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.
38 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
39 checkers that were complaining about end of non-void function with no
42 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
43 This was a regression intruduced in 4.83 by another bugfix.
45 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
47 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
48 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
54 TF/01 Correctly close the server side of TLS when forking for delivery.
56 When a message was received over SMTP with TLS, Exim failed to clear up
57 the incoming connection properly after forking off the child process to
58 deliver the message. In some situations the subsequent outgoing
59 delivery connection happened to have the same fd number as the incoming
60 connection previously had. Exim would try to use TLS and fail, logging
61 a "Bad file descriptor" error.
63 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
64 utilities have not been installed.
66 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
67 temporary space as the ACL may create new global variables.
69 TL/01 LDAP support uses per connection or global context settings, depending
70 upon the detected version of the libraries at build time.
72 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
73 to extract and use the src ip:port in logging and expansions as if it
74 were a direct connection from the outside internet. PPv2 support was
75 updated based on HAProxy spec change in May 2014.
77 JH/02 Add ${listextract {number}{list}{success}{fail}}.
79 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
80 Properly escape header and check for NULL return.
82 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
85 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
87 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
88 characters in header names, implemented as a verify condition.
89 Contributed by Michael Fischer v. Mollard.
91 TL/05 Rename SPF condition results err_perm and err_temp to standardized
92 results permerror and temperror. Previous values are deprecated but
93 still accepted. In a future release, err_perm and err_temp will be
94 completely removed, which will be a backward incompatibility if the
95 ACL tests for either of these two old results. Patch contributed by
96 user bes-internal on the mailing list.
98 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
100 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
101 selectors, in both main and reject logs.
103 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
106 JH/07 Add malware type "sock" for talking to simple daemon.
108 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
110 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
111 routers/transports under cutthrough routing.
113 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
114 numbers. Touch up "bool" conditional to keep the same definition.
116 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
118 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
120 JH/12 Expand items in router/transport headers_add or headers_remove lists
121 individually rather than the list as a whole. Bug 1452.
123 Required for reasonable handling of multiple headers_ options when
124 they may be empty; requires that headers_remove items with embedded
125 colons must have them doubled (or the list-separator changed).
127 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
128 view the policy declared in the DMARC record. Currently, $dmarc_status
129 is a combined value of both the record presence and the result of the
132 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
134 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
135 dnslookup router and the smtp transport (applying to the forward
138 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
139 of ldap servers used for a specific lookup. Patch provided by Heiko
142 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
143 New variable $lookup_dnssec_authenticated for observability.
145 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
146 Patch submitted by Lars Timman.
148 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
150 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
151 Requires trusted mode and valid format message id, aborts otherwise.
152 Patch contributed by Heiko Schlichting.
154 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
155 certextract with support for various fields. Bug 1358.
157 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
158 is requested by default, modifiable by smtp transport option
161 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
162 operate on certificate variables to give certificate fingerprints
163 Also new ${sha256:cert_variable}.
165 JH/23 The PRDR feature is moved from being Experimental into the mainline.
167 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
168 Christian Aistleitner.
170 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
172 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
173 file. Patch from Wolfgang Breyha.
175 JH/25 Expand the coverage of the delivery $host and $host_address to
176 client authenticators run in verify callout. Bug 1476.
178 JH/26 Port service names are now accepted for tls_on_connect_ports, to
179 align with daemon_smtp_ports. Bug 72.
181 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
182 support and error reporting did not work properly.
184 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
185 and is readable. Patch from Andrew Colin Kissa.
187 TL/14 Enhance documentation of ${run expansion and how it parses the
188 commandline after expansion, particularly in the case when an
189 unquoted variable expansion results in an empty value.
191 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
193 PP/02 Fix internal collision of T_APL on systems which support RFC3123
194 by renaming away from it. Addresses GH issue 15, reported by
197 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
199 TL/15 SECURITY: prevent double expansion in math comparison functions
200 (can expand unsanitized data). Not remotely exploitable.
207 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
209 PP/02 Make -n do something, by making it not do something.
210 When combined with -bP, the name of an option is not output.
212 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
215 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
216 $sender_host_name and config options to manage this, and basic check
219 PP/05 DSCP support for outbound connections and control modifier for inbound.
221 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
222 (Only plugin which currently uses this is kerberos4, which nobody should
223 be using, but we should make it available and other future plugins might
224 conceivably use it, even though it would break NAT; stuff *should* be
225 using channel bindings instead).
227 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
228 name; added for Sendmail compatibility; requires admin caller.
229 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
230 just ignore it); requires trusted caller.
231 Also parse but ignore: -Ac -Am -X<logfile>
234 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
236 TL/02 Add +smtp_confirmation as a default logging option.
238 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
239 Patch by Magnus Holmgren from 2007-02-20.
241 TL/04 Bugzilla 1281 - Spec typo.
242 Bugzilla 1283 - Spec typo.
243 Bugzilla 1290 - Spec grammar fixes.
245 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
247 TL/06 Add Experimental DMARC support using libopendmarc libraries.
249 TL/07 Fix an out of order global option causing a segfault. Reported to dev
250 mailing list by by Dmitry Isaikin.
252 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
254 JH/02 Support "G" suffix to numbers in ${if comparisons.
256 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
258 NM/01 Bugzilla 1197 - Spec typo
259 Bugzilla 1196 - Spec examples corrections
261 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
263 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
264 gnutls_enable_pkcs11, but renamed to more accurately indicate its
267 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
268 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
270 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
271 "acl {{name}{arg}...}", and optional args on acl condition
274 JH/05 Permit multiple router/transport headers_add/remove lines.
276 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
278 JH/07 Avoid using a waiting database for a single-message-only transport.
279 Performance patch from Paul Fisher. Bugzilla 1262.
281 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
284 JH/09 Add $headers_added variable, with content from use of ACL modifier
285 add_header (but not yet added to the message). Bugzilla 199.
287 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
288 Pulled from Bugzilla 817 by Wolfgang Breyha.
290 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
292 (nb: this is the same fix as in Exim 4.80.1)
294 JH/11 Add A= logging on delivery lines, and a client_set_id option on
297 JH/12 Add optional authenticated_sender logging to A= and a log_selector
300 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
302 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
303 advertise SMTP AUTH mechanism to us, instead of a generic
304 protocol violation error. Also, make Exim more robust to bad
305 data from the Dovecot auth socket.
307 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
309 When a queue runner is handling a message, Exim first routes the
310 recipient addresses, during which it prunes them based on the retry
311 hints database. After that it attempts to deliver the message to
312 any remaining recipients. It then updates the hints database using
315 So if a recipient address works intermittently, it can get repeatedly
316 deferred at routing time. The retry hints record remains fresh so the
317 address never reaches the final cutoff time.
319 This is a fairly common occurrence when a user is bumping up against
320 their storage quota. Exim had some logic in its local delivery code
321 to deal with this. However it did not apply to per-recipient defers
322 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
324 This change adds a proper retry rule check during routing so that the
325 final cutoff time is checked against the message's age. We only do
326 this check if there is an address retry record and there is not a
327 domain retry record; this implies that previous attempts to handle
328 the address had the retry_use_local_parts option turned on. We use
329 this as an approximation for the destination being like a local
330 delivery, as in LMTP.
332 I suspect this new check makes the old local delivery cutoff check
333 redundant, but I have not verified this so I left the code in place.
335 TF/02 Correct gecos expansion when From: is a prefix of the username.
337 Test 0254 submits a message to Exim with the header
341 When I ran the test suite under the user fanf2, Exim expanded
342 the header to contain my full name, whereas it should have added
343 a Resent-Sender: header. It erroneously treats any prefix of the
344 username as equal to the username.
346 This change corrects that bug.
348 GF/01 DCC debug and logging tidyup
349 Error conditions log to paniclog rather than rejectlog.
350 Debug lines prefixed by "DCC: " to remove any ambiguity.
352 TF/03 Avoid unnecessary rebuilds of lookup-related code.
354 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
355 Bug spotted by Jeremy Harris; was flawed since initial commit.
356 Would have resulted in OCSP responses post-SNI triggering an Exim
357 NULL dereference and crash.
359 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
361 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
362 Bug detection, analysis and fix by Samuel Thibault.
363 Bugzilla 1331, Debian bug #698092.
365 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
367 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
368 Server implementation by Todd Lyons, client by JH.
369 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
370 config variable "prdr_enable" controls whether the server
371 advertises the facility. If the client requests PRDR a new
372 acl_data_smtp_prdr ACL is called once for each recipient, after
373 the body content is received and before the acl_smtp_data ACL.
374 The client is controlled by bolth of: a hosts_try_prdr option
375 on the smtp transport, and the server advertisement.
376 Default client logging of deliveries and rejections involving
377 PRDR are flagged with the string "PRDR".
379 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
380 fclose(). Diagnosis by Todd Lyons.
382 PP/17 Update configure.default to handle IPv6 localhost better.
383 Patch by Alain Williams (plus minor tweaks).
386 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
387 This is now consistent with GnuTLS, and is now documented: the
388 previous undocumented portable approach to treating the option as
389 unset was to force an expansion failure. That still works, and
390 an empty string is now equivalent.
392 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
393 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
394 not performing validation itself.
396 PP/20 Added force_command boolean option to pipe transport.
397 Patch from Nick Koston, of cPanel Inc.
399 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
402 TF/04 Added udpsend ACL modifer and hexquote expansion operator
404 PP/21 Fix eximon continuous updating with timestamped log-files.
405 Broken in a format-string cleanup in 4.80, missed when I repaired the
406 other false fix of the same issue.
407 Report and fix from Heiko Schlichting.
410 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
411 Report from Prashanth Katuri.
413 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
414 It's SecureTransport, so affects any MacOS clients which use the
415 system-integrated TLS libraries, including email clients.
417 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
418 using a MIME ACL for non-SMTP local injection.
419 Report and assistance in diagnosis by Warren Baker.
421 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
423 JH/16 Fix comparisons for 64b. Bugzilla 1385.
425 TL/09 Add expansion variable $authenticated_fail_id to keep track of
426 last id that failed so it may be referenced in subsequent ACL's.
428 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
431 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
432 ldap library initialization, allowing self-signed CA's to be
433 used. Also properly sets require_cert option later in code by
434 using NULL (global ldap config) instead of ldap handle (per
435 session). Bug diagnosis and testing by alxgomz.
437 TL/12 Enhanced documentation in the ratelimit.pl script provided in
438 the src/util/ subdirectory.
440 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
441 renamed to Transport Post Delivery Action by Jeremy Harris, as
444 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
445 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
446 redis_servers = needs to be configured which will be used by the redis
447 lookup. Patch from Warren Baker, of The Packet Hub.
449 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
451 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
452 hostname or reverse DNS when processing a host list. Used suggestions
453 from multiple comments on this bug.
455 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
457 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
458 Missed a few lines, added it to make the runtest require no keyboard
461 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
462 contains upper case chars. Make router use caseful_local_part.
464 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
465 support when GnuTLS has been built with p11-kit.
471 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
473 This, or similar/improved, will also be change PP/11 of 4.82.
479 PP/01 Handle short writes when writing local log-files.
480 In practice, only affects FreeBSD (8 onwards).
481 Bugzilla 1053, with thanks to Dmitry Isaikin.
483 NM/01 Bugzilla 949 - Documentation tweak
485 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
488 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
490 PP/02 Implemented gsasl authenticator.
492 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
494 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
495 `pkg-config foo` for cflags/libs.
497 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
498 with rest of GSASL and with heimdal_gssapi.
500 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
501 `pkg-config foo` for cflags/libs for the TLS implementation.
503 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
504 properties get this fed in as external SSF. A number of robustness
505 and debugging improvements to the cyrus_sasl authenticator.
507 PP/08 cyrus_sasl server now expands the server_realm option.
509 PP/09 Bugzilla 1214 - Log authentication information in reject log.
510 Patch by Jeremy Harris.
512 PP/10 Added dbmjz lookup type.
514 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
516 PP/12 MAIL args handles TAB as well as SP, for better interop with
517 non-compliant senders.
518 Analysis and variant patch by Todd Lyons.
520 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
521 Bug report from Lars Müller <lars@samba.org> (via SUSE),
522 Patch from Dirk Mueller <dmueller@suse.com>
524 PP/13 tls_peerdn now print-escaped for spool files.
525 Observed some $tls_peerdn in wild which contained \n, which resulted
526 in spool file corruption.
528 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
529 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
530 or write after TLS renegotiation, which otherwise led to messages
533 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
534 as a tracking header (ie: a signed header comes before the signature).
535 Patch from Wolfgang Breyha.
537 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
538 comma-sep list; embedded commas doubled.
540 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
542 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
544 Report and patch from Dmitry Banschikov.
546 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
547 Removed SSL_clear() after SSL_new() which led to protocol negotiation
548 failures. We appear to now support TLS1.1+ with Exim.
550 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
551 lets Exim select keys and certificates based upon TLS SNI from client.
552 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
553 before an outbound SMTP session. New log_selector, +tls_sni.
555 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
556 NULL dereference. Report and patch from Alun Jones.
558 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
559 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
560 Not seeing resolver debug output on NetBSD, but suspect this is a
561 resolver implementation change.
563 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
564 Left warnings. Added "eximon gdb" invocation mode.
566 PP/21 Defaulting "accept_8bitmime" to true, not false.
568 PP/22 Added -bw for inetd wait mode support.
570 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
571 locate the relevant includes and libraries. Made this the default.
573 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
574 Bugzilla 1246, report and most of solution from Tomasz Kusy.
576 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
577 This may cause build issues on older platforms.
579 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
580 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
581 gnutls_require_mac & gnutls_require_protocols (no longer supported).
582 Added SNI support via GnuTLS too.
583 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
585 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
587 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
588 Applied second patch from Janne, implementing suggestion to default
589 multiple-strings-in-record handling to match SPF spec.
591 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
593 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
594 read-only, out of scope).
595 Patch from Wolfgang Breyha, report from Stuart Northfield.
597 PP/29 Fix three issues highlighted by clang analyser static analysis.
598 Only crash-plausible issue would require the Cambridge-specific
599 iplookup router and a misconfiguration.
600 Report from Marcin Mirosław.
602 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
604 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
605 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
606 As part of this, removing so much warning spew let me fix some minor
607 real issues in debug logging.
609 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
610 assignment on my part. Fixed.
612 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
613 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
614 Janne Snabb (who went above and beyond: thank you).
616 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
617 string otherwise requires a connection and a bunch more work and it's
618 relatively easy to get wrong. Should also expose TLS library linkage
621 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
622 64-bit ${eval} (JH/03).
624 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
625 GNU libc to support some of the 64-bit stuff, should not lead to
626 conflicts. Defined before os.h is pulled in, so if a given platform
627 needs to override this, it can.
629 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
630 protection layer was required, which is not implemented.
631 Bugzilla 1254, patch from Wolfgang Breyha.
633 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
634 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
635 tls_dhparam take prime identifiers. Also unbreak combination of
636 OpenSSL+DH_params+TLSSNI.
638 PP/39 Disable SSLv2 by default in OpenSSL support.
644 PP/01 Solaris build fix for Oracle's LDAP libraries.
645 Bugzilla 1109, patch from Stephen Usher.
647 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
649 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
652 TF/02 Fix a couple more cases where we did not log the error message
653 when unlink() failed. See also change 4.74-TF/03.
655 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
656 lock up or crash if it happened to be inside a call to libc when it
657 got a SIGUSR1 from exiwhat.
659 The SIGUSR1 handler appends the current process status to the process
660 log which is later printed by exiwhat. It used to use the general
661 purpose logging code to do this, but several functions it calls are
662 not safe for signals.
664 The new output code in the SIGUSR1 handler is specific to the process
665 log, and simple enough that it's easy to inspect for signal safety.
666 Removing some special cases also simplifies the general logging code.
667 Removing the spurious timestamps from the process log simplifies
670 TF/04 Improved ratelimit ACL condition.
672 The /noupdate option has been deprecated in favour of /readonly which
673 has clearer semantics. The /leaky, /strict, and /readonly update modes
674 are mutually exclusive. The update mode is no longer included in the
675 database key; it just determines when the database is updated. (This
676 means that when you upgrde Exim will forget old rate measurements.)
678 Exim now checks that the per_* options are used with an update mode that
679 makes sense for the current ACL. For example, when Exim is processing a
680 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
681 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
682 must specify per_mail/readonly. If you omit the update mode it defaults to
683 /leaky where that makes sense (as before) or /readonly where required.
685 The /noupdate option is now undocumented but still supported for
686 backwards compatibility. It is equivalent to /readonly except that in
687 ACLs where /readonly is required you may specify /leaky/noupdate or
688 /strict/noupdate which are treated the same as /readonly.
690 A useful new feature is the /count= option. This is a generalization
691 of the per_byte option, so that you can measure the throughput of other
692 aggregate values. For example, the per_byte option is now equivalent
693 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
695 The per_rcpt option has been generalized using the /count= mechanism
696 (though it's more complicated than the per_byte equivalence). When it is
697 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
698 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
699 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
700 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
701 Note that using per_rcpt with a non-readonly update mode in more than
702 one ACL will cause the recipients to be double-counted. (The per_mail
703 and per_byte options don't have this problem.)
705 The handling of very low rates has changed slightly. If the computed rate
706 is less than the event's count (usually one) then this event is the first
707 after a long gap. In this case the rate is set to the same as this event's
708 count, so that the first message of a spam run is counted properly.
710 The major new feature is a mechanism for counting the rate of unique
711 events. The new per_addr option counts the number of different
712 recipients that someone has sent messages to in the last time period. It
713 behaves like per_rcpt if all the recipient addresses are different, but
714 duplicate recipient addresses do not increase the measured rate. Like
715 the /count= option this is a general mechanism, so the per_addr option
716 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
717 example, measure the rate that a client uses different sender addresses
718 with the options per_mail/unique=$sender_address. There are further
719 details in the main documentation.
721 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
723 TF/06 Removed a few PCRE remnants.
725 TF/07 Automatically extract Exim's version number from tags in the git
726 repository when doing development or release builds.
728 PP/02 Raise smtp_cmd_buffer_size to 16kB.
729 Bugzilla 879. Patch from Paul Fisher.
731 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
732 Heavily based on revision 40f9a89a from Simon Arlott's tree.
735 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
737 PP/05 Variable $av_failed, true if the AV scanner deferred.
738 Bugzilla 1078. Patch from John Horne.
740 PP/06 Stop make process more reliably on build failure.
741 Bugzilla 1087. Patch from Heiko Schlittermann.
743 PP/07 Make maildir_use_size_file an _expandable_ boolean.
744 Bugzilla 1089. Patch from Heiko Schlittermann.
746 PP/08 Handle ${run} returning more data than OS pipe buffer size.
747 Bugzilla 1131. Patch from Holger Weiß.
749 PP/09 Handle IPv6 addresses with SPF.
750 Bugzilla 860. Patch from Wolfgang Breyha.
752 PP/10 GnuTLS: support TLS 1.2 & 1.1.
754 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
757 PP/11 match_* no longer expand right-hand-side by default.
758 New compile-time build option, EXPAND_LISTMATCH_RHS.
759 New expansion conditions, "inlist", "inlisti".
761 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
763 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
765 PP/14 fix log_write() format string regression from TF/03.
766 Bugzilla 1152. Patch from Dmitry Isaikin.
772 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
774 PP/02 Harmonised TLS library version reporting; only show if debugging.
775 Layout now matches that introduced for other libraries in 4.74 PP/03.
777 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
779 PP/04 New "dns_use_edns0" global option.
781 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
784 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
785 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
787 TK/01 Updated PolarSSL code to 0.14.2.
788 Bugzilla 1097. Patch from Andreas Metzler.
790 PP/07 Catch divide-by-zero in ${eval:...}.
793 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
796 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
797 format-string attack -- SECURITY: remote arbitrary code execution.
799 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
800 time unintentionally subject to list matching rules, letting the header
801 cause arbitrary Exim lookups (of items which can occur in lists, *not*
802 arbitrary string expansion). This allowed for information disclosure.
804 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
805 INT_MIN/-1 -- value coerced to INT_MAX.
811 NM/01 Workround for PCRE version dependency in version reporting
814 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
815 This fixes portability to compilers other than gcc, notably
816 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
818 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
819 makefiles for portability to HP-UX and POSIX correctness.
821 PP/01 Permit LOOKUP_foo enabling on the make command-line.
822 Also via indented variable definition in the Makefile.
823 (Debugging by Oliver Heesakkers).
825 PP/02 Restore caching of spamd results with expanded spamd_address.
826 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
828 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
829 Improves build reliability. Fix from: Frank Elsner
831 NM/02 Fix wide character breakage in the rfc2047 coding
832 Fixes bug 1064. Patch from Andrey N. Oktyabrski
834 NM/03 Allow underscore in dnslist lookups
835 Fixes bug 1026. Patch from Graeme Fowler
837 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
838 Code patches from Adam Ciarcinski of NetBSD.
840 NM/04 Fixed exiqgrep to cope with mailq missing size issue
843 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
844 is logged, to avoid truncation. Patch from John Horne.
846 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
847 Patch from Jakob Hirsch.
849 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
850 SQL string expansion failure details.
851 Patch from Andrey Oktyabrski.
853 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
854 Patch from Simon Arlott.
856 PP/09 New lookups functionality failed to compile on old gcc which rejects
857 extern declarations in function scope.
858 Patch from Oliver Fleischmann
860 PP/10 Use sig_atomic_t for flags set from signal handlers.
861 Check getgroups() return and improve debugging.
862 Fixed developed for diagnosis in bug 927 (which turned out to be
865 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
866 Patch from Mark Zealey.
868 PP/12 Bugzilla 1056: Improved spamd server selection.
869 Patch from Mark Zealey.
871 PP/13 Bugzilla 1086: Deal with maildir quota file races.
872 Based on patch from Heiko Schlittermann.
874 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
875 Patch from Uwe Doering, sign-off by Michael Haardt.
877 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
878 variable declaration deep within a block. Bug and patch from
881 PP/15 lookups-Makefile IRIX compatibilty coercion.
883 PP/16 Make DISABLE_DKIM build knob functional.
885 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
886 Patch by Simon Arlott
888 TF/03 Fix valgrind.h portability to C89 compilers that do not support
889 variable argument macros. Our copy now differs from upstream.
895 TF/01 Failure to get a lock on a hints database can have serious
896 consequences so log it to the panic log.
898 TF/02 Log LMTP confirmation messages in the same way as SMTP,
899 controlled using the smtp_confirmation log selector.
901 TF/03 Include the error message when we fail to unlink a spool file.
903 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
904 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
905 for maintaining out-of-tree patches for some time.
907 PP/01 Bugzilla 139: Documentation and portability issues.
908 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
909 Handle per-OS dynamic-module compilation flags.
911 PP/02 Let /dev/null have normal permissions.
912 The 4.73 fixes were a little too stringent and complained about the
913 permissions on /dev/null. Exempt it from some checks.
914 Reported by Andreas M. Kirchwitz.
916 PP/03 Report version information for many libraries, including
917 Exim version information for dynamically loaded libraries. Created
918 version.h, now support a version extension string for distributors
919 who patch heavily. Dynamic module ABI change.
921 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
922 privilege escalation vulnerability whereby the Exim run-time user
923 can cause root to append content of the attacker's choosing to
926 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
929 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
930 If dropping privileges for untrusted macros, we disabled normal logging
931 on the basis that it would fail; for the Exim run-time user, this is not
932 the case, and it resulted in successful deliveries going unlogged.
933 Fixed. Reported by Andreas Metzler.
939 PP/01 Date: & Message-Id: revert to normally being appended to a message,
940 only prepend for the Resent-* case. Fixes regression introduced in
941 Exim 4.70 by NM/22 for Bugzilla 607.
943 PP/02 Include check_rfc2047_length in configure.default because we're seeing
944 increasing numbers of administrators be bitten by this.
946 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
948 PP/03 Bugzilla 994: added openssl_options main configuration option.
950 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
952 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
954 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
956 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
957 without a peer certificate, leading to a segfault because of an
958 assumption that peers always have certificates. Be a little more
959 paranoid. Problem reported by Martin Tscholak.
961 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
962 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
963 NB: ClamAV planning to remove STREAM in "middle of 2010".
964 CL also introduces -bmalware, various -d+acl logging additions and
965 more caution in buffer sizes.
967 PP/09 Implemented reverse_ip expansion operator.
969 PP/10 Bugzilla 937: provide a "debug" ACL control.
971 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
973 PP/12 Bugzilla 973: Implement --version.
975 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
977 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
979 PP/15 Bugzilla 816: support multiple condition rules on Routers.
981 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
982 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
983 ignore trailing whitespace.
985 JJ/02 prevent non-panic DKIM error from being sent to paniclog
987 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
990 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
991 Notification from Dr Andrew Aitchison.
993 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
994 ExtendedDetectionInfo response format.
995 Notification from John Horne.
997 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1000 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1001 XSL and documented dependency on system catalogs, with examples of how
1004 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1007 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1008 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1009 configuration file which is writeable by the Exim user or group.
1011 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1012 of configuration files to cover files specified with the -C option if
1013 they are going to be used with root privileges, not just the default
1016 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1017 option (effectively making it always true).
1019 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1020 files to be used while preserving root privileges.
1022 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1023 that rogue child processes cannot use them.
1025 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1026 run-time user, instead of root.
1028 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1029 Exim run-time user without dropping privileges.
1031 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1032 result string, instead of calling string_vformat() twice with the same
1035 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1036 for other users. Others should always drop root privileges if they use
1037 -C on the command line, even for a whitelisted configure file.
1039 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1041 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1047 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1048 $data_path, and $header_path variables; fixed documentation bugs and
1051 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1052 exipick to access non-standard spools, including the "frozen" queue
1055 NM/01 Bugzilla 965: Support mysql stored procedures.
1056 Patch from Alain Williams
1058 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1060 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1061 Patch from Andreas Metzler
1063 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1064 Patch from Kirill Miazine
1066 NM/05 Bugzilla 671: Added umask to procmail example.
1068 JJ/03 installed exipick 20100323.0, fixing doc bug
1070 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1071 directory. Notification and patch from Dan Rosenberg.
1073 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1075 TK/02 Improve log output when DKIM signing operation fails.
1077 MH/01 Treat the transport option dkim_domain as a colon separated
1078 list, not as a single string, and sign the message with each element,
1079 omitting multiple occurences of the same signer.
1081 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1082 Bugzilla 985, 986. Patch by Simon Arlott
1084 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1085 Patch by Simon Arlott
1087 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1088 MBX locking. Notification from Dan Rosenberg.
1094 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1096 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1098 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1100 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1102 NM/04 Bugzilla 915: Flag broken perl installation during build.
1108 TK/01 Added patch by Johannes Berg that expands the main option
1109 "spamd_address" if it starts with a dollar sign.
1111 TK/02 Write list of recipients to X-Envelope-Sender header when building
1112 the mbox-format spool file for content scanning (suggested by Jakob
1115 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1116 (http://www.dcc-servers.net/) support via dccifd. Activated by
1117 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1119 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1120 by Mark Daniel Reidel <mr@df.eu>.
1122 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1123 When building exim an external PCRE library is now needed -
1124 PCRE is a system library on the majority of modern systems.
1125 See entry on PCRE_LIBS in EDITME file.
1127 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1128 conversation. Added nologin parameter to request.
1129 Patch contributed by Kirill Miazine.
1131 TF/01 Do not log submission mode rewrites if they do not change the address.
1133 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1135 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1136 log files in place. Contributed by Roberto Lima.
1138 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1140 TF/03 Bugzilla 615: When checking the local_parts router precondition
1141 after a local_part_suffix or local_part_prefix option, Exim now
1142 does not use the address's named list lookup cache, since this
1143 contains cached lookups for the whole local part.
1145 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1146 Robert Millan. Documentation is in experimental-spec.txt.
1148 TF/04 Bugzilla 668: Fix parallel build (make -j).
1150 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1152 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1153 Patch provided by Jan Srzednicki.
1155 TF/05 Leading white space used to be stripped from $spam_report which
1156 wrecked the formatting. Now it is preserved.
1158 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1159 that they are available at delivery time.
1161 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1163 TF/08 TLS error reporting now respects the incoming_interface and
1164 incoming_port log selectors.
1166 TF/09 Produce a more useful error message if an SMTP transport's hosts
1167 setting expands to an empty string.
1169 NM/06 Bugzilla 744: EXPN did not work under TLS.
1170 Patch provided by Phil Pennock.
1172 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1173 Patch provided by Richard Godbee.
1175 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1176 acl_smtp_notquit, added index entry.
1178 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1179 Patch provided by Eugene Bujak.
1181 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1182 accept(). Patch provided by Maxim Dounin.
1184 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1185 Patch provided by Phil Pennock.
1187 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1189 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1190 Patch provided by Brad "anomie" Jorsch.
1192 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1193 Patch provided by Dean Brooks.
1195 TK/05 Add native DKIM support (does not depend on external libraries).
1197 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1198 Patch provided by Graeme Fowler.
1200 NM/16 Bugzilla 851: Documentation example syntax fix.
1202 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1204 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1207 NM/19 Bugzilla 745: TLS version reporting.
1208 Patch provided by Phil Pennock.
1210 NM/20 Bugzilla 167: bool: condition support.
1211 Patch provided by Phil Pennock.
1213 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1214 clients. Patch provided by Phil Pennock.
1216 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1217 Patch provided by Brad "anomie" Jorsch.
1219 NM/23 Bugzilla 687: Fix misparses in eximstats.
1220 Patch provided by Heiko Schlittermann.
1222 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1223 Patch provided by Heiko Schlittermann.
1225 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1226 plus update to original patch.
1228 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1230 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1231 Patch provided by David Brownlee.
1233 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1235 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1237 NM/30 Bugzilla 888: TLS documentation bugfixes.
1239 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1241 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1242 Unlike the original bugzilla I have changed all shell scripts in src tree.
1244 NM/33 Bugzilla 898: Transport filter timeout fix.
1245 Patch by Todd Rinaldo.
1247 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1248 Patch by Serge Demonchaux.
1250 NM/35 Bugzilla 39: Base64 decode bug fixes.
1251 Patch by Jakob Hirsch.
1253 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1255 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1257 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1259 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1265 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1266 ALT-N's libdkim that I have put here:
1267 http://duncanthrax.net/exim-experimental/
1269 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1270 were called 'true' and it seems that C99 defines that as a reserved
1271 keyword to be used with 'bool' variable types. That means you could
1272 not include C99-style headers which use bools without triggering
1273 build errors in sieve.c.
1275 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1276 as mailq or other aliases. Changed the --help handling significantly
1277 to do whats expected. exim_usage() emits usage/help information.
1279 SC/01 Added the -bylocaldomain option to eximstats.
1281 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1283 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1285 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1291 PH/01 Another patch from the Sieve maintainer.
1293 PH/02 When an IPv6 address is converted to a string for single-key lookup
1294 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1295 dots are used instead of colons so that keys in lsearch files need not
1296 contain colons. This was done some time before quoting was made available
1297 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1298 (notated using the quote facility) so as to distinguish them from IPv4
1299 keys. This meant that lookups for IP addresses in host lists did not work
1300 for iplsearch lookups.
1302 This has been fixed by arranging for IPv6 addresses to be expressed with
1303 colons if the lookup type is iplsearch. This is not incompatible, because
1304 previously such lookups could never work.
1306 The situation is now rather anomolous, since one *can* have colons in
1307 ordinary lsearch keys. However, making the change in all cases is
1308 incompatible and would probably break a number of configurations.
1310 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1313 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1314 conversion specification without a maximum field width, thereby enabling
1315 a rogue spamd server to cause a buffer overflow. While nobody in their
1316 right mind would setup Exim to query an untrusted spamd server, an
1317 attacker that gains access to a server running spamd could potentially
1318 exploit this vulnerability to run arbitrary code as the Exim user.
1320 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1321 $primary_hostname instead of what libspf2 thinks the hosts name is.
1323 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1324 a directory entry by the name of the lookup key. Previously, if a
1325 symlink pointed to a non-existing file or a file in a directory that
1326 Exim lacked permissions to read, a lookup for a key matching that
1327 symlink would fail. Now it is enough that a matching directory entry
1328 exists, symlink or not. (Bugzilla 503.)
1330 PH/03 The body_linecount and body_zerocount variables are now exported in the
1333 PH/04 Added the $dnslist_matched variable.
1335 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1336 This means they are set thereafter only if the connection becomes
1339 PH/06 Added the client_condition to authenticators so that some can be skipped
1340 by clients under certain conditions.
1342 PH/07 The error message for a badly-placed control=no_multiline_responses left
1343 "_responses" off the end of the name.
1345 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1347 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1348 (without spaces) instead of just copying the configuration text.
1350 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1352 PH/11 Added $max_received_linelength.
1354 PH/12 Added +ignore_defer and +include_defer to host lists.
1356 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1357 way in which PCRE > 7.0 is built.
1359 PH/14 Implemented queue_only_load_latch.
1361 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1362 MAIL command. The effect was to mangle the value on 64-bit systems.
1364 PH/16 Another patch from the Sieve maintainer.
1366 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1368 PH/18 If a system quota error occurred while trying to create the file for
1369 a maildir delivery, the message "Mailbox is full" was not appended to the
1370 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1371 only to a quota excession during the actual writing of the file.
1373 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1374 characters?) which causes problems in log lines. The DN values are now
1375 passed through string_printing() before being added to log lines.
1377 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1378 and InterBase are left for another time.)
1380 PH/21 Added message_body_newlines option.
1382 PH/22 Guard against possible overflow in moan_check_errorcopy().
1384 PH/23 POSIX allows open() to be a macro; guard against that.
1386 PH/24 If the recipient of an error message contained an @ in the local part
1387 (suitably quoted, of course), incorrect values were put in $domain and
1388 $local_part during the evaluation of errors_copy.
1394 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1395 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1398 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1399 issue a MAIL command.
1401 PH/02 In an ACL statement such as
1403 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1405 if a client was not listed at all, or was listed with a value other than
1406 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1407 the condition was not true (as it should be), so access was not denied.
1408 The bug was that the ! inversion was incorrectly passed on to the second
1409 item. This has been fixed.
1411 PH/03 Added additional dnslists conditions == and =& which are different from
1412 = and & when the dns lookup returns more than one IP address.
1414 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1415 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1417 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1418 FSYNC, which compiles an option called disable_fsync that allows for
1419 bypassing fsync(). The documentation is heavily laced with warnings.
1421 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1423 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1424 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1425 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1426 including adding "make clean"; (3) Added -fPIC when compiling the test
1427 dynamically loaded module, to get rid of a warning.
1429 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1430 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1431 = 0s. The bug is otherwise harmless.
1433 PH/07 There was a bug in the dovecot authenticator such that the value of
1434 $auth1 could be overwritten, and so not correctly preserved, after a
1435 successful authentication. This usually meant that the value preserved by
1436 the server_setid option was incorrect.
1438 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1440 PH/09 Installed PCRE release 7.0.
1442 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1443 run for batched SMTP input. It is now run at the start of every message
1444 in the batch. While fixing this I discovered that the process information
1445 (output by running exiwhat) was not always getting set for -bs and -bS
1446 input. This is fixed, and it now also says "batched" for BSMTP.
1448 PH/11 Added control=no_pipelining.
1450 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1451 patch, slightly modified), and move the expansion of helo_data till after
1452 the connection is made in the smtp transport (so it can use these
1455 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1457 PH/14 Added log_selector = +pid.
1459 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1461 PH/16 Add ${if forany and ${if forall.
1463 PH/17 Added dsn_from option to vary the From: line in DSNs.
1465 PH/18 Flush SMTP output before performing a callout, unless control =
1466 no_callout_flush is set.
1468 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1469 was true (the default) a successful delivery failed to delete the retry
1470 item, thus causing premature timeout of the address. The bug is now
1473 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1475 PH/21 Long custom messages for fakedefer and fakereject are now split up
1476 into multiline reponses in the same way that messages for "deny" and
1477 other ACL rejections are.
1479 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1480 with slight modification.
1482 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1483 draft, changing the syntax and factoring some duplicate code".
1485 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1486 for deliveries of the second and subsequent messages over the same SMTP
1489 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1490 ${reduce, with only minor "tidies".
1492 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1494 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1495 expansion side effects.
1497 PH/27 When a message times out after an over-quota error from an Exim-imposed
1498 quota, the bounce message says "mailbox is full". This message was not
1499 being given when it was a system quota that was exceeded. It now should
1502 MH/03 Made $recipients available in local_scan(). local_scan() already has
1503 better access to the recipient list through recipients_list[], but
1504 $recipients can be useful in postmaster-provided expansion strings.
1506 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1507 in the case of a MAIL command with additional options following the
1508 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1509 were accidentally chopped off.
1511 PH/29 SMTP synchronization checks are implemented when a command is read -
1512 there is a check that no more input is waiting when there shouldn't be
1513 any. However, for some commands, a delay in an ACL can mean that it is
1514 some time before the response is written. In this time, more input might
1515 arrive, invalidly. So now there are extra checks after an ACL has run for
1516 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1517 pipelining has not been advertised.
1519 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1521 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1522 returned in the SMTP response when smtp_return_error_details was false.
1523 This has been fixed.
1525 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1526 instead of the C I/O that was originally supplied, because problems were
1527 reported on Solaris.
1529 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1530 Exim which did not show up earlier: it was assuming that a call to
1531 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1532 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1533 was a macro that became an assignment, so it seemed to work. This has
1534 changed to a proper function call with a void return, hence the compile
1535 error. Exim's code has been fixed.
1537 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1540 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1542 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1544 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1545 because it turns out to be too big or there is a timeout) there is no
1546 "Completed" line in the log. When some messages of this type were
1547 selected by exigrep, they were listed as "not completed". Others were
1548 picked up by some special patterns. I have improved the selection
1549 criteria to be more general.
1551 PH/38 The host_find_failed option in the manualroute router can now be set
1552 to "ignore", to completely ignore a host whose IP address cannot be
1553 found. If all hosts are ignored, the behaviour is controlled by the new
1554 host_all_ignored option.
1556 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1557 homing or because of multiple MX records with /mx) generated more than
1558 one IP address, and the following item turned out to be the local host,
1559 all the secondary addresses of the first item were incorrectly removed
1560 from the list, along with the local host and any following hosts (which
1561 is what is supposed to happen).
1563 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1564 whoever called Exim into the -H file. In the case of the daemon it was
1565 behaving confusingly. When first started, it used values for whoever
1566 started the daemon, but after a SIGHUP it used the Exim user (because it
1567 calls itself on a restart). I have changed the code so that it now always
1570 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1571 message are rejected with the same error (e.g. no authentication or bad
1572 sender address), and a DATA command is nevertheless sent (as can happen
1573 with PIPELINING or a stupid MUA), the error message that was given to the
1574 RCPT commands is included in the rejection of the DATA command. This is
1575 intended to be helpful for MUAs that show only the final error to their
1578 PH/42 Another patch from the Sieve maintainer.
1580 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1581 Eximstats - Fixed some broken HTML links and added missing column headers
1583 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1584 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1586 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1588 PH/43 Yet another patch from the Sieve maintainer.
1590 PH/44 I found a way to check for a TCP/IP connection going away before sending
1591 the response to the final '.' that terminates a message, but only in the
1592 case where the client has not sent further data following the '.'
1593 (unfortunately, this is allowed). However, in many cases there won't be
1594 any further data because there won't be any more messages to send. A call
1595 to select() can be used: if it shows that the input is "ready", there is
1596 either input waiting, or the socket has been closed. An attempt to read
1597 the next input character can distinguish the two cases. Previously, Exim
1598 would have sent an OK response which the client would never have see.
1599 This could lead to message repetition. This fix should cure that, at
1600 least in a lot of common cases.
1602 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1603 advertised in response to EHLO.
1609 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1610 fixed by 4.65/MH/01 (is this a record?) are fixed:
1612 (i) An empty string was always treated as zero by the numeric comparison
1613 operators. This behaviour has been restored.
1615 (ii) It is documented that the numeric comparison operators always treat
1616 their arguments as decimal numbers. This was broken in that numbers
1617 starting with 0 were being interpreted as octal.
1619 While fixing these problems I realized that there was another issue that
1620 hadn't been noticed. Values of message_size_limit (both the global option
1621 and the transport option) were treated as octal if they started with 0.
1622 The documentation was vague. These values are now always treated as
1623 decimal, and I will make that clear in the documentation.
1629 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1630 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1633 MH/01 Don't check that the operands of numeric comparison operators are
1634 integers when their expansion is in "skipping" mode (fixes bug
1635 introduced by 4.64-PH/07).
1637 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1638 child addresses, Exim now panics and dies. Previously, because the count
1639 is held in a short int, deliveries were likely to be lost. As such a
1640 large number of recipients for a single message is ridiculous
1641 (performance will be very, very poor), I have chosen to impose a limit
1642 rather than extend the field.
1648 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1649 leftover -K file (the existence of which was triggered by #402).
1650 While we were at it, introduced process PID as part of the -K
1651 filename. This should rule out race conditions when creating
1654 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1655 processing considerably. Previous code took too long for large mails,
1656 triggering a timeout which in turn triggers #401.
1658 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1659 in the DK code in transports.c. sendfile() is not really portable,
1660 hence the _LINUX specificness.
1662 TF/01 In the add_headers option to the mail command in an Exim filter,
1663 there was a bug that Exim would claim a syntax error in any
1664 header after the first one which had an odd number of characters
1667 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1668 callout verification, Exim cached a "reject" for the entire domain. This
1669 is correct for most verifications, but it is not correct for a recipient
1670 verification with use_sender or use_postmaster set, because in that case
1671 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1672 case of MAIL FROM:<> rejection from other early rejections (e.g.
1673 rejection of HELO). When verifying a recipient using a non-null MAIL
1674 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1675 Whatever the result of the callout, the value of the domain cache is
1676 left unchanged (for any other kind of callout, getting as far as trying
1677 RCPT means that the domain itself is ok).
1679 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1682 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1683 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1684 session, the daemon ignores ECONNECT errors and logs others; it now
1685 ignores EPIPE as well.
1687 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1688 (quoted-printable decoding).
1690 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1691 later the small subsequent patch to fix an introduced bug.
1693 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1695 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1697 PH/08 An error is now given if message_size_limit is specified negative.
1699 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1700 to be given (somewhat) arbitrary names.
1702 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1705 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1706 miscellaneous code fixes
1708 PH/10 Added the log_reject_target ACL modifier to specify where to log
1711 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1712 hostname. This is wrong, because it relates to the incoming message (and
1713 probably the interface on which it is arriving) and not to the outgoing
1714 callout (which could be using a different interface). This has been
1715 changed to use the value of the helo_data option from the smtp transport
1716 instead - this is what is used when a message is actually being sent. If
1717 there is no remote transport (possible with a router that sets up host
1718 addresses), $smtp_active_hostname is used.
1720 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1721 tweaks were necessary in order to get it to work (see also 21 below):
1722 (a) The code assumed that strncpy() returns a negative number on buffer
1723 overflow, which isn't the case. Replaced with Exim's string_format()
1725 (b) There were several signed/unsigned issues. I just did the minimum
1726 hacking in of casts. There is scope for a larger refactoring.
1727 (c) The code used strcasecmp() which is not a standard C function.
1728 Replaced with Exim's strcmpic() function.
1729 (d) The code set only $1; it now sets $auth1 as well.
1730 (e) A simple test gave the error "authentication client didn't specify
1731 service in request". It would seem that Dovecot has changed its
1732 interface. Fortunately there's a specification; I followed it and
1733 changed what the client sends and it appears to be working now.
1735 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1738 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1739 address A is aliased to B and C, where B exists and C does not. Without
1740 -v the output is "A verified" because verification stops after a
1741 successful redirection if more than one address is generated. However,
1742 with -v the child addresses are also verified. Exim was outputting "A
1743 failed to verify" and then showing the successful verification for C,
1744 with its parentage. It now outputs "B failed to verify", showing B's
1745 parentage before showing the successful verification of C.
1747 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1748 look up a TXT record in a specific list after matching in a combined
1751 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1752 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1753 they consult the DNS. I had assumed they would set it the way they
1754 wanted; and indeed my experiments on Linux seem to show that in some
1755 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1756 To be on the safe side, however, I have now made the interface to
1757 host_find_byname() similar to host_find_bydns(), with an argument
1758 containing the DNS resolver options. The host_find_byname() function now
1759 sets these options at its start, just as host_find_bydns() does. The smtp
1760 transport options dns_qualify_single and dns_search_parents are passed to
1761 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1762 of host_find_byname() use the default settings of RES_DEFNAMES
1763 (qualify_single) but not RES_DNSRCH (search_parents).
1765 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1766 spool_read_header() do less string testing, by means of a preliminary
1767 switch on the second character of optional "-foo" lines. (This is
1768 overdue, caused by the large number of possibilities that now exist.
1769 Originally there were few.) While I was there, I also converted the
1770 str(n)cmp tests so they don't re-test the leading "-" and the first
1771 character, in the hope this might squeeze out yet more improvement.
1773 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1774 flag allowing group syntax was set by the header_syntax check but not
1775 turned off, possible causing trouble later; (2) The flag was not being
1776 set at all for the header_verify test, causing "group"-style headers to
1777 be rejected. I have now set it in this case, and also caused header_
1778 verify to ignore an empty address taken from a group. While doing this, I
1779 came across some other cases where the code for allowing group syntax
1780 while scanning a header line wasn't quite right (mostly, not resetting
1781 the flag correctly in the right place). These bugs could have caused
1782 trouble for malformed header lines. I hope it is now all correct.
1784 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1785 with the "reply" argument non-NULL. The code, however (which originally
1786 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1787 but it didn't always do it. This confused somebody who was copying the
1788 code for some other use. I have removed all the tests.
1790 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1791 feature that was used to support insecure browsers during the U.S. crypto
1792 embargo. It requires special client support, and Exim is probably the
1793 only MTA that supported it -- and would never use it because real RSA is
1794 always available. This code has been removed, because it had the bad
1795 effect of slowing Exim down by computing (never used) parameters for the
1796 RSA_EXPORT functionality.
1798 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1799 authenticator to fail if there's a tab character in the incoming data
1800 (there should never be unless someone is messing about, as it's supposed
1801 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1802 if the connection is using TLS or if the remote IP is the same as the
1803 local IP, and the "valid-client-cert option" if a client certificate has
1806 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1807 authenticators. This can be used for authorization after authentication
1808 succeeds. (In the case of plaintext, it servers for both authentication
1811 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1812 if any retry times were supplied.
1814 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1815 connection, where there is no client IP address to check. In this
1816 situation, the verify now always succeeds.
1818 PH/25 Applied John Jetmore's -Mset patch.
1820 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1822 PH/27 In a string expansion for a processed (not raw) header when multiple
1823 headers of the same name were present, leading whitespace was being
1824 removed from all of them, but trailing whitespace was being removed only
1825 from the last one. Now trailing whitespace is removed from each header
1826 before concatenation. Completely empty headers in a concatenation (as
1827 before) are ignored.
1829 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1830 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1832 PH/29 [Removed. This was a change that I later backed out, and forgot to
1833 correct the ChangeLog entry (that I had efficiently created) before
1834 committing the later change.]
1836 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1837 address errors (4xx response to RCPT) over the same connection as other
1838 messages routed to the same hosts. Such deliveries are always "forced",
1839 so retry times are not inspected. This resulted in far too many retries
1840 for the affected addresses. The effect occurred only when there were more
1841 hosts than the hosts_max_try setting in the smtp transport when it had
1842 the 4xx errors. Those hosts that it had tried were not added to the list
1843 of hosts for which the message was waiting, so if all were tried, there
1844 was no problem. Two fixes have been applied:
1846 (i) If there are any address or message errors in an SMTP delivery, none
1847 of the hosts (tried or untried) are now added to the list of hosts
1848 for which the message is waiting, so the message should not be a
1849 candidate for sending over the same connection that was used for a
1850 successful delivery of some other message. This seems entirely
1851 reasonable: after all the message is NOT "waiting for some host".
1852 This is so "obvious" that I'm not sure why it wasn't done
1853 previously. Hope I haven't missed anything, but it can't do any
1854 harm, as the worst effect is to miss an optimization.
1856 (ii) If, despite (i), such a delivery is accidentally attempted, the
1857 routing retry time is respected, so at least it doesn't keep
1858 hammering the server.
1860 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1861 in ${readsocket because some servers need this prod.
1863 PH/32 Added some extra debug output when updating a wait-xxx database.
1865 PH/33 The hint "could be header name not terminated by colon", which has been
1866 given for certain expansion errors for a long time, was not being given
1867 for the ${if def:h_colon_omitted{... case.
1869 PH/34 The spec says: "With one important exception, whenever a domain list is
1870 being scanned, $domain contains the subject domain." There was at least
1871 one case where this was not true.
1873 PH/35 The error "getsockname() failed: connection reset by peer" was being
1874 written to the panic log as well as the main log, but it isn't really
1875 panic-worthy as it just means the connection died rather early on. I have
1876 removed the panic log writing for the ECONNRESET error when getsockname()
1879 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1880 runs only) independently of the message's sender address. This meant
1881 that, if the 4xx error was in fact related to the sender, a different
1882 message to the same recipient with a different sender could confuse
1883 things. In particualar, this can happen when sending to a greylisting
1884 server, but other circumstances could also provoke similar problems.
1885 I have changed the default so that the retry time for these errors is now
1886 based a combination of the sender and recipient addresses. This change
1887 can be overridden by setting address_retry_include_sender=false in the
1890 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1891 remote server are returned as part of bounce messages. This was not
1892 happening for LMTP over a pipe (the lmtp transport), but now it is the
1893 same for both kinds of LMTP.
1895 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1896 in header lines that were in fact CNAMEs. This is no longer the case.
1898 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1899 and queue runs started by the daemon processed all messages. This has
1900 been fixed so that -R and -S can now usefully be given with -q<time>.
1902 PH/40 Import PCRE release 6.7 (fixes some bugs).
1904 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1906 PH/42 Give an error if -q is specified more than once.
1908 PH/43 Renamed the variables $interface_address and $interface_port as
1909 $received_ip_address and $received_port, to make it clear that these
1910 values apply to message reception, and not to the outgoing interface when
1911 a message is delivered. (The old names remain recognized, of course.)
1913 PH/44 There was no timeout on the connect() call when using a Unix domain
1914 socket in the ${readsocket expansion. There now is.
1916 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1917 be meaningful with "accept".
1919 SC/01 Eximstats V1.43
1920 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1922 SC/02 Eximstats V1.44
1923 Use a glob alias rather than an array ref in the generated
1924 parser. This improves both readability and performance.
1926 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1927 Collect SpamAssassin and rejection statistics.
1928 Don't display local sender or destination tables unless
1929 there is data to show.
1930 Added average volumes into the top table text output.
1932 SC/04 Eximstats V1.46
1933 Collect data on the number of addresses (recipients)
1934 as well as the number of messages.
1936 SC/05 Eximstats V1.47
1937 Added 'Message too big' to the list of mail rejection
1938 reasons (thanks to Marco Gaiarin).
1940 SC/06 Eximstats V1.48
1941 Mainlog lines which have GMT offsets and are too short to
1942 have a flag are now skipped.
1944 SC/07 Eximstats V1.49 (Alain Williams)
1945 Added the -emptyok flag.
1947 SC/08 Eximstats V1.50
1948 Fixes for obtaining the IP address from reject messages.
1950 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1951 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1952 whitesspace changes from 4.64-PH/27
1954 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1957 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1958 are found, allow negative numbers in numeric criteria)
1960 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1962 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1963 to match changes made in 4.64-PH/43
1965 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1967 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1968 "domain" is actually an IP address, and force a failure. This locks out
1969 those revolvers/nameservers that support "A-for-A" lookups, in
1970 contravention of the specifications.
1972 PH/48 When a host name was looked up from an IP address, and the subsequent
1973 forward lookup of the name timed out, the host name was left in
1974 $sender_host_name, contrary to the specification.
1976 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1977 restricted to single-key lookups, Exim was not diagnosing an error if
1978 * or *@ was used with a query-style lookup.
1980 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1982 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1983 long ago, but noone interested enough thought of it. Let's just say that
1984 the "1.1" means that there are some new functions that weren't there at
1985 some point in the past.
1987 PH/51 Error processing for expansion failure of helo_data from an smtp
1988 transport during callout processing was broken.
1990 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1991 tested/used via the -bh/-bhc/-bs options.
1993 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1994 bug, fixed in subsequent PCRE releases).
1996 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1997 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1999 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2005 SC/01 Use a glob alias rather than an array ref in eximstats generated
2006 parser. This improves both readability and performance.
2008 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2009 Don't display local sender or destination tables in eximstats unless
2010 there is data to show.
2011 Added average volumes into the eximstats top table text output.
2013 SC/03 Collect data on the number of addresses (recipients) as well
2014 as the number of messages in eximstats.
2016 TF/01 Correct an error in the documentation for the redirect router. Exim
2017 does (usually) call initgroups() when daemonizing.
2019 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2020 with consistent privilege compared to when running as a daemon.
2022 TF/03 Note in the spec that $authenticated_id is not set for local
2023 submissions from trusted users.
2025 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2026 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2028 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2029 by adding some example configuration directives to the default
2030 configuration file. A little bit of work is required to uncomment the
2031 directives and define how usernames and passwords are checked, but
2032 there is now a framework to start from.
2034 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2035 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2036 without this. I don't know how relevant this is to other LDAP libraries.
2038 PH/02 Add the verb name to the "unknown ACL verb" error.
2040 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2042 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2044 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2045 directory not expanded when it should be if an expanded home directory
2046 was set for the address (which is overridden by the transport).
2048 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2051 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2052 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2053 because it is too late at that time, and has no effect.
2055 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2056 security issue with \' (bugzilla #107). I could not use the
2057 PQescapeStringConn() function, because it needs a PGconn value as one of
2060 PH/08 When testing addresses using -bt, indicate those final addresses that
2061 are duplicates that would not cause an additional delivery. At least one
2062 person was confused, thinking that -bt output corresponded to deliveries.
2063 (Suppressing duplicates isn't a good idea as you lose the information
2064 about possibly different redirections that led to the duplicates.)
2066 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2067 systems where poll() doesn't work, in particular OS X.
2069 PH/10 Added more information to debugging output for retry time not reached.
2071 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2072 operations in malware.c.
2074 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2077 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2078 syslog_duplication set false, log lines that would normally be written
2079 both the the main log and to the reject log were not written to syslog at
2082 PH/14 In the default configuration, change the use of "message" in ACL warn
2083 statements to "add_header".
2085 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2086 not followed by a command (e.g. "seen endif").
2088 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2089 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2092 PH/17 Added extra conditions to the default value of delay_warning_condition
2096 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2097 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2098 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2101 The Auto-Submitted: and various List- headers are standardised, whereas I
2102 don't think Precedence: ever was.
2104 PH/18 Refactored debugging code in route_finduser() to show more information,
2105 in particular, the error code if getpwnam() issues one.
2107 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2108 This is apparently needed in addition to the PH/07 change above to avoid
2109 any possible encoding problems.
2111 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2112 but not after initializing Perl.
2114 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2115 output them only if debugging. By default they are written stderr,
2116 apparently, which is not desirable.
2118 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2121 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2124 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2126 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2127 authenticated or an ident call has been made. Suppress the default
2128 values for $authenticated_id and $authenticated_sender (but permit -oMai
2129 and -oMas) when testing with -bh.
2131 PH/24 Re-jigged the order of the tests in the default configuration so that the
2132 tests for valid domains and recipients precede the DNS black list and CSA
2133 tests, on the grounds that those ones are more expensive.
2135 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2136 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2137 This bug exists in every version of Exim that I still have, right back to
2140 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2141 However, an attempt to turn on case-sensitivity in a regex key by
2142 including (?-i) didn't work because the subject string was already
2143 lowercased, and the effects were non-intuitive. It turns out that a
2144 one-line patch can be used to allow (?-i) to work as expected.
2150 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2151 other effects) broke the use of negated acl sub-conditions.
2153 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2156 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2157 "Deny" causes Exim to reject the incoming connection with a 554 error.
2158 Unfortunately, if there is a major crisis, such as a disk failure,
2159 tcp-wrappers gives "deny", whereas what one would like would be some
2160 kind of temporary error. A kludge has been added to help with this.
2161 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2162 554 error is used if errno is still zero or contains ENOENT (which occurs
2163 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2166 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2168 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2169 errors. Otherwise a message that provokes a temporary error (when other
2170 messages do not) can cause a whole host to time out.
2172 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2173 addresses were routed directly to files or pipes from a redirect router.
2174 File deliveries just didn't batch; pipe deliveries might have suffered
2177 PH/06 A failure to get a lock for a hints database would erroneously always say
2178 "Failed to get write lock", even when it was really a read lock.
2180 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2181 of 0600. This has been changed to use the value of the lockfile_mode
2182 option (which defaults to 0600).
2184 PH/08 Applied small patch from the Sieve maintainer.
2186 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2187 folder from quota calculations, a direct delivery into this folder messed
2188 up the contents of the maildirsize file. This was because the regex was
2189 used only to exclude .Trash (or whatever) when the size of the mailbox
2190 was calculated. There was no check that a delivery was happening into an
2191 excluded directory. This bug has been fixed by ignoring all quota
2192 processing for deliveries into excluded directories.
2194 PH/10 Added the maildirfolder_create_regex option to appendfile.
2200 PH/01 The code for finding all the local interface addresses on a FreeBSD
2201 system running IPv6 was broken. This may well have applied to all BSD
2202 systems, as well as to others that have similar system calls. The broken
2203 code found IPv4 interfaces correctly, but gave incorrect values for the
2204 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2205 that it would not match correctly against @[] and not recognize the IPv6
2208 PH/02 The ipliteral router was not recognizing addresses of the form user@
2209 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2211 PH/03 Added disable_ipv6.
2213 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2214 decoded form, because it is most often used to construct To: headers
2215 lines in autoreplies, and the decoded form may well be syntactically
2216 invalid. However, $reply_address has leading white space removed, and all
2217 newlines turned into spaces so that the autoreply transport does not
2220 PH/05 If group was specified without a user on a router, and no group or user
2221 was specified on a transport, the group from the router was ignored.
2223 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2224 for visible compile-time settings that can be used to change these
2225 numbers, for those that want even more. Backwards compatibility with old
2226 spool files has been maintained. However, going back to a previous Exim
2227 release will lost any variables that are in spool files.
2229 PH/07 Two small changes when running in the test harness: increase delay when
2230 passing a TCP/IP connection to a new process, in case the original
2231 process has to generate a bounce, and remove special handling of
2232 127.0.0.2 (sic), which is no longer necessary.
2234 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2235 be the same on different OS.
2237 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2240 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2241 whether --show-vars was specified or not
2243 JJ/02 exipick: Added support for new ACL variable spool format introduced
2246 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2247 syntactically invalid From: or Reply-to: line, and a filter used this to
2248 generate an autoreply, and therefore failed to obtain an address for the
2249 autoreply, Exim could try to deliver to a non-existent relative file
2250 name, causing unrelated and misleading errors. What now happens is that
2251 it logs this as a hard delivery error, but does not attempt to create a
2254 PH/11 The exinext utility has a -C option for testing purposes, but although
2255 the given file was scanned by exinext itself; it wasn't being passed on
2256 when Exim was called.
2258 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2259 an end-of-file indication when reading a command response.
2261 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2262 compiled. In many other places in Exim, IPv6 addresses are always
2263 recognized, so I have changed this. It also means that IPv4 domain
2264 literals of the form [IPV4:n.n.n.n] are now always recognized.
2266 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2267 used if the router is not running as root, for example, when verifying at
2268 ACL time, or when using -bh. The debugging output from this situation was
2269 non-existent - all you got was a failure to exec. I have made two
2272 (a) Failures to set uid/gid, the current directory, or a process leader
2273 in a subprocess such as that created by queryprogram now generate
2274 suitable debugging ouput when -d is set.
2276 (b) The queryprogram router detects when it is not running as root,
2277 outputs suitable debugging information if -d is set, and then runs
2278 the subprocess without attempting to change uid/gid.
2280 PH/15 Minor change to Makefile for building test_host (undocumented testing
2283 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2284 additional section of a DNS packet that returns MX or SRV records.
2285 Instead, it always explicitly searches for A/AAAA records. This avoids
2286 major problems that occur when a DNS server includes only records of one
2287 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2288 fixed another bug: if SRV records were looked up and the corresponding
2289 address records were *not* found in the additional section, the port
2290 values from the SRV records were lost.
2292 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2293 using the correct key (the original address) when searching the retry
2294 rules in order to find which one to use for generating the retry hint.
2296 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2297 adding the default one. Similarly, if it contains a Reply-To: header, the
2298 errors_reply_to option, if set, is not used.
2300 PH/19 When calculating a retry time, Exim used to measure the "time since
2301 failure" by looking at the "first failed" field in the retry record. Now
2302 it does not use this if it is later than than the arrival time of the
2303 message. Instead it uses the arrival time. This makes for better
2304 behaviour in cases where some deliveries succeed, thus re-setting the
2305 "first failed" field. An example is a quota failure for a huge message
2306 when small messages continue to be delivered. Without this change, the
2307 "time since failure" will always be short, possible causing more frequent
2308 delivery attempts for the huge message than are intended.
2309 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2311 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2312 $1, $2, $3) because the numerical variables can be reset during some
2313 expansion items (e.g. "match"), thereby losing the authentication data.
2315 PH/21 Make -bV show the size of off_t variables so that the test suite can
2316 decide whether to run tests for quotas > 2G.
2318 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2319 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2320 filecount value is greater than 2G or if a quota value is greater than 2G
2321 on a system where the size of off_t is not greater than 4, a panic error
2324 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2325 never match. The debug and -bh output now contains an explicit error
2326 message indicating a malformed IPv4 address or mask.
2328 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2329 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2330 PH/23 above applies.
2332 PH/25 Do not write to syslog when running in the test harness. The only
2333 occasion when this arises is a failure to open the main or panic logs
2334 (for which there is an explicit test).
2336 PH/26 Added the /no_tell option to "control=freeze".
2338 PH/27 If a host name lookup failed very early in a connection, for example, if
2339 the IP address matched host_lookup and the reverse lookup yielded a name
2340 that did not have a forward lookup, an error message of the form "no IP
2341 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2342 could be logged. Now it outputs the IP address instead of "NULL".
2344 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2345 allows the sender and the authenticated sender to be set when
2346 submitting a message from within Exim. Since child_open_exim() is
2347 documented for local_scan(), the new function should be too.
2349 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2350 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2351 results in an empty string is now treated as unset.
2353 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2355 PH/31 Added sender_verify_fail logging option.
2357 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2358 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2359 tidied the source and removed it altogether.
2361 PH/33 When a queue run was abandoned because the load average was too high, a
2362 log line was always written; now it is written only if the queue_run log
2363 selector is set. In addition, the log line for abandonment now contains
2364 information about the queue run such as the pid. This is always present
2365 in "start" and "stop" lines but was omitted from the "abandon" line.
2367 PH/34 Omit spaces between a header name and the colon in the error message that
2368 is given when verify = headers_syntax fails (if there are lots of them,
2369 the message gets confusing).
2371 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2372 names, as there are now some PTR records that contain slashes. This check
2373 is only to protect against broken name servers that fall over on strange
2374 characters, so the fact that it applies to all lookups doesn't matter.
2376 PH/36 Now that the new test suite is complete, we can remove some of the
2377 special code in Exim that was needed for the old test suite. For example,
2378 sorting DNS records because real resolvers return them in an arbitrary
2379 order. The new test suite's fake resolver always returns records in the
2382 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2383 bounces) except when queue_only is set, to avoid logging races between
2384 the different processes.
2386 PH/38 Panic-die if .include specifies a non-absolute path.
2388 PH/39 A tweak to the "H" retry rule from its user.
2390 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2391 a label. They prevented compilation on older perls.
2393 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2394 a warning to be raised on newish perls.
2396 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2397 on queue. Changes to match documented behaviour of showing count of
2398 messages matching specified criteria.
2400 PH/40 Changed the default ident timeout from 30s to 5s.
2402 PH/41 Added support for the use of login_cap features, on those BSD systems
2403 that have them, for controlling the resources used by pipe deliveries.
2405 PH/42 The content-scanning code uses fopen() to create files in which to put
2406 message data. Previously it was not paying any attention to the mode of
2407 the files. Exim runs with umask(0) because the rest of the code creates
2408 files with open(), and sets the required mode explicitly. Thus, these
2409 files were ending up world-writeable. This was not a big issue, because,
2410 being within the spool directory, they were not world-accessible. I have
2411 created a function called modefopen, which takes an additional mode
2412 argument. It sets umask(777), creates the file, chmods it to the required
2413 mode, then resets the umask. All the relevant calls to fopen() in the
2414 content scanning code have been changed to use this function.
2416 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2417 to 24 hours. This avoids potential overflow problems when processing G
2418 and H retry rules. I suspect nobody ever tinkers with this value.
2420 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2422 PH/45 When the plaintext authenticator is running as a client, the server's
2423 challenges are checked to ensure they are valid base64 strings. By
2424 default, the authentication attempt is cancelled if an invalid string is
2425 received. Setting client_ignore_invalid_base64 true ignores these errors.
2426 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2427 they are received. Thus, the responses can be made to depend on the
2428 challenges. If an invalid string is ignored, an empty string is placed in
2431 PH/46 Messages that are created by the autoreply transport now contains a
2432 References: header, in accordance with RFCs 2822 and 3834.
2434 PH/47 Added authenticated_sender_force to the smtp transport.
2436 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2438 PH/49 Installed latest patch from the Sieve maintainer.
2440 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2441 was also set, the appendfile transport was unnecessarily scanning a
2442 directory of message files (e.g. for maildir delivery) to find the count
2443 of files (along with the size), even though it did not need this
2444 information. It now does the scan only if it needs to find either the
2445 size of the count of files.
2447 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2449 PH/52 Two bugs concerned with error handling when the smtp transport is
2452 (i) Exim was not creating retry information for temporary errors given
2453 for individual recipients after the DATA command when the smtp transport
2454 was used in LMTP mode. This meant that they could be retried too
2455 frequently, and not timed out correctly.
2457 (ii) Exim was setting the flag that allows error details to be returned
2458 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2459 recipients that were returned after the DATA command.
2461 PH/53 This is related to PH/52, but is more general: for any failing address,
2462 when detailed error information was permitted to be returned to the
2463 sender, but the error was temporary, then after the final timeout, only
2464 "retry timeout exceeded" was returned. Now it returns the full error as
2465 well as "retry timeout exceeded".
2467 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2468 do this, and (what is worse) MTAs that accept it.
2470 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2471 will now be deprecated.
2473 PH/56 New os.c-cygwin from the Cygwin maintainer.
2475 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2476 formats (previously only available in exim formats via -bpr, -bpru,
2477 and -bpra. Now also available in native and exiqgrep formats)
2479 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2480 with very large, slow to parse queues
2482 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2484 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2486 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2487 responses to authentication challenges, though it was showing the
2488 challenges; (ii) I've removed the CR characters from the debug output for
2491 PH/58 Allow for the insertion of a newline as well as a space when a string
2492 is turned into more than one encoded-word during RFC 2047 encoding. The
2493 Sieve code now uses this.
2495 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2496 data_4xx, lost_connection, tls_required.
2498 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2499 message was being sent as an SMTP response.
2501 PH/61 Add -l and -k options to exicyclog.
2503 PH/62 When verifying, if an address was redirected to one new address, so that
2504 verification continued, and the new address failed or deferred after
2505 having set something in $address_data, the value of $address_data was not
2506 passed back to the ACL. This was different to the case when no
2507 redirection occurred. The value is now passed back in both cases.
2509 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2510 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2511 use login_cap.h, so on its own it isn't the distinguishing feature. The
2512 new name refers directly to the setclassresources() function.
2514 PH/65 Added configuration files for NetBSD3.
2516 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2518 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2519 is preferred over IPv4.
2521 PH/68 The bounce_return_message and bounce_return_body options were not being
2522 honoured for bounces generated during the reception of non-SMTP messages.
2523 In particular, this applied to messages rejected by the ACL. This bug has
2524 been fixed. However, if bounce_return_message is true and bounce_return_
2525 body is false, the headers that are returned for a non-SMTP message
2526 include only those that have been read before the error was detected.
2527 (In the case of an ACL rejection, they have all been read.)
2529 PH/69 The HTML version of the specification is now built in a directory called
2530 spec_html instead of spec.html, because the latter looks like a path with
2531 a MIME-type, and this confuses some software.
2533 PH/70 Catch two compiler warnings in sieve.c.
2535 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2536 function verify_get_ident() calls ip_connect() to connect a socket, but
2537 if the "connect()" function timed out, ip_connect() used to close the
2538 socket. However, verify_get_ident() also closes the socket later, and in
2539 between Exim writes to the log, which may get opened at this point. When
2540 the socket was closed in ip_connect(), the log could get the same file
2541 descriptor number as the socket. This naturally causes chaos. The fix is
2542 not to close the socket in ip_connect(); the socket should be closed by
2543 the function that creates it. There was only one place in the code where
2544 this was missing, in the iplookup router, which I don't think anybody now
2545 uses, but I've fixed it anyway.
2547 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2548 well as to direct DNS lookups. Otherwise the handling of names in host
2549 lists is inconsistent and therefore confusing.
2555 PH/01 Two changes to the default runtime configuration:
2557 (1) Move the checks for relay_from_hosts and authenticated clients from
2558 after to before the (commented out) DNS black list checks.
2560 (2) Add control=submission to the relay_from_hosts and authenticated
2561 clients checks, on the grounds that messages accepted by these
2562 statements are most likely to be submissions.
2564 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2566 (1) Generate an error if the third argument for the ${prvs expansion is
2569 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2572 (3) Reset the variables that are obtained from the first argument of
2573 ${prvscheck and used in the second argument before leaving the code,
2574 because their memory is reclaimed, so using them afterwards may do
2577 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2578 one (it's much easier than Tom thought :-).
2580 (5) Because of (4), we can now allow for the use of $prvscheck_result
2581 inside the third argument.
2583 PH/03 For some reason, the default setting of PATH when running a command from
2584 a pipe transport was just "/usr/bin". I have changed it to
2587 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2588 anything to be listed in the output from -bV.
2590 PH/05 When a filter generated an autoreply, the entire To: header line was
2591 quoted in the delivery log line, like this:
2593 => >A.N.Other <ano@some.domain> <original@ddress> ...
2595 This has been changed so that it extracts the operative address. There
2596 may be more than one such address. If so, they are comma-separated, like
2599 => >ano@some.domain,ona@other.domain <original@ddress> ...
2601 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2602 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2603 not being looked up in the rDNS to get a host name, Exim was showing the
2604 IP address twice in Received: lines, even though the IP addresses were
2605 identical. For example:
2607 Received: from [1.2.3.4] (helo=[1.2.3.4])
2609 However, if the real host name was known, it was omitting the HELO data
2610 if it matched the actual IP address. This has been tidied up so that it
2611 doesn't show the same IP address twice.
2613 PH/07 When both +timestamp and +memory debugging was on, the value given by
2614 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2615 called by the debug printing, thereby overwriting the timestamp buffer.
2616 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2618 PH/08 When the original message was included in an autoreply transport, it
2619 always said "this is a copy of the message, including all the headers",
2620 even if body_only or headers_only was set. It now gives an appropriate
2623 PH/09 Applied a patch from the Sieve maintainer which:
2625 o fixes some comments
2626 o adds the (disabled) notify extension core
2627 o adds some debug output for the result of if/elsif tests
2628 o points to the current vacation draft in the documentation
2629 and documents the missing references header update
2633 o fixes a bug in processing the envelope test (when testing
2634 multiple envelope elements, the last element determinted the
2637 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2638 Electronic Mail") by including:
2640 Auto-submitted: auto-generated
2642 in the messages that it generates (bounce messages and others, such as
2643 warnings). In the case of bounce messages for non-SMTP mesages, there was
2644 also a typo: it was using "Auto_submitted" (underscore instead of
2645 hyphen). Since every message generated by Exim is necessarily in response
2646 to another message, thes have all been changed to:
2648 Auto-Submitted: auto-replied
2650 in accordance with these statements in the RFC:
2652 The auto-replied keyword:
2654 - SHOULD be used on messages sent in direct response to another
2655 message by an automatic process,
2657 - MUST NOT be used on manually-generated messages,
2659 - MAY be used on Delivery Status Notifications (DSNs) and Message
2660 Disposition Notifications (MDNs),
2662 - MUST NOT be used on messages generated by automatic or periodic
2663 processes, except for messages which are automatic responses to
2666 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2667 to the default Received: header definition.
2669 PH/12 Added log selector acl_warn_skipped (default on).
2671 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2672 variables because (a) they are in the wrong storage pool and (b) even if
2673 they were copied, it wouldn't work properly because of the caching.
2675 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2676 checking when decoding. Apparently there are clients that generate
2677 overlong encoded strings. Why am I not surprised?
2679 PH/15 If the first argument of "${if match_address" was not empty, but did not
2680 contain an "@" character, Exim crashed. Now it writes a panic log message
2681 and treats the condition as false.
2683 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2685 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2686 extension "envelope-auth". The code is finished and in agreement with
2687 other implementations, but there is no documentation so far and in fact,
2688 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2689 not changing the active code.
2691 Print executed "if" and "elsif" statements when debugging is used. This
2692 helps a great deal to understand what a filter does.
2694 Document more things not specified clearly in RFC3028. I had all this
2695 sorted out, when out of a sudden new issues came to my mind. Oops."
2697 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2700 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2701 canonical form (as documented). However, after a host name lookup from
2702 the IP address, check_host() was doing a simple string comparison with
2703 addresses acquired from the DNS when checking that the found name did
2704 have the original IP as one of its addresses. Since any found IPv6
2705 addresses are likely to be in abbreviated form, the comparison could
2706 fail. Luckily, there already exists a function for doing the comparison
2707 by converting both addresses to binary, so now that is used instead of
2708 the text comparison.
2710 PH/20 There was another similar case to PH/19, when a complete host name was
2711 given in a host list; looking up its IP address could give an abbreviated
2712 form, whereas the current host's name might or might not be abbreviated.
2713 The same fix has been applied.
2719 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2720 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2723 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2724 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2726 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2728 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2729 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2730 Also, for "Auto-Submitted", treat anything other than "no" as
2731 non-personal, in accordance with RFC 3834. (Previously it treated
2732 anything starting "auto-" as non-personal.)
2734 TF/01 The control=submission/name=... option had a problem with syntax
2735 errors if the name included a slash character. The /name= option
2736 now slurps the rest of the string, so it can include any characters
2737 but it must come last in the list of options (after /sender_retain
2740 PH/05 Some modifications to the interface to the fake nameserver for the new
2748 TK/01 Added the "success_on_redirect" address verification option. See
2749 NewStuff for rationale and an example.
2751 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2753 PH/02 Patch to exigrep to allow it to work on syslog lines.
2755 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2756 fread() to skip over the body file's header line, because in Cygwin the
2757 header line is locked and is inaccessible.
2759 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2760 co-exist for some time) to make it clear that it is the Exim ID that is
2761 referenced, not the Message-ID: header line.
2763 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2764 string_format() function, because snprintf() does not exist on all
2767 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2768 ${stat: expansion item.
2770 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2771 protocol synchronization error", to keep the pedants happy.
2773 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2774 well as for IRIX systems, when gcc is being used. See the host.c source
2777 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2779 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2782 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2783 transport and to the smtp transport in LMTP mode.
2785 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2787 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2788 run a filter in a subprocess. This could lead to confusion in subsequent
2789 lookups in the parent process. There should also be a search_tidyup() at
2790 the end of the subprocess.
2792 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2793 only if the host matched helo_try_verify_hosts, which caused the
2794 verification to occur when the EHLO/HELO command was issued. The ACL just
2795 tested the remembered result. Now, if a previous verification attempt has
2796 not happened, "verify = helo" does it there and then.
2798 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2800 TK/03 Fix log output including CR from clamd.
2802 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2803 exist provoked a memory error which could cause a segfault.
2805 PH/15 Installed PCRE 6.2
2807 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2809 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2810 of the problem. Specifically, suggested +O2 rather than +O1 for the
2813 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2815 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2816 record, the port was not correctly logged when the outgoing_port log
2817 selector was set (it logged the transort's default port).
2819 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2820 fallback_hosts, and "hosts" in the smtp transport.
2822 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2823 host errors such as "Connection refused".
2825 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2826 authentication with radiusclient 0.4.9:
2828 - Error returned from rc_read_config was caught wrongly
2829 - Username/password not passed on to radius server due to wrong length.
2831 The presumption is that some radiusclient API changes for 4.51/PH/17
2832 were not taken care of correctly. The code is still untested by me (my
2833 Linux distribution still has 0.3.2 of radiusclient), but it was
2834 contributed by a Radius user.
2836 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2837 expanding the "port" option of the smtp transport.
2839 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2840 while reading a MIME header. Thanks to Tom Hughes for a patch.
2842 PH/24 Include config.h inside local_scan.h so that configuration settings are
2845 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2846 that in an ACL for RCPT (for example), you can examine exactly what was
2849 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2850 commands, but it was not correctly comparing the address with the actual
2851 client host address. Thus, it would show the EHLO address in Received:
2852 header lines when this was not necessary.
2854 PH/27 Added the % operator to ${eval:}.
2856 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2857 it should be ignoring failures (because with -C, for example, it has lost
2858 privilege). It wasn't ignoring creation failures other than "already
2861 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2864 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2865 because an input line was too long, either on its own, or by virtue of
2866 too many continuations, the temporary file was not being removed, and the
2867 return code was incorrect.
2869 PH/31 Missing "BOOL" in function definition in filtertest.c.
2871 PH/32 Applied Sieve patches from the maintainer.
2873 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2875 PH/33 Added "verify = not_blind".
2877 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2878 Local/Makefile (with some defaults set). These are used in built scripts
2879 such as exicyclog, but they have never been used in the exim_install
2880 script (though there are many overriding facilities there). I have
2881 arranged that the exim_install script now takes note of these two
2884 PH/35 Installed configuration files for Dragonfly.
2886 PH/36 When a locally submitted message by a trusted user did not contain a
2887 From: header, and the sender address was obtained from -f or from an SMTP
2888 MAIL command, and the trusted user did not use -F to supply a sender
2889 name, $originator_name was incorrectly used when constructing a From:
2890 header. Furthermore, $originator_name was used for submission mode
2891 messages from external hosts without From: headers in a similar way,
2892 which is clearly wrong.
2894 PH/37 Added control=suppress_local_fixups.
2896 PH/38 When log_selector = +received_sender was set, and the addition of the
2897 sender made the log line's construction buffer exactly full, or one byte
2898 less than full, an overflow happened when the terminating "\n" was
2901 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2902 when the result of a list match is failure because a DNS lookup failed.
2904 PH/40 RM_COMMAND is now used in the building process.
2906 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2907 the "build-* directories that it finds.
2909 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2910 address in a domain literal was a prefix of an interface address.
2912 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2913 when verifying a sender address, unless rewrite_headers is false.
2915 PH/44 Wrote a long comment about why errors_to addresses are verified as
2916 recipients, not senders.
2918 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2919 the ratelimit ACL was added.
2921 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2923 PH/46 Added extra information about PostgreSQL errors to the error string.
2925 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2926 avoiding the need to install special zones in a real server. This is
2927 backwards compatible; if it can't find the fake resolver, it drops back.
2928 Thus, both old and new test suites can be run.
2930 TF/02 Added util/ratelimit.pl
2932 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2933 clock is set back in time.
2935 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2936 Candler <B.Candler@pobox.com>.
2938 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2939 OK for addresses that are the result of redirections.
2941 PH/48 A number of further additions for the benefit of the new test suite,
2942 including a fake gethostbyname() that interfaces to the fake DNS resolver
2945 TF/06 The fix for widen_domains has also been applied to qualify_single and
2946 search_parents which are the other dnslookup options that can cause
2949 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2952 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2954 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2955 service" if it encountered an MX record with an empty target hostname.
2956 The message is now "an MX or SRV record indicated no SMTP service".
2958 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2959 if the DNS of the sending site is misconfigured. This is quite a
2960 common situation. This change restores the behaviour of treating a
2961 helo verification defer as a failure.
2963 PH/51 If self=fail was set on a router, the bounce message did not include the
2964 actual error message.
2970 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2972 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2973 command itself ended in error, the underlying message about the transport
2974 filter timeout was being overwritten with the pipe command error. Now the
2975 underlying error message should be appended to the second error message.
2977 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2979 PH/02 Reduce the amount of output that "make" produces by default. Full output
2980 can still be requested.
2982 PH/03 The warning log line about a condition test deferring for a "warn" verb
2983 was being output only once per connection, rather than after each
2984 occurrence (because it was using the same function as for successful
2985 "warn" verbs). This seems wrong, so I have changed it.
2987 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2988 it should not have, which might have caused a crash in the right
2989 circumstances, but probably never did.
2991 PH/04 Installed a modified version of Tony Finch's patch to make submission
2992 mode fix the return path as well as the Sender: header line, and to
2993 add a /name= option so that you can make the user's friendly name appear
2996 TF/03 Added the control = fakedefer ACL modifier.
2998 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2999 Mark Lowes for thorough testing.
3001 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3003 TK/03 Merged latest SRS patch from Miles Wilton.
3005 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3006 with the definition in sysexits.h (which is #included earlier).
3007 Fortunately, Exim does not actually use EX_OK. The code used to try to
3008 preserve the sysexits.h value, by assumimg that macro definitions were
3009 scanned for macro replacements. I have been disabused of this notion,
3010 so now the code just undefines EX_OK before #including unistd.h.
3012 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3013 in the smtp transport. When a block could not be written in a single
3014 write() function, the timeout was being re-applied to each part-write.
3015 This seems wrong - if the receiver was accepting one byte at a time it
3016 would take for ever. The timeout is now adjusted when this happens. It
3017 doesn't have to be particularly precise.
3019 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3020 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3022 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3023 without a domain if the check to <postmaster@domain> fails.
3025 SC/01 Eximstats: added -xls and the ability to specify output files
3026 (patch written by Frank Heydlauf).
3028 SC/02 Eximstats: use FileHandles for outputing results.
3030 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3032 SC/04 Eximstats: fixed display of large numbers with -nvr option
3034 SC/05 Eximstats: fixed merging of reports with empty tables.
3036 SC/06 Eximstats: added the -include_original_destination flag
3038 SC/07 Eximstats: removed tabs and trailing whitespace.
3040 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3042 TK/06 MBOX spool code: Add real "From " MBOX separator line
3043 so the .eml file is really in mbox format (even though
3044 most programs do not really care). Patch from Alex Miller.
3046 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3047 The latter is generated from $received_to and is only set if the
3048 message has one envelope recipient. SA can use these headers,
3049 obviously out-of-the-box. Patch from Alex Miller.
3051 PH/08 The ${def test on a variable was returning false if the variable's
3052 value was "0", contrary to what the specification has always said!
3053 The result should be true unless the variable is empty.
3055 PH/09 The syntax error of a character other than { following "${if
3056 def:variable_name" (after optional whitespace) was not being diagnosed.
3057 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3058 accidental colon was present, for example, could give incorrect results.
3060 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3061 result is used (not including appendfile, where other changes are about
3064 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3065 This involved changing a lot of size variables from int to off_t. It
3066 should work with maildirs and everything.
3068 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3069 spamd dying while we are connected to it.
3071 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3074 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3075 function for BDB 4.3.
3077 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3079 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3080 the sources for PCRE have been re-arranged and more files are now
3083 PH/15 The code I had for printing potentially long long variables in PH/11
3084 above was not the best (it lost precision). The length of off_t variables
3085 is now inspected at build time, and an appropriate printing format (%ld
3086 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3087 to be "long long int" or "long int". This is needed for the internal
3088 formatting function string_vformat().
3090 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3091 the configuration file to be ":syslog", then the script "guesses" where
3092 the logs files are, rather than using the compiled in default. In our
3093 case the guess is not the same as the compiled default, so the script
3094 suddenly stopped working when I started to use syslog. The patch checks
3095 to see if log_file_path is "". If so, it attempts to read it from exim
3096 with no configuration file to get the compiled in version, before it
3097 falls back to the previous guessing code."
3099 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3100 implementing BATV in an Exim configuration. See NewStuff for the gory
3103 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3104 Makefile that are specific to HP-UX.
3106 PH/18 If the "use_postmaster" option was set for a recipient callout together
3107 with the "random" option, the postmaster address was used as the MAIL
3108 FROM address for the random test, but not for the subsequent recipient
3109 test. It is now used for both.
3111 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3112 patch removes a few documentation additions to RFC 3028, because the
3113 latest draft now contains them. It adds the new en;ascii-case comparator
3114 and a new error check for 8bit text in MIME parts. Comparator and
3115 require names are now matched exactly. I enabled the subaddress
3116 extension, but it is not well tested yet (read: it works for me)."
3118 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3119 rework some of the code of TK/09 above to avoid the hardwired use of
3120 "%lld" and "long long". Replaced the call to snprintf() with a call to
3123 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3124 records point to non-existent hosts", "retry timeout exceeded", and
3125 "retry time not reached for any host after a long failure period".
3127 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3128 experimental DomainKeys support:
3130 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3131 (2) On an error such as an illegally used "control", the wrong name for
3132 the control was given.
3134 These problems did NOT occur unless DomainKeys support was compiled.
3136 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3138 PH/24 Added ${if match_ip condition.
3140 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3141 databases so that it will be absolutely obvious if a crash occurs in the
3142 DB library. This is a regular occurrence (often caused by mis-matched
3145 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3146 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3147 picked up on a user's system that detects such things. There doesn't seem
3148 to be a gcc warning option for this - only an attribute that has to be
3149 put on the function's prototype. It seems that in Fedora Core 4 they have
3150 set this on a number of new functions. No doubt there will be more in due
3153 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3154 specify a transport. However, if an address that was verified by such a
3155 router was the subject of a callout, Exim crashed because it tried to
3156 read the rcpt_include_affixes from the non-existent transport. Now it
3157 just assumes that the setting of that option is false. This bug was
3158 introduced by 4.51/PH/31.
3160 PH/28 Changed -d+all to exclude +memory, because that information is very
3161 rarely of interest, but it makes the output a lot bigger. People tend to
3162 do -d+all out of habit.
3164 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3165 code in os-type was giving problems when libc.so lives in lib64, like on
3168 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3169 aren't the modern standard, and it seems that some systems' include files
3170 don't always have them. Exim was already checking for some of the newer
3171 ones like T_AAAA, and defining it itself. I've added checks for all the
3172 record types that Exim uses.
3174 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3175 not automatically generating a new one, as it is supposed to. This
3176 prevented TLS from working. If the file did exist, but contained invalid
3177 data, a new version was generated, as expected. It was only the case of a
3178 non-existent file that was broken.
3180 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3181 with a change in libdomainkeys > 0.64.
3183 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3184 from DNS. If the selector record carries the flag, it now has
3185 precedence over the domain-wide flag.
3187 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3189 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3190 the use of a port name that isn't defined in /etc/services) occurred, the
3191 message was deferred as in a normal delivery, and thus remained on the
3192 spool, instead of being failed because of the mua_wrapper setting. This
3193 is now fixed, and I tidied up some of the mua_wrapper messages at the
3196 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3197 the messages in a hash of arrays rather than using individual hashes.
3198 This is a bit cleaner and results in dramatic memory savings, albeit
3199 at a slight CPU cost.
3201 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3202 as requested by Marc Sherman.
3204 SC/10 Eximstats: added histograms for user specified patterns as requested
3207 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3209 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3210 fopen() in the content-scanning modules that did not already have it.
3216 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3217 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3219 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3221 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3223 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3224 to test Sieve filters that use "vacation".
3226 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3227 that changes the way the GnuTLS parameters are stored in the cache file.
3228 The new format can be generated externally. For backward compatibility,
3229 if the data in the cache doesn't make sense, Exim assumes it has read an
3230 old-format file, and it generates new data and writes a new file. This
3231 means that you can't go back to an older release without removing the
3234 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3235 work if there are any delivery delays because "one_time" forces the
3236 parent to be marked "delivered", so its unseen clone is never tried
3237 again. For this reason, Exim now forbids the simultaneous setting of
3240 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3241 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3242 entry if you want to know the details. The fix, however, neglected to
3243 consider the case when local delivery batching is involved. The test for
3244 "previously delivered" was not happening when checking to see if an
3245 address could be batched with a previous (undelivered) one; under
3246 certain circumstances this could lead to multiple deliveries to the same
3249 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3250 in its include files, and this causes problems building Exim.
3252 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3253 header_syntax) but Exim was just ignoring anything given after a slash.
3254 In particular, this caused confusion with an attempt to use "verify =
3255 reverse_host_lookup/defer_ok". An error is now given when options are
3256 supplied for verify items that do not have them. (Maybe reverse_host_
3257 lookup should have a defer_ok option, but that's a different point.)
3259 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3260 defined by RFC 821) to 2048, because there were problems with some AUTH
3261 commands, and RFC 1869 says the size should be increased for extended
3262 SMTP commands that take arguments.
3264 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3267 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3268 "unknown" error; now it says that the functionality isn't in the binary.
3270 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3271 an address' error message when a string expansion fails (syntax or
3272 whatever). Otherwise the password may appear in the log. Following change
3273 PH/42 below, there is no longer a chance of it appearing in a bounce
3276 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3278 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3279 of its addresses was ever tried. (Bugzilla bug #2.)
3281 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3282 the result incorrectly in the debug output. (It correctly added a newline
3283 to what was transported.)
3285 TF/01 Added $received_time.
3287 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3288 commented out examples of how to interface to a virus scanner and to
3289 SpamAssassin. Also added commented examples of av_scanner and
3290 spamd_address settings.
3292 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3293 and controls are allowed in which ACLs. There were a couple of minor
3294 errors. Some of the entries in the conditions table (which is a table of
3295 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3296 negation of where the condition IS allowed.
3298 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3300 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3301 header file does not have a version number, so I've had to invent a new
3302 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3303 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3304 radiusclient), but it was contributed by a Radius user.
3306 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3307 files or directories, for OpenSSL.
3309 PH/19 When an Exim process that is running as root has to create an Exim log
3310 file, it does so in a subprocess that runs as exim:exim so as to get the
3311 ownership right at creation (otherwise, other Exim processes might see
3312 the file with the wrong ownership). There was no test for failure of this
3313 fork() call, which would lead to the process getting stuck as it waited
3314 for a non-existent subprocess. Forks do occasionally fail when resources
3315 run out. I reviewed all the other calls to fork(); they all seem to check
3318 PH/20 When checking for unexpected SMTP input at connect time (before writing
3319 the banner), Exim was not dealing correctly with a non-positive return
3320 from the read() function. If the client had disconnected by this time,
3321 the result was a log entry for a synchronization error with an empty
3322 string after "input=" when read() returned zero. If read() returned -1
3323 (an event I could not check), uninitialized data bytes were printed.
3324 There were reports of junk text (parts of files, etc) appearing after
3327 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3329 PH/22 Added support for macro redefinition, and (re)definition in between
3330 driver and ACL definitions.
3332 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3333 forgetting to use the resulting value; it was using the unexpanded value.
3335 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3336 hadn't been configured. The fix is from Juergen Kreileder, who
3337 understands it better than I do:
3339 "Here's what I see happening with three configured cyrus_sasl
3340 authenticators configured (plain, login, cram-md5):
3342 On startup auth_cyrus_sasl_init() gets called for each of these.
3343 This means three calls to sasl_listmech() without a specified mech_list.
3344 => SASL tests which mechs of all available mechs actually work
3345 => three warnings about OTP not working
3346 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3348 With the patch, sasl_listmech() also gets called three times. But now
3349 SASL's mech_list option is set to the server_mech specified in the the
3350 authenticator. Or in other words, the answer from sasl_listmech()
3351 gets limited to just the mech you're testing for (which is different
3353 => the return list contains just 'plain' or 'login', 'cram-md5' or
3354 nothing depending on the value of ob->server_mech.
3356 I've just tested the patch: Authentication still works fine,
3357 unavailable mechs specified in the exim configuration are still
3358 caught, and the auth.log warnings about OTP are gone."
3360 PH/25 When debugging is enabled, the contents of the command line are added
3361 to the debugging output, even when log_selector=+arguments is not
3364 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3365 answer is "GNU", and only if the return is "GNU/something" is the answer
3368 PH/27 $acl_verify_message is now set immediately after the failure of a
3369 verification in an ACL, and so is available in subsequent modifiers. In
3370 particular, the message can be preserved by coding like this:
3372 warn !verify = sender
3373 set acl_m0 = $acl_verify_message
3375 Previously, $acl_verify_message was set only while expanding "message"
3376 and "log_message" when a very denied access.
3378 PH/28 Modified OS/os.c-Linux with
3380 -#ifndef OS_LOAD_AVERAGE
3381 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3383 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3384 nomenclature these days.)
3386 PH/29 Installed patch from the Sieve maintainer that adds the options
3387 sieve_useraddress and sieve_subaddress to the redirect router.
3389 PH/30 In these circumstances:
3390 . Two addresses routed to the same list of hosts;
3391 . First host does not offer TLS;
3392 . First host accepts first address;
3393 . First host gives temporary error to second address;
3394 . Second host offers TLS and a TLS session is established;
3395 . Second host accepts second address.
3396 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3397 and peerdn, if requested) that were in fact used only for the second
3400 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3401 attention to any local part prefix or suffix that was matched by the
3402 router that accepted the address. It now behaves in the same way as it
3403 does for delivery: the affixes are removed from the local part unless
3404 rcpt_include_affixes is set on the transport.
3406 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3407 timeout during the DATA phase of an incoming message.
3409 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3410 applied a patch sanctioned by the Sieve maintainer.
3412 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3413 the uid or gid is negative. A case of a negative gid caused this to be
3414 noticed. The fix allows for either to be negative.
3416 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3417 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3420 PH/36 The change PH/12 above was broken. Fixed it.
3422 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3423 the grounds that routing the same address twice would always produce the
3424 same answer. This might have been true once, but it is certainly no
3425 longer true now. Routing a child address may depend on the previous
3426 routing that produced that child. Some complicated redirection strategies
3427 went wrong when messages had multiple recipients, and made Exim's
3428 behaviour dependent on the order in which the addresses were given.
3430 I have moved the duplicate checking until after the routing is complete.
3431 Exim scans the addresses that are assigned to local and remote
3432 transports, and removes any duplicates. This means that more work will be
3433 done, as duplicates will always all be routed, but duplicates are
3434 presumably rare, so I don't expect this is of any significance.
3436 For deliveries to pipes, files, and autoreplies, the duplicate checking
3437 still happens during the routing process, since they are not going to be
3440 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3441 It corrects a timeout issue with spamd. This is Ian's comment: "The
3442 background is that sometimes spamd either never reads data from a
3443 connection it has accepted, or it never writes response data. The exiscan
3444 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3445 blindly assumes that writes won't block so it may never time out."
3447 PH/39 Allow G after quota size as well as K and M.
3449 PH/40 The value set for $authenticated_id in an authenticator may not contain
3450 binary zeroes or newlines because the value is written to log lines and
3451 to spool files. There was no check on this. Now the value is run through
3452 the string_printing() function so that such characters are converted to
3453 printable escape sequences.
3455 PH/41 $message_linecount is a new variable that contains the total number of
3456 lines in the message. Compare $body_linecount, which is the count for the
3459 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3460 bounce and delay warning messages, except in certain special cases, which
3463 (a) An SMTP error message from a remote host;
3464 (b) A message specified in a :fail: redirection;
3465 (c) A message specified in a "fail" command in a system filter;
3466 (d) A message specified in a FAIL return from the queryprogram router;
3467 (e) A message specified by the cannot_route_message router option.
3469 In these cases only, Exim does include the error details in bounce and
3470 warning messages. There are also a few cases where bland messages such
3471 as "unrouteable address" or "local delivery error" are given.
3473 PH/43 $value is now also set for the "else" part of a ${run expansion.
3475 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3476 being worked on, but at least Exim now implements the latest version to
3479 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3480 process to complete was treated as a delivery failure, a timeout while
3481 writing the message to the pipe was logged, but erroneously treated as a
3482 successful delivery. Such timeouts include transport filter timeouts. For
3483 consistency with the overall process timeout, these timeouts are now
3484 treated as errors, giving rise to delivery failures by default. However,
3485 there is now a new Boolean option for the pipe transport called
3486 timeout_defer, which, if set TRUE, converts the failures into defers for
3487 both kinds of timeout. A transport filter timeout is now identified in
3490 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3491 systems where "make" and "gmake" are different, calling "gmake" at top
3492 level broke things. I've arranged for the value of $(MAKE) to be passed
3493 from the Makefile to this script so that it can call the same version of
3497 A note about Exim versions 4.44 and 4.50
3498 ----------------------------------------
3500 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3501 changes of various kinds. As a consequence, a big documentation update was
3502 needed. This delayed the release for rather longer than seemed good, especially
3503 in the light of a couple of (minor) security issues. Therefore, the changes
3504 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3505 release. So 4.44 and 4.50 are in effect two different branches that both start
3508 I have left the 4.50 change log unchanged; it contains all the changes since
3509 4.43. The change log for 4.44 is below; many of its items are identical to
3510 those for 4.50. This seems to be the most sensible way to preserve the
3511 historical information.
3517 1. Minor wording change to the doc/README.SIEVE file.
3519 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3520 computation of the current number of files was incorrect.
3522 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3523 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3526 4. Give more explanation in the error message when the command for a transport
3527 filter fails to execute.
3529 5. There are several places where Exim runs a non-Exim command in a
3530 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3531 was being done only for the command run by the queryprogram router. It is
3532 now done for all such subprocesses. The other cases are: ${run, transport
3533 filters, and the commands run by the lmtp and pipe transports.
3535 6. Added CONFIGURE_GROUP build-time option.
3537 7. Some older OS have a limit of 256 on the maximum number of file
3538 descriptors. Exim was using setrlimit() to set 1000 as a large value
3539 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3540 systems. I've change it so that if it can't get 1000, it tries for 256.
3542 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3543 was an oversight, and furthermore, ever since the addition of extra
3544 controls (e.g. 4.43/32), the checks on when to allow different forms of
3545 "control" were broken. There should now be diagnostics for all cases when a
3546 control that does not make sense is encountered.
3548 9. Added the /retain_sender option to "control=submission".
3550 10. $recipients is now available in the predata ACL (oversight).
3552 11. Tidy the search cache before the fork to do a delivery from a message
3553 received from the command line. Otherwise the child will trigger a lookup
3554 failure and thereby defer the delivery if it tries to use (for example) a
3555 cached ldap connection that the parent has called unbind on.
3557 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3558 of $address_data from the recipient verification was clobbered by the
3559 sender verification.
3561 13. The value of address_data from a sender verification is now available in
3562 $sender_address_data in subsequent conditions in the ACL statement.
3564 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3566 15. Added a new option "connect=<time>" to callout options, to set a different
3569 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3570 was its contents. (It was OK if the option was not defined at all.)
3572 17. A "Completed" log line is now written for messages that are removed from
3573 the spool by the -Mrm option.
3575 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3576 information about exactly what failed.
3578 19. Added -dd to debug only the daemon process.
3580 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3581 handles timeouts, both on the server side and network timeouts. Renamed the
3582 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3584 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3585 It is now set to "smtps".
3587 22. $host_address is now set to the target address during the checking of
3588 ignore_target_hosts.
3590 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3591 being passed; this would have caused $sender_host_name to have been used if
3592 matching the list had actually called for a host name (not very likely,
3593 since this list is usually IP addresses). A host name is now passed as
3596 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3597 code that specifies a non-restarting handler (typically sigaction() in
3598 modern systems) in an attempt to fix a rare and obscure crash bug.
3600 25. Narrowed the window for a race in the daemon that could cause it to ignore
3601 SIGCHLD signals. This is not a major problem, because they are used only to
3602 wake it up if nothing else does.
3604 26. A malformed maildirsize file could cause Exim to calculate negative values
3605 for the mailbox size or file count. Odd effects could occur as a result.
3606 The maildirsize information is now recalculated if the size or filecount
3609 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3610 support for a long time. Removed HAVE_SYS_VFS_H.
3612 28. Installed the latest version of exipick from John Jetmore.
3614 29. In an address list, if the pattern was not a regular expression, an empty
3615 subject address (from a bounce message) matched only if the pattern was an
3616 empty string. Non-empty patterns were not even tested. This was the wrong
3617 because it is perfectly reasonable to use an empty address as part of a
3618 database query. An empty address is now tested by patterns that are
3619 lookups. However, all the other forms of pattern expect the subject to
3620 contain a local part and a domain, and therefore, for them, an empty
3621 address still always fails if the pattern is not itself empty.
3623 30. Exim went into a mad DNS loop when attempting to do a callout where the
3624 host was specified on an smtp transport, and looking it up yielded more
3625 than one IP address.
3627 31. Re-factored the code for checking spool and log partition space into a
3628 function that finds that data and another that does the check. The former
3629 is then used to implement four new variables: $spool_space, $log_space,
3630 $spool_inodes, and $log_inodes.
3632 32. The RFC2047 encoding function was originally intended for short strings
3633 such as real names; it was not keeping to the 75-character limit for
3634 encoded words that the RFC imposes. It now respects the limit, and
3635 generates multiple encoded words if necessary. To be on the safe side, I
3636 have increased the buffer size for the ${rfc2047: expansion operator from
3639 33. It is now permitted to omit both strings after an "if" condition; if the
3640 condition is true, the result is "true". As before, when the second string
3641 is omitted, a false condition yields an empty string. This makes it less
3642 cumbersome to write custom ACL and router conditions.
3644 34. Failure to deliver a bounce message always caused it to be frozen, even if
3645 there was an errors_to setting on the router. The errors_to setting is now
3648 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3649 canonical form (fully expanded) before being placed in
3650 $sender_host_address.
3652 36. The table in the code that translates DNS record types into text (T_A to
3653 "A" for instance) was missing entries for NS and CNAME. It is just possible
3654 that this could have caused confusion if both these types were looked up
3655 for the same domain, because the text type is used as part of Exim's
3656 per-process caching. But the chance of anyone hitting this buglet seems
3659 37. The dnsdb lookup has been extended in a number of ways.
3661 (1) There is a new type, "zns", which walks up the domain tree until it
3662 finds some nameserver records. It should be used with care.
3664 (2) There is a new type, "mxh", which is like "mx" except that it returns
3665 just the host names, not the priorities.
3667 (3) It is now possible to give a list of domains (or IP addresses) to be
3668 looked up. The behaviour when one of the lookups defers can be
3669 controlled by a keyword.
3671 (4) It is now possible to specify the separator character for use when
3672 multiple records are returned.
3674 38. The dnslists ACL condition has been extended: it is now possible to supply
3675 a list of IP addresses and/or domains to be looked up in a particular DNS
3678 39. Added log_selector=+queue_time_overall.
3680 40. When running the queue in the test harness, wait just a tad after forking a
3681 delivery process, to get repeatability of debugging output.
3683 41. Include certificate and key file names in error message when GnuTLS fails
3684 to set them up, because the GnuTLS error message doesn't include the name
3685 of the failing file when there is a problem reading it.
3687 42. Allow both -bf and -bF in the same test run.
3689 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3691 44. The "Exiscan patch" is now merged into the mainline Exim source.
3693 45. Sometimes the final signoff response after QUIT could fail to get
3694 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3695 < 0 before doing a fflush(). This bug looks as though it goes back to the
3696 introduction of TLS in release 3.20, but "sometimes" must have been rare
3697 because the tests only now provoked it.
3699 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3700 (this can affect the format of dates).
3702 47. exim_tidydb, when checking for the continued existence of a message for
3703 which it has found a message-specific retry record, was not finding
3704 messages that were in split spool directories. Consequently, it was
3705 deleting retry records that should have stayed in existence.
3707 48. Steve fixed some bugs in eximstats.
3709 49. The SPA authentication driver was not abandoning authentication and moving
3710 on to the next authenticator when an expansion was forced to fail,
3711 contradicting the general specification for all authenticators. Instead it
3712 was generating a temporary error. It now behaves as specified.
3714 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3715 (the order specifies the preference for clients). The order is now AES256,
3716 AES128, 3DES, ARCFOUR128.
3718 51. Small patch to Sieve code - explicitly set From: when generating an
3721 52. Exim crashed if a remote delivery caused a very long error message to be
3722 recorded - for instance if somebody sent an entire SpamAssassin report back
3723 as a large number of 550 error lines. This bug was coincidentally fixed by
3724 increasing the size of one of Exim's internal buffers (big_buffer) that
3725 happened as part of the Exiscan merge. However, to be on the safe side, I
3726 have made the code more robust (and fixed the comments that describe what
3729 53. Now that there can be additional text after "Completed" in log lines (if
3730 the queue_time_overall log selector is set), a one-byte patch to exigrep
3731 was needed to allow it to recognize "Completed" as not the last thing in
3734 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3735 patch that reportedly fixes this has been added. I am not expert enough to
3736 create a test for it. This is what the patch creator wrote:
3738 "I found a little strange behaviour of ldap code when working with
3739 Windows 2003 AD Domain, where users was placed in more than one
3740 Organization Units. When I tried to give exim partial DN, the exit code
3741 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3742 But simultaneously result of request was absolutely normal ldap result,
3743 so I produce this patch..."
3745 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3746 REFERENCE, so I have modified the code to exclude the patch when that macro
3749 55. Some experimental protocols are using DNS PTR records for new purposes. The
3750 keys for these records are domain names, not reversed IP addresses. The
3751 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3752 leaves it alone. Component reversal etc. now happens only for IP addresses.
3755 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3757 57. Double the size of the debug message buffer (to 2048) so that more of very
3758 long debug lines gets shown.
3760 58. The exicyclog utility now does better if the number of log files to keep
3761 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3763 59. Two changes related to the smtp_active_hostname option:
3765 (1) $smtp_active_hostname is now available as a variable.
3766 (2) The default for smtp_banner uses $smtp_active_hostname instead
3767 of $primary_hostname.
3769 60. The host_aton() function is supposed to be passed a string that is known
3770 to be a valid IP address. However, in the case of IPv6 addresses, it was
3771 not checking this. This is a hostage to fortune. Exim now panics and dies
3772 if the condition is not met. A case was found where this could be provoked
3773 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3774 components; fortuitously, this particular loophole had already been fixed
3775 by change 4.50/55 above.
3777 If there are any other similar loopholes, the new check in host_aton()
3778 itself should stop them being exploited. The report I received stated that
3779 data on the command line could provoke the exploit when Exim was running as
3780 exim, but did not say which command line option was involved. All I could
3781 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3782 running as the user.
3785 61. There was a buffer overflow vulnerability in the SPA authentication code
3786 (which came originally from the Samba project). I have added a test to the
3787 spa_base64_to_bits() function which I hope fixes it.
3790 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3791 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3793 63. The daemon start-up calls getloadavg() while still root for those OS that
3794 need the first call to be done as root, but it missed one case: when
3795 deliver_queue_load_max is set with deliver_drop_privilege. This is
3796 necessary for the benefit of the queue runner, because there is no re-exec
3797 when deliver_drop_privilege is set.
3799 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3800 This has been fixed.
3802 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3803 was in use, was not putting the data itself into the right store pool;
3804 consequently, it could be overwritten for a subsequent message in the same
3805 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3808 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3810 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3811 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3812 the function were treating the return as a boolean value, which happened to
3813 work because 0=false and not-0=true, but is not correct code.
3815 68. The host_aton() function was not handling scoped IPv6 addresses (those
3816 with, for example, "%eth0" on the end) correctly.
3818 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3819 negated items (that is, ~something) in unsigned ints. Some compilers
3820 apparently mutter when there is no cast.
3822 70. If an address verification called from an ACL failed, and did not produce a
3823 user-specific message (i.e. there was only a "system" message), nothing was
3824 put in $acl_verify_message. In this situation, it now puts the system
3827 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3828 session; change 4.31/43 added the unwanted input to the log line - except
3829 that it did not do this in the start of session case. It now does.
3831 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3832 This is wrong and can cause the other end to generate a synchronization
3833 error if it is another Exim or anything else that does the synchronization
3834 check. A QUIT command is no longer sent after a timeout.
3836 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3837 during host lookups.
3839 74. The defer_ok option of callout verification was not working if it was used
3840 when verifying addresses in header lines, that is, for this case:
3842 verify = header_sender/callout=defer_ok
3844 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3845 those file descriptors could be used for SMTP connections. If anything
3846 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3847 could be sent to the SMTP client, causing chaos. The daemon now opens
3848 stdin, stdout, and stderr to /dev/null when it puts itself into the
3851 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3852 log by default. The user can override this with suitable Perl magic.
3854 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3855 the log message when discard triggers, was not working for the DATA ACL or
3856 for the non-SMTP ACL.
3858 78. Error message wording change in sieve.c.
3860 79. If smtp_accept_max_per_host was set, the number of connections could be
3861 restricted to fewer than expected, because the daemon was trying to set up
3862 a new connection before checking whether the processes handling previous
3863 connections had finished. The check for completed processes is now done
3864 earlier. On busy systems, this bug wouldn't be noticed because something
3865 else would have woken the daemon, and it would have reaped the completed
3868 80. If a message was submitted locally by a user whose login name contained one
3869 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3870 It caused a spool format error. I have fixed the spool reading code. A
3871 related problem was that the "from" clause in the Received: line became
3872 illegal because of the space(s). It is now covered by ${quote_local_part.
3874 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3877 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3879 83. Patch from Sieve maintainer to fix unterminated string problem in
3880 "vacation" handling.
3882 84. Some minor changes to the Linux configuration files to help with other
3883 OS variants using glibc.
3885 85. One more patch for Sieve to update vacation handling to latest spec.
3888 ----------------------------------------------------
3889 See the note above about the 4.44 and 4.50 releases.
3890 ----------------------------------------------------
3896 1. Change 4.43/35 introduced a bug that caused file counts to be
3897 incorrectly computed when quota_filecount was set in an appendfile
3900 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3901 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3904 3. Give more explanation in the error message when the command for a transport
3905 filter fails to execute.
3907 4. There are several places where Exim runs a non-Exim command in a
3908 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3909 was being done only for the command run by the queryprogram router. It is
3910 now done for all such subprocesses. The other cases are: ${run, transport
3911 filters, and the commands run by the lmtp and pipe transports.
3913 5. Some older OS have a limit of 256 on the maximum number of file
3914 descriptors. Exim was using setrlimit() to set 1000 as a large value
3915 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3916 systems. I've change it so that if it can't get 1000, it tries for 256.
3918 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3919 was an oversight, and furthermore, ever since the addition of extra
3920 controls (e.g. 4.43/32), the checks on when to allow different forms of
3921 "control" were broken. There should now be diagnostics for all cases when a
3922 control that does not make sense is encountered.
3924 7. $recipients is now available in the predata ACL (oversight).
3926 8. Tidy the search cache before the fork to do a delivery from a message
3927 received from the command line. Otherwise the child will trigger a lookup
3928 failure and thereby defer the delivery if it tries to use (for example) a
3929 cached ldap connection that the parent has called unbind on.
3931 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3932 of $address_data from the recipient verification was clobbered by the
3933 sender verification.
3935 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3936 was its contents. (It was OK if the option was not defined at all.)
3938 11. A "Completed" log line is now written for messages that are removed from
3939 the spool by the -Mrm option.
3941 12. $host_address is now set to the target address during the checking of
3942 ignore_target_hosts.
3944 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3945 being passed; this would have caused $sender_host_name to have been used if
3946 matching the list had actually called for a host name (not very likely,
3947 since this list is usually IP addresses). A host name is now passed as
3950 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3951 code that specifies a non-restarting handler (typically sigaction() in
3952 modern systems) in an attempt to fix a rare and obscure crash bug.
3954 15. Narrowed the window for a race in the daemon that could cause it to ignore
3955 SIGCHLD signals. This is not a major problem, because they are used only to
3956 wake it up if nothing else does.
3958 16. A malformed maildirsize file could cause Exim to calculate negative values
3959 for the mailbox size or file count. Odd effects could occur as a result.
3960 The maildirsize information is now recalculated if the size or filecount
3963 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3964 support for a long time. Removed HAVE_SYS_VFS_H.
3966 18. Updated exipick to current release from John Jetmore.
3968 19. Allow an empty sender to be matched against a lookup in an address list.
3969 Previously the only cases considered were a regular expression, or an
3972 20. Exim went into a mad DNS lookup loop when doing a callout where the
3973 host was specified on the transport, if the DNS lookup yielded more than
3976 21. The RFC2047 encoding function was originally intended for short strings
3977 such as real names; it was not keeping to the 75-character limit for
3978 encoded words that the RFC imposes. It now respects the limit, and
3979 generates multiple encoded words if necessary. To be on the safe side, I
3980 have increased the buffer size for the ${rfc2047: expansion operator from
3983 22. Failure to deliver a bounce message always caused it to be frozen, even if
3984 there was an errors_to setting on the router. The errors_to setting is now
3987 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3988 canonical form (fully expanded) before being placed in
3989 $sender_host_address.
3991 24. Updated eximstats to version 1.33
3993 25. Include certificate and key file names in error message when GnuTLS fails
3994 to set them up, because the GnuTLS error message doesn't include the name
3995 of the failing file when there is a problem reading it.
3997 26. Expand error message when OpenSSL has problems setting up cert/key files.
4000 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4001 (this can affect the format of dates).
4003 28. exim_tidydb, when checking for the continued existence of a message for
4004 which it has found a message-specific retry record, was not finding
4005 messages that were in split spool directories. Consequently, it was
4006 deleting retry records that should have stayed in existence.
4008 29. eximstats updated to version 1.35
4009 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4010 1.35 - bugfix such that pie charts by volume are generated correctly
4012 30. The SPA authentication driver was not abandoning authentication and moving
4013 on to the next authenticator when an expansion was forced to fail,
4014 contradicting the general specification for all authenticators. Instead it
4015 was generating a temporary error. It now behaves as specified.
4017 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4018 (the order specifies the preference for clients). The order is now AES256,
4019 AES128, 3DES, ARCFOUR128.
4021 31. Small patch to Sieve code - explicitly set From: when generating an
4024 32. Exim crashed if a remote delivery caused a very long error message to be
4025 recorded - for instance if somebody sent an entire SpamAssassin report back
4026 as a large number of 550 error lines. This bug was coincidentally fixed by
4027 increasing the size of one of Exim's internal buffers (big_buffer) that
4028 happened as part of the Exiscan merge. However, to be on the safe side, I
4029 have made the code more robust (and fixed the comments that describe what
4032 33. Some experimental protocols are using DNS PTR records for new purposes. The
4033 keys for these records are domain names, not reversed IP addresses. The
4034 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4035 leaves it alone. Component reversal etc. now happens only for IP addresses.
4038 34. The host_aton() function is supposed to be passed a string that is known
4039 to be a valid IP address. However, in the case of IPv6 addresses, it was
4040 not checking this. This is a hostage to fortune. Exim now panics and dies
4041 if the condition is not met. A case was found where this could be provoked
4042 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4043 components; fortuitously, this particular loophole had already been fixed
4044 by change 4.50/55 or 4.44/33 above.
4046 If there are any other similar loopholes, the new check in host_aton()
4047 itself should stop them being exploited. The report I received stated that
4048 data on the command line could provoke the exploit when Exim was running as
4049 exim, but did not say which command line option was involved. All I could
4050 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4051 running as the user.
4054 35. There was a buffer overflow vulnerability in the SPA authentication code
4055 (which came originally from the Samba project). I have added a test to the
4056 spa_base64_to_bits() function which I hope fixes it.
4059 36. The daemon start-up calls getloadavg() while still root for those OS that
4060 need the first call to be done as root, but it missed one case: when
4061 deliver_queue_load_max is set with deliver_drop_privilege. This is
4062 necessary for the benefit of the queue runner, because there is no re-exec
4063 when deliver_drop_privilege is set.
4065 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4066 was in use, was not putting the data itself into the right store pool;
4067 consequently, it could be overwritten for a subsequent message in the same
4068 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4071 38. Sometimes the final signoff response after QUIT could fail to get
4072 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4073 < 0 before doing a fflush(). This bug looks as though it goes back to the
4074 introduction of TLS in release 3.20, but "sometimes" must have been rare
4075 because the tests only now provoked it.
4081 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4082 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4083 Now it returns an integer. A number of calls were still expecting a T/F
4084 return. Fortuitously, in all cases, the tests worked in OK situations,
4085 which is the norm. However, things would have gone wrong on any write
4086 failures on the smtp file descriptor. This function is used when sending
4087 messages over SMTP and also when doing verify callouts.
4089 2. When Exim is called to do synchronous delivery of a locally submitted
4090 message (the -odf or -odi options), it no longer closes stderr before doing
4093 3. Implemented the mua_wrapper option.
4095 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4097 5. Implemented the functions header_remove(), header_testname(),
4098 header_add_at_position(), and receive_remove_recipient(), and exported them
4101 6. If an ACL "warn" statement specified the addition of headers, Exim already
4102 inserted X-ACL-Warn: at the start if there was no header name. However, it
4103 was not making this test for the second and subsequent header lines if
4104 there were newlines in the string. This meant that an invalid header could
4105 be inserted if Exim was badly configured.
4107 7. Allow an ACL "warn" statement to add header lines at the start or after all
4108 the Received: headers, as well as at the end.
4110 8. Added the rcpt_4xx retry error code.
4112 9. Added postmaster_mailfrom=xxx to callout verification option.
4114 10. Added mailfrom=xxxx to the callout verification option, for verify=
4117 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4118 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4120 12. Inserted some casts to stop certain compilers warning when using pointer
4121 differences as field lengths or precisions in printf-type calls (mostly
4122 affecting debugging statements).
4124 13. Added optional readline() support for -be (dynamically loaded).
4126 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4127 same clock tick as a message's arrival, so that its received time was the
4128 same as the "first fail" time on the retry record, and that message
4129 remained on the queue past the ultimate address timeout, every queue runner
4130 would try a delivery (because it was past the ultimate address timeout) but
4131 after another failure, the ultimate address timeout, which should have then
4132 bounced the address, did not kick in. This was a "< instead of <=" error;
4133 in most cases the first failure would have been in the next clock tick
4134 after the received time, and all would be well.
4136 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4137 being recognized when the domain list was tested by the match_domain
4138 condition in an expansion string.
4140 16. Added the ${str2b64: operator.
4142 17. Exim was always calling setrlimit() to set a large limit for the number of
4143 processes, without checking whether the existing limit was already
4144 adequate. (It did check for the limit on file descriptors.) Furthermore,
4145 errors from getrlimit() and setrlimit() were being ignored. Now they are
4146 logged to the main and panic logs, but Exim does carry on, to try to do its
4147 job under whatever limits there are.
4149 18. Imported PCRE 5.0.
4151 19. Trivial typo in log message " temporarily refused connection" (the leading
4154 20. If the log selector return_path_on_delivery was set and an address was
4155 redirected to /dev/null, the delivery process crashed because it assumed
4156 that a return path would always be set for a "successful" delivery. In this
4157 case, the whole delivery is bypassed as an optimization, and therefore no
4160 21. Internal re-arrangement: the function for sending a challenge and reading
4161 a response while authentication was assuming a zero-terminated challenge
4162 string. It's now changed to take a pointer and a length, to allow for
4163 binary data in such strings.
4165 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4167 23. Exim was not respecting finduser_retries when seeking the login of the
4168 uid under which it was called; it was always trying 10 times. (The default
4169 setting of finduser_retries is zero.) Also, it was sleeping after the final
4170 failure, which is pointless.
4172 24. Implemented tls_on_connect_ports.
4174 25. Implemented acl_smtp_predata.
4176 26. If the domain in control=submission is set empty, Exim assumes that the
4177 authenticated id is a complete email address when it generates From: or
4178 Sender: header lines.
4180 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4181 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4182 chown and chgrp in /bin and hostname in /usr/bin.
4184 28. Exim was keeping the "process log" file open after each use, just as it
4185 does for the main log. This opens the possibility of it remaining open for
4186 long periods when the USR1 signal hits a daemon. Occasional processlog
4187 errors were reported, that could have been caused by this. Anyway, it seems
4188 much more sensible not to leave this file open at all, so that is what now
4191 29. The long-running daemon process does not normally write to the log once it
4192 has entered its main loop, and it closes the log before doing so. This is
4193 so that log files can straightforwardly be renamed and moved. However,
4194 there are a couple of unusual error situations where the daemon does write
4195 log entries, and I had neglected to close the log afterwards.
4197 30. The text of an SMTP error response that was received during a remote
4198 delivery was being truncated at 512 bytes. This is too short for some of
4199 the long messages that one sometimes sees. I've increased the limit to
4202 31. It is now possible to make retry rules that apply only when a message has a
4203 specific sender, in particular, an empty sender.
4205 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4206 it possible to be selective about when SMTP synchronization is enforced.
4208 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4210 32. Implemented hosts_connection_nolog.
4212 33. Added an ACL for QUIT.
4214 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4217 35. Added mailbox_size and mailbox_filecount to appendfile.
4219 36. Added control = no_multiline_responses to ACLs.
4221 37. There was a bug in the logic of the code that waits for the clock to tick
4222 in the case where the clock went backwards by a substantial amount such
4223 that the microsecond fraction of "now" was more than the microsecond
4224 fraction of "then" (but the whole seconds number was less).
4226 38. Added support for the libradius Radius client library this is found on
4227 FreeBSD (previously only the radiusclient library was supported).
4233 1. When certain lookups returned multiple values in the form name=value, the
4234 quoting of the values was not always being done properly. Specifically:
4235 (a) If the value started with a double quote, but contained no whitespace,
4237 (b) If the value contained whitespace other than a space character (i.e.
4238 tabs or newlines or carriage returns) it was not quoted.
4239 This fix has been applied to the mysql and pgsql lookups by writing a
4240 separate quoting function and calling it from the lookup code. The fix
4241 should probably also be applied to nisplus, ibase and oracle lookups, but
4242 since I cannot test any of those, I have not disturbed their existing code.
4244 2. A hit in the callout cache for a specific address caused a log line with no
4245 reason for rejecting RCPT. Now it says "Previous (cached) callout
4246 verification failure".
4248 3. There was an off-by-one bug in the queryprogram router. An over-long
4249 return line was truncated at 256 instead of 255 characters, thereby
4250 overflowing its buffer with the terminating zero. As well as fixing this, I
4251 have increased the buffer size to 1024 (and made a note to document this).
4253 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4254 when Exim is waiting for an SMTP response from a remote server, Exim
4255 restarts its select() call on the socket, thereby resetting its timeout.
4256 This is not a problem when such interrupts are rare. Somebody set up a cron
4257 job to run exiwhat every 2 minutes, which is less than the normal select()
4258 timeout (5 or 10 minutes). This meant that the select() timeout never
4259 kicked in because it was always reset. I have fixed this by comparing the
4260 time when an interrupt arrives with the time at the start of the first call
4261 to select(). If more time than the timeout has elapsed, the interrupt is
4262 treated as a timeout.
4264 5. Some internal re-factoring in preparation for the addition of Sieve
4265 extensions (by MH). In particular, the "personal" test is moved to a
4266 separate function, and given an option for scanning Cc: and Bcc: (which is
4267 not set for Exim filters).
4269 6. When Exim created an email address using the login of the caller as the
4270 local part (e.g. when creating a From: or Sender: header line), it was not
4271 quoting the local part when it contained special characters such as @.
4273 7. Installed new OpenBSD configuration files.
4275 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4276 try to make them clearer.
4278 9. Callout options, other than the timeout value, were being ignored when
4279 verifying sender addresses in header lines. For example, when using
4281 verify = header_sender/callout=no_cache
4283 the cache was (incorrectly) being used.
4285 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4286 only the Cygwin environment.
4288 11. When return_path_on_delivery was set as a log selector, if different remote
4289 addresses in the same message used different return paths and parallel
4290 remote delivery occurred, the wrong values would sometimes be logged.
4291 (Whenever a remote delivery process finished, the return path value from
4292 the most recently started remote delivery process was logged.)
4294 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4295 lines when AUTH and/or TLS are in use. This is the "received protocol"
4296 field. Exim used to use "asmtp" for authenticated SMTP, without any
4297 indication (in the protocol name) for TLS use. Now it follows the RFC and
4298 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4299 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4300 names appear in log lines as well as in Received: header lines.
4302 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4303 extensions, and comparison tests, and to fix some bugs.
4305 14. Changes to the "personal" filter test:
4307 (1) The test was buggy in that it was just doing the equivalent of
4308 "contains" tests on header lines. For example, if a user's address was
4309 anne@some.where, the "personal" test would incorrectly be true for
4311 To: susanne@some.where
4313 This test is now done by extracting each address from the header in turn,
4314 and checking the entire address. Other tests that are part of "personal"
4315 are now done using regular expressions (for example, to check local parts
4316 of addresses in From: header lines).
4318 (2) The list of non-personal local parts in From: addresses has been
4319 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4320 taken from the Sieve specification recommendations.
4322 (3) If the message contains any header line starting with "List-" it is
4323 treated as non-personal.
4325 (4) The test for "circular" in the Subject: header line has been removed
4326 because it now seems ill-conceived.
4328 15. Minor typos in src/EDITME comments corrected.
4330 16. Installed latest exipick from John Jetmore.
4332 17. If headers_add on a router specified a text string that was too long for
4333 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4334 of string_sprintf() is now avoided.
4336 18. $message_body_size was not set (it was always zero) when running the DATA
4337 ACL and the local_scan() function.
4339 19. For the "mail" command in an Exim filter, no default was being set for
4340 the once_repeat time, causing a random time value to be used if "once" was
4341 specified. (If the value happened to be <= 0, no repeat happened.) The
4342 default is now 0s, meaning "never repeat". The "vacation" command was OK
4343 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4344 (I found it when inspecting the code).
4346 20. There is now an overall timeout for performing a callout verification. It
4347 defaults to 4 times the callout timeout, which applies to individual SMTP
4348 commands during the callout. The overall timeout applies when there is more
4349 than one host that can be tried. The timeout is checked before trying the
4350 next host. This prevents very long delays if there are a large number of
4351 hosts and all are timing out (e.g. when the network connections are timing
4352 out). The value of the overall timeout can be changed by specifying an
4353 additional sub-option for "callout", called "maxwait". For example:
4355 verify = sender/callout=5s,maxwait=20s
4357 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4358 to the end before writing, but this should make it even safer).
4360 22. Exim was forgetting that it had advertised PIPELINING for the second and
4361 subsequent messages on an SMTP connection. It was also not resetting its
4362 memory on STARTTLS and an internal HELO.
4364 23. When Exim logs an SMTP synchronization error within a session, it now
4365 records whether PIPELINING has been advertised or not.
4367 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4368 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4369 rather than long int.
4371 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4373 26. Added the never_mail option to autoreply.
4379 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4380 crash if the getsockname() call failed; this can happen if a connection is
4381 closed very soon after it is established. The problem was simply in the
4382 order in which certain operations were done, causing Exim to try to write
4383 to the SMTP stream before it had set up the file descriptor. The bug has
4384 been fixed by making things happen in the correct order.
4390 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4391 before the connection was closed, thus losing the rejection response.
4393 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4394 some early Solaris releases, but causes trouble in current releases where
4395 socklen_t is defined.
4397 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4400 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4403 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4404 that releasing the top of it at the end releases what was used for sub-
4405 expansions (unless the block got too big). However, discard this block if
4406 the first thing is a variable or header, so that we can use its block when
4407 it is dynamic (useful for very large $message_headers, for example).
4409 6. Lookups now cache *every* query, not just the most recent. A new, separate
4410 store pool is used for this. It can be recovered when all lookup caches are
4411 flushed. Lookups now release memory at the end of their result strings.
4412 This has involved some general refactoring of the lookup sources.
4414 7. Some code has been added to the store_xxx() functions to reduce the amount
4415 of flapping under certain conditions.
4417 8. log_incoming_interface used to affect only the <= reception log lines. Now
4418 it causes the local interface and port to be added to several more SMTP log
4419 lines, for example "SMTP connection from", and rejection lines.
4421 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4423 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4425 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4426 could be overwritten at the end of the current message (or the start of a
4427 new message if it was set in a HELO ACL). The value is now preserved for
4428 the duration of the SMTP connection.
4430 12. If a transport had a headers_rewrite setting, and a matching header line
4431 contained an unqualified address, that address was qualified, even if it
4432 did not match any rewriting rules. The underlying bug was that the values
4433 of the flags that permit the existence of unqualified sender and recipient
4434 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4435 non-local messages, and by -bnq for local messages) were not being
4436 preserved with the message after it was received.
4438 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4439 "next input=" as part of the text comprising the host identity instead of
4440 the correct text. The code was using the same buffer for two different
4441 strings. However, depending on which order the printing function evaluated
4442 its arguments, the bug did not always show up. Under Linux, for example, my
4443 test suite worked just fine.
4445 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4446 doesn't work with some older versions of Perl. It has been changed to "my",
4447 which in any case is probably the better facility to use.
4449 15. A really picky compiler found some instances of statements for creating
4450 error messages that either had too many or two few arguments for the format
4453 16. The size of the buffer for calls to the DNS resolver has been increased
4454 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4455 for addresses that have a lot of PTR records. This alleviates a problem; it
4456 does not fully solve it.
4458 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4459 buffer now truncates the list and logs the incident, which is the same
4460 action as happens when Exim is looking up a host name and its aliases.
4461 Previously in this situation something unpredictable would happen;
4462 sometimes it was "internal error: store_reset failed".
4464 18. If a server dropped the connection unexpectedly when an Exim client was
4465 using GnuTLS and trying to read a response, the client delivery process
4466 crashed while trying to generate an error log message.
4468 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4469 string, for example:
4471 warn message = H1: something\nH2: something
4473 the text was added as a single header line from Exim's point of view
4474 though it ended up OK in the delivered message. However, searching for the
4475 second and subsequent header lines using $h_h2: did not work. This has been
4476 fixed. Similarly, if a system filter added multiple headers in this way,
4477 the routers could not see them.
4479 20. Expanded the error message when iplsearch is called with an invalid key to
4480 suggest using net-iplsearch in a host list.
4482 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4483 ACLs are no longer actually imposed (and a message to that effect is
4486 22. If a "gecos" field in a passwd entry contained escaped characters, in
4487 particular, if it contained a \" sequence, Exim got it wrong when building
4488 a From: or a Sender: header from that name. A second bug also caused
4489 incorrect handling when an unquoted " was present following a character
4490 that needed quoting.
4492 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4493 was not being matched caselessly.
4495 24. Arranged for all hyphens in the exim.8 source to be escaped with
4498 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4499 properly. Recipient callout cache records were still being keyed to include
4500 the sender, even when use_sender was set false. This led to far more
4501 callouts that were necessary. The sender is no longer included in the key
4502 when use_sender is false.
4504 26. Added "control = submission" modifier to ACLs.
4506 27. Added the ${base62d: operator to decode base 62 numbers.
4508 28. dnsdb lookups can now access SRV records.
4510 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4511 the configuration file.
4513 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4514 (-v) mode. This makes the output for a verbose queue run more intelligible.
4516 31. Added a use_postmaster feature to recipient callouts.
4518 32. Added the $body_zerocount variable, containing the number of binary zero
4519 bytes in the message body.
4521 33. The time of last modification of the "new" subdirectory is now used as the
4522 "mailbox time last read" when there is a quota error for a maildir
4525 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4527 35. Added +ignore_unknown as a special item in host lists.
4529 36. Code for decoding IPv6 addresses in host lists is now included, even if
4530 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4531 address was recognized as an IP address, but was then not correctly decoded
4532 into binary, causing unexpected and incorrect effects when compared with
4539 1. Very minor rewording of debugging text in manualroute to say "list of
4540 hosts" instead of "hostlist".
4542 2. If verify=header_syntax was set, and a header line with an unqualified
4543 address (no domain) and a large number of spaces between the end of the
4544 name and the colon was received, the reception process suffered a buffer
4545 overflow, and (when I tested it) crashed. This was caused by some obsolete
4546 code that should have been removed. The fix is to remove it!
4548 3. When running in the test harness, delay a bit after writing a bounce
4549 message to get a bit more predictability in the log output.
4551 4. Added a call to search_tidyup() just before forking a reception process. In
4552 theory, someone could use a lookup in the expansion of smtp_accept_max_
4553 per_host which, without the tidyup, could leave open a database connection.
4555 5. Added the variables $recipient_data and $sender_data which get set from a
4556 lookup success in an ACL "recipients" or "senders" condition, or a router
4557 "senders" option, similar to $domain_data and $local_part_data.
4559 6. Moved the writing of debug_print from before to after the "senders" test
4562 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4563 problems for message scanning, either using a data ACL, or using
4564 local_scan() because the Received: header was not generated till after they
4565 were called (in order to set the time as the time of reception completion).
4566 I have revised the way this works. The header is now generated after the
4567 body is received, but before the ACL or local_scan() are called. After they
4568 are run, the timestamp in the header is updated.
4574 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4575 before starting a queue runner without re-exec. This happened only when
4576 deliver_drop_privilege was set or when the Exim user was set to root. The
4577 effect of the bug was that timeouts during subsequent deliveries caused
4578 crashes instead of being properly handled. The handler is now left at its
4579 default (and expected) setting.
4581 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4582 message, again when deliver_drop_privilege is set or Exim is run as root.
4583 The bug described in (1) was not present in this case, but the tidying up
4584 of the other signals was missing. I have made the two cases consistent.
4586 3. The ignore_target_hosts setting on a manualroute router was being ignored
4587 for hosts that were looked up using the /MX notation.
4589 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4592 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4593 operated on the sender address. After changing the $sender_address to <>
4594 for the sender address verify, Exim was re-instated it as the original
4595 (before rewriting) address, but remembering that it had rewritten it, so it
4596 wasn't rewriting it again. This bug also had the effect of breaking the
4597 sender address verification caching when the sender address was rewritten.
4599 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4600 This has been changed so that if the ip literal address matches
4601 ignore_target_hosts, the router declines.
4603 7. Added expansion conditions match_domain, match_address, and match_local_
4604 part (NOT match_host).
4606 8. The placeholder for the Received: header didn't have a length field set.
4608 9. Added code to Exim itself and to exim_lock to test for a specific race
4609 condition that could lead to file corruption when using MBX delivery. The
4610 issue is with the lockfile that is created in /tmp. If this file is removed
4611 after a process has opened it but before that process has acquired a lock,
4612 there is the potential for a second process to recreate the file and also
4613 acquire a lock. This could lead to two Exim processes writing to the file
4614 at the same time. The added code performs the same test as UW imapd; it
4615 checks after acquiring the lock that its file descriptor still refers to
4616 the same named file.
4618 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4619 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4622 11. Added the smtp_active_hostname option. If used, this will typically be made
4623 to depend on the incoming interface address. Because $interface_address is
4624 not set up until the daemon has forked a reception process, error responses
4625 that can happen earlier (such as "too many connections") no longer contain
4628 12. If an expansion in a condition on a "warn" statement fails because a lookup
4629 defers, the "warn" statement is abandoned, and the next ACL statement is
4630 processed. Previously this caused the whole ACL to be aborted.
4632 13. Added the iplsearch lookup type.
4634 14. Added ident_timeout as a log selector.
4636 15. Added tls_certificate_verified as a log selector.
4638 16. Added a global option tls_require_ciphers (compare the smtp transport
4639 option of the same name). This controls incoming TLS connections.
4641 17. I finally figured out how to make tls_require_ciphers do a similar thing
4642 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4643 before starting the TLS session.
4645 18. Tabs are now shown as \t in -bP output.
4647 19. If the log selector return_path_on_delivery was set, Exim crashed when
4648 bouncing a message because it had too many Received: header lines.
4650 20. If two routers both had headers_remove settings, and the first one included
4651 a superfluous trailing colon, the final name in the first list and the
4652 first name in the second list were incorrectly joined into one item (with a
4653 colon in the middle).
4659 1. Added -C and -D options to the exinext utility, mainly to make it easier
4660 to include in the automated testing, but these could be helpful when
4661 multiple configurations are in use.
4663 2. The exinext utility was not formatting the output nicely when there was
4664 an alternate port involved in the retry record key, nor when there was a
4665 message id as well (for retries that were specific to a specific message
4666 and a specific host). It was also confused by IPv6 addresses, because of
4667 the additional colons they contain. I have fixed the IPv4 problem, and
4668 patched it up to do a reasonable job for IPv6.
4670 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4671 delivery, the log line now contains "pipelined" if PIPELINING was used.
4673 4. An SMTP transport process used to panic and die if the bind() call to set
4674 an explicit outgoing interface failed. This has been changed; it is now
4675 treated in the same way as a connect() failure.
4677 5. A reference to $sender_host_name in the part of a conditional expansion
4678 that was being skipped was still causing a DNS lookup. This no longer
4681 6. The def: expansion condition was not recognizing references to header lines
4682 that used bh_ and bheader_.
4684 7. Added the _cache feature to named lists.
4686 8. The code for checking quota_filecount in the appendfile transport was
4687 allowing one more file than it should have been.
4689 9. For compatibility with Sendmail, the command line option
4697 and sets the incoming protocol and host name (for trusted callers). The
4698 host name and its colon can be omitted when only the protocol is to be set.
4699 Note the Exim already has two private options, -pd and -ps, that refer to
4700 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4701 "s", but I don't think that's a major issue.
4703 10. A number of refactoring changes to the code, none of which should affect
4706 (a) The number of logging options was getting close to filling up the
4707 32-bit word that was used as a bit map. I have split them into two classes:
4708 those that are passed in the argument to log_write(), and those that are
4709 only ever tested independently outside of that function. These are now in
4710 separate 32-bit words, so there is plenty of room for expansion again.
4711 There is no change in the user interface or the logging behaviour.
4713 (b) When building, for example, log lines, the code previously used a
4714 macro that called string_cat() twice, in order to add two strings. This is
4715 not really sufficiently general. Furthermore, there was one instance where
4716 it was actually wrong because one of the argument was used twice, and in
4717 one call a function was used. (As it happened, calling the function twice
4718 did not affect the overall behaviour.) The macro has been replaced by a
4719 function that can join an arbitrary number of extra strings onto a growing
4722 (c) The code for expansion conditions now uses a table and a binary chop
4723 instead of a serial search (which was left over from when there were very
4724 few conditions). Also, it now recognizes conditions like "pam" even when
4725 the relevant support is not compiled in: a suitably worded error message is
4726 given if an attempt is made to use such a condition.
4728 11. Added ${time_interval:xxxxx}.
4730 12. A bug was causing one of the ddress fields not to be passed back correctly
4731 from remote delivery subprocesses. The field in question was not being
4732 subsequently used, so this caused to problems in practice.
4734 13. Added new log selectors queue_time and deliver_time.
4736 14. Might have fixed a bug in maildirsizefile handling that threw up
4737 "unexpected character" debug warnings, and recalculated the data
4738 unnecessarily. In any case, I expanded the warning message to give more
4741 15. Added the message "Restricted characters in address" to the statements in
4742 the default ACL that block characters like @ and % in local parts.
4744 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4745 Three changes have been made:
4747 (a) There was a serious bug; a negative response to MAIL caused the whole
4748 recipient domain to be cached as invalid, thereby blocking all messages
4749 to all local parts at the same domain, from all senders. This bug has
4750 been fixed. The domain is no longer cached after a negative response to
4751 MAIL if the sender used is not empty.
4753 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4756 (c) A new callout option, "use_sender" has been added for people who want
4757 the modified behaviour.
4763 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4766 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4767 indeed breaks things for older releases.
4769 3. Added additional logging to the case where there is a problem reading data
4770 from a filter that is running in a subprocess using a pipe, in order to
4771 try to track down a specific problem.
4773 4. Testing facility fudge: when running in the test harness and attempting
4774 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4775 getting "No route to host". Convert this to a timeout.
4777 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4780 6. Some OS don't have socklen_t but use size_t instead. This affects the
4781 fifth argument of getsockopt() amongst other things. This is now
4782 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4783 can be set for individual OS. I have set it for SunOS5, OSF1, and
4784 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4785 some earlier ones do not.
4787 7. Change 4.30/15 was not doing the test caselessly.
4789 8. The standard form for an IPv6 address literal was being rejected by address
4790 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4791 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4792 this, as well as the form without the "IPv6" on the front (but only when
4793 address literals are enabled, of course).
4795 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4797 10. Exim crashed if a message with an empty sender address specified by -f
4798 encountered a router with an errors_to setting. This could be provoked only
4799 by a command such as
4803 where an empty string was supplied; "<>" did not hit this bug.
4805 11. Installed PCRE release 4.5.
4807 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4808 remained set. It is now erased.
4810 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4811 times from message ids (which are base 36 rather than the normal 62).
4813 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4814 were being counted as actual protocol errors, and logged if the log
4815 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4816 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4817 and DATA following a set of rejected RCPTs do not count as protocol errors.
4818 In other words, Exim assumes they were pipelined, though this may not
4819 actually be the case. Of course, in all cases the client gets an
4820 appropriate error code.
4822 15. If a lookup fails in an ACL condition, a message about the failure may
4823 be available; it is used if testing the ACL cannot continue, because most
4824 such messages specify what the cause of the deferral is. However, some
4825 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4826 that caused an old message to be retained and used if a later statement
4827 caused a defer, replacing the real cause of the deferral.
4829 16. If an IP address had so many PTR records that the DNS lookup buffer
4830 was not large enough to hold them, Exim could crash while trying to process
4831 the truncated data. It now detects and logs this case.
4833 17. Further to 4.21/58, another change has been made: if (and only if) the
4834 first line of a message (the first header line) ends with CRLF, a bare LF
4835 in a subsequent header line has a space inserted after it, so as not to
4836 terminate the header.
4838 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4839 unnecessarily, used atoi() instead of strtol(), and didn't check the
4840 termination when getting file sizes from file names by regex.
4842 19. Completely re-implemented the support for maildirsize files, in the light
4843 of a number of problems with the previous contributed implementation
4844 (4.30/29). In particular:
4846 . If the quota is zero, the maildirsize file is maintained, but no quota is
4849 . If the maildir directory does not exist, it is created before any attempt
4850 to write a maildirsize file.
4852 . The quota value in the file is just a cache; if the quota is changed in
4853 the transport, the new value overrides.
4855 . A regular expression is available for excluding directories from the
4858 20. The autoreply transport checks the characters in options that define the
4859 message's headers; it allows continued headers, but it was checking with
4860 isspace() after an embedded newline instead of explicitly looking for a
4863 21. If all the "regular" hosts to which an address was routed had passed their
4864 expiry times, and had not reached their retry times, the address was
4865 bounced, even if fallback hosts were defined. Now Exim should go on to try
4868 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4869 equivalent code in the SMTP transport. Some hosts send humungous responses
4870 to HELO/EHLO, more than 1024 it seems.
4872 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4873 gives compiler warnings in some environments. I've now done it "properly",
4876 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4877 (because of problems with the built-in one) was declared to return uschar *
4878 instead of char *, causing compiler failure.
4880 25. Fixed a file descriptor leak when processing alias/forward files.
4882 26. Fixed a minor format string issue in dbfn.c.
4884 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4886 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4887 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4890 29. When a sender address is verified, it is cached, to save repeating the test
4891 when there is more than one recipient in a message. However, when the
4892 verification involves a callout, it is possible for different callout
4893 options to be set for different recipients. It is too complicated to keep
4894 track of this in the cache, so now Exim always runs a verification when a
4895 callout is required, relying on the callout cache for the optimization.
4896 The overhead is duplication of the address routing, but this should not be
4899 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4900 to be verified with callout=postmaster, and the main callout worked but the
4901 postmaster check failed, the verification correctly failed. However, if a
4902 subsequent RCPT command asked for sender verification *without* the
4903 postmaster check, incorrect caching caused this verification also to fail,
4906 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4907 it was not caching the DNS options (qualify_single, search_parents) that
4908 were used when the lookup failed. A subsequent lookup with different
4909 options therefore always gave the same answer, though there were cases
4910 where it should not have. (Example: a "domains = !$mx_any" option on a
4911 dnslookup router: the "domains" option is always processed without any
4912 widening, but the router might have qualify_single set.) Now Exim uses the
4913 cached value only when the same options are set.
4915 32. Added John Jetmore's "exipick" utility to the distribution.
4917 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4918 than a timeout (e.g. a certificate is required, and is not provided), an
4919 Exim server now closes the connection immediately. Previously it waited for
4920 the client to close - but if the client is SSL, it seems that they each
4921 wait for each other, leading to a delay before one of them times out.
4923 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4924 maintained 0.8.x compatibility because I don't think many are using it, and
4925 it is clearly obsolete.
4927 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4930 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4931 client certificate was expired. A simple patch fixes this, though I don't
4932 understand the full logic of why the verify callback is called multiple
4935 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4936 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4937 which causes problems with some clients (such as the Certicom SSL Plus
4938 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4939 disables the coutermeasure allowing Eudora to connect."
4941 38. Exim was not checking that a write() to a log file succeeded. This could
4942 lead to Bad Things if a log got too big, in particular if it hit a file
4943 size limit. Exim now panics and dies if it cannot write to a log file, just
4944 as it does if it cannot open a log file.
4946 39. Modified OS/Makefile-Linux so that it now contains
4948 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4950 The two -D definitions ensure that Exim is compiled with large file
4951 support, which makes it possible to handle log files that are bigger than
4954 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4955 instance) a domain was checked against a named list that involved a lookup,
4956 causing $domain_data to be set, then another domain was checked against the
4957 same list, then the first domain was re-checked, the value of $domain_data
4958 after the final check could be wrong. In particular, if the second check
4959 failed, it could be set empty. This bug probably also applied to
4962 41. The strip_trailing_dot option was not being applied to the address given
4963 with the -f command-line option.
4965 42. The code for reading a message's header from the spool was incrementing
4966 $received_count, but never initializing it. This meant that the value was
4967 incorrect (doubled) while delivering a message in the same process in which
4968 it was received. In the most common configuration of Exim, this never
4969 happens - a fresh exec is done - but it can happen when
4970 deliver_drop_privilege is set.
4972 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4973 it now includes up to 150 characters of the unexpected data in the log
4976 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4977 and building data strings. The size of both of these buffers was 10 000
4978 bytes - far larger than anybody would *ever* want, thought I. Needless to
4979 say, somebody hit the limit. I have increased the maximum line length to
4980 20 000 and the maximum data length of concatenated lines to 100 000. I have
4981 also fixed two bugs, because there was no checking on these buffers. Tsk,
4982 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4983 buffer is too small.
4985 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4986 lsearch lookups. Now it does.
4988 46. When parsing a route_list item in a manualroute router, a fixed-length
4989 buffer was used for the list of hosts. I made this 1024 bytes long,
4990 thinking that nobody would ever have a list of hosts that long. Wrong.
4991 Somebody had a whole pile of complicated expansion conditions, and the
4992 string was silently truncated, leading to an expansion error. It turns out
4993 that it is easier to change to an unlimited length (owing to other changes
4994 that have happened since this code was originally written) than to build
4995 structure for giving a limitation error. The length of the item that
4996 expands into the list of hosts is now unlimited.
4998 47. The lsearch lookup could not handle data where the length of text line was
4999 more than 4095 characters. Such lines were truncated, leading to shortened
5000 data being returned. It should now handle lines of any length.
5002 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5003 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5006 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5007 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5008 no longer happens in comment lines. A list of replacements is now placed
5009 at the head of all of the source files, except those whose only change is
5010 to replace PERL_COMMAND in the very first #! line.
5012 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5013 messages on the queue, with a bottom-up merge sort, using code contributed
5014 by Michael Haardt. This should make operations like -bp somewhat faster on
5015 large queues. It won't affect queue runners, except when queue_run_in_order
5018 51. Installed eximstats 1.31 in the distribution.
5020 52. Added support for SRV lookups to the dnslookup router.
5022 53. If an ACL referred to $message_body or $message_body_end, the value was not
5023 reset for any messages that followed in the same SMTP session.
5025 54. The store-handling optimization for building very long strings was not
5026 differentiating between the different store pools. I don't think this
5027 actually made any difference in practice, but I've tidied it.
5029 55. While running the routers to verify a sender address, $sender_address
5030 was still set to the sender address. This is wrong, because when routing to
5031 send a bounce to the sender, it would be empty. Therefore, I have changed
5032 it so that, while verifying a sender address, $sender_address is set to <>.
5033 (There is no change to what happens when verifying a recipient address.)
5035 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5036 target A or AAAA records (if not already returned) without resetting the
5037 qualify_single or search_parents options of the DNS resolver. These are
5038 inappropriate in this case because the targets of MX and SRV records must
5039 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5040 target that, when qualified, matched something in the local domain. These
5041 two options are now turned off when doing these lookups.
5043 57. It seems that at least some releases of Reiserfs (which does not have the
5044 concept of a fixed number of inodes) returns zero and not -1 for the
5045 number of available inodes. This interacted badly with check_spool_inodes,
5046 which assumed that -1 was the "no such thing" setting. What I have done is
5047 to check that the total number of inodes is greater than zero before doing
5048 the test of how many are available.
5050 58. When a "warn" ACL statement has a log_message modifier, the message is
5051 remembered, and not repeated. This is to avoid a lot of repetition when a
5052 message has many recipients that cause the same warning to be written.
5053 Howewer, Exim was preserving the list of already written lines for an
5054 entire SMTP session, which doesn't seem right. The memory is now reset if a
5055 new message is started.
5057 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5058 debugging output unless log_rewrite was also set.
5060 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5061 of (int)(handle) when we know that handle contains (void *)(-1).
5063 61. The Exim daemon panic-logs an error return when it closes the incoming
5064 connection. However "connection reset by peer" seems to be common, and
5065 isn't really an error worthy of noting specially, so that particular error
5068 62. When Exim is trying to find all the local interfaces, it used to panic and
5069 die if the ioctl to get the interface flags failed. However, it seems that
5070 on at least one OS (Solaris 9) it is possible to have an interface that is
5071 included in the list of interfaces, but for which you get a failure error
5072 for this call. This happens when the interface is not "plumbed" into a
5073 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5074 failure of the "get flags" call assumes that the interface is down.
5076 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5077 makes life easier for people who are doing arithmetic on fields extracted
5078 from dates, where you often get leading zeros that should not be
5079 interpreted as octal.
5081 64. Added qualify_domain to the redirect router, to override the global
5084 65. If a pathologically long header line contained very many addresses (the
5085 report of this problem mentioned 10 000) and each of them was rewritten,
5086 Exim could use up a very large amount of memory. (It kept on making new
5087 copies of the header line as it rewrote, and never released the old ones.)
5088 At the expense of a bit more processing, the header rewriting function has
5089 been changed so that it no longer eats memory in this way.
5091 66. The generation of the Received: header has been moved from the time that a
5092 message starts to be received, to the time that it finishes. The timestamp
5093 in the Received: header should now be very close to that of the <= log
5094 line. There are two side-effects of this change:
5096 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5097 logged header lines no longer include the local Received: line, because
5098 it has not yet been created. The same applies to a copy of the message
5099 that is returned to a non-SMTP sender when a message is rejected.
5101 (b) When a filter file is tested using -bf, no additional Received: header
5102 is added to the test message. After some thought, I decided that this
5105 This change does not affect the value of $received_for. It is still set
5106 after address rewriting, but before local_scan() is called.
5108 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5110 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5111 gave an unhelpful panic error message, and a defer error. I have managed to
5112 change this behaviour so that it now rejects any supplied certificate,
5113 which seems right, as the list of acceptable certificates is empty.
5115 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5116 gave an unhelpful defer error. I have not managed to make this reject any
5117 supplied certificates, but the error message it gives is "no certificate
5118 supplied", which is not helpful.
5120 70. exigrep's output now also includes lines that are not associated with any
5121 message, but which match the given pattern. Implemented by a patch from
5122 Martin Sluka, which also tidied up the Perl a bit.
5124 71. Recipient callout verification, like sender verification, was using <> in
5125 the MAIL FROM command. This isn't really the right thing, since the actual
5126 sender may affect whether the remote host accepts the recipient or not. I
5127 have changed it to use the actual sender in the callout; this means that
5128 the cache record is now keyed on a recipient/sender pair, not just the
5129 recipient address. There doesn't seem to be a real danger of callout loops,
5130 since a callout by the remote host to check the sender would use <>.
5131 [SEE ABOVE: changed after hitting problems.]
5133 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5134 temporary errors. However, in the case of such a code being given after
5135 the end of a data transmission (i.e. after ".") Exim was failing to write
5136 a retry record for the message. (Yes, there was some broken host that was
5137 actually sending 8xx at this point.)
5139 73. An unknown lookup type in a host list could cause Exim to panic-die when
5140 the list was checked. (An example that provoked this was putting <; in the
5141 middle of a list instead of at the start.) If this happened during a DATA
5142 ACL check, a -D file could be left lying around. This kind of configuration
5143 error no longer causes Exim to die; instead it causes a defer errror. The
5144 incident is still logged to the main and panic logs.
5146 74. Buglet left over from Exim 3 conversion. The message "too many messages
5147 in one connection" was written to the rejectlog but not the mainlog, except
5148 when address rewriting (yes!) was being logged.
5150 75. Added write_rejectlog option.
5152 76. When a system filter was run not as root (that is, when system_filter_user
5153 was set), the values of the $n variables were not being returned to the
5154 main process; thus, they were not subsequently available in the $sn
5157 77. Added +return_path_on_delivery log selector.
5159 78. A connection timeout was being treated differently from recipients deferred
5160 when testing hosts_max_try with a message that was older than the host's
5161 retry timeout. (The host should not be counted, thus allowing all hosts to
5162 be tried at least once before bouncing.) This may have been the cause of an
5163 occasionally reported bug whereby a message would remain on the queue
5164 longer than the retry timeout, but would be bounced if a delivery was
5165 forced. I say "may" because I never totally pinned down the problem;
5166 setting up timeout/retry tests is difficult. See also the next item.
5168 79. The ultimate address timeout was not being applied to errors that involved
5169 a combination of host plus message (for example, a timeout on a MAIL
5170 command). When an address resolved to a number of possible hosts, and they
5171 were not all tried for each delivery (e.g. because of hosts_max_try), a
5172 message could remain on the queue longer than the retry timeout.
5174 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5177 81. Fixed an obscure SMTP outgoing bug which required at least the following
5178 conditions: (a) there was another message waiting for the same server;
5179 (b) the server returned 5xx to all RCPT commands in the first message so
5180 that the message was not completed; (c) the server dropped the connection
5181 or gave a negative response to the RSET that Exim sends to abort the
5182 transaction. The observed case was a dropped connection after DATA that had
5183 been sent in pipelining mode. That is, the server had advertised PIPELINING
5184 but was not implementing it correctly. The effect of the bug was incorrect
5185 behaviour, such as trying another host, and this could lead to a crash.
5191 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5192 and daemon.c were passed as pointers to ints; they should have been
5193 pointers to socklen_t variables (which are typically unsigned ints).
5195 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5198 3. Fixed a really odd bug that affected only the testing scheme; patching a
5199 certain fixed string in the binary changed the value of another string that
5200 happened to be identical to the end of the original first string.
5202 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5203 name", it returns that address as the IP address. On some operating
5204 systems (e.g. Solaris), it also passes back the IP address string as the
5205 "host name". However, on others (e.g. Linux), it passes back an empty
5206 string. Exim wasn't checking for this, and was changing the host name to an
5207 empty string, assuming it had been canonicized.
5209 5. Although rare, it is permitted to have more than one PTR record for a given
5210 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5211 all the names associated with an address, because they do in Solaris.
5212 However, it seems that they do not in Linux for data that comes from the
5213 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5214 I found this out when I moved to a new Linux workstation and tried to run
5215 the Exim test suite.
5217 To get round this problem I have changed the code so that it now does its
5218 own call to the DNS to look up PTR records when searching for a host name.
5219 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5220 addresses that are only in /etc/hosts are still found.
5222 This behaviour is, however, controlled by an option called host_lookup_
5223 order, which defaults to "bydns:byaddr". If people want to use the other
5224 order, or indeed, just use one or the other means of lookup, they can
5225 specify it in this variable.
5227 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5228 some operating systems, this comes back from gethostbyaddr() as an empty
5229 string, and this is what Exim used to test for. However, it seems that in
5230 other systems, "." is yielded. Exim now tests for this case too.
5232 7. The values of check_spool_space and check_log_space are now held internally
5233 as a number of kilobytes instead of an absolute number of bytes. If a
5234 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5235 kilobyte. This means that much larger values can be stored.
5237 8. Exim monitor: an attempt to get the action menu when not actually pointing
5238 at a message produces an empty menu entitled "No message selected". This
5239 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5240 no entries in it ("Shell widget menu has zero width and/or height"). So I
5241 have added a single, blank menu entry in this case.
5243 9. Added ${quote_local_part.
5245 10. MIME decoding is now applied to the contents of Subject: header lines when
5248 11. Now that a reference to $sender_host_address automatically causes a reverse
5249 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5250 host lookup before query-style lookups in lists that might use this
5251 variable. This has therefore been abolished, and the "net-" prefix is no
5252 longer necessary for query-style lookups.
5254 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5255 have been a typo for LFLAGS, so it has been changed.
5257 13. The install script calls Exim with "-C /dev/null" in order to find the
5258 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5259 to be output. Howeve, since Exim outputs its version number before the
5260 error, it didn't break the script. It just looked ugly. I fixed this by
5261 always allowing "-C /dev/null" if the caller is root.
5263 14. Ignore overlarge ACL variable number when reading spool file - insurance
5264 against a later release with more variables having written the file.
5266 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5267 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5268 this, as well as the form without the "IPv6" on the front.
5270 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5271 OS/Makefile-Darwin file.
5273 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5274 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5276 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5277 variable. However, some OS use ints for inodes. I've added cast to long int
5278 to get rid of the compiler warning.
5280 19. I had forgotten to lock out "/../" in configuration file names when
5281 ALT_CONFIG_PREFIX was set.
5283 20. Routers used for verification do not need to specify transports. However,
5284 if such a router generated a host list, and callout was configured, Exim
5285 crashed, because it could not find a port number from the (non-existent)
5286 transport. It now assumes port 25 in this circumstance.
5288 21. Added the -t option to exigrep.
5290 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5291 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5292 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5293 them is set, LOOKUP_LSEARCH is forced.
5295 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5297 24. Added sender and host information to the "rejected by local_scan()" log
5298 line; previously there was no indication of these.
5300 25. Added .include_if_exists.
5302 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5303 Linux. It turns out that not all file systems support this. Apparently some
5304 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5305 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5306 not supported on the file descriptor, is now ignored when Exim is trying to
5307 sync a directory. This applies only to Linux.
5309 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5311 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5312 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5315 29. Added support for maildirsize files from supplied patch (modified a bit).
5317 30. The use of :fail: followed by an empty string could lead Exim to respond to
5318 sender verification failures with (e.g.):
5320 550 Verification failed for <xxx>
5321 550 Sender verify failed
5323 where the first response line was missing the '-' that indicates it is not
5324 the final line of the response.
5326 31. The loop for finding the name of the user that called Exim had a hardwired
5327 limit of 10; it now uses the value of finduser_retries, which is used for
5328 all other user lookups.
5330 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5333 33. Exim was neglecting to zero errno before one call of strtol() when
5334 expanding a string and expecting an integer value. On some systems this
5335 resulted in spurious "integer overflow" errors. Also, it was casting the
5336 result into an int without checking.
5338 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5339 did not work. The code looks as if it has *never* worked, though it appears
5340 to have been documented since at least releast 1.62. I have made it work.
5342 35. The "timeout_DNS" error in retry rules, also documented since at least
5343 1.62, also never worked. As it isn't clear exactly what this means, and
5344 clearly it isn't a major issue, I have abolished the feature by treating it
5345 as "timeout", and writing a warning to the main and panic logs.
5347 36. The display of retry rules for -brt wasn't always showing the error code
5350 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5351 timeout_connect_A, timeout_connect_MX.
5353 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5354 to the empty sender.
5356 39. The daemon was not analysing the content of -oX till after it had closed
5357 stderr and disconnected from the controlling terminal. This meant that any
5358 syntax errors were only noted on the panic log, and the return code from
5359 the command was 0. By re-arranging the code a little, I've made the
5360 decoding happen first, so such errors now appear on stderr, and the return
5361 code is 1. However, the actual setting up of the sockets still happens in
5362 the disconnected process, so errors there are still only recorded on the
5365 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5366 connections (as happens on some IP stacks) was logged at start up time as
5367 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5368 it from "IPv6 and IPv4", which means that two separate sockets are being
5371 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5372 says whether AF_INET or AF_INET6 was passed as an argument.
5374 42. Exiwhat output was messed up when time zones were included in log
5377 43. Exiwhat now gives more information about the daemon's listening ports,
5378 and whether -tls-on-connect was used.
5380 44. The "port" option of the smtp transport is now expanded.
5382 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5383 silently ignored. Now an error message is written to the main and panic
5386 46. There's a new ACL modifier called "logwrite" which writes to a log file
5387 as soon as it is encountered.
5389 47. Added $local_user_uid and $local_user_gid at routing time.
5391 48. Exim crashed when trying to verify a sender address that was being
5394 49. Exim was recognizing only a space character after ".include". It now also
5395 recognizes a tab character.
5397 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5398 extracting the relevant information from the specification. The man page no
5399 longer contains scrambled data for the -d option, and I've added a section
5400 at the front about calling Exim under different names.
5402 51. Added "extra_headers" argument to the "mail" command in filter files.
5404 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5407 53. Installed eximstats 1.29.
5409 54. Added transport_filter_timeout as a generic transport option.
5411 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5412 Cc: header lines. This was required by RFC 822, but it not required by RFC
5415 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5416 incoming messages that did not have them. Now it does so only if the
5417 message originates locally, that is, if there is no associated remote host
5418 address. When Resent- header lines are present, this applies to the Resent-
5419 lines rather than the non-Resent- lines.
5421 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5422 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5424 58. Messages for configuration errors now include the name of the main
5425 configuration file - useful now that there may be more than one file in a
5426 list (.included file names were always shown).
5428 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5429 for those rare installations that do not start the daemon as root or run it
5430 setuid root. I've cut out the call to initgroups() if the daemon is not
5433 60. The Exim user and group can now be bound into the binary as text strings
5434 that are looked up at the start of Exim's processing.
5436 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5438 62. Added $mailstore_basename variable.
5440 63. Installed patch to sieve.c from Michael Haardt.
5442 64. When Exim failed to open the panic log after failing to open the main log,
5443 the original message it was trying to log was written to stderr and debug
5444 output, but if they were not available (the usual case in production), it
5445 was lost. Now it is written to syslog before the two lines that record the
5446 failures to open the logs.
5448 65. Users' Exim filters run in subprocesses under the user's uid. It is
5449 possible for a "deliver" command or an alias in a "personal" command to
5450 provoke an address rewrite. If logging of address rewriting is configured,
5451 this fails because the process is not running as root or exim. There may be
5452 a better way of dealing with this, but for the moment (because 4.30 needs
5453 to be released), I have disabled address rewrite logging when running a
5454 filter in a non-root, non-exim process.
5460 1. The buildconfig auxiliary program wasn't quoting the value set for
5461 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5462 not defined. This bug was masked in 4.22 by the effect that was fixed in
5465 2. Some messages that were rejected after a message id was allocated were
5466 shown as "incomplete" by exigrep. It no longer does this for messages that
5467 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5469 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5470 have allow_domain_literals set, the ID did not get logged in the <= line.
5471 Domain literals are now always recognized in Message-ID: header lines.
5473 4. The first argument for a ${extract expansion item is the key name or field
5474 number. Leading and trailing spaces in this item were not being ignored,
5475 causing some misleading effects.
5477 5. When deliver_drop_privilege was set, single queue runner processes started
5478 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5479 same command in the process it spins off) were not dropping privilege.
5481 6. When the daemon running as "exim" started a queue runner, it always
5482 re-executed Exim in the spun-off process. This is a waste of effort when
5483 deliver_drop_privilege is set. The new process now just calls the
5484 queue-runner function directly.
5490 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5493 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5494 ignored. Though the use of -oP was forcing the writing of a pid file, it
5495 was always written to the default place.
5497 3. If the message "no IP address found for host xxxx" is generated during
5498 incoming verification, it is now followed by identification of the incoming
5499 connection (so you can more easily find what provoked it).
5501 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5503 5. Added some features to "harden" Exim a bit more against certain attacks:
5505 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5506 be put in Local/Makefile. This is like the never_users runtime option,
5507 but it cannot be overridden. The default setting is "root".
5509 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5510 prefix string with which any file named in a -C command line option
5513 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5514 is retained for -C and -D only if the caller of Exim is root. Without
5515 it, the exim user may also use -C and -D and retain privilege.
5517 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5518 command line option is disabled.
5520 6. Macro names set by the -D option must start with an upper case letter, just
5521 like macro names defined in the configuration file.
5523 7. Added "dereference=" facility to LDAP.
5525 8. Two instances of the typo "uknown" in the source files are fixed.
5527 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5528 the Configure-Makefile script screwed up while processing it.
5530 10. Incorporated PCRE 4.4.
5532 11. The SMTP synchronization check was not operating right at the start of an
5533 SMTP session. For example, it could not catch a HELO sent before the client
5534 waited for the greeting. There is now a check for outstanding input at the
5535 point when the greeting is written. Because of the duplex, asynchronous
5536 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5537 way, but not yet received, when the check is performed.
5539 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5540 on TCP/IP sockets, because this apparently causes some broken clients to
5543 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5544 unchanged) from the Cygwin maintainer.
5546 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5547 was supported instead of testing for mbx. Effectively a typo.
5549 15. The spa authenticator server code was not checking that the input it
5550 received was valid base64.
5552 16. The debug output line for the "set" modifier in ACLs was not showing the
5553 name of the variable that was being set.
5555 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5557 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5558 lookup on its own. Something else was needed to trigger the lookup. For
5559 example, a match in host_lookup or the need for a host name in a host list.
5560 Now, if $sender_host_name is referenced and the host name has not yet been
5561 looked up, a lookup is performed. If the lookup fails, the variable remains
5562 empty, and $host_lookup_failed is set to "1".
5564 19. Added "eqi" as a case-independent comparison operator.
5566 20. The saslauthd authentication condition could segfault if neither service
5567 nor realm was specified.
5569 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5570 error message that was logged was misleading, and incoming SMTP
5571 connections were dropped. The message is now more accurate, and temporary
5572 errors are given to SMTP connections.
5574 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5575 (except RSET) with a 421 error, until QUIT is received. However, it was
5576 failing to send a response to QUIT.
5578 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5579 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5580 ACL is now run after the helo_try_verify_hosts code.
5582 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5583 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5584 case-independent, so other case variants are also recognized). Apparently
5585 some systems use these upper case variants.
5587 25. If more than two messages were waiting for the same host, and a transport
5588 filter was specified for the transport, Exim sent two messages over the
5589 same TCP/IP connection, and then failed with "socket operation on non-
5590 socket" when it tried to send the third.
5592 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5594 27. The extern definition of crypt16() in expand.c was not being excluded when
5595 the OS had its own crypt16() function.
5597 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5598 as a preferred synonym for return_size_limit, both as an option and as an
5601 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5603 30. Changed the default configuration ACL to relax the local part checking rule
5604 for addresses that are not in any local domains. For these addresses,
5605 slashes and pipe symbols are allowed within local parts, but the sequence
5606 /../ is explicitly forbidden.
5608 31. SPA server authentication was not clearing the challenge buffer before
5611 32. log_message in a "warn" ACL statement was writing to the reject log as
5612 well as to the main log, which contradicts the documentation and doesn't
5613 seem right (because no rejection is happening). So I have stopped it.
5615 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5616 However, I am unable to do any testing of this.
5618 34. Fixed an infelicity in the appendfile transport. When checking directories
5619 for a mailbox, to see if any needed to be created, it was accidentally
5620 using path names with one or more superfluous leading slashes; tracing
5621 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5623 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5624 contents are added to the log line that is written for every discarded
5625 recipient. (Previously a log_message setting was ignored.)
5627 36. The ${quote: operator now quotes the string if it is empty.
5629 37. The install script runs exim in order to find its version number. If for
5630 some reason other than non-existence or emptiness, which it checks, it
5631 could not run './exim', it was installing it with an empty version number,
5632 i.e. as "exim-". This error state is now caught, and the installation is
5635 38. An argument was missing from the function that creates an error message
5636 when Exim fails to connect to the socket for saslauthd authentication.
5637 This could cause Exim to crash, or give a corrupted message.
5639 39. Added isip, isip4, and isip6 to ${if conditions.
5641 40. The ACL variables $acl_xx are now saved with the message, and can be
5642 accessed later in routers, transports, and filters.
5644 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5645 strings in the file are not string-expanded.
5647 42. If a MAIL command specified a SIZE value that was too large to fit into an
5648 int variable, the check against message_size_limit failed. Such values are
5649 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5650 day this will have to be increased, but I don't think I want to be around
5651 when emails are that large.
5658 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5659 iconv() is not standard in FreeBSD.
5661 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5662 IPv6 enabled. The observed symptom was a segmentation fault on return from
5663 the function os_common_find_running_interfaces() in src/os.c.
5665 3. In the check_special_case() function in daemon.c I had used "errno" as an
5666 argument name, which causes warnings on some systems. This was basically a
5667 typo, since it was named "eno" in the comments!
5669 4. The code that waits for the clock to tick (at a resolution of some fraction
5670 of a second) so as to ensure message-id uniqueness was always waiting for
5671 at least one whole tick, when it could have waited for less. [This is
5672 almost certainly not relevant at current processor speeds, where it is
5673 unlikely to ever wait at all. But we try to future-proof.]
5675 5. The function that sleeps for a time interval that includes fractions of a
5676 second contained a race. It did not block SIGALRM between setting the
5677 timer, and suspending (a couple of lines later). If the interval was short
5678 and the sigsuspend() was delayed until after it had expired, the suspension
5679 never ended. On busy systems this could lead to processes getting stuck for
5682 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5683 process, before it forks any delivery processes. The open lookup caching
5684 mechanism meant that the open file or database connection was passed into
5685 the delivery process. The problem was that delivery processes always tidy
5686 up cached lookup data. This could cause a problem for the next delivery
5687 process started by the queue runner, because the external queue runner
5688 process does not know about the closure. So the next delivery process
5689 still has data in the lookup cache. In the case of a file lookup, there was
5690 no problem because closing a file descriptor in a subprocess doesn't affect
5691 the parent. However, if the lookup was caching a connection to a database,
5692 the connection was closed, and the second delivery process was likely to
5693 see errors such as "PGSQL: query failed: server closed the connection
5694 unexpectedly". The problem has been fixed by closing all cached lookups
5695 in a queue runner before running a delivery process.
5697 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5698 seem to have the "const" qualifier which it has on other OS. I've
5701 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5702 *of the same type* (client or server) with the same public name that an
5703 error should be diagnosed.
5705 9. When Exim looked up a host name for an IP address, but failed to find the
5706 original IP address when looking up the host name (a safety check), it
5707 output the message "<ip address> does not match any IP for NULL", which was
5708 confusing, to say the least. The bug was that the host name should have
5709 appeared instead of "NULL".
5711 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5712 user that is built into the binary, and the -C or -D options is used, root
5713 privilege is dropped before the configuration file is read. In addition,
5714 logging is switched to stderr instead of the normal log files. If the
5715 configuration then re-defines the Exim user, the unprivileged environment
5716 is probably not what is expected, so Exim logs a panic warning message (but
5719 However, if deliver_drop_privilege is set, the unprivileged state may well
5720 be exactly what is intended, so the warning has been cut out in that case,
5721 and Exim is allowed to try to write to its normal log files.
5727 1. smtp_return_error_details was not giving details for temporary sender
5728 or receiver verification errors.
5730 2. Diagnose a configuration error if two authenticators have the same public
5733 3. Exim used not to create the message log file for a message until the first
5734 delivery attempt. This could be confusing when incoming messages were held
5735 for policy or load reasons. The message log file is now created at the time
5736 the message is received, and an initial "Received" line is written to it.
5738 4. The automatically generated man page for command line options had a minor
5739 bug that caused no ill effects; however, a more serious problem was that
5740 the procedure for building the man page automatically didn't always
5741 operate. Consequently, release 4.20 contains an out-of-date version. This
5742 shouldn't happen again.
5744 5. When building Exim with embedded Perl support, the script that builds the
5745 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5746 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5748 6. The freeze_tell option was not being used for messages that were frozen on
5749 arrival, either by an ACL or by local_scan().
5751 7. Added the smtp_incomplete_transaction log selector.
5753 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5754 was accepting AUTH without a new EHLO.
5756 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5757 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5760 10. Logging of TCP/IP connections (when configured) now happens in the main
5761 daemon process instead of the child process, so that the TCP/IP connection
5762 count is more accurate (but it can never be perfect).
5764 11. The use of "drop" in a nested ACL was not being handled correctly in the
5765 outer ACL. Now, if condition failure induced by the nested "drop" causes
5766 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5767 or "require"), the connection is dropped.
5769 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5770 that yield "discard" can now be used with an "accept" or a "discard" verb,
5771 but an error is generated for any others (because I can't see a useful way
5772 to define what should happen).
5774 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5775 are now processed in the same way as lines in the Exim configuration file.
5776 In particular, continuation lines are supported.
5778 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5780 15. Added -ti meaning -t -i.
5782 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5783 domains, and warn by logging if others are found.
5785 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5786 returned from GIFCONF when Exim is trying to find the list of interfaces on
5787 a host. The code in os.c has been modified to copy each ifreq to an aligned
5788 structure in all cases.
5790 Also, in some cases, the returned ifreq's were being copied to a 'struct
5791 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5792 means the last couple of bytes of an IPv6 address could be chopped if the
5793 ifreq contained only a normal sockaddr (14 bytes storage).
5795 18. Named domain lists were not supported in the hosts_treat_as_local option.
5796 An entry such as +xxxx was not recognized, and was treated as a literal
5799 19. Ensure that header lines added by a DATA ACL are included in the reject log
5800 if the ACL subsequently rejects the message.
5802 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5803 MD5 (which is deprecated).
5805 21. When testing a filter file using -bf, Exim was writing a message when it
5806 took the sender from a "From " line in the message, but it was not doing so
5807 when it took $return_path from a Return-Path: header line. It now does.
5809 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5810 with a valid header line field name (a series of printing characters
5811 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5813 23. Changed "disc" in the source to "disk" to conform to the documentation and
5814 the book and for uniformity.
5816 24. Ignore Sendmail's -Ooption=value command line item.
5818 25. When execve() failed while trying to run a command in a pipe transport,
5819 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5820 could be confused with a return value of 69 from the command itself. This
5821 has been changed to 127, the value the shell returns if it is asked to run
5822 a non-existent command. The wording for the related log line suggests a
5823 non-existent command as the problem.
5825 26. If received_header_text expands to an empty string, do not add a Received:
5826 header line to the message. (Well, it adds a token one on the spool, but
5827 marks it "old" so that it doesn't get used or transmitted.)
5829 27. Installed eximstats 1.28 (addition of -nt option).
5831 28. There was no check for failure on the call to getsockname() in the daemon
5832 code. This can fail if there is a shortage of resources on the system, with
5833 ENOMEM, for example. A temporary error is now given on failure.
5835 29. Contrary to the C standard, it seems that in some environments, the
5836 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5837 program. Exim now does this explicitly; it affects the formatting of
5838 timestamps using strftime().
5840 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5841 complaints. I've now initialized all the variables, to avoid this.
5843 32. Header lines added by a system filter were not being "seen" during
5844 transport-time rewrites.
5846 33. The info_callback() function passed to OpenSSL is set up with type void
5847 (*)(SSL *, int, int), as described somewhere. However, when calling the
5848 function (actually a macro) that sets it up, the type void(*)() is
5849 expected. I've put in a cast to prevent warnings from picky compilers.
5851 34. If a DNS black list lookup found a CNAME record, but there were no A
5852 records associated with the domain it pointed at, Exim crashed.
5854 35. If a DNS black list lookup returned more than one A record, Exim ignored
5855 all but the first. It now scans all returned addresses if a particular IP
5856 value is being sought. In this situation, the contents of the
5857 $dnslist_value variable are a list of all the addresses, separated by a
5860 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5861 to accept a host name and all its aliases if the forward lookup for any of
5862 them yielded the IP address of the incoming connection. Now it accepts only
5863 those names whose forward lookup yields the correct IP address. Any other
5864 names are discarded. This closes a loophole whereby a rogue DNS
5865 administrator could create reverse DNS records to break through a
5866 wildcarded host restriction in an ACL.
5868 37. If a user filter or a system filter that ran in a subprocess used any of
5869 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5870 the wrong values were passed to the pipe command ($thisaddress had the
5871 value of $0, $0 had the value of $1, etc). This bug was introduced by
5872 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5874 38. Improved the line breaking for long SMTP error messages from ACLs.
5875 Previously, if there was no break point between 40 and 75 characters, Exim
5876 left the rest of the message alone. Two changes have been made: (a) I've
5877 reduced the minimum length to 35 characters; (b) if it can't find a break
5878 point between 35 and 75 characters, it looks ahead and uses the first one
5879 that it finds. This may give the occasional overlong line, but at least the
5880 remaining text gets split now.
5882 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5883 file descriptors might be low, and that setting 1000 would always raise it.
5884 It turns out that in some environments, the limit is already over 1000 and
5885 that lowering it causes trouble. So now Exim takes care not to decrease it.
5887 40. When delivering a message, the value of $return_path is set to $sender_
5888 address at the start of routing (routers may change the value). By an
5889 oversight, this default was not being set up when an address was tested by
5890 -bt or -bv, which affected the outcome if any router or filter referred to
5893 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5894 log only when "message" or "log_message" are set. However, if one of the
5895 conditions was an address verification, or a call to a nested ACL, the
5896 messages generated by the underlying test were being passed through. This
5897 no longer happens. The underlying message is available in $acl_verify_
5898 message for both "message" and "log_message" expansions, so it can be
5899 passed through if needed.
5901 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5902 new expansion $bh_ to give the encoded byte string without charset
5903 translation. Translation happens only if iconv() is available; HAVE_ICONV
5904 indicates this at build time. HEADERS_CHARSET gives the charset to
5905 translate to; headers_charset can change it in the configuration, and
5906 "headers charset" can change it in an individual filter file.
5908 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5909 that creates RFC 2047 encoded "words" labels them as that charset instead
5910 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5911 expansion operator; (ii) when Exim creates a From: line for a local
5912 message; (iii) when a header line is rewritten to include a "phrase" part.
5914 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5915 buggy, causing it to skip the first lines of messages whose message ID
5916 ended in 'D'. This would not have bitten before Exim release 4.14, because
5917 message IDs were unlikely to end in 'D' before then. The effect was to have
5918 incorrect size information for certain domains.
5920 45. #include "config.h" was missing at the start of the crypt16.c module. This
5921 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5924 46. If there was a timeout during a "random" callout check, Exim treated it as
5925 a failure of the random address, and carried on sending RSET and the real
5926 address. If the delay was just some slowness somewhere, the response to the
5927 original RCPT would be taken as a response to RSET and so on, causing
5928 mayhem of various kinds.
5930 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5931 when I implemented it. It didn't allow for the fact that some option values
5932 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5933 the right test for positive values.
5935 48. Domain names in DNS records are case-independent. Exim always looks them up
5936 in lower case. Some resolvers return domain names in exactly the case they
5937 appear in the zone file, that is, they may contain uppercase letters. Not
5938 all resolvers do this - some return always lower case. Exim was treating a
5939 change of case by a resolver as a change of domain, similar to a widening
5940 of a domain abbreviation. This triggered its re-routing code and so it was
5941 trying to route what was effectively the same domain again. This normally
5942 caused routing to fail (because the router wouldn't handle the domain
5943 twice). Now Exim checks for this case specially, and just changes the
5944 casing of the domain that it ultimately uses when it transmits the message
5947 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5950 50. If a filter generated a file delivery with a non-absolute name (possible if
5951 no home directory exists for the router), the forbid_file option was not
5954 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5955 the existing equality matching.
5957 52. Exim was using ints instead of ino_t variables in some places where it was
5958 dealing with inode numbers.
5960 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5961 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5962 called TMPDIR, and if it finds it is different, it changes its value.
5964 54. The smtp_printf() function is now made available to local_scan() so
5965 additional output lines can be written before returning. There is also an
5966 smtp_fflush() function to enable the detection of a dropped connection.
5967 The variables smtp_input and smtp_batched_input are exported to
5970 55. Changed the default runtime configuration: the message "Unknown user"
5971 has been removed from the ACL, and instead placed on the localuser router,
5972 using the cannot_route_message feature. This means that any verification
5973 failures that generate their own messages won't get overridden. Similarly,
5974 the "Unrouteable address" message that was in the ACL for unverifiable
5975 relay addresses has also been removed.
5977 56. Added hosts_avoid_esmtp to the smtp transport.
5979 57. The exicyclog script was not checking for the esoteric option
5980 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5981 will work only if exicyclog is run under the appropriate euid.
5983 58. Following a discussion on the list, the rules by which Exim recognises line
5984 endings on incoming messages have been changed. The -dropcr and drop_cr
5985 options are now no-ops, retained only for backwards compatibility. The
5986 following line terminators are recognized: LF CRLF CR. However, special
5987 processing applies to CR:
5989 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5990 nor a local message in the state where . is a terminator.
5992 (ii) If a bare CR is encountered in a header line, an extra space is added
5993 after the line terminator so as not to end the header. The reasoning
5994 behind this is that bare CRs in header lines are most likely either
5995 to be mistakes, or people trying to play silly games.
5997 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5998 was being incorrectly given as 18 bytes larger than it should have been.
5999 This is a VOB (very old bug).
6001 60. This may never have affected anything current, but just in case it has:
6002 When the local host is found other than at the start of a list of hosts,
6003 the local host, those with the same MX, and any that follow, are discarded.
6004 When the list in question was part of a longer list of hosts, the following
6005 hosts (not currently being processed) were also being discarded. This no
6006 longer happens. I'm not sure if this situation could ever has previously
6009 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6012 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6013 header. This is something that is recommended in a new Internet Draft, and
6014 is something that is documented as being done by Sendmail. There are two
6015 possible values. For messages generated by the autoreply transport, Exim
6018 Auto-Submitted: auto-replied
6020 whereas for all other generated messages (e.g. bounces) it adds
6022 Auto-Submitted: auto-generated
6024 63. The "personal" condition in filters now includes a test for the
6025 Auto-Submitted: header. If it contains the string "auto-" the message it
6026 not considered personal.
6028 64. Added rcpt_include_affixes as a generic transport option.
6030 65. Added queue_only_override (default true).
6032 66. Added the syslog_duplication option.
6034 67. If what should have been the first header line of a message consisted of
6035 a space followed by a colon, Exim was mis-interpreting it as a header line.
6036 It isn't of course - it is syntactically invalid and should therefore be
6037 treated as the start of the message body. The misbehaviour could have
6038 caused a number of strange effects, including loss of data in subsequent
6039 header lines, and spool format errors.
6041 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6042 client host had authenticated. This control can now be exercised by an ACL
6043 for more flexibility.
6045 69. By default, callouts do not happen when testing with -bh. There is now a
6046 variant, -bhc, which does actually run the callout code, including
6047 consulting and updating the callout cache.
6049 70. Added support for saslauthd authentication, courtesy of Alexander
6052 71. If statvfs() failed on the spool or log directories while checking their
6053 size for availability, Exim confusingly gave the error "space shortage".
6054 Furthermore, in debugging mode it crashed with a floating point exception.
6055 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6056 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6057 serious problem, Exim now writes to the main and panic logs when this
6058 happens, with details of the failure. It then refuses to accept the
6059 incoming message, giving the message "spool directory problem" or "log
6060 directory problem" with a 421 code for SMTP messages.
6062 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6063 0, 1, and 2 exist, because some OS complain for execs without them (see
6064 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6065 descriptors. However, the code omitted to check that the open succeeded,
6066 causing mysterious errors if for some reason the permissions on /dev/null
6067 got screwed. Now Exim writes a message to the main and panic logs, and
6068 bombs out if it can't open /dev/null.
6070 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6071 interact so that it is all more flexible. It is supposed to remain
6072 backwards compatible. Also added extra_local_interfaces.
6074 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6075 to bomb out with an assertion failure - to the client this appears as a
6076 connection drop. This problem occurs in the part of the code that was taken
6077 from the Samba project. Fortunately, the assertion is in a very simple
6078 function, so I have fixed this by reproducing the function inline in the
6079 one place where it is called, and arranging for authentication to fail
6080 instead of killing the process with assert().
6082 75. The SPA client code was not working when the server requested OEM rather
6083 than Unicode encoding.
6085 76. Added code to make require_files with a specific uid setting more usable in
6086 the case where statting the file as root fails - usually a non-root-mounted
6087 NFS file system. When this happens and the failure is EACCES, Exim now
6088 forks a subprocess and does the per-uid checking as the relevant uid.
6090 77. Added process_log_path.
6092 78. If log_file_path was not explicitly set, a setting of check_log_space or
6093 check_log_inodes was ignored.
6095 79. If a space check for the spool or log partitions fails, the incident is now
6096 logged. Of course, in the latter case the data may get lost...
6098 80. Added the %p formatting code to string_format() so that it can be used to
6099 print addresses in debug_print(). Adjusted all the address printing in the
6100 debugging in store.c to use %p rather than %d.
6102 81. There was a concern that a line of code in smtp_in.c could overflow a
6103 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6104 initially expressed, the concern was not well-founded, because trailing
6105 spaces are removed early. However, if the trailing spaces were followed by
6106 a NULL, they did not get removed, so the overflow was possible. Two fixes
6109 (a) I re-wrote the offending code in a cleaner fashion.
6110 (b) If an incoming SMTP command contains a NULL character, it is rejected
6113 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6114 runs initgroups(), so that if the Exim user is in any additional groups,
6115 they will be used during message reception.
6121 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob