1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 TL/06 Add Experimental DMARC support using libopendmarc libraries.
49 TL/07 Fix an out of order global option causing a segfault. Reported to dev
50 mailing list by by Dmitry Isaikin.
52 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
54 JH/02 Support "G" suffix to numbers in ${if comparisons.
56 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
58 NM/01 Bugzilla 1197 - Spec typo
59 Bugzilla 1196 - Spec examples corrections
61 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
63 PP/09 Add gnutls_enable_pkcs11 option.
65 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
66 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
68 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
69 "acl {{name}{arg}...}", and optional args on acl condition
72 JH/05 Permit multiple router/transport headers_add/remove lines.
74 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
76 JH/07 Avoid using a waiting database for a single-message-only transport.
77 Performance patch from Paul Fisher. Bugzilla 1262.
79 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
82 JH/09 Add $headers_added variable, with content from use of ACL modifier
83 add_header (but not yet added to the message). Bugzilla 199.
85 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
86 Pulled from Bugzilla 817 by Wolfgang Breyha.
88 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
90 (nb: this is the same fix as in Exim 4.80.1)
92 JH/11 Add A= logging on delivery lines, and a client_set_id option on
95 JH/12 Add optional authenticated_sender logging to A= and a log_selector
98 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
100 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
101 advertise SMTP AUTH mechanism to us, instead of a generic
102 protocol violation error. Also, make Exim more robust to bad
103 data from the Dovecot auth socket.
105 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
107 When a queue runner is handling a message, Exim first routes the
108 recipient addresses, during which it prunes them based on the retry
109 hints database. After that it attempts to deliver the message to
110 any remaining recipients. It then updates the hints database using
113 So if a recipient address works intermittently, it can get repeatedly
114 deferred at routing time. The retry hints record remains fresh so the
115 address never reaches the final cutoff time.
117 This is a fairly common occurrence when a user is bumping up against
118 their storage quota. Exim had some logic in its local delivery code
119 to deal with this. However it did not apply to per-recipient defers
120 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
122 This change adds a proper retry rule check during routing so that the
123 final cutoff time is checked against the message's age. We only do
124 this check if there is an address retry record and there is not a
125 domain retry record; this implies that previous attempts to handle
126 the address had the retry_use_local_parts option turned on. We use
127 this as an approximation for the destination being like a local
128 delivery, as in LMTP.
130 I suspect this new check makes the old local delivery cutoff check
131 redundant, but I have not verified this so I left the code in place.
133 TF/02 Correct gecos expansion when From: is a prefix of the username.
135 Test 0254 submits a message to Exim with the header
139 When I ran the test suite under the user fanf2, Exim expanded
140 the header to contain my full name, whereas it should have added
141 a Resent-Sender: header. It erroneously treats any prefix of the
142 username as equal to the username.
144 This change corrects that bug.
146 GF/01 DCC debug and logging tidyup
147 Error conditions log to paniclog rather than rejectlog.
148 Debug lines prefixed by "DCC: " to remove any ambiguity.
150 TF/03 Avoid unnecessary rebuilds of lookup-related code.
152 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
153 Bug spotted by Jeremy Harris; was flawed since initial commit.
154 Would have resulted in OCSP responses post-SNI triggering an Exim
155 NULL dereference and crash.
157 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
159 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
160 Bug detection, analysis and fix by Samuel Thibault.
161 Bugzilla 1331, Debian bug #698092.
163 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
165 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
166 Server implementation by Todd Lyons, client by JH.
167 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
168 config variable "prdr_enable" controls whether the server
169 advertises the facility. If the client requests PRDR a new
170 acl_data_smtp_prdr ACL is called once for each recipient, after
171 the body content is received and before the acl_smtp_data ACL.
172 The client is controlled by bolth of: a hosts_try_prdr option
173 on the smtp transport, and the server advertisement.
174 Default client logging of deliveries and rejections involving
175 PRDR are flagged with the string "PRDR".
177 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
178 fclose(). Diagnosis by Todd Lyons.
180 PP/17 Update configure.default to handle IPv6 localhost better.
181 Patch by Alain Williams (plus minor tweaks).
184 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
185 This is now consistent with GnuTLS, and is now documented: the
186 previous undocumented portable approach to treating the option as
187 unset was to force an expansion failure. That still works, and
188 an empty string is now equivalent.
190 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
191 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
192 not performing validation itself.
194 PP/20 Added force_command boolean option to pipe transport.
195 Patch from Nick Koston, of cPanel Inc.
197 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
200 TF/04 Added udpsend ACL modifer and hexquote expansion operator
202 PP/21 Fix eximon continuous updating with timestamped log-files.
203 Broken in a format-string cleanup in 4.80, missed when I repaired the
204 other false fix of the same issue.
205 Report and fix from Heiko Schlichting.
208 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
209 Report from Prashanth Katuri.
211 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
212 It's SecureTransport, so affects any MacOS clients which use the
213 system-integrated TLS libraries, including email clients.
215 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
216 using a MIME ACL for non-SMTP local injection.
217 Report and assistance in diagnosis by Warren Baker.
219 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
221 JH/16 Fix comparisons for 64b. Bugzilla 1385.
223 TL/09 Add expansion variable $authenticated_fail_id to keep track of
224 last id that failed so it may be referenced in subsequent ACL's.
226 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
229 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
230 ldap library initialization, allowing self-signed CA's to be
231 used. Also properly sets require_cert option later in code by
232 using NULL (global ldap config) instead of ldap handle (per
233 session). Bug diagnosis and testing by alxgomz.
235 TL/12 Enhanced documentation in the ratelimit.pl script provided in
236 the src/util/ subdirectory.
238 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
239 renamed to Transport Post Delivery Action by Jeremy Harris, as
242 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
243 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
244 redis_servers = needs to be configured which will be used by the redis
245 lookup. Patch from Warren Baker, of The Packet Hub.
247 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
249 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
250 hostname or reverse DNS when processing a host list. Used suggestions
251 from multiple comments on this bug.
253 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
255 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
256 Missed a few lines, added it to make the runtest require no keyboard
259 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
260 contains upper case chars. Make router use caseful_local_part.
266 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
268 This, or similar/improved, will also be change PP/11 of 4.82.
274 PP/01 Handle short writes when writing local log-files.
275 In practice, only affects FreeBSD (8 onwards).
276 Bugzilla 1053, with thanks to Dmitry Isaikin.
278 NM/01 Bugzilla 949 - Documentation tweak
280 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
283 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
285 PP/02 Implemented gsasl authenticator.
287 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
289 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
290 `pkg-config foo` for cflags/libs.
292 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
293 with rest of GSASL and with heimdal_gssapi.
295 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
296 `pkg-config foo` for cflags/libs for the TLS implementation.
298 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
299 properties get this fed in as external SSF. A number of robustness
300 and debugging improvements to the cyrus_sasl authenticator.
302 PP/08 cyrus_sasl server now expands the server_realm option.
304 PP/09 Bugzilla 1214 - Log authentication information in reject log.
305 Patch by Jeremy Harris.
307 PP/10 Added dbmjz lookup type.
309 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
311 PP/12 MAIL args handles TAB as well as SP, for better interop with
312 non-compliant senders.
313 Analysis and variant patch by Todd Lyons.
315 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
316 Bug report from Lars Müller <lars@samba.org> (via SUSE),
317 Patch from Dirk Mueller <dmueller@suse.com>
319 PP/13 tls_peerdn now print-escaped for spool files.
320 Observed some $tls_peerdn in wild which contained \n, which resulted
321 in spool file corruption.
323 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
324 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
325 or write after TLS renegotiation, which otherwise led to messages
328 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
329 as a tracking header (ie: a signed header comes before the signature).
330 Patch from Wolfgang Breyha.
332 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
333 comma-sep list; embedded commas doubled.
335 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
337 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
339 Report and patch from Dmitry Banschikov.
341 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
342 Removed SSL_clear() after SSL_new() which led to protocol negotiation
343 failures. We appear to now support TLS1.1+ with Exim.
345 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
346 lets Exim select keys and certificates based upon TLS SNI from client.
347 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
348 before an outbound SMTP session. New log_selector, +tls_sni.
350 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
351 NULL dereference. Report and patch from Alun Jones.
353 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
354 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
355 Not seeing resolver debug output on NetBSD, but suspect this is a
356 resolver implementation change.
358 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
359 Left warnings. Added "eximon gdb" invocation mode.
361 PP/21 Defaulting "accept_8bitmime" to true, not false.
363 PP/22 Added -bw for inetd wait mode support.
365 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
366 locate the relevant includes and libraries. Made this the default.
368 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
369 Bugzilla 1246, report and most of solution from Tomasz Kusy.
371 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
372 This may cause build issues on older platforms.
374 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
375 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
376 gnutls_require_mac & gnutls_require_protocols (no longer supported).
377 Added SNI support via GnuTLS too.
378 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
380 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
382 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
383 Applied second patch from Janne, implementing suggestion to default
384 multiple-strings-in-record handling to match SPF spec.
386 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
388 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
389 read-only, out of scope).
390 Patch from Wolfgang Breyha, report from Stuart Northfield.
392 PP/29 Fix three issues highlighted by clang analyser static analysis.
393 Only crash-plausible issue would require the Cambridge-specific
394 iplookup router and a misconfiguration.
395 Report from Marcin Mirosław.
397 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
399 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
400 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
401 As part of this, removing so much warning spew let me fix some minor
402 real issues in debug logging.
404 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
405 assignment on my part. Fixed.
407 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
408 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
409 Janne Snabb (who went above and beyond: thank you).
411 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
412 string otherwise requires a connection and a bunch more work and it's
413 relatively easy to get wrong. Should also expose TLS library linkage
416 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
417 64-bit ${eval} (JH/03).
419 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
420 GNU libc to support some of the 64-bit stuff, should not lead to
421 conflicts. Defined before os.h is pulled in, so if a given platform
422 needs to override this, it can.
424 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
425 protection layer was required, which is not implemented.
426 Bugzilla 1254, patch from Wolfgang Breyha.
428 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
429 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
430 tls_dhparam take prime identifiers. Also unbreak combination of
431 OpenSSL+DH_params+TLSSNI.
433 PP/39 Disable SSLv2 by default in OpenSSL support.
439 PP/01 Solaris build fix for Oracle's LDAP libraries.
440 Bugzilla 1109, patch from Stephen Usher.
442 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
444 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
447 TF/02 Fix a couple more cases where we did not log the error message
448 when unlink() failed. See also change 4.74-TF/03.
450 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
451 lock up or crash if it happened to be inside a call to libc when it
452 got a SIGUSR1 from exiwhat.
454 The SIGUSR1 handler appends the current process status to the process
455 log which is later printed by exiwhat. It used to use the general
456 purpose logging code to do this, but several functions it calls are
457 not safe for signals.
459 The new output code in the SIGUSR1 handler is specific to the process
460 log, and simple enough that it's easy to inspect for signal safety.
461 Removing some special cases also simplifies the general logging code.
462 Removing the spurious timestamps from the process log simplifies
465 TF/04 Improved ratelimit ACL condition.
467 The /noupdate option has been deprecated in favour of /readonly which
468 has clearer semantics. The /leaky, /strict, and /readonly update modes
469 are mutually exclusive. The update mode is no longer included in the
470 database key; it just determines when the database is updated. (This
471 means that when you upgrde Exim will forget old rate measurements.)
473 Exim now checks that the per_* options are used with an update mode that
474 makes sense for the current ACL. For example, when Exim is processing a
475 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
476 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
477 must specify per_mail/readonly. If you omit the update mode it defaults to
478 /leaky where that makes sense (as before) or /readonly where required.
480 The /noupdate option is now undocumented but still supported for
481 backwards compatibility. It is equivalent to /readonly except that in
482 ACLs where /readonly is required you may specify /leaky/noupdate or
483 /strict/noupdate which are treated the same as /readonly.
485 A useful new feature is the /count= option. This is a generalization
486 of the per_byte option, so that you can measure the throughput of other
487 aggregate values. For example, the per_byte option is now equivalent
488 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
490 The per_rcpt option has been generalized using the /count= mechanism
491 (though it's more complicated than the per_byte equivalence). When it is
492 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
493 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
494 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
495 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
496 Note that using per_rcpt with a non-readonly update mode in more than
497 one ACL will cause the recipients to be double-counted. (The per_mail
498 and per_byte options don't have this problem.)
500 The handling of very low rates has changed slightly. If the computed rate
501 is less than the event's count (usually one) then this event is the first
502 after a long gap. In this case the rate is set to the same as this event's
503 count, so that the first message of a spam run is counted properly.
505 The major new feature is a mechanism for counting the rate of unique
506 events. The new per_addr option counts the number of different
507 recipients that someone has sent messages to in the last time period. It
508 behaves like per_rcpt if all the recipient addresses are different, but
509 duplicate recipient addresses do not increase the measured rate. Like
510 the /count= option this is a general mechanism, so the per_addr option
511 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
512 example, measure the rate that a client uses different sender addresses
513 with the options per_mail/unique=$sender_address. There are further
514 details in the main documentation.
516 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
518 TF/06 Removed a few PCRE remnants.
520 TF/07 Automatically extract Exim's version number from tags in the git
521 repository when doing development or release builds.
523 PP/02 Raise smtp_cmd_buffer_size to 16kB.
524 Bugzilla 879. Patch from Paul Fisher.
526 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
527 Heavily based on revision 40f9a89a from Simon Arlott's tree.
530 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
532 PP/05 Variable $av_failed, true if the AV scanner deferred.
533 Bugzilla 1078. Patch from John Horne.
535 PP/06 Stop make process more reliably on build failure.
536 Bugzilla 1087. Patch from Heiko Schlittermann.
538 PP/07 Make maildir_use_size_file an _expandable_ boolean.
539 Bugzilla 1089. Patch from Heiko Schlittermann.
541 PP/08 Handle ${run} returning more data than OS pipe buffer size.
542 Bugzilla 1131. Patch from Holger Weiß.
544 PP/09 Handle IPv6 addresses with SPF.
545 Bugzilla 860. Patch from Wolfgang Breyha.
547 PP/10 GnuTLS: support TLS 1.2 & 1.1.
549 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
552 PP/11 match_* no longer expand right-hand-side by default.
553 New compile-time build option, EXPAND_LISTMATCH_RHS.
554 New expansion conditions, "inlist", "inlisti".
556 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
558 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
560 PP/14 fix log_write() format string regression from TF/03.
561 Bugzilla 1152. Patch from Dmitry Isaikin.
567 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
569 PP/02 Harmonised TLS library version reporting; only show if debugging.
570 Layout now matches that introduced for other libraries in 4.74 PP/03.
572 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
574 PP/04 New "dns_use_edns0" global option.
576 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
579 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
580 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
582 TK/01 Updated PolarSSL code to 0.14.2.
583 Bugzilla 1097. Patch from Andreas Metzler.
585 PP/07 Catch divide-by-zero in ${eval:...}.
588 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
591 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
592 format-string attack -- SECURITY: remote arbitrary code execution.
594 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
595 time unintentionally subject to list matching rules, letting the header
596 cause arbitrary Exim lookups (of items which can occur in lists, *not*
597 arbitrary string expansion). This allowed for information disclosure.
599 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
600 INT_MIN/-1 -- value coerced to INT_MAX.
606 NM/01 Workround for PCRE version dependency in version reporting
609 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
610 This fixes portability to compilers other than gcc, notably
611 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
613 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
614 makefiles for portability to HP-UX and POSIX correctness.
616 PP/01 Permit LOOKUP_foo enabling on the make command-line.
617 Also via indented variable definition in the Makefile.
618 (Debugging by Oliver Heesakkers).
620 PP/02 Restore caching of spamd results with expanded spamd_address.
621 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
623 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
624 Improves build reliability. Fix from: Frank Elsner
626 NM/02 Fix wide character breakage in the rfc2047 coding
627 Fixes bug 1064. Patch from Andrey N. Oktyabrski
629 NM/03 Allow underscore in dnslist lookups
630 Fixes bug 1026. Patch from Graeme Fowler
632 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
633 Code patches from Adam Ciarcinski of NetBSD.
635 NM/04 Fixed exiqgrep to cope with mailq missing size issue
638 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
639 is logged, to avoid truncation. Patch from John Horne.
641 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
642 Patch from Jakob Hirsch.
644 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
645 SQL string expansion failure details.
646 Patch from Andrey Oktyabrski.
648 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
649 Patch from Simon Arlott.
651 PP/09 New lookups functionality failed to compile on old gcc which rejects
652 extern declarations in function scope.
653 Patch from Oliver Fleischmann
655 PP/10 Use sig_atomic_t for flags set from signal handlers.
656 Check getgroups() return and improve debugging.
657 Fixed developed for diagnosis in bug 927 (which turned out to be
660 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
661 Patch from Mark Zealey.
663 PP/12 Bugzilla 1056: Improved spamd server selection.
664 Patch from Mark Zealey.
666 PP/13 Bugzilla 1086: Deal with maildir quota file races.
667 Based on patch from Heiko Schlittermann.
669 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
670 Patch from Uwe Doering, sign-off by Michael Haardt.
672 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
673 variable declaration deep within a block. Bug and patch from
676 PP/15 lookups-Makefile IRIX compatibilty coercion.
678 PP/16 Make DISABLE_DKIM build knob functional.
680 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
681 Patch by Simon Arlott
683 TF/03 Fix valgrind.h portability to C89 compilers that do not support
684 variable argument macros. Our copy now differs from upstream.
690 TF/01 Failure to get a lock on a hints database can have serious
691 consequences so log it to the panic log.
693 TF/02 Log LMTP confirmation messages in the same way as SMTP,
694 controlled using the smtp_confirmation log selector.
696 TF/03 Include the error message when we fail to unlink a spool file.
698 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
699 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
700 for maintaining out-of-tree patches for some time.
702 PP/01 Bugzilla 139: Documentation and portability issues.
703 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
704 Handle per-OS dynamic-module compilation flags.
706 PP/02 Let /dev/null have normal permissions.
707 The 4.73 fixes were a little too stringent and complained about the
708 permissions on /dev/null. Exempt it from some checks.
709 Reported by Andreas M. Kirchwitz.
711 PP/03 Report version information for many libraries, including
712 Exim version information for dynamically loaded libraries. Created
713 version.h, now support a version extension string for distributors
714 who patch heavily. Dynamic module ABI change.
716 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
717 privilege escalation vulnerability whereby the Exim run-time user
718 can cause root to append content of the attacker's choosing to
721 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
724 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
725 If dropping privileges for untrusted macros, we disabled normal logging
726 on the basis that it would fail; for the Exim run-time user, this is not
727 the case, and it resulted in successful deliveries going unlogged.
728 Fixed. Reported by Andreas Metzler.
734 PP/01 Date: & Message-Id: revert to normally being appended to a message,
735 only prepend for the Resent-* case. Fixes regression introduced in
736 Exim 4.70 by NM/22 for Bugzilla 607.
738 PP/02 Include check_rfc2047_length in configure.default because we're seeing
739 increasing numbers of administrators be bitten by this.
741 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
743 PP/03 Bugzilla 994: added openssl_options main configuration option.
745 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
747 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
749 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
751 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
752 without a peer certificate, leading to a segfault because of an
753 assumption that peers always have certificates. Be a little more
754 paranoid. Problem reported by Martin Tscholak.
756 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
757 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
758 NB: ClamAV planning to remove STREAM in "middle of 2010".
759 CL also introduces -bmalware, various -d+acl logging additions and
760 more caution in buffer sizes.
762 PP/09 Implemented reverse_ip expansion operator.
764 PP/10 Bugzilla 937: provide a "debug" ACL control.
766 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
768 PP/12 Bugzilla 973: Implement --version.
770 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
772 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
774 PP/15 Bugzilla 816: support multiple condition rules on Routers.
776 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
777 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
778 ignore trailing whitespace.
780 JJ/02 prevent non-panic DKIM error from being sent to paniclog
782 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
785 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
786 Notification from Dr Andrew Aitchison.
788 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
789 ExtendedDetectionInfo response format.
790 Notification from John Horne.
792 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
795 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
796 XSL and documented dependency on system catalogs, with examples of how
799 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
802 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
803 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
804 configuration file which is writeable by the Exim user or group.
806 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
807 of configuration files to cover files specified with the -C option if
808 they are going to be used with root privileges, not just the default
811 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
812 option (effectively making it always true).
814 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
815 files to be used while preserving root privileges.
817 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
818 that rogue child processes cannot use them.
820 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
821 run-time user, instead of root.
823 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
824 Exim run-time user without dropping privileges.
826 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
827 result string, instead of calling string_vformat() twice with the same
830 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
831 for other users. Others should always drop root privileges if they use
832 -C on the command line, even for a whitelisted configure file.
834 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
836 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
842 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
843 $data_path, and $header_path variables; fixed documentation bugs and
846 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
847 exipick to access non-standard spools, including the "frozen" queue
850 NM/01 Bugzilla 965: Support mysql stored procedures.
851 Patch from Alain Williams
853 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
855 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
856 Patch from Andreas Metzler
858 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
859 Patch from Kirill Miazine
861 NM/05 Bugzilla 671: Added umask to procmail example.
863 JJ/03 installed exipick 20100323.0, fixing doc bug
865 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
866 directory. Notification and patch from Dan Rosenberg.
868 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
870 TK/02 Improve log output when DKIM signing operation fails.
872 MH/01 Treat the transport option dkim_domain as a colon separated
873 list, not as a single string, and sign the message with each element,
874 omitting multiple occurences of the same signer.
876 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
877 Bugzilla 985, 986. Patch by Simon Arlott
879 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
880 Patch by Simon Arlott
882 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
883 MBX locking. Notification from Dan Rosenberg.
889 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
891 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
893 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
895 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
897 NM/04 Bugzilla 915: Flag broken perl installation during build.
903 TK/01 Added patch by Johannes Berg that expands the main option
904 "spamd_address" if it starts with a dollar sign.
906 TK/02 Write list of recipients to X-Envelope-Sender header when building
907 the mbox-format spool file for content scanning (suggested by Jakob
910 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
911 (http://www.dcc-servers.net/) support via dccifd. Activated by
912 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
914 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
915 by Mark Daniel Reidel <mr@df.eu>.
917 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
918 When building exim an external PCRE library is now needed -
919 PCRE is a system library on the majority of modern systems.
920 See entry on PCRE_LIBS in EDITME file.
922 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
923 conversation. Added nologin parameter to request.
924 Patch contributed by Kirill Miazine.
926 TF/01 Do not log submission mode rewrites if they do not change the address.
928 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
930 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
931 log files in place. Contributed by Roberto Lima.
933 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
935 TF/03 Bugzilla 615: When checking the local_parts router precondition
936 after a local_part_suffix or local_part_prefix option, Exim now
937 does not use the address's named list lookup cache, since this
938 contains cached lookups for the whole local part.
940 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
941 Robert Millan. Documentation is in experimental-spec.txt.
943 TF/04 Bugzilla 668: Fix parallel build (make -j).
945 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
947 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
948 Patch provided by Jan Srzednicki.
950 TF/05 Leading white space used to be stripped from $spam_report which
951 wrecked the formatting. Now it is preserved.
953 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
954 that they are available at delivery time.
956 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
958 TF/08 TLS error reporting now respects the incoming_interface and
959 incoming_port log selectors.
961 TF/09 Produce a more useful error message if an SMTP transport's hosts
962 setting expands to an empty string.
964 NM/06 Bugzilla 744: EXPN did not work under TLS.
965 Patch provided by Phil Pennock.
967 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
968 Patch provided by Richard Godbee.
970 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
971 acl_smtp_notquit, added index entry.
973 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
974 Patch provided by Eugene Bujak.
976 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
977 accept(). Patch provided by Maxim Dounin.
979 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
980 Patch provided by Phil Pennock.
982 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
984 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
985 Patch provided by Brad "anomie" Jorsch.
987 NM/14 Bugzilla 622: Added timeout setting to transport filter.
988 Patch provided by Dean Brooks.
990 TK/05 Add native DKIM support (does not depend on external libraries).
992 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
993 Patch provided by Graeme Fowler.
995 NM/16 Bugzilla 851: Documentation example syntax fix.
997 NM/17 Changed NOTICE file to remove references to embedded PCRE.
999 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1002 NM/19 Bugzilla 745: TLS version reporting.
1003 Patch provided by Phil Pennock.
1005 NM/20 Bugzilla 167: bool: condition support.
1006 Patch provided by Phil Pennock.
1008 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1009 clients. Patch provided by Phil Pennock.
1011 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1012 Patch provided by Brad "anomie" Jorsch.
1014 NM/23 Bugzilla 687: Fix misparses in eximstats.
1015 Patch provided by Heiko Schlittermann.
1017 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1018 Patch provided by Heiko Schlittermann.
1020 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1021 plus update to original patch.
1023 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1025 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1026 Patch provided by David Brownlee.
1028 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1030 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1032 NM/30 Bugzilla 888: TLS documentation bugfixes.
1034 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1036 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1037 Unlike the original bugzilla I have changed all shell scripts in src tree.
1039 NM/33 Bugzilla 898: Transport filter timeout fix.
1040 Patch by Todd Rinaldo.
1042 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1043 Patch by Serge Demonchaux.
1045 NM/35 Bugzilla 39: Base64 decode bug fixes.
1046 Patch by Jakob Hirsch.
1048 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1050 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1052 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1054 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1060 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1061 ALT-N's libdkim that I have put here:
1062 http://duncanthrax.net/exim-experimental/
1064 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1065 were called 'true' and it seems that C99 defines that as a reserved
1066 keyword to be used with 'bool' variable types. That means you could
1067 not include C99-style headers which use bools without triggering
1068 build errors in sieve.c.
1070 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1071 as mailq or other aliases. Changed the --help handling significantly
1072 to do whats expected. exim_usage() emits usage/help information.
1074 SC/01 Added the -bylocaldomain option to eximstats.
1076 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1078 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1080 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1086 PH/01 Another patch from the Sieve maintainer.
1088 PH/02 When an IPv6 address is converted to a string for single-key lookup
1089 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1090 dots are used instead of colons so that keys in lsearch files need not
1091 contain colons. This was done some time before quoting was made available
1092 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1093 (notated using the quote facility) so as to distinguish them from IPv4
1094 keys. This meant that lookups for IP addresses in host lists did not work
1095 for iplsearch lookups.
1097 This has been fixed by arranging for IPv6 addresses to be expressed with
1098 colons if the lookup type is iplsearch. This is not incompatible, because
1099 previously such lookups could never work.
1101 The situation is now rather anomolous, since one *can* have colons in
1102 ordinary lsearch keys. However, making the change in all cases is
1103 incompatible and would probably break a number of configurations.
1105 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1108 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1109 conversion specification without a maximum field width, thereby enabling
1110 a rogue spamd server to cause a buffer overflow. While nobody in their
1111 right mind would setup Exim to query an untrusted spamd server, an
1112 attacker that gains access to a server running spamd could potentially
1113 exploit this vulnerability to run arbitrary code as the Exim user.
1115 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1116 $primary_hostname instead of what libspf2 thinks the hosts name is.
1118 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1119 a directory entry by the name of the lookup key. Previously, if a
1120 symlink pointed to a non-existing file or a file in a directory that
1121 Exim lacked permissions to read, a lookup for a key matching that
1122 symlink would fail. Now it is enough that a matching directory entry
1123 exists, symlink or not. (Bugzilla 503.)
1125 PH/03 The body_linecount and body_zerocount variables are now exported in the
1128 PH/04 Added the $dnslist_matched variable.
1130 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1131 This means they are set thereafter only if the connection becomes
1134 PH/06 Added the client_condition to authenticators so that some can be skipped
1135 by clients under certain conditions.
1137 PH/07 The error message for a badly-placed control=no_multiline_responses left
1138 "_responses" off the end of the name.
1140 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1142 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1143 (without spaces) instead of just copying the configuration text.
1145 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1147 PH/11 Added $max_received_linelength.
1149 PH/12 Added +ignore_defer and +include_defer to host lists.
1151 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1152 way in which PCRE > 7.0 is built.
1154 PH/14 Implemented queue_only_load_latch.
1156 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1157 MAIL command. The effect was to mangle the value on 64-bit systems.
1159 PH/16 Another patch from the Sieve maintainer.
1161 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1163 PH/18 If a system quota error occurred while trying to create the file for
1164 a maildir delivery, the message "Mailbox is full" was not appended to the
1165 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1166 only to a quota excession during the actual writing of the file.
1168 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1169 characters?) which causes problems in log lines. The DN values are now
1170 passed through string_printing() before being added to log lines.
1172 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1173 and InterBase are left for another time.)
1175 PH/21 Added message_body_newlines option.
1177 PH/22 Guard against possible overflow in moan_check_errorcopy().
1179 PH/23 POSIX allows open() to be a macro; guard against that.
1181 PH/24 If the recipient of an error message contained an @ in the local part
1182 (suitably quoted, of course), incorrect values were put in $domain and
1183 $local_part during the evaluation of errors_copy.
1189 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1190 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1193 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1194 issue a MAIL command.
1196 PH/02 In an ACL statement such as
1198 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1200 if a client was not listed at all, or was listed with a value other than
1201 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1202 the condition was not true (as it should be), so access was not denied.
1203 The bug was that the ! inversion was incorrectly passed on to the second
1204 item. This has been fixed.
1206 PH/03 Added additional dnslists conditions == and =& which are different from
1207 = and & when the dns lookup returns more than one IP address.
1209 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1210 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1212 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1213 FSYNC, which compiles an option called disable_fsync that allows for
1214 bypassing fsync(). The documentation is heavily laced with warnings.
1216 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1218 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1219 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1220 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1221 including adding "make clean"; (3) Added -fPIC when compiling the test
1222 dynamically loaded module, to get rid of a warning.
1224 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1225 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1226 = 0s. The bug is otherwise harmless.
1228 PH/07 There was a bug in the dovecot authenticator such that the value of
1229 $auth1 could be overwritten, and so not correctly preserved, after a
1230 successful authentication. This usually meant that the value preserved by
1231 the server_setid option was incorrect.
1233 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1235 PH/09 Installed PCRE release 7.0.
1237 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1238 run for batched SMTP input. It is now run at the start of every message
1239 in the batch. While fixing this I discovered that the process information
1240 (output by running exiwhat) was not always getting set for -bs and -bS
1241 input. This is fixed, and it now also says "batched" for BSMTP.
1243 PH/11 Added control=no_pipelining.
1245 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1246 patch, slightly modified), and move the expansion of helo_data till after
1247 the connection is made in the smtp transport (so it can use these
1250 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1252 PH/14 Added log_selector = +pid.
1254 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1256 PH/16 Add ${if forany and ${if forall.
1258 PH/17 Added dsn_from option to vary the From: line in DSNs.
1260 PH/18 Flush SMTP output before performing a callout, unless control =
1261 no_callout_flush is set.
1263 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1264 was true (the default) a successful delivery failed to delete the retry
1265 item, thus causing premature timeout of the address. The bug is now
1268 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1270 PH/21 Long custom messages for fakedefer and fakereject are now split up
1271 into multiline reponses in the same way that messages for "deny" and
1272 other ACL rejections are.
1274 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1275 with slight modification.
1277 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1278 draft, changing the syntax and factoring some duplicate code".
1280 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1281 for deliveries of the second and subsequent messages over the same SMTP
1284 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1285 ${reduce, with only minor "tidies".
1287 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1289 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1290 expansion side effects.
1292 PH/27 When a message times out after an over-quota error from an Exim-imposed
1293 quota, the bounce message says "mailbox is full". This message was not
1294 being given when it was a system quota that was exceeded. It now should
1297 MH/03 Made $recipients available in local_scan(). local_scan() already has
1298 better access to the recipient list through recipients_list[], but
1299 $recipients can be useful in postmaster-provided expansion strings.
1301 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1302 in the case of a MAIL command with additional options following the
1303 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1304 were accidentally chopped off.
1306 PH/29 SMTP synchronization checks are implemented when a command is read -
1307 there is a check that no more input is waiting when there shouldn't be
1308 any. However, for some commands, a delay in an ACL can mean that it is
1309 some time before the response is written. In this time, more input might
1310 arrive, invalidly. So now there are extra checks after an ACL has run for
1311 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1312 pipelining has not been advertised.
1314 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1316 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1317 returned in the SMTP response when smtp_return_error_details was false.
1318 This has been fixed.
1320 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1321 instead of the C I/O that was originally supplied, because problems were
1322 reported on Solaris.
1324 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1325 Exim which did not show up earlier: it was assuming that a call to
1326 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1327 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1328 was a macro that became an assignment, so it seemed to work. This has
1329 changed to a proper function call with a void return, hence the compile
1330 error. Exim's code has been fixed.
1332 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1335 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1337 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1339 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1340 because it turns out to be too big or there is a timeout) there is no
1341 "Completed" line in the log. When some messages of this type were
1342 selected by exigrep, they were listed as "not completed". Others were
1343 picked up by some special patterns. I have improved the selection
1344 criteria to be more general.
1346 PH/38 The host_find_failed option in the manualroute router can now be set
1347 to "ignore", to completely ignore a host whose IP address cannot be
1348 found. If all hosts are ignored, the behaviour is controlled by the new
1349 host_all_ignored option.
1351 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1352 homing or because of multiple MX records with /mx) generated more than
1353 one IP address, and the following item turned out to be the local host,
1354 all the secondary addresses of the first item were incorrectly removed
1355 from the list, along with the local host and any following hosts (which
1356 is what is supposed to happen).
1358 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1359 whoever called Exim into the -H file. In the case of the daemon it was
1360 behaving confusingly. When first started, it used values for whoever
1361 started the daemon, but after a SIGHUP it used the Exim user (because it
1362 calls itself on a restart). I have changed the code so that it now always
1365 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1366 message are rejected with the same error (e.g. no authentication or bad
1367 sender address), and a DATA command is nevertheless sent (as can happen
1368 with PIPELINING or a stupid MUA), the error message that was given to the
1369 RCPT commands is included in the rejection of the DATA command. This is
1370 intended to be helpful for MUAs that show only the final error to their
1373 PH/42 Another patch from the Sieve maintainer.
1375 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1376 Eximstats - Fixed some broken HTML links and added missing column headers
1378 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1379 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1381 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1383 PH/43 Yet another patch from the Sieve maintainer.
1385 PH/44 I found a way to check for a TCP/IP connection going away before sending
1386 the response to the final '.' that terminates a message, but only in the
1387 case where the client has not sent further data following the '.'
1388 (unfortunately, this is allowed). However, in many cases there won't be
1389 any further data because there won't be any more messages to send. A call
1390 to select() can be used: if it shows that the input is "ready", there is
1391 either input waiting, or the socket has been closed. An attempt to read
1392 the next input character can distinguish the two cases. Previously, Exim
1393 would have sent an OK response which the client would never have see.
1394 This could lead to message repetition. This fix should cure that, at
1395 least in a lot of common cases.
1397 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1398 advertised in response to EHLO.
1404 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1405 fixed by 4.65/MH/01 (is this a record?) are fixed:
1407 (i) An empty string was always treated as zero by the numeric comparison
1408 operators. This behaviour has been restored.
1410 (ii) It is documented that the numeric comparison operators always treat
1411 their arguments as decimal numbers. This was broken in that numbers
1412 starting with 0 were being interpreted as octal.
1414 While fixing these problems I realized that there was another issue that
1415 hadn't been noticed. Values of message_size_limit (both the global option
1416 and the transport option) were treated as octal if they started with 0.
1417 The documentation was vague. These values are now always treated as
1418 decimal, and I will make that clear in the documentation.
1424 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1425 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1428 MH/01 Don't check that the operands of numeric comparison operators are
1429 integers when their expansion is in "skipping" mode (fixes bug
1430 introduced by 4.64-PH/07).
1432 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1433 child addresses, Exim now panics and dies. Previously, because the count
1434 is held in a short int, deliveries were likely to be lost. As such a
1435 large number of recipients for a single message is ridiculous
1436 (performance will be very, very poor), I have chosen to impose a limit
1437 rather than extend the field.
1443 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1444 leftover -K file (the existence of which was triggered by #402).
1445 While we were at it, introduced process PID as part of the -K
1446 filename. This should rule out race conditions when creating
1449 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1450 processing considerably. Previous code took too long for large mails,
1451 triggering a timeout which in turn triggers #401.
1453 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1454 in the DK code in transports.c. sendfile() is not really portable,
1455 hence the _LINUX specificness.
1457 TF/01 In the add_headers option to the mail command in an Exim filter,
1458 there was a bug that Exim would claim a syntax error in any
1459 header after the first one which had an odd number of characters
1462 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1463 callout verification, Exim cached a "reject" for the entire domain. This
1464 is correct for most verifications, but it is not correct for a recipient
1465 verification with use_sender or use_postmaster set, because in that case
1466 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1467 case of MAIL FROM:<> rejection from other early rejections (e.g.
1468 rejection of HELO). When verifying a recipient using a non-null MAIL
1469 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1470 Whatever the result of the callout, the value of the domain cache is
1471 left unchanged (for any other kind of callout, getting as far as trying
1472 RCPT means that the domain itself is ok).
1474 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1477 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1478 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1479 session, the daemon ignores ECONNECT errors and logs others; it now
1480 ignores EPIPE as well.
1482 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1483 (quoted-printable decoding).
1485 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1486 later the small subsequent patch to fix an introduced bug.
1488 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1490 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1492 PH/08 An error is now given if message_size_limit is specified negative.
1494 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1495 to be given (somewhat) arbitrary names.
1497 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1500 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1501 miscellaneous code fixes
1503 PH/10 Added the log_reject_target ACL modifier to specify where to log
1506 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1507 hostname. This is wrong, because it relates to the incoming message (and
1508 probably the interface on which it is arriving) and not to the outgoing
1509 callout (which could be using a different interface). This has been
1510 changed to use the value of the helo_data option from the smtp transport
1511 instead - this is what is used when a message is actually being sent. If
1512 there is no remote transport (possible with a router that sets up host
1513 addresses), $smtp_active_hostname is used.
1515 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1516 tweaks were necessary in order to get it to work (see also 21 below):
1517 (a) The code assumed that strncpy() returns a negative number on buffer
1518 overflow, which isn't the case. Replaced with Exim's string_format()
1520 (b) There were several signed/unsigned issues. I just did the minimum
1521 hacking in of casts. There is scope for a larger refactoring.
1522 (c) The code used strcasecmp() which is not a standard C function.
1523 Replaced with Exim's strcmpic() function.
1524 (d) The code set only $1; it now sets $auth1 as well.
1525 (e) A simple test gave the error "authentication client didn't specify
1526 service in request". It would seem that Dovecot has changed its
1527 interface. Fortunately there's a specification; I followed it and
1528 changed what the client sends and it appears to be working now.
1530 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1533 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1534 address A is aliased to B and C, where B exists and C does not. Without
1535 -v the output is "A verified" because verification stops after a
1536 successful redirection if more than one address is generated. However,
1537 with -v the child addresses are also verified. Exim was outputting "A
1538 failed to verify" and then showing the successful verification for C,
1539 with its parentage. It now outputs "B failed to verify", showing B's
1540 parentage before showing the successful verification of C.
1542 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1543 look up a TXT record in a specific list after matching in a combined
1546 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1547 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1548 they consult the DNS. I had assumed they would set it the way they
1549 wanted; and indeed my experiments on Linux seem to show that in some
1550 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1551 To be on the safe side, however, I have now made the interface to
1552 host_find_byname() similar to host_find_bydns(), with an argument
1553 containing the DNS resolver options. The host_find_byname() function now
1554 sets these options at its start, just as host_find_bydns() does. The smtp
1555 transport options dns_qualify_single and dns_search_parents are passed to
1556 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1557 of host_find_byname() use the default settings of RES_DEFNAMES
1558 (qualify_single) but not RES_DNSRCH (search_parents).
1560 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1561 spool_read_header() do less string testing, by means of a preliminary
1562 switch on the second character of optional "-foo" lines. (This is
1563 overdue, caused by the large number of possibilities that now exist.
1564 Originally there were few.) While I was there, I also converted the
1565 str(n)cmp tests so they don't re-test the leading "-" and the first
1566 character, in the hope this might squeeze out yet more improvement.
1568 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1569 flag allowing group syntax was set by the header_syntax check but not
1570 turned off, possible causing trouble later; (2) The flag was not being
1571 set at all for the header_verify test, causing "group"-style headers to
1572 be rejected. I have now set it in this case, and also caused header_
1573 verify to ignore an empty address taken from a group. While doing this, I
1574 came across some other cases where the code for allowing group syntax
1575 while scanning a header line wasn't quite right (mostly, not resetting
1576 the flag correctly in the right place). These bugs could have caused
1577 trouble for malformed header lines. I hope it is now all correct.
1579 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1580 with the "reply" argument non-NULL. The code, however (which originally
1581 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1582 but it didn't always do it. This confused somebody who was copying the
1583 code for some other use. I have removed all the tests.
1585 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1586 feature that was used to support insecure browsers during the U.S. crypto
1587 embargo. It requires special client support, and Exim is probably the
1588 only MTA that supported it -- and would never use it because real RSA is
1589 always available. This code has been removed, because it had the bad
1590 effect of slowing Exim down by computing (never used) parameters for the
1591 RSA_EXPORT functionality.
1593 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1594 authenticator to fail if there's a tab character in the incoming data
1595 (there should never be unless someone is messing about, as it's supposed
1596 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1597 if the connection is using TLS or if the remote IP is the same as the
1598 local IP, and the "valid-client-cert option" if a client certificate has
1601 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1602 authenticators. This can be used for authorization after authentication
1603 succeeds. (In the case of plaintext, it servers for both authentication
1606 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1607 if any retry times were supplied.
1609 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1610 connection, where there is no client IP address to check. In this
1611 situation, the verify now always succeeds.
1613 PH/25 Applied John Jetmore's -Mset patch.
1615 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1617 PH/27 In a string expansion for a processed (not raw) header when multiple
1618 headers of the same name were present, leading whitespace was being
1619 removed from all of them, but trailing whitespace was being removed only
1620 from the last one. Now trailing whitespace is removed from each header
1621 before concatenation. Completely empty headers in a concatenation (as
1622 before) are ignored.
1624 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1625 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1627 PH/29 [Removed. This was a change that I later backed out, and forgot to
1628 correct the ChangeLog entry (that I had efficiently created) before
1629 committing the later change.]
1631 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1632 address errors (4xx response to RCPT) over the same connection as other
1633 messages routed to the same hosts. Such deliveries are always "forced",
1634 so retry times are not inspected. This resulted in far too many retries
1635 for the affected addresses. The effect occurred only when there were more
1636 hosts than the hosts_max_try setting in the smtp transport when it had
1637 the 4xx errors. Those hosts that it had tried were not added to the list
1638 of hosts for which the message was waiting, so if all were tried, there
1639 was no problem. Two fixes have been applied:
1641 (i) If there are any address or message errors in an SMTP delivery, none
1642 of the hosts (tried or untried) are now added to the list of hosts
1643 for which the message is waiting, so the message should not be a
1644 candidate for sending over the same connection that was used for a
1645 successful delivery of some other message. This seems entirely
1646 reasonable: after all the message is NOT "waiting for some host".
1647 This is so "obvious" that I'm not sure why it wasn't done
1648 previously. Hope I haven't missed anything, but it can't do any
1649 harm, as the worst effect is to miss an optimization.
1651 (ii) If, despite (i), such a delivery is accidentally attempted, the
1652 routing retry time is respected, so at least it doesn't keep
1653 hammering the server.
1655 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1656 in ${readsocket because some servers need this prod.
1658 PH/32 Added some extra debug output when updating a wait-xxx database.
1660 PH/33 The hint "could be header name not terminated by colon", which has been
1661 given for certain expansion errors for a long time, was not being given
1662 for the ${if def:h_colon_omitted{... case.
1664 PH/34 The spec says: "With one important exception, whenever a domain list is
1665 being scanned, $domain contains the subject domain." There was at least
1666 one case where this was not true.
1668 PH/35 The error "getsockname() failed: connection reset by peer" was being
1669 written to the panic log as well as the main log, but it isn't really
1670 panic-worthy as it just means the connection died rather early on. I have
1671 removed the panic log writing for the ECONNRESET error when getsockname()
1674 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1675 runs only) independently of the message's sender address. This meant
1676 that, if the 4xx error was in fact related to the sender, a different
1677 message to the same recipient with a different sender could confuse
1678 things. In particualar, this can happen when sending to a greylisting
1679 server, but other circumstances could also provoke similar problems.
1680 I have changed the default so that the retry time for these errors is now
1681 based a combination of the sender and recipient addresses. This change
1682 can be overridden by setting address_retry_include_sender=false in the
1685 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1686 remote server are returned as part of bounce messages. This was not
1687 happening for LMTP over a pipe (the lmtp transport), but now it is the
1688 same for both kinds of LMTP.
1690 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1691 in header lines that were in fact CNAMEs. This is no longer the case.
1693 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1694 and queue runs started by the daemon processed all messages. This has
1695 been fixed so that -R and -S can now usefully be given with -q<time>.
1697 PH/40 Import PCRE release 6.7 (fixes some bugs).
1699 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1701 PH/42 Give an error if -q is specified more than once.
1703 PH/43 Renamed the variables $interface_address and $interface_port as
1704 $received_ip_address and $received_port, to make it clear that these
1705 values apply to message reception, and not to the outgoing interface when
1706 a message is delivered. (The old names remain recognized, of course.)
1708 PH/44 There was no timeout on the connect() call when using a Unix domain
1709 socket in the ${readsocket expansion. There now is.
1711 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1712 be meaningful with "accept".
1714 SC/01 Eximstats V1.43
1715 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1717 SC/02 Eximstats V1.44
1718 Use a glob alias rather than an array ref in the generated
1719 parser. This improves both readability and performance.
1721 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1722 Collect SpamAssassin and rejection statistics.
1723 Don't display local sender or destination tables unless
1724 there is data to show.
1725 Added average volumes into the top table text output.
1727 SC/04 Eximstats V1.46
1728 Collect data on the number of addresses (recipients)
1729 as well as the number of messages.
1731 SC/05 Eximstats V1.47
1732 Added 'Message too big' to the list of mail rejection
1733 reasons (thanks to Marco Gaiarin).
1735 SC/06 Eximstats V1.48
1736 Mainlog lines which have GMT offsets and are too short to
1737 have a flag are now skipped.
1739 SC/07 Eximstats V1.49 (Alain Williams)
1740 Added the -emptyok flag.
1742 SC/08 Eximstats V1.50
1743 Fixes for obtaining the IP address from reject messages.
1745 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1746 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1747 whitesspace changes from 4.64-PH/27
1749 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1752 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1753 are found, allow negative numbers in numeric criteria)
1755 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1757 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1758 to match changes made in 4.64-PH/43
1760 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1762 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1763 "domain" is actually an IP address, and force a failure. This locks out
1764 those revolvers/nameservers that support "A-for-A" lookups, in
1765 contravention of the specifications.
1767 PH/48 When a host name was looked up from an IP address, and the subsequent
1768 forward lookup of the name timed out, the host name was left in
1769 $sender_host_name, contrary to the specification.
1771 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1772 restricted to single-key lookups, Exim was not diagnosing an error if
1773 * or *@ was used with a query-style lookup.
1775 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1777 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1778 long ago, but noone interested enough thought of it. Let's just say that
1779 the "1.1" means that there are some new functions that weren't there at
1780 some point in the past.
1782 PH/51 Error processing for expansion failure of helo_data from an smtp
1783 transport during callout processing was broken.
1785 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1786 tested/used via the -bh/-bhc/-bs options.
1788 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1789 bug, fixed in subsequent PCRE releases).
1791 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1792 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1794 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1800 SC/01 Use a glob alias rather than an array ref in eximstats generated
1801 parser. This improves both readability and performance.
1803 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1804 Don't display local sender or destination tables in eximstats unless
1805 there is data to show.
1806 Added average volumes into the eximstats top table text output.
1808 SC/03 Collect data on the number of addresses (recipients) as well
1809 as the number of messages in eximstats.
1811 TF/01 Correct an error in the documentation for the redirect router. Exim
1812 does (usually) call initgroups() when daemonizing.
1814 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1815 with consistent privilege compared to when running as a daemon.
1817 TF/03 Note in the spec that $authenticated_id is not set for local
1818 submissions from trusted users.
1820 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1821 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1823 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1824 by adding some example configuration directives to the default
1825 configuration file. A little bit of work is required to uncomment the
1826 directives and define how usernames and passwords are checked, but
1827 there is now a framework to start from.
1829 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1830 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1831 without this. I don't know how relevant this is to other LDAP libraries.
1833 PH/02 Add the verb name to the "unknown ACL verb" error.
1835 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1837 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1839 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1840 directory not expanded when it should be if an expanded home directory
1841 was set for the address (which is overridden by the transport).
1843 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1846 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1847 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1848 because it is too late at that time, and has no effect.
1850 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1851 security issue with \' (bugzilla #107). I could not use the
1852 PQescapeStringConn() function, because it needs a PGconn value as one of
1855 PH/08 When testing addresses using -bt, indicate those final addresses that
1856 are duplicates that would not cause an additional delivery. At least one
1857 person was confused, thinking that -bt output corresponded to deliveries.
1858 (Suppressing duplicates isn't a good idea as you lose the information
1859 about possibly different redirections that led to the duplicates.)
1861 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1862 systems where poll() doesn't work, in particular OS X.
1864 PH/10 Added more information to debugging output for retry time not reached.
1866 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1867 operations in malware.c.
1869 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1872 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1873 syslog_duplication set false, log lines that would normally be written
1874 both the the main log and to the reject log were not written to syslog at
1877 PH/14 In the default configuration, change the use of "message" in ACL warn
1878 statements to "add_header".
1880 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1881 not followed by a command (e.g. "seen endif").
1883 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1884 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1887 PH/17 Added extra conditions to the default value of delay_warning_condition
1891 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1892 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1893 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1896 The Auto-Submitted: and various List- headers are standardised, whereas I
1897 don't think Precedence: ever was.
1899 PH/18 Refactored debugging code in route_finduser() to show more information,
1900 in particular, the error code if getpwnam() issues one.
1902 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1903 This is apparently needed in addition to the PH/07 change above to avoid
1904 any possible encoding problems.
1906 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1907 but not after initializing Perl.
1909 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1910 output them only if debugging. By default they are written stderr,
1911 apparently, which is not desirable.
1913 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1916 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1919 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1921 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1922 authenticated or an ident call has been made. Suppress the default
1923 values for $authenticated_id and $authenticated_sender (but permit -oMai
1924 and -oMas) when testing with -bh.
1926 PH/24 Re-jigged the order of the tests in the default configuration so that the
1927 tests for valid domains and recipients precede the DNS black list and CSA
1928 tests, on the grounds that those ones are more expensive.
1930 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1931 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1932 This bug exists in every version of Exim that I still have, right back to
1935 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1936 However, an attempt to turn on case-sensitivity in a regex key by
1937 including (?-i) didn't work because the subject string was already
1938 lowercased, and the effects were non-intuitive. It turns out that a
1939 one-line patch can be used to allow (?-i) to work as expected.
1945 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1946 other effects) broke the use of negated acl sub-conditions.
1948 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1951 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1952 "Deny" causes Exim to reject the incoming connection with a 554 error.
1953 Unfortunately, if there is a major crisis, such as a disk failure,
1954 tcp-wrappers gives "deny", whereas what one would like would be some
1955 kind of temporary error. A kludge has been added to help with this.
1956 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1957 554 error is used if errno is still zero or contains ENOENT (which occurs
1958 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1961 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1963 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1964 errors. Otherwise a message that provokes a temporary error (when other
1965 messages do not) can cause a whole host to time out.
1967 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1968 addresses were routed directly to files or pipes from a redirect router.
1969 File deliveries just didn't batch; pipe deliveries might have suffered
1972 PH/06 A failure to get a lock for a hints database would erroneously always say
1973 "Failed to get write lock", even when it was really a read lock.
1975 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1976 of 0600. This has been changed to use the value of the lockfile_mode
1977 option (which defaults to 0600).
1979 PH/08 Applied small patch from the Sieve maintainer.
1981 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1982 folder from quota calculations, a direct delivery into this folder messed
1983 up the contents of the maildirsize file. This was because the regex was
1984 used only to exclude .Trash (or whatever) when the size of the mailbox
1985 was calculated. There was no check that a delivery was happening into an
1986 excluded directory. This bug has been fixed by ignoring all quota
1987 processing for deliveries into excluded directories.
1989 PH/10 Added the maildirfolder_create_regex option to appendfile.
1995 PH/01 The code for finding all the local interface addresses on a FreeBSD
1996 system running IPv6 was broken. This may well have applied to all BSD
1997 systems, as well as to others that have similar system calls. The broken
1998 code found IPv4 interfaces correctly, but gave incorrect values for the
1999 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2000 that it would not match correctly against @[] and not recognize the IPv6
2003 PH/02 The ipliteral router was not recognizing addresses of the form user@
2004 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2006 PH/03 Added disable_ipv6.
2008 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2009 decoded form, because it is most often used to construct To: headers
2010 lines in autoreplies, and the decoded form may well be syntactically
2011 invalid. However, $reply_address has leading white space removed, and all
2012 newlines turned into spaces so that the autoreply transport does not
2015 PH/05 If group was specified without a user on a router, and no group or user
2016 was specified on a transport, the group from the router was ignored.
2018 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2019 for visible compile-time settings that can be used to change these
2020 numbers, for those that want even more. Backwards compatibility with old
2021 spool files has been maintained. However, going back to a previous Exim
2022 release will lost any variables that are in spool files.
2024 PH/07 Two small changes when running in the test harness: increase delay when
2025 passing a TCP/IP connection to a new process, in case the original
2026 process has to generate a bounce, and remove special handling of
2027 127.0.0.2 (sic), which is no longer necessary.
2029 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2030 be the same on different OS.
2032 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2035 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2036 whether --show-vars was specified or not
2038 JJ/02 exipick: Added support for new ACL variable spool format introduced
2041 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2042 syntactically invalid From: or Reply-to: line, and a filter used this to
2043 generate an autoreply, and therefore failed to obtain an address for the
2044 autoreply, Exim could try to deliver to a non-existent relative file
2045 name, causing unrelated and misleading errors. What now happens is that
2046 it logs this as a hard delivery error, but does not attempt to create a
2049 PH/11 The exinext utility has a -C option for testing purposes, but although
2050 the given file was scanned by exinext itself; it wasn't being passed on
2051 when Exim was called.
2053 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2054 an end-of-file indication when reading a command response.
2056 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2057 compiled. In many other places in Exim, IPv6 addresses are always
2058 recognized, so I have changed this. It also means that IPv4 domain
2059 literals of the form [IPV4:n.n.n.n] are now always recognized.
2061 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2062 used if the router is not running as root, for example, when verifying at
2063 ACL time, or when using -bh. The debugging output from this situation was
2064 non-existent - all you got was a failure to exec. I have made two
2067 (a) Failures to set uid/gid, the current directory, or a process leader
2068 in a subprocess such as that created by queryprogram now generate
2069 suitable debugging ouput when -d is set.
2071 (b) The queryprogram router detects when it is not running as root,
2072 outputs suitable debugging information if -d is set, and then runs
2073 the subprocess without attempting to change uid/gid.
2075 PH/15 Minor change to Makefile for building test_host (undocumented testing
2078 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2079 additional section of a DNS packet that returns MX or SRV records.
2080 Instead, it always explicitly searches for A/AAAA records. This avoids
2081 major problems that occur when a DNS server includes only records of one
2082 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2083 fixed another bug: if SRV records were looked up and the corresponding
2084 address records were *not* found in the additional section, the port
2085 values from the SRV records were lost.
2087 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2088 using the correct key (the original address) when searching the retry
2089 rules in order to find which one to use for generating the retry hint.
2091 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2092 adding the default one. Similarly, if it contains a Reply-To: header, the
2093 errors_reply_to option, if set, is not used.
2095 PH/19 When calculating a retry time, Exim used to measure the "time since
2096 failure" by looking at the "first failed" field in the retry record. Now
2097 it does not use this if it is later than than the arrival time of the
2098 message. Instead it uses the arrival time. This makes for better
2099 behaviour in cases where some deliveries succeed, thus re-setting the
2100 "first failed" field. An example is a quota failure for a huge message
2101 when small messages continue to be delivered. Without this change, the
2102 "time since failure" will always be short, possible causing more frequent
2103 delivery attempts for the huge message than are intended.
2104 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2106 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2107 $1, $2, $3) because the numerical variables can be reset during some
2108 expansion items (e.g. "match"), thereby losing the authentication data.
2110 PH/21 Make -bV show the size of off_t variables so that the test suite can
2111 decide whether to run tests for quotas > 2G.
2113 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2114 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2115 filecount value is greater than 2G or if a quota value is greater than 2G
2116 on a system where the size of off_t is not greater than 4, a panic error
2119 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2120 never match. The debug and -bh output now contains an explicit error
2121 message indicating a malformed IPv4 address or mask.
2123 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2124 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2125 PH/23 above applies.
2127 PH/25 Do not write to syslog when running in the test harness. The only
2128 occasion when this arises is a failure to open the main or panic logs
2129 (for which there is an explicit test).
2131 PH/26 Added the /no_tell option to "control=freeze".
2133 PH/27 If a host name lookup failed very early in a connection, for example, if
2134 the IP address matched host_lookup and the reverse lookup yielded a name
2135 that did not have a forward lookup, an error message of the form "no IP
2136 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2137 could be logged. Now it outputs the IP address instead of "NULL".
2139 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2140 allows the sender and the authenticated sender to be set when
2141 submitting a message from within Exim. Since child_open_exim() is
2142 documented for local_scan(), the new function should be too.
2144 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2145 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2146 results in an empty string is now treated as unset.
2148 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2150 PH/31 Added sender_verify_fail logging option.
2152 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2153 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2154 tidied the source and removed it altogether.
2156 PH/33 When a queue run was abandoned because the load average was too high, a
2157 log line was always written; now it is written only if the queue_run log
2158 selector is set. In addition, the log line for abandonment now contains
2159 information about the queue run such as the pid. This is always present
2160 in "start" and "stop" lines but was omitted from the "abandon" line.
2162 PH/34 Omit spaces between a header name and the colon in the error message that
2163 is given when verify = headers_syntax fails (if there are lots of them,
2164 the message gets confusing).
2166 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2167 names, as there are now some PTR records that contain slashes. This check
2168 is only to protect against broken name servers that fall over on strange
2169 characters, so the fact that it applies to all lookups doesn't matter.
2171 PH/36 Now that the new test suite is complete, we can remove some of the
2172 special code in Exim that was needed for the old test suite. For example,
2173 sorting DNS records because real resolvers return them in an arbitrary
2174 order. The new test suite's fake resolver always returns records in the
2177 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2178 bounces) except when queue_only is set, to avoid logging races between
2179 the different processes.
2181 PH/38 Panic-die if .include specifies a non-absolute path.
2183 PH/39 A tweak to the "H" retry rule from its user.
2185 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2186 a label. They prevented compilation on older perls.
2188 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2189 a warning to be raised on newish perls.
2191 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2192 on queue. Changes to match documented behaviour of showing count of
2193 messages matching specified criteria.
2195 PH/40 Changed the default ident timeout from 30s to 5s.
2197 PH/41 Added support for the use of login_cap features, on those BSD systems
2198 that have them, for controlling the resources used by pipe deliveries.
2200 PH/42 The content-scanning code uses fopen() to create files in which to put
2201 message data. Previously it was not paying any attention to the mode of
2202 the files. Exim runs with umask(0) because the rest of the code creates
2203 files with open(), and sets the required mode explicitly. Thus, these
2204 files were ending up world-writeable. This was not a big issue, because,
2205 being within the spool directory, they were not world-accessible. I have
2206 created a function called modefopen, which takes an additional mode
2207 argument. It sets umask(777), creates the file, chmods it to the required
2208 mode, then resets the umask. All the relevant calls to fopen() in the
2209 content scanning code have been changed to use this function.
2211 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2212 to 24 hours. This avoids potential overflow problems when processing G
2213 and H retry rules. I suspect nobody ever tinkers with this value.
2215 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2217 PH/45 When the plaintext authenticator is running as a client, the server's
2218 challenges are checked to ensure they are valid base64 strings. By
2219 default, the authentication attempt is cancelled if an invalid string is
2220 received. Setting client_ignore_invalid_base64 true ignores these errors.
2221 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2222 they are received. Thus, the responses can be made to depend on the
2223 challenges. If an invalid string is ignored, an empty string is placed in
2226 PH/46 Messages that are created by the autoreply transport now contains a
2227 References: header, in accordance with RFCs 2822 and 3834.
2229 PH/47 Added authenticated_sender_force to the smtp transport.
2231 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2233 PH/49 Installed latest patch from the Sieve maintainer.
2235 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2236 was also set, the appendfile transport was unnecessarily scanning a
2237 directory of message files (e.g. for maildir delivery) to find the count
2238 of files (along with the size), even though it did not need this
2239 information. It now does the scan only if it needs to find either the
2240 size of the count of files.
2242 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2244 PH/52 Two bugs concerned with error handling when the smtp transport is
2247 (i) Exim was not creating retry information for temporary errors given
2248 for individual recipients after the DATA command when the smtp transport
2249 was used in LMTP mode. This meant that they could be retried too
2250 frequently, and not timed out correctly.
2252 (ii) Exim was setting the flag that allows error details to be returned
2253 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2254 recipients that were returned after the DATA command.
2256 PH/53 This is related to PH/52, but is more general: for any failing address,
2257 when detailed error information was permitted to be returned to the
2258 sender, but the error was temporary, then after the final timeout, only
2259 "retry timeout exceeded" was returned. Now it returns the full error as
2260 well as "retry timeout exceeded".
2262 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2263 do this, and (what is worse) MTAs that accept it.
2265 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2266 will now be deprecated.
2268 PH/56 New os.c-cygwin from the Cygwin maintainer.
2270 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2271 formats (previously only available in exim formats via -bpr, -bpru,
2272 and -bpra. Now also available in native and exiqgrep formats)
2274 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2275 with very large, slow to parse queues
2277 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2279 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2281 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2282 responses to authentication challenges, though it was showing the
2283 challenges; (ii) I've removed the CR characters from the debug output for
2286 PH/58 Allow for the insertion of a newline as well as a space when a string
2287 is turned into more than one encoded-word during RFC 2047 encoding. The
2288 Sieve code now uses this.
2290 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2291 data_4xx, lost_connection, tls_required.
2293 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2294 message was being sent as an SMTP response.
2296 PH/61 Add -l and -k options to exicyclog.
2298 PH/62 When verifying, if an address was redirected to one new address, so that
2299 verification continued, and the new address failed or deferred after
2300 having set something in $address_data, the value of $address_data was not
2301 passed back to the ACL. This was different to the case when no
2302 redirection occurred. The value is now passed back in both cases.
2304 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2305 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2306 use login_cap.h, so on its own it isn't the distinguishing feature. The
2307 new name refers directly to the setclassresources() function.
2309 PH/65 Added configuration files for NetBSD3.
2311 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2313 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2314 is preferred over IPv4.
2316 PH/68 The bounce_return_message and bounce_return_body options were not being
2317 honoured for bounces generated during the reception of non-SMTP messages.
2318 In particular, this applied to messages rejected by the ACL. This bug has
2319 been fixed. However, if bounce_return_message is true and bounce_return_
2320 body is false, the headers that are returned for a non-SMTP message
2321 include only those that have been read before the error was detected.
2322 (In the case of an ACL rejection, they have all been read.)
2324 PH/69 The HTML version of the specification is now built in a directory called
2325 spec_html instead of spec.html, because the latter looks like a path with
2326 a MIME-type, and this confuses some software.
2328 PH/70 Catch two compiler warnings in sieve.c.
2330 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2331 function verify_get_ident() calls ip_connect() to connect a socket, but
2332 if the "connect()" function timed out, ip_connect() used to close the
2333 socket. However, verify_get_ident() also closes the socket later, and in
2334 between Exim writes to the log, which may get opened at this point. When
2335 the socket was closed in ip_connect(), the log could get the same file
2336 descriptor number as the socket. This naturally causes chaos. The fix is
2337 not to close the socket in ip_connect(); the socket should be closed by
2338 the function that creates it. There was only one place in the code where
2339 this was missing, in the iplookup router, which I don't think anybody now
2340 uses, but I've fixed it anyway.
2342 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2343 well as to direct DNS lookups. Otherwise the handling of names in host
2344 lists is inconsistent and therefore confusing.
2350 PH/01 Two changes to the default runtime configuration:
2352 (1) Move the checks for relay_from_hosts and authenticated clients from
2353 after to before the (commented out) DNS black list checks.
2355 (2) Add control=submission to the relay_from_hosts and authenticated
2356 clients checks, on the grounds that messages accepted by these
2357 statements are most likely to be submissions.
2359 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2361 (1) Generate an error if the third argument for the ${prvs expansion is
2364 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2367 (3) Reset the variables that are obtained from the first argument of
2368 ${prvscheck and used in the second argument before leaving the code,
2369 because their memory is reclaimed, so using them afterwards may do
2372 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2373 one (it's much easier than Tom thought :-).
2375 (5) Because of (4), we can now allow for the use of $prvscheck_result
2376 inside the third argument.
2378 PH/03 For some reason, the default setting of PATH when running a command from
2379 a pipe transport was just "/usr/bin". I have changed it to
2382 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2383 anything to be listed in the output from -bV.
2385 PH/05 When a filter generated an autoreply, the entire To: header line was
2386 quoted in the delivery log line, like this:
2388 => >A.N.Other <ano@some.domain> <original@ddress> ...
2390 This has been changed so that it extracts the operative address. There
2391 may be more than one such address. If so, they are comma-separated, like
2394 => >ano@some.domain,ona@other.domain <original@ddress> ...
2396 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2397 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2398 not being looked up in the rDNS to get a host name, Exim was showing the
2399 IP address twice in Received: lines, even though the IP addresses were
2400 identical. For example:
2402 Received: from [1.2.3.4] (helo=[1.2.3.4])
2404 However, if the real host name was known, it was omitting the HELO data
2405 if it matched the actual IP address. This has been tidied up so that it
2406 doesn't show the same IP address twice.
2408 PH/07 When both +timestamp and +memory debugging was on, the value given by
2409 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2410 called by the debug printing, thereby overwriting the timestamp buffer.
2411 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2413 PH/08 When the original message was included in an autoreply transport, it
2414 always said "this is a copy of the message, including all the headers",
2415 even if body_only or headers_only was set. It now gives an appropriate
2418 PH/09 Applied a patch from the Sieve maintainer which:
2420 o fixes some comments
2421 o adds the (disabled) notify extension core
2422 o adds some debug output for the result of if/elsif tests
2423 o points to the current vacation draft in the documentation
2424 and documents the missing references header update
2428 o fixes a bug in processing the envelope test (when testing
2429 multiple envelope elements, the last element determinted the
2432 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2433 Electronic Mail") by including:
2435 Auto-submitted: auto-generated
2437 in the messages that it generates (bounce messages and others, such as
2438 warnings). In the case of bounce messages for non-SMTP mesages, there was
2439 also a typo: it was using "Auto_submitted" (underscore instead of
2440 hyphen). Since every message generated by Exim is necessarily in response
2441 to another message, thes have all been changed to:
2443 Auto-Submitted: auto-replied
2445 in accordance with these statements in the RFC:
2447 The auto-replied keyword:
2449 - SHOULD be used on messages sent in direct response to another
2450 message by an automatic process,
2452 - MUST NOT be used on manually-generated messages,
2454 - MAY be used on Delivery Status Notifications (DSNs) and Message
2455 Disposition Notifications (MDNs),
2457 - MUST NOT be used on messages generated by automatic or periodic
2458 processes, except for messages which are automatic responses to
2461 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2462 to the default Received: header definition.
2464 PH/12 Added log selector acl_warn_skipped (default on).
2466 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2467 variables because (a) they are in the wrong storage pool and (b) even if
2468 they were copied, it wouldn't work properly because of the caching.
2470 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2471 checking when decoding. Apparently there are clients that generate
2472 overlong encoded strings. Why am I not surprised?
2474 PH/15 If the first argument of "${if match_address" was not empty, but did not
2475 contain an "@" character, Exim crashed. Now it writes a panic log message
2476 and treats the condition as false.
2478 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2480 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2481 extension "envelope-auth". The code is finished and in agreement with
2482 other implementations, but there is no documentation so far and in fact,
2483 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2484 not changing the active code.
2486 Print executed "if" and "elsif" statements when debugging is used. This
2487 helps a great deal to understand what a filter does.
2489 Document more things not specified clearly in RFC3028. I had all this
2490 sorted out, when out of a sudden new issues came to my mind. Oops."
2492 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2495 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2496 canonical form (as documented). However, after a host name lookup from
2497 the IP address, check_host() was doing a simple string comparison with
2498 addresses acquired from the DNS when checking that the found name did
2499 have the original IP as one of its addresses. Since any found IPv6
2500 addresses are likely to be in abbreviated form, the comparison could
2501 fail. Luckily, there already exists a function for doing the comparison
2502 by converting both addresses to binary, so now that is used instead of
2503 the text comparison.
2505 PH/20 There was another similar case to PH/19, when a complete host name was
2506 given in a host list; looking up its IP address could give an abbreviated
2507 form, whereas the current host's name might or might not be abbreviated.
2508 The same fix has been applied.
2514 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2515 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2518 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2519 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2521 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2523 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2524 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2525 Also, for "Auto-Submitted", treat anything other than "no" as
2526 non-personal, in accordance with RFC 3834. (Previously it treated
2527 anything starting "auto-" as non-personal.)
2529 TF/01 The control=submission/name=... option had a problem with syntax
2530 errors if the name included a slash character. The /name= option
2531 now slurps the rest of the string, so it can include any characters
2532 but it must come last in the list of options (after /sender_retain
2535 PH/05 Some modifications to the interface to the fake nameserver for the new
2543 TK/01 Added the "success_on_redirect" address verification option. See
2544 NewStuff for rationale and an example.
2546 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2548 PH/02 Patch to exigrep to allow it to work on syslog lines.
2550 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2551 fread() to skip over the body file's header line, because in Cygwin the
2552 header line is locked and is inaccessible.
2554 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2555 co-exist for some time) to make it clear that it is the Exim ID that is
2556 referenced, not the Message-ID: header line.
2558 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2559 string_format() function, because snprintf() does not exist on all
2562 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2563 ${stat: expansion item.
2565 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2566 protocol synchronization error", to keep the pedants happy.
2568 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2569 well as for IRIX systems, when gcc is being used. See the host.c source
2572 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2574 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2577 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2578 transport and to the smtp transport in LMTP mode.
2580 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2582 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2583 run a filter in a subprocess. This could lead to confusion in subsequent
2584 lookups in the parent process. There should also be a search_tidyup() at
2585 the end of the subprocess.
2587 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2588 only if the host matched helo_try_verify_hosts, which caused the
2589 verification to occur when the EHLO/HELO command was issued. The ACL just
2590 tested the remembered result. Now, if a previous verification attempt has
2591 not happened, "verify = helo" does it there and then.
2593 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2595 TK/03 Fix log output including CR from clamd.
2597 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2598 exist provoked a memory error which could cause a segfault.
2600 PH/15 Installed PCRE 6.2
2602 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2604 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2605 of the problem. Specifically, suggested +O2 rather than +O1 for the
2608 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2610 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2611 record, the port was not correctly logged when the outgoing_port log
2612 selector was set (it logged the transort's default port).
2614 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2615 fallback_hosts, and "hosts" in the smtp transport.
2617 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2618 host errors such as "Connection refused".
2620 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2621 authentication with radiusclient 0.4.9:
2623 - Error returned from rc_read_config was caught wrongly
2624 - Username/password not passed on to radius server due to wrong length.
2626 The presumption is that some radiusclient API changes for 4.51/PH/17
2627 were not taken care of correctly. The code is still untested by me (my
2628 Linux distribution still has 0.3.2 of radiusclient), but it was
2629 contributed by a Radius user.
2631 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2632 expanding the "port" option of the smtp transport.
2634 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2635 while reading a MIME header. Thanks to Tom Hughes for a patch.
2637 PH/24 Include config.h inside local_scan.h so that configuration settings are
2640 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2641 that in an ACL for RCPT (for example), you can examine exactly what was
2644 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2645 commands, but it was not correctly comparing the address with the actual
2646 client host address. Thus, it would show the EHLO address in Received:
2647 header lines when this was not necessary.
2649 PH/27 Added the % operator to ${eval:}.
2651 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2652 it should be ignoring failures (because with -C, for example, it has lost
2653 privilege). It wasn't ignoring creation failures other than "already
2656 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2659 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2660 because an input line was too long, either on its own, or by virtue of
2661 too many continuations, the temporary file was not being removed, and the
2662 return code was incorrect.
2664 PH/31 Missing "BOOL" in function definition in filtertest.c.
2666 PH/32 Applied Sieve patches from the maintainer.
2668 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2670 PH/33 Added "verify = not_blind".
2672 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2673 Local/Makefile (with some defaults set). These are used in built scripts
2674 such as exicyclog, but they have never been used in the exim_install
2675 script (though there are many overriding facilities there). I have
2676 arranged that the exim_install script now takes note of these two
2679 PH/35 Installed configuration files for Dragonfly.
2681 PH/36 When a locally submitted message by a trusted user did not contain a
2682 From: header, and the sender address was obtained from -f or from an SMTP
2683 MAIL command, and the trusted user did not use -F to supply a sender
2684 name, $originator_name was incorrectly used when constructing a From:
2685 header. Furthermore, $originator_name was used for submission mode
2686 messages from external hosts without From: headers in a similar way,
2687 which is clearly wrong.
2689 PH/37 Added control=suppress_local_fixups.
2691 PH/38 When log_selector = +received_sender was set, and the addition of the
2692 sender made the log line's construction buffer exactly full, or one byte
2693 less than full, an overflow happened when the terminating "\n" was
2696 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2697 when the result of a list match is failure because a DNS lookup failed.
2699 PH/40 RM_COMMAND is now used in the building process.
2701 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2702 the "build-* directories that it finds.
2704 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2705 address in a domain literal was a prefix of an interface address.
2707 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2708 when verifying a sender address, unless rewrite_headers is false.
2710 PH/44 Wrote a long comment about why errors_to addresses are verified as
2711 recipients, not senders.
2713 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2714 the ratelimit ACL was added.
2716 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2718 PH/46 Added extra information about PostgreSQL errors to the error string.
2720 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2721 avoiding the need to install special zones in a real server. This is
2722 backwards compatible; if it can't find the fake resolver, it drops back.
2723 Thus, both old and new test suites can be run.
2725 TF/02 Added util/ratelimit.pl
2727 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2728 clock is set back in time.
2730 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2731 Candler <B.Candler@pobox.com>.
2733 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2734 OK for addresses that are the result of redirections.
2736 PH/48 A number of further additions for the benefit of the new test suite,
2737 including a fake gethostbyname() that interfaces to the fake DNS resolver
2740 TF/06 The fix for widen_domains has also been applied to qualify_single and
2741 search_parents which are the other dnslookup options that can cause
2744 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2747 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2749 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2750 service" if it encountered an MX record with an empty target hostname.
2751 The message is now "an MX or SRV record indicated no SMTP service".
2753 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2754 if the DNS of the sending site is misconfigured. This is quite a
2755 common situation. This change restores the behaviour of treating a
2756 helo verification defer as a failure.
2758 PH/51 If self=fail was set on a router, the bounce message did not include the
2759 actual error message.
2765 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2767 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2768 command itself ended in error, the underlying message about the transport
2769 filter timeout was being overwritten with the pipe command error. Now the
2770 underlying error message should be appended to the second error message.
2772 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2774 PH/02 Reduce the amount of output that "make" produces by default. Full output
2775 can still be requested.
2777 PH/03 The warning log line about a condition test deferring for a "warn" verb
2778 was being output only once per connection, rather than after each
2779 occurrence (because it was using the same function as for successful
2780 "warn" verbs). This seems wrong, so I have changed it.
2782 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2783 it should not have, which might have caused a crash in the right
2784 circumstances, but probably never did.
2786 PH/04 Installed a modified version of Tony Finch's patch to make submission
2787 mode fix the return path as well as the Sender: header line, and to
2788 add a /name= option so that you can make the user's friendly name appear
2791 TF/03 Added the control = fakedefer ACL modifier.
2793 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2794 Mark Lowes for thorough testing.
2796 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2798 TK/03 Merged latest SRS patch from Miles Wilton.
2800 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2801 with the definition in sysexits.h (which is #included earlier).
2802 Fortunately, Exim does not actually use EX_OK. The code used to try to
2803 preserve the sysexits.h value, by assumimg that macro definitions were
2804 scanned for macro replacements. I have been disabused of this notion,
2805 so now the code just undefines EX_OK before #including unistd.h.
2807 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2808 in the smtp transport. When a block could not be written in a single
2809 write() function, the timeout was being re-applied to each part-write.
2810 This seems wrong - if the receiver was accepting one byte at a time it
2811 would take for ever. The timeout is now adjusted when this happens. It
2812 doesn't have to be particularly precise.
2814 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2815 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2817 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2818 without a domain if the check to <postmaster@domain> fails.
2820 SC/01 Eximstats: added -xls and the ability to specify output files
2821 (patch written by Frank Heydlauf).
2823 SC/02 Eximstats: use FileHandles for outputing results.
2825 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2827 SC/04 Eximstats: fixed display of large numbers with -nvr option
2829 SC/05 Eximstats: fixed merging of reports with empty tables.
2831 SC/06 Eximstats: added the -include_original_destination flag
2833 SC/07 Eximstats: removed tabs and trailing whitespace.
2835 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2837 TK/06 MBOX spool code: Add real "From " MBOX separator line
2838 so the .eml file is really in mbox format (even though
2839 most programs do not really care). Patch from Alex Miller.
2841 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2842 The latter is generated from $received_to and is only set if the
2843 message has one envelope recipient. SA can use these headers,
2844 obviously out-of-the-box. Patch from Alex Miller.
2846 PH/08 The ${def test on a variable was returning false if the variable's
2847 value was "0", contrary to what the specification has always said!
2848 The result should be true unless the variable is empty.
2850 PH/09 The syntax error of a character other than { following "${if
2851 def:variable_name" (after optional whitespace) was not being diagnosed.
2852 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2853 accidental colon was present, for example, could give incorrect results.
2855 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2856 result is used (not including appendfile, where other changes are about
2859 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2860 This involved changing a lot of size variables from int to off_t. It
2861 should work with maildirs and everything.
2863 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2864 spamd dying while we are connected to it.
2866 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2869 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2870 function for BDB 4.3.
2872 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2874 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2875 the sources for PCRE have been re-arranged and more files are now
2878 PH/15 The code I had for printing potentially long long variables in PH/11
2879 above was not the best (it lost precision). The length of off_t variables
2880 is now inspected at build time, and an appropriate printing format (%ld
2881 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2882 to be "long long int" or "long int". This is needed for the internal
2883 formatting function string_vformat().
2885 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2886 the configuration file to be ":syslog", then the script "guesses" where
2887 the logs files are, rather than using the compiled in default. In our
2888 case the guess is not the same as the compiled default, so the script
2889 suddenly stopped working when I started to use syslog. The patch checks
2890 to see if log_file_path is "". If so, it attempts to read it from exim
2891 with no configuration file to get the compiled in version, before it
2892 falls back to the previous guessing code."
2894 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2895 implementing BATV in an Exim configuration. See NewStuff for the gory
2898 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2899 Makefile that are specific to HP-UX.
2901 PH/18 If the "use_postmaster" option was set for a recipient callout together
2902 with the "random" option, the postmaster address was used as the MAIL
2903 FROM address for the random test, but not for the subsequent recipient
2904 test. It is now used for both.
2906 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2907 patch removes a few documentation additions to RFC 3028, because the
2908 latest draft now contains them. It adds the new en;ascii-case comparator
2909 and a new error check for 8bit text in MIME parts. Comparator and
2910 require names are now matched exactly. I enabled the subaddress
2911 extension, but it is not well tested yet (read: it works for me)."
2913 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2914 rework some of the code of TK/09 above to avoid the hardwired use of
2915 "%lld" and "long long". Replaced the call to snprintf() with a call to
2918 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2919 records point to non-existent hosts", "retry timeout exceeded", and
2920 "retry time not reached for any host after a long failure period".
2922 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2923 experimental DomainKeys support:
2925 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2926 (2) On an error such as an illegally used "control", the wrong name for
2927 the control was given.
2929 These problems did NOT occur unless DomainKeys support was compiled.
2931 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2933 PH/24 Added ${if match_ip condition.
2935 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2936 databases so that it will be absolutely obvious if a crash occurs in the
2937 DB library. This is a regular occurrence (often caused by mis-matched
2940 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2941 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2942 picked up on a user's system that detects such things. There doesn't seem
2943 to be a gcc warning option for this - only an attribute that has to be
2944 put on the function's prototype. It seems that in Fedora Core 4 they have
2945 set this on a number of new functions. No doubt there will be more in due
2948 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2949 specify a transport. However, if an address that was verified by such a
2950 router was the subject of a callout, Exim crashed because it tried to
2951 read the rcpt_include_affixes from the non-existent transport. Now it
2952 just assumes that the setting of that option is false. This bug was
2953 introduced by 4.51/PH/31.
2955 PH/28 Changed -d+all to exclude +memory, because that information is very
2956 rarely of interest, but it makes the output a lot bigger. People tend to
2957 do -d+all out of habit.
2959 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2960 code in os-type was giving problems when libc.so lives in lib64, like on
2963 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2964 aren't the modern standard, and it seems that some systems' include files
2965 don't always have them. Exim was already checking for some of the newer
2966 ones like T_AAAA, and defining it itself. I've added checks for all the
2967 record types that Exim uses.
2969 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2970 not automatically generating a new one, as it is supposed to. This
2971 prevented TLS from working. If the file did exist, but contained invalid
2972 data, a new version was generated, as expected. It was only the case of a
2973 non-existent file that was broken.
2975 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2976 with a change in libdomainkeys > 0.64.
2978 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2979 from DNS. If the selector record carries the flag, it now has
2980 precedence over the domain-wide flag.
2982 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2984 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2985 the use of a port name that isn't defined in /etc/services) occurred, the
2986 message was deferred as in a normal delivery, and thus remained on the
2987 spool, instead of being failed because of the mua_wrapper setting. This
2988 is now fixed, and I tidied up some of the mua_wrapper messages at the
2991 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2992 the messages in a hash of arrays rather than using individual hashes.
2993 This is a bit cleaner and results in dramatic memory savings, albeit
2994 at a slight CPU cost.
2996 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2997 as requested by Marc Sherman.
2999 SC/10 Eximstats: added histograms for user specified patterns as requested
3002 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3004 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3005 fopen() in the content-scanning modules that did not already have it.
3011 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3012 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3014 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3016 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3018 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3019 to test Sieve filters that use "vacation".
3021 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3022 that changes the way the GnuTLS parameters are stored in the cache file.
3023 The new format can be generated externally. For backward compatibility,
3024 if the data in the cache doesn't make sense, Exim assumes it has read an
3025 old-format file, and it generates new data and writes a new file. This
3026 means that you can't go back to an older release without removing the
3029 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3030 work if there are any delivery delays because "one_time" forces the
3031 parent to be marked "delivered", so its unseen clone is never tried
3032 again. For this reason, Exim now forbids the simultaneous setting of
3035 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3036 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3037 entry if you want to know the details. The fix, however, neglected to
3038 consider the case when local delivery batching is involved. The test for
3039 "previously delivered" was not happening when checking to see if an
3040 address could be batched with a previous (undelivered) one; under
3041 certain circumstances this could lead to multiple deliveries to the same
3044 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3045 in its include files, and this causes problems building Exim.
3047 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3048 header_syntax) but Exim was just ignoring anything given after a slash.
3049 In particular, this caused confusion with an attempt to use "verify =
3050 reverse_host_lookup/defer_ok". An error is now given when options are
3051 supplied for verify items that do not have them. (Maybe reverse_host_
3052 lookup should have a defer_ok option, but that's a different point.)
3054 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3055 defined by RFC 821) to 2048, because there were problems with some AUTH
3056 commands, and RFC 1869 says the size should be increased for extended
3057 SMTP commands that take arguments.
3059 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3062 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3063 "unknown" error; now it says that the functionality isn't in the binary.
3065 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3066 an address' error message when a string expansion fails (syntax or
3067 whatever). Otherwise the password may appear in the log. Following change
3068 PH/42 below, there is no longer a chance of it appearing in a bounce
3071 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3073 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3074 of its addresses was ever tried. (Bugzilla bug #2.)
3076 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3077 the result incorrectly in the debug output. (It correctly added a newline
3078 to what was transported.)
3080 TF/01 Added $received_time.
3082 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3083 commented out examples of how to interface to a virus scanner and to
3084 SpamAssassin. Also added commented examples of av_scanner and
3085 spamd_address settings.
3087 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3088 and controls are allowed in which ACLs. There were a couple of minor
3089 errors. Some of the entries in the conditions table (which is a table of
3090 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3091 negation of where the condition IS allowed.
3093 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3095 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3096 header file does not have a version number, so I've had to invent a new
3097 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3098 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3099 radiusclient), but it was contributed by a Radius user.
3101 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3102 files or directories, for OpenSSL.
3104 PH/19 When an Exim process that is running as root has to create an Exim log
3105 file, it does so in a subprocess that runs as exim:exim so as to get the
3106 ownership right at creation (otherwise, other Exim processes might see
3107 the file with the wrong ownership). There was no test for failure of this
3108 fork() call, which would lead to the process getting stuck as it waited
3109 for a non-existent subprocess. Forks do occasionally fail when resources
3110 run out. I reviewed all the other calls to fork(); they all seem to check
3113 PH/20 When checking for unexpected SMTP input at connect time (before writing
3114 the banner), Exim was not dealing correctly with a non-positive return
3115 from the read() function. If the client had disconnected by this time,
3116 the result was a log entry for a synchronization error with an empty
3117 string after "input=" when read() returned zero. If read() returned -1
3118 (an event I could not check), uninitialized data bytes were printed.
3119 There were reports of junk text (parts of files, etc) appearing after
3122 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3124 PH/22 Added support for macro redefinition, and (re)definition in between
3125 driver and ACL definitions.
3127 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3128 forgetting to use the resulting value; it was using the unexpanded value.
3130 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3131 hadn't been configured. The fix is from Juergen Kreileder, who
3132 understands it better than I do:
3134 "Here's what I see happening with three configured cyrus_sasl
3135 authenticators configured (plain, login, cram-md5):
3137 On startup auth_cyrus_sasl_init() gets called for each of these.
3138 This means three calls to sasl_listmech() without a specified mech_list.
3139 => SASL tests which mechs of all available mechs actually work
3140 => three warnings about OTP not working
3141 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3143 With the patch, sasl_listmech() also gets called three times. But now
3144 SASL's mech_list option is set to the server_mech specified in the the
3145 authenticator. Or in other words, the answer from sasl_listmech()
3146 gets limited to just the mech you're testing for (which is different
3148 => the return list contains just 'plain' or 'login', 'cram-md5' or
3149 nothing depending on the value of ob->server_mech.
3151 I've just tested the patch: Authentication still works fine,
3152 unavailable mechs specified in the exim configuration are still
3153 caught, and the auth.log warnings about OTP are gone."
3155 PH/25 When debugging is enabled, the contents of the command line are added
3156 to the debugging output, even when log_selector=+arguments is not
3159 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3160 answer is "GNU", and only if the return is "GNU/something" is the answer
3163 PH/27 $acl_verify_message is now set immediately after the failure of a
3164 verification in an ACL, and so is available in subsequent modifiers. In
3165 particular, the message can be preserved by coding like this:
3167 warn !verify = sender
3168 set acl_m0 = $acl_verify_message
3170 Previously, $acl_verify_message was set only while expanding "message"
3171 and "log_message" when a very denied access.
3173 PH/28 Modified OS/os.c-Linux with
3175 -#ifndef OS_LOAD_AVERAGE
3176 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3178 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3179 nomenclature these days.)
3181 PH/29 Installed patch from the Sieve maintainer that adds the options
3182 sieve_useraddress and sieve_subaddress to the redirect router.
3184 PH/30 In these circumstances:
3185 . Two addresses routed to the same list of hosts;
3186 . First host does not offer TLS;
3187 . First host accepts first address;
3188 . First host gives temporary error to second address;
3189 . Second host offers TLS and a TLS session is established;
3190 . Second host accepts second address.
3191 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3192 and peerdn, if requested) that were in fact used only for the second
3195 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3196 attention to any local part prefix or suffix that was matched by the
3197 router that accepted the address. It now behaves in the same way as it
3198 does for delivery: the affixes are removed from the local part unless
3199 rcpt_include_affixes is set on the transport.
3201 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3202 timeout during the DATA phase of an incoming message.
3204 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3205 applied a patch sanctioned by the Sieve maintainer.
3207 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3208 the uid or gid is negative. A case of a negative gid caused this to be
3209 noticed. The fix allows for either to be negative.
3211 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3212 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3215 PH/36 The change PH/12 above was broken. Fixed it.
3217 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3218 the grounds that routing the same address twice would always produce the
3219 same answer. This might have been true once, but it is certainly no
3220 longer true now. Routing a child address may depend on the previous
3221 routing that produced that child. Some complicated redirection strategies
3222 went wrong when messages had multiple recipients, and made Exim's
3223 behaviour dependent on the order in which the addresses were given.
3225 I have moved the duplicate checking until after the routing is complete.
3226 Exim scans the addresses that are assigned to local and remote
3227 transports, and removes any duplicates. This means that more work will be
3228 done, as duplicates will always all be routed, but duplicates are
3229 presumably rare, so I don't expect this is of any significance.
3231 For deliveries to pipes, files, and autoreplies, the duplicate checking
3232 still happens during the routing process, since they are not going to be
3235 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3236 It corrects a timeout issue with spamd. This is Ian's comment: "The
3237 background is that sometimes spamd either never reads data from a
3238 connection it has accepted, or it never writes response data. The exiscan
3239 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3240 blindly assumes that writes won't block so it may never time out."
3242 PH/39 Allow G after quota size as well as K and M.
3244 PH/40 The value set for $authenticated_id in an authenticator may not contain
3245 binary zeroes or newlines because the value is written to log lines and
3246 to spool files. There was no check on this. Now the value is run through
3247 the string_printing() function so that such characters are converted to
3248 printable escape sequences.
3250 PH/41 $message_linecount is a new variable that contains the total number of
3251 lines in the message. Compare $body_linecount, which is the count for the
3254 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3255 bounce and delay warning messages, except in certain special cases, which
3258 (a) An SMTP error message from a remote host;
3259 (b) A message specified in a :fail: redirection;
3260 (c) A message specified in a "fail" command in a system filter;
3261 (d) A message specified in a FAIL return from the queryprogram router;
3262 (e) A message specified by the cannot_route_message router option.
3264 In these cases only, Exim does include the error details in bounce and
3265 warning messages. There are also a few cases where bland messages such
3266 as "unrouteable address" or "local delivery error" are given.
3268 PH/43 $value is now also set for the "else" part of a ${run expansion.
3270 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3271 being worked on, but at least Exim now implements the latest version to
3274 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3275 process to complete was treated as a delivery failure, a timeout while
3276 writing the message to the pipe was logged, but erroneously treated as a
3277 successful delivery. Such timeouts include transport filter timeouts. For
3278 consistency with the overall process timeout, these timeouts are now
3279 treated as errors, giving rise to delivery failures by default. However,
3280 there is now a new Boolean option for the pipe transport called
3281 timeout_defer, which, if set TRUE, converts the failures into defers for
3282 both kinds of timeout. A transport filter timeout is now identified in
3285 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3286 systems where "make" and "gmake" are different, calling "gmake" at top
3287 level broke things. I've arranged for the value of $(MAKE) to be passed
3288 from the Makefile to this script so that it can call the same version of
3292 A note about Exim versions 4.44 and 4.50
3293 ----------------------------------------
3295 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3296 changes of various kinds. As a consequence, a big documentation update was
3297 needed. This delayed the release for rather longer than seemed good, especially
3298 in the light of a couple of (minor) security issues. Therefore, the changes
3299 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3300 release. So 4.44 and 4.50 are in effect two different branches that both start
3303 I have left the 4.50 change log unchanged; it contains all the changes since
3304 4.43. The change log for 4.44 is below; many of its items are identical to
3305 those for 4.50. This seems to be the most sensible way to preserve the
3306 historical information.
3312 1. Minor wording change to the doc/README.SIEVE file.
3314 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3315 computation of the current number of files was incorrect.
3317 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3318 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3321 4. Give more explanation in the error message when the command for a transport
3322 filter fails to execute.
3324 5. There are several places where Exim runs a non-Exim command in a
3325 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3326 was being done only for the command run by the queryprogram router. It is
3327 now done for all such subprocesses. The other cases are: ${run, transport
3328 filters, and the commands run by the lmtp and pipe transports.
3330 6. Added CONFIGURE_GROUP build-time option.
3332 7. Some older OS have a limit of 256 on the maximum number of file
3333 descriptors. Exim was using setrlimit() to set 1000 as a large value
3334 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3335 systems. I've change it so that if it can't get 1000, it tries for 256.
3337 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3338 was an oversight, and furthermore, ever since the addition of extra
3339 controls (e.g. 4.43/32), the checks on when to allow different forms of
3340 "control" were broken. There should now be diagnostics for all cases when a
3341 control that does not make sense is encountered.
3343 9. Added the /retain_sender option to "control=submission".
3345 10. $recipients is now available in the predata ACL (oversight).
3347 11. Tidy the search cache before the fork to do a delivery from a message
3348 received from the command line. Otherwise the child will trigger a lookup
3349 failure and thereby defer the delivery if it tries to use (for example) a
3350 cached ldap connection that the parent has called unbind on.
3352 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3353 of $address_data from the recipient verification was clobbered by the
3354 sender verification.
3356 13. The value of address_data from a sender verification is now available in
3357 $sender_address_data in subsequent conditions in the ACL statement.
3359 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3361 15. Added a new option "connect=<time>" to callout options, to set a different
3364 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3365 was its contents. (It was OK if the option was not defined at all.)
3367 17. A "Completed" log line is now written for messages that are removed from
3368 the spool by the -Mrm option.
3370 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3371 information about exactly what failed.
3373 19. Added -dd to debug only the daemon process.
3375 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3376 handles timeouts, both on the server side and network timeouts. Renamed the
3377 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3379 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3380 It is now set to "smtps".
3382 22. $host_address is now set to the target address during the checking of
3383 ignore_target_hosts.
3385 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3386 being passed; this would have caused $sender_host_name to have been used if
3387 matching the list had actually called for a host name (not very likely,
3388 since this list is usually IP addresses). A host name is now passed as
3391 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3392 code that specifies a non-restarting handler (typically sigaction() in
3393 modern systems) in an attempt to fix a rare and obscure crash bug.
3395 25. Narrowed the window for a race in the daemon that could cause it to ignore
3396 SIGCHLD signals. This is not a major problem, because they are used only to
3397 wake it up if nothing else does.
3399 26. A malformed maildirsize file could cause Exim to calculate negative values
3400 for the mailbox size or file count. Odd effects could occur as a result.
3401 The maildirsize information is now recalculated if the size or filecount
3404 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3405 support for a long time. Removed HAVE_SYS_VFS_H.
3407 28. Installed the latest version of exipick from John Jetmore.
3409 29. In an address list, if the pattern was not a regular expression, an empty
3410 subject address (from a bounce message) matched only if the pattern was an
3411 empty string. Non-empty patterns were not even tested. This was the wrong
3412 because it is perfectly reasonable to use an empty address as part of a
3413 database query. An empty address is now tested by patterns that are
3414 lookups. However, all the other forms of pattern expect the subject to
3415 contain a local part and a domain, and therefore, for them, an empty
3416 address still always fails if the pattern is not itself empty.
3418 30. Exim went into a mad DNS loop when attempting to do a callout where the
3419 host was specified on an smtp transport, and looking it up yielded more
3420 than one IP address.
3422 31. Re-factored the code for checking spool and log partition space into a
3423 function that finds that data and another that does the check. The former
3424 is then used to implement four new variables: $spool_space, $log_space,
3425 $spool_inodes, and $log_inodes.
3427 32. The RFC2047 encoding function was originally intended for short strings
3428 such as real names; it was not keeping to the 75-character limit for
3429 encoded words that the RFC imposes. It now respects the limit, and
3430 generates multiple encoded words if necessary. To be on the safe side, I
3431 have increased the buffer size for the ${rfc2047: expansion operator from
3434 33. It is now permitted to omit both strings after an "if" condition; if the
3435 condition is true, the result is "true". As before, when the second string
3436 is omitted, a false condition yields an empty string. This makes it less
3437 cumbersome to write custom ACL and router conditions.
3439 34. Failure to deliver a bounce message always caused it to be frozen, even if
3440 there was an errors_to setting on the router. The errors_to setting is now
3443 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3444 canonical form (fully expanded) before being placed in
3445 $sender_host_address.
3447 36. The table in the code that translates DNS record types into text (T_A to
3448 "A" for instance) was missing entries for NS and CNAME. It is just possible
3449 that this could have caused confusion if both these types were looked up
3450 for the same domain, because the text type is used as part of Exim's
3451 per-process caching. But the chance of anyone hitting this buglet seems
3454 37. The dnsdb lookup has been extended in a number of ways.
3456 (1) There is a new type, "zns", which walks up the domain tree until it
3457 finds some nameserver records. It should be used with care.
3459 (2) There is a new type, "mxh", which is like "mx" except that it returns
3460 just the host names, not the priorities.
3462 (3) It is now possible to give a list of domains (or IP addresses) to be
3463 looked up. The behaviour when one of the lookups defers can be
3464 controlled by a keyword.
3466 (4) It is now possible to specify the separator character for use when
3467 multiple records are returned.
3469 38. The dnslists ACL condition has been extended: it is now possible to supply
3470 a list of IP addresses and/or domains to be looked up in a particular DNS
3473 39. Added log_selector=+queue_time_overall.
3475 40. When running the queue in the test harness, wait just a tad after forking a
3476 delivery process, to get repeatability of debugging output.
3478 41. Include certificate and key file names in error message when GnuTLS fails
3479 to set them up, because the GnuTLS error message doesn't include the name
3480 of the failing file when there is a problem reading it.
3482 42. Allow both -bf and -bF in the same test run.
3484 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3486 44. The "Exiscan patch" is now merged into the mainline Exim source.
3488 45. Sometimes the final signoff response after QUIT could fail to get
3489 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3490 < 0 before doing a fflush(). This bug looks as though it goes back to the
3491 introduction of TLS in release 3.20, but "sometimes" must have been rare
3492 because the tests only now provoked it.
3494 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3495 (this can affect the format of dates).
3497 47. exim_tidydb, when checking for the continued existence of a message for
3498 which it has found a message-specific retry record, was not finding
3499 messages that were in split spool directories. Consequently, it was
3500 deleting retry records that should have stayed in existence.
3502 48. Steve fixed some bugs in eximstats.
3504 49. The SPA authentication driver was not abandoning authentication and moving
3505 on to the next authenticator when an expansion was forced to fail,
3506 contradicting the general specification for all authenticators. Instead it
3507 was generating a temporary error. It now behaves as specified.
3509 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3510 (the order specifies the preference for clients). The order is now AES256,
3511 AES128, 3DES, ARCFOUR128.
3513 51. Small patch to Sieve code - explicitly set From: when generating an
3516 52. Exim crashed if a remote delivery caused a very long error message to be
3517 recorded - for instance if somebody sent an entire SpamAssassin report back
3518 as a large number of 550 error lines. This bug was coincidentally fixed by
3519 increasing the size of one of Exim's internal buffers (big_buffer) that
3520 happened as part of the Exiscan merge. However, to be on the safe side, I
3521 have made the code more robust (and fixed the comments that describe what
3524 53. Now that there can be additional text after "Completed" in log lines (if
3525 the queue_time_overall log selector is set), a one-byte patch to exigrep
3526 was needed to allow it to recognize "Completed" as not the last thing in
3529 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3530 patch that reportedly fixes this has been added. I am not expert enough to
3531 create a test for it. This is what the patch creator wrote:
3533 "I found a little strange behaviour of ldap code when working with
3534 Windows 2003 AD Domain, where users was placed in more than one
3535 Organization Units. When I tried to give exim partial DN, the exit code
3536 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3537 But simultaneously result of request was absolutely normal ldap result,
3538 so I produce this patch..."
3540 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3541 REFERENCE, so I have modified the code to exclude the patch when that macro
3544 55. Some experimental protocols are using DNS PTR records for new purposes. The
3545 keys for these records are domain names, not reversed IP addresses. The
3546 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3547 leaves it alone. Component reversal etc. now happens only for IP addresses.
3550 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3552 57. Double the size of the debug message buffer (to 2048) so that more of very
3553 long debug lines gets shown.
3555 58. The exicyclog utility now does better if the number of log files to keep
3556 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3558 59. Two changes related to the smtp_active_hostname option:
3560 (1) $smtp_active_hostname is now available as a variable.
3561 (2) The default for smtp_banner uses $smtp_active_hostname instead
3562 of $primary_hostname.
3564 60. The host_aton() function is supposed to be passed a string that is known
3565 to be a valid IP address. However, in the case of IPv6 addresses, it was
3566 not checking this. This is a hostage to fortune. Exim now panics and dies
3567 if the condition is not met. A case was found where this could be provoked
3568 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3569 components; fortuitously, this particular loophole had already been fixed
3570 by change 4.50/55 above.
3572 If there are any other similar loopholes, the new check in host_aton()
3573 itself should stop them being exploited. The report I received stated that
3574 data on the command line could provoke the exploit when Exim was running as
3575 exim, but did not say which command line option was involved. All I could
3576 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3577 running as the user.
3580 61. There was a buffer overflow vulnerability in the SPA authentication code
3581 (which came originally from the Samba project). I have added a test to the
3582 spa_base64_to_bits() function which I hope fixes it.
3585 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3586 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3588 63. The daemon start-up calls getloadavg() while still root for those OS that
3589 need the first call to be done as root, but it missed one case: when
3590 deliver_queue_load_max is set with deliver_drop_privilege. This is
3591 necessary for the benefit of the queue runner, because there is no re-exec
3592 when deliver_drop_privilege is set.
3594 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3595 This has been fixed.
3597 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3598 was in use, was not putting the data itself into the right store pool;
3599 consequently, it could be overwritten for a subsequent message in the same
3600 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3603 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3605 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3606 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3607 the function were treating the return as a boolean value, which happened to
3608 work because 0=false and not-0=true, but is not correct code.
3610 68. The host_aton() function was not handling scoped IPv6 addresses (those
3611 with, for example, "%eth0" on the end) correctly.
3613 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3614 negated items (that is, ~something) in unsigned ints. Some compilers
3615 apparently mutter when there is no cast.
3617 70. If an address verification called from an ACL failed, and did not produce a
3618 user-specific message (i.e. there was only a "system" message), nothing was
3619 put in $acl_verify_message. In this situation, it now puts the system
3622 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3623 session; change 4.31/43 added the unwanted input to the log line - except
3624 that it did not do this in the start of session case. It now does.
3626 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3627 This is wrong and can cause the other end to generate a synchronization
3628 error if it is another Exim or anything else that does the synchronization
3629 check. A QUIT command is no longer sent after a timeout.
3631 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3632 during host lookups.
3634 74. The defer_ok option of callout verification was not working if it was used
3635 when verifying addresses in header lines, that is, for this case:
3637 verify = header_sender/callout=defer_ok
3639 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3640 those file descriptors could be used for SMTP connections. If anything
3641 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3642 could be sent to the SMTP client, causing chaos. The daemon now opens
3643 stdin, stdout, and stderr to /dev/null when it puts itself into the
3646 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3647 log by default. The user can override this with suitable Perl magic.
3649 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3650 the log message when discard triggers, was not working for the DATA ACL or
3651 for the non-SMTP ACL.
3653 78. Error message wording change in sieve.c.
3655 79. If smtp_accept_max_per_host was set, the number of connections could be
3656 restricted to fewer than expected, because the daemon was trying to set up
3657 a new connection before checking whether the processes handling previous
3658 connections had finished. The check for completed processes is now done
3659 earlier. On busy systems, this bug wouldn't be noticed because something
3660 else would have woken the daemon, and it would have reaped the completed
3663 80. If a message was submitted locally by a user whose login name contained one
3664 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3665 It caused a spool format error. I have fixed the spool reading code. A
3666 related problem was that the "from" clause in the Received: line became
3667 illegal because of the space(s). It is now covered by ${quote_local_part.
3669 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3672 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3674 83. Patch from Sieve maintainer to fix unterminated string problem in
3675 "vacation" handling.
3677 84. Some minor changes to the Linux configuration files to help with other
3678 OS variants using glibc.
3680 85. One more patch for Sieve to update vacation handling to latest spec.
3683 ----------------------------------------------------
3684 See the note above about the 4.44 and 4.50 releases.
3685 ----------------------------------------------------
3691 1. Change 4.43/35 introduced a bug that caused file counts to be
3692 incorrectly computed when quota_filecount was set in an appendfile
3695 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3696 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3699 3. Give more explanation in the error message when the command for a transport
3700 filter fails to execute.
3702 4. There are several places where Exim runs a non-Exim command in a
3703 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3704 was being done only for the command run by the queryprogram router. It is
3705 now done for all such subprocesses. The other cases are: ${run, transport
3706 filters, and the commands run by the lmtp and pipe transports.
3708 5. Some older OS have a limit of 256 on the maximum number of file
3709 descriptors. Exim was using setrlimit() to set 1000 as a large value
3710 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3711 systems. I've change it so that if it can't get 1000, it tries for 256.
3713 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3714 was an oversight, and furthermore, ever since the addition of extra
3715 controls (e.g. 4.43/32), the checks on when to allow different forms of
3716 "control" were broken. There should now be diagnostics for all cases when a
3717 control that does not make sense is encountered.
3719 7. $recipients is now available in the predata ACL (oversight).
3721 8. Tidy the search cache before the fork to do a delivery from a message
3722 received from the command line. Otherwise the child will trigger a lookup
3723 failure and thereby defer the delivery if it tries to use (for example) a
3724 cached ldap connection that the parent has called unbind on.
3726 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3727 of $address_data from the recipient verification was clobbered by the
3728 sender verification.
3730 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3731 was its contents. (It was OK if the option was not defined at all.)
3733 11. A "Completed" log line is now written for messages that are removed from
3734 the spool by the -Mrm option.
3736 12. $host_address is now set to the target address during the checking of
3737 ignore_target_hosts.
3739 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3740 being passed; this would have caused $sender_host_name to have been used if
3741 matching the list had actually called for a host name (not very likely,
3742 since this list is usually IP addresses). A host name is now passed as
3745 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3746 code that specifies a non-restarting handler (typically sigaction() in
3747 modern systems) in an attempt to fix a rare and obscure crash bug.
3749 15. Narrowed the window for a race in the daemon that could cause it to ignore
3750 SIGCHLD signals. This is not a major problem, because they are used only to
3751 wake it up if nothing else does.
3753 16. A malformed maildirsize file could cause Exim to calculate negative values
3754 for the mailbox size or file count. Odd effects could occur as a result.
3755 The maildirsize information is now recalculated if the size or filecount
3758 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3759 support for a long time. Removed HAVE_SYS_VFS_H.
3761 18. Updated exipick to current release from John Jetmore.
3763 19. Allow an empty sender to be matched against a lookup in an address list.
3764 Previously the only cases considered were a regular expression, or an
3767 20. Exim went into a mad DNS lookup loop when doing a callout where the
3768 host was specified on the transport, if the DNS lookup yielded more than
3771 21. The RFC2047 encoding function was originally intended for short strings
3772 such as real names; it was not keeping to the 75-character limit for
3773 encoded words that the RFC imposes. It now respects the limit, and
3774 generates multiple encoded words if necessary. To be on the safe side, I
3775 have increased the buffer size for the ${rfc2047: expansion operator from
3778 22. Failure to deliver a bounce message always caused it to be frozen, even if
3779 there was an errors_to setting on the router. The errors_to setting is now
3782 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3783 canonical form (fully expanded) before being placed in
3784 $sender_host_address.
3786 24. Updated eximstats to version 1.33
3788 25. Include certificate and key file names in error message when GnuTLS fails
3789 to set them up, because the GnuTLS error message doesn't include the name
3790 of the failing file when there is a problem reading it.
3792 26. Expand error message when OpenSSL has problems setting up cert/key files.
3795 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3796 (this can affect the format of dates).
3798 28. exim_tidydb, when checking for the continued existence of a message for
3799 which it has found a message-specific retry record, was not finding
3800 messages that were in split spool directories. Consequently, it was
3801 deleting retry records that should have stayed in existence.
3803 29. eximstats updated to version 1.35
3804 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3805 1.35 - bugfix such that pie charts by volume are generated correctly
3807 30. The SPA authentication driver was not abandoning authentication and moving
3808 on to the next authenticator when an expansion was forced to fail,
3809 contradicting the general specification for all authenticators. Instead it
3810 was generating a temporary error. It now behaves as specified.
3812 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3813 (the order specifies the preference for clients). The order is now AES256,
3814 AES128, 3DES, ARCFOUR128.
3816 31. Small patch to Sieve code - explicitly set From: when generating an
3819 32. Exim crashed if a remote delivery caused a very long error message to be
3820 recorded - for instance if somebody sent an entire SpamAssassin report back
3821 as a large number of 550 error lines. This bug was coincidentally fixed by
3822 increasing the size of one of Exim's internal buffers (big_buffer) that
3823 happened as part of the Exiscan merge. However, to be on the safe side, I
3824 have made the code more robust (and fixed the comments that describe what
3827 33. Some experimental protocols are using DNS PTR records for new purposes. The
3828 keys for these records are domain names, not reversed IP addresses. The
3829 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3830 leaves it alone. Component reversal etc. now happens only for IP addresses.
3833 34. The host_aton() function is supposed to be passed a string that is known
3834 to be a valid IP address. However, in the case of IPv6 addresses, it was
3835 not checking this. This is a hostage to fortune. Exim now panics and dies
3836 if the condition is not met. A case was found where this could be provoked
3837 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3838 components; fortuitously, this particular loophole had already been fixed
3839 by change 4.50/55 or 4.44/33 above.
3841 If there are any other similar loopholes, the new check in host_aton()
3842 itself should stop them being exploited. The report I received stated that
3843 data on the command line could provoke the exploit when Exim was running as
3844 exim, but did not say which command line option was involved. All I could
3845 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3846 running as the user.
3849 35. There was a buffer overflow vulnerability in the SPA authentication code
3850 (which came originally from the Samba project). I have added a test to the
3851 spa_base64_to_bits() function which I hope fixes it.
3854 36. The daemon start-up calls getloadavg() while still root for those OS that
3855 need the first call to be done as root, but it missed one case: when
3856 deliver_queue_load_max is set with deliver_drop_privilege. This is
3857 necessary for the benefit of the queue runner, because there is no re-exec
3858 when deliver_drop_privilege is set.
3860 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3861 was in use, was not putting the data itself into the right store pool;
3862 consequently, it could be overwritten for a subsequent message in the same
3863 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3866 38. Sometimes the final signoff response after QUIT could fail to get
3867 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3868 < 0 before doing a fflush(). This bug looks as though it goes back to the
3869 introduction of TLS in release 3.20, but "sometimes" must have been rare
3870 because the tests only now provoked it.
3876 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3877 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3878 Now it returns an integer. A number of calls were still expecting a T/F
3879 return. Fortuitously, in all cases, the tests worked in OK situations,
3880 which is the norm. However, things would have gone wrong on any write
3881 failures on the smtp file descriptor. This function is used when sending
3882 messages over SMTP and also when doing verify callouts.
3884 2. When Exim is called to do synchronous delivery of a locally submitted
3885 message (the -odf or -odi options), it no longer closes stderr before doing
3888 3. Implemented the mua_wrapper option.
3890 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3892 5. Implemented the functions header_remove(), header_testname(),
3893 header_add_at_position(), and receive_remove_recipient(), and exported them
3896 6. If an ACL "warn" statement specified the addition of headers, Exim already
3897 inserted X-ACL-Warn: at the start if there was no header name. However, it
3898 was not making this test for the second and subsequent header lines if
3899 there were newlines in the string. This meant that an invalid header could
3900 be inserted if Exim was badly configured.
3902 7. Allow an ACL "warn" statement to add header lines at the start or after all
3903 the Received: headers, as well as at the end.
3905 8. Added the rcpt_4xx retry error code.
3907 9. Added postmaster_mailfrom=xxx to callout verification option.
3909 10. Added mailfrom=xxxx to the callout verification option, for verify=
3912 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3913 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3915 12. Inserted some casts to stop certain compilers warning when using pointer
3916 differences as field lengths or precisions in printf-type calls (mostly
3917 affecting debugging statements).
3919 13. Added optional readline() support for -be (dynamically loaded).
3921 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3922 same clock tick as a message's arrival, so that its received time was the
3923 same as the "first fail" time on the retry record, and that message
3924 remained on the queue past the ultimate address timeout, every queue runner
3925 would try a delivery (because it was past the ultimate address timeout) but
3926 after another failure, the ultimate address timeout, which should have then
3927 bounced the address, did not kick in. This was a "< instead of <=" error;
3928 in most cases the first failure would have been in the next clock tick
3929 after the received time, and all would be well.
3931 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3932 being recognized when the domain list was tested by the match_domain
3933 condition in an expansion string.
3935 16. Added the ${str2b64: operator.
3937 17. Exim was always calling setrlimit() to set a large limit for the number of
3938 processes, without checking whether the existing limit was already
3939 adequate. (It did check for the limit on file descriptors.) Furthermore,
3940 errors from getrlimit() and setrlimit() were being ignored. Now they are
3941 logged to the main and panic logs, but Exim does carry on, to try to do its
3942 job under whatever limits there are.
3944 18. Imported PCRE 5.0.
3946 19. Trivial typo in log message " temporarily refused connection" (the leading
3949 20. If the log selector return_path_on_delivery was set and an address was
3950 redirected to /dev/null, the delivery process crashed because it assumed
3951 that a return path would always be set for a "successful" delivery. In this
3952 case, the whole delivery is bypassed as an optimization, and therefore no
3955 21. Internal re-arrangement: the function for sending a challenge and reading
3956 a response while authentication was assuming a zero-terminated challenge
3957 string. It's now changed to take a pointer and a length, to allow for
3958 binary data in such strings.
3960 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3962 23. Exim was not respecting finduser_retries when seeking the login of the
3963 uid under which it was called; it was always trying 10 times. (The default
3964 setting of finduser_retries is zero.) Also, it was sleeping after the final
3965 failure, which is pointless.
3967 24. Implemented tls_on_connect_ports.
3969 25. Implemented acl_smtp_predata.
3971 26. If the domain in control=submission is set empty, Exim assumes that the
3972 authenticated id is a complete email address when it generates From: or
3973 Sender: header lines.
3975 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3976 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3977 chown and chgrp in /bin and hostname in /usr/bin.
3979 28. Exim was keeping the "process log" file open after each use, just as it
3980 does for the main log. This opens the possibility of it remaining open for
3981 long periods when the USR1 signal hits a daemon. Occasional processlog
3982 errors were reported, that could have been caused by this. Anyway, it seems
3983 much more sensible not to leave this file open at all, so that is what now
3986 29. The long-running daemon process does not normally write to the log once it
3987 has entered its main loop, and it closes the log before doing so. This is
3988 so that log files can straightforwardly be renamed and moved. However,
3989 there are a couple of unusual error situations where the daemon does write
3990 log entries, and I had neglected to close the log afterwards.
3992 30. The text of an SMTP error response that was received during a remote
3993 delivery was being truncated at 512 bytes. This is too short for some of
3994 the long messages that one sometimes sees. I've increased the limit to
3997 31. It is now possible to make retry rules that apply only when a message has a
3998 specific sender, in particular, an empty sender.
4000 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4001 it possible to be selective about when SMTP synchronization is enforced.
4003 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4005 32. Implemented hosts_connection_nolog.
4007 33. Added an ACL for QUIT.
4009 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4012 35. Added mailbox_size and mailbox_filecount to appendfile.
4014 36. Added control = no_multiline_responses to ACLs.
4016 37. There was a bug in the logic of the code that waits for the clock to tick
4017 in the case where the clock went backwards by a substantial amount such
4018 that the microsecond fraction of "now" was more than the microsecond
4019 fraction of "then" (but the whole seconds number was less).
4021 38. Added support for the libradius Radius client library this is found on
4022 FreeBSD (previously only the radiusclient library was supported).
4028 1. When certain lookups returned multiple values in the form name=value, the
4029 quoting of the values was not always being done properly. Specifically:
4030 (a) If the value started with a double quote, but contained no whitespace,
4032 (b) If the value contained whitespace other than a space character (i.e.
4033 tabs or newlines or carriage returns) it was not quoted.
4034 This fix has been applied to the mysql and pgsql lookups by writing a
4035 separate quoting function and calling it from the lookup code. The fix
4036 should probably also be applied to nisplus, ibase and oracle lookups, but
4037 since I cannot test any of those, I have not disturbed their existing code.
4039 2. A hit in the callout cache for a specific address caused a log line with no
4040 reason for rejecting RCPT. Now it says "Previous (cached) callout
4041 verification failure".
4043 3. There was an off-by-one bug in the queryprogram router. An over-long
4044 return line was truncated at 256 instead of 255 characters, thereby
4045 overflowing its buffer with the terminating zero. As well as fixing this, I
4046 have increased the buffer size to 1024 (and made a note to document this).
4048 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4049 when Exim is waiting for an SMTP response from a remote server, Exim
4050 restarts its select() call on the socket, thereby resetting its timeout.
4051 This is not a problem when such interrupts are rare. Somebody set up a cron
4052 job to run exiwhat every 2 minutes, which is less than the normal select()
4053 timeout (5 or 10 minutes). This meant that the select() timeout never
4054 kicked in because it was always reset. I have fixed this by comparing the
4055 time when an interrupt arrives with the time at the start of the first call
4056 to select(). If more time than the timeout has elapsed, the interrupt is
4057 treated as a timeout.
4059 5. Some internal re-factoring in preparation for the addition of Sieve
4060 extensions (by MH). In particular, the "personal" test is moved to a
4061 separate function, and given an option for scanning Cc: and Bcc: (which is
4062 not set for Exim filters).
4064 6. When Exim created an email address using the login of the caller as the
4065 local part (e.g. when creating a From: or Sender: header line), it was not
4066 quoting the local part when it contained special characters such as @.
4068 7. Installed new OpenBSD configuration files.
4070 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4071 try to make them clearer.
4073 9. Callout options, other than the timeout value, were being ignored when
4074 verifying sender addresses in header lines. For example, when using
4076 verify = header_sender/callout=no_cache
4078 the cache was (incorrectly) being used.
4080 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4081 only the Cygwin environment.
4083 11. When return_path_on_delivery was set as a log selector, if different remote
4084 addresses in the same message used different return paths and parallel
4085 remote delivery occurred, the wrong values would sometimes be logged.
4086 (Whenever a remote delivery process finished, the return path value from
4087 the most recently started remote delivery process was logged.)
4089 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4090 lines when AUTH and/or TLS are in use. This is the "received protocol"
4091 field. Exim used to use "asmtp" for authenticated SMTP, without any
4092 indication (in the protocol name) for TLS use. Now it follows the RFC and
4093 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4094 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4095 names appear in log lines as well as in Received: header lines.
4097 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4098 extensions, and comparison tests, and to fix some bugs.
4100 14. Changes to the "personal" filter test:
4102 (1) The test was buggy in that it was just doing the equivalent of
4103 "contains" tests on header lines. For example, if a user's address was
4104 anne@some.where, the "personal" test would incorrectly be true for
4106 To: susanne@some.where
4108 This test is now done by extracting each address from the header in turn,
4109 and checking the entire address. Other tests that are part of "personal"
4110 are now done using regular expressions (for example, to check local parts
4111 of addresses in From: header lines).
4113 (2) The list of non-personal local parts in From: addresses has been
4114 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4115 taken from the Sieve specification recommendations.
4117 (3) If the message contains any header line starting with "List-" it is
4118 treated as non-personal.
4120 (4) The test for "circular" in the Subject: header line has been removed
4121 because it now seems ill-conceived.
4123 15. Minor typos in src/EDITME comments corrected.
4125 16. Installed latest exipick from John Jetmore.
4127 17. If headers_add on a router specified a text string that was too long for
4128 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4129 of string_sprintf() is now avoided.
4131 18. $message_body_size was not set (it was always zero) when running the DATA
4132 ACL and the local_scan() function.
4134 19. For the "mail" command in an Exim filter, no default was being set for
4135 the once_repeat time, causing a random time value to be used if "once" was
4136 specified. (If the value happened to be <= 0, no repeat happened.) The
4137 default is now 0s, meaning "never repeat". The "vacation" command was OK
4138 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4139 (I found it when inspecting the code).
4141 20. There is now an overall timeout for performing a callout verification. It
4142 defaults to 4 times the callout timeout, which applies to individual SMTP
4143 commands during the callout. The overall timeout applies when there is more
4144 than one host that can be tried. The timeout is checked before trying the
4145 next host. This prevents very long delays if there are a large number of
4146 hosts and all are timing out (e.g. when the network connections are timing
4147 out). The value of the overall timeout can be changed by specifying an
4148 additional sub-option for "callout", called "maxwait". For example:
4150 verify = sender/callout=5s,maxwait=20s
4152 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4153 to the end before writing, but this should make it even safer).
4155 22. Exim was forgetting that it had advertised PIPELINING for the second and
4156 subsequent messages on an SMTP connection. It was also not resetting its
4157 memory on STARTTLS and an internal HELO.
4159 23. When Exim logs an SMTP synchronization error within a session, it now
4160 records whether PIPELINING has been advertised or not.
4162 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4163 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4164 rather than long int.
4166 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4168 26. Added the never_mail option to autoreply.
4174 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4175 crash if the getsockname() call failed; this can happen if a connection is
4176 closed very soon after it is established. The problem was simply in the
4177 order in which certain operations were done, causing Exim to try to write
4178 to the SMTP stream before it had set up the file descriptor. The bug has
4179 been fixed by making things happen in the correct order.
4185 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4186 before the connection was closed, thus losing the rejection response.
4188 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4189 some early Solaris releases, but causes trouble in current releases where
4190 socklen_t is defined.
4192 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4195 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4198 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4199 that releasing the top of it at the end releases what was used for sub-
4200 expansions (unless the block got too big). However, discard this block if
4201 the first thing is a variable or header, so that we can use its block when
4202 it is dynamic (useful for very large $message_headers, for example).
4204 6. Lookups now cache *every* query, not just the most recent. A new, separate
4205 store pool is used for this. It can be recovered when all lookup caches are
4206 flushed. Lookups now release memory at the end of their result strings.
4207 This has involved some general refactoring of the lookup sources.
4209 7. Some code has been added to the store_xxx() functions to reduce the amount
4210 of flapping under certain conditions.
4212 8. log_incoming_interface used to affect only the <= reception log lines. Now
4213 it causes the local interface and port to be added to several more SMTP log
4214 lines, for example "SMTP connection from", and rejection lines.
4216 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4218 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4220 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4221 could be overwritten at the end of the current message (or the start of a
4222 new message if it was set in a HELO ACL). The value is now preserved for
4223 the duration of the SMTP connection.
4225 12. If a transport had a headers_rewrite setting, and a matching header line
4226 contained an unqualified address, that address was qualified, even if it
4227 did not match any rewriting rules. The underlying bug was that the values
4228 of the flags that permit the existence of unqualified sender and recipient
4229 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4230 non-local messages, and by -bnq for local messages) were not being
4231 preserved with the message after it was received.
4233 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4234 "next input=" as part of the text comprising the host identity instead of
4235 the correct text. The code was using the same buffer for two different
4236 strings. However, depending on which order the printing function evaluated
4237 its arguments, the bug did not always show up. Under Linux, for example, my
4238 test suite worked just fine.
4240 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4241 doesn't work with some older versions of Perl. It has been changed to "my",
4242 which in any case is probably the better facility to use.
4244 15. A really picky compiler found some instances of statements for creating
4245 error messages that either had too many or two few arguments for the format
4248 16. The size of the buffer for calls to the DNS resolver has been increased
4249 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4250 for addresses that have a lot of PTR records. This alleviates a problem; it
4251 does not fully solve it.
4253 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4254 buffer now truncates the list and logs the incident, which is the same
4255 action as happens when Exim is looking up a host name and its aliases.
4256 Previously in this situation something unpredictable would happen;
4257 sometimes it was "internal error: store_reset failed".
4259 18. If a server dropped the connection unexpectedly when an Exim client was
4260 using GnuTLS and trying to read a response, the client delivery process
4261 crashed while trying to generate an error log message.
4263 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4264 string, for example:
4266 warn message = H1: something\nH2: something
4268 the text was added as a single header line from Exim's point of view
4269 though it ended up OK in the delivered message. However, searching for the
4270 second and subsequent header lines using $h_h2: did not work. This has been
4271 fixed. Similarly, if a system filter added multiple headers in this way,
4272 the routers could not see them.
4274 20. Expanded the error message when iplsearch is called with an invalid key to
4275 suggest using net-iplsearch in a host list.
4277 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4278 ACLs are no longer actually imposed (and a message to that effect is
4281 22. If a "gecos" field in a passwd entry contained escaped characters, in
4282 particular, if it contained a \" sequence, Exim got it wrong when building
4283 a From: or a Sender: header from that name. A second bug also caused
4284 incorrect handling when an unquoted " was present following a character
4285 that needed quoting.
4287 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4288 was not being matched caselessly.
4290 24. Arranged for all hyphens in the exim.8 source to be escaped with
4293 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4294 properly. Recipient callout cache records were still being keyed to include
4295 the sender, even when use_sender was set false. This led to far more
4296 callouts that were necessary. The sender is no longer included in the key
4297 when use_sender is false.
4299 26. Added "control = submission" modifier to ACLs.
4301 27. Added the ${base62d: operator to decode base 62 numbers.
4303 28. dnsdb lookups can now access SRV records.
4305 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4306 the configuration file.
4308 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4309 (-v) mode. This makes the output for a verbose queue run more intelligible.
4311 31. Added a use_postmaster feature to recipient callouts.
4313 32. Added the $body_zerocount variable, containing the number of binary zero
4314 bytes in the message body.
4316 33. The time of last modification of the "new" subdirectory is now used as the
4317 "mailbox time last read" when there is a quota error for a maildir
4320 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4322 35. Added +ignore_unknown as a special item in host lists.
4324 36. Code for decoding IPv6 addresses in host lists is now included, even if
4325 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4326 address was recognized as an IP address, but was then not correctly decoded
4327 into binary, causing unexpected and incorrect effects when compared with
4334 1. Very minor rewording of debugging text in manualroute to say "list of
4335 hosts" instead of "hostlist".
4337 2. If verify=header_syntax was set, and a header line with an unqualified
4338 address (no domain) and a large number of spaces between the end of the
4339 name and the colon was received, the reception process suffered a buffer
4340 overflow, and (when I tested it) crashed. This was caused by some obsolete
4341 code that should have been removed. The fix is to remove it!
4343 3. When running in the test harness, delay a bit after writing a bounce
4344 message to get a bit more predictability in the log output.
4346 4. Added a call to search_tidyup() just before forking a reception process. In
4347 theory, someone could use a lookup in the expansion of smtp_accept_max_
4348 per_host which, without the tidyup, could leave open a database connection.
4350 5. Added the variables $recipient_data and $sender_data which get set from a
4351 lookup success in an ACL "recipients" or "senders" condition, or a router
4352 "senders" option, similar to $domain_data and $local_part_data.
4354 6. Moved the writing of debug_print from before to after the "senders" test
4357 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4358 problems for message scanning, either using a data ACL, or using
4359 local_scan() because the Received: header was not generated till after they
4360 were called (in order to set the time as the time of reception completion).
4361 I have revised the way this works. The header is now generated after the
4362 body is received, but before the ACL or local_scan() are called. After they
4363 are run, the timestamp in the header is updated.
4369 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4370 before starting a queue runner without re-exec. This happened only when
4371 deliver_drop_privilege was set or when the Exim user was set to root. The
4372 effect of the bug was that timeouts during subsequent deliveries caused
4373 crashes instead of being properly handled. The handler is now left at its
4374 default (and expected) setting.
4376 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4377 message, again when deliver_drop_privilege is set or Exim is run as root.
4378 The bug described in (1) was not present in this case, but the tidying up
4379 of the other signals was missing. I have made the two cases consistent.
4381 3. The ignore_target_hosts setting on a manualroute router was being ignored
4382 for hosts that were looked up using the /MX notation.
4384 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4387 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4388 operated on the sender address. After changing the $sender_address to <>
4389 for the sender address verify, Exim was re-instated it as the original
4390 (before rewriting) address, but remembering that it had rewritten it, so it
4391 wasn't rewriting it again. This bug also had the effect of breaking the
4392 sender address verification caching when the sender address was rewritten.
4394 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4395 This has been changed so that if the ip literal address matches
4396 ignore_target_hosts, the router declines.
4398 7. Added expansion conditions match_domain, match_address, and match_local_
4399 part (NOT match_host).
4401 8. The placeholder for the Received: header didn't have a length field set.
4403 9. Added code to Exim itself and to exim_lock to test for a specific race
4404 condition that could lead to file corruption when using MBX delivery. The
4405 issue is with the lockfile that is created in /tmp. If this file is removed
4406 after a process has opened it but before that process has acquired a lock,
4407 there is the potential for a second process to recreate the file and also
4408 acquire a lock. This could lead to two Exim processes writing to the file
4409 at the same time. The added code performs the same test as UW imapd; it
4410 checks after acquiring the lock that its file descriptor still refers to
4411 the same named file.
4413 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4414 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4417 11. Added the smtp_active_hostname option. If used, this will typically be made
4418 to depend on the incoming interface address. Because $interface_address is
4419 not set up until the daemon has forked a reception process, error responses
4420 that can happen earlier (such as "too many connections") no longer contain
4423 12. If an expansion in a condition on a "warn" statement fails because a lookup
4424 defers, the "warn" statement is abandoned, and the next ACL statement is
4425 processed. Previously this caused the whole ACL to be aborted.
4427 13. Added the iplsearch lookup type.
4429 14. Added ident_timeout as a log selector.
4431 15. Added tls_certificate_verified as a log selector.
4433 16. Added a global option tls_require_ciphers (compare the smtp transport
4434 option of the same name). This controls incoming TLS connections.
4436 17. I finally figured out how to make tls_require_ciphers do a similar thing
4437 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4438 before starting the TLS session.
4440 18. Tabs are now shown as \t in -bP output.
4442 19. If the log selector return_path_on_delivery was set, Exim crashed when
4443 bouncing a message because it had too many Received: header lines.
4445 20. If two routers both had headers_remove settings, and the first one included
4446 a superfluous trailing colon, the final name in the first list and the
4447 first name in the second list were incorrectly joined into one item (with a
4448 colon in the middle).
4454 1. Added -C and -D options to the exinext utility, mainly to make it easier
4455 to include in the automated testing, but these could be helpful when
4456 multiple configurations are in use.
4458 2. The exinext utility was not formatting the output nicely when there was
4459 an alternate port involved in the retry record key, nor when there was a
4460 message id as well (for retries that were specific to a specific message
4461 and a specific host). It was also confused by IPv6 addresses, because of
4462 the additional colons they contain. I have fixed the IPv4 problem, and
4463 patched it up to do a reasonable job for IPv6.
4465 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4466 delivery, the log line now contains "pipelined" if PIPELINING was used.
4468 4. An SMTP transport process used to panic and die if the bind() call to set
4469 an explicit outgoing interface failed. This has been changed; it is now
4470 treated in the same way as a connect() failure.
4472 5. A reference to $sender_host_name in the part of a conditional expansion
4473 that was being skipped was still causing a DNS lookup. This no longer
4476 6. The def: expansion condition was not recognizing references to header lines
4477 that used bh_ and bheader_.
4479 7. Added the _cache feature to named lists.
4481 8. The code for checking quota_filecount in the appendfile transport was
4482 allowing one more file than it should have been.
4484 9. For compatibility with Sendmail, the command line option
4492 and sets the incoming protocol and host name (for trusted callers). The
4493 host name and its colon can be omitted when only the protocol is to be set.
4494 Note the Exim already has two private options, -pd and -ps, that refer to
4495 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4496 "s", but I don't think that's a major issue.
4498 10. A number of refactoring changes to the code, none of which should affect
4501 (a) The number of logging options was getting close to filling up the
4502 32-bit word that was used as a bit map. I have split them into two classes:
4503 those that are passed in the argument to log_write(), and those that are
4504 only ever tested independently outside of that function. These are now in
4505 separate 32-bit words, so there is plenty of room for expansion again.
4506 There is no change in the user interface or the logging behaviour.
4508 (b) When building, for example, log lines, the code previously used a
4509 macro that called string_cat() twice, in order to add two strings. This is
4510 not really sufficiently general. Furthermore, there was one instance where
4511 it was actually wrong because one of the argument was used twice, and in
4512 one call a function was used. (As it happened, calling the function twice
4513 did not affect the overall behaviour.) The macro has been replaced by a
4514 function that can join an arbitrary number of extra strings onto a growing
4517 (c) The code for expansion conditions now uses a table and a binary chop
4518 instead of a serial search (which was left over from when there were very
4519 few conditions). Also, it now recognizes conditions like "pam" even when
4520 the relevant support is not compiled in: a suitably worded error message is
4521 given if an attempt is made to use such a condition.
4523 11. Added ${time_interval:xxxxx}.
4525 12. A bug was causing one of the ddress fields not to be passed back correctly
4526 from remote delivery subprocesses. The field in question was not being
4527 subsequently used, so this caused to problems in practice.
4529 13. Added new log selectors queue_time and deliver_time.
4531 14. Might have fixed a bug in maildirsizefile handling that threw up
4532 "unexpected character" debug warnings, and recalculated the data
4533 unnecessarily. In any case, I expanded the warning message to give more
4536 15. Added the message "Restricted characters in address" to the statements in
4537 the default ACL that block characters like @ and % in local parts.
4539 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4540 Three changes have been made:
4542 (a) There was a serious bug; a negative response to MAIL caused the whole
4543 recipient domain to be cached as invalid, thereby blocking all messages
4544 to all local parts at the same domain, from all senders. This bug has
4545 been fixed. The domain is no longer cached after a negative response to
4546 MAIL if the sender used is not empty.
4548 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4551 (c) A new callout option, "use_sender" has been added for people who want
4552 the modified behaviour.
4558 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4561 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4562 indeed breaks things for older releases.
4564 3. Added additional logging to the case where there is a problem reading data
4565 from a filter that is running in a subprocess using a pipe, in order to
4566 try to track down a specific problem.
4568 4. Testing facility fudge: when running in the test harness and attempting
4569 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4570 getting "No route to host". Convert this to a timeout.
4572 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4575 6. Some OS don't have socklen_t but use size_t instead. This affects the
4576 fifth argument of getsockopt() amongst other things. This is now
4577 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4578 can be set for individual OS. I have set it for SunOS5, OSF1, and
4579 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4580 some earlier ones do not.
4582 7. Change 4.30/15 was not doing the test caselessly.
4584 8. The standard form for an IPv6 address literal was being rejected by address
4585 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4586 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4587 this, as well as the form without the "IPv6" on the front (but only when
4588 address literals are enabled, of course).
4590 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4592 10. Exim crashed if a message with an empty sender address specified by -f
4593 encountered a router with an errors_to setting. This could be provoked only
4594 by a command such as
4598 where an empty string was supplied; "<>" did not hit this bug.
4600 11. Installed PCRE release 4.5.
4602 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4603 remained set. It is now erased.
4605 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4606 times from message ids (which are base 36 rather than the normal 62).
4608 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4609 were being counted as actual protocol errors, and logged if the log
4610 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4611 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4612 and DATA following a set of rejected RCPTs do not count as protocol errors.
4613 In other words, Exim assumes they were pipelined, though this may not
4614 actually be the case. Of course, in all cases the client gets an
4615 appropriate error code.
4617 15. If a lookup fails in an ACL condition, a message about the failure may
4618 be available; it is used if testing the ACL cannot continue, because most
4619 such messages specify what the cause of the deferral is. However, some
4620 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4621 that caused an old message to be retained and used if a later statement
4622 caused a defer, replacing the real cause of the deferral.
4624 16. If an IP address had so many PTR records that the DNS lookup buffer
4625 was not large enough to hold them, Exim could crash while trying to process
4626 the truncated data. It now detects and logs this case.
4628 17. Further to 4.21/58, another change has been made: if (and only if) the
4629 first line of a message (the first header line) ends with CRLF, a bare LF
4630 in a subsequent header line has a space inserted after it, so as not to
4631 terminate the header.
4633 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4634 unnecessarily, used atoi() instead of strtol(), and didn't check the
4635 termination when getting file sizes from file names by regex.
4637 19. Completely re-implemented the support for maildirsize files, in the light
4638 of a number of problems with the previous contributed implementation
4639 (4.30/29). In particular:
4641 . If the quota is zero, the maildirsize file is maintained, but no quota is
4644 . If the maildir directory does not exist, it is created before any attempt
4645 to write a maildirsize file.
4647 . The quota value in the file is just a cache; if the quota is changed in
4648 the transport, the new value overrides.
4650 . A regular expression is available for excluding directories from the
4653 20. The autoreply transport checks the characters in options that define the
4654 message's headers; it allows continued headers, but it was checking with
4655 isspace() after an embedded newline instead of explicitly looking for a
4658 21. If all the "regular" hosts to which an address was routed had passed their
4659 expiry times, and had not reached their retry times, the address was
4660 bounced, even if fallback hosts were defined. Now Exim should go on to try
4663 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4664 equivalent code in the SMTP transport. Some hosts send humungous responses
4665 to HELO/EHLO, more than 1024 it seems.
4667 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4668 gives compiler warnings in some environments. I've now done it "properly",
4671 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4672 (because of problems with the built-in one) was declared to return uschar *
4673 instead of char *, causing compiler failure.
4675 25. Fixed a file descriptor leak when processing alias/forward files.
4677 26. Fixed a minor format string issue in dbfn.c.
4679 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4681 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4682 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4685 29. When a sender address is verified, it is cached, to save repeating the test
4686 when there is more than one recipient in a message. However, when the
4687 verification involves a callout, it is possible for different callout
4688 options to be set for different recipients. It is too complicated to keep
4689 track of this in the cache, so now Exim always runs a verification when a
4690 callout is required, relying on the callout cache for the optimization.
4691 The overhead is duplication of the address routing, but this should not be
4694 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4695 to be verified with callout=postmaster, and the main callout worked but the
4696 postmaster check failed, the verification correctly failed. However, if a
4697 subsequent RCPT command asked for sender verification *without* the
4698 postmaster check, incorrect caching caused this verification also to fail,
4701 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4702 it was not caching the DNS options (qualify_single, search_parents) that
4703 were used when the lookup failed. A subsequent lookup with different
4704 options therefore always gave the same answer, though there were cases
4705 where it should not have. (Example: a "domains = !$mx_any" option on a
4706 dnslookup router: the "domains" option is always processed without any
4707 widening, but the router might have qualify_single set.) Now Exim uses the
4708 cached value only when the same options are set.
4710 32. Added John Jetmore's "exipick" utility to the distribution.
4712 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4713 than a timeout (e.g. a certificate is required, and is not provided), an
4714 Exim server now closes the connection immediately. Previously it waited for
4715 the client to close - but if the client is SSL, it seems that they each
4716 wait for each other, leading to a delay before one of them times out.
4718 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4719 maintained 0.8.x compatibility because I don't think many are using it, and
4720 it is clearly obsolete.
4722 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4725 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4726 client certificate was expired. A simple patch fixes this, though I don't
4727 understand the full logic of why the verify callback is called multiple
4730 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4731 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4732 which causes problems with some clients (such as the Certicom SSL Plus
4733 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4734 disables the coutermeasure allowing Eudora to connect."
4736 38. Exim was not checking that a write() to a log file succeeded. This could
4737 lead to Bad Things if a log got too big, in particular if it hit a file
4738 size limit. Exim now panics and dies if it cannot write to a log file, just
4739 as it does if it cannot open a log file.
4741 39. Modified OS/Makefile-Linux so that it now contains
4743 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4745 The two -D definitions ensure that Exim is compiled with large file
4746 support, which makes it possible to handle log files that are bigger than
4749 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4750 instance) a domain was checked against a named list that involved a lookup,
4751 causing $domain_data to be set, then another domain was checked against the
4752 same list, then the first domain was re-checked, the value of $domain_data
4753 after the final check could be wrong. In particular, if the second check
4754 failed, it could be set empty. This bug probably also applied to
4757 41. The strip_trailing_dot option was not being applied to the address given
4758 with the -f command-line option.
4760 42. The code for reading a message's header from the spool was incrementing
4761 $received_count, but never initializing it. This meant that the value was
4762 incorrect (doubled) while delivering a message in the same process in which
4763 it was received. In the most common configuration of Exim, this never
4764 happens - a fresh exec is done - but it can happen when
4765 deliver_drop_privilege is set.
4767 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4768 it now includes up to 150 characters of the unexpected data in the log
4771 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4772 and building data strings. The size of both of these buffers was 10 000
4773 bytes - far larger than anybody would *ever* want, thought I. Needless to
4774 say, somebody hit the limit. I have increased the maximum line length to
4775 20 000 and the maximum data length of concatenated lines to 100 000. I have
4776 also fixed two bugs, because there was no checking on these buffers. Tsk,
4777 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4778 buffer is too small.
4780 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4781 lsearch lookups. Now it does.
4783 46. When parsing a route_list item in a manualroute router, a fixed-length
4784 buffer was used for the list of hosts. I made this 1024 bytes long,
4785 thinking that nobody would ever have a list of hosts that long. Wrong.
4786 Somebody had a whole pile of complicated expansion conditions, and the
4787 string was silently truncated, leading to an expansion error. It turns out
4788 that it is easier to change to an unlimited length (owing to other changes
4789 that have happened since this code was originally written) than to build
4790 structure for giving a limitation error. The length of the item that
4791 expands into the list of hosts is now unlimited.
4793 47. The lsearch lookup could not handle data where the length of text line was
4794 more than 4095 characters. Such lines were truncated, leading to shortened
4795 data being returned. It should now handle lines of any length.
4797 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4798 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4801 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4802 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4803 no longer happens in comment lines. A list of replacements is now placed
4804 at the head of all of the source files, except those whose only change is
4805 to replace PERL_COMMAND in the very first #! line.
4807 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4808 messages on the queue, with a bottom-up merge sort, using code contributed
4809 by Michael Haardt. This should make operations like -bp somewhat faster on
4810 large queues. It won't affect queue runners, except when queue_run_in_order
4813 51. Installed eximstats 1.31 in the distribution.
4815 52. Added support for SRV lookups to the dnslookup router.
4817 53. If an ACL referred to $message_body or $message_body_end, the value was not
4818 reset for any messages that followed in the same SMTP session.
4820 54. The store-handling optimization for building very long strings was not
4821 differentiating between the different store pools. I don't think this
4822 actually made any difference in practice, but I've tidied it.
4824 55. While running the routers to verify a sender address, $sender_address
4825 was still set to the sender address. This is wrong, because when routing to
4826 send a bounce to the sender, it would be empty. Therefore, I have changed
4827 it so that, while verifying a sender address, $sender_address is set to <>.
4828 (There is no change to what happens when verifying a recipient address.)
4830 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4831 target A or AAAA records (if not already returned) without resetting the
4832 qualify_single or search_parents options of the DNS resolver. These are
4833 inappropriate in this case because the targets of MX and SRV records must
4834 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4835 target that, when qualified, matched something in the local domain. These
4836 two options are now turned off when doing these lookups.
4838 57. It seems that at least some releases of Reiserfs (which does not have the
4839 concept of a fixed number of inodes) returns zero and not -1 for the
4840 number of available inodes. This interacted badly with check_spool_inodes,
4841 which assumed that -1 was the "no such thing" setting. What I have done is
4842 to check that the total number of inodes is greater than zero before doing
4843 the test of how many are available.
4845 58. When a "warn" ACL statement has a log_message modifier, the message is
4846 remembered, and not repeated. This is to avoid a lot of repetition when a
4847 message has many recipients that cause the same warning to be written.
4848 Howewer, Exim was preserving the list of already written lines for an
4849 entire SMTP session, which doesn't seem right. The memory is now reset if a
4850 new message is started.
4852 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4853 debugging output unless log_rewrite was also set.
4855 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4856 of (int)(handle) when we know that handle contains (void *)(-1).
4858 61. The Exim daemon panic-logs an error return when it closes the incoming
4859 connection. However "connection reset by peer" seems to be common, and
4860 isn't really an error worthy of noting specially, so that particular error
4863 62. When Exim is trying to find all the local interfaces, it used to panic and
4864 die if the ioctl to get the interface flags failed. However, it seems that
4865 on at least one OS (Solaris 9) it is possible to have an interface that is
4866 included in the list of interfaces, but for which you get a failure error
4867 for this call. This happens when the interface is not "plumbed" into a
4868 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4869 failure of the "get flags" call assumes that the interface is down.
4871 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4872 makes life easier for people who are doing arithmetic on fields extracted
4873 from dates, where you often get leading zeros that should not be
4874 interpreted as octal.
4876 64. Added qualify_domain to the redirect router, to override the global
4879 65. If a pathologically long header line contained very many addresses (the
4880 report of this problem mentioned 10 000) and each of them was rewritten,
4881 Exim could use up a very large amount of memory. (It kept on making new
4882 copies of the header line as it rewrote, and never released the old ones.)
4883 At the expense of a bit more processing, the header rewriting function has
4884 been changed so that it no longer eats memory in this way.
4886 66. The generation of the Received: header has been moved from the time that a
4887 message starts to be received, to the time that it finishes. The timestamp
4888 in the Received: header should now be very close to that of the <= log
4889 line. There are two side-effects of this change:
4891 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4892 logged header lines no longer include the local Received: line, because
4893 it has not yet been created. The same applies to a copy of the message
4894 that is returned to a non-SMTP sender when a message is rejected.
4896 (b) When a filter file is tested using -bf, no additional Received: header
4897 is added to the test message. After some thought, I decided that this
4900 This change does not affect the value of $received_for. It is still set
4901 after address rewriting, but before local_scan() is called.
4903 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4905 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4906 gave an unhelpful panic error message, and a defer error. I have managed to
4907 change this behaviour so that it now rejects any supplied certificate,
4908 which seems right, as the list of acceptable certificates is empty.
4910 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4911 gave an unhelpful defer error. I have not managed to make this reject any
4912 supplied certificates, but the error message it gives is "no certificate
4913 supplied", which is not helpful.
4915 70. exigrep's output now also includes lines that are not associated with any
4916 message, but which match the given pattern. Implemented by a patch from
4917 Martin Sluka, which also tidied up the Perl a bit.
4919 71. Recipient callout verification, like sender verification, was using <> in
4920 the MAIL FROM command. This isn't really the right thing, since the actual
4921 sender may affect whether the remote host accepts the recipient or not. I
4922 have changed it to use the actual sender in the callout; this means that
4923 the cache record is now keyed on a recipient/sender pair, not just the
4924 recipient address. There doesn't seem to be a real danger of callout loops,
4925 since a callout by the remote host to check the sender would use <>.
4926 [SEE ABOVE: changed after hitting problems.]
4928 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4929 temporary errors. However, in the case of such a code being given after
4930 the end of a data transmission (i.e. after ".") Exim was failing to write
4931 a retry record for the message. (Yes, there was some broken host that was
4932 actually sending 8xx at this point.)
4934 73. An unknown lookup type in a host list could cause Exim to panic-die when
4935 the list was checked. (An example that provoked this was putting <; in the
4936 middle of a list instead of at the start.) If this happened during a DATA
4937 ACL check, a -D file could be left lying around. This kind of configuration
4938 error no longer causes Exim to die; instead it causes a defer errror. The
4939 incident is still logged to the main and panic logs.
4941 74. Buglet left over from Exim 3 conversion. The message "too many messages
4942 in one connection" was written to the rejectlog but not the mainlog, except
4943 when address rewriting (yes!) was being logged.
4945 75. Added write_rejectlog option.
4947 76. When a system filter was run not as root (that is, when system_filter_user
4948 was set), the values of the $n variables were not being returned to the
4949 main process; thus, they were not subsequently available in the $sn
4952 77. Added +return_path_on_delivery log selector.
4954 78. A connection timeout was being treated differently from recipients deferred
4955 when testing hosts_max_try with a message that was older than the host's
4956 retry timeout. (The host should not be counted, thus allowing all hosts to
4957 be tried at least once before bouncing.) This may have been the cause of an
4958 occasionally reported bug whereby a message would remain on the queue
4959 longer than the retry timeout, but would be bounced if a delivery was
4960 forced. I say "may" because I never totally pinned down the problem;
4961 setting up timeout/retry tests is difficult. See also the next item.
4963 79. The ultimate address timeout was not being applied to errors that involved
4964 a combination of host plus message (for example, a timeout on a MAIL
4965 command). When an address resolved to a number of possible hosts, and they
4966 were not all tried for each delivery (e.g. because of hosts_max_try), a
4967 message could remain on the queue longer than the retry timeout.
4969 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4972 81. Fixed an obscure SMTP outgoing bug which required at least the following
4973 conditions: (a) there was another message waiting for the same server;
4974 (b) the server returned 5xx to all RCPT commands in the first message so
4975 that the message was not completed; (c) the server dropped the connection
4976 or gave a negative response to the RSET that Exim sends to abort the
4977 transaction. The observed case was a dropped connection after DATA that had
4978 been sent in pipelining mode. That is, the server had advertised PIPELINING
4979 but was not implementing it correctly. The effect of the bug was incorrect
4980 behaviour, such as trying another host, and this could lead to a crash.
4986 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4987 and daemon.c were passed as pointers to ints; they should have been
4988 pointers to socklen_t variables (which are typically unsigned ints).
4990 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4993 3. Fixed a really odd bug that affected only the testing scheme; patching a
4994 certain fixed string in the binary changed the value of another string that
4995 happened to be identical to the end of the original first string.
4997 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4998 name", it returns that address as the IP address. On some operating
4999 systems (e.g. Solaris), it also passes back the IP address string as the
5000 "host name". However, on others (e.g. Linux), it passes back an empty
5001 string. Exim wasn't checking for this, and was changing the host name to an
5002 empty string, assuming it had been canonicized.
5004 5. Although rare, it is permitted to have more than one PTR record for a given
5005 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5006 all the names associated with an address, because they do in Solaris.
5007 However, it seems that they do not in Linux for data that comes from the
5008 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5009 I found this out when I moved to a new Linux workstation and tried to run
5010 the Exim test suite.
5012 To get round this problem I have changed the code so that it now does its
5013 own call to the DNS to look up PTR records when searching for a host name.
5014 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5015 addresses that are only in /etc/hosts are still found.
5017 This behaviour is, however, controlled by an option called host_lookup_
5018 order, which defaults to "bydns:byaddr". If people want to use the other
5019 order, or indeed, just use one or the other means of lookup, they can
5020 specify it in this variable.
5022 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5023 some operating systems, this comes back from gethostbyaddr() as an empty
5024 string, and this is what Exim used to test for. However, it seems that in
5025 other systems, "." is yielded. Exim now tests for this case too.
5027 7. The values of check_spool_space and check_log_space are now held internally
5028 as a number of kilobytes instead of an absolute number of bytes. If a
5029 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5030 kilobyte. This means that much larger values can be stored.
5032 8. Exim monitor: an attempt to get the action menu when not actually pointing
5033 at a message produces an empty menu entitled "No message selected". This
5034 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5035 no entries in it ("Shell widget menu has zero width and/or height"). So I
5036 have added a single, blank menu entry in this case.
5038 9. Added ${quote_local_part.
5040 10. MIME decoding is now applied to the contents of Subject: header lines when
5043 11. Now that a reference to $sender_host_address automatically causes a reverse
5044 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5045 host lookup before query-style lookups in lists that might use this
5046 variable. This has therefore been abolished, and the "net-" prefix is no
5047 longer necessary for query-style lookups.
5049 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5050 have been a typo for LFLAGS, so it has been changed.
5052 13. The install script calls Exim with "-C /dev/null" in order to find the
5053 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5054 to be output. Howeve, since Exim outputs its version number before the
5055 error, it didn't break the script. It just looked ugly. I fixed this by
5056 always allowing "-C /dev/null" if the caller is root.
5058 14. Ignore overlarge ACL variable number when reading spool file - insurance
5059 against a later release with more variables having written the file.
5061 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5062 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5063 this, as well as the form without the "IPv6" on the front.
5065 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5066 OS/Makefile-Darwin file.
5068 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5069 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5071 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5072 variable. However, some OS use ints for inodes. I've added cast to long int
5073 to get rid of the compiler warning.
5075 19. I had forgotten to lock out "/../" in configuration file names when
5076 ALT_CONFIG_PREFIX was set.
5078 20. Routers used for verification do not need to specify transports. However,
5079 if such a router generated a host list, and callout was configured, Exim
5080 crashed, because it could not find a port number from the (non-existent)
5081 transport. It now assumes port 25 in this circumstance.
5083 21. Added the -t option to exigrep.
5085 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5086 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5087 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5088 them is set, LOOKUP_LSEARCH is forced.
5090 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5092 24. Added sender and host information to the "rejected by local_scan()" log
5093 line; previously there was no indication of these.
5095 25. Added .include_if_exists.
5097 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5098 Linux. It turns out that not all file systems support this. Apparently some
5099 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5100 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5101 not supported on the file descriptor, is now ignored when Exim is trying to
5102 sync a directory. This applies only to Linux.
5104 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5106 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5107 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5110 29. Added support for maildirsize files from supplied patch (modified a bit).
5112 30. The use of :fail: followed by an empty string could lead Exim to respond to
5113 sender verification failures with (e.g.):
5115 550 Verification failed for <xxx>
5116 550 Sender verify failed
5118 where the first response line was missing the '-' that indicates it is not
5119 the final line of the response.
5121 31. The loop for finding the name of the user that called Exim had a hardwired
5122 limit of 10; it now uses the value of finduser_retries, which is used for
5123 all other user lookups.
5125 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5128 33. Exim was neglecting to zero errno before one call of strtol() when
5129 expanding a string and expecting an integer value. On some systems this
5130 resulted in spurious "integer overflow" errors. Also, it was casting the
5131 result into an int without checking.
5133 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5134 did not work. The code looks as if it has *never* worked, though it appears
5135 to have been documented since at least releast 1.62. I have made it work.
5137 35. The "timeout_DNS" error in retry rules, also documented since at least
5138 1.62, also never worked. As it isn't clear exactly what this means, and
5139 clearly it isn't a major issue, I have abolished the feature by treating it
5140 as "timeout", and writing a warning to the main and panic logs.
5142 36. The display of retry rules for -brt wasn't always showing the error code
5145 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5146 timeout_connect_A, timeout_connect_MX.
5148 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5149 to the empty sender.
5151 39. The daemon was not analysing the content of -oX till after it had closed
5152 stderr and disconnected from the controlling terminal. This meant that any
5153 syntax errors were only noted on the panic log, and the return code from
5154 the command was 0. By re-arranging the code a little, I've made the
5155 decoding happen first, so such errors now appear on stderr, and the return
5156 code is 1. However, the actual setting up of the sockets still happens in
5157 the disconnected process, so errors there are still only recorded on the
5160 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5161 connections (as happens on some IP stacks) was logged at start up time as
5162 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5163 it from "IPv6 and IPv4", which means that two separate sockets are being
5166 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5167 says whether AF_INET or AF_INET6 was passed as an argument.
5169 42. Exiwhat output was messed up when time zones were included in log
5172 43. Exiwhat now gives more information about the daemon's listening ports,
5173 and whether -tls-on-connect was used.
5175 44. The "port" option of the smtp transport is now expanded.
5177 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5178 silently ignored. Now an error message is written to the main and panic
5181 46. There's a new ACL modifier called "logwrite" which writes to a log file
5182 as soon as it is encountered.
5184 47. Added $local_user_uid and $local_user_gid at routing time.
5186 48. Exim crashed when trying to verify a sender address that was being
5189 49. Exim was recognizing only a space character after ".include". It now also
5190 recognizes a tab character.
5192 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5193 extracting the relevant information from the specification. The man page no
5194 longer contains scrambled data for the -d option, and I've added a section
5195 at the front about calling Exim under different names.
5197 51. Added "extra_headers" argument to the "mail" command in filter files.
5199 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5202 53. Installed eximstats 1.29.
5204 54. Added transport_filter_timeout as a generic transport option.
5206 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5207 Cc: header lines. This was required by RFC 822, but it not required by RFC
5210 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5211 incoming messages that did not have them. Now it does so only if the
5212 message originates locally, that is, if there is no associated remote host
5213 address. When Resent- header lines are present, this applies to the Resent-
5214 lines rather than the non-Resent- lines.
5216 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5217 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5219 58. Messages for configuration errors now include the name of the main
5220 configuration file - useful now that there may be more than one file in a
5221 list (.included file names were always shown).
5223 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5224 for those rare installations that do not start the daemon as root or run it
5225 setuid root. I've cut out the call to initgroups() if the daemon is not
5228 60. The Exim user and group can now be bound into the binary as text strings
5229 that are looked up at the start of Exim's processing.
5231 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5233 62. Added $mailstore_basename variable.
5235 63. Installed patch to sieve.c from Michael Haardt.
5237 64. When Exim failed to open the panic log after failing to open the main log,
5238 the original message it was trying to log was written to stderr and debug
5239 output, but if they were not available (the usual case in production), it
5240 was lost. Now it is written to syslog before the two lines that record the
5241 failures to open the logs.
5243 65. Users' Exim filters run in subprocesses under the user's uid. It is
5244 possible for a "deliver" command or an alias in a "personal" command to
5245 provoke an address rewrite. If logging of address rewriting is configured,
5246 this fails because the process is not running as root or exim. There may be
5247 a better way of dealing with this, but for the moment (because 4.30 needs
5248 to be released), I have disabled address rewrite logging when running a
5249 filter in a non-root, non-exim process.
5255 1. The buildconfig auxiliary program wasn't quoting the value set for
5256 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5257 not defined. This bug was masked in 4.22 by the effect that was fixed in
5260 2. Some messages that were rejected after a message id was allocated were
5261 shown as "incomplete" by exigrep. It no longer does this for messages that
5262 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5264 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5265 have allow_domain_literals set, the ID did not get logged in the <= line.
5266 Domain literals are now always recognized in Message-ID: header lines.
5268 4. The first argument for a ${extract expansion item is the key name or field
5269 number. Leading and trailing spaces in this item were not being ignored,
5270 causing some misleading effects.
5272 5. When deliver_drop_privilege was set, single queue runner processes started
5273 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5274 same command in the process it spins off) were not dropping privilege.
5276 6. When the daemon running as "exim" started a queue runner, it always
5277 re-executed Exim in the spun-off process. This is a waste of effort when
5278 deliver_drop_privilege is set. The new process now just calls the
5279 queue-runner function directly.
5285 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5288 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5289 ignored. Though the use of -oP was forcing the writing of a pid file, it
5290 was always written to the default place.
5292 3. If the message "no IP address found for host xxxx" is generated during
5293 incoming verification, it is now followed by identification of the incoming
5294 connection (so you can more easily find what provoked it).
5296 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5298 5. Added some features to "harden" Exim a bit more against certain attacks:
5300 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5301 be put in Local/Makefile. This is like the never_users runtime option,
5302 but it cannot be overridden. The default setting is "root".
5304 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5305 prefix string with which any file named in a -C command line option
5308 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5309 is retained for -C and -D only if the caller of Exim is root. Without
5310 it, the exim user may also use -C and -D and retain privilege.
5312 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5313 command line option is disabled.
5315 6. Macro names set by the -D option must start with an upper case letter, just
5316 like macro names defined in the configuration file.
5318 7. Added "dereference=" facility to LDAP.
5320 8. Two instances of the typo "uknown" in the source files are fixed.
5322 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5323 the Configure-Makefile script screwed up while processing it.
5325 10. Incorporated PCRE 4.4.
5327 11. The SMTP synchronization check was not operating right at the start of an
5328 SMTP session. For example, it could not catch a HELO sent before the client
5329 waited for the greeting. There is now a check for outstanding input at the
5330 point when the greeting is written. Because of the duplex, asynchronous
5331 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5332 way, but not yet received, when the check is performed.
5334 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5335 on TCP/IP sockets, because this apparently causes some broken clients to
5338 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5339 unchanged) from the Cygwin maintainer.
5341 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5342 was supported instead of testing for mbx. Effectively a typo.
5344 15. The spa authenticator server code was not checking that the input it
5345 received was valid base64.
5347 16. The debug output line for the "set" modifier in ACLs was not showing the
5348 name of the variable that was being set.
5350 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5352 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5353 lookup on its own. Something else was needed to trigger the lookup. For
5354 example, a match in host_lookup or the need for a host name in a host list.
5355 Now, if $sender_host_name is referenced and the host name has not yet been
5356 looked up, a lookup is performed. If the lookup fails, the variable remains
5357 empty, and $host_lookup_failed is set to "1".
5359 19. Added "eqi" as a case-independent comparison operator.
5361 20. The saslauthd authentication condition could segfault if neither service
5362 nor realm was specified.
5364 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5365 error message that was logged was misleading, and incoming SMTP
5366 connections were dropped. The message is now more accurate, and temporary
5367 errors are given to SMTP connections.
5369 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5370 (except RSET) with a 421 error, until QUIT is received. However, it was
5371 failing to send a response to QUIT.
5373 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5374 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5375 ACL is now run after the helo_try_verify_hosts code.
5377 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5378 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5379 case-independent, so other case variants are also recognized). Apparently
5380 some systems use these upper case variants.
5382 25. If more than two messages were waiting for the same host, and a transport
5383 filter was specified for the transport, Exim sent two messages over the
5384 same TCP/IP connection, and then failed with "socket operation on non-
5385 socket" when it tried to send the third.
5387 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5389 27. The extern definition of crypt16() in expand.c was not being excluded when
5390 the OS had its own crypt16() function.
5392 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5393 as a preferred synonym for return_size_limit, both as an option and as an
5396 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5398 30. Changed the default configuration ACL to relax the local part checking rule
5399 for addresses that are not in any local domains. For these addresses,
5400 slashes and pipe symbols are allowed within local parts, but the sequence
5401 /../ is explicitly forbidden.
5403 31. SPA server authentication was not clearing the challenge buffer before
5406 32. log_message in a "warn" ACL statement was writing to the reject log as
5407 well as to the main log, which contradicts the documentation and doesn't
5408 seem right (because no rejection is happening). So I have stopped it.
5410 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5411 However, I am unable to do any testing of this.
5413 34. Fixed an infelicity in the appendfile transport. When checking directories
5414 for a mailbox, to see if any needed to be created, it was accidentally
5415 using path names with one or more superfluous leading slashes; tracing
5416 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5418 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5419 contents are added to the log line that is written for every discarded
5420 recipient. (Previously a log_message setting was ignored.)
5422 36. The ${quote: operator now quotes the string if it is empty.
5424 37. The install script runs exim in order to find its version number. If for
5425 some reason other than non-existence or emptiness, which it checks, it
5426 could not run './exim', it was installing it with an empty version number,
5427 i.e. as "exim-". This error state is now caught, and the installation is
5430 38. An argument was missing from the function that creates an error message
5431 when Exim fails to connect to the socket for saslauthd authentication.
5432 This could cause Exim to crash, or give a corrupted message.
5434 39. Added isip, isip4, and isip6 to ${if conditions.
5436 40. The ACL variables $acl_xx are now saved with the message, and can be
5437 accessed later in routers, transports, and filters.
5439 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5440 strings in the file are not string-expanded.
5442 42. If a MAIL command specified a SIZE value that was too large to fit into an
5443 int variable, the check against message_size_limit failed. Such values are
5444 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5445 day this will have to be increased, but I don't think I want to be around
5446 when emails are that large.
5453 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5454 iconv() is not standard in FreeBSD.
5456 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5457 IPv6 enabled. The observed symptom was a segmentation fault on return from
5458 the function os_common_find_running_interfaces() in src/os.c.
5460 3. In the check_special_case() function in daemon.c I had used "errno" as an
5461 argument name, which causes warnings on some systems. This was basically a
5462 typo, since it was named "eno" in the comments!
5464 4. The code that waits for the clock to tick (at a resolution of some fraction
5465 of a second) so as to ensure message-id uniqueness was always waiting for
5466 at least one whole tick, when it could have waited for less. [This is
5467 almost certainly not relevant at current processor speeds, where it is
5468 unlikely to ever wait at all. But we try to future-proof.]
5470 5. The function that sleeps for a time interval that includes fractions of a
5471 second contained a race. It did not block SIGALRM between setting the
5472 timer, and suspending (a couple of lines later). If the interval was short
5473 and the sigsuspend() was delayed until after it had expired, the suspension
5474 never ended. On busy systems this could lead to processes getting stuck for
5477 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5478 process, before it forks any delivery processes. The open lookup caching
5479 mechanism meant that the open file or database connection was passed into
5480 the delivery process. The problem was that delivery processes always tidy
5481 up cached lookup data. This could cause a problem for the next delivery
5482 process started by the queue runner, because the external queue runner
5483 process does not know about the closure. So the next delivery process
5484 still has data in the lookup cache. In the case of a file lookup, there was
5485 no problem because closing a file descriptor in a subprocess doesn't affect
5486 the parent. However, if the lookup was caching a connection to a database,
5487 the connection was closed, and the second delivery process was likely to
5488 see errors such as "PGSQL: query failed: server closed the connection
5489 unexpectedly". The problem has been fixed by closing all cached lookups
5490 in a queue runner before running a delivery process.
5492 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5493 seem to have the "const" qualifier which it has on other OS. I've
5496 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5497 *of the same type* (client or server) with the same public name that an
5498 error should be diagnosed.
5500 9. When Exim looked up a host name for an IP address, but failed to find the
5501 original IP address when looking up the host name (a safety check), it
5502 output the message "<ip address> does not match any IP for NULL", which was
5503 confusing, to say the least. The bug was that the host name should have
5504 appeared instead of "NULL".
5506 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5507 user that is built into the binary, and the -C or -D options is used, root
5508 privilege is dropped before the configuration file is read. In addition,
5509 logging is switched to stderr instead of the normal log files. If the
5510 configuration then re-defines the Exim user, the unprivileged environment
5511 is probably not what is expected, so Exim logs a panic warning message (but
5514 However, if deliver_drop_privilege is set, the unprivileged state may well
5515 be exactly what is intended, so the warning has been cut out in that case,
5516 and Exim is allowed to try to write to its normal log files.
5522 1. smtp_return_error_details was not giving details for temporary sender
5523 or receiver verification errors.
5525 2. Diagnose a configuration error if two authenticators have the same public
5528 3. Exim used not to create the message log file for a message until the first
5529 delivery attempt. This could be confusing when incoming messages were held
5530 for policy or load reasons. The message log file is now created at the time
5531 the message is received, and an initial "Received" line is written to it.
5533 4. The automatically generated man page for command line options had a minor
5534 bug that caused no ill effects; however, a more serious problem was that
5535 the procedure for building the man page automatically didn't always
5536 operate. Consequently, release 4.20 contains an out-of-date version. This
5537 shouldn't happen again.
5539 5. When building Exim with embedded Perl support, the script that builds the
5540 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5541 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5543 6. The freeze_tell option was not being used for messages that were frozen on
5544 arrival, either by an ACL or by local_scan().
5546 7. Added the smtp_incomplete_transaction log selector.
5548 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5549 was accepting AUTH without a new EHLO.
5551 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5552 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5555 10. Logging of TCP/IP connections (when configured) now happens in the main
5556 daemon process instead of the child process, so that the TCP/IP connection
5557 count is more accurate (but it can never be perfect).
5559 11. The use of "drop" in a nested ACL was not being handled correctly in the
5560 outer ACL. Now, if condition failure induced by the nested "drop" causes
5561 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5562 or "require"), the connection is dropped.
5564 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5565 that yield "discard" can now be used with an "accept" or a "discard" verb,
5566 but an error is generated for any others (because I can't see a useful way
5567 to define what should happen).
5569 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5570 are now processed in the same way as lines in the Exim configuration file.
5571 In particular, continuation lines are supported.
5573 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5575 15. Added -ti meaning -t -i.
5577 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5578 domains, and warn by logging if others are found.
5580 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5581 returned from GIFCONF when Exim is trying to find the list of interfaces on
5582 a host. The code in os.c has been modified to copy each ifreq to an aligned
5583 structure in all cases.
5585 Also, in some cases, the returned ifreq's were being copied to a 'struct
5586 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5587 means the last couple of bytes of an IPv6 address could be chopped if the
5588 ifreq contained only a normal sockaddr (14 bytes storage).
5590 18. Named domain lists were not supported in the hosts_treat_as_local option.
5591 An entry such as +xxxx was not recognized, and was treated as a literal
5594 19. Ensure that header lines added by a DATA ACL are included in the reject log
5595 if the ACL subsequently rejects the message.
5597 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5598 MD5 (which is deprecated).
5600 21. When testing a filter file using -bf, Exim was writing a message when it
5601 took the sender from a "From " line in the message, but it was not doing so
5602 when it took $return_path from a Return-Path: header line. It now does.
5604 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5605 with a valid header line field name (a series of printing characters
5606 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5608 23. Changed "disc" in the source to "disk" to conform to the documentation and
5609 the book and for uniformity.
5611 24. Ignore Sendmail's -Ooption=value command line item.
5613 25. When execve() failed while trying to run a command in a pipe transport,
5614 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5615 could be confused with a return value of 69 from the command itself. This
5616 has been changed to 127, the value the shell returns if it is asked to run
5617 a non-existent command. The wording for the related log line suggests a
5618 non-existent command as the problem.
5620 26. If received_header_text expands to an empty string, do not add a Received:
5621 header line to the message. (Well, it adds a token one on the spool, but
5622 marks it "old" so that it doesn't get used or transmitted.)
5624 27. Installed eximstats 1.28 (addition of -nt option).
5626 28. There was no check for failure on the call to getsockname() in the daemon
5627 code. This can fail if there is a shortage of resources on the system, with
5628 ENOMEM, for example. A temporary error is now given on failure.
5630 29. Contrary to the C standard, it seems that in some environments, the
5631 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5632 program. Exim now does this explicitly; it affects the formatting of
5633 timestamps using strftime().
5635 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5636 complaints. I've now initialized all the variables, to avoid this.
5638 32. Header lines added by a system filter were not being "seen" during
5639 transport-time rewrites.
5641 33. The info_callback() function passed to OpenSSL is set up with type void
5642 (*)(SSL *, int, int), as described somewhere. However, when calling the
5643 function (actually a macro) that sets it up, the type void(*)() is
5644 expected. I've put in a cast to prevent warnings from picky compilers.
5646 34. If a DNS black list lookup found a CNAME record, but there were no A
5647 records associated with the domain it pointed at, Exim crashed.
5649 35. If a DNS black list lookup returned more than one A record, Exim ignored
5650 all but the first. It now scans all returned addresses if a particular IP
5651 value is being sought. In this situation, the contents of the
5652 $dnslist_value variable are a list of all the addresses, separated by a
5655 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5656 to accept a host name and all its aliases if the forward lookup for any of
5657 them yielded the IP address of the incoming connection. Now it accepts only
5658 those names whose forward lookup yields the correct IP address. Any other
5659 names are discarded. This closes a loophole whereby a rogue DNS
5660 administrator could create reverse DNS records to break through a
5661 wildcarded host restriction in an ACL.
5663 37. If a user filter or a system filter that ran in a subprocess used any of
5664 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5665 the wrong values were passed to the pipe command ($thisaddress had the
5666 value of $0, $0 had the value of $1, etc). This bug was introduced by
5667 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5669 38. Improved the line breaking for long SMTP error messages from ACLs.
5670 Previously, if there was no break point between 40 and 75 characters, Exim
5671 left the rest of the message alone. Two changes have been made: (a) I've
5672 reduced the minimum length to 35 characters; (b) if it can't find a break
5673 point between 35 and 75 characters, it looks ahead and uses the first one
5674 that it finds. This may give the occasional overlong line, but at least the
5675 remaining text gets split now.
5677 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5678 file descriptors might be low, and that setting 1000 would always raise it.
5679 It turns out that in some environments, the limit is already over 1000 and
5680 that lowering it causes trouble. So now Exim takes care not to decrease it.
5682 40. When delivering a message, the value of $return_path is set to $sender_
5683 address at the start of routing (routers may change the value). By an
5684 oversight, this default was not being set up when an address was tested by
5685 -bt or -bv, which affected the outcome if any router or filter referred to
5688 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5689 log only when "message" or "log_message" are set. However, if one of the
5690 conditions was an address verification, or a call to a nested ACL, the
5691 messages generated by the underlying test were being passed through. This
5692 no longer happens. The underlying message is available in $acl_verify_
5693 message for both "message" and "log_message" expansions, so it can be
5694 passed through if needed.
5696 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5697 new expansion $bh_ to give the encoded byte string without charset
5698 translation. Translation happens only if iconv() is available; HAVE_ICONV
5699 indicates this at build time. HEADERS_CHARSET gives the charset to
5700 translate to; headers_charset can change it in the configuration, and
5701 "headers charset" can change it in an individual filter file.
5703 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5704 that creates RFC 2047 encoded "words" labels them as that charset instead
5705 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5706 expansion operator; (ii) when Exim creates a From: line for a local
5707 message; (iii) when a header line is rewritten to include a "phrase" part.
5709 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5710 buggy, causing it to skip the first lines of messages whose message ID
5711 ended in 'D'. This would not have bitten before Exim release 4.14, because
5712 message IDs were unlikely to end in 'D' before then. The effect was to have
5713 incorrect size information for certain domains.
5715 45. #include "config.h" was missing at the start of the crypt16.c module. This
5716 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5719 46. If there was a timeout during a "random" callout check, Exim treated it as
5720 a failure of the random address, and carried on sending RSET and the real
5721 address. If the delay was just some slowness somewhere, the response to the
5722 original RCPT would be taken as a response to RSET and so on, causing
5723 mayhem of various kinds.
5725 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5726 when I implemented it. It didn't allow for the fact that some option values
5727 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5728 the right test for positive values.
5730 48. Domain names in DNS records are case-independent. Exim always looks them up
5731 in lower case. Some resolvers return domain names in exactly the case they
5732 appear in the zone file, that is, they may contain uppercase letters. Not
5733 all resolvers do this - some return always lower case. Exim was treating a
5734 change of case by a resolver as a change of domain, similar to a widening
5735 of a domain abbreviation. This triggered its re-routing code and so it was
5736 trying to route what was effectively the same domain again. This normally
5737 caused routing to fail (because the router wouldn't handle the domain
5738 twice). Now Exim checks for this case specially, and just changes the
5739 casing of the domain that it ultimately uses when it transmits the message
5742 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5745 50. If a filter generated a file delivery with a non-absolute name (possible if
5746 no home directory exists for the router), the forbid_file option was not
5749 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5750 the existing equality matching.
5752 52. Exim was using ints instead of ino_t variables in some places where it was
5753 dealing with inode numbers.
5755 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5756 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5757 called TMPDIR, and if it finds it is different, it changes its value.
5759 54. The smtp_printf() function is now made available to local_scan() so
5760 additional output lines can be written before returning. There is also an
5761 smtp_fflush() function to enable the detection of a dropped connection.
5762 The variables smtp_input and smtp_batched_input are exported to
5765 55. Changed the default runtime configuration: the message "Unknown user"
5766 has been removed from the ACL, and instead placed on the localuser router,
5767 using the cannot_route_message feature. This means that any verification
5768 failures that generate their own messages won't get overridden. Similarly,
5769 the "Unrouteable address" message that was in the ACL for unverifiable
5770 relay addresses has also been removed.
5772 56. Added hosts_avoid_esmtp to the smtp transport.
5774 57. The exicyclog script was not checking for the esoteric option
5775 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5776 will work only if exicyclog is run under the appropriate euid.
5778 58. Following a discussion on the list, the rules by which Exim recognises line
5779 endings on incoming messages have been changed. The -dropcr and drop_cr
5780 options are now no-ops, retained only for backwards compatibility. The
5781 following line terminators are recognized: LF CRLF CR. However, special
5782 processing applies to CR:
5784 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5785 nor a local message in the state where . is a terminator.
5787 (ii) If a bare CR is encountered in a header line, an extra space is added
5788 after the line terminator so as not to end the header. The reasoning
5789 behind this is that bare CRs in header lines are most likely either
5790 to be mistakes, or people trying to play silly games.
5792 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5793 was being incorrectly given as 18 bytes larger than it should have been.
5794 This is a VOB (very old bug).
5796 60. This may never have affected anything current, but just in case it has:
5797 When the local host is found other than at the start of a list of hosts,
5798 the local host, those with the same MX, and any that follow, are discarded.
5799 When the list in question was part of a longer list of hosts, the following
5800 hosts (not currently being processed) were also being discarded. This no
5801 longer happens. I'm not sure if this situation could ever has previously
5804 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5807 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5808 header. This is something that is recommended in a new Internet Draft, and
5809 is something that is documented as being done by Sendmail. There are two
5810 possible values. For messages generated by the autoreply transport, Exim
5813 Auto-Submitted: auto-replied
5815 whereas for all other generated messages (e.g. bounces) it adds
5817 Auto-Submitted: auto-generated
5819 63. The "personal" condition in filters now includes a test for the
5820 Auto-Submitted: header. If it contains the string "auto-" the message it
5821 not considered personal.
5823 64. Added rcpt_include_affixes as a generic transport option.
5825 65. Added queue_only_override (default true).
5827 66. Added the syslog_duplication option.
5829 67. If what should have been the first header line of a message consisted of
5830 a space followed by a colon, Exim was mis-interpreting it as a header line.
5831 It isn't of course - it is syntactically invalid and should therefore be
5832 treated as the start of the message body. The misbehaviour could have
5833 caused a number of strange effects, including loss of data in subsequent
5834 header lines, and spool format errors.
5836 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5837 client host had authenticated. This control can now be exercised by an ACL
5838 for more flexibility.
5840 69. By default, callouts do not happen when testing with -bh. There is now a
5841 variant, -bhc, which does actually run the callout code, including
5842 consulting and updating the callout cache.
5844 70. Added support for saslauthd authentication, courtesy of Alexander
5847 71. If statvfs() failed on the spool or log directories while checking their
5848 size for availability, Exim confusingly gave the error "space shortage".
5849 Furthermore, in debugging mode it crashed with a floating point exception.
5850 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5851 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5852 serious problem, Exim now writes to the main and panic logs when this
5853 happens, with details of the failure. It then refuses to accept the
5854 incoming message, giving the message "spool directory problem" or "log
5855 directory problem" with a 421 code for SMTP messages.
5857 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5858 0, 1, and 2 exist, because some OS complain for execs without them (see
5859 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5860 descriptors. However, the code omitted to check that the open succeeded,
5861 causing mysterious errors if for some reason the permissions on /dev/null
5862 got screwed. Now Exim writes a message to the main and panic logs, and
5863 bombs out if it can't open /dev/null.
5865 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5866 interact so that it is all more flexible. It is supposed to remain
5867 backwards compatible. Also added extra_local_interfaces.
5869 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5870 to bomb out with an assertion failure - to the client this appears as a
5871 connection drop. This problem occurs in the part of the code that was taken
5872 from the Samba project. Fortunately, the assertion is in a very simple
5873 function, so I have fixed this by reproducing the function inline in the
5874 one place where it is called, and arranging for authentication to fail
5875 instead of killing the process with assert().
5877 75. The SPA client code was not working when the server requested OEM rather
5878 than Unicode encoding.
5880 76. Added code to make require_files with a specific uid setting more usable in
5881 the case where statting the file as root fails - usually a non-root-mounted
5882 NFS file system. When this happens and the failure is EACCES, Exim now
5883 forks a subprocess and does the per-uid checking as the relevant uid.
5885 77. Added process_log_path.
5887 78. If log_file_path was not explicitly set, a setting of check_log_space or
5888 check_log_inodes was ignored.
5890 79. If a space check for the spool or log partitions fails, the incident is now
5891 logged. Of course, in the latter case the data may get lost...
5893 80. Added the %p formatting code to string_format() so that it can be used to
5894 print addresses in debug_print(). Adjusted all the address printing in the
5895 debugging in store.c to use %p rather than %d.
5897 81. There was a concern that a line of code in smtp_in.c could overflow a
5898 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5899 initially expressed, the concern was not well-founded, because trailing
5900 spaces are removed early. However, if the trailing spaces were followed by
5901 a NULL, they did not get removed, so the overflow was possible. Two fixes
5904 (a) I re-wrote the offending code in a cleaner fashion.
5905 (b) If an incoming SMTP command contains a NULL character, it is rejected
5908 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5909 runs initgroups(), so that if the Exim user is in any additional groups,
5910 they will be used during message reception.
5916 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob