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.
31 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
32 checkers that were complaining about end of non-void function with no
35 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
36 This was a regression intruduced in 4.83 by another bugfix.
38 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
40 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
41 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
47 TF/01 Correctly close the server side of TLS when forking for delivery.
49 When a message was received over SMTP with TLS, Exim failed to clear up
50 the incoming connection properly after forking off the child process to
51 deliver the message. In some situations the subsequent outgoing
52 delivery connection happened to have the same fd number as the incoming
53 connection previously had. Exim would try to use TLS and fail, logging
54 a "Bad file descriptor" error.
56 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
57 utilities have not been installed.
59 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
60 temporary space as the ACL may create new global variables.
62 TL/01 LDAP support uses per connection or global context settings, depending
63 upon the detected version of the libraries at build time.
65 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
66 to extract and use the src ip:port in logging and expansions as if it
67 were a direct connection from the outside internet. PPv2 support was
68 updated based on HAProxy spec change in May 2014.
70 JH/02 Add ${listextract {number}{list}{success}{fail}}.
72 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
73 Properly escape header and check for NULL return.
75 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
78 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
80 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
81 characters in header names, implemented as a verify condition.
82 Contributed by Michael Fischer v. Mollard.
84 TL/05 Rename SPF condition results err_perm and err_temp to standardized
85 results permerror and temperror. Previous values are deprecated but
86 still accepted. In a future release, err_perm and err_temp will be
87 completely removed, which will be a backward incompatibility if the
88 ACL tests for either of these two old results. Patch contributed by
89 user bes-internal on the mailing list.
91 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
93 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
94 selectors, in both main and reject logs.
96 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
99 JH/07 Add malware type "sock" for talking to simple daemon.
101 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
103 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
104 routers/transports under cutthrough routing.
106 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
107 numbers. Touch up "bool" conditional to keep the same definition.
109 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
111 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
113 JH/12 Expand items in router/transport headers_add or headers_remove lists
114 individually rather than the list as a whole. Bug 1452.
116 Required for reasonable handling of multiple headers_ options when
117 they may be empty; requires that headers_remove items with embedded
118 colons must have them doubled (or the list-separator changed).
120 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
121 view the policy declared in the DMARC record. Currently, $dmarc_status
122 is a combined value of both the record presence and the result of the
125 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
127 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
128 dnslookup router and the smtp transport (applying to the forward
131 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
132 of ldap servers used for a specific lookup. Patch provided by Heiko
135 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
136 New variable $lookup_dnssec_authenticated for observability.
138 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
139 Patch submitted by Lars Timman.
141 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
143 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
144 Requires trusted mode and valid format message id, aborts otherwise.
145 Patch contributed by Heiko Schlichting.
147 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
148 certextract with support for various fields. Bug 1358.
150 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
151 is requested by default, modifiable by smtp transport option
154 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
155 operate on certificate variables to give certificate fingerprints
156 Also new ${sha256:cert_variable}.
158 JH/23 The PRDR feature is moved from being Experimental into the mainline.
160 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
161 Christian Aistleitner.
163 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
165 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
166 file. Patch from Wolfgang Breyha.
168 JH/25 Expand the coverage of the delivery $host and $host_address to
169 client authenticators run in verify callout. Bug 1476.
171 JH/26 Port service names are now accepted for tls_on_connect_ports, to
172 align with daemon_smtp_ports. Bug 72.
174 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
175 support and error reporting did not work properly.
177 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
178 and is readable. Patch from Andrew Colin Kissa.
180 TL/14 Enhance documentation of ${run expansion and how it parses the
181 commandline after expansion, particularly in the case when an
182 unquoted variable expansion results in an empty value.
184 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
186 PP/02 Fix internal collision of T_APL on systems which support RFC3123
187 by renaming away from it. Addresses GH issue 15, reported by
190 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
192 TL/15 SECURITY: prevent double expansion in math comparison functions
193 (can expand unsanitized data). Not remotely exploitable.
200 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
202 PP/02 Make -n do something, by making it not do something.
203 When combined with -bP, the name of an option is not output.
205 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
208 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
209 $sender_host_name and config options to manage this, and basic check
212 PP/05 DSCP support for outbound connections and control modifier for inbound.
214 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
215 (Only plugin which currently uses this is kerberos4, which nobody should
216 be using, but we should make it available and other future plugins might
217 conceivably use it, even though it would break NAT; stuff *should* be
218 using channel bindings instead).
220 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
221 name; added for Sendmail compatibility; requires admin caller.
222 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
223 just ignore it); requires trusted caller.
224 Also parse but ignore: -Ac -Am -X<logfile>
227 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
229 TL/02 Add +smtp_confirmation as a default logging option.
231 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
232 Patch by Magnus Holmgren from 2007-02-20.
234 TL/04 Bugzilla 1281 - Spec typo.
235 Bugzilla 1283 - Spec typo.
236 Bugzilla 1290 - Spec grammar fixes.
238 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
240 TL/06 Add Experimental DMARC support using libopendmarc libraries.
242 TL/07 Fix an out of order global option causing a segfault. Reported to dev
243 mailing list by by Dmitry Isaikin.
245 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
247 JH/02 Support "G" suffix to numbers in ${if comparisons.
249 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
251 NM/01 Bugzilla 1197 - Spec typo
252 Bugzilla 1196 - Spec examples corrections
254 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
256 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
257 gnutls_enable_pkcs11, but renamed to more accurately indicate its
260 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
261 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
263 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
264 "acl {{name}{arg}...}", and optional args on acl condition
267 JH/05 Permit multiple router/transport headers_add/remove lines.
269 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
271 JH/07 Avoid using a waiting database for a single-message-only transport.
272 Performance patch from Paul Fisher. Bugzilla 1262.
274 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
277 JH/09 Add $headers_added variable, with content from use of ACL modifier
278 add_header (but not yet added to the message). Bugzilla 199.
280 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
281 Pulled from Bugzilla 817 by Wolfgang Breyha.
283 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
285 (nb: this is the same fix as in Exim 4.80.1)
287 JH/11 Add A= logging on delivery lines, and a client_set_id option on
290 JH/12 Add optional authenticated_sender logging to A= and a log_selector
293 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
295 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
296 advertise SMTP AUTH mechanism to us, instead of a generic
297 protocol violation error. Also, make Exim more robust to bad
298 data from the Dovecot auth socket.
300 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
302 When a queue runner is handling a message, Exim first routes the
303 recipient addresses, during which it prunes them based on the retry
304 hints database. After that it attempts to deliver the message to
305 any remaining recipients. It then updates the hints database using
308 So if a recipient address works intermittently, it can get repeatedly
309 deferred at routing time. The retry hints record remains fresh so the
310 address never reaches the final cutoff time.
312 This is a fairly common occurrence when a user is bumping up against
313 their storage quota. Exim had some logic in its local delivery code
314 to deal with this. However it did not apply to per-recipient defers
315 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
317 This change adds a proper retry rule check during routing so that the
318 final cutoff time is checked against the message's age. We only do
319 this check if there is an address retry record and there is not a
320 domain retry record; this implies that previous attempts to handle
321 the address had the retry_use_local_parts option turned on. We use
322 this as an approximation for the destination being like a local
323 delivery, as in LMTP.
325 I suspect this new check makes the old local delivery cutoff check
326 redundant, but I have not verified this so I left the code in place.
328 TF/02 Correct gecos expansion when From: is a prefix of the username.
330 Test 0254 submits a message to Exim with the header
334 When I ran the test suite under the user fanf2, Exim expanded
335 the header to contain my full name, whereas it should have added
336 a Resent-Sender: header. It erroneously treats any prefix of the
337 username as equal to the username.
339 This change corrects that bug.
341 GF/01 DCC debug and logging tidyup
342 Error conditions log to paniclog rather than rejectlog.
343 Debug lines prefixed by "DCC: " to remove any ambiguity.
345 TF/03 Avoid unnecessary rebuilds of lookup-related code.
347 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
348 Bug spotted by Jeremy Harris; was flawed since initial commit.
349 Would have resulted in OCSP responses post-SNI triggering an Exim
350 NULL dereference and crash.
352 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
354 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
355 Bug detection, analysis and fix by Samuel Thibault.
356 Bugzilla 1331, Debian bug #698092.
358 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
360 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
361 Server implementation by Todd Lyons, client by JH.
362 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
363 config variable "prdr_enable" controls whether the server
364 advertises the facility. If the client requests PRDR a new
365 acl_data_smtp_prdr ACL is called once for each recipient, after
366 the body content is received and before the acl_smtp_data ACL.
367 The client is controlled by bolth of: a hosts_try_prdr option
368 on the smtp transport, and the server advertisement.
369 Default client logging of deliveries and rejections involving
370 PRDR are flagged with the string "PRDR".
372 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
373 fclose(). Diagnosis by Todd Lyons.
375 PP/17 Update configure.default to handle IPv6 localhost better.
376 Patch by Alain Williams (plus minor tweaks).
379 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
380 This is now consistent with GnuTLS, and is now documented: the
381 previous undocumented portable approach to treating the option as
382 unset was to force an expansion failure. That still works, and
383 an empty string is now equivalent.
385 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
386 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
387 not performing validation itself.
389 PP/20 Added force_command boolean option to pipe transport.
390 Patch from Nick Koston, of cPanel Inc.
392 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
395 TF/04 Added udpsend ACL modifer and hexquote expansion operator
397 PP/21 Fix eximon continuous updating with timestamped log-files.
398 Broken in a format-string cleanup in 4.80, missed when I repaired the
399 other false fix of the same issue.
400 Report and fix from Heiko Schlichting.
403 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
404 Report from Prashanth Katuri.
406 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
407 It's SecureTransport, so affects any MacOS clients which use the
408 system-integrated TLS libraries, including email clients.
410 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
411 using a MIME ACL for non-SMTP local injection.
412 Report and assistance in diagnosis by Warren Baker.
414 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
416 JH/16 Fix comparisons for 64b. Bugzilla 1385.
418 TL/09 Add expansion variable $authenticated_fail_id to keep track of
419 last id that failed so it may be referenced in subsequent ACL's.
421 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
424 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
425 ldap library initialization, allowing self-signed CA's to be
426 used. Also properly sets require_cert option later in code by
427 using NULL (global ldap config) instead of ldap handle (per
428 session). Bug diagnosis and testing by alxgomz.
430 TL/12 Enhanced documentation in the ratelimit.pl script provided in
431 the src/util/ subdirectory.
433 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
434 renamed to Transport Post Delivery Action by Jeremy Harris, as
437 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
438 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
439 redis_servers = needs to be configured which will be used by the redis
440 lookup. Patch from Warren Baker, of The Packet Hub.
442 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
444 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
445 hostname or reverse DNS when processing a host list. Used suggestions
446 from multiple comments on this bug.
448 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
450 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
451 Missed a few lines, added it to make the runtest require no keyboard
454 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
455 contains upper case chars. Make router use caseful_local_part.
457 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
458 support when GnuTLS has been built with p11-kit.
464 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
466 This, or similar/improved, will also be change PP/11 of 4.82.
472 PP/01 Handle short writes when writing local log-files.
473 In practice, only affects FreeBSD (8 onwards).
474 Bugzilla 1053, with thanks to Dmitry Isaikin.
476 NM/01 Bugzilla 949 - Documentation tweak
478 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
481 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
483 PP/02 Implemented gsasl authenticator.
485 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
487 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
488 `pkg-config foo` for cflags/libs.
490 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
491 with rest of GSASL and with heimdal_gssapi.
493 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
494 `pkg-config foo` for cflags/libs for the TLS implementation.
496 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
497 properties get this fed in as external SSF. A number of robustness
498 and debugging improvements to the cyrus_sasl authenticator.
500 PP/08 cyrus_sasl server now expands the server_realm option.
502 PP/09 Bugzilla 1214 - Log authentication information in reject log.
503 Patch by Jeremy Harris.
505 PP/10 Added dbmjz lookup type.
507 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
509 PP/12 MAIL args handles TAB as well as SP, for better interop with
510 non-compliant senders.
511 Analysis and variant patch by Todd Lyons.
513 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
514 Bug report from Lars Müller <lars@samba.org> (via SUSE),
515 Patch from Dirk Mueller <dmueller@suse.com>
517 PP/13 tls_peerdn now print-escaped for spool files.
518 Observed some $tls_peerdn in wild which contained \n, which resulted
519 in spool file corruption.
521 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
522 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
523 or write after TLS renegotiation, which otherwise led to messages
526 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
527 as a tracking header (ie: a signed header comes before the signature).
528 Patch from Wolfgang Breyha.
530 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
531 comma-sep list; embedded commas doubled.
533 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
535 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
537 Report and patch from Dmitry Banschikov.
539 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
540 Removed SSL_clear() after SSL_new() which led to protocol negotiation
541 failures. We appear to now support TLS1.1+ with Exim.
543 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
544 lets Exim select keys and certificates based upon TLS SNI from client.
545 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
546 before an outbound SMTP session. New log_selector, +tls_sni.
548 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
549 NULL dereference. Report and patch from Alun Jones.
551 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
552 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
553 Not seeing resolver debug output on NetBSD, but suspect this is a
554 resolver implementation change.
556 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
557 Left warnings. Added "eximon gdb" invocation mode.
559 PP/21 Defaulting "accept_8bitmime" to true, not false.
561 PP/22 Added -bw for inetd wait mode support.
563 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
564 locate the relevant includes and libraries. Made this the default.
566 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
567 Bugzilla 1246, report and most of solution from Tomasz Kusy.
569 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
570 This may cause build issues on older platforms.
572 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
573 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
574 gnutls_require_mac & gnutls_require_protocols (no longer supported).
575 Added SNI support via GnuTLS too.
576 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
578 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
580 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
581 Applied second patch from Janne, implementing suggestion to default
582 multiple-strings-in-record handling to match SPF spec.
584 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
586 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
587 read-only, out of scope).
588 Patch from Wolfgang Breyha, report from Stuart Northfield.
590 PP/29 Fix three issues highlighted by clang analyser static analysis.
591 Only crash-plausible issue would require the Cambridge-specific
592 iplookup router and a misconfiguration.
593 Report from Marcin Mirosław.
595 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
597 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
598 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
599 As part of this, removing so much warning spew let me fix some minor
600 real issues in debug logging.
602 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
603 assignment on my part. Fixed.
605 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
606 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
607 Janne Snabb (who went above and beyond: thank you).
609 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
610 string otherwise requires a connection and a bunch more work and it's
611 relatively easy to get wrong. Should also expose TLS library linkage
614 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
615 64-bit ${eval} (JH/03).
617 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
618 GNU libc to support some of the 64-bit stuff, should not lead to
619 conflicts. Defined before os.h is pulled in, so if a given platform
620 needs to override this, it can.
622 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
623 protection layer was required, which is not implemented.
624 Bugzilla 1254, patch from Wolfgang Breyha.
626 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
627 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
628 tls_dhparam take prime identifiers. Also unbreak combination of
629 OpenSSL+DH_params+TLSSNI.
631 PP/39 Disable SSLv2 by default in OpenSSL support.
637 PP/01 Solaris build fix for Oracle's LDAP libraries.
638 Bugzilla 1109, patch from Stephen Usher.
640 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
642 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
645 TF/02 Fix a couple more cases where we did not log the error message
646 when unlink() failed. See also change 4.74-TF/03.
648 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
649 lock up or crash if it happened to be inside a call to libc when it
650 got a SIGUSR1 from exiwhat.
652 The SIGUSR1 handler appends the current process status to the process
653 log which is later printed by exiwhat. It used to use the general
654 purpose logging code to do this, but several functions it calls are
655 not safe for signals.
657 The new output code in the SIGUSR1 handler is specific to the process
658 log, and simple enough that it's easy to inspect for signal safety.
659 Removing some special cases also simplifies the general logging code.
660 Removing the spurious timestamps from the process log simplifies
663 TF/04 Improved ratelimit ACL condition.
665 The /noupdate option has been deprecated in favour of /readonly which
666 has clearer semantics. The /leaky, /strict, and /readonly update modes
667 are mutually exclusive. The update mode is no longer included in the
668 database key; it just determines when the database is updated. (This
669 means that when you upgrde Exim will forget old rate measurements.)
671 Exim now checks that the per_* options are used with an update mode that
672 makes sense for the current ACL. For example, when Exim is processing a
673 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
674 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
675 must specify per_mail/readonly. If you omit the update mode it defaults to
676 /leaky where that makes sense (as before) or /readonly where required.
678 The /noupdate option is now undocumented but still supported for
679 backwards compatibility. It is equivalent to /readonly except that in
680 ACLs where /readonly is required you may specify /leaky/noupdate or
681 /strict/noupdate which are treated the same as /readonly.
683 A useful new feature is the /count= option. This is a generalization
684 of the per_byte option, so that you can measure the throughput of other
685 aggregate values. For example, the per_byte option is now equivalent
686 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
688 The per_rcpt option has been generalized using the /count= mechanism
689 (though it's more complicated than the per_byte equivalence). When it is
690 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
691 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
692 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
693 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
694 Note that using per_rcpt with a non-readonly update mode in more than
695 one ACL will cause the recipients to be double-counted. (The per_mail
696 and per_byte options don't have this problem.)
698 The handling of very low rates has changed slightly. If the computed rate
699 is less than the event's count (usually one) then this event is the first
700 after a long gap. In this case the rate is set to the same as this event's
701 count, so that the first message of a spam run is counted properly.
703 The major new feature is a mechanism for counting the rate of unique
704 events. The new per_addr option counts the number of different
705 recipients that someone has sent messages to in the last time period. It
706 behaves like per_rcpt if all the recipient addresses are different, but
707 duplicate recipient addresses do not increase the measured rate. Like
708 the /count= option this is a general mechanism, so the per_addr option
709 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
710 example, measure the rate that a client uses different sender addresses
711 with the options per_mail/unique=$sender_address. There are further
712 details in the main documentation.
714 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
716 TF/06 Removed a few PCRE remnants.
718 TF/07 Automatically extract Exim's version number from tags in the git
719 repository when doing development or release builds.
721 PP/02 Raise smtp_cmd_buffer_size to 16kB.
722 Bugzilla 879. Patch from Paul Fisher.
724 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
725 Heavily based on revision 40f9a89a from Simon Arlott's tree.
728 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
730 PP/05 Variable $av_failed, true if the AV scanner deferred.
731 Bugzilla 1078. Patch from John Horne.
733 PP/06 Stop make process more reliably on build failure.
734 Bugzilla 1087. Patch from Heiko Schlittermann.
736 PP/07 Make maildir_use_size_file an _expandable_ boolean.
737 Bugzilla 1089. Patch from Heiko Schlittermann.
739 PP/08 Handle ${run} returning more data than OS pipe buffer size.
740 Bugzilla 1131. Patch from Holger Weiß.
742 PP/09 Handle IPv6 addresses with SPF.
743 Bugzilla 860. Patch from Wolfgang Breyha.
745 PP/10 GnuTLS: support TLS 1.2 & 1.1.
747 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
750 PP/11 match_* no longer expand right-hand-side by default.
751 New compile-time build option, EXPAND_LISTMATCH_RHS.
752 New expansion conditions, "inlist", "inlisti".
754 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
756 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
758 PP/14 fix log_write() format string regression from TF/03.
759 Bugzilla 1152. Patch from Dmitry Isaikin.
765 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
767 PP/02 Harmonised TLS library version reporting; only show if debugging.
768 Layout now matches that introduced for other libraries in 4.74 PP/03.
770 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
772 PP/04 New "dns_use_edns0" global option.
774 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
777 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
778 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
780 TK/01 Updated PolarSSL code to 0.14.2.
781 Bugzilla 1097. Patch from Andreas Metzler.
783 PP/07 Catch divide-by-zero in ${eval:...}.
786 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
789 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
790 format-string attack -- SECURITY: remote arbitrary code execution.
792 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
793 time unintentionally subject to list matching rules, letting the header
794 cause arbitrary Exim lookups (of items which can occur in lists, *not*
795 arbitrary string expansion). This allowed for information disclosure.
797 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
798 INT_MIN/-1 -- value coerced to INT_MAX.
804 NM/01 Workround for PCRE version dependency in version reporting
807 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
808 This fixes portability to compilers other than gcc, notably
809 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
811 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
812 makefiles for portability to HP-UX and POSIX correctness.
814 PP/01 Permit LOOKUP_foo enabling on the make command-line.
815 Also via indented variable definition in the Makefile.
816 (Debugging by Oliver Heesakkers).
818 PP/02 Restore caching of spamd results with expanded spamd_address.
819 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
821 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
822 Improves build reliability. Fix from: Frank Elsner
824 NM/02 Fix wide character breakage in the rfc2047 coding
825 Fixes bug 1064. Patch from Andrey N. Oktyabrski
827 NM/03 Allow underscore in dnslist lookups
828 Fixes bug 1026. Patch from Graeme Fowler
830 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
831 Code patches from Adam Ciarcinski of NetBSD.
833 NM/04 Fixed exiqgrep to cope with mailq missing size issue
836 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
837 is logged, to avoid truncation. Patch from John Horne.
839 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
840 Patch from Jakob Hirsch.
842 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
843 SQL string expansion failure details.
844 Patch from Andrey Oktyabrski.
846 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
847 Patch from Simon Arlott.
849 PP/09 New lookups functionality failed to compile on old gcc which rejects
850 extern declarations in function scope.
851 Patch from Oliver Fleischmann
853 PP/10 Use sig_atomic_t for flags set from signal handlers.
854 Check getgroups() return and improve debugging.
855 Fixed developed for diagnosis in bug 927 (which turned out to be
858 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
859 Patch from Mark Zealey.
861 PP/12 Bugzilla 1056: Improved spamd server selection.
862 Patch from Mark Zealey.
864 PP/13 Bugzilla 1086: Deal with maildir quota file races.
865 Based on patch from Heiko Schlittermann.
867 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
868 Patch from Uwe Doering, sign-off by Michael Haardt.
870 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
871 variable declaration deep within a block. Bug and patch from
874 PP/15 lookups-Makefile IRIX compatibilty coercion.
876 PP/16 Make DISABLE_DKIM build knob functional.
878 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
879 Patch by Simon Arlott
881 TF/03 Fix valgrind.h portability to C89 compilers that do not support
882 variable argument macros. Our copy now differs from upstream.
888 TF/01 Failure to get a lock on a hints database can have serious
889 consequences so log it to the panic log.
891 TF/02 Log LMTP confirmation messages in the same way as SMTP,
892 controlled using the smtp_confirmation log selector.
894 TF/03 Include the error message when we fail to unlink a spool file.
896 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
897 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
898 for maintaining out-of-tree patches for some time.
900 PP/01 Bugzilla 139: Documentation and portability issues.
901 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
902 Handle per-OS dynamic-module compilation flags.
904 PP/02 Let /dev/null have normal permissions.
905 The 4.73 fixes were a little too stringent and complained about the
906 permissions on /dev/null. Exempt it from some checks.
907 Reported by Andreas M. Kirchwitz.
909 PP/03 Report version information for many libraries, including
910 Exim version information for dynamically loaded libraries. Created
911 version.h, now support a version extension string for distributors
912 who patch heavily. Dynamic module ABI change.
914 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
915 privilege escalation vulnerability whereby the Exim run-time user
916 can cause root to append content of the attacker's choosing to
919 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
922 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
923 If dropping privileges for untrusted macros, we disabled normal logging
924 on the basis that it would fail; for the Exim run-time user, this is not
925 the case, and it resulted in successful deliveries going unlogged.
926 Fixed. Reported by Andreas Metzler.
932 PP/01 Date: & Message-Id: revert to normally being appended to a message,
933 only prepend for the Resent-* case. Fixes regression introduced in
934 Exim 4.70 by NM/22 for Bugzilla 607.
936 PP/02 Include check_rfc2047_length in configure.default because we're seeing
937 increasing numbers of administrators be bitten by this.
939 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
941 PP/03 Bugzilla 994: added openssl_options main configuration option.
943 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
945 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
947 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
949 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
950 without a peer certificate, leading to a segfault because of an
951 assumption that peers always have certificates. Be a little more
952 paranoid. Problem reported by Martin Tscholak.
954 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
955 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
956 NB: ClamAV planning to remove STREAM in "middle of 2010".
957 CL also introduces -bmalware, various -d+acl logging additions and
958 more caution in buffer sizes.
960 PP/09 Implemented reverse_ip expansion operator.
962 PP/10 Bugzilla 937: provide a "debug" ACL control.
964 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
966 PP/12 Bugzilla 973: Implement --version.
968 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
970 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
972 PP/15 Bugzilla 816: support multiple condition rules on Routers.
974 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
975 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
976 ignore trailing whitespace.
978 JJ/02 prevent non-panic DKIM error from being sent to paniclog
980 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
983 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
984 Notification from Dr Andrew Aitchison.
986 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
987 ExtendedDetectionInfo response format.
988 Notification from John Horne.
990 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
993 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
994 XSL and documented dependency on system catalogs, with examples of how
997 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1000 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1001 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1002 configuration file which is writeable by the Exim user or group.
1004 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1005 of configuration files to cover files specified with the -C option if
1006 they are going to be used with root privileges, not just the default
1009 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1010 option (effectively making it always true).
1012 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1013 files to be used while preserving root privileges.
1015 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1016 that rogue child processes cannot use them.
1018 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1019 run-time user, instead of root.
1021 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1022 Exim run-time user without dropping privileges.
1024 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1025 result string, instead of calling string_vformat() twice with the same
1028 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1029 for other users. Others should always drop root privileges if they use
1030 -C on the command line, even for a whitelisted configure file.
1032 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1034 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1040 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1041 $data_path, and $header_path variables; fixed documentation bugs and
1044 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1045 exipick to access non-standard spools, including the "frozen" queue
1048 NM/01 Bugzilla 965: Support mysql stored procedures.
1049 Patch from Alain Williams
1051 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1053 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1054 Patch from Andreas Metzler
1056 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1057 Patch from Kirill Miazine
1059 NM/05 Bugzilla 671: Added umask to procmail example.
1061 JJ/03 installed exipick 20100323.0, fixing doc bug
1063 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1064 directory. Notification and patch from Dan Rosenberg.
1066 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1068 TK/02 Improve log output when DKIM signing operation fails.
1070 MH/01 Treat the transport option dkim_domain as a colon separated
1071 list, not as a single string, and sign the message with each element,
1072 omitting multiple occurences of the same signer.
1074 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1075 Bugzilla 985, 986. Patch by Simon Arlott
1077 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1078 Patch by Simon Arlott
1080 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1081 MBX locking. Notification from Dan Rosenberg.
1087 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1089 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1091 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1093 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1095 NM/04 Bugzilla 915: Flag broken perl installation during build.
1101 TK/01 Added patch by Johannes Berg that expands the main option
1102 "spamd_address" if it starts with a dollar sign.
1104 TK/02 Write list of recipients to X-Envelope-Sender header when building
1105 the mbox-format spool file for content scanning (suggested by Jakob
1108 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1109 (http://www.dcc-servers.net/) support via dccifd. Activated by
1110 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1112 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1113 by Mark Daniel Reidel <mr@df.eu>.
1115 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1116 When building exim an external PCRE library is now needed -
1117 PCRE is a system library on the majority of modern systems.
1118 See entry on PCRE_LIBS in EDITME file.
1120 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1121 conversation. Added nologin parameter to request.
1122 Patch contributed by Kirill Miazine.
1124 TF/01 Do not log submission mode rewrites if they do not change the address.
1126 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1128 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1129 log files in place. Contributed by Roberto Lima.
1131 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1133 TF/03 Bugzilla 615: When checking the local_parts router precondition
1134 after a local_part_suffix or local_part_prefix option, Exim now
1135 does not use the address's named list lookup cache, since this
1136 contains cached lookups for the whole local part.
1138 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1139 Robert Millan. Documentation is in experimental-spec.txt.
1141 TF/04 Bugzilla 668: Fix parallel build (make -j).
1143 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1145 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1146 Patch provided by Jan Srzednicki.
1148 TF/05 Leading white space used to be stripped from $spam_report which
1149 wrecked the formatting. Now it is preserved.
1151 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1152 that they are available at delivery time.
1154 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1156 TF/08 TLS error reporting now respects the incoming_interface and
1157 incoming_port log selectors.
1159 TF/09 Produce a more useful error message if an SMTP transport's hosts
1160 setting expands to an empty string.
1162 NM/06 Bugzilla 744: EXPN did not work under TLS.
1163 Patch provided by Phil Pennock.
1165 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1166 Patch provided by Richard Godbee.
1168 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1169 acl_smtp_notquit, added index entry.
1171 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1172 Patch provided by Eugene Bujak.
1174 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1175 accept(). Patch provided by Maxim Dounin.
1177 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1178 Patch provided by Phil Pennock.
1180 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1182 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1183 Patch provided by Brad "anomie" Jorsch.
1185 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1186 Patch provided by Dean Brooks.
1188 TK/05 Add native DKIM support (does not depend on external libraries).
1190 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1191 Patch provided by Graeme Fowler.
1193 NM/16 Bugzilla 851: Documentation example syntax fix.
1195 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1197 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1200 NM/19 Bugzilla 745: TLS version reporting.
1201 Patch provided by Phil Pennock.
1203 NM/20 Bugzilla 167: bool: condition support.
1204 Patch provided by Phil Pennock.
1206 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1207 clients. Patch provided by Phil Pennock.
1209 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1210 Patch provided by Brad "anomie" Jorsch.
1212 NM/23 Bugzilla 687: Fix misparses in eximstats.
1213 Patch provided by Heiko Schlittermann.
1215 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1216 Patch provided by Heiko Schlittermann.
1218 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1219 plus update to original patch.
1221 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1223 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1224 Patch provided by David Brownlee.
1226 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1228 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1230 NM/30 Bugzilla 888: TLS documentation bugfixes.
1232 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1234 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1235 Unlike the original bugzilla I have changed all shell scripts in src tree.
1237 NM/33 Bugzilla 898: Transport filter timeout fix.
1238 Patch by Todd Rinaldo.
1240 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1241 Patch by Serge Demonchaux.
1243 NM/35 Bugzilla 39: Base64 decode bug fixes.
1244 Patch by Jakob Hirsch.
1246 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1248 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1250 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1252 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1258 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1259 ALT-N's libdkim that I have put here:
1260 http://duncanthrax.net/exim-experimental/
1262 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1263 were called 'true' and it seems that C99 defines that as a reserved
1264 keyword to be used with 'bool' variable types. That means you could
1265 not include C99-style headers which use bools without triggering
1266 build errors in sieve.c.
1268 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1269 as mailq or other aliases. Changed the --help handling significantly
1270 to do whats expected. exim_usage() emits usage/help information.
1272 SC/01 Added the -bylocaldomain option to eximstats.
1274 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1276 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1278 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1284 PH/01 Another patch from the Sieve maintainer.
1286 PH/02 When an IPv6 address is converted to a string for single-key lookup
1287 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1288 dots are used instead of colons so that keys in lsearch files need not
1289 contain colons. This was done some time before quoting was made available
1290 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1291 (notated using the quote facility) so as to distinguish them from IPv4
1292 keys. This meant that lookups for IP addresses in host lists did not work
1293 for iplsearch lookups.
1295 This has been fixed by arranging for IPv6 addresses to be expressed with
1296 colons if the lookup type is iplsearch. This is not incompatible, because
1297 previously such lookups could never work.
1299 The situation is now rather anomolous, since one *can* have colons in
1300 ordinary lsearch keys. However, making the change in all cases is
1301 incompatible and would probably break a number of configurations.
1303 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1306 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1307 conversion specification without a maximum field width, thereby enabling
1308 a rogue spamd server to cause a buffer overflow. While nobody in their
1309 right mind would setup Exim to query an untrusted spamd server, an
1310 attacker that gains access to a server running spamd could potentially
1311 exploit this vulnerability to run arbitrary code as the Exim user.
1313 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1314 $primary_hostname instead of what libspf2 thinks the hosts name is.
1316 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1317 a directory entry by the name of the lookup key. Previously, if a
1318 symlink pointed to a non-existing file or a file in a directory that
1319 Exim lacked permissions to read, a lookup for a key matching that
1320 symlink would fail. Now it is enough that a matching directory entry
1321 exists, symlink or not. (Bugzilla 503.)
1323 PH/03 The body_linecount and body_zerocount variables are now exported in the
1326 PH/04 Added the $dnslist_matched variable.
1328 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1329 This means they are set thereafter only if the connection becomes
1332 PH/06 Added the client_condition to authenticators so that some can be skipped
1333 by clients under certain conditions.
1335 PH/07 The error message for a badly-placed control=no_multiline_responses left
1336 "_responses" off the end of the name.
1338 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1340 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1341 (without spaces) instead of just copying the configuration text.
1343 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1345 PH/11 Added $max_received_linelength.
1347 PH/12 Added +ignore_defer and +include_defer to host lists.
1349 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1350 way in which PCRE > 7.0 is built.
1352 PH/14 Implemented queue_only_load_latch.
1354 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1355 MAIL command. The effect was to mangle the value on 64-bit systems.
1357 PH/16 Another patch from the Sieve maintainer.
1359 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1361 PH/18 If a system quota error occurred while trying to create the file for
1362 a maildir delivery, the message "Mailbox is full" was not appended to the
1363 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1364 only to a quota excession during the actual writing of the file.
1366 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1367 characters?) which causes problems in log lines. The DN values are now
1368 passed through string_printing() before being added to log lines.
1370 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1371 and InterBase are left for another time.)
1373 PH/21 Added message_body_newlines option.
1375 PH/22 Guard against possible overflow in moan_check_errorcopy().
1377 PH/23 POSIX allows open() to be a macro; guard against that.
1379 PH/24 If the recipient of an error message contained an @ in the local part
1380 (suitably quoted, of course), incorrect values were put in $domain and
1381 $local_part during the evaluation of errors_copy.
1387 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1388 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1391 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1392 issue a MAIL command.
1394 PH/02 In an ACL statement such as
1396 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1398 if a client was not listed at all, or was listed with a value other than
1399 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1400 the condition was not true (as it should be), so access was not denied.
1401 The bug was that the ! inversion was incorrectly passed on to the second
1402 item. This has been fixed.
1404 PH/03 Added additional dnslists conditions == and =& which are different from
1405 = and & when the dns lookup returns more than one IP address.
1407 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1408 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1410 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1411 FSYNC, which compiles an option called disable_fsync that allows for
1412 bypassing fsync(). The documentation is heavily laced with warnings.
1414 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1416 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1417 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1418 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1419 including adding "make clean"; (3) Added -fPIC when compiling the test
1420 dynamically loaded module, to get rid of a warning.
1422 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1423 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1424 = 0s. The bug is otherwise harmless.
1426 PH/07 There was a bug in the dovecot authenticator such that the value of
1427 $auth1 could be overwritten, and so not correctly preserved, after a
1428 successful authentication. This usually meant that the value preserved by
1429 the server_setid option was incorrect.
1431 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1433 PH/09 Installed PCRE release 7.0.
1435 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1436 run for batched SMTP input. It is now run at the start of every message
1437 in the batch. While fixing this I discovered that the process information
1438 (output by running exiwhat) was not always getting set for -bs and -bS
1439 input. This is fixed, and it now also says "batched" for BSMTP.
1441 PH/11 Added control=no_pipelining.
1443 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1444 patch, slightly modified), and move the expansion of helo_data till after
1445 the connection is made in the smtp transport (so it can use these
1448 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1450 PH/14 Added log_selector = +pid.
1452 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1454 PH/16 Add ${if forany and ${if forall.
1456 PH/17 Added dsn_from option to vary the From: line in DSNs.
1458 PH/18 Flush SMTP output before performing a callout, unless control =
1459 no_callout_flush is set.
1461 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1462 was true (the default) a successful delivery failed to delete the retry
1463 item, thus causing premature timeout of the address. The bug is now
1466 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1468 PH/21 Long custom messages for fakedefer and fakereject are now split up
1469 into multiline reponses in the same way that messages for "deny" and
1470 other ACL rejections are.
1472 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1473 with slight modification.
1475 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1476 draft, changing the syntax and factoring some duplicate code".
1478 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1479 for deliveries of the second and subsequent messages over the same SMTP
1482 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1483 ${reduce, with only minor "tidies".
1485 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1487 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1488 expansion side effects.
1490 PH/27 When a message times out after an over-quota error from an Exim-imposed
1491 quota, the bounce message says "mailbox is full". This message was not
1492 being given when it was a system quota that was exceeded. It now should
1495 MH/03 Made $recipients available in local_scan(). local_scan() already has
1496 better access to the recipient list through recipients_list[], but
1497 $recipients can be useful in postmaster-provided expansion strings.
1499 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1500 in the case of a MAIL command with additional options following the
1501 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1502 were accidentally chopped off.
1504 PH/29 SMTP synchronization checks are implemented when a command is read -
1505 there is a check that no more input is waiting when there shouldn't be
1506 any. However, for some commands, a delay in an ACL can mean that it is
1507 some time before the response is written. In this time, more input might
1508 arrive, invalidly. So now there are extra checks after an ACL has run for
1509 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1510 pipelining has not been advertised.
1512 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1514 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1515 returned in the SMTP response when smtp_return_error_details was false.
1516 This has been fixed.
1518 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1519 instead of the C I/O that was originally supplied, because problems were
1520 reported on Solaris.
1522 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1523 Exim which did not show up earlier: it was assuming that a call to
1524 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1525 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1526 was a macro that became an assignment, so it seemed to work. This has
1527 changed to a proper function call with a void return, hence the compile
1528 error. Exim's code has been fixed.
1530 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1533 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1535 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1537 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1538 because it turns out to be too big or there is a timeout) there is no
1539 "Completed" line in the log. When some messages of this type were
1540 selected by exigrep, they were listed as "not completed". Others were
1541 picked up by some special patterns. I have improved the selection
1542 criteria to be more general.
1544 PH/38 The host_find_failed option in the manualroute router can now be set
1545 to "ignore", to completely ignore a host whose IP address cannot be
1546 found. If all hosts are ignored, the behaviour is controlled by the new
1547 host_all_ignored option.
1549 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1550 homing or because of multiple MX records with /mx) generated more than
1551 one IP address, and the following item turned out to be the local host,
1552 all the secondary addresses of the first item were incorrectly removed
1553 from the list, along with the local host and any following hosts (which
1554 is what is supposed to happen).
1556 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1557 whoever called Exim into the -H file. In the case of the daemon it was
1558 behaving confusingly. When first started, it used values for whoever
1559 started the daemon, but after a SIGHUP it used the Exim user (because it
1560 calls itself on a restart). I have changed the code so that it now always
1563 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1564 message are rejected with the same error (e.g. no authentication or bad
1565 sender address), and a DATA command is nevertheless sent (as can happen
1566 with PIPELINING or a stupid MUA), the error message that was given to the
1567 RCPT commands is included in the rejection of the DATA command. This is
1568 intended to be helpful for MUAs that show only the final error to their
1571 PH/42 Another patch from the Sieve maintainer.
1573 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1574 Eximstats - Fixed some broken HTML links and added missing column headers
1576 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1577 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1579 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1581 PH/43 Yet another patch from the Sieve maintainer.
1583 PH/44 I found a way to check for a TCP/IP connection going away before sending
1584 the response to the final '.' that terminates a message, but only in the
1585 case where the client has not sent further data following the '.'
1586 (unfortunately, this is allowed). However, in many cases there won't be
1587 any further data because there won't be any more messages to send. A call
1588 to select() can be used: if it shows that the input is "ready", there is
1589 either input waiting, or the socket has been closed. An attempt to read
1590 the next input character can distinguish the two cases. Previously, Exim
1591 would have sent an OK response which the client would never have see.
1592 This could lead to message repetition. This fix should cure that, at
1593 least in a lot of common cases.
1595 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1596 advertised in response to EHLO.
1602 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1603 fixed by 4.65/MH/01 (is this a record?) are fixed:
1605 (i) An empty string was always treated as zero by the numeric comparison
1606 operators. This behaviour has been restored.
1608 (ii) It is documented that the numeric comparison operators always treat
1609 their arguments as decimal numbers. This was broken in that numbers
1610 starting with 0 were being interpreted as octal.
1612 While fixing these problems I realized that there was another issue that
1613 hadn't been noticed. Values of message_size_limit (both the global option
1614 and the transport option) were treated as octal if they started with 0.
1615 The documentation was vague. These values are now always treated as
1616 decimal, and I will make that clear in the documentation.
1622 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1623 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1626 MH/01 Don't check that the operands of numeric comparison operators are
1627 integers when their expansion is in "skipping" mode (fixes bug
1628 introduced by 4.64-PH/07).
1630 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1631 child addresses, Exim now panics and dies. Previously, because the count
1632 is held in a short int, deliveries were likely to be lost. As such a
1633 large number of recipients for a single message is ridiculous
1634 (performance will be very, very poor), I have chosen to impose a limit
1635 rather than extend the field.
1641 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1642 leftover -K file (the existence of which was triggered by #402).
1643 While we were at it, introduced process PID as part of the -K
1644 filename. This should rule out race conditions when creating
1647 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1648 processing considerably. Previous code took too long for large mails,
1649 triggering a timeout which in turn triggers #401.
1651 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1652 in the DK code in transports.c. sendfile() is not really portable,
1653 hence the _LINUX specificness.
1655 TF/01 In the add_headers option to the mail command in an Exim filter,
1656 there was a bug that Exim would claim a syntax error in any
1657 header after the first one which had an odd number of characters
1660 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1661 callout verification, Exim cached a "reject" for the entire domain. This
1662 is correct for most verifications, but it is not correct for a recipient
1663 verification with use_sender or use_postmaster set, because in that case
1664 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1665 case of MAIL FROM:<> rejection from other early rejections (e.g.
1666 rejection of HELO). When verifying a recipient using a non-null MAIL
1667 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1668 Whatever the result of the callout, the value of the domain cache is
1669 left unchanged (for any other kind of callout, getting as far as trying
1670 RCPT means that the domain itself is ok).
1672 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1675 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1676 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1677 session, the daemon ignores ECONNECT errors and logs others; it now
1678 ignores EPIPE as well.
1680 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1681 (quoted-printable decoding).
1683 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1684 later the small subsequent patch to fix an introduced bug.
1686 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1688 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1690 PH/08 An error is now given if message_size_limit is specified negative.
1692 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1693 to be given (somewhat) arbitrary names.
1695 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1698 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1699 miscellaneous code fixes
1701 PH/10 Added the log_reject_target ACL modifier to specify where to log
1704 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1705 hostname. This is wrong, because it relates to the incoming message (and
1706 probably the interface on which it is arriving) and not to the outgoing
1707 callout (which could be using a different interface). This has been
1708 changed to use the value of the helo_data option from the smtp transport
1709 instead - this is what is used when a message is actually being sent. If
1710 there is no remote transport (possible with a router that sets up host
1711 addresses), $smtp_active_hostname is used.
1713 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1714 tweaks were necessary in order to get it to work (see also 21 below):
1715 (a) The code assumed that strncpy() returns a negative number on buffer
1716 overflow, which isn't the case. Replaced with Exim's string_format()
1718 (b) There were several signed/unsigned issues. I just did the minimum
1719 hacking in of casts. There is scope for a larger refactoring.
1720 (c) The code used strcasecmp() which is not a standard C function.
1721 Replaced with Exim's strcmpic() function.
1722 (d) The code set only $1; it now sets $auth1 as well.
1723 (e) A simple test gave the error "authentication client didn't specify
1724 service in request". It would seem that Dovecot has changed its
1725 interface. Fortunately there's a specification; I followed it and
1726 changed what the client sends and it appears to be working now.
1728 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1731 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1732 address A is aliased to B and C, where B exists and C does not. Without
1733 -v the output is "A verified" because verification stops after a
1734 successful redirection if more than one address is generated. However,
1735 with -v the child addresses are also verified. Exim was outputting "A
1736 failed to verify" and then showing the successful verification for C,
1737 with its parentage. It now outputs "B failed to verify", showing B's
1738 parentage before showing the successful verification of C.
1740 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1741 look up a TXT record in a specific list after matching in a combined
1744 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1745 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1746 they consult the DNS. I had assumed they would set it the way they
1747 wanted; and indeed my experiments on Linux seem to show that in some
1748 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1749 To be on the safe side, however, I have now made the interface to
1750 host_find_byname() similar to host_find_bydns(), with an argument
1751 containing the DNS resolver options. The host_find_byname() function now
1752 sets these options at its start, just as host_find_bydns() does. The smtp
1753 transport options dns_qualify_single and dns_search_parents are passed to
1754 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1755 of host_find_byname() use the default settings of RES_DEFNAMES
1756 (qualify_single) but not RES_DNSRCH (search_parents).
1758 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1759 spool_read_header() do less string testing, by means of a preliminary
1760 switch on the second character of optional "-foo" lines. (This is
1761 overdue, caused by the large number of possibilities that now exist.
1762 Originally there were few.) While I was there, I also converted the
1763 str(n)cmp tests so they don't re-test the leading "-" and the first
1764 character, in the hope this might squeeze out yet more improvement.
1766 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1767 flag allowing group syntax was set by the header_syntax check but not
1768 turned off, possible causing trouble later; (2) The flag was not being
1769 set at all for the header_verify test, causing "group"-style headers to
1770 be rejected. I have now set it in this case, and also caused header_
1771 verify to ignore an empty address taken from a group. While doing this, I
1772 came across some other cases where the code for allowing group syntax
1773 while scanning a header line wasn't quite right (mostly, not resetting
1774 the flag correctly in the right place). These bugs could have caused
1775 trouble for malformed header lines. I hope it is now all correct.
1777 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1778 with the "reply" argument non-NULL. The code, however (which originally
1779 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1780 but it didn't always do it. This confused somebody who was copying the
1781 code for some other use. I have removed all the tests.
1783 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1784 feature that was used to support insecure browsers during the U.S. crypto
1785 embargo. It requires special client support, and Exim is probably the
1786 only MTA that supported it -- and would never use it because real RSA is
1787 always available. This code has been removed, because it had the bad
1788 effect of slowing Exim down by computing (never used) parameters for the
1789 RSA_EXPORT functionality.
1791 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1792 authenticator to fail if there's a tab character in the incoming data
1793 (there should never be unless someone is messing about, as it's supposed
1794 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1795 if the connection is using TLS or if the remote IP is the same as the
1796 local IP, and the "valid-client-cert option" if a client certificate has
1799 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1800 authenticators. This can be used for authorization after authentication
1801 succeeds. (In the case of plaintext, it servers for both authentication
1804 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1805 if any retry times were supplied.
1807 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1808 connection, where there is no client IP address to check. In this
1809 situation, the verify now always succeeds.
1811 PH/25 Applied John Jetmore's -Mset patch.
1813 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1815 PH/27 In a string expansion for a processed (not raw) header when multiple
1816 headers of the same name were present, leading whitespace was being
1817 removed from all of them, but trailing whitespace was being removed only
1818 from the last one. Now trailing whitespace is removed from each header
1819 before concatenation. Completely empty headers in a concatenation (as
1820 before) are ignored.
1822 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1823 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1825 PH/29 [Removed. This was a change that I later backed out, and forgot to
1826 correct the ChangeLog entry (that I had efficiently created) before
1827 committing the later change.]
1829 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1830 address errors (4xx response to RCPT) over the same connection as other
1831 messages routed to the same hosts. Such deliveries are always "forced",
1832 so retry times are not inspected. This resulted in far too many retries
1833 for the affected addresses. The effect occurred only when there were more
1834 hosts than the hosts_max_try setting in the smtp transport when it had
1835 the 4xx errors. Those hosts that it had tried were not added to the list
1836 of hosts for which the message was waiting, so if all were tried, there
1837 was no problem. Two fixes have been applied:
1839 (i) If there are any address or message errors in an SMTP delivery, none
1840 of the hosts (tried or untried) are now added to the list of hosts
1841 for which the message is waiting, so the message should not be a
1842 candidate for sending over the same connection that was used for a
1843 successful delivery of some other message. This seems entirely
1844 reasonable: after all the message is NOT "waiting for some host".
1845 This is so "obvious" that I'm not sure why it wasn't done
1846 previously. Hope I haven't missed anything, but it can't do any
1847 harm, as the worst effect is to miss an optimization.
1849 (ii) If, despite (i), such a delivery is accidentally attempted, the
1850 routing retry time is respected, so at least it doesn't keep
1851 hammering the server.
1853 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1854 in ${readsocket because some servers need this prod.
1856 PH/32 Added some extra debug output when updating a wait-xxx database.
1858 PH/33 The hint "could be header name not terminated by colon", which has been
1859 given for certain expansion errors for a long time, was not being given
1860 for the ${if def:h_colon_omitted{... case.
1862 PH/34 The spec says: "With one important exception, whenever a domain list is
1863 being scanned, $domain contains the subject domain." There was at least
1864 one case where this was not true.
1866 PH/35 The error "getsockname() failed: connection reset by peer" was being
1867 written to the panic log as well as the main log, but it isn't really
1868 panic-worthy as it just means the connection died rather early on. I have
1869 removed the panic log writing for the ECONNRESET error when getsockname()
1872 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1873 runs only) independently of the message's sender address. This meant
1874 that, if the 4xx error was in fact related to the sender, a different
1875 message to the same recipient with a different sender could confuse
1876 things. In particualar, this can happen when sending to a greylisting
1877 server, but other circumstances could also provoke similar problems.
1878 I have changed the default so that the retry time for these errors is now
1879 based a combination of the sender and recipient addresses. This change
1880 can be overridden by setting address_retry_include_sender=false in the
1883 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1884 remote server are returned as part of bounce messages. This was not
1885 happening for LMTP over a pipe (the lmtp transport), but now it is the
1886 same for both kinds of LMTP.
1888 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1889 in header lines that were in fact CNAMEs. This is no longer the case.
1891 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1892 and queue runs started by the daemon processed all messages. This has
1893 been fixed so that -R and -S can now usefully be given with -q<time>.
1895 PH/40 Import PCRE release 6.7 (fixes some bugs).
1897 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1899 PH/42 Give an error if -q is specified more than once.
1901 PH/43 Renamed the variables $interface_address and $interface_port as
1902 $received_ip_address and $received_port, to make it clear that these
1903 values apply to message reception, and not to the outgoing interface when
1904 a message is delivered. (The old names remain recognized, of course.)
1906 PH/44 There was no timeout on the connect() call when using a Unix domain
1907 socket in the ${readsocket expansion. There now is.
1909 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1910 be meaningful with "accept".
1912 SC/01 Eximstats V1.43
1913 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1915 SC/02 Eximstats V1.44
1916 Use a glob alias rather than an array ref in the generated
1917 parser. This improves both readability and performance.
1919 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1920 Collect SpamAssassin and rejection statistics.
1921 Don't display local sender or destination tables unless
1922 there is data to show.
1923 Added average volumes into the top table text output.
1925 SC/04 Eximstats V1.46
1926 Collect data on the number of addresses (recipients)
1927 as well as the number of messages.
1929 SC/05 Eximstats V1.47
1930 Added 'Message too big' to the list of mail rejection
1931 reasons (thanks to Marco Gaiarin).
1933 SC/06 Eximstats V1.48
1934 Mainlog lines which have GMT offsets and are too short to
1935 have a flag are now skipped.
1937 SC/07 Eximstats V1.49 (Alain Williams)
1938 Added the -emptyok flag.
1940 SC/08 Eximstats V1.50
1941 Fixes for obtaining the IP address from reject messages.
1943 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1944 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1945 whitesspace changes from 4.64-PH/27
1947 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1950 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1951 are found, allow negative numbers in numeric criteria)
1953 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1955 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1956 to match changes made in 4.64-PH/43
1958 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1960 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1961 "domain" is actually an IP address, and force a failure. This locks out
1962 those revolvers/nameservers that support "A-for-A" lookups, in
1963 contravention of the specifications.
1965 PH/48 When a host name was looked up from an IP address, and the subsequent
1966 forward lookup of the name timed out, the host name was left in
1967 $sender_host_name, contrary to the specification.
1969 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1970 restricted to single-key lookups, Exim was not diagnosing an error if
1971 * or *@ was used with a query-style lookup.
1973 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1975 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1976 long ago, but noone interested enough thought of it. Let's just say that
1977 the "1.1" means that there are some new functions that weren't there at
1978 some point in the past.
1980 PH/51 Error processing for expansion failure of helo_data from an smtp
1981 transport during callout processing was broken.
1983 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1984 tested/used via the -bh/-bhc/-bs options.
1986 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1987 bug, fixed in subsequent PCRE releases).
1989 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1990 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1992 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1998 SC/01 Use a glob alias rather than an array ref in eximstats generated
1999 parser. This improves both readability and performance.
2001 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2002 Don't display local sender or destination tables in eximstats unless
2003 there is data to show.
2004 Added average volumes into the eximstats top table text output.
2006 SC/03 Collect data on the number of addresses (recipients) as well
2007 as the number of messages in eximstats.
2009 TF/01 Correct an error in the documentation for the redirect router. Exim
2010 does (usually) call initgroups() when daemonizing.
2012 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2013 with consistent privilege compared to when running as a daemon.
2015 TF/03 Note in the spec that $authenticated_id is not set for local
2016 submissions from trusted users.
2018 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2019 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2021 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2022 by adding some example configuration directives to the default
2023 configuration file. A little bit of work is required to uncomment the
2024 directives and define how usernames and passwords are checked, but
2025 there is now a framework to start from.
2027 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2028 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2029 without this. I don't know how relevant this is to other LDAP libraries.
2031 PH/02 Add the verb name to the "unknown ACL verb" error.
2033 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2035 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2037 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2038 directory not expanded when it should be if an expanded home directory
2039 was set for the address (which is overridden by the transport).
2041 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2044 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2045 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2046 because it is too late at that time, and has no effect.
2048 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2049 security issue with \' (bugzilla #107). I could not use the
2050 PQescapeStringConn() function, because it needs a PGconn value as one of
2053 PH/08 When testing addresses using -bt, indicate those final addresses that
2054 are duplicates that would not cause an additional delivery. At least one
2055 person was confused, thinking that -bt output corresponded to deliveries.
2056 (Suppressing duplicates isn't a good idea as you lose the information
2057 about possibly different redirections that led to the duplicates.)
2059 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2060 systems where poll() doesn't work, in particular OS X.
2062 PH/10 Added more information to debugging output for retry time not reached.
2064 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2065 operations in malware.c.
2067 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2070 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2071 syslog_duplication set false, log lines that would normally be written
2072 both the the main log and to the reject log were not written to syslog at
2075 PH/14 In the default configuration, change the use of "message" in ACL warn
2076 statements to "add_header".
2078 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2079 not followed by a command (e.g. "seen endif").
2081 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2082 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2085 PH/17 Added extra conditions to the default value of delay_warning_condition
2089 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2090 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2091 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2094 The Auto-Submitted: and various List- headers are standardised, whereas I
2095 don't think Precedence: ever was.
2097 PH/18 Refactored debugging code in route_finduser() to show more information,
2098 in particular, the error code if getpwnam() issues one.
2100 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2101 This is apparently needed in addition to the PH/07 change above to avoid
2102 any possible encoding problems.
2104 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2105 but not after initializing Perl.
2107 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2108 output them only if debugging. By default they are written stderr,
2109 apparently, which is not desirable.
2111 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2114 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2117 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2119 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2120 authenticated or an ident call has been made. Suppress the default
2121 values for $authenticated_id and $authenticated_sender (but permit -oMai
2122 and -oMas) when testing with -bh.
2124 PH/24 Re-jigged the order of the tests in the default configuration so that the
2125 tests for valid domains and recipients precede the DNS black list and CSA
2126 tests, on the grounds that those ones are more expensive.
2128 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2129 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2130 This bug exists in every version of Exim that I still have, right back to
2133 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2134 However, an attempt to turn on case-sensitivity in a regex key by
2135 including (?-i) didn't work because the subject string was already
2136 lowercased, and the effects were non-intuitive. It turns out that a
2137 one-line patch can be used to allow (?-i) to work as expected.
2143 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2144 other effects) broke the use of negated acl sub-conditions.
2146 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2149 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2150 "Deny" causes Exim to reject the incoming connection with a 554 error.
2151 Unfortunately, if there is a major crisis, such as a disk failure,
2152 tcp-wrappers gives "deny", whereas what one would like would be some
2153 kind of temporary error. A kludge has been added to help with this.
2154 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2155 554 error is used if errno is still zero or contains ENOENT (which occurs
2156 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2159 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2161 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2162 errors. Otherwise a message that provokes a temporary error (when other
2163 messages do not) can cause a whole host to time out.
2165 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2166 addresses were routed directly to files or pipes from a redirect router.
2167 File deliveries just didn't batch; pipe deliveries might have suffered
2170 PH/06 A failure to get a lock for a hints database would erroneously always say
2171 "Failed to get write lock", even when it was really a read lock.
2173 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2174 of 0600. This has been changed to use the value of the lockfile_mode
2175 option (which defaults to 0600).
2177 PH/08 Applied small patch from the Sieve maintainer.
2179 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2180 folder from quota calculations, a direct delivery into this folder messed
2181 up the contents of the maildirsize file. This was because the regex was
2182 used only to exclude .Trash (or whatever) when the size of the mailbox
2183 was calculated. There was no check that a delivery was happening into an
2184 excluded directory. This bug has been fixed by ignoring all quota
2185 processing for deliveries into excluded directories.
2187 PH/10 Added the maildirfolder_create_regex option to appendfile.
2193 PH/01 The code for finding all the local interface addresses on a FreeBSD
2194 system running IPv6 was broken. This may well have applied to all BSD
2195 systems, as well as to others that have similar system calls. The broken
2196 code found IPv4 interfaces correctly, but gave incorrect values for the
2197 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2198 that it would not match correctly against @[] and not recognize the IPv6
2201 PH/02 The ipliteral router was not recognizing addresses of the form user@
2202 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2204 PH/03 Added disable_ipv6.
2206 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2207 decoded form, because it is most often used to construct To: headers
2208 lines in autoreplies, and the decoded form may well be syntactically
2209 invalid. However, $reply_address has leading white space removed, and all
2210 newlines turned into spaces so that the autoreply transport does not
2213 PH/05 If group was specified without a user on a router, and no group or user
2214 was specified on a transport, the group from the router was ignored.
2216 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2217 for visible compile-time settings that can be used to change these
2218 numbers, for those that want even more. Backwards compatibility with old
2219 spool files has been maintained. However, going back to a previous Exim
2220 release will lost any variables that are in spool files.
2222 PH/07 Two small changes when running in the test harness: increase delay when
2223 passing a TCP/IP connection to a new process, in case the original
2224 process has to generate a bounce, and remove special handling of
2225 127.0.0.2 (sic), which is no longer necessary.
2227 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2228 be the same on different OS.
2230 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2233 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2234 whether --show-vars was specified or not
2236 JJ/02 exipick: Added support for new ACL variable spool format introduced
2239 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2240 syntactically invalid From: or Reply-to: line, and a filter used this to
2241 generate an autoreply, and therefore failed to obtain an address for the
2242 autoreply, Exim could try to deliver to a non-existent relative file
2243 name, causing unrelated and misleading errors. What now happens is that
2244 it logs this as a hard delivery error, but does not attempt to create a
2247 PH/11 The exinext utility has a -C option for testing purposes, but although
2248 the given file was scanned by exinext itself; it wasn't being passed on
2249 when Exim was called.
2251 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2252 an end-of-file indication when reading a command response.
2254 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2255 compiled. In many other places in Exim, IPv6 addresses are always
2256 recognized, so I have changed this. It also means that IPv4 domain
2257 literals of the form [IPV4:n.n.n.n] are now always recognized.
2259 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2260 used if the router is not running as root, for example, when verifying at
2261 ACL time, or when using -bh. The debugging output from this situation was
2262 non-existent - all you got was a failure to exec. I have made two
2265 (a) Failures to set uid/gid, the current directory, or a process leader
2266 in a subprocess such as that created by queryprogram now generate
2267 suitable debugging ouput when -d is set.
2269 (b) The queryprogram router detects when it is not running as root,
2270 outputs suitable debugging information if -d is set, and then runs
2271 the subprocess without attempting to change uid/gid.
2273 PH/15 Minor change to Makefile for building test_host (undocumented testing
2276 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2277 additional section of a DNS packet that returns MX or SRV records.
2278 Instead, it always explicitly searches for A/AAAA records. This avoids
2279 major problems that occur when a DNS server includes only records of one
2280 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2281 fixed another bug: if SRV records were looked up and the corresponding
2282 address records were *not* found in the additional section, the port
2283 values from the SRV records were lost.
2285 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2286 using the correct key (the original address) when searching the retry
2287 rules in order to find which one to use for generating the retry hint.
2289 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2290 adding the default one. Similarly, if it contains a Reply-To: header, the
2291 errors_reply_to option, if set, is not used.
2293 PH/19 When calculating a retry time, Exim used to measure the "time since
2294 failure" by looking at the "first failed" field in the retry record. Now
2295 it does not use this if it is later than than the arrival time of the
2296 message. Instead it uses the arrival time. This makes for better
2297 behaviour in cases where some deliveries succeed, thus re-setting the
2298 "first failed" field. An example is a quota failure for a huge message
2299 when small messages continue to be delivered. Without this change, the
2300 "time since failure" will always be short, possible causing more frequent
2301 delivery attempts for the huge message than are intended.
2302 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2304 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2305 $1, $2, $3) because the numerical variables can be reset during some
2306 expansion items (e.g. "match"), thereby losing the authentication data.
2308 PH/21 Make -bV show the size of off_t variables so that the test suite can
2309 decide whether to run tests for quotas > 2G.
2311 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2312 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2313 filecount value is greater than 2G or if a quota value is greater than 2G
2314 on a system where the size of off_t is not greater than 4, a panic error
2317 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2318 never match. The debug and -bh output now contains an explicit error
2319 message indicating a malformed IPv4 address or mask.
2321 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2322 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2323 PH/23 above applies.
2325 PH/25 Do not write to syslog when running in the test harness. The only
2326 occasion when this arises is a failure to open the main or panic logs
2327 (for which there is an explicit test).
2329 PH/26 Added the /no_tell option to "control=freeze".
2331 PH/27 If a host name lookup failed very early in a connection, for example, if
2332 the IP address matched host_lookup and the reverse lookup yielded a name
2333 that did not have a forward lookup, an error message of the form "no IP
2334 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2335 could be logged. Now it outputs the IP address instead of "NULL".
2337 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2338 allows the sender and the authenticated sender to be set when
2339 submitting a message from within Exim. Since child_open_exim() is
2340 documented for local_scan(), the new function should be too.
2342 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2343 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2344 results in an empty string is now treated as unset.
2346 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2348 PH/31 Added sender_verify_fail logging option.
2350 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2351 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2352 tidied the source and removed it altogether.
2354 PH/33 When a queue run was abandoned because the load average was too high, a
2355 log line was always written; now it is written only if the queue_run log
2356 selector is set. In addition, the log line for abandonment now contains
2357 information about the queue run such as the pid. This is always present
2358 in "start" and "stop" lines but was omitted from the "abandon" line.
2360 PH/34 Omit spaces between a header name and the colon in the error message that
2361 is given when verify = headers_syntax fails (if there are lots of them,
2362 the message gets confusing).
2364 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2365 names, as there are now some PTR records that contain slashes. This check
2366 is only to protect against broken name servers that fall over on strange
2367 characters, so the fact that it applies to all lookups doesn't matter.
2369 PH/36 Now that the new test suite is complete, we can remove some of the
2370 special code in Exim that was needed for the old test suite. For example,
2371 sorting DNS records because real resolvers return them in an arbitrary
2372 order. The new test suite's fake resolver always returns records in the
2375 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2376 bounces) except when queue_only is set, to avoid logging races between
2377 the different processes.
2379 PH/38 Panic-die if .include specifies a non-absolute path.
2381 PH/39 A tweak to the "H" retry rule from its user.
2383 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2384 a label. They prevented compilation on older perls.
2386 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2387 a warning to be raised on newish perls.
2389 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2390 on queue. Changes to match documented behaviour of showing count of
2391 messages matching specified criteria.
2393 PH/40 Changed the default ident timeout from 30s to 5s.
2395 PH/41 Added support for the use of login_cap features, on those BSD systems
2396 that have them, for controlling the resources used by pipe deliveries.
2398 PH/42 The content-scanning code uses fopen() to create files in which to put
2399 message data. Previously it was not paying any attention to the mode of
2400 the files. Exim runs with umask(0) because the rest of the code creates
2401 files with open(), and sets the required mode explicitly. Thus, these
2402 files were ending up world-writeable. This was not a big issue, because,
2403 being within the spool directory, they were not world-accessible. I have
2404 created a function called modefopen, which takes an additional mode
2405 argument. It sets umask(777), creates the file, chmods it to the required
2406 mode, then resets the umask. All the relevant calls to fopen() in the
2407 content scanning code have been changed to use this function.
2409 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2410 to 24 hours. This avoids potential overflow problems when processing G
2411 and H retry rules. I suspect nobody ever tinkers with this value.
2413 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2415 PH/45 When the plaintext authenticator is running as a client, the server's
2416 challenges are checked to ensure they are valid base64 strings. By
2417 default, the authentication attempt is cancelled if an invalid string is
2418 received. Setting client_ignore_invalid_base64 true ignores these errors.
2419 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2420 they are received. Thus, the responses can be made to depend on the
2421 challenges. If an invalid string is ignored, an empty string is placed in
2424 PH/46 Messages that are created by the autoreply transport now contains a
2425 References: header, in accordance with RFCs 2822 and 3834.
2427 PH/47 Added authenticated_sender_force to the smtp transport.
2429 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2431 PH/49 Installed latest patch from the Sieve maintainer.
2433 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2434 was also set, the appendfile transport was unnecessarily scanning a
2435 directory of message files (e.g. for maildir delivery) to find the count
2436 of files (along with the size), even though it did not need this
2437 information. It now does the scan only if it needs to find either the
2438 size of the count of files.
2440 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2442 PH/52 Two bugs concerned with error handling when the smtp transport is
2445 (i) Exim was not creating retry information for temporary errors given
2446 for individual recipients after the DATA command when the smtp transport
2447 was used in LMTP mode. This meant that they could be retried too
2448 frequently, and not timed out correctly.
2450 (ii) Exim was setting the flag that allows error details to be returned
2451 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2452 recipients that were returned after the DATA command.
2454 PH/53 This is related to PH/52, but is more general: for any failing address,
2455 when detailed error information was permitted to be returned to the
2456 sender, but the error was temporary, then after the final timeout, only
2457 "retry timeout exceeded" was returned. Now it returns the full error as
2458 well as "retry timeout exceeded".
2460 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2461 do this, and (what is worse) MTAs that accept it.
2463 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2464 will now be deprecated.
2466 PH/56 New os.c-cygwin from the Cygwin maintainer.
2468 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2469 formats (previously only available in exim formats via -bpr, -bpru,
2470 and -bpra. Now also available in native and exiqgrep formats)
2472 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2473 with very large, slow to parse queues
2475 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2477 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2479 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2480 responses to authentication challenges, though it was showing the
2481 challenges; (ii) I've removed the CR characters from the debug output for
2484 PH/58 Allow for the insertion of a newline as well as a space when a string
2485 is turned into more than one encoded-word during RFC 2047 encoding. The
2486 Sieve code now uses this.
2488 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2489 data_4xx, lost_connection, tls_required.
2491 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2492 message was being sent as an SMTP response.
2494 PH/61 Add -l and -k options to exicyclog.
2496 PH/62 When verifying, if an address was redirected to one new address, so that
2497 verification continued, and the new address failed or deferred after
2498 having set something in $address_data, the value of $address_data was not
2499 passed back to the ACL. This was different to the case when no
2500 redirection occurred. The value is now passed back in both cases.
2502 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2503 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2504 use login_cap.h, so on its own it isn't the distinguishing feature. The
2505 new name refers directly to the setclassresources() function.
2507 PH/65 Added configuration files for NetBSD3.
2509 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2511 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2512 is preferred over IPv4.
2514 PH/68 The bounce_return_message and bounce_return_body options were not being
2515 honoured for bounces generated during the reception of non-SMTP messages.
2516 In particular, this applied to messages rejected by the ACL. This bug has
2517 been fixed. However, if bounce_return_message is true and bounce_return_
2518 body is false, the headers that are returned for a non-SMTP message
2519 include only those that have been read before the error was detected.
2520 (In the case of an ACL rejection, they have all been read.)
2522 PH/69 The HTML version of the specification is now built in a directory called
2523 spec_html instead of spec.html, because the latter looks like a path with
2524 a MIME-type, and this confuses some software.
2526 PH/70 Catch two compiler warnings in sieve.c.
2528 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2529 function verify_get_ident() calls ip_connect() to connect a socket, but
2530 if the "connect()" function timed out, ip_connect() used to close the
2531 socket. However, verify_get_ident() also closes the socket later, and in
2532 between Exim writes to the log, which may get opened at this point. When
2533 the socket was closed in ip_connect(), the log could get the same file
2534 descriptor number as the socket. This naturally causes chaos. The fix is
2535 not to close the socket in ip_connect(); the socket should be closed by
2536 the function that creates it. There was only one place in the code where
2537 this was missing, in the iplookup router, which I don't think anybody now
2538 uses, but I've fixed it anyway.
2540 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2541 well as to direct DNS lookups. Otherwise the handling of names in host
2542 lists is inconsistent and therefore confusing.
2548 PH/01 Two changes to the default runtime configuration:
2550 (1) Move the checks for relay_from_hosts and authenticated clients from
2551 after to before the (commented out) DNS black list checks.
2553 (2) Add control=submission to the relay_from_hosts and authenticated
2554 clients checks, on the grounds that messages accepted by these
2555 statements are most likely to be submissions.
2557 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2559 (1) Generate an error if the third argument for the ${prvs expansion is
2562 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2565 (3) Reset the variables that are obtained from the first argument of
2566 ${prvscheck and used in the second argument before leaving the code,
2567 because their memory is reclaimed, so using them afterwards may do
2570 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2571 one (it's much easier than Tom thought :-).
2573 (5) Because of (4), we can now allow for the use of $prvscheck_result
2574 inside the third argument.
2576 PH/03 For some reason, the default setting of PATH when running a command from
2577 a pipe transport was just "/usr/bin". I have changed it to
2580 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2581 anything to be listed in the output from -bV.
2583 PH/05 When a filter generated an autoreply, the entire To: header line was
2584 quoted in the delivery log line, like this:
2586 => >A.N.Other <ano@some.domain> <original@ddress> ...
2588 This has been changed so that it extracts the operative address. There
2589 may be more than one such address. If so, they are comma-separated, like
2592 => >ano@some.domain,ona@other.domain <original@ddress> ...
2594 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2595 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2596 not being looked up in the rDNS to get a host name, Exim was showing the
2597 IP address twice in Received: lines, even though the IP addresses were
2598 identical. For example:
2600 Received: from [1.2.3.4] (helo=[1.2.3.4])
2602 However, if the real host name was known, it was omitting the HELO data
2603 if it matched the actual IP address. This has been tidied up so that it
2604 doesn't show the same IP address twice.
2606 PH/07 When both +timestamp and +memory debugging was on, the value given by
2607 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2608 called by the debug printing, thereby overwriting the timestamp buffer.
2609 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2611 PH/08 When the original message was included in an autoreply transport, it
2612 always said "this is a copy of the message, including all the headers",
2613 even if body_only or headers_only was set. It now gives an appropriate
2616 PH/09 Applied a patch from the Sieve maintainer which:
2618 o fixes some comments
2619 o adds the (disabled) notify extension core
2620 o adds some debug output for the result of if/elsif tests
2621 o points to the current vacation draft in the documentation
2622 and documents the missing references header update
2626 o fixes a bug in processing the envelope test (when testing
2627 multiple envelope elements, the last element determinted the
2630 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2631 Electronic Mail") by including:
2633 Auto-submitted: auto-generated
2635 in the messages that it generates (bounce messages and others, such as
2636 warnings). In the case of bounce messages for non-SMTP mesages, there was
2637 also a typo: it was using "Auto_submitted" (underscore instead of
2638 hyphen). Since every message generated by Exim is necessarily in response
2639 to another message, thes have all been changed to:
2641 Auto-Submitted: auto-replied
2643 in accordance with these statements in the RFC:
2645 The auto-replied keyword:
2647 - SHOULD be used on messages sent in direct response to another
2648 message by an automatic process,
2650 - MUST NOT be used on manually-generated messages,
2652 - MAY be used on Delivery Status Notifications (DSNs) and Message
2653 Disposition Notifications (MDNs),
2655 - MUST NOT be used on messages generated by automatic or periodic
2656 processes, except for messages which are automatic responses to
2659 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2660 to the default Received: header definition.
2662 PH/12 Added log selector acl_warn_skipped (default on).
2664 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2665 variables because (a) they are in the wrong storage pool and (b) even if
2666 they were copied, it wouldn't work properly because of the caching.
2668 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2669 checking when decoding. Apparently there are clients that generate
2670 overlong encoded strings. Why am I not surprised?
2672 PH/15 If the first argument of "${if match_address" was not empty, but did not
2673 contain an "@" character, Exim crashed. Now it writes a panic log message
2674 and treats the condition as false.
2676 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2678 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2679 extension "envelope-auth". The code is finished and in agreement with
2680 other implementations, but there is no documentation so far and in fact,
2681 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2682 not changing the active code.
2684 Print executed "if" and "elsif" statements when debugging is used. This
2685 helps a great deal to understand what a filter does.
2687 Document more things not specified clearly in RFC3028. I had all this
2688 sorted out, when out of a sudden new issues came to my mind. Oops."
2690 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2693 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2694 canonical form (as documented). However, after a host name lookup from
2695 the IP address, check_host() was doing a simple string comparison with
2696 addresses acquired from the DNS when checking that the found name did
2697 have the original IP as one of its addresses. Since any found IPv6
2698 addresses are likely to be in abbreviated form, the comparison could
2699 fail. Luckily, there already exists a function for doing the comparison
2700 by converting both addresses to binary, so now that is used instead of
2701 the text comparison.
2703 PH/20 There was another similar case to PH/19, when a complete host name was
2704 given in a host list; looking up its IP address could give an abbreviated
2705 form, whereas the current host's name might or might not be abbreviated.
2706 The same fix has been applied.
2712 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2713 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2716 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2717 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2719 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2721 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2722 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2723 Also, for "Auto-Submitted", treat anything other than "no" as
2724 non-personal, in accordance with RFC 3834. (Previously it treated
2725 anything starting "auto-" as non-personal.)
2727 TF/01 The control=submission/name=... option had a problem with syntax
2728 errors if the name included a slash character. The /name= option
2729 now slurps the rest of the string, so it can include any characters
2730 but it must come last in the list of options (after /sender_retain
2733 PH/05 Some modifications to the interface to the fake nameserver for the new
2741 TK/01 Added the "success_on_redirect" address verification option. See
2742 NewStuff for rationale and an example.
2744 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2746 PH/02 Patch to exigrep to allow it to work on syslog lines.
2748 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2749 fread() to skip over the body file's header line, because in Cygwin the
2750 header line is locked and is inaccessible.
2752 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2753 co-exist for some time) to make it clear that it is the Exim ID that is
2754 referenced, not the Message-ID: header line.
2756 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2757 string_format() function, because snprintf() does not exist on all
2760 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2761 ${stat: expansion item.
2763 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2764 protocol synchronization error", to keep the pedants happy.
2766 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2767 well as for IRIX systems, when gcc is being used. See the host.c source
2770 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2772 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2775 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2776 transport and to the smtp transport in LMTP mode.
2778 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2780 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2781 run a filter in a subprocess. This could lead to confusion in subsequent
2782 lookups in the parent process. There should also be a search_tidyup() at
2783 the end of the subprocess.
2785 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2786 only if the host matched helo_try_verify_hosts, which caused the
2787 verification to occur when the EHLO/HELO command was issued. The ACL just
2788 tested the remembered result. Now, if a previous verification attempt has
2789 not happened, "verify = helo" does it there and then.
2791 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2793 TK/03 Fix log output including CR from clamd.
2795 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2796 exist provoked a memory error which could cause a segfault.
2798 PH/15 Installed PCRE 6.2
2800 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2802 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2803 of the problem. Specifically, suggested +O2 rather than +O1 for the
2806 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2808 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2809 record, the port was not correctly logged when the outgoing_port log
2810 selector was set (it logged the transort's default port).
2812 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2813 fallback_hosts, and "hosts" in the smtp transport.
2815 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2816 host errors such as "Connection refused".
2818 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2819 authentication with radiusclient 0.4.9:
2821 - Error returned from rc_read_config was caught wrongly
2822 - Username/password not passed on to radius server due to wrong length.
2824 The presumption is that some radiusclient API changes for 4.51/PH/17
2825 were not taken care of correctly. The code is still untested by me (my
2826 Linux distribution still has 0.3.2 of radiusclient), but it was
2827 contributed by a Radius user.
2829 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2830 expanding the "port" option of the smtp transport.
2832 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2833 while reading a MIME header. Thanks to Tom Hughes for a patch.
2835 PH/24 Include config.h inside local_scan.h so that configuration settings are
2838 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2839 that in an ACL for RCPT (for example), you can examine exactly what was
2842 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2843 commands, but it was not correctly comparing the address with the actual
2844 client host address. Thus, it would show the EHLO address in Received:
2845 header lines when this was not necessary.
2847 PH/27 Added the % operator to ${eval:}.
2849 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2850 it should be ignoring failures (because with -C, for example, it has lost
2851 privilege). It wasn't ignoring creation failures other than "already
2854 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2857 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2858 because an input line was too long, either on its own, or by virtue of
2859 too many continuations, the temporary file was not being removed, and the
2860 return code was incorrect.
2862 PH/31 Missing "BOOL" in function definition in filtertest.c.
2864 PH/32 Applied Sieve patches from the maintainer.
2866 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2868 PH/33 Added "verify = not_blind".
2870 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2871 Local/Makefile (with some defaults set). These are used in built scripts
2872 such as exicyclog, but they have never been used in the exim_install
2873 script (though there are many overriding facilities there). I have
2874 arranged that the exim_install script now takes note of these two
2877 PH/35 Installed configuration files for Dragonfly.
2879 PH/36 When a locally submitted message by a trusted user did not contain a
2880 From: header, and the sender address was obtained from -f or from an SMTP
2881 MAIL command, and the trusted user did not use -F to supply a sender
2882 name, $originator_name was incorrectly used when constructing a From:
2883 header. Furthermore, $originator_name was used for submission mode
2884 messages from external hosts without From: headers in a similar way,
2885 which is clearly wrong.
2887 PH/37 Added control=suppress_local_fixups.
2889 PH/38 When log_selector = +received_sender was set, and the addition of the
2890 sender made the log line's construction buffer exactly full, or one byte
2891 less than full, an overflow happened when the terminating "\n" was
2894 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2895 when the result of a list match is failure because a DNS lookup failed.
2897 PH/40 RM_COMMAND is now used in the building process.
2899 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2900 the "build-* directories that it finds.
2902 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2903 address in a domain literal was a prefix of an interface address.
2905 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2906 when verifying a sender address, unless rewrite_headers is false.
2908 PH/44 Wrote a long comment about why errors_to addresses are verified as
2909 recipients, not senders.
2911 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2912 the ratelimit ACL was added.
2914 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2916 PH/46 Added extra information about PostgreSQL errors to the error string.
2918 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2919 avoiding the need to install special zones in a real server. This is
2920 backwards compatible; if it can't find the fake resolver, it drops back.
2921 Thus, both old and new test suites can be run.
2923 TF/02 Added util/ratelimit.pl
2925 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2926 clock is set back in time.
2928 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2929 Candler <B.Candler@pobox.com>.
2931 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2932 OK for addresses that are the result of redirections.
2934 PH/48 A number of further additions for the benefit of the new test suite,
2935 including a fake gethostbyname() that interfaces to the fake DNS resolver
2938 TF/06 The fix for widen_domains has also been applied to qualify_single and
2939 search_parents which are the other dnslookup options that can cause
2942 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2945 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2947 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2948 service" if it encountered an MX record with an empty target hostname.
2949 The message is now "an MX or SRV record indicated no SMTP service".
2951 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2952 if the DNS of the sending site is misconfigured. This is quite a
2953 common situation. This change restores the behaviour of treating a
2954 helo verification defer as a failure.
2956 PH/51 If self=fail was set on a router, the bounce message did not include the
2957 actual error message.
2963 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2965 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2966 command itself ended in error, the underlying message about the transport
2967 filter timeout was being overwritten with the pipe command error. Now the
2968 underlying error message should be appended to the second error message.
2970 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2972 PH/02 Reduce the amount of output that "make" produces by default. Full output
2973 can still be requested.
2975 PH/03 The warning log line about a condition test deferring for a "warn" verb
2976 was being output only once per connection, rather than after each
2977 occurrence (because it was using the same function as for successful
2978 "warn" verbs). This seems wrong, so I have changed it.
2980 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2981 it should not have, which might have caused a crash in the right
2982 circumstances, but probably never did.
2984 PH/04 Installed a modified version of Tony Finch's patch to make submission
2985 mode fix the return path as well as the Sender: header line, and to
2986 add a /name= option so that you can make the user's friendly name appear
2989 TF/03 Added the control = fakedefer ACL modifier.
2991 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2992 Mark Lowes for thorough testing.
2994 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2996 TK/03 Merged latest SRS patch from Miles Wilton.
2998 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2999 with the definition in sysexits.h (which is #included earlier).
3000 Fortunately, Exim does not actually use EX_OK. The code used to try to
3001 preserve the sysexits.h value, by assumimg that macro definitions were
3002 scanned for macro replacements. I have been disabused of this notion,
3003 so now the code just undefines EX_OK before #including unistd.h.
3005 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3006 in the smtp transport. When a block could not be written in a single
3007 write() function, the timeout was being re-applied to each part-write.
3008 This seems wrong - if the receiver was accepting one byte at a time it
3009 would take for ever. The timeout is now adjusted when this happens. It
3010 doesn't have to be particularly precise.
3012 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3013 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3015 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3016 without a domain if the check to <postmaster@domain> fails.
3018 SC/01 Eximstats: added -xls and the ability to specify output files
3019 (patch written by Frank Heydlauf).
3021 SC/02 Eximstats: use FileHandles for outputing results.
3023 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3025 SC/04 Eximstats: fixed display of large numbers with -nvr option
3027 SC/05 Eximstats: fixed merging of reports with empty tables.
3029 SC/06 Eximstats: added the -include_original_destination flag
3031 SC/07 Eximstats: removed tabs and trailing whitespace.
3033 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3035 TK/06 MBOX spool code: Add real "From " MBOX separator line
3036 so the .eml file is really in mbox format (even though
3037 most programs do not really care). Patch from Alex Miller.
3039 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3040 The latter is generated from $received_to and is only set if the
3041 message has one envelope recipient. SA can use these headers,
3042 obviously out-of-the-box. Patch from Alex Miller.
3044 PH/08 The ${def test on a variable was returning false if the variable's
3045 value was "0", contrary to what the specification has always said!
3046 The result should be true unless the variable is empty.
3048 PH/09 The syntax error of a character other than { following "${if
3049 def:variable_name" (after optional whitespace) was not being diagnosed.
3050 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3051 accidental colon was present, for example, could give incorrect results.
3053 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3054 result is used (not including appendfile, where other changes are about
3057 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3058 This involved changing a lot of size variables from int to off_t. It
3059 should work with maildirs and everything.
3061 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3062 spamd dying while we are connected to it.
3064 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3067 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3068 function for BDB 4.3.
3070 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3072 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3073 the sources for PCRE have been re-arranged and more files are now
3076 PH/15 The code I had for printing potentially long long variables in PH/11
3077 above was not the best (it lost precision). The length of off_t variables
3078 is now inspected at build time, and an appropriate printing format (%ld
3079 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3080 to be "long long int" or "long int". This is needed for the internal
3081 formatting function string_vformat().
3083 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3084 the configuration file to be ":syslog", then the script "guesses" where
3085 the logs files are, rather than using the compiled in default. In our
3086 case the guess is not the same as the compiled default, so the script
3087 suddenly stopped working when I started to use syslog. The patch checks
3088 to see if log_file_path is "". If so, it attempts to read it from exim
3089 with no configuration file to get the compiled in version, before it
3090 falls back to the previous guessing code."
3092 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3093 implementing BATV in an Exim configuration. See NewStuff for the gory
3096 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3097 Makefile that are specific to HP-UX.
3099 PH/18 If the "use_postmaster" option was set for a recipient callout together
3100 with the "random" option, the postmaster address was used as the MAIL
3101 FROM address for the random test, but not for the subsequent recipient
3102 test. It is now used for both.
3104 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3105 patch removes a few documentation additions to RFC 3028, because the
3106 latest draft now contains them. It adds the new en;ascii-case comparator
3107 and a new error check for 8bit text in MIME parts. Comparator and
3108 require names are now matched exactly. I enabled the subaddress
3109 extension, but it is not well tested yet (read: it works for me)."
3111 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3112 rework some of the code of TK/09 above to avoid the hardwired use of
3113 "%lld" and "long long". Replaced the call to snprintf() with a call to
3116 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3117 records point to non-existent hosts", "retry timeout exceeded", and
3118 "retry time not reached for any host after a long failure period".
3120 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3121 experimental DomainKeys support:
3123 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3124 (2) On an error such as an illegally used "control", the wrong name for
3125 the control was given.
3127 These problems did NOT occur unless DomainKeys support was compiled.
3129 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3131 PH/24 Added ${if match_ip condition.
3133 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3134 databases so that it will be absolutely obvious if a crash occurs in the
3135 DB library. This is a regular occurrence (often caused by mis-matched
3138 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3139 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3140 picked up on a user's system that detects such things. There doesn't seem
3141 to be a gcc warning option for this - only an attribute that has to be
3142 put on the function's prototype. It seems that in Fedora Core 4 they have
3143 set this on a number of new functions. No doubt there will be more in due
3146 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3147 specify a transport. However, if an address that was verified by such a
3148 router was the subject of a callout, Exim crashed because it tried to
3149 read the rcpt_include_affixes from the non-existent transport. Now it
3150 just assumes that the setting of that option is false. This bug was
3151 introduced by 4.51/PH/31.
3153 PH/28 Changed -d+all to exclude +memory, because that information is very
3154 rarely of interest, but it makes the output a lot bigger. People tend to
3155 do -d+all out of habit.
3157 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3158 code in os-type was giving problems when libc.so lives in lib64, like on
3161 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3162 aren't the modern standard, and it seems that some systems' include files
3163 don't always have them. Exim was already checking for some of the newer
3164 ones like T_AAAA, and defining it itself. I've added checks for all the
3165 record types that Exim uses.
3167 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3168 not automatically generating a new one, as it is supposed to. This
3169 prevented TLS from working. If the file did exist, but contained invalid
3170 data, a new version was generated, as expected. It was only the case of a
3171 non-existent file that was broken.
3173 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3174 with a change in libdomainkeys > 0.64.
3176 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3177 from DNS. If the selector record carries the flag, it now has
3178 precedence over the domain-wide flag.
3180 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3182 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3183 the use of a port name that isn't defined in /etc/services) occurred, the
3184 message was deferred as in a normal delivery, and thus remained on the
3185 spool, instead of being failed because of the mua_wrapper setting. This
3186 is now fixed, and I tidied up some of the mua_wrapper messages at the
3189 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3190 the messages in a hash of arrays rather than using individual hashes.
3191 This is a bit cleaner and results in dramatic memory savings, albeit
3192 at a slight CPU cost.
3194 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3195 as requested by Marc Sherman.
3197 SC/10 Eximstats: added histograms for user specified patterns as requested
3200 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3202 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3203 fopen() in the content-scanning modules that did not already have it.
3209 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3210 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3212 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3214 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3216 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3217 to test Sieve filters that use "vacation".
3219 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3220 that changes the way the GnuTLS parameters are stored in the cache file.
3221 The new format can be generated externally. For backward compatibility,
3222 if the data in the cache doesn't make sense, Exim assumes it has read an
3223 old-format file, and it generates new data and writes a new file. This
3224 means that you can't go back to an older release without removing the
3227 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3228 work if there are any delivery delays because "one_time" forces the
3229 parent to be marked "delivered", so its unseen clone is never tried
3230 again. For this reason, Exim now forbids the simultaneous setting of
3233 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3234 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3235 entry if you want to know the details. The fix, however, neglected to
3236 consider the case when local delivery batching is involved. The test for
3237 "previously delivered" was not happening when checking to see if an
3238 address could be batched with a previous (undelivered) one; under
3239 certain circumstances this could lead to multiple deliveries to the same
3242 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3243 in its include files, and this causes problems building Exim.
3245 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3246 header_syntax) but Exim was just ignoring anything given after a slash.
3247 In particular, this caused confusion with an attempt to use "verify =
3248 reverse_host_lookup/defer_ok". An error is now given when options are
3249 supplied for verify items that do not have them. (Maybe reverse_host_
3250 lookup should have a defer_ok option, but that's a different point.)
3252 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3253 defined by RFC 821) to 2048, because there were problems with some AUTH
3254 commands, and RFC 1869 says the size should be increased for extended
3255 SMTP commands that take arguments.
3257 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3260 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3261 "unknown" error; now it says that the functionality isn't in the binary.
3263 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3264 an address' error message when a string expansion fails (syntax or
3265 whatever). Otherwise the password may appear in the log. Following change
3266 PH/42 below, there is no longer a chance of it appearing in a bounce
3269 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3271 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3272 of its addresses was ever tried. (Bugzilla bug #2.)
3274 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3275 the result incorrectly in the debug output. (It correctly added a newline
3276 to what was transported.)
3278 TF/01 Added $received_time.
3280 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3281 commented out examples of how to interface to a virus scanner and to
3282 SpamAssassin. Also added commented examples of av_scanner and
3283 spamd_address settings.
3285 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3286 and controls are allowed in which ACLs. There were a couple of minor
3287 errors. Some of the entries in the conditions table (which is a table of
3288 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3289 negation of where the condition IS allowed.
3291 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3293 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3294 header file does not have a version number, so I've had to invent a new
3295 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3296 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3297 radiusclient), but it was contributed by a Radius user.
3299 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3300 files or directories, for OpenSSL.
3302 PH/19 When an Exim process that is running as root has to create an Exim log
3303 file, it does so in a subprocess that runs as exim:exim so as to get the
3304 ownership right at creation (otherwise, other Exim processes might see
3305 the file with the wrong ownership). There was no test for failure of this
3306 fork() call, which would lead to the process getting stuck as it waited
3307 for a non-existent subprocess. Forks do occasionally fail when resources
3308 run out. I reviewed all the other calls to fork(); they all seem to check
3311 PH/20 When checking for unexpected SMTP input at connect time (before writing
3312 the banner), Exim was not dealing correctly with a non-positive return
3313 from the read() function. If the client had disconnected by this time,
3314 the result was a log entry for a synchronization error with an empty
3315 string after "input=" when read() returned zero. If read() returned -1
3316 (an event I could not check), uninitialized data bytes were printed.
3317 There were reports of junk text (parts of files, etc) appearing after
3320 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3322 PH/22 Added support for macro redefinition, and (re)definition in between
3323 driver and ACL definitions.
3325 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3326 forgetting to use the resulting value; it was using the unexpanded value.
3328 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3329 hadn't been configured. The fix is from Juergen Kreileder, who
3330 understands it better than I do:
3332 "Here's what I see happening with three configured cyrus_sasl
3333 authenticators configured (plain, login, cram-md5):
3335 On startup auth_cyrus_sasl_init() gets called for each of these.
3336 This means three calls to sasl_listmech() without a specified mech_list.
3337 => SASL tests which mechs of all available mechs actually work
3338 => three warnings about OTP not working
3339 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3341 With the patch, sasl_listmech() also gets called three times. But now
3342 SASL's mech_list option is set to the server_mech specified in the the
3343 authenticator. Or in other words, the answer from sasl_listmech()
3344 gets limited to just the mech you're testing for (which is different
3346 => the return list contains just 'plain' or 'login', 'cram-md5' or
3347 nothing depending on the value of ob->server_mech.
3349 I've just tested the patch: Authentication still works fine,
3350 unavailable mechs specified in the exim configuration are still
3351 caught, and the auth.log warnings about OTP are gone."
3353 PH/25 When debugging is enabled, the contents of the command line are added
3354 to the debugging output, even when log_selector=+arguments is not
3357 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3358 answer is "GNU", and only if the return is "GNU/something" is the answer
3361 PH/27 $acl_verify_message is now set immediately after the failure of a
3362 verification in an ACL, and so is available in subsequent modifiers. In
3363 particular, the message can be preserved by coding like this:
3365 warn !verify = sender
3366 set acl_m0 = $acl_verify_message
3368 Previously, $acl_verify_message was set only while expanding "message"
3369 and "log_message" when a very denied access.
3371 PH/28 Modified OS/os.c-Linux with
3373 -#ifndef OS_LOAD_AVERAGE
3374 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3376 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3377 nomenclature these days.)
3379 PH/29 Installed patch from the Sieve maintainer that adds the options
3380 sieve_useraddress and sieve_subaddress to the redirect router.
3382 PH/30 In these circumstances:
3383 . Two addresses routed to the same list of hosts;
3384 . First host does not offer TLS;
3385 . First host accepts first address;
3386 . First host gives temporary error to second address;
3387 . Second host offers TLS and a TLS session is established;
3388 . Second host accepts second address.
3389 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3390 and peerdn, if requested) that were in fact used only for the second
3393 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3394 attention to any local part prefix or suffix that was matched by the
3395 router that accepted the address. It now behaves in the same way as it
3396 does for delivery: the affixes are removed from the local part unless
3397 rcpt_include_affixes is set on the transport.
3399 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3400 timeout during the DATA phase of an incoming message.
3402 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3403 applied a patch sanctioned by the Sieve maintainer.
3405 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3406 the uid or gid is negative. A case of a negative gid caused this to be
3407 noticed. The fix allows for either to be negative.
3409 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3410 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3413 PH/36 The change PH/12 above was broken. Fixed it.
3415 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3416 the grounds that routing the same address twice would always produce the
3417 same answer. This might have been true once, but it is certainly no
3418 longer true now. Routing a child address may depend on the previous
3419 routing that produced that child. Some complicated redirection strategies
3420 went wrong when messages had multiple recipients, and made Exim's
3421 behaviour dependent on the order in which the addresses were given.
3423 I have moved the duplicate checking until after the routing is complete.
3424 Exim scans the addresses that are assigned to local and remote
3425 transports, and removes any duplicates. This means that more work will be
3426 done, as duplicates will always all be routed, but duplicates are
3427 presumably rare, so I don't expect this is of any significance.
3429 For deliveries to pipes, files, and autoreplies, the duplicate checking
3430 still happens during the routing process, since they are not going to be
3433 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3434 It corrects a timeout issue with spamd. This is Ian's comment: "The
3435 background is that sometimes spamd either never reads data from a
3436 connection it has accepted, or it never writes response data. The exiscan
3437 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3438 blindly assumes that writes won't block so it may never time out."
3440 PH/39 Allow G after quota size as well as K and M.
3442 PH/40 The value set for $authenticated_id in an authenticator may not contain
3443 binary zeroes or newlines because the value is written to log lines and
3444 to spool files. There was no check on this. Now the value is run through
3445 the string_printing() function so that such characters are converted to
3446 printable escape sequences.
3448 PH/41 $message_linecount is a new variable that contains the total number of
3449 lines in the message. Compare $body_linecount, which is the count for the
3452 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3453 bounce and delay warning messages, except in certain special cases, which
3456 (a) An SMTP error message from a remote host;
3457 (b) A message specified in a :fail: redirection;
3458 (c) A message specified in a "fail" command in a system filter;
3459 (d) A message specified in a FAIL return from the queryprogram router;
3460 (e) A message specified by the cannot_route_message router option.
3462 In these cases only, Exim does include the error details in bounce and
3463 warning messages. There are also a few cases where bland messages such
3464 as "unrouteable address" or "local delivery error" are given.
3466 PH/43 $value is now also set for the "else" part of a ${run expansion.
3468 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3469 being worked on, but at least Exim now implements the latest version to
3472 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3473 process to complete was treated as a delivery failure, a timeout while
3474 writing the message to the pipe was logged, but erroneously treated as a
3475 successful delivery. Such timeouts include transport filter timeouts. For
3476 consistency with the overall process timeout, these timeouts are now
3477 treated as errors, giving rise to delivery failures by default. However,
3478 there is now a new Boolean option for the pipe transport called
3479 timeout_defer, which, if set TRUE, converts the failures into defers for
3480 both kinds of timeout. A transport filter timeout is now identified in
3483 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3484 systems where "make" and "gmake" are different, calling "gmake" at top
3485 level broke things. I've arranged for the value of $(MAKE) to be passed
3486 from the Makefile to this script so that it can call the same version of
3490 A note about Exim versions 4.44 and 4.50
3491 ----------------------------------------
3493 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3494 changes of various kinds. As a consequence, a big documentation update was
3495 needed. This delayed the release for rather longer than seemed good, especially
3496 in the light of a couple of (minor) security issues. Therefore, the changes
3497 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3498 release. So 4.44 and 4.50 are in effect two different branches that both start
3501 I have left the 4.50 change log unchanged; it contains all the changes since
3502 4.43. The change log for 4.44 is below; many of its items are identical to
3503 those for 4.50. This seems to be the most sensible way to preserve the
3504 historical information.
3510 1. Minor wording change to the doc/README.SIEVE file.
3512 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3513 computation of the current number of files was incorrect.
3515 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3516 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3519 4. Give more explanation in the error message when the command for a transport
3520 filter fails to execute.
3522 5. There are several places where Exim runs a non-Exim command in a
3523 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3524 was being done only for the command run by the queryprogram router. It is
3525 now done for all such subprocesses. The other cases are: ${run, transport
3526 filters, and the commands run by the lmtp and pipe transports.
3528 6. Added CONFIGURE_GROUP build-time option.
3530 7. Some older OS have a limit of 256 on the maximum number of file
3531 descriptors. Exim was using setrlimit() to set 1000 as a large value
3532 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3533 systems. I've change it so that if it can't get 1000, it tries for 256.
3535 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3536 was an oversight, and furthermore, ever since the addition of extra
3537 controls (e.g. 4.43/32), the checks on when to allow different forms of
3538 "control" were broken. There should now be diagnostics for all cases when a
3539 control that does not make sense is encountered.
3541 9. Added the /retain_sender option to "control=submission".
3543 10. $recipients is now available in the predata ACL (oversight).
3545 11. Tidy the search cache before the fork to do a delivery from a message
3546 received from the command line. Otherwise the child will trigger a lookup
3547 failure and thereby defer the delivery if it tries to use (for example) a
3548 cached ldap connection that the parent has called unbind on.
3550 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3551 of $address_data from the recipient verification was clobbered by the
3552 sender verification.
3554 13. The value of address_data from a sender verification is now available in
3555 $sender_address_data in subsequent conditions in the ACL statement.
3557 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3559 15. Added a new option "connect=<time>" to callout options, to set a different
3562 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3563 was its contents. (It was OK if the option was not defined at all.)
3565 17. A "Completed" log line is now written for messages that are removed from
3566 the spool by the -Mrm option.
3568 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3569 information about exactly what failed.
3571 19. Added -dd to debug only the daemon process.
3573 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3574 handles timeouts, both on the server side and network timeouts. Renamed the
3575 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3577 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3578 It is now set to "smtps".
3580 22. $host_address is now set to the target address during the checking of
3581 ignore_target_hosts.
3583 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3584 being passed; this would have caused $sender_host_name to have been used if
3585 matching the list had actually called for a host name (not very likely,
3586 since this list is usually IP addresses). A host name is now passed as
3589 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3590 code that specifies a non-restarting handler (typically sigaction() in
3591 modern systems) in an attempt to fix a rare and obscure crash bug.
3593 25. Narrowed the window for a race in the daemon that could cause it to ignore
3594 SIGCHLD signals. This is not a major problem, because they are used only to
3595 wake it up if nothing else does.
3597 26. A malformed maildirsize file could cause Exim to calculate negative values
3598 for the mailbox size or file count. Odd effects could occur as a result.
3599 The maildirsize information is now recalculated if the size or filecount
3602 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3603 support for a long time. Removed HAVE_SYS_VFS_H.
3605 28. Installed the latest version of exipick from John Jetmore.
3607 29. In an address list, if the pattern was not a regular expression, an empty
3608 subject address (from a bounce message) matched only if the pattern was an
3609 empty string. Non-empty patterns were not even tested. This was the wrong
3610 because it is perfectly reasonable to use an empty address as part of a
3611 database query. An empty address is now tested by patterns that are
3612 lookups. However, all the other forms of pattern expect the subject to
3613 contain a local part and a domain, and therefore, for them, an empty
3614 address still always fails if the pattern is not itself empty.
3616 30. Exim went into a mad DNS loop when attempting to do a callout where the
3617 host was specified on an smtp transport, and looking it up yielded more
3618 than one IP address.
3620 31. Re-factored the code for checking spool and log partition space into a
3621 function that finds that data and another that does the check. The former
3622 is then used to implement four new variables: $spool_space, $log_space,
3623 $spool_inodes, and $log_inodes.
3625 32. The RFC2047 encoding function was originally intended for short strings
3626 such as real names; it was not keeping to the 75-character limit for
3627 encoded words that the RFC imposes. It now respects the limit, and
3628 generates multiple encoded words if necessary. To be on the safe side, I
3629 have increased the buffer size for the ${rfc2047: expansion operator from
3632 33. It is now permitted to omit both strings after an "if" condition; if the
3633 condition is true, the result is "true". As before, when the second string
3634 is omitted, a false condition yields an empty string. This makes it less
3635 cumbersome to write custom ACL and router conditions.
3637 34. Failure to deliver a bounce message always caused it to be frozen, even if
3638 there was an errors_to setting on the router. The errors_to setting is now
3641 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3642 canonical form (fully expanded) before being placed in
3643 $sender_host_address.
3645 36. The table in the code that translates DNS record types into text (T_A to
3646 "A" for instance) was missing entries for NS and CNAME. It is just possible
3647 that this could have caused confusion if both these types were looked up
3648 for the same domain, because the text type is used as part of Exim's
3649 per-process caching. But the chance of anyone hitting this buglet seems
3652 37. The dnsdb lookup has been extended in a number of ways.
3654 (1) There is a new type, "zns", which walks up the domain tree until it
3655 finds some nameserver records. It should be used with care.
3657 (2) There is a new type, "mxh", which is like "mx" except that it returns
3658 just the host names, not the priorities.
3660 (3) It is now possible to give a list of domains (or IP addresses) to be
3661 looked up. The behaviour when one of the lookups defers can be
3662 controlled by a keyword.
3664 (4) It is now possible to specify the separator character for use when
3665 multiple records are returned.
3667 38. The dnslists ACL condition has been extended: it is now possible to supply
3668 a list of IP addresses and/or domains to be looked up in a particular DNS
3671 39. Added log_selector=+queue_time_overall.
3673 40. When running the queue in the test harness, wait just a tad after forking a
3674 delivery process, to get repeatability of debugging output.
3676 41. Include certificate and key file names in error message when GnuTLS fails
3677 to set them up, because the GnuTLS error message doesn't include the name
3678 of the failing file when there is a problem reading it.
3680 42. Allow both -bf and -bF in the same test run.
3682 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3684 44. The "Exiscan patch" is now merged into the mainline Exim source.
3686 45. Sometimes the final signoff response after QUIT could fail to get
3687 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3688 < 0 before doing a fflush(). This bug looks as though it goes back to the
3689 introduction of TLS in release 3.20, but "sometimes" must have been rare
3690 because the tests only now provoked it.
3692 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3693 (this can affect the format of dates).
3695 47. exim_tidydb, when checking for the continued existence of a message for
3696 which it has found a message-specific retry record, was not finding
3697 messages that were in split spool directories. Consequently, it was
3698 deleting retry records that should have stayed in existence.
3700 48. Steve fixed some bugs in eximstats.
3702 49. The SPA authentication driver was not abandoning authentication and moving
3703 on to the next authenticator when an expansion was forced to fail,
3704 contradicting the general specification for all authenticators. Instead it
3705 was generating a temporary error. It now behaves as specified.
3707 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3708 (the order specifies the preference for clients). The order is now AES256,
3709 AES128, 3DES, ARCFOUR128.
3711 51. Small patch to Sieve code - explicitly set From: when generating an
3714 52. Exim crashed if a remote delivery caused a very long error message to be
3715 recorded - for instance if somebody sent an entire SpamAssassin report back
3716 as a large number of 550 error lines. This bug was coincidentally fixed by
3717 increasing the size of one of Exim's internal buffers (big_buffer) that
3718 happened as part of the Exiscan merge. However, to be on the safe side, I
3719 have made the code more robust (and fixed the comments that describe what
3722 53. Now that there can be additional text after "Completed" in log lines (if
3723 the queue_time_overall log selector is set), a one-byte patch to exigrep
3724 was needed to allow it to recognize "Completed" as not the last thing in
3727 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3728 patch that reportedly fixes this has been added. I am not expert enough to
3729 create a test for it. This is what the patch creator wrote:
3731 "I found a little strange behaviour of ldap code when working with
3732 Windows 2003 AD Domain, where users was placed in more than one
3733 Organization Units. When I tried to give exim partial DN, the exit code
3734 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3735 But simultaneously result of request was absolutely normal ldap result,
3736 so I produce this patch..."
3738 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3739 REFERENCE, so I have modified the code to exclude the patch when that macro
3742 55. Some experimental protocols are using DNS PTR records for new purposes. The
3743 keys for these records are domain names, not reversed IP addresses. The
3744 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3745 leaves it alone. Component reversal etc. now happens only for IP addresses.
3748 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3750 57. Double the size of the debug message buffer (to 2048) so that more of very
3751 long debug lines gets shown.
3753 58. The exicyclog utility now does better if the number of log files to keep
3754 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3756 59. Two changes related to the smtp_active_hostname option:
3758 (1) $smtp_active_hostname is now available as a variable.
3759 (2) The default for smtp_banner uses $smtp_active_hostname instead
3760 of $primary_hostname.
3762 60. The host_aton() function is supposed to be passed a string that is known
3763 to be a valid IP address. However, in the case of IPv6 addresses, it was
3764 not checking this. This is a hostage to fortune. Exim now panics and dies
3765 if the condition is not met. A case was found where this could be provoked
3766 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3767 components; fortuitously, this particular loophole had already been fixed
3768 by change 4.50/55 above.
3770 If there are any other similar loopholes, the new check in host_aton()
3771 itself should stop them being exploited. The report I received stated that
3772 data on the command line could provoke the exploit when Exim was running as
3773 exim, but did not say which command line option was involved. All I could
3774 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3775 running as the user.
3778 61. There was a buffer overflow vulnerability in the SPA authentication code
3779 (which came originally from the Samba project). I have added a test to the
3780 spa_base64_to_bits() function which I hope fixes it.
3783 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3784 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3786 63. The daemon start-up calls getloadavg() while still root for those OS that
3787 need the first call to be done as root, but it missed one case: when
3788 deliver_queue_load_max is set with deliver_drop_privilege. This is
3789 necessary for the benefit of the queue runner, because there is no re-exec
3790 when deliver_drop_privilege is set.
3792 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3793 This has been fixed.
3795 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3796 was in use, was not putting the data itself into the right store pool;
3797 consequently, it could be overwritten for a subsequent message in the same
3798 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3801 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3803 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3804 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3805 the function were treating the return as a boolean value, which happened to
3806 work because 0=false and not-0=true, but is not correct code.
3808 68. The host_aton() function was not handling scoped IPv6 addresses (those
3809 with, for example, "%eth0" on the end) correctly.
3811 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3812 negated items (that is, ~something) in unsigned ints. Some compilers
3813 apparently mutter when there is no cast.
3815 70. If an address verification called from an ACL failed, and did not produce a
3816 user-specific message (i.e. there was only a "system" message), nothing was
3817 put in $acl_verify_message. In this situation, it now puts the system
3820 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3821 session; change 4.31/43 added the unwanted input to the log line - except
3822 that it did not do this in the start of session case. It now does.
3824 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3825 This is wrong and can cause the other end to generate a synchronization
3826 error if it is another Exim or anything else that does the synchronization
3827 check. A QUIT command is no longer sent after a timeout.
3829 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3830 during host lookups.
3832 74. The defer_ok option of callout verification was not working if it was used
3833 when verifying addresses in header lines, that is, for this case:
3835 verify = header_sender/callout=defer_ok
3837 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3838 those file descriptors could be used for SMTP connections. If anything
3839 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3840 could be sent to the SMTP client, causing chaos. The daemon now opens
3841 stdin, stdout, and stderr to /dev/null when it puts itself into the
3844 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3845 log by default. The user can override this with suitable Perl magic.
3847 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3848 the log message when discard triggers, was not working for the DATA ACL or
3849 for the non-SMTP ACL.
3851 78. Error message wording change in sieve.c.
3853 79. If smtp_accept_max_per_host was set, the number of connections could be
3854 restricted to fewer than expected, because the daemon was trying to set up
3855 a new connection before checking whether the processes handling previous
3856 connections had finished. The check for completed processes is now done
3857 earlier. On busy systems, this bug wouldn't be noticed because something
3858 else would have woken the daemon, and it would have reaped the completed
3861 80. If a message was submitted locally by a user whose login name contained one
3862 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3863 It caused a spool format error. I have fixed the spool reading code. A
3864 related problem was that the "from" clause in the Received: line became
3865 illegal because of the space(s). It is now covered by ${quote_local_part.
3867 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3870 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3872 83. Patch from Sieve maintainer to fix unterminated string problem in
3873 "vacation" handling.
3875 84. Some minor changes to the Linux configuration files to help with other
3876 OS variants using glibc.
3878 85. One more patch for Sieve to update vacation handling to latest spec.
3881 ----------------------------------------------------
3882 See the note above about the 4.44 and 4.50 releases.
3883 ----------------------------------------------------
3889 1. Change 4.43/35 introduced a bug that caused file counts to be
3890 incorrectly computed when quota_filecount was set in an appendfile
3893 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3894 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3897 3. Give more explanation in the error message when the command for a transport
3898 filter fails to execute.
3900 4. There are several places where Exim runs a non-Exim command in a
3901 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3902 was being done only for the command run by the queryprogram router. It is
3903 now done for all such subprocesses. The other cases are: ${run, transport
3904 filters, and the commands run by the lmtp and pipe transports.
3906 5. Some older OS have a limit of 256 on the maximum number of file
3907 descriptors. Exim was using setrlimit() to set 1000 as a large value
3908 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3909 systems. I've change it so that if it can't get 1000, it tries for 256.
3911 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3912 was an oversight, and furthermore, ever since the addition of extra
3913 controls (e.g. 4.43/32), the checks on when to allow different forms of
3914 "control" were broken. There should now be diagnostics for all cases when a
3915 control that does not make sense is encountered.
3917 7. $recipients is now available in the predata ACL (oversight).
3919 8. Tidy the search cache before the fork to do a delivery from a message
3920 received from the command line. Otherwise the child will trigger a lookup
3921 failure and thereby defer the delivery if it tries to use (for example) a
3922 cached ldap connection that the parent has called unbind on.
3924 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3925 of $address_data from the recipient verification was clobbered by the
3926 sender verification.
3928 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3929 was its contents. (It was OK if the option was not defined at all.)
3931 11. A "Completed" log line is now written for messages that are removed from
3932 the spool by the -Mrm option.
3934 12. $host_address is now set to the target address during the checking of
3935 ignore_target_hosts.
3937 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3938 being passed; this would have caused $sender_host_name to have been used if
3939 matching the list had actually called for a host name (not very likely,
3940 since this list is usually IP addresses). A host name is now passed as
3943 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3944 code that specifies a non-restarting handler (typically sigaction() in
3945 modern systems) in an attempt to fix a rare and obscure crash bug.
3947 15. Narrowed the window for a race in the daemon that could cause it to ignore
3948 SIGCHLD signals. This is not a major problem, because they are used only to
3949 wake it up if nothing else does.
3951 16. A malformed maildirsize file could cause Exim to calculate negative values
3952 for the mailbox size or file count. Odd effects could occur as a result.
3953 The maildirsize information is now recalculated if the size or filecount
3956 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3957 support for a long time. Removed HAVE_SYS_VFS_H.
3959 18. Updated exipick to current release from John Jetmore.
3961 19. Allow an empty sender to be matched against a lookup in an address list.
3962 Previously the only cases considered were a regular expression, or an
3965 20. Exim went into a mad DNS lookup loop when doing a callout where the
3966 host was specified on the transport, if the DNS lookup yielded more than
3969 21. The RFC2047 encoding function was originally intended for short strings
3970 such as real names; it was not keeping to the 75-character limit for
3971 encoded words that the RFC imposes. It now respects the limit, and
3972 generates multiple encoded words if necessary. To be on the safe side, I
3973 have increased the buffer size for the ${rfc2047: expansion operator from
3976 22. Failure to deliver a bounce message always caused it to be frozen, even if
3977 there was an errors_to setting on the router. The errors_to setting is now
3980 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3981 canonical form (fully expanded) before being placed in
3982 $sender_host_address.
3984 24. Updated eximstats to version 1.33
3986 25. Include certificate and key file names in error message when GnuTLS fails
3987 to set them up, because the GnuTLS error message doesn't include the name
3988 of the failing file when there is a problem reading it.
3990 26. Expand error message when OpenSSL has problems setting up cert/key files.
3993 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3994 (this can affect the format of dates).
3996 28. exim_tidydb, when checking for the continued existence of a message for
3997 which it has found a message-specific retry record, was not finding
3998 messages that were in split spool directories. Consequently, it was
3999 deleting retry records that should have stayed in existence.
4001 29. eximstats updated to version 1.35
4002 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4003 1.35 - bugfix such that pie charts by volume are generated correctly
4005 30. The SPA authentication driver was not abandoning authentication and moving
4006 on to the next authenticator when an expansion was forced to fail,
4007 contradicting the general specification for all authenticators. Instead it
4008 was generating a temporary error. It now behaves as specified.
4010 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4011 (the order specifies the preference for clients). The order is now AES256,
4012 AES128, 3DES, ARCFOUR128.
4014 31. Small patch to Sieve code - explicitly set From: when generating an
4017 32. Exim crashed if a remote delivery caused a very long error message to be
4018 recorded - for instance if somebody sent an entire SpamAssassin report back
4019 as a large number of 550 error lines. This bug was coincidentally fixed by
4020 increasing the size of one of Exim's internal buffers (big_buffer) that
4021 happened as part of the Exiscan merge. However, to be on the safe side, I
4022 have made the code more robust (and fixed the comments that describe what
4025 33. Some experimental protocols are using DNS PTR records for new purposes. The
4026 keys for these records are domain names, not reversed IP addresses. The
4027 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4028 leaves it alone. Component reversal etc. now happens only for IP addresses.
4031 34. The host_aton() function is supposed to be passed a string that is known
4032 to be a valid IP address. However, in the case of IPv6 addresses, it was
4033 not checking this. This is a hostage to fortune. Exim now panics and dies
4034 if the condition is not met. A case was found where this could be provoked
4035 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4036 components; fortuitously, this particular loophole had already been fixed
4037 by change 4.50/55 or 4.44/33 above.
4039 If there are any other similar loopholes, the new check in host_aton()
4040 itself should stop them being exploited. The report I received stated that
4041 data on the command line could provoke the exploit when Exim was running as
4042 exim, but did not say which command line option was involved. All I could
4043 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4044 running as the user.
4047 35. There was a buffer overflow vulnerability in the SPA authentication code
4048 (which came originally from the Samba project). I have added a test to the
4049 spa_base64_to_bits() function which I hope fixes it.
4052 36. The daemon start-up calls getloadavg() while still root for those OS that
4053 need the first call to be done as root, but it missed one case: when
4054 deliver_queue_load_max is set with deliver_drop_privilege. This is
4055 necessary for the benefit of the queue runner, because there is no re-exec
4056 when deliver_drop_privilege is set.
4058 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4059 was in use, was not putting the data itself into the right store pool;
4060 consequently, it could be overwritten for a subsequent message in the same
4061 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4064 38. Sometimes the final signoff response after QUIT could fail to get
4065 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4066 < 0 before doing a fflush(). This bug looks as though it goes back to the
4067 introduction of TLS in release 3.20, but "sometimes" must have been rare
4068 because the tests only now provoked it.
4074 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4075 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4076 Now it returns an integer. A number of calls were still expecting a T/F
4077 return. Fortuitously, in all cases, the tests worked in OK situations,
4078 which is the norm. However, things would have gone wrong on any write
4079 failures on the smtp file descriptor. This function is used when sending
4080 messages over SMTP and also when doing verify callouts.
4082 2. When Exim is called to do synchronous delivery of a locally submitted
4083 message (the -odf or -odi options), it no longer closes stderr before doing
4086 3. Implemented the mua_wrapper option.
4088 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4090 5. Implemented the functions header_remove(), header_testname(),
4091 header_add_at_position(), and receive_remove_recipient(), and exported them
4094 6. If an ACL "warn" statement specified the addition of headers, Exim already
4095 inserted X-ACL-Warn: at the start if there was no header name. However, it
4096 was not making this test for the second and subsequent header lines if
4097 there were newlines in the string. This meant that an invalid header could
4098 be inserted if Exim was badly configured.
4100 7. Allow an ACL "warn" statement to add header lines at the start or after all
4101 the Received: headers, as well as at the end.
4103 8. Added the rcpt_4xx retry error code.
4105 9. Added postmaster_mailfrom=xxx to callout verification option.
4107 10. Added mailfrom=xxxx to the callout verification option, for verify=
4110 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4111 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4113 12. Inserted some casts to stop certain compilers warning when using pointer
4114 differences as field lengths or precisions in printf-type calls (mostly
4115 affecting debugging statements).
4117 13. Added optional readline() support for -be (dynamically loaded).
4119 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4120 same clock tick as a message's arrival, so that its received time was the
4121 same as the "first fail" time on the retry record, and that message
4122 remained on the queue past the ultimate address timeout, every queue runner
4123 would try a delivery (because it was past the ultimate address timeout) but
4124 after another failure, the ultimate address timeout, which should have then
4125 bounced the address, did not kick in. This was a "< instead of <=" error;
4126 in most cases the first failure would have been in the next clock tick
4127 after the received time, and all would be well.
4129 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4130 being recognized when the domain list was tested by the match_domain
4131 condition in an expansion string.
4133 16. Added the ${str2b64: operator.
4135 17. Exim was always calling setrlimit() to set a large limit for the number of
4136 processes, without checking whether the existing limit was already
4137 adequate. (It did check for the limit on file descriptors.) Furthermore,
4138 errors from getrlimit() and setrlimit() were being ignored. Now they are
4139 logged to the main and panic logs, but Exim does carry on, to try to do its
4140 job under whatever limits there are.
4142 18. Imported PCRE 5.0.
4144 19. Trivial typo in log message " temporarily refused connection" (the leading
4147 20. If the log selector return_path_on_delivery was set and an address was
4148 redirected to /dev/null, the delivery process crashed because it assumed
4149 that a return path would always be set for a "successful" delivery. In this
4150 case, the whole delivery is bypassed as an optimization, and therefore no
4153 21. Internal re-arrangement: the function for sending a challenge and reading
4154 a response while authentication was assuming a zero-terminated challenge
4155 string. It's now changed to take a pointer and a length, to allow for
4156 binary data in such strings.
4158 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4160 23. Exim was not respecting finduser_retries when seeking the login of the
4161 uid under which it was called; it was always trying 10 times. (The default
4162 setting of finduser_retries is zero.) Also, it was sleeping after the final
4163 failure, which is pointless.
4165 24. Implemented tls_on_connect_ports.
4167 25. Implemented acl_smtp_predata.
4169 26. If the domain in control=submission is set empty, Exim assumes that the
4170 authenticated id is a complete email address when it generates From: or
4171 Sender: header lines.
4173 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4174 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4175 chown and chgrp in /bin and hostname in /usr/bin.
4177 28. Exim was keeping the "process log" file open after each use, just as it
4178 does for the main log. This opens the possibility of it remaining open for
4179 long periods when the USR1 signal hits a daemon. Occasional processlog
4180 errors were reported, that could have been caused by this. Anyway, it seems
4181 much more sensible not to leave this file open at all, so that is what now
4184 29. The long-running daemon process does not normally write to the log once it
4185 has entered its main loop, and it closes the log before doing so. This is
4186 so that log files can straightforwardly be renamed and moved. However,
4187 there are a couple of unusual error situations where the daemon does write
4188 log entries, and I had neglected to close the log afterwards.
4190 30. The text of an SMTP error response that was received during a remote
4191 delivery was being truncated at 512 bytes. This is too short for some of
4192 the long messages that one sometimes sees. I've increased the limit to
4195 31. It is now possible to make retry rules that apply only when a message has a
4196 specific sender, in particular, an empty sender.
4198 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4199 it possible to be selective about when SMTP synchronization is enforced.
4201 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4203 32. Implemented hosts_connection_nolog.
4205 33. Added an ACL for QUIT.
4207 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4210 35. Added mailbox_size and mailbox_filecount to appendfile.
4212 36. Added control = no_multiline_responses to ACLs.
4214 37. There was a bug in the logic of the code that waits for the clock to tick
4215 in the case where the clock went backwards by a substantial amount such
4216 that the microsecond fraction of "now" was more than the microsecond
4217 fraction of "then" (but the whole seconds number was less).
4219 38. Added support for the libradius Radius client library this is found on
4220 FreeBSD (previously only the radiusclient library was supported).
4226 1. When certain lookups returned multiple values in the form name=value, the
4227 quoting of the values was not always being done properly. Specifically:
4228 (a) If the value started with a double quote, but contained no whitespace,
4230 (b) If the value contained whitespace other than a space character (i.e.
4231 tabs or newlines or carriage returns) it was not quoted.
4232 This fix has been applied to the mysql and pgsql lookups by writing a
4233 separate quoting function and calling it from the lookup code. The fix
4234 should probably also be applied to nisplus, ibase and oracle lookups, but
4235 since I cannot test any of those, I have not disturbed their existing code.
4237 2. A hit in the callout cache for a specific address caused a log line with no
4238 reason for rejecting RCPT. Now it says "Previous (cached) callout
4239 verification failure".
4241 3. There was an off-by-one bug in the queryprogram router. An over-long
4242 return line was truncated at 256 instead of 255 characters, thereby
4243 overflowing its buffer with the terminating zero. As well as fixing this, I
4244 have increased the buffer size to 1024 (and made a note to document this).
4246 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4247 when Exim is waiting for an SMTP response from a remote server, Exim
4248 restarts its select() call on the socket, thereby resetting its timeout.
4249 This is not a problem when such interrupts are rare. Somebody set up a cron
4250 job to run exiwhat every 2 minutes, which is less than the normal select()
4251 timeout (5 or 10 minutes). This meant that the select() timeout never
4252 kicked in because it was always reset. I have fixed this by comparing the
4253 time when an interrupt arrives with the time at the start of the first call
4254 to select(). If more time than the timeout has elapsed, the interrupt is
4255 treated as a timeout.
4257 5. Some internal re-factoring in preparation for the addition of Sieve
4258 extensions (by MH). In particular, the "personal" test is moved to a
4259 separate function, and given an option for scanning Cc: and Bcc: (which is
4260 not set for Exim filters).
4262 6. When Exim created an email address using the login of the caller as the
4263 local part (e.g. when creating a From: or Sender: header line), it was not
4264 quoting the local part when it contained special characters such as @.
4266 7. Installed new OpenBSD configuration files.
4268 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4269 try to make them clearer.
4271 9. Callout options, other than the timeout value, were being ignored when
4272 verifying sender addresses in header lines. For example, when using
4274 verify = header_sender/callout=no_cache
4276 the cache was (incorrectly) being used.
4278 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4279 only the Cygwin environment.
4281 11. When return_path_on_delivery was set as a log selector, if different remote
4282 addresses in the same message used different return paths and parallel
4283 remote delivery occurred, the wrong values would sometimes be logged.
4284 (Whenever a remote delivery process finished, the return path value from
4285 the most recently started remote delivery process was logged.)
4287 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4288 lines when AUTH and/or TLS are in use. This is the "received protocol"
4289 field. Exim used to use "asmtp" for authenticated SMTP, without any
4290 indication (in the protocol name) for TLS use. Now it follows the RFC and
4291 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4292 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4293 names appear in log lines as well as in Received: header lines.
4295 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4296 extensions, and comparison tests, and to fix some bugs.
4298 14. Changes to the "personal" filter test:
4300 (1) The test was buggy in that it was just doing the equivalent of
4301 "contains" tests on header lines. For example, if a user's address was
4302 anne@some.where, the "personal" test would incorrectly be true for
4304 To: susanne@some.where
4306 This test is now done by extracting each address from the header in turn,
4307 and checking the entire address. Other tests that are part of "personal"
4308 are now done using regular expressions (for example, to check local parts
4309 of addresses in From: header lines).
4311 (2) The list of non-personal local parts in From: addresses has been
4312 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4313 taken from the Sieve specification recommendations.
4315 (3) If the message contains any header line starting with "List-" it is
4316 treated as non-personal.
4318 (4) The test for "circular" in the Subject: header line has been removed
4319 because it now seems ill-conceived.
4321 15. Minor typos in src/EDITME comments corrected.
4323 16. Installed latest exipick from John Jetmore.
4325 17. If headers_add on a router specified a text string that was too long for
4326 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4327 of string_sprintf() is now avoided.
4329 18. $message_body_size was not set (it was always zero) when running the DATA
4330 ACL and the local_scan() function.
4332 19. For the "mail" command in an Exim filter, no default was being set for
4333 the once_repeat time, causing a random time value to be used if "once" was
4334 specified. (If the value happened to be <= 0, no repeat happened.) The
4335 default is now 0s, meaning "never repeat". The "vacation" command was OK
4336 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4337 (I found it when inspecting the code).
4339 20. There is now an overall timeout for performing a callout verification. It
4340 defaults to 4 times the callout timeout, which applies to individual SMTP
4341 commands during the callout. The overall timeout applies when there is more
4342 than one host that can be tried. The timeout is checked before trying the
4343 next host. This prevents very long delays if there are a large number of
4344 hosts and all are timing out (e.g. when the network connections are timing
4345 out). The value of the overall timeout can be changed by specifying an
4346 additional sub-option for "callout", called "maxwait". For example:
4348 verify = sender/callout=5s,maxwait=20s
4350 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4351 to the end before writing, but this should make it even safer).
4353 22. Exim was forgetting that it had advertised PIPELINING for the second and
4354 subsequent messages on an SMTP connection. It was also not resetting its
4355 memory on STARTTLS and an internal HELO.
4357 23. When Exim logs an SMTP synchronization error within a session, it now
4358 records whether PIPELINING has been advertised or not.
4360 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4361 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4362 rather than long int.
4364 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4366 26. Added the never_mail option to autoreply.
4372 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4373 crash if the getsockname() call failed; this can happen if a connection is
4374 closed very soon after it is established. The problem was simply in the
4375 order in which certain operations were done, causing Exim to try to write
4376 to the SMTP stream before it had set up the file descriptor. The bug has
4377 been fixed by making things happen in the correct order.
4383 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4384 before the connection was closed, thus losing the rejection response.
4386 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4387 some early Solaris releases, but causes trouble in current releases where
4388 socklen_t is defined.
4390 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4393 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4396 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4397 that releasing the top of it at the end releases what was used for sub-
4398 expansions (unless the block got too big). However, discard this block if
4399 the first thing is a variable or header, so that we can use its block when
4400 it is dynamic (useful for very large $message_headers, for example).
4402 6. Lookups now cache *every* query, not just the most recent. A new, separate
4403 store pool is used for this. It can be recovered when all lookup caches are
4404 flushed. Lookups now release memory at the end of their result strings.
4405 This has involved some general refactoring of the lookup sources.
4407 7. Some code has been added to the store_xxx() functions to reduce the amount
4408 of flapping under certain conditions.
4410 8. log_incoming_interface used to affect only the <= reception log lines. Now
4411 it causes the local interface and port to be added to several more SMTP log
4412 lines, for example "SMTP connection from", and rejection lines.
4414 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4416 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4418 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4419 could be overwritten at the end of the current message (or the start of a
4420 new message if it was set in a HELO ACL). The value is now preserved for
4421 the duration of the SMTP connection.
4423 12. If a transport had a headers_rewrite setting, and a matching header line
4424 contained an unqualified address, that address was qualified, even if it
4425 did not match any rewriting rules. The underlying bug was that the values
4426 of the flags that permit the existence of unqualified sender and recipient
4427 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4428 non-local messages, and by -bnq for local messages) were not being
4429 preserved with the message after it was received.
4431 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4432 "next input=" as part of the text comprising the host identity instead of
4433 the correct text. The code was using the same buffer for two different
4434 strings. However, depending on which order the printing function evaluated
4435 its arguments, the bug did not always show up. Under Linux, for example, my
4436 test suite worked just fine.
4438 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4439 doesn't work with some older versions of Perl. It has been changed to "my",
4440 which in any case is probably the better facility to use.
4442 15. A really picky compiler found some instances of statements for creating
4443 error messages that either had too many or two few arguments for the format
4446 16. The size of the buffer for calls to the DNS resolver has been increased
4447 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4448 for addresses that have a lot of PTR records. This alleviates a problem; it
4449 does not fully solve it.
4451 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4452 buffer now truncates the list and logs the incident, which is the same
4453 action as happens when Exim is looking up a host name and its aliases.
4454 Previously in this situation something unpredictable would happen;
4455 sometimes it was "internal error: store_reset failed".
4457 18. If a server dropped the connection unexpectedly when an Exim client was
4458 using GnuTLS and trying to read a response, the client delivery process
4459 crashed while trying to generate an error log message.
4461 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4462 string, for example:
4464 warn message = H1: something\nH2: something
4466 the text was added as a single header line from Exim's point of view
4467 though it ended up OK in the delivered message. However, searching for the
4468 second and subsequent header lines using $h_h2: did not work. This has been
4469 fixed. Similarly, if a system filter added multiple headers in this way,
4470 the routers could not see them.
4472 20. Expanded the error message when iplsearch is called with an invalid key to
4473 suggest using net-iplsearch in a host list.
4475 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4476 ACLs are no longer actually imposed (and a message to that effect is
4479 22. If a "gecos" field in a passwd entry contained escaped characters, in
4480 particular, if it contained a \" sequence, Exim got it wrong when building
4481 a From: or a Sender: header from that name. A second bug also caused
4482 incorrect handling when an unquoted " was present following a character
4483 that needed quoting.
4485 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4486 was not being matched caselessly.
4488 24. Arranged for all hyphens in the exim.8 source to be escaped with
4491 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4492 properly. Recipient callout cache records were still being keyed to include
4493 the sender, even when use_sender was set false. This led to far more
4494 callouts that were necessary. The sender is no longer included in the key
4495 when use_sender is false.
4497 26. Added "control = submission" modifier to ACLs.
4499 27. Added the ${base62d: operator to decode base 62 numbers.
4501 28. dnsdb lookups can now access SRV records.
4503 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4504 the configuration file.
4506 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4507 (-v) mode. This makes the output for a verbose queue run more intelligible.
4509 31. Added a use_postmaster feature to recipient callouts.
4511 32. Added the $body_zerocount variable, containing the number of binary zero
4512 bytes in the message body.
4514 33. The time of last modification of the "new" subdirectory is now used as the
4515 "mailbox time last read" when there is a quota error for a maildir
4518 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4520 35. Added +ignore_unknown as a special item in host lists.
4522 36. Code for decoding IPv6 addresses in host lists is now included, even if
4523 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4524 address was recognized as an IP address, but was then not correctly decoded
4525 into binary, causing unexpected and incorrect effects when compared with
4532 1. Very minor rewording of debugging text in manualroute to say "list of
4533 hosts" instead of "hostlist".
4535 2. If verify=header_syntax was set, and a header line with an unqualified
4536 address (no domain) and a large number of spaces between the end of the
4537 name and the colon was received, the reception process suffered a buffer
4538 overflow, and (when I tested it) crashed. This was caused by some obsolete
4539 code that should have been removed. The fix is to remove it!
4541 3. When running in the test harness, delay a bit after writing a bounce
4542 message to get a bit more predictability in the log output.
4544 4. Added a call to search_tidyup() just before forking a reception process. In
4545 theory, someone could use a lookup in the expansion of smtp_accept_max_
4546 per_host which, without the tidyup, could leave open a database connection.
4548 5. Added the variables $recipient_data and $sender_data which get set from a
4549 lookup success in an ACL "recipients" or "senders" condition, or a router
4550 "senders" option, similar to $domain_data and $local_part_data.
4552 6. Moved the writing of debug_print from before to after the "senders" test
4555 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4556 problems for message scanning, either using a data ACL, or using
4557 local_scan() because the Received: header was not generated till after they
4558 were called (in order to set the time as the time of reception completion).
4559 I have revised the way this works. The header is now generated after the
4560 body is received, but before the ACL or local_scan() are called. After they
4561 are run, the timestamp in the header is updated.
4567 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4568 before starting a queue runner without re-exec. This happened only when
4569 deliver_drop_privilege was set or when the Exim user was set to root. The
4570 effect of the bug was that timeouts during subsequent deliveries caused
4571 crashes instead of being properly handled. The handler is now left at its
4572 default (and expected) setting.
4574 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4575 message, again when deliver_drop_privilege is set or Exim is run as root.
4576 The bug described in (1) was not present in this case, but the tidying up
4577 of the other signals was missing. I have made the two cases consistent.
4579 3. The ignore_target_hosts setting on a manualroute router was being ignored
4580 for hosts that were looked up using the /MX notation.
4582 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4585 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4586 operated on the sender address. After changing the $sender_address to <>
4587 for the sender address verify, Exim was re-instated it as the original
4588 (before rewriting) address, but remembering that it had rewritten it, so it
4589 wasn't rewriting it again. This bug also had the effect of breaking the
4590 sender address verification caching when the sender address was rewritten.
4592 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4593 This has been changed so that if the ip literal address matches
4594 ignore_target_hosts, the router declines.
4596 7. Added expansion conditions match_domain, match_address, and match_local_
4597 part (NOT match_host).
4599 8. The placeholder for the Received: header didn't have a length field set.
4601 9. Added code to Exim itself and to exim_lock to test for a specific race
4602 condition that could lead to file corruption when using MBX delivery. The
4603 issue is with the lockfile that is created in /tmp. If this file is removed
4604 after a process has opened it but before that process has acquired a lock,
4605 there is the potential for a second process to recreate the file and also
4606 acquire a lock. This could lead to two Exim processes writing to the file
4607 at the same time. The added code performs the same test as UW imapd; it
4608 checks after acquiring the lock that its file descriptor still refers to
4609 the same named file.
4611 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4612 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4615 11. Added the smtp_active_hostname option. If used, this will typically be made
4616 to depend on the incoming interface address. Because $interface_address is
4617 not set up until the daemon has forked a reception process, error responses
4618 that can happen earlier (such as "too many connections") no longer contain
4621 12. If an expansion in a condition on a "warn" statement fails because a lookup
4622 defers, the "warn" statement is abandoned, and the next ACL statement is
4623 processed. Previously this caused the whole ACL to be aborted.
4625 13. Added the iplsearch lookup type.
4627 14. Added ident_timeout as a log selector.
4629 15. Added tls_certificate_verified as a log selector.
4631 16. Added a global option tls_require_ciphers (compare the smtp transport
4632 option of the same name). This controls incoming TLS connections.
4634 17. I finally figured out how to make tls_require_ciphers do a similar thing
4635 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4636 before starting the TLS session.
4638 18. Tabs are now shown as \t in -bP output.
4640 19. If the log selector return_path_on_delivery was set, Exim crashed when
4641 bouncing a message because it had too many Received: header lines.
4643 20. If two routers both had headers_remove settings, and the first one included
4644 a superfluous trailing colon, the final name in the first list and the
4645 first name in the second list were incorrectly joined into one item (with a
4646 colon in the middle).
4652 1. Added -C and -D options to the exinext utility, mainly to make it easier
4653 to include in the automated testing, but these could be helpful when
4654 multiple configurations are in use.
4656 2. The exinext utility was not formatting the output nicely when there was
4657 an alternate port involved in the retry record key, nor when there was a
4658 message id as well (for retries that were specific to a specific message
4659 and a specific host). It was also confused by IPv6 addresses, because of
4660 the additional colons they contain. I have fixed the IPv4 problem, and
4661 patched it up to do a reasonable job for IPv6.
4663 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4664 delivery, the log line now contains "pipelined" if PIPELINING was used.
4666 4. An SMTP transport process used to panic and die if the bind() call to set
4667 an explicit outgoing interface failed. This has been changed; it is now
4668 treated in the same way as a connect() failure.
4670 5. A reference to $sender_host_name in the part of a conditional expansion
4671 that was being skipped was still causing a DNS lookup. This no longer
4674 6. The def: expansion condition was not recognizing references to header lines
4675 that used bh_ and bheader_.
4677 7. Added the _cache feature to named lists.
4679 8. The code for checking quota_filecount in the appendfile transport was
4680 allowing one more file than it should have been.
4682 9. For compatibility with Sendmail, the command line option
4690 and sets the incoming protocol and host name (for trusted callers). The
4691 host name and its colon can be omitted when only the protocol is to be set.
4692 Note the Exim already has two private options, -pd and -ps, that refer to
4693 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4694 "s", but I don't think that's a major issue.
4696 10. A number of refactoring changes to the code, none of which should affect
4699 (a) The number of logging options was getting close to filling up the
4700 32-bit word that was used as a bit map. I have split them into two classes:
4701 those that are passed in the argument to log_write(), and those that are
4702 only ever tested independently outside of that function. These are now in
4703 separate 32-bit words, so there is plenty of room for expansion again.
4704 There is no change in the user interface or the logging behaviour.
4706 (b) When building, for example, log lines, the code previously used a
4707 macro that called string_cat() twice, in order to add two strings. This is
4708 not really sufficiently general. Furthermore, there was one instance where
4709 it was actually wrong because one of the argument was used twice, and in
4710 one call a function was used. (As it happened, calling the function twice
4711 did not affect the overall behaviour.) The macro has been replaced by a
4712 function that can join an arbitrary number of extra strings onto a growing
4715 (c) The code for expansion conditions now uses a table and a binary chop
4716 instead of a serial search (which was left over from when there were very
4717 few conditions). Also, it now recognizes conditions like "pam" even when
4718 the relevant support is not compiled in: a suitably worded error message is
4719 given if an attempt is made to use such a condition.
4721 11. Added ${time_interval:xxxxx}.
4723 12. A bug was causing one of the ddress fields not to be passed back correctly
4724 from remote delivery subprocesses. The field in question was not being
4725 subsequently used, so this caused to problems in practice.
4727 13. Added new log selectors queue_time and deliver_time.
4729 14. Might have fixed a bug in maildirsizefile handling that threw up
4730 "unexpected character" debug warnings, and recalculated the data
4731 unnecessarily. In any case, I expanded the warning message to give more
4734 15. Added the message "Restricted characters in address" to the statements in
4735 the default ACL that block characters like @ and % in local parts.
4737 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4738 Three changes have been made:
4740 (a) There was a serious bug; a negative response to MAIL caused the whole
4741 recipient domain to be cached as invalid, thereby blocking all messages
4742 to all local parts at the same domain, from all senders. This bug has
4743 been fixed. The domain is no longer cached after a negative response to
4744 MAIL if the sender used is not empty.
4746 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4749 (c) A new callout option, "use_sender" has been added for people who want
4750 the modified behaviour.
4756 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4759 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4760 indeed breaks things for older releases.
4762 3. Added additional logging to the case where there is a problem reading data
4763 from a filter that is running in a subprocess using a pipe, in order to
4764 try to track down a specific problem.
4766 4. Testing facility fudge: when running in the test harness and attempting
4767 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4768 getting "No route to host". Convert this to a timeout.
4770 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4773 6. Some OS don't have socklen_t but use size_t instead. This affects the
4774 fifth argument of getsockopt() amongst other things. This is now
4775 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4776 can be set for individual OS. I have set it for SunOS5, OSF1, and
4777 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4778 some earlier ones do not.
4780 7. Change 4.30/15 was not doing the test caselessly.
4782 8. The standard form for an IPv6 address literal was being rejected by address
4783 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4784 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4785 this, as well as the form without the "IPv6" on the front (but only when
4786 address literals are enabled, of course).
4788 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4790 10. Exim crashed if a message with an empty sender address specified by -f
4791 encountered a router with an errors_to setting. This could be provoked only
4792 by a command such as
4796 where an empty string was supplied; "<>" did not hit this bug.
4798 11. Installed PCRE release 4.5.
4800 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4801 remained set. It is now erased.
4803 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4804 times from message ids (which are base 36 rather than the normal 62).
4806 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4807 were being counted as actual protocol errors, and logged if the log
4808 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4809 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4810 and DATA following a set of rejected RCPTs do not count as protocol errors.
4811 In other words, Exim assumes they were pipelined, though this may not
4812 actually be the case. Of course, in all cases the client gets an
4813 appropriate error code.
4815 15. If a lookup fails in an ACL condition, a message about the failure may
4816 be available; it is used if testing the ACL cannot continue, because most
4817 such messages specify what the cause of the deferral is. However, some
4818 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4819 that caused an old message to be retained and used if a later statement
4820 caused a defer, replacing the real cause of the deferral.
4822 16. If an IP address had so many PTR records that the DNS lookup buffer
4823 was not large enough to hold them, Exim could crash while trying to process
4824 the truncated data. It now detects and logs this case.
4826 17. Further to 4.21/58, another change has been made: if (and only if) the
4827 first line of a message (the first header line) ends with CRLF, a bare LF
4828 in a subsequent header line has a space inserted after it, so as not to
4829 terminate the header.
4831 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4832 unnecessarily, used atoi() instead of strtol(), and didn't check the
4833 termination when getting file sizes from file names by regex.
4835 19. Completely re-implemented the support for maildirsize files, in the light
4836 of a number of problems with the previous contributed implementation
4837 (4.30/29). In particular:
4839 . If the quota is zero, the maildirsize file is maintained, but no quota is
4842 . If the maildir directory does not exist, it is created before any attempt
4843 to write a maildirsize file.
4845 . The quota value in the file is just a cache; if the quota is changed in
4846 the transport, the new value overrides.
4848 . A regular expression is available for excluding directories from the
4851 20. The autoreply transport checks the characters in options that define the
4852 message's headers; it allows continued headers, but it was checking with
4853 isspace() after an embedded newline instead of explicitly looking for a
4856 21. If all the "regular" hosts to which an address was routed had passed their
4857 expiry times, and had not reached their retry times, the address was
4858 bounced, even if fallback hosts were defined. Now Exim should go on to try
4861 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4862 equivalent code in the SMTP transport. Some hosts send humungous responses
4863 to HELO/EHLO, more than 1024 it seems.
4865 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4866 gives compiler warnings in some environments. I've now done it "properly",
4869 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4870 (because of problems with the built-in one) was declared to return uschar *
4871 instead of char *, causing compiler failure.
4873 25. Fixed a file descriptor leak when processing alias/forward files.
4875 26. Fixed a minor format string issue in dbfn.c.
4877 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4879 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4880 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4883 29. When a sender address is verified, it is cached, to save repeating the test
4884 when there is more than one recipient in a message. However, when the
4885 verification involves a callout, it is possible for different callout
4886 options to be set for different recipients. It is too complicated to keep
4887 track of this in the cache, so now Exim always runs a verification when a
4888 callout is required, relying on the callout cache for the optimization.
4889 The overhead is duplication of the address routing, but this should not be
4892 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4893 to be verified with callout=postmaster, and the main callout worked but the
4894 postmaster check failed, the verification correctly failed. However, if a
4895 subsequent RCPT command asked for sender verification *without* the
4896 postmaster check, incorrect caching caused this verification also to fail,
4899 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4900 it was not caching the DNS options (qualify_single, search_parents) that
4901 were used when the lookup failed. A subsequent lookup with different
4902 options therefore always gave the same answer, though there were cases
4903 where it should not have. (Example: a "domains = !$mx_any" option on a
4904 dnslookup router: the "domains" option is always processed without any
4905 widening, but the router might have qualify_single set.) Now Exim uses the
4906 cached value only when the same options are set.
4908 32. Added John Jetmore's "exipick" utility to the distribution.
4910 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4911 than a timeout (e.g. a certificate is required, and is not provided), an
4912 Exim server now closes the connection immediately. Previously it waited for
4913 the client to close - but if the client is SSL, it seems that they each
4914 wait for each other, leading to a delay before one of them times out.
4916 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4917 maintained 0.8.x compatibility because I don't think many are using it, and
4918 it is clearly obsolete.
4920 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4923 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4924 client certificate was expired. A simple patch fixes this, though I don't
4925 understand the full logic of why the verify callback is called multiple
4928 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4929 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4930 which causes problems with some clients (such as the Certicom SSL Plus
4931 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4932 disables the coutermeasure allowing Eudora to connect."
4934 38. Exim was not checking that a write() to a log file succeeded. This could
4935 lead to Bad Things if a log got too big, in particular if it hit a file
4936 size limit. Exim now panics and dies if it cannot write to a log file, just
4937 as it does if it cannot open a log file.
4939 39. Modified OS/Makefile-Linux so that it now contains
4941 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4943 The two -D definitions ensure that Exim is compiled with large file
4944 support, which makes it possible to handle log files that are bigger than
4947 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4948 instance) a domain was checked against a named list that involved a lookup,
4949 causing $domain_data to be set, then another domain was checked against the
4950 same list, then the first domain was re-checked, the value of $domain_data
4951 after the final check could be wrong. In particular, if the second check
4952 failed, it could be set empty. This bug probably also applied to
4955 41. The strip_trailing_dot option was not being applied to the address given
4956 with the -f command-line option.
4958 42. The code for reading a message's header from the spool was incrementing
4959 $received_count, but never initializing it. This meant that the value was
4960 incorrect (doubled) while delivering a message in the same process in which
4961 it was received. In the most common configuration of Exim, this never
4962 happens - a fresh exec is done - but it can happen when
4963 deliver_drop_privilege is set.
4965 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4966 it now includes up to 150 characters of the unexpected data in the log
4969 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4970 and building data strings. The size of both of these buffers was 10 000
4971 bytes - far larger than anybody would *ever* want, thought I. Needless to
4972 say, somebody hit the limit. I have increased the maximum line length to
4973 20 000 and the maximum data length of concatenated lines to 100 000. I have
4974 also fixed two bugs, because there was no checking on these buffers. Tsk,
4975 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4976 buffer is too small.
4978 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4979 lsearch lookups. Now it does.
4981 46. When parsing a route_list item in a manualroute router, a fixed-length
4982 buffer was used for the list of hosts. I made this 1024 bytes long,
4983 thinking that nobody would ever have a list of hosts that long. Wrong.
4984 Somebody had a whole pile of complicated expansion conditions, and the
4985 string was silently truncated, leading to an expansion error. It turns out
4986 that it is easier to change to an unlimited length (owing to other changes
4987 that have happened since this code was originally written) than to build
4988 structure for giving a limitation error. The length of the item that
4989 expands into the list of hosts is now unlimited.
4991 47. The lsearch lookup could not handle data where the length of text line was
4992 more than 4095 characters. Such lines were truncated, leading to shortened
4993 data being returned. It should now handle lines of any length.
4995 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4996 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4999 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5000 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5001 no longer happens in comment lines. A list of replacements is now placed
5002 at the head of all of the source files, except those whose only change is
5003 to replace PERL_COMMAND in the very first #! line.
5005 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5006 messages on the queue, with a bottom-up merge sort, using code contributed
5007 by Michael Haardt. This should make operations like -bp somewhat faster on
5008 large queues. It won't affect queue runners, except when queue_run_in_order
5011 51. Installed eximstats 1.31 in the distribution.
5013 52. Added support for SRV lookups to the dnslookup router.
5015 53. If an ACL referred to $message_body or $message_body_end, the value was not
5016 reset for any messages that followed in the same SMTP session.
5018 54. The store-handling optimization for building very long strings was not
5019 differentiating between the different store pools. I don't think this
5020 actually made any difference in practice, but I've tidied it.
5022 55. While running the routers to verify a sender address, $sender_address
5023 was still set to the sender address. This is wrong, because when routing to
5024 send a bounce to the sender, it would be empty. Therefore, I have changed
5025 it so that, while verifying a sender address, $sender_address is set to <>.
5026 (There is no change to what happens when verifying a recipient address.)
5028 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5029 target A or AAAA records (if not already returned) without resetting the
5030 qualify_single or search_parents options of the DNS resolver. These are
5031 inappropriate in this case because the targets of MX and SRV records must
5032 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5033 target that, when qualified, matched something in the local domain. These
5034 two options are now turned off when doing these lookups.
5036 57. It seems that at least some releases of Reiserfs (which does not have the
5037 concept of a fixed number of inodes) returns zero and not -1 for the
5038 number of available inodes. This interacted badly with check_spool_inodes,
5039 which assumed that -1 was the "no such thing" setting. What I have done is
5040 to check that the total number of inodes is greater than zero before doing
5041 the test of how many are available.
5043 58. When a "warn" ACL statement has a log_message modifier, the message is
5044 remembered, and not repeated. This is to avoid a lot of repetition when a
5045 message has many recipients that cause the same warning to be written.
5046 Howewer, Exim was preserving the list of already written lines for an
5047 entire SMTP session, which doesn't seem right. The memory is now reset if a
5048 new message is started.
5050 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5051 debugging output unless log_rewrite was also set.
5053 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5054 of (int)(handle) when we know that handle contains (void *)(-1).
5056 61. The Exim daemon panic-logs an error return when it closes the incoming
5057 connection. However "connection reset by peer" seems to be common, and
5058 isn't really an error worthy of noting specially, so that particular error
5061 62. When Exim is trying to find all the local interfaces, it used to panic and
5062 die if the ioctl to get the interface flags failed. However, it seems that
5063 on at least one OS (Solaris 9) it is possible to have an interface that is
5064 included in the list of interfaces, but for which you get a failure error
5065 for this call. This happens when the interface is not "plumbed" into a
5066 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5067 failure of the "get flags" call assumes that the interface is down.
5069 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5070 makes life easier for people who are doing arithmetic on fields extracted
5071 from dates, where you often get leading zeros that should not be
5072 interpreted as octal.
5074 64. Added qualify_domain to the redirect router, to override the global
5077 65. If a pathologically long header line contained very many addresses (the
5078 report of this problem mentioned 10 000) and each of them was rewritten,
5079 Exim could use up a very large amount of memory. (It kept on making new
5080 copies of the header line as it rewrote, and never released the old ones.)
5081 At the expense of a bit more processing, the header rewriting function has
5082 been changed so that it no longer eats memory in this way.
5084 66. The generation of the Received: header has been moved from the time that a
5085 message starts to be received, to the time that it finishes. The timestamp
5086 in the Received: header should now be very close to that of the <= log
5087 line. There are two side-effects of this change:
5089 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5090 logged header lines no longer include the local Received: line, because
5091 it has not yet been created. The same applies to a copy of the message
5092 that is returned to a non-SMTP sender when a message is rejected.
5094 (b) When a filter file is tested using -bf, no additional Received: header
5095 is added to the test message. After some thought, I decided that this
5098 This change does not affect the value of $received_for. It is still set
5099 after address rewriting, but before local_scan() is called.
5101 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5103 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5104 gave an unhelpful panic error message, and a defer error. I have managed to
5105 change this behaviour so that it now rejects any supplied certificate,
5106 which seems right, as the list of acceptable certificates is empty.
5108 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5109 gave an unhelpful defer error. I have not managed to make this reject any
5110 supplied certificates, but the error message it gives is "no certificate
5111 supplied", which is not helpful.
5113 70. exigrep's output now also includes lines that are not associated with any
5114 message, but which match the given pattern. Implemented by a patch from
5115 Martin Sluka, which also tidied up the Perl a bit.
5117 71. Recipient callout verification, like sender verification, was using <> in
5118 the MAIL FROM command. This isn't really the right thing, since the actual
5119 sender may affect whether the remote host accepts the recipient or not. I
5120 have changed it to use the actual sender in the callout; this means that
5121 the cache record is now keyed on a recipient/sender pair, not just the
5122 recipient address. There doesn't seem to be a real danger of callout loops,
5123 since a callout by the remote host to check the sender would use <>.
5124 [SEE ABOVE: changed after hitting problems.]
5126 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5127 temporary errors. However, in the case of such a code being given after
5128 the end of a data transmission (i.e. after ".") Exim was failing to write
5129 a retry record for the message. (Yes, there was some broken host that was
5130 actually sending 8xx at this point.)
5132 73. An unknown lookup type in a host list could cause Exim to panic-die when
5133 the list was checked. (An example that provoked this was putting <; in the
5134 middle of a list instead of at the start.) If this happened during a DATA
5135 ACL check, a -D file could be left lying around. This kind of configuration
5136 error no longer causes Exim to die; instead it causes a defer errror. The
5137 incident is still logged to the main and panic logs.
5139 74. Buglet left over from Exim 3 conversion. The message "too many messages
5140 in one connection" was written to the rejectlog but not the mainlog, except
5141 when address rewriting (yes!) was being logged.
5143 75. Added write_rejectlog option.
5145 76. When a system filter was run not as root (that is, when system_filter_user
5146 was set), the values of the $n variables were not being returned to the
5147 main process; thus, they were not subsequently available in the $sn
5150 77. Added +return_path_on_delivery log selector.
5152 78. A connection timeout was being treated differently from recipients deferred
5153 when testing hosts_max_try with a message that was older than the host's
5154 retry timeout. (The host should not be counted, thus allowing all hosts to
5155 be tried at least once before bouncing.) This may have been the cause of an
5156 occasionally reported bug whereby a message would remain on the queue
5157 longer than the retry timeout, but would be bounced if a delivery was
5158 forced. I say "may" because I never totally pinned down the problem;
5159 setting up timeout/retry tests is difficult. See also the next item.
5161 79. The ultimate address timeout was not being applied to errors that involved
5162 a combination of host plus message (for example, a timeout on a MAIL
5163 command). When an address resolved to a number of possible hosts, and they
5164 were not all tried for each delivery (e.g. because of hosts_max_try), a
5165 message could remain on the queue longer than the retry timeout.
5167 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5170 81. Fixed an obscure SMTP outgoing bug which required at least the following
5171 conditions: (a) there was another message waiting for the same server;
5172 (b) the server returned 5xx to all RCPT commands in the first message so
5173 that the message was not completed; (c) the server dropped the connection
5174 or gave a negative response to the RSET that Exim sends to abort the
5175 transaction. The observed case was a dropped connection after DATA that had
5176 been sent in pipelining mode. That is, the server had advertised PIPELINING
5177 but was not implementing it correctly. The effect of the bug was incorrect
5178 behaviour, such as trying another host, and this could lead to a crash.
5184 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5185 and daemon.c were passed as pointers to ints; they should have been
5186 pointers to socklen_t variables (which are typically unsigned ints).
5188 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5191 3. Fixed a really odd bug that affected only the testing scheme; patching a
5192 certain fixed string in the binary changed the value of another string that
5193 happened to be identical to the end of the original first string.
5195 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5196 name", it returns that address as the IP address. On some operating
5197 systems (e.g. Solaris), it also passes back the IP address string as the
5198 "host name". However, on others (e.g. Linux), it passes back an empty
5199 string. Exim wasn't checking for this, and was changing the host name to an
5200 empty string, assuming it had been canonicized.
5202 5. Although rare, it is permitted to have more than one PTR record for a given
5203 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5204 all the names associated with an address, because they do in Solaris.
5205 However, it seems that they do not in Linux for data that comes from the
5206 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5207 I found this out when I moved to a new Linux workstation and tried to run
5208 the Exim test suite.
5210 To get round this problem I have changed the code so that it now does its
5211 own call to the DNS to look up PTR records when searching for a host name.
5212 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5213 addresses that are only in /etc/hosts are still found.
5215 This behaviour is, however, controlled by an option called host_lookup_
5216 order, which defaults to "bydns:byaddr". If people want to use the other
5217 order, or indeed, just use one or the other means of lookup, they can
5218 specify it in this variable.
5220 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5221 some operating systems, this comes back from gethostbyaddr() as an empty
5222 string, and this is what Exim used to test for. However, it seems that in
5223 other systems, "." is yielded. Exim now tests for this case too.
5225 7. The values of check_spool_space and check_log_space are now held internally
5226 as a number of kilobytes instead of an absolute number of bytes. If a
5227 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5228 kilobyte. This means that much larger values can be stored.
5230 8. Exim monitor: an attempt to get the action menu when not actually pointing
5231 at a message produces an empty menu entitled "No message selected". This
5232 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5233 no entries in it ("Shell widget menu has zero width and/or height"). So I
5234 have added a single, blank menu entry in this case.
5236 9. Added ${quote_local_part.
5238 10. MIME decoding is now applied to the contents of Subject: header lines when
5241 11. Now that a reference to $sender_host_address automatically causes a reverse
5242 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5243 host lookup before query-style lookups in lists that might use this
5244 variable. This has therefore been abolished, and the "net-" prefix is no
5245 longer necessary for query-style lookups.
5247 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5248 have been a typo for LFLAGS, so it has been changed.
5250 13. The install script calls Exim with "-C /dev/null" in order to find the
5251 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5252 to be output. Howeve, since Exim outputs its version number before the
5253 error, it didn't break the script. It just looked ugly. I fixed this by
5254 always allowing "-C /dev/null" if the caller is root.
5256 14. Ignore overlarge ACL variable number when reading spool file - insurance
5257 against a later release with more variables having written the file.
5259 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5260 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5261 this, as well as the form without the "IPv6" on the front.
5263 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5264 OS/Makefile-Darwin file.
5266 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5267 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5269 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5270 variable. However, some OS use ints for inodes. I've added cast to long int
5271 to get rid of the compiler warning.
5273 19. I had forgotten to lock out "/../" in configuration file names when
5274 ALT_CONFIG_PREFIX was set.
5276 20. Routers used for verification do not need to specify transports. However,
5277 if such a router generated a host list, and callout was configured, Exim
5278 crashed, because it could not find a port number from the (non-existent)
5279 transport. It now assumes port 25 in this circumstance.
5281 21. Added the -t option to exigrep.
5283 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5284 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5285 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5286 them is set, LOOKUP_LSEARCH is forced.
5288 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5290 24. Added sender and host information to the "rejected by local_scan()" log
5291 line; previously there was no indication of these.
5293 25. Added .include_if_exists.
5295 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5296 Linux. It turns out that not all file systems support this. Apparently some
5297 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5298 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5299 not supported on the file descriptor, is now ignored when Exim is trying to
5300 sync a directory. This applies only to Linux.
5302 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5304 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5305 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5308 29. Added support for maildirsize files from supplied patch (modified a bit).
5310 30. The use of :fail: followed by an empty string could lead Exim to respond to
5311 sender verification failures with (e.g.):
5313 550 Verification failed for <xxx>
5314 550 Sender verify failed
5316 where the first response line was missing the '-' that indicates it is not
5317 the final line of the response.
5319 31. The loop for finding the name of the user that called Exim had a hardwired
5320 limit of 10; it now uses the value of finduser_retries, which is used for
5321 all other user lookups.
5323 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5326 33. Exim was neglecting to zero errno before one call of strtol() when
5327 expanding a string and expecting an integer value. On some systems this
5328 resulted in spurious "integer overflow" errors. Also, it was casting the
5329 result into an int without checking.
5331 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5332 did not work. The code looks as if it has *never* worked, though it appears
5333 to have been documented since at least releast 1.62. I have made it work.
5335 35. The "timeout_DNS" error in retry rules, also documented since at least
5336 1.62, also never worked. As it isn't clear exactly what this means, and
5337 clearly it isn't a major issue, I have abolished the feature by treating it
5338 as "timeout", and writing a warning to the main and panic logs.
5340 36. The display of retry rules for -brt wasn't always showing the error code
5343 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5344 timeout_connect_A, timeout_connect_MX.
5346 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5347 to the empty sender.
5349 39. The daemon was not analysing the content of -oX till after it had closed
5350 stderr and disconnected from the controlling terminal. This meant that any
5351 syntax errors were only noted on the panic log, and the return code from
5352 the command was 0. By re-arranging the code a little, I've made the
5353 decoding happen first, so such errors now appear on stderr, and the return
5354 code is 1. However, the actual setting up of the sockets still happens in
5355 the disconnected process, so errors there are still only recorded on the
5358 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5359 connections (as happens on some IP stacks) was logged at start up time as
5360 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5361 it from "IPv6 and IPv4", which means that two separate sockets are being
5364 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5365 says whether AF_INET or AF_INET6 was passed as an argument.
5367 42. Exiwhat output was messed up when time zones were included in log
5370 43. Exiwhat now gives more information about the daemon's listening ports,
5371 and whether -tls-on-connect was used.
5373 44. The "port" option of the smtp transport is now expanded.
5375 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5376 silently ignored. Now an error message is written to the main and panic
5379 46. There's a new ACL modifier called "logwrite" which writes to a log file
5380 as soon as it is encountered.
5382 47. Added $local_user_uid and $local_user_gid at routing time.
5384 48. Exim crashed when trying to verify a sender address that was being
5387 49. Exim was recognizing only a space character after ".include". It now also
5388 recognizes a tab character.
5390 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5391 extracting the relevant information from the specification. The man page no
5392 longer contains scrambled data for the -d option, and I've added a section
5393 at the front about calling Exim under different names.
5395 51. Added "extra_headers" argument to the "mail" command in filter files.
5397 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5400 53. Installed eximstats 1.29.
5402 54. Added transport_filter_timeout as a generic transport option.
5404 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5405 Cc: header lines. This was required by RFC 822, but it not required by RFC
5408 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5409 incoming messages that did not have them. Now it does so only if the
5410 message originates locally, that is, if there is no associated remote host
5411 address. When Resent- header lines are present, this applies to the Resent-
5412 lines rather than the non-Resent- lines.
5414 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5415 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5417 58. Messages for configuration errors now include the name of the main
5418 configuration file - useful now that there may be more than one file in a
5419 list (.included file names were always shown).
5421 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5422 for those rare installations that do not start the daemon as root or run it
5423 setuid root. I've cut out the call to initgroups() if the daemon is not
5426 60. The Exim user and group can now be bound into the binary as text strings
5427 that are looked up at the start of Exim's processing.
5429 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5431 62. Added $mailstore_basename variable.
5433 63. Installed patch to sieve.c from Michael Haardt.
5435 64. When Exim failed to open the panic log after failing to open the main log,
5436 the original message it was trying to log was written to stderr and debug
5437 output, but if they were not available (the usual case in production), it
5438 was lost. Now it is written to syslog before the two lines that record the
5439 failures to open the logs.
5441 65. Users' Exim filters run in subprocesses under the user's uid. It is
5442 possible for a "deliver" command or an alias in a "personal" command to
5443 provoke an address rewrite. If logging of address rewriting is configured,
5444 this fails because the process is not running as root or exim. There may be
5445 a better way of dealing with this, but for the moment (because 4.30 needs
5446 to be released), I have disabled address rewrite logging when running a
5447 filter in a non-root, non-exim process.
5453 1. The buildconfig auxiliary program wasn't quoting the value set for
5454 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5455 not defined. This bug was masked in 4.22 by the effect that was fixed in
5458 2. Some messages that were rejected after a message id was allocated were
5459 shown as "incomplete" by exigrep. It no longer does this for messages that
5460 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5462 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5463 have allow_domain_literals set, the ID did not get logged in the <= line.
5464 Domain literals are now always recognized in Message-ID: header lines.
5466 4. The first argument for a ${extract expansion item is the key name or field
5467 number. Leading and trailing spaces in this item were not being ignored,
5468 causing some misleading effects.
5470 5. When deliver_drop_privilege was set, single queue runner processes started
5471 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5472 same command in the process it spins off) were not dropping privilege.
5474 6. When the daemon running as "exim" started a queue runner, it always
5475 re-executed Exim in the spun-off process. This is a waste of effort when
5476 deliver_drop_privilege is set. The new process now just calls the
5477 queue-runner function directly.
5483 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5486 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5487 ignored. Though the use of -oP was forcing the writing of a pid file, it
5488 was always written to the default place.
5490 3. If the message "no IP address found for host xxxx" is generated during
5491 incoming verification, it is now followed by identification of the incoming
5492 connection (so you can more easily find what provoked it).
5494 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5496 5. Added some features to "harden" Exim a bit more against certain attacks:
5498 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5499 be put in Local/Makefile. This is like the never_users runtime option,
5500 but it cannot be overridden. The default setting is "root".
5502 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5503 prefix string with which any file named in a -C command line option
5506 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5507 is retained for -C and -D only if the caller of Exim is root. Without
5508 it, the exim user may also use -C and -D and retain privilege.
5510 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5511 command line option is disabled.
5513 6. Macro names set by the -D option must start with an upper case letter, just
5514 like macro names defined in the configuration file.
5516 7. Added "dereference=" facility to LDAP.
5518 8. Two instances of the typo "uknown" in the source files are fixed.
5520 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5521 the Configure-Makefile script screwed up while processing it.
5523 10. Incorporated PCRE 4.4.
5525 11. The SMTP synchronization check was not operating right at the start of an
5526 SMTP session. For example, it could not catch a HELO sent before the client
5527 waited for the greeting. There is now a check for outstanding input at the
5528 point when the greeting is written. Because of the duplex, asynchronous
5529 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5530 way, but not yet received, when the check is performed.
5532 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5533 on TCP/IP sockets, because this apparently causes some broken clients to
5536 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5537 unchanged) from the Cygwin maintainer.
5539 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5540 was supported instead of testing for mbx. Effectively a typo.
5542 15. The spa authenticator server code was not checking that the input it
5543 received was valid base64.
5545 16. The debug output line for the "set" modifier in ACLs was not showing the
5546 name of the variable that was being set.
5548 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5550 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5551 lookup on its own. Something else was needed to trigger the lookup. For
5552 example, a match in host_lookup or the need for a host name in a host list.
5553 Now, if $sender_host_name is referenced and the host name has not yet been
5554 looked up, a lookup is performed. If the lookup fails, the variable remains
5555 empty, and $host_lookup_failed is set to "1".
5557 19. Added "eqi" as a case-independent comparison operator.
5559 20. The saslauthd authentication condition could segfault if neither service
5560 nor realm was specified.
5562 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5563 error message that was logged was misleading, and incoming SMTP
5564 connections were dropped. The message is now more accurate, and temporary
5565 errors are given to SMTP connections.
5567 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5568 (except RSET) with a 421 error, until QUIT is received. However, it was
5569 failing to send a response to QUIT.
5571 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5572 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5573 ACL is now run after the helo_try_verify_hosts code.
5575 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5576 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5577 case-independent, so other case variants are also recognized). Apparently
5578 some systems use these upper case variants.
5580 25. If more than two messages were waiting for the same host, and a transport
5581 filter was specified for the transport, Exim sent two messages over the
5582 same TCP/IP connection, and then failed with "socket operation on non-
5583 socket" when it tried to send the third.
5585 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5587 27. The extern definition of crypt16() in expand.c was not being excluded when
5588 the OS had its own crypt16() function.
5590 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5591 as a preferred synonym for return_size_limit, both as an option and as an
5594 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5596 30. Changed the default configuration ACL to relax the local part checking rule
5597 for addresses that are not in any local domains. For these addresses,
5598 slashes and pipe symbols are allowed within local parts, but the sequence
5599 /../ is explicitly forbidden.
5601 31. SPA server authentication was not clearing the challenge buffer before
5604 32. log_message in a "warn" ACL statement was writing to the reject log as
5605 well as to the main log, which contradicts the documentation and doesn't
5606 seem right (because no rejection is happening). So I have stopped it.
5608 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5609 However, I am unable to do any testing of this.
5611 34. Fixed an infelicity in the appendfile transport. When checking directories
5612 for a mailbox, to see if any needed to be created, it was accidentally
5613 using path names with one or more superfluous leading slashes; tracing
5614 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5616 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5617 contents are added to the log line that is written for every discarded
5618 recipient. (Previously a log_message setting was ignored.)
5620 36. The ${quote: operator now quotes the string if it is empty.
5622 37. The install script runs exim in order to find its version number. If for
5623 some reason other than non-existence or emptiness, which it checks, it
5624 could not run './exim', it was installing it with an empty version number,
5625 i.e. as "exim-". This error state is now caught, and the installation is
5628 38. An argument was missing from the function that creates an error message
5629 when Exim fails to connect to the socket for saslauthd authentication.
5630 This could cause Exim to crash, or give a corrupted message.
5632 39. Added isip, isip4, and isip6 to ${if conditions.
5634 40. The ACL variables $acl_xx are now saved with the message, and can be
5635 accessed later in routers, transports, and filters.
5637 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5638 strings in the file are not string-expanded.
5640 42. If a MAIL command specified a SIZE value that was too large to fit into an
5641 int variable, the check against message_size_limit failed. Such values are
5642 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5643 day this will have to be increased, but I don't think I want to be around
5644 when emails are that large.
5651 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5652 iconv() is not standard in FreeBSD.
5654 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5655 IPv6 enabled. The observed symptom was a segmentation fault on return from
5656 the function os_common_find_running_interfaces() in src/os.c.
5658 3. In the check_special_case() function in daemon.c I had used "errno" as an
5659 argument name, which causes warnings on some systems. This was basically a
5660 typo, since it was named "eno" in the comments!
5662 4. The code that waits for the clock to tick (at a resolution of some fraction
5663 of a second) so as to ensure message-id uniqueness was always waiting for
5664 at least one whole tick, when it could have waited for less. [This is
5665 almost certainly not relevant at current processor speeds, where it is
5666 unlikely to ever wait at all. But we try to future-proof.]
5668 5. The function that sleeps for a time interval that includes fractions of a
5669 second contained a race. It did not block SIGALRM between setting the
5670 timer, and suspending (a couple of lines later). If the interval was short
5671 and the sigsuspend() was delayed until after it had expired, the suspension
5672 never ended. On busy systems this could lead to processes getting stuck for
5675 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5676 process, before it forks any delivery processes. The open lookup caching
5677 mechanism meant that the open file or database connection was passed into
5678 the delivery process. The problem was that delivery processes always tidy
5679 up cached lookup data. This could cause a problem for the next delivery
5680 process started by the queue runner, because the external queue runner
5681 process does not know about the closure. So the next delivery process
5682 still has data in the lookup cache. In the case of a file lookup, there was
5683 no problem because closing a file descriptor in a subprocess doesn't affect
5684 the parent. However, if the lookup was caching a connection to a database,
5685 the connection was closed, and the second delivery process was likely to
5686 see errors such as "PGSQL: query failed: server closed the connection
5687 unexpectedly". The problem has been fixed by closing all cached lookups
5688 in a queue runner before running a delivery process.
5690 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5691 seem to have the "const" qualifier which it has on other OS. I've
5694 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5695 *of the same type* (client or server) with the same public name that an
5696 error should be diagnosed.
5698 9. When Exim looked up a host name for an IP address, but failed to find the
5699 original IP address when looking up the host name (a safety check), it
5700 output the message "<ip address> does not match any IP for NULL", which was
5701 confusing, to say the least. The bug was that the host name should have
5702 appeared instead of "NULL".
5704 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5705 user that is built into the binary, and the -C or -D options is used, root
5706 privilege is dropped before the configuration file is read. In addition,
5707 logging is switched to stderr instead of the normal log files. If the
5708 configuration then re-defines the Exim user, the unprivileged environment
5709 is probably not what is expected, so Exim logs a panic warning message (but
5712 However, if deliver_drop_privilege is set, the unprivileged state may well
5713 be exactly what is intended, so the warning has been cut out in that case,
5714 and Exim is allowed to try to write to its normal log files.
5720 1. smtp_return_error_details was not giving details for temporary sender
5721 or receiver verification errors.
5723 2. Diagnose a configuration error if two authenticators have the same public
5726 3. Exim used not to create the message log file for a message until the first
5727 delivery attempt. This could be confusing when incoming messages were held
5728 for policy or load reasons. The message log file is now created at the time
5729 the message is received, and an initial "Received" line is written to it.
5731 4. The automatically generated man page for command line options had a minor
5732 bug that caused no ill effects; however, a more serious problem was that
5733 the procedure for building the man page automatically didn't always
5734 operate. Consequently, release 4.20 contains an out-of-date version. This
5735 shouldn't happen again.
5737 5. When building Exim with embedded Perl support, the script that builds the
5738 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5739 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5741 6. The freeze_tell option was not being used for messages that were frozen on
5742 arrival, either by an ACL or by local_scan().
5744 7. Added the smtp_incomplete_transaction log selector.
5746 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5747 was accepting AUTH without a new EHLO.
5749 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5750 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5753 10. Logging of TCP/IP connections (when configured) now happens in the main
5754 daemon process instead of the child process, so that the TCP/IP connection
5755 count is more accurate (but it can never be perfect).
5757 11. The use of "drop" in a nested ACL was not being handled correctly in the
5758 outer ACL. Now, if condition failure induced by the nested "drop" causes
5759 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5760 or "require"), the connection is dropped.
5762 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5763 that yield "discard" can now be used with an "accept" or a "discard" verb,
5764 but an error is generated for any others (because I can't see a useful way
5765 to define what should happen).
5767 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5768 are now processed in the same way as lines in the Exim configuration file.
5769 In particular, continuation lines are supported.
5771 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5773 15. Added -ti meaning -t -i.
5775 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5776 domains, and warn by logging if others are found.
5778 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5779 returned from GIFCONF when Exim is trying to find the list of interfaces on
5780 a host. The code in os.c has been modified to copy each ifreq to an aligned
5781 structure in all cases.
5783 Also, in some cases, the returned ifreq's were being copied to a 'struct
5784 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5785 means the last couple of bytes of an IPv6 address could be chopped if the
5786 ifreq contained only a normal sockaddr (14 bytes storage).
5788 18. Named domain lists were not supported in the hosts_treat_as_local option.
5789 An entry such as +xxxx was not recognized, and was treated as a literal
5792 19. Ensure that header lines added by a DATA ACL are included in the reject log
5793 if the ACL subsequently rejects the message.
5795 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5796 MD5 (which is deprecated).
5798 21. When testing a filter file using -bf, Exim was writing a message when it
5799 took the sender from a "From " line in the message, but it was not doing so
5800 when it took $return_path from a Return-Path: header line. It now does.
5802 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5803 with a valid header line field name (a series of printing characters
5804 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5806 23. Changed "disc" in the source to "disk" to conform to the documentation and
5807 the book and for uniformity.
5809 24. Ignore Sendmail's -Ooption=value command line item.
5811 25. When execve() failed while trying to run a command in a pipe transport,
5812 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5813 could be confused with a return value of 69 from the command itself. This
5814 has been changed to 127, the value the shell returns if it is asked to run
5815 a non-existent command. The wording for the related log line suggests a
5816 non-existent command as the problem.
5818 26. If received_header_text expands to an empty string, do not add a Received:
5819 header line to the message. (Well, it adds a token one on the spool, but
5820 marks it "old" so that it doesn't get used or transmitted.)
5822 27. Installed eximstats 1.28 (addition of -nt option).
5824 28. There was no check for failure on the call to getsockname() in the daemon
5825 code. This can fail if there is a shortage of resources on the system, with
5826 ENOMEM, for example. A temporary error is now given on failure.
5828 29. Contrary to the C standard, it seems that in some environments, the
5829 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5830 program. Exim now does this explicitly; it affects the formatting of
5831 timestamps using strftime().
5833 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5834 complaints. I've now initialized all the variables, to avoid this.
5836 32. Header lines added by a system filter were not being "seen" during
5837 transport-time rewrites.
5839 33. The info_callback() function passed to OpenSSL is set up with type void
5840 (*)(SSL *, int, int), as described somewhere. However, when calling the
5841 function (actually a macro) that sets it up, the type void(*)() is
5842 expected. I've put in a cast to prevent warnings from picky compilers.
5844 34. If a DNS black list lookup found a CNAME record, but there were no A
5845 records associated with the domain it pointed at, Exim crashed.
5847 35. If a DNS black list lookup returned more than one A record, Exim ignored
5848 all but the first. It now scans all returned addresses if a particular IP
5849 value is being sought. In this situation, the contents of the
5850 $dnslist_value variable are a list of all the addresses, separated by a
5853 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5854 to accept a host name and all its aliases if the forward lookup for any of
5855 them yielded the IP address of the incoming connection. Now it accepts only
5856 those names whose forward lookup yields the correct IP address. Any other
5857 names are discarded. This closes a loophole whereby a rogue DNS
5858 administrator could create reverse DNS records to break through a
5859 wildcarded host restriction in an ACL.
5861 37. If a user filter or a system filter that ran in a subprocess used any of
5862 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5863 the wrong values were passed to the pipe command ($thisaddress had the
5864 value of $0, $0 had the value of $1, etc). This bug was introduced by
5865 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5867 38. Improved the line breaking for long SMTP error messages from ACLs.
5868 Previously, if there was no break point between 40 and 75 characters, Exim
5869 left the rest of the message alone. Two changes have been made: (a) I've
5870 reduced the minimum length to 35 characters; (b) if it can't find a break
5871 point between 35 and 75 characters, it looks ahead and uses the first one
5872 that it finds. This may give the occasional overlong line, but at least the
5873 remaining text gets split now.
5875 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5876 file descriptors might be low, and that setting 1000 would always raise it.
5877 It turns out that in some environments, the limit is already over 1000 and
5878 that lowering it causes trouble. So now Exim takes care not to decrease it.
5880 40. When delivering a message, the value of $return_path is set to $sender_
5881 address at the start of routing (routers may change the value). By an
5882 oversight, this default was not being set up when an address was tested by
5883 -bt or -bv, which affected the outcome if any router or filter referred to
5886 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5887 log only when "message" or "log_message" are set. However, if one of the
5888 conditions was an address verification, or a call to a nested ACL, the
5889 messages generated by the underlying test were being passed through. This
5890 no longer happens. The underlying message is available in $acl_verify_
5891 message for both "message" and "log_message" expansions, so it can be
5892 passed through if needed.
5894 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5895 new expansion $bh_ to give the encoded byte string without charset
5896 translation. Translation happens only if iconv() is available; HAVE_ICONV
5897 indicates this at build time. HEADERS_CHARSET gives the charset to
5898 translate to; headers_charset can change it in the configuration, and
5899 "headers charset" can change it in an individual filter file.
5901 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5902 that creates RFC 2047 encoded "words" labels them as that charset instead
5903 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5904 expansion operator; (ii) when Exim creates a From: line for a local
5905 message; (iii) when a header line is rewritten to include a "phrase" part.
5907 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5908 buggy, causing it to skip the first lines of messages whose message ID
5909 ended in 'D'. This would not have bitten before Exim release 4.14, because
5910 message IDs were unlikely to end in 'D' before then. The effect was to have
5911 incorrect size information for certain domains.
5913 45. #include "config.h" was missing at the start of the crypt16.c module. This
5914 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5917 46. If there was a timeout during a "random" callout check, Exim treated it as
5918 a failure of the random address, and carried on sending RSET and the real
5919 address. If the delay was just some slowness somewhere, the response to the
5920 original RCPT would be taken as a response to RSET and so on, causing
5921 mayhem of various kinds.
5923 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5924 when I implemented it. It didn't allow for the fact that some option values
5925 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5926 the right test for positive values.
5928 48. Domain names in DNS records are case-independent. Exim always looks them up
5929 in lower case. Some resolvers return domain names in exactly the case they
5930 appear in the zone file, that is, they may contain uppercase letters. Not
5931 all resolvers do this - some return always lower case. Exim was treating a
5932 change of case by a resolver as a change of domain, similar to a widening
5933 of a domain abbreviation. This triggered its re-routing code and so it was
5934 trying to route what was effectively the same domain again. This normally
5935 caused routing to fail (because the router wouldn't handle the domain
5936 twice). Now Exim checks for this case specially, and just changes the
5937 casing of the domain that it ultimately uses when it transmits the message
5940 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5943 50. If a filter generated a file delivery with a non-absolute name (possible if
5944 no home directory exists for the router), the forbid_file option was not
5947 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5948 the existing equality matching.
5950 52. Exim was using ints instead of ino_t variables in some places where it was
5951 dealing with inode numbers.
5953 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5954 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5955 called TMPDIR, and if it finds it is different, it changes its value.
5957 54. The smtp_printf() function is now made available to local_scan() so
5958 additional output lines can be written before returning. There is also an
5959 smtp_fflush() function to enable the detection of a dropped connection.
5960 The variables smtp_input and smtp_batched_input are exported to
5963 55. Changed the default runtime configuration: the message "Unknown user"
5964 has been removed from the ACL, and instead placed on the localuser router,
5965 using the cannot_route_message feature. This means that any verification
5966 failures that generate their own messages won't get overridden. Similarly,
5967 the "Unrouteable address" message that was in the ACL for unverifiable
5968 relay addresses has also been removed.
5970 56. Added hosts_avoid_esmtp to the smtp transport.
5972 57. The exicyclog script was not checking for the esoteric option
5973 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5974 will work only if exicyclog is run under the appropriate euid.
5976 58. Following a discussion on the list, the rules by which Exim recognises line
5977 endings on incoming messages have been changed. The -dropcr and drop_cr
5978 options are now no-ops, retained only for backwards compatibility. The
5979 following line terminators are recognized: LF CRLF CR. However, special
5980 processing applies to CR:
5982 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5983 nor a local message in the state where . is a terminator.
5985 (ii) If a bare CR is encountered in a header line, an extra space is added
5986 after the line terminator so as not to end the header. The reasoning
5987 behind this is that bare CRs in header lines are most likely either
5988 to be mistakes, or people trying to play silly games.
5990 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5991 was being incorrectly given as 18 bytes larger than it should have been.
5992 This is a VOB (very old bug).
5994 60. This may never have affected anything current, but just in case it has:
5995 When the local host is found other than at the start of a list of hosts,
5996 the local host, those with the same MX, and any that follow, are discarded.
5997 When the list in question was part of a longer list of hosts, the following
5998 hosts (not currently being processed) were also being discarded. This no
5999 longer happens. I'm not sure if this situation could ever has previously
6002 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6005 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6006 header. This is something that is recommended in a new Internet Draft, and
6007 is something that is documented as being done by Sendmail. There are two
6008 possible values. For messages generated by the autoreply transport, Exim
6011 Auto-Submitted: auto-replied
6013 whereas for all other generated messages (e.g. bounces) it adds
6015 Auto-Submitted: auto-generated
6017 63. The "personal" condition in filters now includes a test for the
6018 Auto-Submitted: header. If it contains the string "auto-" the message it
6019 not considered personal.
6021 64. Added rcpt_include_affixes as a generic transport option.
6023 65. Added queue_only_override (default true).
6025 66. Added the syslog_duplication option.
6027 67. If what should have been the first header line of a message consisted of
6028 a space followed by a colon, Exim was mis-interpreting it as a header line.
6029 It isn't of course - it is syntactically invalid and should therefore be
6030 treated as the start of the message body. The misbehaviour could have
6031 caused a number of strange effects, including loss of data in subsequent
6032 header lines, and spool format errors.
6034 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6035 client host had authenticated. This control can now be exercised by an ACL
6036 for more flexibility.
6038 69. By default, callouts do not happen when testing with -bh. There is now a
6039 variant, -bhc, which does actually run the callout code, including
6040 consulting and updating the callout cache.
6042 70. Added support for saslauthd authentication, courtesy of Alexander
6045 71. If statvfs() failed on the spool or log directories while checking their
6046 size for availability, Exim confusingly gave the error "space shortage".
6047 Furthermore, in debugging mode it crashed with a floating point exception.
6048 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6049 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6050 serious problem, Exim now writes to the main and panic logs when this
6051 happens, with details of the failure. It then refuses to accept the
6052 incoming message, giving the message "spool directory problem" or "log
6053 directory problem" with a 421 code for SMTP messages.
6055 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6056 0, 1, and 2 exist, because some OS complain for execs without them (see
6057 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6058 descriptors. However, the code omitted to check that the open succeeded,
6059 causing mysterious errors if for some reason the permissions on /dev/null
6060 got screwed. Now Exim writes a message to the main and panic logs, and
6061 bombs out if it can't open /dev/null.
6063 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6064 interact so that it is all more flexible. It is supposed to remain
6065 backwards compatible. Also added extra_local_interfaces.
6067 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6068 to bomb out with an assertion failure - to the client this appears as a
6069 connection drop. This problem occurs in the part of the code that was taken
6070 from the Samba project. Fortunately, the assertion is in a very simple
6071 function, so I have fixed this by reproducing the function inline in the
6072 one place where it is called, and arranging for authentication to fail
6073 instead of killing the process with assert().
6075 75. The SPA client code was not working when the server requested OEM rather
6076 than Unicode encoding.
6078 76. Added code to make require_files with a specific uid setting more usable in
6079 the case where statting the file as root fails - usually a non-root-mounted
6080 NFS file system. When this happens and the failure is EACCES, Exim now
6081 forks a subprocess and does the per-uid checking as the relevant uid.
6083 77. Added process_log_path.
6085 78. If log_file_path was not explicitly set, a setting of check_log_space or
6086 check_log_inodes was ignored.
6088 79. If a space check for the spool or log partitions fails, the incident is now
6089 logged. Of course, in the latter case the data may get lost...
6091 80. Added the %p formatting code to string_format() so that it can be used to
6092 print addresses in debug_print(). Adjusted all the address printing in the
6093 debugging in store.c to use %p rather than %d.
6095 81. There was a concern that a line of code in smtp_in.c could overflow a
6096 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6097 initially expressed, the concern was not well-founded, because trailing
6098 spaces are removed early. However, if the trailing spaces were followed by
6099 a NULL, they did not get removed, so the overflow was possible. Two fixes
6102 (a) I re-wrote the offending code in a cleaner fashion.
6103 (b) If an incoming SMTP command contains a NULL character, it is rejected
6106 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6107 runs initgroups(), so that if the Exim user is in any additional groups,
6108 they will be used during message reception.
6114 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob