662 However, if deliver_drop_privilege is set, the unprivileged state may well
6663 be exactly what is intended, so the warning has been cut out in that case,
6664 and Exim is allowed to try to write to its normal log files.
6670 1. smtp_return_error_details was not giving details for temporary sender
6671 or receiver verification errors.
6673 2. Diagnose a configuration error if two authenticators have the same public
6676 3. Exim used not to create the message log file for a message until the first
6677 delivery attempt. This could be confusing when incoming messages were held
6678 for policy or load reasons. The message log file is now created at the time
6679 the message is received, and an initial "Received" line is written to it.
6681 4. The automatically generated man page for command line options had a minor
6682 bug that caused no ill effects; however, a more serious problem was that
6683 the procedure for building the man page automatically didn't always
6684 operate. Consequently, release 4.20 contains an out-of-date version. This
6685 shouldn't happen again.
6687 5. When building Exim with embedded Perl support, the script that builds the
6688 Makefile was calling 'perl' to find its compile-time parameters, ignoring
6689 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
6691 6. The freeze_tell option was not being used for messages that were frozen on
6692 arrival, either by an ACL or by local_scan().
6694 7. Added the smtp_incomplete_transaction log selector.
6696 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
6697 was accepting AUTH without a new EHLO.
6699 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
6700 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
6703 10. Logging of TCP/IP connections (when configured) now happens in the main
6704 daemon process instead of the child process, so that the TCP/IP connection
6705 count is more accurate (but it can never be perfect).
6707 11. The use of "drop" in a nested ACL was not being handled correctly in the
6708 outer ACL. Now, if condition failure induced by the nested "drop" causes
6709 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
6710 or "require"), the connection is dropped.
6712 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
6713 that yield "discard" can now be used with an "accept" or a "discard" verb,
6714 but an error is generated for any others (because I can't see a useful way
6715 to define what should happen).
6717 13. When an ACL is read dynamically from a file (or anywhere else), the lines
6718 are now processed in the same way as lines in the Exim configuration file.
6719 In particular, continuation lines are supported.
6721 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
6723 15. Added -ti meaning -t -i.
6725 16. Check for letters, digits, hyphens, and dots in the names of dnslist
6726 domains, and warn by logging if others are found.
6728 17. At least on BSD, alignment is not guaranteed for the array of ifreq's
6729 returned from GIFCONF when Exim is trying to find the list of interfaces on
6730 a host. The code in os.c has been modified to copy each ifreq to an aligned
6731 structure in all cases.
6733 Also, in some cases, the returned ifreq's were being copied to a 'struct
6734 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
6735 means the last couple of bytes of an IPv6 address could be chopped if the
6736 ifreq contained only a normal sockaddr (14 bytes storage).
6738 18. Named domain lists were not supported in the hosts_treat_as_local option.
6739 An entry such as +xxxx was not recognized, and was treated as a literal
6742 19. Ensure that header lines added by a DATA ACL are included in the reject log
6743 if the ACL subsequently rejects the message.
6745 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
6746 MD5 (which is deprecated).
6748 21. When testing a filter file using -bf, Exim was writing a message when it
6749 took the sender from a "From " line in the message, but it was not doing so
6750 when it took $return_path from a Return-Path: header line. It now does.
6752 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
6753 with a valid header line field name (a series of printing characters
6754 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
6756 23. Changed "disc" in the source to "disk" to conform to the documentation and
6757 the book and for uniformity.
6759 24. Ignore Sendmail's -Ooption=value command line item.
6761 25. When execve() failed while trying to run a command in a pipe transport,
6762 Exim was returning EX_UNAVAILABLE (69) from the subprocess. However, this
6763 could be confused with a return value of 69 from the command itself. This
6764 has been changed to 127, the value the shell returns if it is asked to run
6765 a non-existent command. The wording for the related log line suggests a
6766 non-existent command as the problem.
6768 26. If received_header_text expands to an empty string, do not add a Received:
6769 header line to the message. (Well, it adds a token one on the spool, but
6770 marks it "old" so that it doesn't get used or transmitted.)
6772 27. Installed eximstats 1.28 (addition of -nt option).
6774 28. There was no check for failure on the call to getsockname() in the daemon
6775 code. This can fail if there is a shortage of resources on the system, with
6776 ENOMEM, for example. A temporary error is now given on failure.
6778 29. Contrary to the C standard, it seems that in some environments, the
6779 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
6780 program. Exim now does this explicitly; it affects the formatting of
6781 timestamps using strftime().
6783 30. If exiqsumm was given junk data, it threw up some uninitialized variable
6784 complaints. I've now initialized all the variables, to avoid this.
6786 32. Header lines added by a system filter were not being "seen" during
6787 transport-time rewrites.
6789 33. The info_callback() function passed to OpenSSL is set up with type void
6790 (*)(SSL *, int, int), as described somewhere. However, when calling the
6791 function (actually a macro) that sets it up, the type void(*)() is
6792 expected. I've put in a cast to prevent warnings from picky compilers.
6794 34. If a DNS black list lookup found a CNAME record, but there were no A
6795 records associated with the domain it pointed at, Exim crashed.
6797 35. If a DNS black list lookup returned more than one A record, Exim ignored
6798 all but the first. It now scans all returned addresses if a particular IP
6799 value is being sought. In this situation, the contents of the
6800 $dnslist_value variable are a list of all the addresses, separated by a
6803 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
6804 to accept a host name and all its aliases if the forward lookup for any of
6805 them yielded the IP address of the incoming connection. Now it accepts only
6806 those names whose forward lookup yields the correct IP address. Any other
6807 names are discarded. This closes a loophole whereby a rogue DNS
6808 administrator could create reverse DNS records to break through a
6809 wildcarded host restriction in an ACL.
6811 37. If a user filter or a system filter that ran in a subprocess used any of
6812 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
6813 the wrong values were passed to the pipe command ($thisaddress had the
6814 value of $0, $0 had the value of $1, etc). This bug was introduced by
6815 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
6817 38. Improved the line breaking for long SMTP error messages from ACLs.
6818 Previously, if there was no break point between 40 and 75 characters, Exim
6819 left the rest of the message alone. Two changes have been made: (a) I've
6820 reduced the minimum length to 35 characters; (b) if it can't find a break
6821 point between 35 and 75 characters, it looks ahead and uses the first one
6822 that it finds. This may give the occasional overlong line, but at least the
6823 remaining text gets split now.
6825 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
6826 file descriptors might be low, and that setting 1000 would always raise it.
6827 It turns out that in some environments, the limit is already over 1000 and
6828 that lowering it causes trouble. So now Exim takes care not to decrease it.
6830 40. When delivering a message, the value of $return_path is set to $sender_
6831 address at the start of routing (routers may change the value). By an
6832 oversight, this default was not being set up when an address was tested by
6833 -bt or -bv, which affected the outcome if any router or filter referred to
6836 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
6837 log only when "message" or "log_message" are set. However, if one of the
6838 conditions was an address verification, or a call to a nested ACL, the
6839 messages generated by the underlying test were being passed through. This
6840 no longer happens. The underlying message is available in $acl_verify_
6841 message for both "message" and "log_message" expansions, so it can be
6842 passed through if needed.
6844 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
6845 new expansion $bh_ to give the encoded byte string without charset
6846 translation. Translation happens only if iconv() is available; HAVE_ICONV
6847 indicates this at build time. HEADERS_CHARSET gives the charset to
6848 translate to; headers_charset can change it in the configuration, and
6849 "headers charset" can change it in an individual filter file.
6851 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
6852 that creates RFC 2047 encoded "words" labels them as that charset instead
6853 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
6854 expansion operator; (ii) when Exim creates a From: line for a local
6855 message; (iii) when a header line is rewritten to include a "phrase" part.
6857 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
6858 buggy, causing it to skip the first lines of messages whose message ID
6859 ended in 'D'. This would not have bitten before Exim release 4.14, because
6860 message IDs were unlikely to end in 'D' before then. The effect was to have
6861 incorrect size information for certain domains.
6863 45. #include "config.h" was missing at the start of the crypt16.c module. This
6864 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
6867 46. If there was a timeout during a "random" callout check, Exim treated it as
6868 a failure of the random address, and carried on sending RSET and the real
6869 address. If the delay was just some slowness somewhere, the response to the
6870 original RCPT would be taken as a response to RSET and so on, causing
6871 mayhem of various kinds.
6873 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
6874 when I implemented it. It didn't allow for the fact that some option values
6875 may legitimately be negative (e.g. size_addition), and it didn't even do
6876 the right test for positive values.
6878 48. Domain names in DNS records are case-independent. Exim always looks them up
6879 in lower case. Some resolvers return domain names in exactly the case they
6880 appear in the zone file, that is, they may contain uppercase letters. Not
6881 all resolvers do this - some return always lower case. Exim was treating a
6882 change of case by a resolver as a change of domain, similar to a widening
6883 of a domain abbreviation. This triggered its re-routing code and so it was
6884 trying to route what was effectively the same domain again. This normally
6885 caused routing to fail (because the router wouldn't handle the domain
6886 twice). Now Exim checks for this case specially, and just changes the
6887 casing of the domain that it ultimately uses when it transmits the message
6890 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
6893 50. If a filter generated a file delivery with a non-absolute name (possible if
6894 no home directory exists for the router), the forbid_file option was not
6897 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
6898 the existing equality matching.
6900 52. Exim was using ints instead of ino_t variables in some places where it was
6901 dealing with inode numbers.
6903 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
6904 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
6905 called TMPDIR, and if it finds it is different, it changes its value.
6907 54. The smtp_printf() function is now made available to local_scan() so
6908 additional output lines can be written before returning. There is also an
6909 smtp_fflush() function to enable the detection of a dropped connection.
6910 The variables smtp_input and smtp_batched_input are exported to
6913 55. Changed the default runtime configuration: the message "Unknown user"
6914 has been removed from the ACL, and instead placed on the localuser router,
6915 using the cannot_route_message feature. This means that any verification
6916 failures that generate their own messages won't get overridden. Similarly,
6917 the "Unrouteable address" message that was in the ACL for unverifiable
6918 relay addresses has also been removed.
6920 56. Added hosts_avoid_esmtp to the smtp transport.
6922 57. The exicyclog script was not checking for the esoteric option
6923 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
6924 will work only if exicyclog is run under the appropriate euid.
6926 58. Following a discussion on the list, the rules by which Exim recognises line
6927 endings on incoming messages have been changed. The -dropcr and drop_cr
6928 options are now no-ops, retained only for backwards compatibility. The
6929 following line terminators are recognized: LF CRLF CR. However, special
6930 processing applies to CR:
6932 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
6933 nor a local message in the state where . is a terminator.
6935 (ii) If a bare CR is encountered in a header line, an extra space is added
6936 after the line terminator so as not to end the header. The reasoning
6937 behind this is that bare CRs in header lines are most likely either
6938 to be mistakes, or people trying to play silly games.
6940 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6941 was being incorrectly given as 18 bytes larger than it should have been.
6942 This is a VOB (very old bug).
6944 60. This may never have affected anything current, but just in case it has:
6945 When the local host is found other than at the start of a list of hosts,
6946 the local host, those with the same MX, and any that follow, are discarded.
6947 When the list in question was part of a longer list of hosts, the following
6948 hosts (not currently being processed) were also being discarded. This no
6949 longer happens. I'm not sure if this situation could ever has previously
6952 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6955 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6956 header. This is something that is recommended in a new Internet Draft, and
6957 is something that is documented as being done by Sendmail. There are two
6958 possible values. For messages generated by the autoreply transport, Exim
6961 Auto-Submitted: auto-replied
6963 whereas for all other generated messages (e.g. bounces) it adds
6965 Auto-Submitted: auto-generated
6967 63. The "personal" condition in filters now includes a test for the
6968 Auto-Submitted: header. If it contains the string "auto-" the message it
6969 not considered personal.
6971 64. Added rcpt_include_affixes as a generic transport option.
6973 65. Added queue_only_override (default true).
6975 66. Added the syslog_duplication option.
6977 67. If what should have been the first header line of a message consisted of
6978 a space followed by a colon, Exim was mis-interpreting it as a header line.
6979 It isn't of course - it is syntactically invalid and should therefore be
6980 treated as the start of the message body. The misbehaviour could have
6981 caused a number of strange effects, including loss of data in subsequent
6982 header lines, and spool format errors.
6984 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6985 client host had authenticated. This control can now be exercised by an ACL
6986 for more flexibility.
6988 69. By default, callouts do not happen when testing with -bh. There is now a
6989 variant, -bhc, which does actually run the callout code, including
6990 consulting and updating the callout cache.
6992 70. Added support for saslauthd authentication, courtesy of Alexander
6995 71. If statvfs() failed on the spool or log directories while checking their
6996 size for availability, Exim confusingly gave the error "space shortage".
6997 Furthermore, in debugging mode it crashed with a floating point exception.
6998 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6999 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
7000 serious problem, Exim now writes to the main and panic logs when this
7001 happens, with details of the failure. It then refuses to accept the
7002 incoming message, giving the message "spool directory problem" or "log
7003 directory problem" with a 421 code for SMTP messages.
7005 72. When Exim is about to re-exec itself, it ensures that the file descriptors
7006 0, 1, and 2 exist, because some OS complain for execs without them (see
7007 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
7008 descriptors. However, the code omitted to check that the open succeeded,
7009 causing mysterious errors if for some reason the permissions on /dev/null
7010 got screwed. Now Exim writes a message to the main and panic logs, and
7011 bombs out if it can't open /dev/null.
7013 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
7014 interact so that it is all more flexible. It is supposed to remain
7015 backwards compatible. Also added extra_local_interfaces.
7017 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
7018 to bomb out with an assertion failure - to the client this appears as a
7019 connection drop. This problem occurs in the part of the code that was taken
7020 from the Samba project. Fortunately, the assertion is in a very simple
7021 function, so I have fixed this by reproducing the function inline in the
7022 one place where it is called, and arranging for authentication to fail
7023 instead of killing the process with assert().
7025 75. The SPA client code was not working when the server requested OEM rather
7026 than Unicode encoding.
7028 76. Added code to make require_files with a specific uid setting more usable in
7029 the case where statting the file as root fails - usually a non-root-mounted
7030 NFS file system. When this happens and the failure is EACCES, Exim now
7031 forks a subprocess and does the per-uid checking as the relevant uid.
7033 77. Added process_log_path.
7035 78. If log_file_path was not explicitly set, a setting of check_log_space or
7036 check_log_inodes was ignored.
7038 79. If a space check for the spool or log partitions fails, the incident is now
7039 logged. Of course, in the latter case the data may get lost...
7041 80. Added the %p formatting code to string_format() so that it can be used to
7042 print addresses in debug_print(). Adjusted all the address printing in the
7043 debugging in store.c to use %p rather than %d.
7045 81. There was a concern that a line of code in smtp_in.c could overflow a
7046 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
7047 initially expressed, the concern was not well-founded, because trailing
7048 spaces are removed early. However, if the trailing spaces were followed by
7049 a NULL, they did not get removed, so the overflow was possible. Two fixes
7052 (a) I re-wrote the offending code in a cleaner fashion.
7053 (b) If an incoming SMTP command contains a NULL character, it is rejected
7056 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
7057 runs initgroups(), so that if the Exim user is in any additional groups,
7058 they will be used during message reception.
7064 The change log for 4.20 and earlier releases has been archived.