1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
40 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
41 checkers that were complaining about end of non-void function with no
44 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
45 This was a regression intruduced in 4.83 by another bugfix.
47 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
49 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
50 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
56 TF/01 Correctly close the server side of TLS when forking for delivery.
58 When a message was received over SMTP with TLS, Exim failed to clear up
59 the incoming connection properly after forking off the child process to
60 deliver the message. In some situations the subsequent outgoing
61 delivery connection happened to have the same fd number as the incoming
62 connection previously had. Exim would try to use TLS and fail, logging
63 a "Bad file descriptor" error.
65 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
66 utilities have not been installed.
68 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
69 temporary space as the ACL may create new global variables.
71 TL/01 LDAP support uses per connection or global context settings, depending
72 upon the detected version of the libraries at build time.
74 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
75 to extract and use the src ip:port in logging and expansions as if it
76 were a direct connection from the outside internet. PPv2 support was
77 updated based on HAProxy spec change in May 2014.
79 JH/02 Add ${listextract {number}{list}{success}{fail}}.
81 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
82 Properly escape header and check for NULL return.
84 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
87 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
89 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
90 characters in header names, implemented as a verify condition.
91 Contributed by Michael Fischer v. Mollard.
93 TL/05 Rename SPF condition results err_perm and err_temp to standardized
94 results permerror and temperror. Previous values are deprecated but
95 still accepted. In a future release, err_perm and err_temp will be
96 completely removed, which will be a backward incompatibility if the
97 ACL tests for either of these two old results. Patch contributed by
98 user bes-internal on the mailing list.
100 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
102 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
103 selectors, in both main and reject logs.
105 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
108 JH/07 Add malware type "sock" for talking to simple daemon.
110 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
112 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
113 routers/transports under cutthrough routing.
115 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
116 numbers. Touch up "bool" conditional to keep the same definition.
118 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
120 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
122 JH/12 Expand items in router/transport headers_add or headers_remove lists
123 individually rather than the list as a whole. Bug 1452.
125 Required for reasonable handling of multiple headers_ options when
126 they may be empty; requires that headers_remove items with embedded
127 colons must have them doubled (or the list-separator changed).
129 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
130 view the policy declared in the DMARC record. Currently, $dmarc_status
131 is a combined value of both the record presence and the result of the
134 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
136 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
137 dnslookup router and the smtp transport (applying to the forward
140 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
141 of ldap servers used for a specific lookup. Patch provided by Heiko
144 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
145 New variable $lookup_dnssec_authenticated for observability.
147 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
148 Patch submitted by Lars Timman.
150 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
152 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
153 Requires trusted mode and valid format message id, aborts otherwise.
154 Patch contributed by Heiko Schlichting.
156 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
157 certextract with support for various fields. Bug 1358.
159 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
160 is requested by default, modifiable by smtp transport option
163 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
164 operate on certificate variables to give certificate fingerprints
165 Also new ${sha256:cert_variable}.
167 JH/23 The PRDR feature is moved from being Experimental into the mainline.
169 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
170 Christian Aistleitner.
172 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
174 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
175 file. Patch from Wolfgang Breyha.
177 JH/25 Expand the coverage of the delivery $host and $host_address to
178 client authenticators run in verify callout. Bug 1476.
180 JH/26 Port service names are now accepted for tls_on_connect_ports, to
181 align with daemon_smtp_ports. Bug 72.
183 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
184 support and error reporting did not work properly.
186 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
187 and is readable. Patch from Andrew Colin Kissa.
189 TL/14 Enhance documentation of ${run expansion and how it parses the
190 commandline after expansion, particularly in the case when an
191 unquoted variable expansion results in an empty value.
193 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
195 PP/02 Fix internal collision of T_APL on systems which support RFC3123
196 by renaming away from it. Addresses GH issue 15, reported by
199 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
201 TL/15 SECURITY: prevent double expansion in math comparison functions
202 (can expand unsanitized data). Not remotely exploitable.
209 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
211 PP/02 Make -n do something, by making it not do something.
212 When combined with -bP, the name of an option is not output.
214 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
217 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
218 $sender_host_name and config options to manage this, and basic check
221 PP/05 DSCP support for outbound connections and control modifier for inbound.
223 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
224 (Only plugin which currently uses this is kerberos4, which nobody should
225 be using, but we should make it available and other future plugins might
226 conceivably use it, even though it would break NAT; stuff *should* be
227 using channel bindings instead).
229 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
230 name; added for Sendmail compatibility; requires admin caller.
231 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
232 just ignore it); requires trusted caller.
233 Also parse but ignore: -Ac -Am -X<logfile>
236 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
238 TL/02 Add +smtp_confirmation as a default logging option.
240 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
241 Patch by Magnus Holmgren from 2007-02-20.
243 TL/04 Bugzilla 1281 - Spec typo.
244 Bugzilla 1283 - Spec typo.
245 Bugzilla 1290 - Spec grammar fixes.
247 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
249 TL/06 Add Experimental DMARC support using libopendmarc libraries.
251 TL/07 Fix an out of order global option causing a segfault. Reported to dev
252 mailing list by by Dmitry Isaikin.
254 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
256 JH/02 Support "G" suffix to numbers in ${if comparisons.
258 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
260 NM/01 Bugzilla 1197 - Spec typo
261 Bugzilla 1196 - Spec examples corrections
263 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
265 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
266 gnutls_enable_pkcs11, but renamed to more accurately indicate its
269 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
270 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
272 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
273 "acl {{name}{arg}...}", and optional args on acl condition
276 JH/05 Permit multiple router/transport headers_add/remove lines.
278 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
280 JH/07 Avoid using a waiting database for a single-message-only transport.
281 Performance patch from Paul Fisher. Bugzilla 1262.
283 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
286 JH/09 Add $headers_added variable, with content from use of ACL modifier
287 add_header (but not yet added to the message). Bugzilla 199.
289 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
290 Pulled from Bugzilla 817 by Wolfgang Breyha.
292 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
294 (nb: this is the same fix as in Exim 4.80.1)
296 JH/11 Add A= logging on delivery lines, and a client_set_id option on
299 JH/12 Add optional authenticated_sender logging to A= and a log_selector
302 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
304 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
305 advertise SMTP AUTH mechanism to us, instead of a generic
306 protocol violation error. Also, make Exim more robust to bad
307 data from the Dovecot auth socket.
309 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
311 When a queue runner is handling a message, Exim first routes the
312 recipient addresses, during which it prunes them based on the retry
313 hints database. After that it attempts to deliver the message to
314 any remaining recipients. It then updates the hints database using
317 So if a recipient address works intermittently, it can get repeatedly
318 deferred at routing time. The retry hints record remains fresh so the
319 address never reaches the final cutoff time.
321 This is a fairly common occurrence when a user is bumping up against
322 their storage quota. Exim had some logic in its local delivery code
323 to deal with this. However it did not apply to per-recipient defers
324 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
326 This change adds a proper retry rule check during routing so that the
327 final cutoff time is checked against the message's age. We only do
328 this check if there is an address retry record and there is not a
329 domain retry record; this implies that previous attempts to handle
330 the address had the retry_use_local_parts option turned on. We use
331 this as an approximation for the destination being like a local
332 delivery, as in LMTP.
334 I suspect this new check makes the old local delivery cutoff check
335 redundant, but I have not verified this so I left the code in place.
337 TF/02 Correct gecos expansion when From: is a prefix of the username.
339 Test 0254 submits a message to Exim with the header
343 When I ran the test suite under the user fanf2, Exim expanded
344 the header to contain my full name, whereas it should have added
345 a Resent-Sender: header. It erroneously treats any prefix of the
346 username as equal to the username.
348 This change corrects that bug.
350 GF/01 DCC debug and logging tidyup
351 Error conditions log to paniclog rather than rejectlog.
352 Debug lines prefixed by "DCC: " to remove any ambiguity.
354 TF/03 Avoid unnecessary rebuilds of lookup-related code.
356 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
357 Bug spotted by Jeremy Harris; was flawed since initial commit.
358 Would have resulted in OCSP responses post-SNI triggering an Exim
359 NULL dereference and crash.
361 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
363 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
364 Bug detection, analysis and fix by Samuel Thibault.
365 Bugzilla 1331, Debian bug #698092.
367 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
369 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
370 Server implementation by Todd Lyons, client by JH.
371 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
372 config variable "prdr_enable" controls whether the server
373 advertises the facility. If the client requests PRDR a new
374 acl_data_smtp_prdr ACL is called once for each recipient, after
375 the body content is received and before the acl_smtp_data ACL.
376 The client is controlled by bolth of: a hosts_try_prdr option
377 on the smtp transport, and the server advertisement.
378 Default client logging of deliveries and rejections involving
379 PRDR are flagged with the string "PRDR".
381 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
382 fclose(). Diagnosis by Todd Lyons.
384 PP/17 Update configure.default to handle IPv6 localhost better.
385 Patch by Alain Williams (plus minor tweaks).
388 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
389 This is now consistent with GnuTLS, and is now documented: the
390 previous undocumented portable approach to treating the option as
391 unset was to force an expansion failure. That still works, and
392 an empty string is now equivalent.
394 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
395 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
396 not performing validation itself.
398 PP/20 Added force_command boolean option to pipe transport.
399 Patch from Nick Koston, of cPanel Inc.
401 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
404 TF/04 Added udpsend ACL modifer and hexquote expansion operator
406 PP/21 Fix eximon continuous updating with timestamped log-files.
407 Broken in a format-string cleanup in 4.80, missed when I repaired the
408 other false fix of the same issue.
409 Report and fix from Heiko Schlichting.
412 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
413 Report from Prashanth Katuri.
415 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
416 It's SecureTransport, so affects any MacOS clients which use the
417 system-integrated TLS libraries, including email clients.
419 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
420 using a MIME ACL for non-SMTP local injection.
421 Report and assistance in diagnosis by Warren Baker.
423 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
425 JH/16 Fix comparisons for 64b. Bugzilla 1385.
427 TL/09 Add expansion variable $authenticated_fail_id to keep track of
428 last id that failed so it may be referenced in subsequent ACL's.
430 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
433 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
434 ldap library initialization, allowing self-signed CA's to be
435 used. Also properly sets require_cert option later in code by
436 using NULL (global ldap config) instead of ldap handle (per
437 session). Bug diagnosis and testing by alxgomz.
439 TL/12 Enhanced documentation in the ratelimit.pl script provided in
440 the src/util/ subdirectory.
442 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
443 renamed to Transport Post Delivery Action by Jeremy Harris, as
446 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
447 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
448 redis_servers = needs to be configured which will be used by the redis
449 lookup. Patch from Warren Baker, of The Packet Hub.
451 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
453 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
454 hostname or reverse DNS when processing a host list. Used suggestions
455 from multiple comments on this bug.
457 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
459 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
460 Missed a few lines, added it to make the runtest require no keyboard
463 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
464 contains upper case chars. Make router use caseful_local_part.
466 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
467 support when GnuTLS has been built with p11-kit.
473 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
475 This, or similar/improved, will also be change PP/11 of 4.82.
481 PP/01 Handle short writes when writing local log-files.
482 In practice, only affects FreeBSD (8 onwards).
483 Bugzilla 1053, with thanks to Dmitry Isaikin.
485 NM/01 Bugzilla 949 - Documentation tweak
487 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
490 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
492 PP/02 Implemented gsasl authenticator.
494 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
496 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
497 `pkg-config foo` for cflags/libs.
499 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
500 with rest of GSASL and with heimdal_gssapi.
502 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
503 `pkg-config foo` for cflags/libs for the TLS implementation.
505 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
506 properties get this fed in as external SSF. A number of robustness
507 and debugging improvements to the cyrus_sasl authenticator.
509 PP/08 cyrus_sasl server now expands the server_realm option.
511 PP/09 Bugzilla 1214 - Log authentication information in reject log.
512 Patch by Jeremy Harris.
514 PP/10 Added dbmjz lookup type.
516 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
518 PP/12 MAIL args handles TAB as well as SP, for better interop with
519 non-compliant senders.
520 Analysis and variant patch by Todd Lyons.
522 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
523 Bug report from Lars Müller <lars@samba.org> (via SUSE),
524 Patch from Dirk Mueller <dmueller@suse.com>
526 PP/13 tls_peerdn now print-escaped for spool files.
527 Observed some $tls_peerdn in wild which contained \n, which resulted
528 in spool file corruption.
530 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
531 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
532 or write after TLS renegotiation, which otherwise led to messages
535 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
536 as a tracking header (ie: a signed header comes before the signature).
537 Patch from Wolfgang Breyha.
539 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
540 comma-sep list; embedded commas doubled.
542 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
544 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
546 Report and patch from Dmitry Banschikov.
548 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
549 Removed SSL_clear() after SSL_new() which led to protocol negotiation
550 failures. We appear to now support TLS1.1+ with Exim.
552 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
553 lets Exim select keys and certificates based upon TLS SNI from client.
554 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
555 before an outbound SMTP session. New log_selector, +tls_sni.
557 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
558 NULL dereference. Report and patch from Alun Jones.
560 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
561 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
562 Not seeing resolver debug output on NetBSD, but suspect this is a
563 resolver implementation change.
565 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
566 Left warnings. Added "eximon gdb" invocation mode.
568 PP/21 Defaulting "accept_8bitmime" to true, not false.
570 PP/22 Added -bw for inetd wait mode support.
572 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
573 locate the relevant includes and libraries. Made this the default.
575 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
576 Bugzilla 1246, report and most of solution from Tomasz Kusy.
578 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
579 This may cause build issues on older platforms.
581 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
582 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
583 gnutls_require_mac & gnutls_require_protocols (no longer supported).
584 Added SNI support via GnuTLS too.
585 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
587 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
589 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
590 Applied second patch from Janne, implementing suggestion to default
591 multiple-strings-in-record handling to match SPF spec.
593 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
595 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
596 read-only, out of scope).
597 Patch from Wolfgang Breyha, report from Stuart Northfield.
599 PP/29 Fix three issues highlighted by clang analyser static analysis.
600 Only crash-plausible issue would require the Cambridge-specific
601 iplookup router and a misconfiguration.
602 Report from Marcin Mirosław.
604 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
606 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
607 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
608 As part of this, removing so much warning spew let me fix some minor
609 real issues in debug logging.
611 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
612 assignment on my part. Fixed.
614 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
615 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
616 Janne Snabb (who went above and beyond: thank you).
618 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
619 string otherwise requires a connection and a bunch more work and it's
620 relatively easy to get wrong. Should also expose TLS library linkage
623 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
624 64-bit ${eval} (JH/03).
626 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
627 GNU libc to support some of the 64-bit stuff, should not lead to
628 conflicts. Defined before os.h is pulled in, so if a given platform
629 needs to override this, it can.
631 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
632 protection layer was required, which is not implemented.
633 Bugzilla 1254, patch from Wolfgang Breyha.
635 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
636 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
637 tls_dhparam take prime identifiers. Also unbreak combination of
638 OpenSSL+DH_params+TLSSNI.
640 PP/39 Disable SSLv2 by default in OpenSSL support.
646 PP/01 Solaris build fix for Oracle's LDAP libraries.
647 Bugzilla 1109, patch from Stephen Usher.
649 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
651 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
654 TF/02 Fix a couple more cases where we did not log the error message
655 when unlink() failed. See also change 4.74-TF/03.
657 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
658 lock up or crash if it happened to be inside a call to libc when it
659 got a SIGUSR1 from exiwhat.
661 The SIGUSR1 handler appends the current process status to the process
662 log which is later printed by exiwhat. It used to use the general
663 purpose logging code to do this, but several functions it calls are
664 not safe for signals.
666 The new output code in the SIGUSR1 handler is specific to the process
667 log, and simple enough that it's easy to inspect for signal safety.
668 Removing some special cases also simplifies the general logging code.
669 Removing the spurious timestamps from the process log simplifies
672 TF/04 Improved ratelimit ACL condition.
674 The /noupdate option has been deprecated in favour of /readonly which
675 has clearer semantics. The /leaky, /strict, and /readonly update modes
676 are mutually exclusive. The update mode is no longer included in the
677 database key; it just determines when the database is updated. (This
678 means that when you upgrde Exim will forget old rate measurements.)
680 Exim now checks that the per_* options are used with an update mode that
681 makes sense for the current ACL. For example, when Exim is processing a
682 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
683 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
684 must specify per_mail/readonly. If you omit the update mode it defaults to
685 /leaky where that makes sense (as before) or /readonly where required.
687 The /noupdate option is now undocumented but still supported for
688 backwards compatibility. It is equivalent to /readonly except that in
689 ACLs where /readonly is required you may specify /leaky/noupdate or
690 /strict/noupdate which are treated the same as /readonly.
692 A useful new feature is the /count= option. This is a generalization
693 of the per_byte option, so that you can measure the throughput of other
694 aggregate values. For example, the per_byte option is now equivalent
695 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
697 The per_rcpt option has been generalized using the /count= mechanism
698 (though it's more complicated than the per_byte equivalence). When it is
699 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
700 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
701 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
702 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
703 Note that using per_rcpt with a non-readonly update mode in more than
704 one ACL will cause the recipients to be double-counted. (The per_mail
705 and per_byte options don't have this problem.)
707 The handling of very low rates has changed slightly. If the computed rate
708 is less than the event's count (usually one) then this event is the first
709 after a long gap. In this case the rate is set to the same as this event's
710 count, so that the first message of a spam run is counted properly.
712 The major new feature is a mechanism for counting the rate of unique
713 events. The new per_addr option counts the number of different
714 recipients that someone has sent messages to in the last time period. It
715 behaves like per_rcpt if all the recipient addresses are different, but
716 duplicate recipient addresses do not increase the measured rate. Like
717 the /count= option this is a general mechanism, so the per_addr option
718 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
719 example, measure the rate that a client uses different sender addresses
720 with the options per_mail/unique=$sender_address. There are further
721 details in the main documentation.
723 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
725 TF/06 Removed a few PCRE remnants.
727 TF/07 Automatically extract Exim's version number from tags in the git
728 repository when doing development or release builds.
730 PP/02 Raise smtp_cmd_buffer_size to 16kB.
731 Bugzilla 879. Patch from Paul Fisher.
733 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
734 Heavily based on revision 40f9a89a from Simon Arlott's tree.
737 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
739 PP/05 Variable $av_failed, true if the AV scanner deferred.
740 Bugzilla 1078. Patch from John Horne.
742 PP/06 Stop make process more reliably on build failure.
743 Bugzilla 1087. Patch from Heiko Schlittermann.
745 PP/07 Make maildir_use_size_file an _expandable_ boolean.
746 Bugzilla 1089. Patch from Heiko Schlittermann.
748 PP/08 Handle ${run} returning more data than OS pipe buffer size.
749 Bugzilla 1131. Patch from Holger Weiß.
751 PP/09 Handle IPv6 addresses with SPF.
752 Bugzilla 860. Patch from Wolfgang Breyha.
754 PP/10 GnuTLS: support TLS 1.2 & 1.1.
756 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
759 PP/11 match_* no longer expand right-hand-side by default.
760 New compile-time build option, EXPAND_LISTMATCH_RHS.
761 New expansion conditions, "inlist", "inlisti".
763 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
765 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
767 PP/14 fix log_write() format string regression from TF/03.
768 Bugzilla 1152. Patch from Dmitry Isaikin.
774 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
776 PP/02 Harmonised TLS library version reporting; only show if debugging.
777 Layout now matches that introduced for other libraries in 4.74 PP/03.
779 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
781 PP/04 New "dns_use_edns0" global option.
783 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
786 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
787 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
789 TK/01 Updated PolarSSL code to 0.14.2.
790 Bugzilla 1097. Patch from Andreas Metzler.
792 PP/07 Catch divide-by-zero in ${eval:...}.
795 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
798 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
799 format-string attack -- SECURITY: remote arbitrary code execution.
801 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
802 time unintentionally subject to list matching rules, letting the header
803 cause arbitrary Exim lookups (of items which can occur in lists, *not*
804 arbitrary string expansion). This allowed for information disclosure.
806 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
807 INT_MIN/-1 -- value coerced to INT_MAX.
813 NM/01 Workround for PCRE version dependency in version reporting
816 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
817 This fixes portability to compilers other than gcc, notably
818 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
820 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
821 makefiles for portability to HP-UX and POSIX correctness.
823 PP/01 Permit LOOKUP_foo enabling on the make command-line.
824 Also via indented variable definition in the Makefile.
825 (Debugging by Oliver Heesakkers).
827 PP/02 Restore caching of spamd results with expanded spamd_address.
828 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
830 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
831 Improves build reliability. Fix from: Frank Elsner
833 NM/02 Fix wide character breakage in the rfc2047 coding
834 Fixes bug 1064. Patch from Andrey N. Oktyabrski
836 NM/03 Allow underscore in dnslist lookups
837 Fixes bug 1026. Patch from Graeme Fowler
839 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
840 Code patches from Adam Ciarcinski of NetBSD.
842 NM/04 Fixed exiqgrep to cope with mailq missing size issue
845 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
846 is logged, to avoid truncation. Patch from John Horne.
848 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
849 Patch from Jakob Hirsch.
851 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
852 SQL string expansion failure details.
853 Patch from Andrey Oktyabrski.
855 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
856 Patch from Simon Arlott.
858 PP/09 New lookups functionality failed to compile on old gcc which rejects
859 extern declarations in function scope.
860 Patch from Oliver Fleischmann
862 PP/10 Use sig_atomic_t for flags set from signal handlers.
863 Check getgroups() return and improve debugging.
864 Fixed developed for diagnosis in bug 927 (which turned out to be
867 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
868 Patch from Mark Zealey.
870 PP/12 Bugzilla 1056: Improved spamd server selection.
871 Patch from Mark Zealey.
873 PP/13 Bugzilla 1086: Deal with maildir quota file races.
874 Based on patch from Heiko Schlittermann.
876 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
877 Patch from Uwe Doering, sign-off by Michael Haardt.
879 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
880 variable declaration deep within a block. Bug and patch from
883 PP/15 lookups-Makefile IRIX compatibilty coercion.
885 PP/16 Make DISABLE_DKIM build knob functional.
887 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
888 Patch by Simon Arlott
890 TF/03 Fix valgrind.h portability to C89 compilers that do not support
891 variable argument macros. Our copy now differs from upstream.
897 TF/01 Failure to get a lock on a hints database can have serious
898 consequences so log it to the panic log.
900 TF/02 Log LMTP confirmation messages in the same way as SMTP,
901 controlled using the smtp_confirmation log selector.
903 TF/03 Include the error message when we fail to unlink a spool file.
905 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
906 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
907 for maintaining out-of-tree patches for some time.
909 PP/01 Bugzilla 139: Documentation and portability issues.
910 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
911 Handle per-OS dynamic-module compilation flags.
913 PP/02 Let /dev/null have normal permissions.
914 The 4.73 fixes were a little too stringent and complained about the
915 permissions on /dev/null. Exempt it from some checks.
916 Reported by Andreas M. Kirchwitz.
918 PP/03 Report version information for many libraries, including
919 Exim version information for dynamically loaded libraries. Created
920 version.h, now support a version extension string for distributors
921 who patch heavily. Dynamic module ABI change.
923 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
924 privilege escalation vulnerability whereby the Exim run-time user
925 can cause root to append content of the attacker's choosing to
928 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
931 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
932 If dropping privileges for untrusted macros, we disabled normal logging
933 on the basis that it would fail; for the Exim run-time user, this is not
934 the case, and it resulted in successful deliveries going unlogged.
935 Fixed. Reported by Andreas Metzler.
941 PP/01 Date: & Message-Id: revert to normally being appended to a message,
942 only prepend for the Resent-* case. Fixes regression introduced in
943 Exim 4.70 by NM/22 for Bugzilla 607.
945 PP/02 Include check_rfc2047_length in configure.default because we're seeing
946 increasing numbers of administrators be bitten by this.
948 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
950 PP/03 Bugzilla 994: added openssl_options main configuration option.
952 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
954 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
956 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
958 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
959 without a peer certificate, leading to a segfault because of an
960 assumption that peers always have certificates. Be a little more
961 paranoid. Problem reported by Martin Tscholak.
963 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
964 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
965 NB: ClamAV planning to remove STREAM in "middle of 2010".
966 CL also introduces -bmalware, various -d+acl logging additions and
967 more caution in buffer sizes.
969 PP/09 Implemented reverse_ip expansion operator.
971 PP/10 Bugzilla 937: provide a "debug" ACL control.
973 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
975 PP/12 Bugzilla 973: Implement --version.
977 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
979 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
981 PP/15 Bugzilla 816: support multiple condition rules on Routers.
983 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
984 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
985 ignore trailing whitespace.
987 JJ/02 prevent non-panic DKIM error from being sent to paniclog
989 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
992 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
993 Notification from Dr Andrew Aitchison.
995 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
996 ExtendedDetectionInfo response format.
997 Notification from John Horne.
999 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1002 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1003 XSL and documented dependency on system catalogs, with examples of how
1006 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1009 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1010 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1011 configuration file which is writeable by the Exim user or group.
1013 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1014 of configuration files to cover files specified with the -C option if
1015 they are going to be used with root privileges, not just the default
1018 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1019 option (effectively making it always true).
1021 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1022 files to be used while preserving root privileges.
1024 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1025 that rogue child processes cannot use them.
1027 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1028 run-time user, instead of root.
1030 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1031 Exim run-time user without dropping privileges.
1033 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1034 result string, instead of calling string_vformat() twice with the same
1037 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1038 for other users. Others should always drop root privileges if they use
1039 -C on the command line, even for a whitelisted configure file.
1041 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1043 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1049 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1050 $data_path, and $header_path variables; fixed documentation bugs and
1053 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1054 exipick to access non-standard spools, including the "frozen" queue
1057 NM/01 Bugzilla 965: Support mysql stored procedures.
1058 Patch from Alain Williams
1060 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1062 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1063 Patch from Andreas Metzler
1065 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1066 Patch from Kirill Miazine
1068 NM/05 Bugzilla 671: Added umask to procmail example.
1070 JJ/03 installed exipick 20100323.0, fixing doc bug
1072 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1073 directory. Notification and patch from Dan Rosenberg.
1075 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1077 TK/02 Improve log output when DKIM signing operation fails.
1079 MH/01 Treat the transport option dkim_domain as a colon separated
1080 list, not as a single string, and sign the message with each element,
1081 omitting multiple occurences of the same signer.
1083 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1084 Bugzilla 985, 986. Patch by Simon Arlott
1086 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1087 Patch by Simon Arlott
1089 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1090 MBX locking. Notification from Dan Rosenberg.
1096 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1098 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1100 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1102 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1104 NM/04 Bugzilla 915: Flag broken perl installation during build.
1110 TK/01 Added patch by Johannes Berg that expands the main option
1111 "spamd_address" if it starts with a dollar sign.
1113 TK/02 Write list of recipients to X-Envelope-Sender header when building
1114 the mbox-format spool file for content scanning (suggested by Jakob
1117 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1118 (http://www.dcc-servers.net/) support via dccifd. Activated by
1119 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1121 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1122 by Mark Daniel Reidel <mr@df.eu>.
1124 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1125 When building exim an external PCRE library is now needed -
1126 PCRE is a system library on the majority of modern systems.
1127 See entry on PCRE_LIBS in EDITME file.
1129 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1130 conversation. Added nologin parameter to request.
1131 Patch contributed by Kirill Miazine.
1133 TF/01 Do not log submission mode rewrites if they do not change the address.
1135 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1137 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1138 log files in place. Contributed by Roberto Lima.
1140 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1142 TF/03 Bugzilla 615: When checking the local_parts router precondition
1143 after a local_part_suffix or local_part_prefix option, Exim now
1144 does not use the address's named list lookup cache, since this
1145 contains cached lookups for the whole local part.
1147 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1148 Robert Millan. Documentation is in experimental-spec.txt.
1150 TF/04 Bugzilla 668: Fix parallel build (make -j).
1152 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1154 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1155 Patch provided by Jan Srzednicki.
1157 TF/05 Leading white space used to be stripped from $spam_report which
1158 wrecked the formatting. Now it is preserved.
1160 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1161 that they are available at delivery time.
1163 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1165 TF/08 TLS error reporting now respects the incoming_interface and
1166 incoming_port log selectors.
1168 TF/09 Produce a more useful error message if an SMTP transport's hosts
1169 setting expands to an empty string.
1171 NM/06 Bugzilla 744: EXPN did not work under TLS.
1172 Patch provided by Phil Pennock.
1174 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1175 Patch provided by Richard Godbee.
1177 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1178 acl_smtp_notquit, added index entry.
1180 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1181 Patch provided by Eugene Bujak.
1183 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1184 accept(). Patch provided by Maxim Dounin.
1186 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1187 Patch provided by Phil Pennock.
1189 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1191 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1192 Patch provided by Brad "anomie" Jorsch.
1194 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1195 Patch provided by Dean Brooks.
1197 TK/05 Add native DKIM support (does not depend on external libraries).
1199 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1200 Patch provided by Graeme Fowler.
1202 NM/16 Bugzilla 851: Documentation example syntax fix.
1204 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1206 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1209 NM/19 Bugzilla 745: TLS version reporting.
1210 Patch provided by Phil Pennock.
1212 NM/20 Bugzilla 167: bool: condition support.
1213 Patch provided by Phil Pennock.
1215 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1216 clients. Patch provided by Phil Pennock.
1218 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1219 Patch provided by Brad "anomie" Jorsch.
1221 NM/23 Bugzilla 687: Fix misparses in eximstats.
1222 Patch provided by Heiko Schlittermann.
1224 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1225 Patch provided by Heiko Schlittermann.
1227 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1228 plus update to original patch.
1230 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1232 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1233 Patch provided by David Brownlee.
1235 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1237 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1239 NM/30 Bugzilla 888: TLS documentation bugfixes.
1241 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1243 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1244 Unlike the original bugzilla I have changed all shell scripts in src tree.
1246 NM/33 Bugzilla 898: Transport filter timeout fix.
1247 Patch by Todd Rinaldo.
1249 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1250 Patch by Serge Demonchaux.
1252 NM/35 Bugzilla 39: Base64 decode bug fixes.
1253 Patch by Jakob Hirsch.
1255 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1257 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1259 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1261 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1267 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1268 ALT-N's libdkim that I have put here:
1269 http://duncanthrax.net/exim-experimental/
1271 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1272 were called 'true' and it seems that C99 defines that as a reserved
1273 keyword to be used with 'bool' variable types. That means you could
1274 not include C99-style headers which use bools without triggering
1275 build errors in sieve.c.
1277 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1278 as mailq or other aliases. Changed the --help handling significantly
1279 to do whats expected. exim_usage() emits usage/help information.
1281 SC/01 Added the -bylocaldomain option to eximstats.
1283 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1285 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1287 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1293 PH/01 Another patch from the Sieve maintainer.
1295 PH/02 When an IPv6 address is converted to a string for single-key lookup
1296 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1297 dots are used instead of colons so that keys in lsearch files need not
1298 contain colons. This was done some time before quoting was made available
1299 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1300 (notated using the quote facility) so as to distinguish them from IPv4
1301 keys. This meant that lookups for IP addresses in host lists did not work
1302 for iplsearch lookups.
1304 This has been fixed by arranging for IPv6 addresses to be expressed with
1305 colons if the lookup type is iplsearch. This is not incompatible, because
1306 previously such lookups could never work.
1308 The situation is now rather anomolous, since one *can* have colons in
1309 ordinary lsearch keys. However, making the change in all cases is
1310 incompatible and would probably break a number of configurations.
1312 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1315 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1316 conversion specification without a maximum field width, thereby enabling
1317 a rogue spamd server to cause a buffer overflow. While nobody in their
1318 right mind would setup Exim to query an untrusted spamd server, an
1319 attacker that gains access to a server running spamd could potentially
1320 exploit this vulnerability to run arbitrary code as the Exim user.
1322 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1323 $primary_hostname instead of what libspf2 thinks the hosts name is.
1325 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1326 a directory entry by the name of the lookup key. Previously, if a
1327 symlink pointed to a non-existing file or a file in a directory that
1328 Exim lacked permissions to read, a lookup for a key matching that
1329 symlink would fail. Now it is enough that a matching directory entry
1330 exists, symlink or not. (Bugzilla 503.)
1332 PH/03 The body_linecount and body_zerocount variables are now exported in the
1335 PH/04 Added the $dnslist_matched variable.
1337 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1338 This means they are set thereafter only if the connection becomes
1341 PH/06 Added the client_condition to authenticators so that some can be skipped
1342 by clients under certain conditions.
1344 PH/07 The error message for a badly-placed control=no_multiline_responses left
1345 "_responses" off the end of the name.
1347 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1349 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1350 (without spaces) instead of just copying the configuration text.
1352 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1354 PH/11 Added $max_received_linelength.
1356 PH/12 Added +ignore_defer and +include_defer to host lists.
1358 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1359 way in which PCRE > 7.0 is built.
1361 PH/14 Implemented queue_only_load_latch.
1363 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1364 MAIL command. The effect was to mangle the value on 64-bit systems.
1366 PH/16 Another patch from the Sieve maintainer.
1368 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1370 PH/18 If a system quota error occurred while trying to create the file for
1371 a maildir delivery, the message "Mailbox is full" was not appended to the
1372 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1373 only to a quota excession during the actual writing of the file.
1375 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1376 characters?) which causes problems in log lines. The DN values are now
1377 passed through string_printing() before being added to log lines.
1379 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1380 and InterBase are left for another time.)
1382 PH/21 Added message_body_newlines option.
1384 PH/22 Guard against possible overflow in moan_check_errorcopy().
1386 PH/23 POSIX allows open() to be a macro; guard against that.
1388 PH/24 If the recipient of an error message contained an @ in the local part
1389 (suitably quoted, of course), incorrect values were put in $domain and
1390 $local_part during the evaluation of errors_copy.
1396 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1397 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1400 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1401 issue a MAIL command.
1403 PH/02 In an ACL statement such as
1405 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1407 if a client was not listed at all, or was listed with a value other than
1408 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1409 the condition was not true (as it should be), so access was not denied.
1410 The bug was that the ! inversion was incorrectly passed on to the second
1411 item. This has been fixed.
1413 PH/03 Added additional dnslists conditions == and =& which are different from
1414 = and & when the dns lookup returns more than one IP address.
1416 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1417 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1419 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1420 FSYNC, which compiles an option called disable_fsync that allows for
1421 bypassing fsync(). The documentation is heavily laced with warnings.
1423 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1425 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1426 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1427 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1428 including adding "make clean"; (3) Added -fPIC when compiling the test
1429 dynamically loaded module, to get rid of a warning.
1431 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1432 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1433 = 0s. The bug is otherwise harmless.
1435 PH/07 There was a bug in the dovecot authenticator such that the value of
1436 $auth1 could be overwritten, and so not correctly preserved, after a
1437 successful authentication. This usually meant that the value preserved by
1438 the server_setid option was incorrect.
1440 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1442 PH/09 Installed PCRE release 7.0.
1444 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1445 run for batched SMTP input. It is now run at the start of every message
1446 in the batch. While fixing this I discovered that the process information
1447 (output by running exiwhat) was not always getting set for -bs and -bS
1448 input. This is fixed, and it now also says "batched" for BSMTP.
1450 PH/11 Added control=no_pipelining.
1452 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1453 patch, slightly modified), and move the expansion of helo_data till after
1454 the connection is made in the smtp transport (so it can use these
1457 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1459 PH/14 Added log_selector = +pid.
1461 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1463 PH/16 Add ${if forany and ${if forall.
1465 PH/17 Added dsn_from option to vary the From: line in DSNs.
1467 PH/18 Flush SMTP output before performing a callout, unless control =
1468 no_callout_flush is set.
1470 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1471 was true (the default) a successful delivery failed to delete the retry
1472 item, thus causing premature timeout of the address. The bug is now
1475 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1477 PH/21 Long custom messages for fakedefer and fakereject are now split up
1478 into multiline reponses in the same way that messages for "deny" and
1479 other ACL rejections are.
1481 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1482 with slight modification.
1484 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1485 draft, changing the syntax and factoring some duplicate code".
1487 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1488 for deliveries of the second and subsequent messages over the same SMTP
1491 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1492 ${reduce, with only minor "tidies".
1494 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1496 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1497 expansion side effects.
1499 PH/27 When a message times out after an over-quota error from an Exim-imposed
1500 quota, the bounce message says "mailbox is full". This message was not
1501 being given when it was a system quota that was exceeded. It now should
1504 MH/03 Made $recipients available in local_scan(). local_scan() already has
1505 better access to the recipient list through recipients_list[], but
1506 $recipients can be useful in postmaster-provided expansion strings.
1508 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1509 in the case of a MAIL command with additional options following the
1510 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1511 were accidentally chopped off.
1513 PH/29 SMTP synchronization checks are implemented when a command is read -
1514 there is a check that no more input is waiting when there shouldn't be
1515 any. However, for some commands, a delay in an ACL can mean that it is
1516 some time before the response is written. In this time, more input might
1517 arrive, invalidly. So now there are extra checks after an ACL has run for
1518 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1519 pipelining has not been advertised.
1521 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1523 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1524 returned in the SMTP response when smtp_return_error_details was false.
1525 This has been fixed.
1527 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1528 instead of the C I/O that was originally supplied, because problems were
1529 reported on Solaris.
1531 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1532 Exim which did not show up earlier: it was assuming that a call to
1533 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1534 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1535 was a macro that became an assignment, so it seemed to work. This has
1536 changed to a proper function call with a void return, hence the compile
1537 error. Exim's code has been fixed.
1539 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1542 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1544 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1546 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1547 because it turns out to be too big or there is a timeout) there is no
1548 "Completed" line in the log. When some messages of this type were
1549 selected by exigrep, they were listed as "not completed". Others were
1550 picked up by some special patterns. I have improved the selection
1551 criteria to be more general.
1553 PH/38 The host_find_failed option in the manualroute router can now be set
1554 to "ignore", to completely ignore a host whose IP address cannot be
1555 found. If all hosts are ignored, the behaviour is controlled by the new
1556 host_all_ignored option.
1558 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1559 homing or because of multiple MX records with /mx) generated more than
1560 one IP address, and the following item turned out to be the local host,
1561 all the secondary addresses of the first item were incorrectly removed
1562 from the list, along with the local host and any following hosts (which
1563 is what is supposed to happen).
1565 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1566 whoever called Exim into the -H file. In the case of the daemon it was
1567 behaving confusingly. When first started, it used values for whoever
1568 started the daemon, but after a SIGHUP it used the Exim user (because it
1569 calls itself on a restart). I have changed the code so that it now always
1572 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1573 message are rejected with the same error (e.g. no authentication or bad
1574 sender address), and a DATA command is nevertheless sent (as can happen
1575 with PIPELINING or a stupid MUA), the error message that was given to the
1576 RCPT commands is included in the rejection of the DATA command. This is
1577 intended to be helpful for MUAs that show only the final error to their
1580 PH/42 Another patch from the Sieve maintainer.
1582 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1583 Eximstats - Fixed some broken HTML links and added missing column headers
1585 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1586 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1588 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1590 PH/43 Yet another patch from the Sieve maintainer.
1592 PH/44 I found a way to check for a TCP/IP connection going away before sending
1593 the response to the final '.' that terminates a message, but only in the
1594 case where the client has not sent further data following the '.'
1595 (unfortunately, this is allowed). However, in many cases there won't be
1596 any further data because there won't be any more messages to send. A call
1597 to select() can be used: if it shows that the input is "ready", there is
1598 either input waiting, or the socket has been closed. An attempt to read
1599 the next input character can distinguish the two cases. Previously, Exim
1600 would have sent an OK response which the client would never have see.
1601 This could lead to message repetition. This fix should cure that, at
1602 least in a lot of common cases.
1604 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1605 advertised in response to EHLO.
1611 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1612 fixed by 4.65/MH/01 (is this a record?) are fixed:
1614 (i) An empty string was always treated as zero by the numeric comparison
1615 operators. This behaviour has been restored.
1617 (ii) It is documented that the numeric comparison operators always treat
1618 their arguments as decimal numbers. This was broken in that numbers
1619 starting with 0 were being interpreted as octal.
1621 While fixing these problems I realized that there was another issue that
1622 hadn't been noticed. Values of message_size_limit (both the global option
1623 and the transport option) were treated as octal if they started with 0.
1624 The documentation was vague. These values are now always treated as
1625 decimal, and I will make that clear in the documentation.
1631 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1632 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1635 MH/01 Don't check that the operands of numeric comparison operators are
1636 integers when their expansion is in "skipping" mode (fixes bug
1637 introduced by 4.64-PH/07).
1639 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1640 child addresses, Exim now panics and dies. Previously, because the count
1641 is held in a short int, deliveries were likely to be lost. As such a
1642 large number of recipients for a single message is ridiculous
1643 (performance will be very, very poor), I have chosen to impose a limit
1644 rather than extend the field.
1650 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1651 leftover -K file (the existence of which was triggered by #402).
1652 While we were at it, introduced process PID as part of the -K
1653 filename. This should rule out race conditions when creating
1656 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1657 processing considerably. Previous code took too long for large mails,
1658 triggering a timeout which in turn triggers #401.
1660 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1661 in the DK code in transports.c. sendfile() is not really portable,
1662 hence the _LINUX specificness.
1664 TF/01 In the add_headers option to the mail command in an Exim filter,
1665 there was a bug that Exim would claim a syntax error in any
1666 header after the first one which had an odd number of characters
1669 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1670 callout verification, Exim cached a "reject" for the entire domain. This
1671 is correct for most verifications, but it is not correct for a recipient
1672 verification with use_sender or use_postmaster set, because in that case
1673 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1674 case of MAIL FROM:<> rejection from other early rejections (e.g.
1675 rejection of HELO). When verifying a recipient using a non-null MAIL
1676 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1677 Whatever the result of the callout, the value of the domain cache is
1678 left unchanged (for any other kind of callout, getting as far as trying
1679 RCPT means that the domain itself is ok).
1681 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1684 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1685 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1686 session, the daemon ignores ECONNECT errors and logs others; it now
1687 ignores EPIPE as well.
1689 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1690 (quoted-printable decoding).
1692 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1693 later the small subsequent patch to fix an introduced bug.
1695 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1697 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1699 PH/08 An error is now given if message_size_limit is specified negative.
1701 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1702 to be given (somewhat) arbitrary names.
1704 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1707 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1708 miscellaneous code fixes
1710 PH/10 Added the log_reject_target ACL modifier to specify where to log
1713 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1714 hostname. This is wrong, because it relates to the incoming message (and
1715 probably the interface on which it is arriving) and not to the outgoing
1716 callout (which could be using a different interface). This has been
1717 changed to use the value of the helo_data option from the smtp transport
1718 instead - this is what is used when a message is actually being sent. If
1719 there is no remote transport (possible with a router that sets up host
1720 addresses), $smtp_active_hostname is used.
1722 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1723 tweaks were necessary in order to get it to work (see also 21 below):
1724 (a) The code assumed that strncpy() returns a negative number on buffer
1725 overflow, which isn't the case. Replaced with Exim's string_format()
1727 (b) There were several signed/unsigned issues. I just did the minimum
1728 hacking in of casts. There is scope for a larger refactoring.
1729 (c) The code used strcasecmp() which is not a standard C function.
1730 Replaced with Exim's strcmpic() function.
1731 (d) The code set only $1; it now sets $auth1 as well.
1732 (e) A simple test gave the error "authentication client didn't specify
1733 service in request". It would seem that Dovecot has changed its
1734 interface. Fortunately there's a specification; I followed it and
1735 changed what the client sends and it appears to be working now.
1737 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1740 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1741 address A is aliased to B and C, where B exists and C does not. Without
1742 -v the output is "A verified" because verification stops after a
1743 successful redirection if more than one address is generated. However,
1744 with -v the child addresses are also verified. Exim was outputting "A
1745 failed to verify" and then showing the successful verification for C,
1746 with its parentage. It now outputs "B failed to verify", showing B's
1747 parentage before showing the successful verification of C.
1749 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1750 look up a TXT record in a specific list after matching in a combined
1753 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1754 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1755 they consult the DNS. I had assumed they would set it the way they
1756 wanted; and indeed my experiments on Linux seem to show that in some
1757 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1758 To be on the safe side, however, I have now made the interface to
1759 host_find_byname() similar to host_find_bydns(), with an argument
1760 containing the DNS resolver options. The host_find_byname() function now
1761 sets these options at its start, just as host_find_bydns() does. The smtp
1762 transport options dns_qualify_single and dns_search_parents are passed to
1763 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1764 of host_find_byname() use the default settings of RES_DEFNAMES
1765 (qualify_single) but not RES_DNSRCH (search_parents).
1767 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1768 spool_read_header() do less string testing, by means of a preliminary
1769 switch on the second character of optional "-foo" lines. (This is
1770 overdue, caused by the large number of possibilities that now exist.
1771 Originally there were few.) While I was there, I also converted the
1772 str(n)cmp tests so they don't re-test the leading "-" and the first
1773 character, in the hope this might squeeze out yet more improvement.
1775 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1776 flag allowing group syntax was set by the header_syntax check but not
1777 turned off, possible causing trouble later; (2) The flag was not being
1778 set at all for the header_verify test, causing "group"-style headers to
1779 be rejected. I have now set it in this case, and also caused header_
1780 verify to ignore an empty address taken from a group. While doing this, I
1781 came across some other cases where the code for allowing group syntax
1782 while scanning a header line wasn't quite right (mostly, not resetting
1783 the flag correctly in the right place). These bugs could have caused
1784 trouble for malformed header lines. I hope it is now all correct.
1786 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1787 with the "reply" argument non-NULL. The code, however (which originally
1788 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1789 but it didn't always do it. This confused somebody who was copying the
1790 code for some other use. I have removed all the tests.
1792 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1793 feature that was used to support insecure browsers during the U.S. crypto
1794 embargo. It requires special client support, and Exim is probably the
1795 only MTA that supported it -- and would never use it because real RSA is
1796 always available. This code has been removed, because it had the bad
1797 effect of slowing Exim down by computing (never used) parameters for the
1798 RSA_EXPORT functionality.
1800 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1801 authenticator to fail if there's a tab character in the incoming data
1802 (there should never be unless someone is messing about, as it's supposed
1803 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1804 if the connection is using TLS or if the remote IP is the same as the
1805 local IP, and the "valid-client-cert option" if a client certificate has
1808 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1809 authenticators. This can be used for authorization after authentication
1810 succeeds. (In the case of plaintext, it servers for both authentication
1813 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1814 if any retry times were supplied.
1816 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1817 connection, where there is no client IP address to check. In this
1818 situation, the verify now always succeeds.
1820 PH/25 Applied John Jetmore's -Mset patch.
1822 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1824 PH/27 In a string expansion for a processed (not raw) header when multiple
1825 headers of the same name were present, leading whitespace was being
1826 removed from all of them, but trailing whitespace was being removed only
1827 from the last one. Now trailing whitespace is removed from each header
1828 before concatenation. Completely empty headers in a concatenation (as
1829 before) are ignored.
1831 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1832 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1834 PH/29 [Removed. This was a change that I later backed out, and forgot to
1835 correct the ChangeLog entry (that I had efficiently created) before
1836 committing the later change.]
1838 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1839 address errors (4xx response to RCPT) over the same connection as other
1840 messages routed to the same hosts. Such deliveries are always "forced",
1841 so retry times are not inspected. This resulted in far too many retries
1842 for the affected addresses. The effect occurred only when there were more
1843 hosts than the hosts_max_try setting in the smtp transport when it had
1844 the 4xx errors. Those hosts that it had tried were not added to the list
1845 of hosts for which the message was waiting, so if all were tried, there
1846 was no problem. Two fixes have been applied:
1848 (i) If there are any address or message errors in an SMTP delivery, none
1849 of the hosts (tried or untried) are now added to the list of hosts
1850 for which the message is waiting, so the message should not be a
1851 candidate for sending over the same connection that was used for a
1852 successful delivery of some other message. This seems entirely
1853 reasonable: after all the message is NOT "waiting for some host".
1854 This is so "obvious" that I'm not sure why it wasn't done
1855 previously. Hope I haven't missed anything, but it can't do any
1856 harm, as the worst effect is to miss an optimization.
1858 (ii) If, despite (i), such a delivery is accidentally attempted, the
1859 routing retry time is respected, so at least it doesn't keep
1860 hammering the server.
1862 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1863 in ${readsocket because some servers need this prod.
1865 PH/32 Added some extra debug output when updating a wait-xxx database.
1867 PH/33 The hint "could be header name not terminated by colon", which has been
1868 given for certain expansion errors for a long time, was not being given
1869 for the ${if def:h_colon_omitted{... case.
1871 PH/34 The spec says: "With one important exception, whenever a domain list is
1872 being scanned, $domain contains the subject domain." There was at least
1873 one case where this was not true.
1875 PH/35 The error "getsockname() failed: connection reset by peer" was being
1876 written to the panic log as well as the main log, but it isn't really
1877 panic-worthy as it just means the connection died rather early on. I have
1878 removed the panic log writing for the ECONNRESET error when getsockname()
1881 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1882 runs only) independently of the message's sender address. This meant
1883 that, if the 4xx error was in fact related to the sender, a different
1884 message to the same recipient with a different sender could confuse
1885 things. In particualar, this can happen when sending to a greylisting
1886 server, but other circumstances could also provoke similar problems.
1887 I have changed the default so that the retry time for these errors is now
1888 based a combination of the sender and recipient addresses. This change
1889 can be overridden by setting address_retry_include_sender=false in the
1892 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1893 remote server are returned as part of bounce messages. This was not
1894 happening for LMTP over a pipe (the lmtp transport), but now it is the
1895 same for both kinds of LMTP.
1897 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1898 in header lines that were in fact CNAMEs. This is no longer the case.
1900 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1901 and queue runs started by the daemon processed all messages. This has
1902 been fixed so that -R and -S can now usefully be given with -q<time>.
1904 PH/40 Import PCRE release 6.7 (fixes some bugs).
1906 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1908 PH/42 Give an error if -q is specified more than once.
1910 PH/43 Renamed the variables $interface_address and $interface_port as
1911 $received_ip_address and $received_port, to make it clear that these
1912 values apply to message reception, and not to the outgoing interface when
1913 a message is delivered. (The old names remain recognized, of course.)
1915 PH/44 There was no timeout on the connect() call when using a Unix domain
1916 socket in the ${readsocket expansion. There now is.
1918 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1919 be meaningful with "accept".
1921 SC/01 Eximstats V1.43
1922 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1924 SC/02 Eximstats V1.44
1925 Use a glob alias rather than an array ref in the generated
1926 parser. This improves both readability and performance.
1928 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1929 Collect SpamAssassin and rejection statistics.
1930 Don't display local sender or destination tables unless
1931 there is data to show.
1932 Added average volumes into the top table text output.
1934 SC/04 Eximstats V1.46
1935 Collect data on the number of addresses (recipients)
1936 as well as the number of messages.
1938 SC/05 Eximstats V1.47
1939 Added 'Message too big' to the list of mail rejection
1940 reasons (thanks to Marco Gaiarin).
1942 SC/06 Eximstats V1.48
1943 Mainlog lines which have GMT offsets and are too short to
1944 have a flag are now skipped.
1946 SC/07 Eximstats V1.49 (Alain Williams)
1947 Added the -emptyok flag.
1949 SC/08 Eximstats V1.50
1950 Fixes for obtaining the IP address from reject messages.
1952 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1953 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1954 whitesspace changes from 4.64-PH/27
1956 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1959 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1960 are found, allow negative numbers in numeric criteria)
1962 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1964 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1965 to match changes made in 4.64-PH/43
1967 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1969 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1970 "domain" is actually an IP address, and force a failure. This locks out
1971 those revolvers/nameservers that support "A-for-A" lookups, in
1972 contravention of the specifications.
1974 PH/48 When a host name was looked up from an IP address, and the subsequent
1975 forward lookup of the name timed out, the host name was left in
1976 $sender_host_name, contrary to the specification.
1978 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1979 restricted to single-key lookups, Exim was not diagnosing an error if
1980 * or *@ was used with a query-style lookup.
1982 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1984 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1985 long ago, but noone interested enough thought of it. Let's just say that
1986 the "1.1" means that there are some new functions that weren't there at
1987 some point in the past.
1989 PH/51 Error processing for expansion failure of helo_data from an smtp
1990 transport during callout processing was broken.
1992 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1993 tested/used via the -bh/-bhc/-bs options.
1995 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1996 bug, fixed in subsequent PCRE releases).
1998 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1999 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2001 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2007 SC/01 Use a glob alias rather than an array ref in eximstats generated
2008 parser. This improves both readability and performance.
2010 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2011 Don't display local sender or destination tables in eximstats unless
2012 there is data to show.
2013 Added average volumes into the eximstats top table text output.
2015 SC/03 Collect data on the number of addresses (recipients) as well
2016 as the number of messages in eximstats.
2018 TF/01 Correct an error in the documentation for the redirect router. Exim
2019 does (usually) call initgroups() when daemonizing.
2021 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2022 with consistent privilege compared to when running as a daemon.
2024 TF/03 Note in the spec that $authenticated_id is not set for local
2025 submissions from trusted users.
2027 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2028 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2030 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2031 by adding some example configuration directives to the default
2032 configuration file. A little bit of work is required to uncomment the
2033 directives and define how usernames and passwords are checked, but
2034 there is now a framework to start from.
2036 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2037 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2038 without this. I don't know how relevant this is to other LDAP libraries.
2040 PH/02 Add the verb name to the "unknown ACL verb" error.
2042 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2044 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2046 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2047 directory not expanded when it should be if an expanded home directory
2048 was set for the address (which is overridden by the transport).
2050 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2053 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2054 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2055 because it is too late at that time, and has no effect.
2057 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2058 security issue with \' (bugzilla #107). I could not use the
2059 PQescapeStringConn() function, because it needs a PGconn value as one of
2062 PH/08 When testing addresses using -bt, indicate those final addresses that
2063 are duplicates that would not cause an additional delivery. At least one
2064 person was confused, thinking that -bt output corresponded to deliveries.
2065 (Suppressing duplicates isn't a good idea as you lose the information
2066 about possibly different redirections that led to the duplicates.)
2068 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2069 systems where poll() doesn't work, in particular OS X.
2071 PH/10 Added more information to debugging output for retry time not reached.
2073 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2074 operations in malware.c.
2076 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2079 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2080 syslog_duplication set false, log lines that would normally be written
2081 both the the main log and to the reject log were not written to syslog at
2084 PH/14 In the default configuration, change the use of "message" in ACL warn
2085 statements to "add_header".
2087 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2088 not followed by a command (e.g. "seen endif").
2090 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2091 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2094 PH/17 Added extra conditions to the default value of delay_warning_condition
2098 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2099 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2100 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2103 The Auto-Submitted: and various List- headers are standardised, whereas I
2104 don't think Precedence: ever was.
2106 PH/18 Refactored debugging code in route_finduser() to show more information,
2107 in particular, the error code if getpwnam() issues one.
2109 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2110 This is apparently needed in addition to the PH/07 change above to avoid
2111 any possible encoding problems.
2113 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2114 but not after initializing Perl.
2116 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2117 output them only if debugging. By default they are written stderr,
2118 apparently, which is not desirable.
2120 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2123 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2126 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2128 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2129 authenticated or an ident call has been made. Suppress the default
2130 values for $authenticated_id and $authenticated_sender (but permit -oMai
2131 and -oMas) when testing with -bh.
2133 PH/24 Re-jigged the order of the tests in the default configuration so that the
2134 tests for valid domains and recipients precede the DNS black list and CSA
2135 tests, on the grounds that those ones are more expensive.
2137 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2138 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2139 This bug exists in every version of Exim that I still have, right back to
2142 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2143 However, an attempt to turn on case-sensitivity in a regex key by
2144 including (?-i) didn't work because the subject string was already
2145 lowercased, and the effects were non-intuitive. It turns out that a
2146 one-line patch can be used to allow (?-i) to work as expected.
2152 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2153 other effects) broke the use of negated acl sub-conditions.
2155 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2158 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2159 "Deny" causes Exim to reject the incoming connection with a 554 error.
2160 Unfortunately, if there is a major crisis, such as a disk failure,
2161 tcp-wrappers gives "deny", whereas what one would like would be some
2162 kind of temporary error. A kludge has been added to help with this.
2163 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2164 554 error is used if errno is still zero or contains ENOENT (which occurs
2165 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2168 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2170 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2171 errors. Otherwise a message that provokes a temporary error (when other
2172 messages do not) can cause a whole host to time out.
2174 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2175 addresses were routed directly to files or pipes from a redirect router.
2176 File deliveries just didn't batch; pipe deliveries might have suffered
2179 PH/06 A failure to get a lock for a hints database would erroneously always say
2180 "Failed to get write lock", even when it was really a read lock.
2182 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2183 of 0600. This has been changed to use the value of the lockfile_mode
2184 option (which defaults to 0600).
2186 PH/08 Applied small patch from the Sieve maintainer.
2188 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2189 folder from quota calculations, a direct delivery into this folder messed
2190 up the contents of the maildirsize file. This was because the regex was
2191 used only to exclude .Trash (or whatever) when the size of the mailbox
2192 was calculated. There was no check that a delivery was happening into an
2193 excluded directory. This bug has been fixed by ignoring all quota
2194 processing for deliveries into excluded directories.
2196 PH/10 Added the maildirfolder_create_regex option to appendfile.
2202 PH/01 The code for finding all the local interface addresses on a FreeBSD
2203 system running IPv6 was broken. This may well have applied to all BSD
2204 systems, as well as to others that have similar system calls. The broken
2205 code found IPv4 interfaces correctly, but gave incorrect values for the
2206 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2207 that it would not match correctly against @[] and not recognize the IPv6
2210 PH/02 The ipliteral router was not recognizing addresses of the form user@
2211 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2213 PH/03 Added disable_ipv6.
2215 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2216 decoded form, because it is most often used to construct To: headers
2217 lines in autoreplies, and the decoded form may well be syntactically
2218 invalid. However, $reply_address has leading white space removed, and all
2219 newlines turned into spaces so that the autoreply transport does not
2222 PH/05 If group was specified without a user on a router, and no group or user
2223 was specified on a transport, the group from the router was ignored.
2225 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2226 for visible compile-time settings that can be used to change these
2227 numbers, for those that want even more. Backwards compatibility with old
2228 spool files has been maintained. However, going back to a previous Exim
2229 release will lost any variables that are in spool files.
2231 PH/07 Two small changes when running in the test harness: increase delay when
2232 passing a TCP/IP connection to a new process, in case the original
2233 process has to generate a bounce, and remove special handling of
2234 127.0.0.2 (sic), which is no longer necessary.
2236 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2237 be the same on different OS.
2239 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2242 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2243 whether --show-vars was specified or not
2245 JJ/02 exipick: Added support for new ACL variable spool format introduced
2248 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2249 syntactically invalid From: or Reply-to: line, and a filter used this to
2250 generate an autoreply, and therefore failed to obtain an address for the
2251 autoreply, Exim could try to deliver to a non-existent relative file
2252 name, causing unrelated and misleading errors. What now happens is that
2253 it logs this as a hard delivery error, but does not attempt to create a
2256 PH/11 The exinext utility has a -C option for testing purposes, but although
2257 the given file was scanned by exinext itself; it wasn't being passed on
2258 when Exim was called.
2260 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2261 an end-of-file indication when reading a command response.
2263 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2264 compiled. In many other places in Exim, IPv6 addresses are always
2265 recognized, so I have changed this. It also means that IPv4 domain
2266 literals of the form [IPV4:n.n.n.n] are now always recognized.
2268 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2269 used if the router is not running as root, for example, when verifying at
2270 ACL time, or when using -bh. The debugging output from this situation was
2271 non-existent - all you got was a failure to exec. I have made two
2274 (a) Failures to set uid/gid, the current directory, or a process leader
2275 in a subprocess such as that created by queryprogram now generate
2276 suitable debugging ouput when -d is set.
2278 (b) The queryprogram router detects when it is not running as root,
2279 outputs suitable debugging information if -d is set, and then runs
2280 the subprocess without attempting to change uid/gid.
2282 PH/15 Minor change to Makefile for building test_host (undocumented testing
2285 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2286 additional section of a DNS packet that returns MX or SRV records.
2287 Instead, it always explicitly searches for A/AAAA records. This avoids
2288 major problems that occur when a DNS server includes only records of one
2289 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2290 fixed another bug: if SRV records were looked up and the corresponding
2291 address records were *not* found in the additional section, the port
2292 values from the SRV records were lost.
2294 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2295 using the correct key (the original address) when searching the retry
2296 rules in order to find which one to use for generating the retry hint.
2298 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2299 adding the default one. Similarly, if it contains a Reply-To: header, the
2300 errors_reply_to option, if set, is not used.
2302 PH/19 When calculating a retry time, Exim used to measure the "time since
2303 failure" by looking at the "first failed" field in the retry record. Now
2304 it does not use this if it is later than than the arrival time of the
2305 message. Instead it uses the arrival time. This makes for better
2306 behaviour in cases where some deliveries succeed, thus re-setting the
2307 "first failed" field. An example is a quota failure for a huge message
2308 when small messages continue to be delivered. Without this change, the
2309 "time since failure" will always be short, possible causing more frequent
2310 delivery attempts for the huge message than are intended.
2311 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2313 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2314 $1, $2, $3) because the numerical variables can be reset during some
2315 expansion items (e.g. "match"), thereby losing the authentication data.
2317 PH/21 Make -bV show the size of off_t variables so that the test suite can
2318 decide whether to run tests for quotas > 2G.
2320 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2321 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2322 filecount value is greater than 2G or if a quota value is greater than 2G
2323 on a system where the size of off_t is not greater than 4, a panic error
2326 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2327 never match. The debug and -bh output now contains an explicit error
2328 message indicating a malformed IPv4 address or mask.
2330 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2331 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2332 PH/23 above applies.
2334 PH/25 Do not write to syslog when running in the test harness. The only
2335 occasion when this arises is a failure to open the main or panic logs
2336 (for which there is an explicit test).
2338 PH/26 Added the /no_tell option to "control=freeze".
2340 PH/27 If a host name lookup failed very early in a connection, for example, if
2341 the IP address matched host_lookup and the reverse lookup yielded a name
2342 that did not have a forward lookup, an error message of the form "no IP
2343 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2344 could be logged. Now it outputs the IP address instead of "NULL".
2346 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2347 allows the sender and the authenticated sender to be set when
2348 submitting a message from within Exim. Since child_open_exim() is
2349 documented for local_scan(), the new function should be too.
2351 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2352 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2353 results in an empty string is now treated as unset.
2355 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2357 PH/31 Added sender_verify_fail logging option.
2359 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2360 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2361 tidied the source and removed it altogether.
2363 PH/33 When a queue run was abandoned because the load average was too high, a
2364 log line was always written; now it is written only if the queue_run log
2365 selector is set. In addition, the log line for abandonment now contains
2366 information about the queue run such as the pid. This is always present
2367 in "start" and "stop" lines but was omitted from the "abandon" line.
2369 PH/34 Omit spaces between a header name and the colon in the error message that
2370 is given when verify = headers_syntax fails (if there are lots of them,
2371 the message gets confusing).
2373 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2374 names, as there are now some PTR records that contain slashes. This check
2375 is only to protect against broken name servers that fall over on strange
2376 characters, so the fact that it applies to all lookups doesn't matter.
2378 PH/36 Now that the new test suite is complete, we can remove some of the
2379 special code in Exim that was needed for the old test suite. For example,
2380 sorting DNS records because real resolvers return them in an arbitrary
2381 order. The new test suite's fake resolver always returns records in the
2384 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2385 bounces) except when queue_only is set, to avoid logging races between
2386 the different processes.
2388 PH/38 Panic-die if .include specifies a non-absolute path.
2390 PH/39 A tweak to the "H" retry rule from its user.
2392 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2393 a label. They prevented compilation on older perls.
2395 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2396 a warning to be raised on newish perls.
2398 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2399 on queue. Changes to match documented behaviour of showing count of
2400 messages matching specified criteria.
2402 PH/40 Changed the default ident timeout from 30s to 5s.
2404 PH/41 Added support for the use of login_cap features, on those BSD systems
2405 that have them, for controlling the resources used by pipe deliveries.
2407 PH/42 The content-scanning code uses fopen() to create files in which to put
2408 message data. Previously it was not paying any attention to the mode of
2409 the files. Exim runs with umask(0) because the rest of the code creates
2410 files with open(), and sets the required mode explicitly. Thus, these
2411 files were ending up world-writeable. This was not a big issue, because,
2412 being within the spool directory, they were not world-accessible. I have
2413 created a function called modefopen, which takes an additional mode
2414 argument. It sets umask(777), creates the file, chmods it to the required
2415 mode, then resets the umask. All the relevant calls to fopen() in the
2416 content scanning code have been changed to use this function.
2418 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2419 to 24 hours. This avoids potential overflow problems when processing G
2420 and H retry rules. I suspect nobody ever tinkers with this value.
2422 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2424 PH/45 When the plaintext authenticator is running as a client, the server's
2425 challenges are checked to ensure they are valid base64 strings. By
2426 default, the authentication attempt is cancelled if an invalid string is
2427 received. Setting client_ignore_invalid_base64 true ignores these errors.
2428 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2429 they are received. Thus, the responses can be made to depend on the
2430 challenges. If an invalid string is ignored, an empty string is placed in
2433 PH/46 Messages that are created by the autoreply transport now contains a
2434 References: header, in accordance with RFCs 2822 and 3834.
2436 PH/47 Added authenticated_sender_force to the smtp transport.
2438 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2440 PH/49 Installed latest patch from the Sieve maintainer.
2442 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2443 was also set, the appendfile transport was unnecessarily scanning a
2444 directory of message files (e.g. for maildir delivery) to find the count
2445 of files (along with the size), even though it did not need this
2446 information. It now does the scan only if it needs to find either the
2447 size of the count of files.
2449 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2451 PH/52 Two bugs concerned with error handling when the smtp transport is
2454 (i) Exim was not creating retry information for temporary errors given
2455 for individual recipients after the DATA command when the smtp transport
2456 was used in LMTP mode. This meant that they could be retried too
2457 frequently, and not timed out correctly.
2459 (ii) Exim was setting the flag that allows error details to be returned
2460 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2461 recipients that were returned after the DATA command.
2463 PH/53 This is related to PH/52, but is more general: for any failing address,
2464 when detailed error information was permitted to be returned to the
2465 sender, but the error was temporary, then after the final timeout, only
2466 "retry timeout exceeded" was returned. Now it returns the full error as
2467 well as "retry timeout exceeded".
2469 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2470 do this, and (what is worse) MTAs that accept it.
2472 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2473 will now be deprecated.
2475 PH/56 New os.c-cygwin from the Cygwin maintainer.
2477 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2478 formats (previously only available in exim formats via -bpr, -bpru,
2479 and -bpra. Now also available in native and exiqgrep formats)
2481 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2482 with very large, slow to parse queues
2484 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2486 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2488 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2489 responses to authentication challenges, though it was showing the
2490 challenges; (ii) I've removed the CR characters from the debug output for
2493 PH/58 Allow for the insertion of a newline as well as a space when a string
2494 is turned into more than one encoded-word during RFC 2047 encoding. The
2495 Sieve code now uses this.
2497 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2498 data_4xx, lost_connection, tls_required.
2500 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2501 message was being sent as an SMTP response.
2503 PH/61 Add -l and -k options to exicyclog.
2505 PH/62 When verifying, if an address was redirected to one new address, so that
2506 verification continued, and the new address failed or deferred after
2507 having set something in $address_data, the value of $address_data was not
2508 passed back to the ACL. This was different to the case when no
2509 redirection occurred. The value is now passed back in both cases.
2511 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2512 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2513 use login_cap.h, so on its own it isn't the distinguishing feature. The
2514 new name refers directly to the setclassresources() function.
2516 PH/65 Added configuration files for NetBSD3.
2518 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2520 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2521 is preferred over IPv4.
2523 PH/68 The bounce_return_message and bounce_return_body options were not being
2524 honoured for bounces generated during the reception of non-SMTP messages.
2525 In particular, this applied to messages rejected by the ACL. This bug has
2526 been fixed. However, if bounce_return_message is true and bounce_return_
2527 body is false, the headers that are returned for a non-SMTP message
2528 include only those that have been read before the error was detected.
2529 (In the case of an ACL rejection, they have all been read.)
2531 PH/69 The HTML version of the specification is now built in a directory called
2532 spec_html instead of spec.html, because the latter looks like a path with
2533 a MIME-type, and this confuses some software.
2535 PH/70 Catch two compiler warnings in sieve.c.
2537 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2538 function verify_get_ident() calls ip_connect() to connect a socket, but
2539 if the "connect()" function timed out, ip_connect() used to close the
2540 socket. However, verify_get_ident() also closes the socket later, and in
2541 between Exim writes to the log, which may get opened at this point. When
2542 the socket was closed in ip_connect(), the log could get the same file
2543 descriptor number as the socket. This naturally causes chaos. The fix is
2544 not to close the socket in ip_connect(); the socket should be closed by
2545 the function that creates it. There was only one place in the code where
2546 this was missing, in the iplookup router, which I don't think anybody now
2547 uses, but I've fixed it anyway.
2549 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2550 well as to direct DNS lookups. Otherwise the handling of names in host
2551 lists is inconsistent and therefore confusing.
2557 PH/01 Two changes to the default runtime configuration:
2559 (1) Move the checks for relay_from_hosts and authenticated clients from
2560 after to before the (commented out) DNS black list checks.
2562 (2) Add control=submission to the relay_from_hosts and authenticated
2563 clients checks, on the grounds that messages accepted by these
2564 statements are most likely to be submissions.
2566 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2568 (1) Generate an error if the third argument for the ${prvs expansion is
2571 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2574 (3) Reset the variables that are obtained from the first argument of
2575 ${prvscheck and used in the second argument before leaving the code,
2576 because their memory is reclaimed, so using them afterwards may do
2579 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2580 one (it's much easier than Tom thought :-).
2582 (5) Because of (4), we can now allow for the use of $prvscheck_result
2583 inside the third argument.
2585 PH/03 For some reason, the default setting of PATH when running a command from
2586 a pipe transport was just "/usr/bin". I have changed it to
2589 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2590 anything to be listed in the output from -bV.
2592 PH/05 When a filter generated an autoreply, the entire To: header line was
2593 quoted in the delivery log line, like this:
2595 => >A.N.Other <ano@some.domain> <original@ddress> ...
2597 This has been changed so that it extracts the operative address. There
2598 may be more than one such address. If so, they are comma-separated, like
2601 => >ano@some.domain,ona@other.domain <original@ddress> ...
2603 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2604 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2605 not being looked up in the rDNS to get a host name, Exim was showing the
2606 IP address twice in Received: lines, even though the IP addresses were
2607 identical. For example:
2609 Received: from [1.2.3.4] (helo=[1.2.3.4])
2611 However, if the real host name was known, it was omitting the HELO data
2612 if it matched the actual IP address. This has been tidied up so that it
2613 doesn't show the same IP address twice.
2615 PH/07 When both +timestamp and +memory debugging was on, the value given by
2616 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2617 called by the debug printing, thereby overwriting the timestamp buffer.
2618 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2620 PH/08 When the original message was included in an autoreply transport, it
2621 always said "this is a copy of the message, including all the headers",
2622 even if body_only or headers_only was set. It now gives an appropriate
2625 PH/09 Applied a patch from the Sieve maintainer which:
2627 o fixes some comments
2628 o adds the (disabled) notify extension core
2629 o adds some debug output for the result of if/elsif tests
2630 o points to the current vacation draft in the documentation
2631 and documents the missing references header update
2635 o fixes a bug in processing the envelope test (when testing
2636 multiple envelope elements, the last element determinted the
2639 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2640 Electronic Mail") by including:
2642 Auto-submitted: auto-generated
2644 in the messages that it generates (bounce messages and others, such as
2645 warnings). In the case of bounce messages for non-SMTP mesages, there was
2646 also a typo: it was using "Auto_submitted" (underscore instead of
2647 hyphen). Since every message generated by Exim is necessarily in response
2648 to another message, thes have all been changed to:
2650 Auto-Submitted: auto-replied
2652 in accordance with these statements in the RFC:
2654 The auto-replied keyword:
2656 - SHOULD be used on messages sent in direct response to another
2657 message by an automatic process,
2659 - MUST NOT be used on manually-generated messages,
2661 - MAY be used on Delivery Status Notifications (DSNs) and Message
2662 Disposition Notifications (MDNs),
2664 - MUST NOT be used on messages generated by automatic or periodic
2665 processes, except for messages which are automatic responses to
2668 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2669 to the default Received: header definition.
2671 PH/12 Added log selector acl_warn_skipped (default on).
2673 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2674 variables because (a) they are in the wrong storage pool and (b) even if
2675 they were copied, it wouldn't work properly because of the caching.
2677 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2678 checking when decoding. Apparently there are clients that generate
2679 overlong encoded strings. Why am I not surprised?
2681 PH/15 If the first argument of "${if match_address" was not empty, but did not
2682 contain an "@" character, Exim crashed. Now it writes a panic log message
2683 and treats the condition as false.
2685 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2687 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2688 extension "envelope-auth". The code is finished and in agreement with
2689 other implementations, but there is no documentation so far and in fact,
2690 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2691 not changing the active code.
2693 Print executed "if" and "elsif" statements when debugging is used. This
2694 helps a great deal to understand what a filter does.
2696 Document more things not specified clearly in RFC3028. I had all this
2697 sorted out, when out of a sudden new issues came to my mind. Oops."
2699 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2702 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2703 canonical form (as documented). However, after a host name lookup from
2704 the IP address, check_host() was doing a simple string comparison with
2705 addresses acquired from the DNS when checking that the found name did
2706 have the original IP as one of its addresses. Since any found IPv6
2707 addresses are likely to be in abbreviated form, the comparison could
2708 fail. Luckily, there already exists a function for doing the comparison
2709 by converting both addresses to binary, so now that is used instead of
2710 the text comparison.
2712 PH/20 There was another similar case to PH/19, when a complete host name was
2713 given in a host list; looking up its IP address could give an abbreviated
2714 form, whereas the current host's name might or might not be abbreviated.
2715 The same fix has been applied.
2721 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2722 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2725 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2726 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2728 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2730 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2731 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2732 Also, for "Auto-Submitted", treat anything other than "no" as
2733 non-personal, in accordance with RFC 3834. (Previously it treated
2734 anything starting "auto-" as non-personal.)
2736 TF/01 The control=submission/name=... option had a problem with syntax
2737 errors if the name included a slash character. The /name= option
2738 now slurps the rest of the string, so it can include any characters
2739 but it must come last in the list of options (after /sender_retain
2742 PH/05 Some modifications to the interface to the fake nameserver for the new
2750 TK/01 Added the "success_on_redirect" address verification option. See
2751 NewStuff for rationale and an example.
2753 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2755 PH/02 Patch to exigrep to allow it to work on syslog lines.
2757 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2758 fread() to skip over the body file's header line, because in Cygwin the
2759 header line is locked and is inaccessible.
2761 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2762 co-exist for some time) to make it clear that it is the Exim ID that is
2763 referenced, not the Message-ID: header line.
2765 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2766 string_format() function, because snprintf() does not exist on all
2769 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2770 ${stat: expansion item.
2772 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2773 protocol synchronization error", to keep the pedants happy.
2775 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2776 well as for IRIX systems, when gcc is being used. See the host.c source
2779 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2781 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2784 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2785 transport and to the smtp transport in LMTP mode.
2787 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2789 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2790 run a filter in a subprocess. This could lead to confusion in subsequent
2791 lookups in the parent process. There should also be a search_tidyup() at
2792 the end of the subprocess.
2794 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2795 only if the host matched helo_try_verify_hosts, which caused the
2796 verification to occur when the EHLO/HELO command was issued. The ACL just
2797 tested the remembered result. Now, if a previous verification attempt has
2798 not happened, "verify = helo" does it there and then.
2800 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2802 TK/03 Fix log output including CR from clamd.
2804 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2805 exist provoked a memory error which could cause a segfault.
2807 PH/15 Installed PCRE 6.2
2809 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2811 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2812 of the problem. Specifically, suggested +O2 rather than +O1 for the
2815 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2817 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2818 record, the port was not correctly logged when the outgoing_port log
2819 selector was set (it logged the transort's default port).
2821 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2822 fallback_hosts, and "hosts" in the smtp transport.
2824 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2825 host errors such as "Connection refused".
2827 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2828 authentication with radiusclient 0.4.9:
2830 - Error returned from rc_read_config was caught wrongly
2831 - Username/password not passed on to radius server due to wrong length.
2833 The presumption is that some radiusclient API changes for 4.51/PH/17
2834 were not taken care of correctly. The code is still untested by me (my
2835 Linux distribution still has 0.3.2 of radiusclient), but it was
2836 contributed by a Radius user.
2838 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2839 expanding the "port" option of the smtp transport.
2841 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2842 while reading a MIME header. Thanks to Tom Hughes for a patch.
2844 PH/24 Include config.h inside local_scan.h so that configuration settings are
2847 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2848 that in an ACL for RCPT (for example), you can examine exactly what was
2851 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2852 commands, but it was not correctly comparing the address with the actual
2853 client host address. Thus, it would show the EHLO address in Received:
2854 header lines when this was not necessary.
2856 PH/27 Added the % operator to ${eval:}.
2858 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2859 it should be ignoring failures (because with -C, for example, it has lost
2860 privilege). It wasn't ignoring creation failures other than "already
2863 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2866 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2867 because an input line was too long, either on its own, or by virtue of
2868 too many continuations, the temporary file was not being removed, and the
2869 return code was incorrect.
2871 PH/31 Missing "BOOL" in function definition in filtertest.c.
2873 PH/32 Applied Sieve patches from the maintainer.
2875 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2877 PH/33 Added "verify = not_blind".
2879 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2880 Local/Makefile (with some defaults set). These are used in built scripts
2881 such as exicyclog, but they have never been used in the exim_install
2882 script (though there are many overriding facilities there). I have
2883 arranged that the exim_install script now takes note of these two
2886 PH/35 Installed configuration files for Dragonfly.
2888 PH/36 When a locally submitted message by a trusted user did not contain a
2889 From: header, and the sender address was obtained from -f or from an SMTP
2890 MAIL command, and the trusted user did not use -F to supply a sender
2891 name, $originator_name was incorrectly used when constructing a From:
2892 header. Furthermore, $originator_name was used for submission mode
2893 messages from external hosts without From: headers in a similar way,
2894 which is clearly wrong.
2896 PH/37 Added control=suppress_local_fixups.
2898 PH/38 When log_selector = +received_sender was set, and the addition of the
2899 sender made the log line's construction buffer exactly full, or one byte
2900 less than full, an overflow happened when the terminating "\n" was
2903 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2904 when the result of a list match is failure because a DNS lookup failed.
2906 PH/40 RM_COMMAND is now used in the building process.
2908 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2909 the "build-* directories that it finds.
2911 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2912 address in a domain literal was a prefix of an interface address.
2914 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2915 when verifying a sender address, unless rewrite_headers is false.
2917 PH/44 Wrote a long comment about why errors_to addresses are verified as
2918 recipients, not senders.
2920 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2921 the ratelimit ACL was added.
2923 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2925 PH/46 Added extra information about PostgreSQL errors to the error string.
2927 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2928 avoiding the need to install special zones in a real server. This is
2929 backwards compatible; if it can't find the fake resolver, it drops back.
2930 Thus, both old and new test suites can be run.
2932 TF/02 Added util/ratelimit.pl
2934 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2935 clock is set back in time.
2937 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2938 Candler <B.Candler@pobox.com>.
2940 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2941 OK for addresses that are the result of redirections.
2943 PH/48 A number of further additions for the benefit of the new test suite,
2944 including a fake gethostbyname() that interfaces to the fake DNS resolver
2947 TF/06 The fix for widen_domains has also been applied to qualify_single and
2948 search_parents which are the other dnslookup options that can cause
2951 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2954 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2956 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2957 service" if it encountered an MX record with an empty target hostname.
2958 The message is now "an MX or SRV record indicated no SMTP service".
2960 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2961 if the DNS of the sending site is misconfigured. This is quite a
2962 common situation. This change restores the behaviour of treating a
2963 helo verification defer as a failure.
2965 PH/51 If self=fail was set on a router, the bounce message did not include the
2966 actual error message.
2972 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2974 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2975 command itself ended in error, the underlying message about the transport
2976 filter timeout was being overwritten with the pipe command error. Now the
2977 underlying error message should be appended to the second error message.
2979 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2981 PH/02 Reduce the amount of output that "make" produces by default. Full output
2982 can still be requested.
2984 PH/03 The warning log line about a condition test deferring for a "warn" verb
2985 was being output only once per connection, rather than after each
2986 occurrence (because it was using the same function as for successful
2987 "warn" verbs). This seems wrong, so I have changed it.
2989 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2990 it should not have, which might have caused a crash in the right
2991 circumstances, but probably never did.
2993 PH/04 Installed a modified version of Tony Finch's patch to make submission
2994 mode fix the return path as well as the Sender: header line, and to
2995 add a /name= option so that you can make the user's friendly name appear
2998 TF/03 Added the control = fakedefer ACL modifier.
3000 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3001 Mark Lowes for thorough testing.
3003 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3005 TK/03 Merged latest SRS patch from Miles Wilton.
3007 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3008 with the definition in sysexits.h (which is #included earlier).
3009 Fortunately, Exim does not actually use EX_OK. The code used to try to
3010 preserve the sysexits.h value, by assumimg that macro definitions were
3011 scanned for macro replacements. I have been disabused of this notion,
3012 so now the code just undefines EX_OK before #including unistd.h.
3014 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3015 in the smtp transport. When a block could not be written in a single
3016 write() function, the timeout was being re-applied to each part-write.
3017 This seems wrong - if the receiver was accepting one byte at a time it
3018 would take for ever. The timeout is now adjusted when this happens. It
3019 doesn't have to be particularly precise.
3021 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3022 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3024 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3025 without a domain if the check to <postmaster@domain> fails.
3027 SC/01 Eximstats: added -xls and the ability to specify output files
3028 (patch written by Frank Heydlauf).
3030 SC/02 Eximstats: use FileHandles for outputing results.
3032 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3034 SC/04 Eximstats: fixed display of large numbers with -nvr option
3036 SC/05 Eximstats: fixed merging of reports with empty tables.
3038 SC/06 Eximstats: added the -include_original_destination flag
3040 SC/07 Eximstats: removed tabs and trailing whitespace.
3042 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3044 TK/06 MBOX spool code: Add real "From " MBOX separator line
3045 so the .eml file is really in mbox format (even though
3046 most programs do not really care). Patch from Alex Miller.
3048 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3049 The latter is generated from $received_to and is only set if the
3050 message has one envelope recipient. SA can use these headers,
3051 obviously out-of-the-box. Patch from Alex Miller.
3053 PH/08 The ${def test on a variable was returning false if the variable's
3054 value was "0", contrary to what the specification has always said!
3055 The result should be true unless the variable is empty.
3057 PH/09 The syntax error of a character other than { following "${if
3058 def:variable_name" (after optional whitespace) was not being diagnosed.
3059 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3060 accidental colon was present, for example, could give incorrect results.
3062 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3063 result is used (not including appendfile, where other changes are about
3066 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3067 This involved changing a lot of size variables from int to off_t. It
3068 should work with maildirs and everything.
3070 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3071 spamd dying while we are connected to it.
3073 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3076 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3077 function for BDB 4.3.
3079 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3081 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3082 the sources for PCRE have been re-arranged and more files are now
3085 PH/15 The code I had for printing potentially long long variables in PH/11
3086 above was not the best (it lost precision). The length of off_t variables
3087 is now inspected at build time, and an appropriate printing format (%ld
3088 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3089 to be "long long int" or "long int". This is needed for the internal
3090 formatting function string_vformat().
3092 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3093 the configuration file to be ":syslog", then the script "guesses" where
3094 the logs files are, rather than using the compiled in default. In our
3095 case the guess is not the same as the compiled default, so the script
3096 suddenly stopped working when I started to use syslog. The patch checks
3097 to see if log_file_path is "". If so, it attempts to read it from exim
3098 with no configuration file to get the compiled in version, before it
3099 falls back to the previous guessing code."
3101 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3102 implementing BATV in an Exim configuration. See NewStuff for the gory
3105 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3106 Makefile that are specific to HP-UX.
3108 PH/18 If the "use_postmaster" option was set for a recipient callout together
3109 with the "random" option, the postmaster address was used as the MAIL
3110 FROM address for the random test, but not for the subsequent recipient
3111 test. It is now used for both.
3113 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3114 patch removes a few documentation additions to RFC 3028, because the
3115 latest draft now contains them. It adds the new en;ascii-case comparator
3116 and a new error check for 8bit text in MIME parts. Comparator and
3117 require names are now matched exactly. I enabled the subaddress
3118 extension, but it is not well tested yet (read: it works for me)."
3120 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3121 rework some of the code of TK/09 above to avoid the hardwired use of
3122 "%lld" and "long long". Replaced the call to snprintf() with a call to
3125 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3126 records point to non-existent hosts", "retry timeout exceeded", and
3127 "retry time not reached for any host after a long failure period".
3129 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3130 experimental DomainKeys support:
3132 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3133 (2) On an error such as an illegally used "control", the wrong name for
3134 the control was given.
3136 These problems did NOT occur unless DomainKeys support was compiled.
3138 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3140 PH/24 Added ${if match_ip condition.
3142 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3143 databases so that it will be absolutely obvious if a crash occurs in the
3144 DB library. This is a regular occurrence (often caused by mis-matched
3147 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3148 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3149 picked up on a user's system that detects such things. There doesn't seem
3150 to be a gcc warning option for this - only an attribute that has to be
3151 put on the function's prototype. It seems that in Fedora Core 4 they have
3152 set this on a number of new functions. No doubt there will be more in due
3155 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3156 specify a transport. However, if an address that was verified by such a
3157 router was the subject of a callout, Exim crashed because it tried to
3158 read the rcpt_include_affixes from the non-existent transport. Now it
3159 just assumes that the setting of that option is false. This bug was
3160 introduced by 4.51/PH/31.
3162 PH/28 Changed -d+all to exclude +memory, because that information is very
3163 rarely of interest, but it makes the output a lot bigger. People tend to
3164 do -d+all out of habit.
3166 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3167 code in os-type was giving problems when libc.so lives in lib64, like on
3170 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3171 aren't the modern standard, and it seems that some systems' include files
3172 don't always have them. Exim was already checking for some of the newer
3173 ones like T_AAAA, and defining it itself. I've added checks for all the
3174 record types that Exim uses.
3176 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3177 not automatically generating a new one, as it is supposed to. This
3178 prevented TLS from working. If the file did exist, but contained invalid
3179 data, a new version was generated, as expected. It was only the case of a
3180 non-existent file that was broken.
3182 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3183 with a change in libdomainkeys > 0.64.
3185 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3186 from DNS. If the selector record carries the flag, it now has
3187 precedence over the domain-wide flag.
3189 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3191 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3192 the use of a port name that isn't defined in /etc/services) occurred, the
3193 message was deferred as in a normal delivery, and thus remained on the
3194 spool, instead of being failed because of the mua_wrapper setting. This
3195 is now fixed, and I tidied up some of the mua_wrapper messages at the
3198 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3199 the messages in a hash of arrays rather than using individual hashes.
3200 This is a bit cleaner and results in dramatic memory savings, albeit
3201 at a slight CPU cost.
3203 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3204 as requested by Marc Sherman.
3206 SC/10 Eximstats: added histograms for user specified patterns as requested
3209 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3211 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3212 fopen() in the content-scanning modules that did not already have it.
3218 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3219 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3221 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3223 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3225 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3226 to test Sieve filters that use "vacation".
3228 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3229 that changes the way the GnuTLS parameters are stored in the cache file.
3230 The new format can be generated externally. For backward compatibility,
3231 if the data in the cache doesn't make sense, Exim assumes it has read an
3232 old-format file, and it generates new data and writes a new file. This
3233 means that you can't go back to an older release without removing the
3236 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3237 work if there are any delivery delays because "one_time" forces the
3238 parent to be marked "delivered", so its unseen clone is never tried
3239 again. For this reason, Exim now forbids the simultaneous setting of
3242 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3243 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3244 entry if you want to know the details. The fix, however, neglected to
3245 consider the case when local delivery batching is involved. The test for
3246 "previously delivered" was not happening when checking to see if an
3247 address could be batched with a previous (undelivered) one; under
3248 certain circumstances this could lead to multiple deliveries to the same
3251 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3252 in its include files, and this causes problems building Exim.
3254 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3255 header_syntax) but Exim was just ignoring anything given after a slash.
3256 In particular, this caused confusion with an attempt to use "verify =
3257 reverse_host_lookup/defer_ok". An error is now given when options are
3258 supplied for verify items that do not have them. (Maybe reverse_host_
3259 lookup should have a defer_ok option, but that's a different point.)
3261 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3262 defined by RFC 821) to 2048, because there were problems with some AUTH
3263 commands, and RFC 1869 says the size should be increased for extended
3264 SMTP commands that take arguments.
3266 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3269 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3270 "unknown" error; now it says that the functionality isn't in the binary.
3272 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3273 an address' error message when a string expansion fails (syntax or
3274 whatever). Otherwise the password may appear in the log. Following change
3275 PH/42 below, there is no longer a chance of it appearing in a bounce
3278 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3280 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3281 of its addresses was ever tried. (Bugzilla bug #2.)
3283 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3284 the result incorrectly in the debug output. (It correctly added a newline
3285 to what was transported.)
3287 TF/01 Added $received_time.
3289 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3290 commented out examples of how to interface to a virus scanner and to
3291 SpamAssassin. Also added commented examples of av_scanner and
3292 spamd_address settings.
3294 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3295 and controls are allowed in which ACLs. There were a couple of minor
3296 errors. Some of the entries in the conditions table (which is a table of
3297 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3298 negation of where the condition IS allowed.
3300 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3302 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3303 header file does not have a version number, so I've had to invent a new
3304 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3305 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3306 radiusclient), but it was contributed by a Radius user.
3308 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3309 files or directories, for OpenSSL.
3311 PH/19 When an Exim process that is running as root has to create an Exim log
3312 file, it does so in a subprocess that runs as exim:exim so as to get the
3313 ownership right at creation (otherwise, other Exim processes might see
3314 the file with the wrong ownership). There was no test for failure of this
3315 fork() call, which would lead to the process getting stuck as it waited
3316 for a non-existent subprocess. Forks do occasionally fail when resources
3317 run out. I reviewed all the other calls to fork(); they all seem to check
3320 PH/20 When checking for unexpected SMTP input at connect time (before writing
3321 the banner), Exim was not dealing correctly with a non-positive return
3322 from the read() function. If the client had disconnected by this time,
3323 the result was a log entry for a synchronization error with an empty
3324 string after "input=" when read() returned zero. If read() returned -1
3325 (an event I could not check), uninitialized data bytes were printed.
3326 There were reports of junk text (parts of files, etc) appearing after
3329 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3331 PH/22 Added support for macro redefinition, and (re)definition in between
3332 driver and ACL definitions.
3334 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3335 forgetting to use the resulting value; it was using the unexpanded value.
3337 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3338 hadn't been configured. The fix is from Juergen Kreileder, who
3339 understands it better than I do:
3341 "Here's what I see happening with three configured cyrus_sasl
3342 authenticators configured (plain, login, cram-md5):
3344 On startup auth_cyrus_sasl_init() gets called for each of these.
3345 This means three calls to sasl_listmech() without a specified mech_list.
3346 => SASL tests which mechs of all available mechs actually work
3347 => three warnings about OTP not working
3348 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3350 With the patch, sasl_listmech() also gets called three times. But now
3351 SASL's mech_list option is set to the server_mech specified in the the
3352 authenticator. Or in other words, the answer from sasl_listmech()
3353 gets limited to just the mech you're testing for (which is different
3355 => the return list contains just 'plain' or 'login', 'cram-md5' or
3356 nothing depending on the value of ob->server_mech.
3358 I've just tested the patch: Authentication still works fine,
3359 unavailable mechs specified in the exim configuration are still
3360 caught, and the auth.log warnings about OTP are gone."
3362 PH/25 When debugging is enabled, the contents of the command line are added
3363 to the debugging output, even when log_selector=+arguments is not
3366 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3367 answer is "GNU", and only if the return is "GNU/something" is the answer
3370 PH/27 $acl_verify_message is now set immediately after the failure of a
3371 verification in an ACL, and so is available in subsequent modifiers. In
3372 particular, the message can be preserved by coding like this:
3374 warn !verify = sender
3375 set acl_m0 = $acl_verify_message
3377 Previously, $acl_verify_message was set only while expanding "message"
3378 and "log_message" when a very denied access.
3380 PH/28 Modified OS/os.c-Linux with
3382 -#ifndef OS_LOAD_AVERAGE
3383 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3385 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3386 nomenclature these days.)
3388 PH/29 Installed patch from the Sieve maintainer that adds the options
3389 sieve_useraddress and sieve_subaddress to the redirect router.
3391 PH/30 In these circumstances:
3392 . Two addresses routed to the same list of hosts;
3393 . First host does not offer TLS;
3394 . First host accepts first address;
3395 . First host gives temporary error to second address;
3396 . Second host offers TLS and a TLS session is established;
3397 . Second host accepts second address.
3398 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3399 and peerdn, if requested) that were in fact used only for the second
3402 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3403 attention to any local part prefix or suffix that was matched by the
3404 router that accepted the address. It now behaves in the same way as it
3405 does for delivery: the affixes are removed from the local part unless
3406 rcpt_include_affixes is set on the transport.
3408 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3409 timeout during the DATA phase of an incoming message.
3411 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3412 applied a patch sanctioned by the Sieve maintainer.
3414 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3415 the uid or gid is negative. A case of a negative gid caused this to be
3416 noticed. The fix allows for either to be negative.
3418 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3419 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3422 PH/36 The change PH/12 above was broken. Fixed it.
3424 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3425 the grounds that routing the same address twice would always produce the
3426 same answer. This might have been true once, but it is certainly no
3427 longer true now. Routing a child address may depend on the previous
3428 routing that produced that child. Some complicated redirection strategies
3429 went wrong when messages had multiple recipients, and made Exim's
3430 behaviour dependent on the order in which the addresses were given.
3432 I have moved the duplicate checking until after the routing is complete.
3433 Exim scans the addresses that are assigned to local and remote
3434 transports, and removes any duplicates. This means that more work will be
3435 done, as duplicates will always all be routed, but duplicates are
3436 presumably rare, so I don't expect this is of any significance.
3438 For deliveries to pipes, files, and autoreplies, the duplicate checking
3439 still happens during the routing process, since they are not going to be
3442 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3443 It corrects a timeout issue with spamd. This is Ian's comment: "The
3444 background is that sometimes spamd either never reads data from a
3445 connection it has accepted, or it never writes response data. The exiscan
3446 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3447 blindly assumes that writes won't block so it may never time out."
3449 PH/39 Allow G after quota size as well as K and M.
3451 PH/40 The value set for $authenticated_id in an authenticator may not contain
3452 binary zeroes or newlines because the value is written to log lines and
3453 to spool files. There was no check on this. Now the value is run through
3454 the string_printing() function so that such characters are converted to
3455 printable escape sequences.
3457 PH/41 $message_linecount is a new variable that contains the total number of
3458 lines in the message. Compare $body_linecount, which is the count for the
3461 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3462 bounce and delay warning messages, except in certain special cases, which
3465 (a) An SMTP error message from a remote host;
3466 (b) A message specified in a :fail: redirection;
3467 (c) A message specified in a "fail" command in a system filter;
3468 (d) A message specified in a FAIL return from the queryprogram router;
3469 (e) A message specified by the cannot_route_message router option.
3471 In these cases only, Exim does include the error details in bounce and
3472 warning messages. There are also a few cases where bland messages such
3473 as "unrouteable address" or "local delivery error" are given.
3475 PH/43 $value is now also set for the "else" part of a ${run expansion.
3477 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3478 being worked on, but at least Exim now implements the latest version to
3481 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3482 process to complete was treated as a delivery failure, a timeout while
3483 writing the message to the pipe was logged, but erroneously treated as a
3484 successful delivery. Such timeouts include transport filter timeouts. For
3485 consistency with the overall process timeout, these timeouts are now
3486 treated as errors, giving rise to delivery failures by default. However,
3487 there is now a new Boolean option for the pipe transport called
3488 timeout_defer, which, if set TRUE, converts the failures into defers for
3489 both kinds of timeout. A transport filter timeout is now identified in
3492 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3493 systems where "make" and "gmake" are different, calling "gmake" at top
3494 level broke things. I've arranged for the value of $(MAKE) to be passed
3495 from the Makefile to this script so that it can call the same version of
3499 A note about Exim versions 4.44 and 4.50
3500 ----------------------------------------
3502 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3503 changes of various kinds. As a consequence, a big documentation update was
3504 needed. This delayed the release for rather longer than seemed good, especially
3505 in the light of a couple of (minor) security issues. Therefore, the changes
3506 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3507 release. So 4.44 and 4.50 are in effect two different branches that both start
3510 I have left the 4.50 change log unchanged; it contains all the changes since
3511 4.43. The change log for 4.44 is below; many of its items are identical to
3512 those for 4.50. This seems to be the most sensible way to preserve the
3513 historical information.
3519 1. Minor wording change to the doc/README.SIEVE file.
3521 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3522 computation of the current number of files was incorrect.
3524 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3525 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3528 4. Give more explanation in the error message when the command for a transport
3529 filter fails to execute.
3531 5. There are several places where Exim runs a non-Exim command in a
3532 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3533 was being done only for the command run by the queryprogram router. It is
3534 now done for all such subprocesses. The other cases are: ${run, transport
3535 filters, and the commands run by the lmtp and pipe transports.
3537 6. Added CONFIGURE_GROUP build-time option.
3539 7. Some older OS have a limit of 256 on the maximum number of file
3540 descriptors. Exim was using setrlimit() to set 1000 as a large value
3541 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3542 systems. I've change it so that if it can't get 1000, it tries for 256.
3544 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3545 was an oversight, and furthermore, ever since the addition of extra
3546 controls (e.g. 4.43/32), the checks on when to allow different forms of
3547 "control" were broken. There should now be diagnostics for all cases when a
3548 control that does not make sense is encountered.
3550 9. Added the /retain_sender option to "control=submission".
3552 10. $recipients is now available in the predata ACL (oversight).
3554 11. Tidy the search cache before the fork to do a delivery from a message
3555 received from the command line. Otherwise the child will trigger a lookup
3556 failure and thereby defer the delivery if it tries to use (for example) a
3557 cached ldap connection that the parent has called unbind on.
3559 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3560 of $address_data from the recipient verification was clobbered by the
3561 sender verification.
3563 13. The value of address_data from a sender verification is now available in
3564 $sender_address_data in subsequent conditions in the ACL statement.
3566 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3568 15. Added a new option "connect=<time>" to callout options, to set a different
3571 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3572 was its contents. (It was OK if the option was not defined at all.)
3574 17. A "Completed" log line is now written for messages that are removed from
3575 the spool by the -Mrm option.
3577 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3578 information about exactly what failed.
3580 19. Added -dd to debug only the daemon process.
3582 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3583 handles timeouts, both on the server side and network timeouts. Renamed the
3584 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3586 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3587 It is now set to "smtps".
3589 22. $host_address is now set to the target address during the checking of
3590 ignore_target_hosts.
3592 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3593 being passed; this would have caused $sender_host_name to have been used if
3594 matching the list had actually called for a host name (not very likely,
3595 since this list is usually IP addresses). A host name is now passed as
3598 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3599 code that specifies a non-restarting handler (typically sigaction() in
3600 modern systems) in an attempt to fix a rare and obscure crash bug.
3602 25. Narrowed the window for a race in the daemon that could cause it to ignore
3603 SIGCHLD signals. This is not a major problem, because they are used only to
3604 wake it up if nothing else does.
3606 26. A malformed maildirsize file could cause Exim to calculate negative values
3607 for the mailbox size or file count. Odd effects could occur as a result.
3608 The maildirsize information is now recalculated if the size or filecount
3611 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3612 support for a long time. Removed HAVE_SYS_VFS_H.
3614 28. Installed the latest version of exipick from John Jetmore.
3616 29. In an address list, if the pattern was not a regular expression, an empty
3617 subject address (from a bounce message) matched only if the pattern was an
3618 empty string. Non-empty patterns were not even tested. This was the wrong
3619 because it is perfectly reasonable to use an empty address as part of a
3620 database query. An empty address is now tested by patterns that are
3621 lookups. However, all the other forms of pattern expect the subject to
3622 contain a local part and a domain, and therefore, for them, an empty
3623 address still always fails if the pattern is not itself empty.
3625 30. Exim went into a mad DNS loop when attempting to do a callout where the
3626 host was specified on an smtp transport, and looking it up yielded more
3627 than one IP address.
3629 31. Re-factored the code for checking spool and log partition space into a
3630 function that finds that data and another that does the check. The former
3631 is then used to implement four new variables: $spool_space, $log_space,
3632 $spool_inodes, and $log_inodes.
3634 32. The RFC2047 encoding function was originally intended for short strings
3635 such as real names; it was not keeping to the 75-character limit for
3636 encoded words that the RFC imposes. It now respects the limit, and
3637 generates multiple encoded words if necessary. To be on the safe side, I
3638 have increased the buffer size for the ${rfc2047: expansion operator from
3641 33. It is now permitted to omit both strings after an "if" condition; if the
3642 condition is true, the result is "true". As before, when the second string
3643 is omitted, a false condition yields an empty string. This makes it less
3644 cumbersome to write custom ACL and router conditions.
3646 34. Failure to deliver a bounce message always caused it to be frozen, even if
3647 there was an errors_to setting on the router. The errors_to setting is now
3650 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3651 canonical form (fully expanded) before being placed in
3652 $sender_host_address.
3654 36. The table in the code that translates DNS record types into text (T_A to
3655 "A" for instance) was missing entries for NS and CNAME. It is just possible
3656 that this could have caused confusion if both these types were looked up
3657 for the same domain, because the text type is used as part of Exim's
3658 per-process caching. But the chance of anyone hitting this buglet seems
3661 37. The dnsdb lookup has been extended in a number of ways.
3663 (1) There is a new type, "zns", which walks up the domain tree until it
3664 finds some nameserver records. It should be used with care.
3666 (2) There is a new type, "mxh", which is like "mx" except that it returns
3667 just the host names, not the priorities.
3669 (3) It is now possible to give a list of domains (or IP addresses) to be
3670 looked up. The behaviour when one of the lookups defers can be
3671 controlled by a keyword.
3673 (4) It is now possible to specify the separator character for use when
3674 multiple records are returned.
3676 38. The dnslists ACL condition has been extended: it is now possible to supply
3677 a list of IP addresses and/or domains to be looked up in a particular DNS
3680 39. Added log_selector=+queue_time_overall.
3682 40. When running the queue in the test harness, wait just a tad after forking a
3683 delivery process, to get repeatability of debugging output.
3685 41. Include certificate and key file names in error message when GnuTLS fails
3686 to set them up, because the GnuTLS error message doesn't include the name
3687 of the failing file when there is a problem reading it.
3689 42. Allow both -bf and -bF in the same test run.
3691 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3693 44. The "Exiscan patch" is now merged into the mainline Exim source.
3695 45. Sometimes the final signoff response after QUIT could fail to get
3696 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3697 < 0 before doing a fflush(). This bug looks as though it goes back to the
3698 introduction of TLS in release 3.20, but "sometimes" must have been rare
3699 because the tests only now provoked it.
3701 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3702 (this can affect the format of dates).
3704 47. exim_tidydb, when checking for the continued existence of a message for
3705 which it has found a message-specific retry record, was not finding
3706 messages that were in split spool directories. Consequently, it was
3707 deleting retry records that should have stayed in existence.
3709 48. Steve fixed some bugs in eximstats.
3711 49. The SPA authentication driver was not abandoning authentication and moving
3712 on to the next authenticator when an expansion was forced to fail,
3713 contradicting the general specification for all authenticators. Instead it
3714 was generating a temporary error. It now behaves as specified.
3716 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3717 (the order specifies the preference for clients). The order is now AES256,
3718 AES128, 3DES, ARCFOUR128.
3720 51. Small patch to Sieve code - explicitly set From: when generating an
3723 52. Exim crashed if a remote delivery caused a very long error message to be
3724 recorded - for instance if somebody sent an entire SpamAssassin report back
3725 as a large number of 550 error lines. This bug was coincidentally fixed by
3726 increasing the size of one of Exim's internal buffers (big_buffer) that
3727 happened as part of the Exiscan merge. However, to be on the safe side, I
3728 have made the code more robust (and fixed the comments that describe what
3731 53. Now that there can be additional text after "Completed" in log lines (if
3732 the queue_time_overall log selector is set), a one-byte patch to exigrep
3733 was needed to allow it to recognize "Completed" as not the last thing in
3736 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3737 patch that reportedly fixes this has been added. I am not expert enough to
3738 create a test for it. This is what the patch creator wrote:
3740 "I found a little strange behaviour of ldap code when working with
3741 Windows 2003 AD Domain, where users was placed in more than one
3742 Organization Units. When I tried to give exim partial DN, the exit code
3743 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3744 But simultaneously result of request was absolutely normal ldap result,
3745 so I produce this patch..."
3747 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3748 REFERENCE, so I have modified the code to exclude the patch when that macro
3751 55. Some experimental protocols are using DNS PTR records for new purposes. The
3752 keys for these records are domain names, not reversed IP addresses. The
3753 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3754 leaves it alone. Component reversal etc. now happens only for IP addresses.
3757 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3759 57. Double the size of the debug message buffer (to 2048) so that more of very
3760 long debug lines gets shown.
3762 58. The exicyclog utility now does better if the number of log files to keep
3763 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3765 59. Two changes related to the smtp_active_hostname option:
3767 (1) $smtp_active_hostname is now available as a variable.
3768 (2) The default for smtp_banner uses $smtp_active_hostname instead
3769 of $primary_hostname.
3771 60. The host_aton() function is supposed to be passed a string that is known
3772 to be a valid IP address. However, in the case of IPv6 addresses, it was
3773 not checking this. This is a hostage to fortune. Exim now panics and dies
3774 if the condition is not met. A case was found where this could be provoked
3775 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3776 components; fortuitously, this particular loophole had already been fixed
3777 by change 4.50/55 above.
3779 If there are any other similar loopholes, the new check in host_aton()
3780 itself should stop them being exploited. The report I received stated that
3781 data on the command line could provoke the exploit when Exim was running as
3782 exim, but did not say which command line option was involved. All I could
3783 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3784 running as the user.
3787 61. There was a buffer overflow vulnerability in the SPA authentication code
3788 (which came originally from the Samba project). I have added a test to the
3789 spa_base64_to_bits() function which I hope fixes it.
3792 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3793 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3795 63. The daemon start-up calls getloadavg() while still root for those OS that
3796 need the first call to be done as root, but it missed one case: when
3797 deliver_queue_load_max is set with deliver_drop_privilege. This is
3798 necessary for the benefit of the queue runner, because there is no re-exec
3799 when deliver_drop_privilege is set.
3801 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3802 This has been fixed.
3804 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3805 was in use, was not putting the data itself into the right store pool;
3806 consequently, it could be overwritten for a subsequent message in the same
3807 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3810 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3812 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3813 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3814 the function were treating the return as a boolean value, which happened to
3815 work because 0=false and not-0=true, but is not correct code.
3817 68. The host_aton() function was not handling scoped IPv6 addresses (those
3818 with, for example, "%eth0" on the end) correctly.
3820 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3821 negated items (that is, ~something) in unsigned ints. Some compilers
3822 apparently mutter when there is no cast.
3824 70. If an address verification called from an ACL failed, and did not produce a
3825 user-specific message (i.e. there was only a "system" message), nothing was
3826 put in $acl_verify_message. In this situation, it now puts the system
3829 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3830 session; change 4.31/43 added the unwanted input to the log line - except
3831 that it did not do this in the start of session case. It now does.
3833 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3834 This is wrong and can cause the other end to generate a synchronization
3835 error if it is another Exim or anything else that does the synchronization
3836 check. A QUIT command is no longer sent after a timeout.
3838 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3839 during host lookups.
3841 74. The defer_ok option of callout verification was not working if it was used
3842 when verifying addresses in header lines, that is, for this case:
3844 verify = header_sender/callout=defer_ok
3846 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3847 those file descriptors could be used for SMTP connections. If anything
3848 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3849 could be sent to the SMTP client, causing chaos. The daemon now opens
3850 stdin, stdout, and stderr to /dev/null when it puts itself into the
3853 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3854 log by default. The user can override this with suitable Perl magic.
3856 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3857 the log message when discard triggers, was not working for the DATA ACL or
3858 for the non-SMTP ACL.
3860 78. Error message wording change in sieve.c.
3862 79. If smtp_accept_max_per_host was set, the number of connections could be
3863 restricted to fewer than expected, because the daemon was trying to set up
3864 a new connection before checking whether the processes handling previous
3865 connections had finished. The check for completed processes is now done
3866 earlier. On busy systems, this bug wouldn't be noticed because something
3867 else would have woken the daemon, and it would have reaped the completed
3870 80. If a message was submitted locally by a user whose login name contained one
3871 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3872 It caused a spool format error. I have fixed the spool reading code. A
3873 related problem was that the "from" clause in the Received: line became
3874 illegal because of the space(s). It is now covered by ${quote_local_part.
3876 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3879 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3881 83. Patch from Sieve maintainer to fix unterminated string problem in
3882 "vacation" handling.
3884 84. Some minor changes to the Linux configuration files to help with other
3885 OS variants using glibc.
3887 85. One more patch for Sieve to update vacation handling to latest spec.
3890 ----------------------------------------------------
3891 See the note above about the 4.44 and 4.50 releases.
3892 ----------------------------------------------------
3898 1. Change 4.43/35 introduced a bug that caused file counts to be
3899 incorrectly computed when quota_filecount was set in an appendfile
3902 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3903 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3906 3. Give more explanation in the error message when the command for a transport
3907 filter fails to execute.
3909 4. There are several places where Exim runs a non-Exim command in a
3910 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3911 was being done only for the command run by the queryprogram router. It is
3912 now done for all such subprocesses. The other cases are: ${run, transport
3913 filters, and the commands run by the lmtp and pipe transports.
3915 5. Some older OS have a limit of 256 on the maximum number of file
3916 descriptors. Exim was using setrlimit() to set 1000 as a large value
3917 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3918 systems. I've change it so that if it can't get 1000, it tries for 256.
3920 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3921 was an oversight, and furthermore, ever since the addition of extra
3922 controls (e.g. 4.43/32), the checks on when to allow different forms of
3923 "control" were broken. There should now be diagnostics for all cases when a
3924 control that does not make sense is encountered.
3926 7. $recipients is now available in the predata ACL (oversight).
3928 8. Tidy the search cache before the fork to do a delivery from a message
3929 received from the command line. Otherwise the child will trigger a lookup
3930 failure and thereby defer the delivery if it tries to use (for example) a
3931 cached ldap connection that the parent has called unbind on.
3933 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3934 of $address_data from the recipient verification was clobbered by the
3935 sender verification.
3937 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3938 was its contents. (It was OK if the option was not defined at all.)
3940 11. A "Completed" log line is now written for messages that are removed from
3941 the spool by the -Mrm option.
3943 12. $host_address is now set to the target address during the checking of
3944 ignore_target_hosts.
3946 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3947 being passed; this would have caused $sender_host_name to have been used if
3948 matching the list had actually called for a host name (not very likely,
3949 since this list is usually IP addresses). A host name is now passed as
3952 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3953 code that specifies a non-restarting handler (typically sigaction() in
3954 modern systems) in an attempt to fix a rare and obscure crash bug.
3956 15. Narrowed the window for a race in the daemon that could cause it to ignore
3957 SIGCHLD signals. This is not a major problem, because they are used only to
3958 wake it up if nothing else does.
3960 16. A malformed maildirsize file could cause Exim to calculate negative values
3961 for the mailbox size or file count. Odd effects could occur as a result.
3962 The maildirsize information is now recalculated if the size or filecount
3965 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3966 support for a long time. Removed HAVE_SYS_VFS_H.
3968 18. Updated exipick to current release from John Jetmore.
3970 19. Allow an empty sender to be matched against a lookup in an address list.
3971 Previously the only cases considered were a regular expression, or an
3974 20. Exim went into a mad DNS lookup loop when doing a callout where the
3975 host was specified on the transport, if the DNS lookup yielded more than
3978 21. The RFC2047 encoding function was originally intended for short strings
3979 such as real names; it was not keeping to the 75-character limit for
3980 encoded words that the RFC imposes. It now respects the limit, and
3981 generates multiple encoded words if necessary. To be on the safe side, I
3982 have increased the buffer size for the ${rfc2047: expansion operator from
3985 22. Failure to deliver a bounce message always caused it to be frozen, even if
3986 there was an errors_to setting on the router. The errors_to setting is now
3989 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3990 canonical form (fully expanded) before being placed in
3991 $sender_host_address.
3993 24. Updated eximstats to version 1.33
3995 25. Include certificate and key file names in error message when GnuTLS fails
3996 to set them up, because the GnuTLS error message doesn't include the name
3997 of the failing file when there is a problem reading it.
3999 26. Expand error message when OpenSSL has problems setting up cert/key files.
4002 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4003 (this can affect the format of dates).
4005 28. exim_tidydb, when checking for the continued existence of a message for
4006 which it has found a message-specific retry record, was not finding
4007 messages that were in split spool directories. Consequently, it was
4008 deleting retry records that should have stayed in existence.
4010 29. eximstats updated to version 1.35
4011 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4012 1.35 - bugfix such that pie charts by volume are generated correctly
4014 30. The SPA authentication driver was not abandoning authentication and moving
4015 on to the next authenticator when an expansion was forced to fail,
4016 contradicting the general specification for all authenticators. Instead it
4017 was generating a temporary error. It now behaves as specified.
4019 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4020 (the order specifies the preference for clients). The order is now AES256,
4021 AES128, 3DES, ARCFOUR128.
4023 31. Small patch to Sieve code - explicitly set From: when generating an
4026 32. Exim crashed if a remote delivery caused a very long error message to be
4027 recorded - for instance if somebody sent an entire SpamAssassin report back
4028 as a large number of 550 error lines. This bug was coincidentally fixed by
4029 increasing the size of one of Exim's internal buffers (big_buffer) that
4030 happened as part of the Exiscan merge. However, to be on the safe side, I
4031 have made the code more robust (and fixed the comments that describe what
4034 33. Some experimental protocols are using DNS PTR records for new purposes. The
4035 keys for these records are domain names, not reversed IP addresses. The
4036 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4037 leaves it alone. Component reversal etc. now happens only for IP addresses.
4040 34. The host_aton() function is supposed to be passed a string that is known
4041 to be a valid IP address. However, in the case of IPv6 addresses, it was
4042 not checking this. This is a hostage to fortune. Exim now panics and dies
4043 if the condition is not met. A case was found where this could be provoked
4044 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4045 components; fortuitously, this particular loophole had already been fixed
4046 by change 4.50/55 or 4.44/33 above.
4048 If there are any other similar loopholes, the new check in host_aton()
4049 itself should stop them being exploited. The report I received stated that
4050 data on the command line could provoke the exploit when Exim was running as
4051 exim, but did not say which command line option was involved. All I could
4052 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4053 running as the user.
4056 35. There was a buffer overflow vulnerability in the SPA authentication code
4057 (which came originally from the Samba project). I have added a test to the
4058 spa_base64_to_bits() function which I hope fixes it.
4061 36. The daemon start-up calls getloadavg() while still root for those OS that
4062 need the first call to be done as root, but it missed one case: when
4063 deliver_queue_load_max is set with deliver_drop_privilege. This is
4064 necessary for the benefit of the queue runner, because there is no re-exec
4065 when deliver_drop_privilege is set.
4067 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4068 was in use, was not putting the data itself into the right store pool;
4069 consequently, it could be overwritten for a subsequent message in the same
4070 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4073 38. Sometimes the final signoff response after QUIT could fail to get
4074 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4075 < 0 before doing a fflush(). This bug looks as though it goes back to the
4076 introduction of TLS in release 3.20, but "sometimes" must have been rare
4077 because the tests only now provoked it.
4083 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4084 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4085 Now it returns an integer. A number of calls were still expecting a T/F
4086 return. Fortuitously, in all cases, the tests worked in OK situations,
4087 which is the norm. However, things would have gone wrong on any write
4088 failures on the smtp file descriptor. This function is used when sending
4089 messages over SMTP and also when doing verify callouts.
4091 2. When Exim is called to do synchronous delivery of a locally submitted
4092 message (the -odf or -odi options), it no longer closes stderr before doing
4095 3. Implemented the mua_wrapper option.
4097 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4099 5. Implemented the functions header_remove(), header_testname(),
4100 header_add_at_position(), and receive_remove_recipient(), and exported them
4103 6. If an ACL "warn" statement specified the addition of headers, Exim already
4104 inserted X-ACL-Warn: at the start if there was no header name. However, it
4105 was not making this test for the second and subsequent header lines if
4106 there were newlines in the string. This meant that an invalid header could
4107 be inserted if Exim was badly configured.
4109 7. Allow an ACL "warn" statement to add header lines at the start or after all
4110 the Received: headers, as well as at the end.
4112 8. Added the rcpt_4xx retry error code.
4114 9. Added postmaster_mailfrom=xxx to callout verification option.
4116 10. Added mailfrom=xxxx to the callout verification option, for verify=
4119 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4120 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4122 12. Inserted some casts to stop certain compilers warning when using pointer
4123 differences as field lengths or precisions in printf-type calls (mostly
4124 affecting debugging statements).
4126 13. Added optional readline() support for -be (dynamically loaded).
4128 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4129 same clock tick as a message's arrival, so that its received time was the
4130 same as the "first fail" time on the retry record, and that message
4131 remained on the queue past the ultimate address timeout, every queue runner
4132 would try a delivery (because it was past the ultimate address timeout) but
4133 after another failure, the ultimate address timeout, which should have then
4134 bounced the address, did not kick in. This was a "< instead of <=" error;
4135 in most cases the first failure would have been in the next clock tick
4136 after the received time, and all would be well.
4138 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4139 being recognized when the domain list was tested by the match_domain
4140 condition in an expansion string.
4142 16. Added the ${str2b64: operator.
4144 17. Exim was always calling setrlimit() to set a large limit for the number of
4145 processes, without checking whether the existing limit was already
4146 adequate. (It did check for the limit on file descriptors.) Furthermore,
4147 errors from getrlimit() and setrlimit() were being ignored. Now they are
4148 logged to the main and panic logs, but Exim does carry on, to try to do its
4149 job under whatever limits there are.
4151 18. Imported PCRE 5.0.
4153 19. Trivial typo in log message " temporarily refused connection" (the leading
4156 20. If the log selector return_path_on_delivery was set and an address was
4157 redirected to /dev/null, the delivery process crashed because it assumed
4158 that a return path would always be set for a "successful" delivery. In this
4159 case, the whole delivery is bypassed as an optimization, and therefore no
4162 21. Internal re-arrangement: the function for sending a challenge and reading
4163 a response while authentication was assuming a zero-terminated challenge
4164 string. It's now changed to take a pointer and a length, to allow for
4165 binary data in such strings.
4167 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4169 23. Exim was not respecting finduser_retries when seeking the login of the
4170 uid under which it was called; it was always trying 10 times. (The default
4171 setting of finduser_retries is zero.) Also, it was sleeping after the final
4172 failure, which is pointless.
4174 24. Implemented tls_on_connect_ports.
4176 25. Implemented acl_smtp_predata.
4178 26. If the domain in control=submission is set empty, Exim assumes that the
4179 authenticated id is a complete email address when it generates From: or
4180 Sender: header lines.
4182 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4183 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4184 chown and chgrp in /bin and hostname in /usr/bin.
4186 28. Exim was keeping the "process log" file open after each use, just as it
4187 does for the main log. This opens the possibility of it remaining open for
4188 long periods when the USR1 signal hits a daemon. Occasional processlog
4189 errors were reported, that could have been caused by this. Anyway, it seems
4190 much more sensible not to leave this file open at all, so that is what now
4193 29. The long-running daemon process does not normally write to the log once it
4194 has entered its main loop, and it closes the log before doing so. This is
4195 so that log files can straightforwardly be renamed and moved. However,
4196 there are a couple of unusual error situations where the daemon does write
4197 log entries, and I had neglected to close the log afterwards.
4199 30. The text of an SMTP error response that was received during a remote
4200 delivery was being truncated at 512 bytes. This is too short for some of
4201 the long messages that one sometimes sees. I've increased the limit to
4204 31. It is now possible to make retry rules that apply only when a message has a
4205 specific sender, in particular, an empty sender.
4207 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4208 it possible to be selective about when SMTP synchronization is enforced.
4210 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4212 32. Implemented hosts_connection_nolog.
4214 33. Added an ACL for QUIT.
4216 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4219 35. Added mailbox_size and mailbox_filecount to appendfile.
4221 36. Added control = no_multiline_responses to ACLs.
4223 37. There was a bug in the logic of the code that waits for the clock to tick
4224 in the case where the clock went backwards by a substantial amount such
4225 that the microsecond fraction of "now" was more than the microsecond
4226 fraction of "then" (but the whole seconds number was less).
4228 38. Added support for the libradius Radius client library this is found on
4229 FreeBSD (previously only the radiusclient library was supported).
4235 1. When certain lookups returned multiple values in the form name=value, the
4236 quoting of the values was not always being done properly. Specifically:
4237 (a) If the value started with a double quote, but contained no whitespace,
4239 (b) If the value contained whitespace other than a space character (i.e.
4240 tabs or newlines or carriage returns) it was not quoted.
4241 This fix has been applied to the mysql and pgsql lookups by writing a
4242 separate quoting function and calling it from the lookup code. The fix
4243 should probably also be applied to nisplus, ibase and oracle lookups, but
4244 since I cannot test any of those, I have not disturbed their existing code.
4246 2. A hit in the callout cache for a specific address caused a log line with no
4247 reason for rejecting RCPT. Now it says "Previous (cached) callout
4248 verification failure".
4250 3. There was an off-by-one bug in the queryprogram router. An over-long
4251 return line was truncated at 256 instead of 255 characters, thereby
4252 overflowing its buffer with the terminating zero. As well as fixing this, I
4253 have increased the buffer size to 1024 (and made a note to document this).
4255 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4256 when Exim is waiting for an SMTP response from a remote server, Exim
4257 restarts its select() call on the socket, thereby resetting its timeout.
4258 This is not a problem when such interrupts are rare. Somebody set up a cron
4259 job to run exiwhat every 2 minutes, which is less than the normal select()
4260 timeout (5 or 10 minutes). This meant that the select() timeout never
4261 kicked in because it was always reset. I have fixed this by comparing the
4262 time when an interrupt arrives with the time at the start of the first call
4263 to select(). If more time than the timeout has elapsed, the interrupt is
4264 treated as a timeout.
4266 5. Some internal re-factoring in preparation for the addition of Sieve
4267 extensions (by MH). In particular, the "personal" test is moved to a
4268 separate function, and given an option for scanning Cc: and Bcc: (which is
4269 not set for Exim filters).
4271 6. When Exim created an email address using the login of the caller as the
4272 local part (e.g. when creating a From: or Sender: header line), it was not
4273 quoting the local part when it contained special characters such as @.
4275 7. Installed new OpenBSD configuration files.
4277 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4278 try to make them clearer.
4280 9. Callout options, other than the timeout value, were being ignored when
4281 verifying sender addresses in header lines. For example, when using
4283 verify = header_sender/callout=no_cache
4285 the cache was (incorrectly) being used.
4287 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4288 only the Cygwin environment.
4290 11. When return_path_on_delivery was set as a log selector, if different remote
4291 addresses in the same message used different return paths and parallel
4292 remote delivery occurred, the wrong values would sometimes be logged.
4293 (Whenever a remote delivery process finished, the return path value from
4294 the most recently started remote delivery process was logged.)
4296 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4297 lines when AUTH and/or TLS are in use. This is the "received protocol"
4298 field. Exim used to use "asmtp" for authenticated SMTP, without any
4299 indication (in the protocol name) for TLS use. Now it follows the RFC and
4300 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4301 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4302 names appear in log lines as well as in Received: header lines.
4304 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4305 extensions, and comparison tests, and to fix some bugs.
4307 14. Changes to the "personal" filter test:
4309 (1) The test was buggy in that it was just doing the equivalent of
4310 "contains" tests on header lines. For example, if a user's address was
4311 anne@some.where, the "personal" test would incorrectly be true for
4313 To: susanne@some.where
4315 This test is now done by extracting each address from the header in turn,
4316 and checking the entire address. Other tests that are part of "personal"
4317 are now done using regular expressions (for example, to check local parts
4318 of addresses in From: header lines).
4320 (2) The list of non-personal local parts in From: addresses has been
4321 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4322 taken from the Sieve specification recommendations.
4324 (3) If the message contains any header line starting with "List-" it is
4325 treated as non-personal.
4327 (4) The test for "circular" in the Subject: header line has been removed
4328 because it now seems ill-conceived.
4330 15. Minor typos in src/EDITME comments corrected.
4332 16. Installed latest exipick from John Jetmore.
4334 17. If headers_add on a router specified a text string that was too long for
4335 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4336 of string_sprintf() is now avoided.
4338 18. $message_body_size was not set (it was always zero) when running the DATA
4339 ACL and the local_scan() function.
4341 19. For the "mail" command in an Exim filter, no default was being set for
4342 the once_repeat time, causing a random time value to be used if "once" was
4343 specified. (If the value happened to be <= 0, no repeat happened.) The
4344 default is now 0s, meaning "never repeat". The "vacation" command was OK
4345 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4346 (I found it when inspecting the code).
4348 20. There is now an overall timeout for performing a callout verification. It
4349 defaults to 4 times the callout timeout, which applies to individual SMTP
4350 commands during the callout. The overall timeout applies when there is more
4351 than one host that can be tried. The timeout is checked before trying the
4352 next host. This prevents very long delays if there are a large number of
4353 hosts and all are timing out (e.g. when the network connections are timing
4354 out). The value of the overall timeout can be changed by specifying an
4355 additional sub-option for "callout", called "maxwait". For example:
4357 verify = sender/callout=5s,maxwait=20s
4359 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4360 to the end before writing, but this should make it even safer).
4362 22. Exim was forgetting that it had advertised PIPELINING for the second and
4363 subsequent messages on an SMTP connection. It was also not resetting its
4364 memory on STARTTLS and an internal HELO.
4366 23. When Exim logs an SMTP synchronization error within a session, it now
4367 records whether PIPELINING has been advertised or not.
4369 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4370 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4371 rather than long int.
4373 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4375 26. Added the never_mail option to autoreply.
4381 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4382 crash if the getsockname() call failed; this can happen if a connection is
4383 closed very soon after it is established. The problem was simply in the
4384 order in which certain operations were done, causing Exim to try to write
4385 to the SMTP stream before it had set up the file descriptor. The bug has
4386 been fixed by making things happen in the correct order.
4392 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4393 before the connection was closed, thus losing the rejection response.
4395 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4396 some early Solaris releases, but causes trouble in current releases where
4397 socklen_t is defined.
4399 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4402 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4405 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4406 that releasing the top of it at the end releases what was used for sub-
4407 expansions (unless the block got too big). However, discard this block if
4408 the first thing is a variable or header, so that we can use its block when
4409 it is dynamic (useful for very large $message_headers, for example).
4411 6. Lookups now cache *every* query, not just the most recent. A new, separate
4412 store pool is used for this. It can be recovered when all lookup caches are
4413 flushed. Lookups now release memory at the end of their result strings.
4414 This has involved some general refactoring of the lookup sources.
4416 7. Some code has been added to the store_xxx() functions to reduce the amount
4417 of flapping under certain conditions.
4419 8. log_incoming_interface used to affect only the <= reception log lines. Now
4420 it causes the local interface and port to be added to several more SMTP log
4421 lines, for example "SMTP connection from", and rejection lines.
4423 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4425 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4427 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4428 could be overwritten at the end of the current message (or the start of a
4429 new message if it was set in a HELO ACL). The value is now preserved for
4430 the duration of the SMTP connection.
4432 12. If a transport had a headers_rewrite setting, and a matching header line
4433 contained an unqualified address, that address was qualified, even if it
4434 did not match any rewriting rules. The underlying bug was that the values
4435 of the flags that permit the existence of unqualified sender and recipient
4436 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4437 non-local messages, and by -bnq for local messages) were not being
4438 preserved with the message after it was received.
4440 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4441 "next input=" as part of the text comprising the host identity instead of
4442 the correct text. The code was using the same buffer for two different
4443 strings. However, depending on which order the printing function evaluated
4444 its arguments, the bug did not always show up. Under Linux, for example, my
4445 test suite worked just fine.
4447 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4448 doesn't work with some older versions of Perl. It has been changed to "my",
4449 which in any case is probably the better facility to use.
4451 15. A really picky compiler found some instances of statements for creating
4452 error messages that either had too many or two few arguments for the format
4455 16. The size of the buffer for calls to the DNS resolver has been increased
4456 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4457 for addresses that have a lot of PTR records. This alleviates a problem; it
4458 does not fully solve it.
4460 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4461 buffer now truncates the list and logs the incident, which is the same
4462 action as happens when Exim is looking up a host name and its aliases.
4463 Previously in this situation something unpredictable would happen;
4464 sometimes it was "internal error: store_reset failed".
4466 18. If a server dropped the connection unexpectedly when an Exim client was
4467 using GnuTLS and trying to read a response, the client delivery process
4468 crashed while trying to generate an error log message.
4470 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4471 string, for example:
4473 warn message = H1: something\nH2: something
4475 the text was added as a single header line from Exim's point of view
4476 though it ended up OK in the delivered message. However, searching for the
4477 second and subsequent header lines using $h_h2: did not work. This has been
4478 fixed. Similarly, if a system filter added multiple headers in this way,
4479 the routers could not see them.
4481 20. Expanded the error message when iplsearch is called with an invalid key to
4482 suggest using net-iplsearch in a host list.
4484 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4485 ACLs are no longer actually imposed (and a message to that effect is
4488 22. If a "gecos" field in a passwd entry contained escaped characters, in
4489 particular, if it contained a \" sequence, Exim got it wrong when building
4490 a From: or a Sender: header from that name. A second bug also caused
4491 incorrect handling when an unquoted " was present following a character
4492 that needed quoting.
4494 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4495 was not being matched caselessly.
4497 24. Arranged for all hyphens in the exim.8 source to be escaped with
4500 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4501 properly. Recipient callout cache records were still being keyed to include
4502 the sender, even when use_sender was set false. This led to far more
4503 callouts that were necessary. The sender is no longer included in the key
4504 when use_sender is false.
4506 26. Added "control = submission" modifier to ACLs.
4508 27. Added the ${base62d: operator to decode base 62 numbers.
4510 28. dnsdb lookups can now access SRV records.
4512 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4513 the configuration file.
4515 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4516 (-v) mode. This makes the output for a verbose queue run more intelligible.
4518 31. Added a use_postmaster feature to recipient callouts.
4520 32. Added the $body_zerocount variable, containing the number of binary zero
4521 bytes in the message body.
4523 33. The time of last modification of the "new" subdirectory is now used as the
4524 "mailbox time last read" when there is a quota error for a maildir
4527 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4529 35. Added +ignore_unknown as a special item in host lists.
4531 36. Code for decoding IPv6 addresses in host lists is now included, even if
4532 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4533 address was recognized as an IP address, but was then not correctly decoded
4534 into binary, causing unexpected and incorrect effects when compared with
4541 1. Very minor rewording of debugging text in manualroute to say "list of
4542 hosts" instead of "hostlist".
4544 2. If verify=header_syntax was set, and a header line with an unqualified
4545 address (no domain) and a large number of spaces between the end of the
4546 name and the colon was received, the reception process suffered a buffer
4547 overflow, and (when I tested it) crashed. This was caused by some obsolete
4548 code that should have been removed. The fix is to remove it!
4550 3. When running in the test harness, delay a bit after writing a bounce
4551 message to get a bit more predictability in the log output.
4553 4. Added a call to search_tidyup() just before forking a reception process. In
4554 theory, someone could use a lookup in the expansion of smtp_accept_max_
4555 per_host which, without the tidyup, could leave open a database connection.
4557 5. Added the variables $recipient_data and $sender_data which get set from a
4558 lookup success in an ACL "recipients" or "senders" condition, or a router
4559 "senders" option, similar to $domain_data and $local_part_data.
4561 6. Moved the writing of debug_print from before to after the "senders" test
4564 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4565 problems for message scanning, either using a data ACL, or using
4566 local_scan() because the Received: header was not generated till after they
4567 were called (in order to set the time as the time of reception completion).
4568 I have revised the way this works. The header is now generated after the
4569 body is received, but before the ACL or local_scan() are called. After they
4570 are run, the timestamp in the header is updated.
4576 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4577 before starting a queue runner without re-exec. This happened only when
4578 deliver_drop_privilege was set or when the Exim user was set to root. The
4579 effect of the bug was that timeouts during subsequent deliveries caused
4580 crashes instead of being properly handled. The handler is now left at its
4581 default (and expected) setting.
4583 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4584 message, again when deliver_drop_privilege is set or Exim is run as root.
4585 The bug described in (1) was not present in this case, but the tidying up
4586 of the other signals was missing. I have made the two cases consistent.
4588 3. The ignore_target_hosts setting on a manualroute router was being ignored
4589 for hosts that were looked up using the /MX notation.
4591 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4594 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4595 operated on the sender address. After changing the $sender_address to <>
4596 for the sender address verify, Exim was re-instated it as the original
4597 (before rewriting) address, but remembering that it had rewritten it, so it
4598 wasn't rewriting it again. This bug also had the effect of breaking the
4599 sender address verification caching when the sender address was rewritten.
4601 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4602 This has been changed so that if the ip literal address matches
4603 ignore_target_hosts, the router declines.
4605 7. Added expansion conditions match_domain, match_address, and match_local_
4606 part (NOT match_host).
4608 8. The placeholder for the Received: header didn't have a length field set.
4610 9. Added code to Exim itself and to exim_lock to test for a specific race
4611 condition that could lead to file corruption when using MBX delivery. The
4612 issue is with the lockfile that is created in /tmp. If this file is removed
4613 after a process has opened it but before that process has acquired a lock,
4614 there is the potential for a second process to recreate the file and also
4615 acquire a lock. This could lead to two Exim processes writing to the file
4616 at the same time. The added code performs the same test as UW imapd; it
4617 checks after acquiring the lock that its file descriptor still refers to
4618 the same named file.
4620 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4621 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4624 11. Added the smtp_active_hostname option. If used, this will typically be made
4625 to depend on the incoming interface address. Because $interface_address is
4626 not set up until the daemon has forked a reception process, error responses
4627 that can happen earlier (such as "too many connections") no longer contain
4630 12. If an expansion in a condition on a "warn" statement fails because a lookup
4631 defers, the "warn" statement is abandoned, and the next ACL statement is
4632 processed. Previously this caused the whole ACL to be aborted.
4634 13. Added the iplsearch lookup type.
4636 14. Added ident_timeout as a log selector.
4638 15. Added tls_certificate_verified as a log selector.
4640 16. Added a global option tls_require_ciphers (compare the smtp transport
4641 option of the same name). This controls incoming TLS connections.
4643 17. I finally figured out how to make tls_require_ciphers do a similar thing
4644 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4645 before starting the TLS session.
4647 18. Tabs are now shown as \t in -bP output.
4649 19. If the log selector return_path_on_delivery was set, Exim crashed when
4650 bouncing a message because it had too many Received: header lines.
4652 20. If two routers both had headers_remove settings, and the first one included
4653 a superfluous trailing colon, the final name in the first list and the
4654 first name in the second list were incorrectly joined into one item (with a
4655 colon in the middle).
4661 1. Added -C and -D options to the exinext utility, mainly to make it easier
4662 to include in the automated testing, but these could be helpful when
4663 multiple configurations are in use.
4665 2. The exinext utility was not formatting the output nicely when there was
4666 an alternate port involved in the retry record key, nor when there was a
4667 message id as well (for retries that were specific to a specific message
4668 and a specific host). It was also confused by IPv6 addresses, because of
4669 the additional colons they contain. I have fixed the IPv4 problem, and
4670 patched it up to do a reasonable job for IPv6.
4672 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4673 delivery, the log line now contains "pipelined" if PIPELINING was used.
4675 4. An SMTP transport process used to panic and die if the bind() call to set
4676 an explicit outgoing interface failed. This has been changed; it is now
4677 treated in the same way as a connect() failure.
4679 5. A reference to $sender_host_name in the part of a conditional expansion
4680 that was being skipped was still causing a DNS lookup. This no longer
4683 6. The def: expansion condition was not recognizing references to header lines
4684 that used bh_ and bheader_.
4686 7. Added the _cache feature to named lists.
4688 8. The code for checking quota_filecount in the appendfile transport was
4689 allowing one more file than it should have been.
4691 9. For compatibility with Sendmail, the command line option
4699 and sets the incoming protocol and host name (for trusted callers). The
4700 host name and its colon can be omitted when only the protocol is to be set.
4701 Note the Exim already has two private options, -pd and -ps, that refer to
4702 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4703 "s", but I don't think that's a major issue.
4705 10. A number of refactoring changes to the code, none of which should affect
4708 (a) The number of logging options was getting close to filling up the
4709 32-bit word that was used as a bit map. I have split them into two classes:
4710 those that are passed in the argument to log_write(), and those that are
4711 only ever tested independently outside of that function. These are now in
4712 separate 32-bit words, so there is plenty of room for expansion again.
4713 There is no change in the user interface or the logging behaviour.
4715 (b) When building, for example, log lines, the code previously used a
4716 macro that called string_cat() twice, in order to add two strings. This is
4717 not really sufficiently general. Furthermore, there was one instance where
4718 it was actually wrong because one of the argument was used twice, and in
4719 one call a function was used. (As it happened, calling the function twice
4720 did not affect the overall behaviour.) The macro has been replaced by a
4721 function that can join an arbitrary number of extra strings onto a growing
4724 (c) The code for expansion conditions now uses a table and a binary chop
4725 instead of a serial search (which was left over from when there were very
4726 few conditions). Also, it now recognizes conditions like "pam" even when
4727 the relevant support is not compiled in: a suitably worded error message is
4728 given if an attempt is made to use such a condition.
4730 11. Added ${time_interval:xxxxx}.
4732 12. A bug was causing one of the ddress fields not to be passed back correctly
4733 from remote delivery subprocesses. The field in question was not being
4734 subsequently used, so this caused to problems in practice.
4736 13. Added new log selectors queue_time and deliver_time.
4738 14. Might have fixed a bug in maildirsizefile handling that threw up
4739 "unexpected character" debug warnings, and recalculated the data
4740 unnecessarily. In any case, I expanded the warning message to give more
4743 15. Added the message "Restricted characters in address" to the statements in
4744 the default ACL that block characters like @ and % in local parts.
4746 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4747 Three changes have been made:
4749 (a) There was a serious bug; a negative response to MAIL caused the whole
4750 recipient domain to be cached as invalid, thereby blocking all messages
4751 to all local parts at the same domain, from all senders. This bug has
4752 been fixed. The domain is no longer cached after a negative response to
4753 MAIL if the sender used is not empty.
4755 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4758 (c) A new callout option, "use_sender" has been added for people who want
4759 the modified behaviour.
4765 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4768 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4769 indeed breaks things for older releases.
4771 3. Added additional logging to the case where there is a problem reading data
4772 from a filter that is running in a subprocess using a pipe, in order to
4773 try to track down a specific problem.
4775 4. Testing facility fudge: when running in the test harness and attempting
4776 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4777 getting "No route to host". Convert this to a timeout.
4779 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4782 6. Some OS don't have socklen_t but use size_t instead. This affects the
4783 fifth argument of getsockopt() amongst other things. This is now
4784 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4785 can be set for individual OS. I have set it for SunOS5, OSF1, and
4786 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4787 some earlier ones do not.
4789 7. Change 4.30/15 was not doing the test caselessly.
4791 8. The standard form for an IPv6 address literal was being rejected by address
4792 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4793 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4794 this, as well as the form without the "IPv6" on the front (but only when
4795 address literals are enabled, of course).
4797 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4799 10. Exim crashed if a message with an empty sender address specified by -f
4800 encountered a router with an errors_to setting. This could be provoked only
4801 by a command such as
4805 where an empty string was supplied; "<>" did not hit this bug.
4807 11. Installed PCRE release 4.5.
4809 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4810 remained set. It is now erased.
4812 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4813 times from message ids (which are base 36 rather than the normal 62).
4815 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4816 were being counted as actual protocol errors, and logged if the log
4817 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4818 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4819 and DATA following a set of rejected RCPTs do not count as protocol errors.
4820 In other words, Exim assumes they were pipelined, though this may not
4821 actually be the case. Of course, in all cases the client gets an
4822 appropriate error code.
4824 15. If a lookup fails in an ACL condition, a message about the failure may
4825 be available; it is used if testing the ACL cannot continue, because most
4826 such messages specify what the cause of the deferral is. However, some
4827 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4828 that caused an old message to be retained and used if a later statement
4829 caused a defer, replacing the real cause of the deferral.
4831 16. If an IP address had so many PTR records that the DNS lookup buffer
4832 was not large enough to hold them, Exim could crash while trying to process
4833 the truncated data. It now detects and logs this case.
4835 17. Further to 4.21/58, another change has been made: if (and only if) the
4836 first line of a message (the first header line) ends with CRLF, a bare LF
4837 in a subsequent header line has a space inserted after it, so as not to
4838 terminate the header.
4840 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4841 unnecessarily, used atoi() instead of strtol(), and didn't check the
4842 termination when getting file sizes from file names by regex.
4844 19. Completely re-implemented the support for maildirsize files, in the light
4845 of a number of problems with the previous contributed implementation
4846 (4.30/29). In particular:
4848 . If the quota is zero, the maildirsize file is maintained, but no quota is
4851 . If the maildir directory does not exist, it is created before any attempt
4852 to write a maildirsize file.
4854 . The quota value in the file is just a cache; if the quota is changed in
4855 the transport, the new value overrides.
4857 . A regular expression is available for excluding directories from the
4860 20. The autoreply transport checks the characters in options that define the
4861 message's headers; it allows continued headers, but it was checking with
4862 isspace() after an embedded newline instead of explicitly looking for a
4865 21. If all the "regular" hosts to which an address was routed had passed their
4866 expiry times, and had not reached their retry times, the address was
4867 bounced, even if fallback hosts were defined. Now Exim should go on to try
4870 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4871 equivalent code in the SMTP transport. Some hosts send humungous responses
4872 to HELO/EHLO, more than 1024 it seems.
4874 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4875 gives compiler warnings in some environments. I've now done it "properly",
4878 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4879 (because of problems with the built-in one) was declared to return uschar *
4880 instead of char *, causing compiler failure.
4882 25. Fixed a file descriptor leak when processing alias/forward files.
4884 26. Fixed a minor format string issue in dbfn.c.
4886 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4888 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4889 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4892 29. When a sender address is verified, it is cached, to save repeating the test
4893 when there is more than one recipient in a message. However, when the
4894 verification involves a callout, it is possible for different callout
4895 options to be set for different recipients. It is too complicated to keep
4896 track of this in the cache, so now Exim always runs a verification when a
4897 callout is required, relying on the callout cache for the optimization.
4898 The overhead is duplication of the address routing, but this should not be
4901 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4902 to be verified with callout=postmaster, and the main callout worked but the
4903 postmaster check failed, the verification correctly failed. However, if a
4904 subsequent RCPT command asked for sender verification *without* the
4905 postmaster check, incorrect caching caused this verification also to fail,
4908 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4909 it was not caching the DNS options (qualify_single, search_parents) that
4910 were used when the lookup failed. A subsequent lookup with different
4911 options therefore always gave the same answer, though there were cases
4912 where it should not have. (Example: a "domains = !$mx_any" option on a
4913 dnslookup router: the "domains" option is always processed without any
4914 widening, but the router might have qualify_single set.) Now Exim uses the
4915 cached value only when the same options are set.
4917 32. Added John Jetmore's "exipick" utility to the distribution.
4919 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4920 than a timeout (e.g. a certificate is required, and is not provided), an
4921 Exim server now closes the connection immediately. Previously it waited for
4922 the client to close - but if the client is SSL, it seems that they each
4923 wait for each other, leading to a delay before one of them times out.
4925 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4926 maintained 0.8.x compatibility because I don't think many are using it, and
4927 it is clearly obsolete.
4929 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4932 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4933 client certificate was expired. A simple patch fixes this, though I don't
4934 understand the full logic of why the verify callback is called multiple
4937 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4938 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4939 which causes problems with some clients (such as the Certicom SSL Plus
4940 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4941 disables the coutermeasure allowing Eudora to connect."
4943 38. Exim was not checking that a write() to a log file succeeded. This could
4944 lead to Bad Things if a log got too big, in particular if it hit a file
4945 size limit. Exim now panics and dies if it cannot write to a log file, just
4946 as it does if it cannot open a log file.
4948 39. Modified OS/Makefile-Linux so that it now contains
4950 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4952 The two -D definitions ensure that Exim is compiled with large file
4953 support, which makes it possible to handle log files that are bigger than
4956 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4957 instance) a domain was checked against a named list that involved a lookup,
4958 causing $domain_data to be set, then another domain was checked against the
4959 same list, then the first domain was re-checked, the value of $domain_data
4960 after the final check could be wrong. In particular, if the second check
4961 failed, it could be set empty. This bug probably also applied to
4964 41. The strip_trailing_dot option was not being applied to the address given
4965 with the -f command-line option.
4967 42. The code for reading a message's header from the spool was incrementing
4968 $received_count, but never initializing it. This meant that the value was
4969 incorrect (doubled) while delivering a message in the same process in which
4970 it was received. In the most common configuration of Exim, this never
4971 happens - a fresh exec is done - but it can happen when
4972 deliver_drop_privilege is set.
4974 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4975 it now includes up to 150 characters of the unexpected data in the log
4978 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4979 and building data strings. The size of both of these buffers was 10 000
4980 bytes - far larger than anybody would *ever* want, thought I. Needless to
4981 say, somebody hit the limit. I have increased the maximum line length to
4982 20 000 and the maximum data length of concatenated lines to 100 000. I have
4983 also fixed two bugs, because there was no checking on these buffers. Tsk,
4984 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4985 buffer is too small.
4987 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4988 lsearch lookups. Now it does.
4990 46. When parsing a route_list item in a manualroute router, a fixed-length
4991 buffer was used for the list of hosts. I made this 1024 bytes long,
4992 thinking that nobody would ever have a list of hosts that long. Wrong.
4993 Somebody had a whole pile of complicated expansion conditions, and the
4994 string was silently truncated, leading to an expansion error. It turns out
4995 that it is easier to change to an unlimited length (owing to other changes
4996 that have happened since this code was originally written) than to build
4997 structure for giving a limitation error. The length of the item that
4998 expands into the list of hosts is now unlimited.
5000 47. The lsearch lookup could not handle data where the length of text line was
5001 more than 4095 characters. Such lines were truncated, leading to shortened
5002 data being returned. It should now handle lines of any length.
5004 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5005 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5008 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5009 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5010 no longer happens in comment lines. A list of replacements is now placed
5011 at the head of all of the source files, except those whose only change is
5012 to replace PERL_COMMAND in the very first #! line.
5014 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5015 messages on the queue, with a bottom-up merge sort, using code contributed
5016 by Michael Haardt. This should make operations like -bp somewhat faster on
5017 large queues. It won't affect queue runners, except when queue_run_in_order
5020 51. Installed eximstats 1.31 in the distribution.
5022 52. Added support for SRV lookups to the dnslookup router.
5024 53. If an ACL referred to $message_body or $message_body_end, the value was not
5025 reset for any messages that followed in the same SMTP session.
5027 54. The store-handling optimization for building very long strings was not
5028 differentiating between the different store pools. I don't think this
5029 actually made any difference in practice, but I've tidied it.
5031 55. While running the routers to verify a sender address, $sender_address
5032 was still set to the sender address. This is wrong, because when routing to
5033 send a bounce to the sender, it would be empty. Therefore, I have changed
5034 it so that, while verifying a sender address, $sender_address is set to <>.
5035 (There is no change to what happens when verifying a recipient address.)
5037 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5038 target A or AAAA records (if not already returned) without resetting the
5039 qualify_single or search_parents options of the DNS resolver. These are
5040 inappropriate in this case because the targets of MX and SRV records must
5041 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5042 target that, when qualified, matched something in the local domain. These
5043 two options are now turned off when doing these lookups.
5045 57. It seems that at least some releases of Reiserfs (which does not have the
5046 concept of a fixed number of inodes) returns zero and not -1 for the
5047 number of available inodes. This interacted badly with check_spool_inodes,
5048 which assumed that -1 was the "no such thing" setting. What I have done is
5049 to check that the total number of inodes is greater than zero before doing
5050 the test of how many are available.
5052 58. When a "warn" ACL statement has a log_message modifier, the message is
5053 remembered, and not repeated. This is to avoid a lot of repetition when a
5054 message has many recipients that cause the same warning to be written.
5055 Howewer, Exim was preserving the list of already written lines for an
5056 entire SMTP session, which doesn't seem right. The memory is now reset if a
5057 new message is started.
5059 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5060 debugging output unless log_rewrite was also set.
5062 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5063 of (int)(handle) when we know that handle contains (void *)(-1).
5065 61. The Exim daemon panic-logs an error return when it closes the incoming
5066 connection. However "connection reset by peer" seems to be common, and
5067 isn't really an error worthy of noting specially, so that particular error
5070 62. When Exim is trying to find all the local interfaces, it used to panic and
5071 die if the ioctl to get the interface flags failed. However, it seems that
5072 on at least one OS (Solaris 9) it is possible to have an interface that is
5073 included in the list of interfaces, but for which you get a failure error
5074 for this call. This happens when the interface is not "plumbed" into a
5075 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5076 failure of the "get flags" call assumes that the interface is down.
5078 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5079 makes life easier for people who are doing arithmetic on fields extracted
5080 from dates, where you often get leading zeros that should not be
5081 interpreted as octal.
5083 64. Added qualify_domain to the redirect router, to override the global
5086 65. If a pathologically long header line contained very many addresses (the
5087 report of this problem mentioned 10 000) and each of them was rewritten,
5088 Exim could use up a very large amount of memory. (It kept on making new
5089 copies of the header line as it rewrote, and never released the old ones.)
5090 At the expense of a bit more processing, the header rewriting function has
5091 been changed so that it no longer eats memory in this way.
5093 66. The generation of the Received: header has been moved from the time that a
5094 message starts to be received, to the time that it finishes. The timestamp
5095 in the Received: header should now be very close to that of the <= log
5096 line. There are two side-effects of this change:
5098 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5099 logged header lines no longer include the local Received: line, because
5100 it has not yet been created. The same applies to a copy of the message
5101 that is returned to a non-SMTP sender when a message is rejected.
5103 (b) When a filter file is tested using -bf, no additional Received: header
5104 is added to the test message. After some thought, I decided that this
5107 This change does not affect the value of $received_for. It is still set
5108 after address rewriting, but before local_scan() is called.
5110 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5112 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5113 gave an unhelpful panic error message, and a defer error. I have managed to
5114 change this behaviour so that it now rejects any supplied certificate,
5115 which seems right, as the list of acceptable certificates is empty.
5117 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5118 gave an unhelpful defer error. I have not managed to make this reject any
5119 supplied certificates, but the error message it gives is "no certificate
5120 supplied", which is not helpful.
5122 70. exigrep's output now also includes lines that are not associated with any
5123 message, but which match the given pattern. Implemented by a patch from
5124 Martin Sluka, which also tidied up the Perl a bit.
5126 71. Recipient callout verification, like sender verification, was using <> in
5127 the MAIL FROM command. This isn't really the right thing, since the actual
5128 sender may affect whether the remote host accepts the recipient or not. I
5129 have changed it to use the actual sender in the callout; this means that
5130 the cache record is now keyed on a recipient/sender pair, not just the
5131 recipient address. There doesn't seem to be a real danger of callout loops,
5132 since a callout by the remote host to check the sender would use <>.
5133 [SEE ABOVE: changed after hitting problems.]
5135 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5136 temporary errors. However, in the case of such a code being given after
5137 the end of a data transmission (i.e. after ".") Exim was failing to write
5138 a retry record for the message. (Yes, there was some broken host that was
5139 actually sending 8xx at this point.)
5141 73. An unknown lookup type in a host list could cause Exim to panic-die when
5142 the list was checked. (An example that provoked this was putting <; in the
5143 middle of a list instead of at the start.) If this happened during a DATA
5144 ACL check, a -D file could be left lying around. This kind of configuration
5145 error no longer causes Exim to die; instead it causes a defer errror. The
5146 incident is still logged to the main and panic logs.
5148 74. Buglet left over from Exim 3 conversion. The message "too many messages
5149 in one connection" was written to the rejectlog but not the mainlog, except
5150 when address rewriting (yes!) was being logged.
5152 75. Added write_rejectlog option.
5154 76. When a system filter was run not as root (that is, when system_filter_user
5155 was set), the values of the $n variables were not being returned to the
5156 main process; thus, they were not subsequently available in the $sn
5159 77. Added +return_path_on_delivery log selector.
5161 78. A connection timeout was being treated differently from recipients deferred
5162 when testing hosts_max_try with a message that was older than the host's
5163 retry timeout. (The host should not be counted, thus allowing all hosts to
5164 be tried at least once before bouncing.) This may have been the cause of an
5165 occasionally reported bug whereby a message would remain on the queue
5166 longer than the retry timeout, but would be bounced if a delivery was
5167 forced. I say "may" because I never totally pinned down the problem;
5168 setting up timeout/retry tests is difficult. See also the next item.
5170 79. The ultimate address timeout was not being applied to errors that involved
5171 a combination of host plus message (for example, a timeout on a MAIL
5172 command). When an address resolved to a number of possible hosts, and they
5173 were not all tried for each delivery (e.g. because of hosts_max_try), a
5174 message could remain on the queue longer than the retry timeout.
5176 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5179 81. Fixed an obscure SMTP outgoing bug which required at least the following
5180 conditions: (a) there was another message waiting for the same server;
5181 (b) the server returned 5xx to all RCPT commands in the first message so
5182 that the message was not completed; (c) the server dropped the connection
5183 or gave a negative response to the RSET that Exim sends to abort the
5184 transaction. The observed case was a dropped connection after DATA that had
5185 been sent in pipelining mode. That is, the server had advertised PIPELINING
5186 but was not implementing it correctly. The effect of the bug was incorrect
5187 behaviour, such as trying another host, and this could lead to a crash.
5193 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5194 and daemon.c were passed as pointers to ints; they should have been
5195 pointers to socklen_t variables (which are typically unsigned ints).
5197 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5200 3. Fixed a really odd bug that affected only the testing scheme; patching a
5201 certain fixed string in the binary changed the value of another string that
5202 happened to be identical to the end of the original first string.
5204 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5205 name", it returns that address as the IP address. On some operating
5206 systems (e.g. Solaris), it also passes back the IP address string as the
5207 "host name". However, on others (e.g. Linux), it passes back an empty
5208 string. Exim wasn't checking for this, and was changing the host name to an
5209 empty string, assuming it had been canonicized.
5211 5. Although rare, it is permitted to have more than one PTR record for a given
5212 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5213 all the names associated with an address, because they do in Solaris.
5214 However, it seems that they do not in Linux for data that comes from the
5215 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5216 I found this out when I moved to a new Linux workstation and tried to run
5217 the Exim test suite.
5219 To get round this problem I have changed the code so that it now does its
5220 own call to the DNS to look up PTR records when searching for a host name.
5221 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5222 addresses that are only in /etc/hosts are still found.
5224 This behaviour is, however, controlled by an option called host_lookup_
5225 order, which defaults to "bydns:byaddr". If people want to use the other
5226 order, or indeed, just use one or the other means of lookup, they can
5227 specify it in this variable.
5229 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5230 some operating systems, this comes back from gethostbyaddr() as an empty
5231 string, and this is what Exim used to test for. However, it seems that in
5232 other systems, "." is yielded. Exim now tests for this case too.
5234 7. The values of check_spool_space and check_log_space are now held internally
5235 as a number of kilobytes instead of an absolute number of bytes. If a
5236 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5237 kilobyte. This means that much larger values can be stored.
5239 8. Exim monitor: an attempt to get the action menu when not actually pointing
5240 at a message produces an empty menu entitled "No message selected". This
5241 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5242 no entries in it ("Shell widget menu has zero width and/or height"). So I
5243 have added a single, blank menu entry in this case.
5245 9. Added ${quote_local_part.
5247 10. MIME decoding is now applied to the contents of Subject: header lines when
5250 11. Now that a reference to $sender_host_address automatically causes a reverse
5251 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5252 host lookup before query-style lookups in lists that might use this
5253 variable. This has therefore been abolished, and the "net-" prefix is no
5254 longer necessary for query-style lookups.
5256 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5257 have been a typo for LFLAGS, so it has been changed.
5259 13. The install script calls Exim with "-C /dev/null" in order to find the
5260 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5261 to be output. Howeve, since Exim outputs its version number before the
5262 error, it didn't break the script. It just looked ugly. I fixed this by
5263 always allowing "-C /dev/null" if the caller is root.
5265 14. Ignore overlarge ACL variable number when reading spool file - insurance
5266 against a later release with more variables having written the file.
5268 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5269 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5270 this, as well as the form without the "IPv6" on the front.
5272 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5273 OS/Makefile-Darwin file.
5275 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5276 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5278 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5279 variable. However, some OS use ints for inodes. I've added cast to long int
5280 to get rid of the compiler warning.
5282 19. I had forgotten to lock out "/../" in configuration file names when
5283 ALT_CONFIG_PREFIX was set.
5285 20. Routers used for verification do not need to specify transports. However,
5286 if such a router generated a host list, and callout was configured, Exim
5287 crashed, because it could not find a port number from the (non-existent)
5288 transport. It now assumes port 25 in this circumstance.
5290 21. Added the -t option to exigrep.
5292 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5293 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5294 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5295 them is set, LOOKUP_LSEARCH is forced.
5297 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5299 24. Added sender and host information to the "rejected by local_scan()" log
5300 line; previously there was no indication of these.
5302 25. Added .include_if_exists.
5304 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5305 Linux. It turns out that not all file systems support this. Apparently some
5306 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5307 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5308 not supported on the file descriptor, is now ignored when Exim is trying to
5309 sync a directory. This applies only to Linux.
5311 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5313 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5314 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5317 29. Added support for maildirsize files from supplied patch (modified a bit).
5319 30. The use of :fail: followed by an empty string could lead Exim to respond to
5320 sender verification failures with (e.g.):
5322 550 Verification failed for <xxx>
5323 550 Sender verify failed
5325 where the first response line was missing the '-' that indicates it is not
5326 the final line of the response.
5328 31. The loop for finding the name of the user that called Exim had a hardwired
5329 limit of 10; it now uses the value of finduser_retries, which is used for
5330 all other user lookups.
5332 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5335 33. Exim was neglecting to zero errno before one call of strtol() when
5336 expanding a string and expecting an integer value. On some systems this
5337 resulted in spurious "integer overflow" errors. Also, it was casting the
5338 result into an int without checking.
5340 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5341 did not work. The code looks as if it has *never* worked, though it appears
5342 to have been documented since at least releast 1.62. I have made it work.
5344 35. The "timeout_DNS" error in retry rules, also documented since at least
5345 1.62, also never worked. As it isn't clear exactly what this means, and
5346 clearly it isn't a major issue, I have abolished the feature by treating it
5347 as "timeout", and writing a warning to the main and panic logs.
5349 36. The display of retry rules for -brt wasn't always showing the error code
5352 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5353 timeout_connect_A, timeout_connect_MX.
5355 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5356 to the empty sender.
5358 39. The daemon was not analysing the content of -oX till after it had closed
5359 stderr and disconnected from the controlling terminal. This meant that any
5360 syntax errors were only noted on the panic log, and the return code from
5361 the command was 0. By re-arranging the code a little, I've made the
5362 decoding happen first, so such errors now appear on stderr, and the return
5363 code is 1. However, the actual setting up of the sockets still happens in
5364 the disconnected process, so errors there are still only recorded on the
5367 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5368 connections (as happens on some IP stacks) was logged at start up time as
5369 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5370 it from "IPv6 and IPv4", which means that two separate sockets are being
5373 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5374 says whether AF_INET or AF_INET6 was passed as an argument.
5376 42. Exiwhat output was messed up when time zones were included in log
5379 43. Exiwhat now gives more information about the daemon's listening ports,
5380 and whether -tls-on-connect was used.
5382 44. The "port" option of the smtp transport is now expanded.
5384 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5385 silently ignored. Now an error message is written to the main and panic
5388 46. There's a new ACL modifier called "logwrite" which writes to a log file
5389 as soon as it is encountered.
5391 47. Added $local_user_uid and $local_user_gid at routing time.
5393 48. Exim crashed when trying to verify a sender address that was being
5396 49. Exim was recognizing only a space character after ".include". It now also
5397 recognizes a tab character.
5399 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5400 extracting the relevant information from the specification. The man page no
5401 longer contains scrambled data for the -d option, and I've added a section
5402 at the front about calling Exim under different names.
5404 51. Added "extra_headers" argument to the "mail" command in filter files.
5406 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5409 53. Installed eximstats 1.29.
5411 54. Added transport_filter_timeout as a generic transport option.
5413 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5414 Cc: header lines. This was required by RFC 822, but it not required by RFC
5417 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5418 incoming messages that did not have them. Now it does so only if the
5419 message originates locally, that is, if there is no associated remote host
5420 address. When Resent- header lines are present, this applies to the Resent-
5421 lines rather than the non-Resent- lines.
5423 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5424 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5426 58. Messages for configuration errors now include the name of the main
5427 configuration file - useful now that there may be more than one file in a
5428 list (.included file names were always shown).
5430 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5431 for those rare installations that do not start the daemon as root or run it
5432 setuid root. I've cut out the call to initgroups() if the daemon is not
5435 60. The Exim user and group can now be bound into the binary as text strings
5436 that are looked up at the start of Exim's processing.
5438 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5440 62. Added $mailstore_basename variable.
5442 63. Installed patch to sieve.c from Michael Haardt.
5444 64. When Exim failed to open the panic log after failing to open the main log,
5445 the original message it was trying to log was written to stderr and debug
5446 output, but if they were not available (the usual case in production), it
5447 was lost. Now it is written to syslog before the two lines that record the
5448 failures to open the logs.
5450 65. Users' Exim filters run in subprocesses under the user's uid. It is
5451 possible for a "deliver" command or an alias in a "personal" command to
5452 provoke an address rewrite. If logging of address rewriting is configured,
5453 this fails because the process is not running as root or exim. There may be
5454 a better way of dealing with this, but for the moment (because 4.30 needs
5455 to be released), I have disabled address rewrite logging when running a
5456 filter in a non-root, non-exim process.
5462 1. The buildconfig auxiliary program wasn't quoting the value set for
5463 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5464 not defined. This bug was masked in 4.22 by the effect that was fixed in
5467 2. Some messages that were rejected after a message id was allocated were
5468 shown as "incomplete" by exigrep. It no longer does this for messages that
5469 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5471 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5472 have allow_domain_literals set, the ID did not get logged in the <= line.
5473 Domain literals are now always recognized in Message-ID: header lines.
5475 4. The first argument for a ${extract expansion item is the key name or field
5476 number. Leading and trailing spaces in this item were not being ignored,
5477 causing some misleading effects.
5479 5. When deliver_drop_privilege was set, single queue runner processes started
5480 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5481 same command in the process it spins off) were not dropping privilege.
5483 6. When the daemon running as "exim" started a queue runner, it always
5484 re-executed Exim in the spun-off process. This is a waste of effort when
5485 deliver_drop_privilege is set. The new process now just calls the
5486 queue-runner function directly.
5492 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5495 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5496 ignored. Though the use of -oP was forcing the writing of a pid file, it
5497 was always written to the default place.
5499 3. If the message "no IP address found for host xxxx" is generated during
5500 incoming verification, it is now followed by identification of the incoming
5501 connection (so you can more easily find what provoked it).
5503 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5505 5. Added some features to "harden" Exim a bit more against certain attacks:
5507 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5508 be put in Local/Makefile. This is like the never_users runtime option,
5509 but it cannot be overridden. The default setting is "root".
5511 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5512 prefix string with which any file named in a -C command line option
5515 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5516 is retained for -C and -D only if the caller of Exim is root. Without
5517 it, the exim user may also use -C and -D and retain privilege.
5519 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5520 command line option is disabled.
5522 6. Macro names set by the -D option must start with an upper case letter, just
5523 like macro names defined in the configuration file.
5525 7. Added "dereference=" facility to LDAP.
5527 8. Two instances of the typo "uknown" in the source files are fixed.
5529 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5530 the Configure-Makefile script screwed up while processing it.
5532 10. Incorporated PCRE 4.4.
5534 11. The SMTP synchronization check was not operating right at the start of an
5535 SMTP session. For example, it could not catch a HELO sent before the client
5536 waited for the greeting. There is now a check for outstanding input at the
5537 point when the greeting is written. Because of the duplex, asynchronous
5538 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5539 way, but not yet received, when the check is performed.
5541 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5542 on TCP/IP sockets, because this apparently causes some broken clients to
5545 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5546 unchanged) from the Cygwin maintainer.
5548 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5549 was supported instead of testing for mbx. Effectively a typo.
5551 15. The spa authenticator server code was not checking that the input it
5552 received was valid base64.
5554 16. The debug output line for the "set" modifier in ACLs was not showing the
5555 name of the variable that was being set.
5557 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5559 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5560 lookup on its own. Something else was needed to trigger the lookup. For
5561 example, a match in host_lookup or the need for a host name in a host list.
5562 Now, if $sender_host_name is referenced and the host name has not yet been
5563 looked up, a lookup is performed. If the lookup fails, the variable remains
5564 empty, and $host_lookup_failed is set to "1".
5566 19. Added "eqi" as a case-independent comparison operator.
5568 20. The saslauthd authentication condition could segfault if neither service
5569 nor realm was specified.
5571 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5572 error message that was logged was misleading, and incoming SMTP
5573 connections were dropped. The message is now more accurate, and temporary
5574 errors are given to SMTP connections.
5576 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5577 (except RSET) with a 421 error, until QUIT is received. However, it was
5578 failing to send a response to QUIT.
5580 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5581 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5582 ACL is now run after the helo_try_verify_hosts code.
5584 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5585 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5586 case-independent, so other case variants are also recognized). Apparently
5587 some systems use these upper case variants.
5589 25. If more than two messages were waiting for the same host, and a transport
5590 filter was specified for the transport, Exim sent two messages over the
5591 same TCP/IP connection, and then failed with "socket operation on non-
5592 socket" when it tried to send the third.
5594 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5596 27. The extern definition of crypt16() in expand.c was not being excluded when
5597 the OS had its own crypt16() function.
5599 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5600 as a preferred synonym for return_size_limit, both as an option and as an
5603 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5605 30. Changed the default configuration ACL to relax the local part checking rule
5606 for addresses that are not in any local domains. For these addresses,
5607 slashes and pipe symbols are allowed within local parts, but the sequence
5608 /../ is explicitly forbidden.
5610 31. SPA server authentication was not clearing the challenge buffer before
5613 32. log_message in a "warn" ACL statement was writing to the reject log as
5614 well as to the main log, which contradicts the documentation and doesn't
5615 seem right (because no rejection is happening). So I have stopped it.
5617 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5618 However, I am unable to do any testing of this.
5620 34. Fixed an infelicity in the appendfile transport. When checking directories
5621 for a mailbox, to see if any needed to be created, it was accidentally
5622 using path names with one or more superfluous leading slashes; tracing
5623 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5625 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5626 contents are added to the log line that is written for every discarded
5627 recipient. (Previously a log_message setting was ignored.)
5629 36. The ${quote: operator now quotes the string if it is empty.
5631 37. The install script runs exim in order to find its version number. If for
5632 some reason other than non-existence or emptiness, which it checks, it
5633 could not run './exim', it was installing it with an empty version number,
5634 i.e. as "exim-". This error state is now caught, and the installation is
5637 38. An argument was missing from the function that creates an error message
5638 when Exim fails to connect to the socket for saslauthd authentication.
5639 This could cause Exim to crash, or give a corrupted message.
5641 39. Added isip, isip4, and isip6 to ${if conditions.
5643 40. The ACL variables $acl_xx are now saved with the message, and can be
5644 accessed later in routers, transports, and filters.
5646 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5647 strings in the file are not string-expanded.
5649 42. If a MAIL command specified a SIZE value that was too large to fit into an
5650 int variable, the check against message_size_limit failed. Such values are
5651 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5652 day this will have to be increased, but I don't think I want to be around
5653 when emails are that large.
5660 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5661 iconv() is not standard in FreeBSD.
5663 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5664 IPv6 enabled. The observed symptom was a segmentation fault on return from
5665 the function os_common_find_running_interfaces() in src/os.c.
5667 3. In the check_special_case() function in daemon.c I had used "errno" as an
5668 argument name, which causes warnings on some systems. This was basically a
5669 typo, since it was named "eno" in the comments!
5671 4. The code that waits for the clock to tick (at a resolution of some fraction
5672 of a second) so as to ensure message-id uniqueness was always waiting for
5673 at least one whole tick, when it could have waited for less. [This is
5674 almost certainly not relevant at current processor speeds, where it is
5675 unlikely to ever wait at all. But we try to future-proof.]
5677 5. The function that sleeps for a time interval that includes fractions of a
5678 second contained a race. It did not block SIGALRM between setting the
5679 timer, and suspending (a couple of lines later). If the interval was short
5680 and the sigsuspend() was delayed until after it had expired, the suspension
5681 never ended. On busy systems this could lead to processes getting stuck for
5684 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5685 process, before it forks any delivery processes. The open lookup caching
5686 mechanism meant that the open file or database connection was passed into
5687 the delivery process. The problem was that delivery processes always tidy
5688 up cached lookup data. This could cause a problem for the next delivery
5689 process started by the queue runner, because the external queue runner
5690 process does not know about the closure. So the next delivery process
5691 still has data in the lookup cache. In the case of a file lookup, there was
5692 no problem because closing a file descriptor in a subprocess doesn't affect
5693 the parent. However, if the lookup was caching a connection to a database,
5694 the connection was closed, and the second delivery process was likely to
5695 see errors such as "PGSQL: query failed: server closed the connection
5696 unexpectedly". The problem has been fixed by closing all cached lookups
5697 in a queue runner before running a delivery process.
5699 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5700 seem to have the "const" qualifier which it has on other OS. I've
5703 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5704 *of the same type* (client or server) with the same public name that an
5705 error should be diagnosed.
5707 9. When Exim looked up a host name for an IP address, but failed to find the
5708 original IP address when looking up the host name (a safety check), it
5709 output the message "<ip address> does not match any IP for NULL", which was
5710 confusing, to say the least. The bug was that the host name should have
5711 appeared instead of "NULL".
5713 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5714 user that is built into the binary, and the -C or -D options is used, root
5715 privilege is dropped before the configuration file is read. In addition,
5716 logging is switched to stderr instead of the normal log files. If the
5717 configuration then re-defines the Exim user, the unprivileged environment
5718 is probably not what is expected, so Exim logs a panic warning message (but
5721 However, if deliver_drop_privilege is set, the unprivileged state may well
5722 be exactly what is intended, so the warning has been cut out in that case,
5723 and Exim is allowed to try to write to its normal log files.
5729 1. smtp_return_error_details was not giving details for temporary sender
5730 or receiver verification errors.
5732 2. Diagnose a configuration error if two authenticators have the same public
5735 3. Exim used not to create the message log file for a message until the first
5736 delivery attempt. This could be confusing when incoming messages were held
5737 for policy or load reasons. The message log file is now created at the time
5738 the message is received, and an initial "Received" line is written to it.
5740 4. The automatically generated man page for command line options had a minor
5741 bug that caused no ill effects; however, a more serious problem was that
5742 the procedure for building the man page automatically didn't always
5743 operate. Consequently, release 4.20 contains an out-of-date version. This
5744 shouldn't happen again.
5746 5. When building Exim with embedded Perl support, the script that builds the
5747 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5748 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5750 6. The freeze_tell option was not being used for messages that were frozen on
5751 arrival, either by an ACL or by local_scan().
5753 7. Added the smtp_incomplete_transaction log selector.
5755 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5756 was accepting AUTH without a new EHLO.
5758 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5759 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5762 10. Logging of TCP/IP connections (when configured) now happens in the main
5763 daemon process instead of the child process, so that the TCP/IP connection
5764 count is more accurate (but it can never be perfect).
5766 11. The use of "drop" in a nested ACL was not being handled correctly in the
5767 outer ACL. Now, if condition failure induced by the nested "drop" causes
5768 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5769 or "require"), the connection is dropped.
5771 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5772 that yield "discard" can now be used with an "accept" or a "discard" verb,
5773 but an error is generated for any others (because I can't see a useful way
5774 to define what should happen).
5776 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5777 are now processed in the same way as lines in the Exim configuration file.
5778 In particular, continuation lines are supported.
5780 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5782 15. Added -ti meaning -t -i.
5784 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5785 domains, and warn by logging if others are found.
5787 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5788 returned from GIFCONF when Exim is trying to find the list of interfaces on
5789 a host. The code in os.c has been modified to copy each ifreq to an aligned
5790 structure in all cases.
5792 Also, in some cases, the returned ifreq's were being copied to a 'struct
5793 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5794 means the last couple of bytes of an IPv6 address could be chopped if the
5795 ifreq contained only a normal sockaddr (14 bytes storage).
5797 18. Named domain lists were not supported in the hosts_treat_as_local option.
5798 An entry such as +xxxx was not recognized, and was treated as a literal
5801 19. Ensure that header lines added by a DATA ACL are included in the reject log
5802 if the ACL subsequently rejects the message.
5804 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5805 MD5 (which is deprecated).
5807 21. When testing a filter file using -bf, Exim was writing a message when it
5808 took the sender from a "From " line in the message, but it was not doing so
5809 when it took $return_path from a Return-Path: header line. It now does.
5811 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5812 with a valid header line field name (a series of printing characters
5813 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5815 23. Changed "disc" in the source to "disk" to conform to the documentation and
5816 the book and for uniformity.
5818 24. Ignore Sendmail's -Ooption=value command line item.
5820 25. When execve() failed while trying to run a command in a pipe transport,
5821 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5822 could be confused with a return value of 69 from the command itself. This
5823 has been changed to 127, the value the shell returns if it is asked to run
5824 a non-existent command. The wording for the related log line suggests a
5825 non-existent command as the problem.
5827 26. If received_header_text expands to an empty string, do not add a Received:
5828 header line to the message. (Well, it adds a token one on the spool, but
5829 marks it "old" so that it doesn't get used or transmitted.)
5831 27. Installed eximstats 1.28 (addition of -nt option).
5833 28. There was no check for failure on the call to getsockname() in the daemon
5834 code. This can fail if there is a shortage of resources on the system, with
5835 ENOMEM, for example. A temporary error is now given on failure.
5837 29. Contrary to the C standard, it seems that in some environments, the
5838 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5839 program. Exim now does this explicitly; it affects the formatting of
5840 timestamps using strftime().
5842 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5843 complaints. I've now initialized all the variables, to avoid this.
5845 32. Header lines added by a system filter were not being "seen" during
5846 transport-time rewrites.
5848 33. The info_callback() function passed to OpenSSL is set up with type void
5849 (*)(SSL *, int, int), as described somewhere. However, when calling the
5850 function (actually a macro) that sets it up, the type void(*)() is
5851 expected. I've put in a cast to prevent warnings from picky compilers.
5853 34. If a DNS black list lookup found a CNAME record, but there were no A
5854 records associated with the domain it pointed at, Exim crashed.
5856 35. If a DNS black list lookup returned more than one A record, Exim ignored
5857 all but the first. It now scans all returned addresses if a particular IP
5858 value is being sought. In this situation, the contents of the
5859 $dnslist_value variable are a list of all the addresses, separated by a
5862 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5863 to accept a host name and all its aliases if the forward lookup for any of
5864 them yielded the IP address of the incoming connection. Now it accepts only
5865 those names whose forward lookup yields the correct IP address. Any other
5866 names are discarded. This closes a loophole whereby a rogue DNS
5867 administrator could create reverse DNS records to break through a
5868 wildcarded host restriction in an ACL.
5870 37. If a user filter or a system filter that ran in a subprocess used any of
5871 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5872 the wrong values were passed to the pipe command ($thisaddress had the
5873 value of $0, $0 had the value of $1, etc). This bug was introduced by
5874 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5876 38. Improved the line breaking for long SMTP error messages from ACLs.
5877 Previously, if there was no break point between 40 and 75 characters, Exim
5878 left the rest of the message alone. Two changes have been made: (a) I've
5879 reduced the minimum length to 35 characters; (b) if it can't find a break
5880 point between 35 and 75 characters, it looks ahead and uses the first one
5881 that it finds. This may give the occasional overlong line, but at least the
5882 remaining text gets split now.
5884 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5885 file descriptors might be low, and that setting 1000 would always raise it.
5886 It turns out that in some environments, the limit is already over 1000 and
5887 that lowering it causes trouble. So now Exim takes care not to decrease it.
5889 40. When delivering a message, the value of $return_path is set to $sender_
5890 address at the start of routing (routers may change the value). By an
5891 oversight, this default was not being set up when an address was tested by
5892 -bt or -bv, which affected the outcome if any router or filter referred to
5895 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5896 log only when "message" or "log_message" are set. However, if one of the
5897 conditions was an address verification, or a call to a nested ACL, the
5898 messages generated by the underlying test were being passed through. This
5899 no longer happens. The underlying message is available in $acl_verify_
5900 message for both "message" and "log_message" expansions, so it can be
5901 passed through if needed.
5903 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5904 new expansion $bh_ to give the encoded byte string without charset
5905 translation. Translation happens only if iconv() is available; HAVE_ICONV
5906 indicates this at build time. HEADERS_CHARSET gives the charset to
5907 translate to; headers_charset can change it in the configuration, and
5908 "headers charset" can change it in an individual filter file.
5910 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5911 that creates RFC 2047 encoded "words" labels them as that charset instead
5912 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5913 expansion operator; (ii) when Exim creates a From: line for a local
5914 message; (iii) when a header line is rewritten to include a "phrase" part.
5916 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5917 buggy, causing it to skip the first lines of messages whose message ID
5918 ended in 'D'. This would not have bitten before Exim release 4.14, because
5919 message IDs were unlikely to end in 'D' before then. The effect was to have
5920 incorrect size information for certain domains.
5922 45. #include "config.h" was missing at the start of the crypt16.c module. This
5923 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5926 46. If there was a timeout during a "random" callout check, Exim treated it as
5927 a failure of the random address, and carried on sending RSET and the real
5928 address. If the delay was just some slowness somewhere, the response to the
5929 original RCPT would be taken as a response to RSET and so on, causing
5930 mayhem of various kinds.
5932 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5933 when I implemented it. It didn't allow for the fact that some option values
5934 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5935 the right test for positive values.
5937 48. Domain names in DNS records are case-independent. Exim always looks them up
5938 in lower case. Some resolvers return domain names in exactly the case they
5939 appear in the zone file, that is, they may contain uppercase letters. Not
5940 all resolvers do this - some return always lower case. Exim was treating a
5941 change of case by a resolver as a change of domain, similar to a widening
5942 of a domain abbreviation. This triggered its re-routing code and so it was
5943 trying to route what was effectively the same domain again. This normally
5944 caused routing to fail (because the router wouldn't handle the domain
5945 twice). Now Exim checks for this case specially, and just changes the
5946 casing of the domain that it ultimately uses when it transmits the message
5949 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5952 50. If a filter generated a file delivery with a non-absolute name (possible if
5953 no home directory exists for the router), the forbid_file option was not
5956 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5957 the existing equality matching.
5959 52. Exim was using ints instead of ino_t variables in some places where it was
5960 dealing with inode numbers.
5962 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5963 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5964 called TMPDIR, and if it finds it is different, it changes its value.
5966 54. The smtp_printf() function is now made available to local_scan() so
5967 additional output lines can be written before returning. There is also an
5968 smtp_fflush() function to enable the detection of a dropped connection.
5969 The variables smtp_input and smtp_batched_input are exported to
5972 55. Changed the default runtime configuration: the message "Unknown user"
5973 has been removed from the ACL, and instead placed on the localuser router,
5974 using the cannot_route_message feature. This means that any verification
5975 failures that generate their own messages won't get overridden. Similarly,
5976 the "Unrouteable address" message that was in the ACL for unverifiable
5977 relay addresses has also been removed.
5979 56. Added hosts_avoid_esmtp to the smtp transport.
5981 57. The exicyclog script was not checking for the esoteric option
5982 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5983 will work only if exicyclog is run under the appropriate euid.
5985 58. Following a discussion on the list, the rules by which Exim recognises line
5986 endings on incoming messages have been changed. The -dropcr and drop_cr
5987 options are now no-ops, retained only for backwards compatibility. The
5988 following line terminators are recognized: LF CRLF CR. However, special
5989 processing applies to CR:
5991 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5992 nor a local message in the state where . is a terminator.
5994 (ii) If a bare CR is encountered in a header line, an extra space is added
5995 after the line terminator so as not to end the header. The reasoning
5996 behind this is that bare CRs in header lines are most likely either
5997 to be mistakes, or people trying to play silly games.
5999 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6000 was being incorrectly given as 18 bytes larger than it should have been.
6001 This is a VOB (very old bug).
6003 60. This may never have affected anything current, but just in case it has:
6004 When the local host is found other than at the start of a list of hosts,
6005 the local host, those with the same MX, and any that follow, are discarded.
6006 When the list in question was part of a longer list of hosts, the following
6007 hosts (not currently being processed) were also being discarded. This no
6008 longer happens. I'm not sure if this situation could ever has previously
6011 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6014 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6015 header. This is something that is recommended in a new Internet Draft, and
6016 is something that is documented as being done by Sendmail. There are two
6017 possible values. For messages generated by the autoreply transport, Exim
6020 Auto-Submitted: auto-replied
6022 whereas for all other generated messages (e.g. bounces) it adds
6024 Auto-Submitted: auto-generated
6026 63. The "personal" condition in filters now includes a test for the
6027 Auto-Submitted: header. If it contains the string "auto-" the message it
6028 not considered personal.
6030 64. Added rcpt_include_affixes as a generic transport option.
6032 65. Added queue_only_override (default true).
6034 66. Added the syslog_duplication option.
6036 67. If what should have been the first header line of a message consisted of
6037 a space followed by a colon, Exim was mis-interpreting it as a header line.
6038 It isn't of course - it is syntactically invalid and should therefore be
6039 treated as the start of the message body. The misbehaviour could have
6040 caused a number of strange effects, including loss of data in subsequent
6041 header lines, and spool format errors.
6043 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6044 client host had authenticated. This control can now be exercised by an ACL
6045 for more flexibility.
6047 69. By default, callouts do not happen when testing with -bh. There is now a
6048 variant, -bhc, which does actually run the callout code, including
6049 consulting and updating the callout cache.
6051 70. Added support for saslauthd authentication, courtesy of Alexander
6054 71. If statvfs() failed on the spool or log directories while checking their
6055 size for availability, Exim confusingly gave the error "space shortage".
6056 Furthermore, in debugging mode it crashed with a floating point exception.
6057 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6058 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6059 serious problem, Exim now writes to the main and panic logs when this
6060 happens, with details of the failure. It then refuses to accept the
6061 incoming message, giving the message "spool directory problem" or "log
6062 directory problem" with a 421 code for SMTP messages.
6064 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6065 0, 1, and 2 exist, because some OS complain for execs without them (see
6066 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6067 descriptors. However, the code omitted to check that the open succeeded,
6068 causing mysterious errors if for some reason the permissions on /dev/null
6069 got screwed. Now Exim writes a message to the main and panic logs, and
6070 bombs out if it can't open /dev/null.
6072 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6073 interact so that it is all more flexible. It is supposed to remain
6074 backwards compatible. Also added extra_local_interfaces.
6076 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6077 to bomb out with an assertion failure - to the client this appears as a
6078 connection drop. This problem occurs in the part of the code that was taken
6079 from the Samba project. Fortunately, the assertion is in a very simple
6080 function, so I have fixed this by reproducing the function inline in the
6081 one place where it is called, and arranging for authentication to fail
6082 instead of killing the process with assert().
6084 75. The SPA client code was not working when the server requested OEM rather
6085 than Unicode encoding.
6087 76. Added code to make require_files with a specific uid setting more usable in
6088 the case where statting the file as root fails - usually a non-root-mounted
6089 NFS file system. When this happens and the failure is EACCES, Exim now
6090 forks a subprocess and does the per-uid checking as the relevant uid.
6092 77. Added process_log_path.
6094 78. If log_file_path was not explicitly set, a setting of check_log_space or
6095 check_log_inodes was ignored.
6097 79. If a space check for the spool or log partitions fails, the incident is now
6098 logged. Of course, in the latter case the data may get lost...
6100 80. Added the %p formatting code to string_format() so that it can be used to
6101 print addresses in debug_print(). Adjusted all the address printing in the
6102 debugging in store.c to use %p rather than %d.
6104 81. There was a concern that a line of code in smtp_in.c could overflow a
6105 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6106 initially expressed, the concern was not well-founded, because trailing
6107 spaces are removed early. However, if the trailing spaces were followed by
6108 a NULL, they did not get removed, so the overflow was possible. Two fixes
6111 (a) I re-wrote the offending code in a cleaner fashion.
6112 (b) If an incoming SMTP command contains a NULL character, it is rejected
6115 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6116 runs initgroups(), so that if the Exim user is in any additional groups,
6117 they will be used during message reception.
6123 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob