1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
8 checkers that were complaining about end of non-void function with no
15 TF/01 Correctly close the server side of TLS when forking for delivery.
17 When a message was received over SMTP with TLS, Exim failed to clear up
18 the incoming connection properly after forking off the child process to
19 deliver the message. In some situations the subsequent outgoing
20 delivery connection happened to have the same fd number as the incoming
21 connection previously had. Exim would try to use TLS and fail, logging
22 a "Bad file descriptor" error.
24 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
25 utilities have not been installed.
27 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
28 temporary space as the ACL may create new global variables.
30 TL/01 LDAP support uses per connection or global context settings, depending
31 upon the detected version of the libraries at build time.
33 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
34 to extract and use the src ip:port in logging and expansions as if it
35 were a direct connection from the outside internet. PPv2 support was
36 updated based on HAProxy spec change in May 2014.
38 JH/02 Add ${listextract {number}{list}{success}{fail}}.
40 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
41 Properly escape header and check for NULL return.
43 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
46 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
48 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
49 characters in header names, implemented as a verify condition.
50 Contributed by Michael Fischer v. Mollard.
52 TL/05 Rename SPF condition results err_perm and err_temp to standardized
53 results permerror and temperror. Previous values are deprecated but
54 still accepted. In a future release, err_perm and err_temp will be
55 completely removed, which will be a backward incompatibility if the
56 ACL tests for either of these two old results. Patch contributed by
57 user bes-internal on the mailing list.
59 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
61 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
62 selectors, in both main and reject logs.
64 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
67 JH/07 Add malware type "sock" for talking to simple daemon.
69 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
71 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
72 routers/transports under cutthrough routing.
74 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
75 numbers. Touch up "bool" conditional to keep the same definition.
77 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
79 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
81 JH/12 Expand items in router/transport headers_add or headers_remove lists
82 individually rather than the list as a whole. Bug 1452.
84 Required for reasonable handling of multiple headers_ options when
85 they may be empty; requires that headers_remove items with embedded
86 colons must have them doubled (or the list-separator changed).
88 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
89 view the policy declared in the DMARC record. Currently, $dmarc_status
90 is a combined value of both the record presence and the result of the
93 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
95 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
96 dnslookup router and the smtp transport (applying to the forward
99 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
100 of ldap servers used for a specific lookup. Patch provided by Heiko
103 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
104 New variable $lookup_dnssec_authenticated for observability.
106 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
107 Patch submitted by Lars Timman.
109 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
111 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
112 Requires trusted mode and valid format message id, aborts otherwise.
113 Patch contributed by Heiko Schlichting.
115 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
116 certextract with support for various fields. Bug 1358.
118 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
119 is requested by default, modifiable by smtp transport option
122 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
123 operate on certificate variables to give certificate fingerprints
124 Also new ${sha256:cert_variable}.
126 JH/23 The PRDR feature is moved from being Experimental into the mainline.
128 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
129 Christian Aistleitner.
131 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
133 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
134 file. Patch from Wolfgang Breyha.
136 JH/25 Expand the coverage of the delivery $host and $host_address to
137 client authenticators run in verify callout. Bug 1476.
139 JH/26 Port service names are now accepted for tls_on_connect_ports, to
140 align with daemon_smtp_ports. Bug 72.
142 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
143 support and error reporting did not work properly.
145 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
146 and is readable. Patch from Andrew Colin Kissa.
148 TL/14 Enhance documentation of ${run expansion and how it parses the
149 commandline after expansion, particularly in the case when an
150 unquoted variable expansion results in an empty value.
152 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
154 PP/02 Fix internal collision of T_APL on systems which support RFC3123
155 by renaming away from it. Addresses GH issue 15, reported by
158 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
160 TL/15 SECURITY: prevent double expansion in math comparison functions
161 (can expand unsanitized data). Not remotely exploitable.
168 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
170 PP/02 Make -n do something, by making it not do something.
171 When combined with -bP, the name of an option is not output.
173 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
176 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
177 $sender_host_name and config options to manage this, and basic check
180 PP/05 DSCP support for outbound connections and control modifier for inbound.
182 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
183 (Only plugin which currently uses this is kerberos4, which nobody should
184 be using, but we should make it available and other future plugins might
185 conceivably use it, even though it would break NAT; stuff *should* be
186 using channel bindings instead).
188 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
189 name; added for Sendmail compatibility; requires admin caller.
190 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
191 just ignore it); requires trusted caller.
192 Also parse but ignore: -Ac -Am -X<logfile>
195 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
197 TL/02 Add +smtp_confirmation as a default logging option.
199 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
200 Patch by Magnus Holmgren from 2007-02-20.
202 TL/04 Bugzilla 1281 - Spec typo.
203 Bugzilla 1283 - Spec typo.
204 Bugzilla 1290 - Spec grammar fixes.
206 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
208 TL/06 Add Experimental DMARC support using libopendmarc libraries.
210 TL/07 Fix an out of order global option causing a segfault. Reported to dev
211 mailing list by by Dmitry Isaikin.
213 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
215 JH/02 Support "G" suffix to numbers in ${if comparisons.
217 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
219 NM/01 Bugzilla 1197 - Spec typo
220 Bugzilla 1196 - Spec examples corrections
222 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
224 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
225 gnutls_enable_pkcs11, but renamed to more accurately indicate its
228 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
229 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
231 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
232 "acl {{name}{arg}...}", and optional args on acl condition
235 JH/05 Permit multiple router/transport headers_add/remove lines.
237 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
239 JH/07 Avoid using a waiting database for a single-message-only transport.
240 Performance patch from Paul Fisher. Bugzilla 1262.
242 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
245 JH/09 Add $headers_added variable, with content from use of ACL modifier
246 add_header (but not yet added to the message). Bugzilla 199.
248 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
249 Pulled from Bugzilla 817 by Wolfgang Breyha.
251 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
253 (nb: this is the same fix as in Exim 4.80.1)
255 JH/11 Add A= logging on delivery lines, and a client_set_id option on
258 JH/12 Add optional authenticated_sender logging to A= and a log_selector
261 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
263 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
264 advertise SMTP AUTH mechanism to us, instead of a generic
265 protocol violation error. Also, make Exim more robust to bad
266 data from the Dovecot auth socket.
268 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
270 When a queue runner is handling a message, Exim first routes the
271 recipient addresses, during which it prunes them based on the retry
272 hints database. After that it attempts to deliver the message to
273 any remaining recipients. It then updates the hints database using
276 So if a recipient address works intermittently, it can get repeatedly
277 deferred at routing time. The retry hints record remains fresh so the
278 address never reaches the final cutoff time.
280 This is a fairly common occurrence when a user is bumping up against
281 their storage quota. Exim had some logic in its local delivery code
282 to deal with this. However it did not apply to per-recipient defers
283 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
285 This change adds a proper retry rule check during routing so that the
286 final cutoff time is checked against the message's age. We only do
287 this check if there is an address retry record and there is not a
288 domain retry record; this implies that previous attempts to handle
289 the address had the retry_use_local_parts option turned on. We use
290 this as an approximation for the destination being like a local
291 delivery, as in LMTP.
293 I suspect this new check makes the old local delivery cutoff check
294 redundant, but I have not verified this so I left the code in place.
296 TF/02 Correct gecos expansion when From: is a prefix of the username.
298 Test 0254 submits a message to Exim with the header
302 When I ran the test suite under the user fanf2, Exim expanded
303 the header to contain my full name, whereas it should have added
304 a Resent-Sender: header. It erroneously treats any prefix of the
305 username as equal to the username.
307 This change corrects that bug.
309 GF/01 DCC debug and logging tidyup
310 Error conditions log to paniclog rather than rejectlog.
311 Debug lines prefixed by "DCC: " to remove any ambiguity.
313 TF/03 Avoid unnecessary rebuilds of lookup-related code.
315 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
316 Bug spotted by Jeremy Harris; was flawed since initial commit.
317 Would have resulted in OCSP responses post-SNI triggering an Exim
318 NULL dereference and crash.
320 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
322 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
323 Bug detection, analysis and fix by Samuel Thibault.
324 Bugzilla 1331, Debian bug #698092.
326 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
328 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
329 Server implementation by Todd Lyons, client by JH.
330 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
331 config variable "prdr_enable" controls whether the server
332 advertises the facility. If the client requests PRDR a new
333 acl_data_smtp_prdr ACL is called once for each recipient, after
334 the body content is received and before the acl_smtp_data ACL.
335 The client is controlled by bolth of: a hosts_try_prdr option
336 on the smtp transport, and the server advertisement.
337 Default client logging of deliveries and rejections involving
338 PRDR are flagged with the string "PRDR".
340 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
341 fclose(). Diagnosis by Todd Lyons.
343 PP/17 Update configure.default to handle IPv6 localhost better.
344 Patch by Alain Williams (plus minor tweaks).
347 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
348 This is now consistent with GnuTLS, and is now documented: the
349 previous undocumented portable approach to treating the option as
350 unset was to force an expansion failure. That still works, and
351 an empty string is now equivalent.
353 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
354 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
355 not performing validation itself.
357 PP/20 Added force_command boolean option to pipe transport.
358 Patch from Nick Koston, of cPanel Inc.
360 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
363 TF/04 Added udpsend ACL modifer and hexquote expansion operator
365 PP/21 Fix eximon continuous updating with timestamped log-files.
366 Broken in a format-string cleanup in 4.80, missed when I repaired the
367 other false fix of the same issue.
368 Report and fix from Heiko Schlichting.
371 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
372 Report from Prashanth Katuri.
374 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
375 It's SecureTransport, so affects any MacOS clients which use the
376 system-integrated TLS libraries, including email clients.
378 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
379 using a MIME ACL for non-SMTP local injection.
380 Report and assistance in diagnosis by Warren Baker.
382 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
384 JH/16 Fix comparisons for 64b. Bugzilla 1385.
386 TL/09 Add expansion variable $authenticated_fail_id to keep track of
387 last id that failed so it may be referenced in subsequent ACL's.
389 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
392 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
393 ldap library initialization, allowing self-signed CA's to be
394 used. Also properly sets require_cert option later in code by
395 using NULL (global ldap config) instead of ldap handle (per
396 session). Bug diagnosis and testing by alxgomz.
398 TL/12 Enhanced documentation in the ratelimit.pl script provided in
399 the src/util/ subdirectory.
401 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
402 renamed to Transport Post Delivery Action by Jeremy Harris, as
405 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
406 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
407 redis_servers = needs to be configured which will be used by the redis
408 lookup. Patch from Warren Baker, of The Packet Hub.
410 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
412 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
413 hostname or reverse DNS when processing a host list. Used suggestions
414 from multiple comments on this bug.
416 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
418 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
419 Missed a few lines, added it to make the runtest require no keyboard
422 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
423 contains upper case chars. Make router use caseful_local_part.
425 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
426 support when GnuTLS has been built with p11-kit.
432 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
434 This, or similar/improved, will also be change PP/11 of 4.82.
440 PP/01 Handle short writes when writing local log-files.
441 In practice, only affects FreeBSD (8 onwards).
442 Bugzilla 1053, with thanks to Dmitry Isaikin.
444 NM/01 Bugzilla 949 - Documentation tweak
446 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
449 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
451 PP/02 Implemented gsasl authenticator.
453 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
455 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
456 `pkg-config foo` for cflags/libs.
458 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
459 with rest of GSASL and with heimdal_gssapi.
461 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
462 `pkg-config foo` for cflags/libs for the TLS implementation.
464 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
465 properties get this fed in as external SSF. A number of robustness
466 and debugging improvements to the cyrus_sasl authenticator.
468 PP/08 cyrus_sasl server now expands the server_realm option.
470 PP/09 Bugzilla 1214 - Log authentication information in reject log.
471 Patch by Jeremy Harris.
473 PP/10 Added dbmjz lookup type.
475 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
477 PP/12 MAIL args handles TAB as well as SP, for better interop with
478 non-compliant senders.
479 Analysis and variant patch by Todd Lyons.
481 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
482 Bug report from Lars Müller <lars@samba.org> (via SUSE),
483 Patch from Dirk Mueller <dmueller@suse.com>
485 PP/13 tls_peerdn now print-escaped for spool files.
486 Observed some $tls_peerdn in wild which contained \n, which resulted
487 in spool file corruption.
489 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
490 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
491 or write after TLS renegotiation, which otherwise led to messages
494 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
495 as a tracking header (ie: a signed header comes before the signature).
496 Patch from Wolfgang Breyha.
498 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
499 comma-sep list; embedded commas doubled.
501 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
503 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
505 Report and patch from Dmitry Banschikov.
507 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
508 Removed SSL_clear() after SSL_new() which led to protocol negotiation
509 failures. We appear to now support TLS1.1+ with Exim.
511 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
512 lets Exim select keys and certificates based upon TLS SNI from client.
513 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
514 before an outbound SMTP session. New log_selector, +tls_sni.
516 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
517 NULL dereference. Report and patch from Alun Jones.
519 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
520 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
521 Not seeing resolver debug output on NetBSD, but suspect this is a
522 resolver implementation change.
524 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
525 Left warnings. Added "eximon gdb" invocation mode.
527 PP/21 Defaulting "accept_8bitmime" to true, not false.
529 PP/22 Added -bw for inetd wait mode support.
531 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
532 locate the relevant includes and libraries. Made this the default.
534 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
535 Bugzilla 1246, report and most of solution from Tomasz Kusy.
537 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
538 This may cause build issues on older platforms.
540 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
541 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
542 gnutls_require_mac & gnutls_require_protocols (no longer supported).
543 Added SNI support via GnuTLS too.
544 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
546 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
548 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
549 Applied second patch from Janne, implementing suggestion to default
550 multiple-strings-in-record handling to match SPF spec.
552 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
554 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
555 read-only, out of scope).
556 Patch from Wolfgang Breyha, report from Stuart Northfield.
558 PP/29 Fix three issues highlighted by clang analyser static analysis.
559 Only crash-plausible issue would require the Cambridge-specific
560 iplookup router and a misconfiguration.
561 Report from Marcin Mirosław.
563 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
565 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
566 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
567 As part of this, removing so much warning spew let me fix some minor
568 real issues in debug logging.
570 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
571 assignment on my part. Fixed.
573 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
574 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
575 Janne Snabb (who went above and beyond: thank you).
577 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
578 string otherwise requires a connection and a bunch more work and it's
579 relatively easy to get wrong. Should also expose TLS library linkage
582 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
583 64-bit ${eval} (JH/03).
585 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
586 GNU libc to support some of the 64-bit stuff, should not lead to
587 conflicts. Defined before os.h is pulled in, so if a given platform
588 needs to override this, it can.
590 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
591 protection layer was required, which is not implemented.
592 Bugzilla 1254, patch from Wolfgang Breyha.
594 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
595 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
596 tls_dhparam take prime identifiers. Also unbreak combination of
597 OpenSSL+DH_params+TLSSNI.
599 PP/39 Disable SSLv2 by default in OpenSSL support.
605 PP/01 Solaris build fix for Oracle's LDAP libraries.
606 Bugzilla 1109, patch from Stephen Usher.
608 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
610 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
613 TF/02 Fix a couple more cases where we did not log the error message
614 when unlink() failed. See also change 4.74-TF/03.
616 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
617 lock up or crash if it happened to be inside a call to libc when it
618 got a SIGUSR1 from exiwhat.
620 The SIGUSR1 handler appends the current process status to the process
621 log which is later printed by exiwhat. It used to use the general
622 purpose logging code to do this, but several functions it calls are
623 not safe for signals.
625 The new output code in the SIGUSR1 handler is specific to the process
626 log, and simple enough that it's easy to inspect for signal safety.
627 Removing some special cases also simplifies the general logging code.
628 Removing the spurious timestamps from the process log simplifies
631 TF/04 Improved ratelimit ACL condition.
633 The /noupdate option has been deprecated in favour of /readonly which
634 has clearer semantics. The /leaky, /strict, and /readonly update modes
635 are mutually exclusive. The update mode is no longer included in the
636 database key; it just determines when the database is updated. (This
637 means that when you upgrde Exim will forget old rate measurements.)
639 Exim now checks that the per_* options are used with an update mode that
640 makes sense for the current ACL. For example, when Exim is processing a
641 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
642 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
643 must specify per_mail/readonly. If you omit the update mode it defaults to
644 /leaky where that makes sense (as before) or /readonly where required.
646 The /noupdate option is now undocumented but still supported for
647 backwards compatibility. It is equivalent to /readonly except that in
648 ACLs where /readonly is required you may specify /leaky/noupdate or
649 /strict/noupdate which are treated the same as /readonly.
651 A useful new feature is the /count= option. This is a generalization
652 of the per_byte option, so that you can measure the throughput of other
653 aggregate values. For example, the per_byte option is now equivalent
654 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
656 The per_rcpt option has been generalized using the /count= mechanism
657 (though it's more complicated than the per_byte equivalence). When it is
658 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
659 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
660 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
661 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
662 Note that using per_rcpt with a non-readonly update mode in more than
663 one ACL will cause the recipients to be double-counted. (The per_mail
664 and per_byte options don't have this problem.)
666 The handling of very low rates has changed slightly. If the computed rate
667 is less than the event's count (usually one) then this event is the first
668 after a long gap. In this case the rate is set to the same as this event's
669 count, so that the first message of a spam run is counted properly.
671 The major new feature is a mechanism for counting the rate of unique
672 events. The new per_addr option counts the number of different
673 recipients that someone has sent messages to in the last time period. It
674 behaves like per_rcpt if all the recipient addresses are different, but
675 duplicate recipient addresses do not increase the measured rate. Like
676 the /count= option this is a general mechanism, so the per_addr option
677 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
678 example, measure the rate that a client uses different sender addresses
679 with the options per_mail/unique=$sender_address. There are further
680 details in the main documentation.
682 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
684 TF/06 Removed a few PCRE remnants.
686 TF/07 Automatically extract Exim's version number from tags in the git
687 repository when doing development or release builds.
689 PP/02 Raise smtp_cmd_buffer_size to 16kB.
690 Bugzilla 879. Patch from Paul Fisher.
692 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
693 Heavily based on revision 40f9a89a from Simon Arlott's tree.
696 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
698 PP/05 Variable $av_failed, true if the AV scanner deferred.
699 Bugzilla 1078. Patch from John Horne.
701 PP/06 Stop make process more reliably on build failure.
702 Bugzilla 1087. Patch from Heiko Schlittermann.
704 PP/07 Make maildir_use_size_file an _expandable_ boolean.
705 Bugzilla 1089. Patch from Heiko Schlittermann.
707 PP/08 Handle ${run} returning more data than OS pipe buffer size.
708 Bugzilla 1131. Patch from Holger Weiß.
710 PP/09 Handle IPv6 addresses with SPF.
711 Bugzilla 860. Patch from Wolfgang Breyha.
713 PP/10 GnuTLS: support TLS 1.2 & 1.1.
715 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
718 PP/11 match_* no longer expand right-hand-side by default.
719 New compile-time build option, EXPAND_LISTMATCH_RHS.
720 New expansion conditions, "inlist", "inlisti".
722 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
724 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
726 PP/14 fix log_write() format string regression from TF/03.
727 Bugzilla 1152. Patch from Dmitry Isaikin.
733 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
735 PP/02 Harmonised TLS library version reporting; only show if debugging.
736 Layout now matches that introduced for other libraries in 4.74 PP/03.
738 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
740 PP/04 New "dns_use_edns0" global option.
742 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
745 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
746 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
748 TK/01 Updated PolarSSL code to 0.14.2.
749 Bugzilla 1097. Patch from Andreas Metzler.
751 PP/07 Catch divide-by-zero in ${eval:...}.
754 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
757 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
758 format-string attack -- SECURITY: remote arbitrary code execution.
760 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
761 time unintentionally subject to list matching rules, letting the header
762 cause arbitrary Exim lookups (of items which can occur in lists, *not*
763 arbitrary string expansion). This allowed for information disclosure.
765 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
766 INT_MIN/-1 -- value coerced to INT_MAX.
772 NM/01 Workround for PCRE version dependency in version reporting
775 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
776 This fixes portability to compilers other than gcc, notably
777 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
779 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
780 makefiles for portability to HP-UX and POSIX correctness.
782 PP/01 Permit LOOKUP_foo enabling on the make command-line.
783 Also via indented variable definition in the Makefile.
784 (Debugging by Oliver Heesakkers).
786 PP/02 Restore caching of spamd results with expanded spamd_address.
787 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
789 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
790 Improves build reliability. Fix from: Frank Elsner
792 NM/02 Fix wide character breakage in the rfc2047 coding
793 Fixes bug 1064. Patch from Andrey N. Oktyabrski
795 NM/03 Allow underscore in dnslist lookups
796 Fixes bug 1026. Patch from Graeme Fowler
798 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
799 Code patches from Adam Ciarcinski of NetBSD.
801 NM/04 Fixed exiqgrep to cope with mailq missing size issue
804 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
805 is logged, to avoid truncation. Patch from John Horne.
807 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
808 Patch from Jakob Hirsch.
810 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
811 SQL string expansion failure details.
812 Patch from Andrey Oktyabrski.
814 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
815 Patch from Simon Arlott.
817 PP/09 New lookups functionality failed to compile on old gcc which rejects
818 extern declarations in function scope.
819 Patch from Oliver Fleischmann
821 PP/10 Use sig_atomic_t for flags set from signal handlers.
822 Check getgroups() return and improve debugging.
823 Fixed developed for diagnosis in bug 927 (which turned out to be
826 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
827 Patch from Mark Zealey.
829 PP/12 Bugzilla 1056: Improved spamd server selection.
830 Patch from Mark Zealey.
832 PP/13 Bugzilla 1086: Deal with maildir quota file races.
833 Based on patch from Heiko Schlittermann.
835 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
836 Patch from Uwe Doering, sign-off by Michael Haardt.
838 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
839 variable declaration deep within a block. Bug and patch from
842 PP/15 lookups-Makefile IRIX compatibilty coercion.
844 PP/16 Make DISABLE_DKIM build knob functional.
846 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
847 Patch by Simon Arlott
849 TF/03 Fix valgrind.h portability to C89 compilers that do not support
850 variable argument macros. Our copy now differs from upstream.
856 TF/01 Failure to get a lock on a hints database can have serious
857 consequences so log it to the panic log.
859 TF/02 Log LMTP confirmation messages in the same way as SMTP,
860 controlled using the smtp_confirmation log selector.
862 TF/03 Include the error message when we fail to unlink a spool file.
864 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
865 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
866 for maintaining out-of-tree patches for some time.
868 PP/01 Bugzilla 139: Documentation and portability issues.
869 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
870 Handle per-OS dynamic-module compilation flags.
872 PP/02 Let /dev/null have normal permissions.
873 The 4.73 fixes were a little too stringent and complained about the
874 permissions on /dev/null. Exempt it from some checks.
875 Reported by Andreas M. Kirchwitz.
877 PP/03 Report version information for many libraries, including
878 Exim version information for dynamically loaded libraries. Created
879 version.h, now support a version extension string for distributors
880 who patch heavily. Dynamic module ABI change.
882 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
883 privilege escalation vulnerability whereby the Exim run-time user
884 can cause root to append content of the attacker's choosing to
887 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
890 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
891 If dropping privileges for untrusted macros, we disabled normal logging
892 on the basis that it would fail; for the Exim run-time user, this is not
893 the case, and it resulted in successful deliveries going unlogged.
894 Fixed. Reported by Andreas Metzler.
900 PP/01 Date: & Message-Id: revert to normally being appended to a message,
901 only prepend for the Resent-* case. Fixes regression introduced in
902 Exim 4.70 by NM/22 for Bugzilla 607.
904 PP/02 Include check_rfc2047_length in configure.default because we're seeing
905 increasing numbers of administrators be bitten by this.
907 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
909 PP/03 Bugzilla 994: added openssl_options main configuration option.
911 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
913 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
915 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
917 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
918 without a peer certificate, leading to a segfault because of an
919 assumption that peers always have certificates. Be a little more
920 paranoid. Problem reported by Martin Tscholak.
922 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
923 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
924 NB: ClamAV planning to remove STREAM in "middle of 2010".
925 CL also introduces -bmalware, various -d+acl logging additions and
926 more caution in buffer sizes.
928 PP/09 Implemented reverse_ip expansion operator.
930 PP/10 Bugzilla 937: provide a "debug" ACL control.
932 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
934 PP/12 Bugzilla 973: Implement --version.
936 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
938 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
940 PP/15 Bugzilla 816: support multiple condition rules on Routers.
942 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
943 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
944 ignore trailing whitespace.
946 JJ/02 prevent non-panic DKIM error from being sent to paniclog
948 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
951 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
952 Notification from Dr Andrew Aitchison.
954 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
955 ExtendedDetectionInfo response format.
956 Notification from John Horne.
958 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
961 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
962 XSL and documented dependency on system catalogs, with examples of how
965 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
968 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
969 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
970 configuration file which is writeable by the Exim user or group.
972 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
973 of configuration files to cover files specified with the -C option if
974 they are going to be used with root privileges, not just the default
977 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
978 option (effectively making it always true).
980 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
981 files to be used while preserving root privileges.
983 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
984 that rogue child processes cannot use them.
986 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
987 run-time user, instead of root.
989 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
990 Exim run-time user without dropping privileges.
992 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
993 result string, instead of calling string_vformat() twice with the same
996 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
997 for other users. Others should always drop root privileges if they use
998 -C on the command line, even for a whitelisted configure file.
1000 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1002 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1008 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1009 $data_path, and $header_path variables; fixed documentation bugs and
1012 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1013 exipick to access non-standard spools, including the "frozen" queue
1016 NM/01 Bugzilla 965: Support mysql stored procedures.
1017 Patch from Alain Williams
1019 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1021 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1022 Patch from Andreas Metzler
1024 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1025 Patch from Kirill Miazine
1027 NM/05 Bugzilla 671: Added umask to procmail example.
1029 JJ/03 installed exipick 20100323.0, fixing doc bug
1031 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1032 directory. Notification and patch from Dan Rosenberg.
1034 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1036 TK/02 Improve log output when DKIM signing operation fails.
1038 MH/01 Treat the transport option dkim_domain as a colon separated
1039 list, not as a single string, and sign the message with each element,
1040 omitting multiple occurences of the same signer.
1042 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1043 Bugzilla 985, 986. Patch by Simon Arlott
1045 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1046 Patch by Simon Arlott
1048 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1049 MBX locking. Notification from Dan Rosenberg.
1055 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1057 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1059 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1061 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1063 NM/04 Bugzilla 915: Flag broken perl installation during build.
1069 TK/01 Added patch by Johannes Berg that expands the main option
1070 "spamd_address" if it starts with a dollar sign.
1072 TK/02 Write list of recipients to X-Envelope-Sender header when building
1073 the mbox-format spool file for content scanning (suggested by Jakob
1076 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1077 (http://www.dcc-servers.net/) support via dccifd. Activated by
1078 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1080 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1081 by Mark Daniel Reidel <mr@df.eu>.
1083 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1084 When building exim an external PCRE library is now needed -
1085 PCRE is a system library on the majority of modern systems.
1086 See entry on PCRE_LIBS in EDITME file.
1088 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1089 conversation. Added nologin parameter to request.
1090 Patch contributed by Kirill Miazine.
1092 TF/01 Do not log submission mode rewrites if they do not change the address.
1094 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1096 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1097 log files in place. Contributed by Roberto Lima.
1099 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1101 TF/03 Bugzilla 615: When checking the local_parts router precondition
1102 after a local_part_suffix or local_part_prefix option, Exim now
1103 does not use the address's named list lookup cache, since this
1104 contains cached lookups for the whole local part.
1106 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1107 Robert Millan. Documentation is in experimental-spec.txt.
1109 TF/04 Bugzilla 668: Fix parallel build (make -j).
1111 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1113 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1114 Patch provided by Jan Srzednicki.
1116 TF/05 Leading white space used to be stripped from $spam_report which
1117 wrecked the formatting. Now it is preserved.
1119 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1120 that they are available at delivery time.
1122 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1124 TF/08 TLS error reporting now respects the incoming_interface and
1125 incoming_port log selectors.
1127 TF/09 Produce a more useful error message if an SMTP transport's hosts
1128 setting expands to an empty string.
1130 NM/06 Bugzilla 744: EXPN did not work under TLS.
1131 Patch provided by Phil Pennock.
1133 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1134 Patch provided by Richard Godbee.
1136 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1137 acl_smtp_notquit, added index entry.
1139 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1140 Patch provided by Eugene Bujak.
1142 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1143 accept(). Patch provided by Maxim Dounin.
1145 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1146 Patch provided by Phil Pennock.
1148 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1150 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1151 Patch provided by Brad "anomie" Jorsch.
1153 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1154 Patch provided by Dean Brooks.
1156 TK/05 Add native DKIM support (does not depend on external libraries).
1158 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1159 Patch provided by Graeme Fowler.
1161 NM/16 Bugzilla 851: Documentation example syntax fix.
1163 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1165 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1168 NM/19 Bugzilla 745: TLS version reporting.
1169 Patch provided by Phil Pennock.
1171 NM/20 Bugzilla 167: bool: condition support.
1172 Patch provided by Phil Pennock.
1174 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1175 clients. Patch provided by Phil Pennock.
1177 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1178 Patch provided by Brad "anomie" Jorsch.
1180 NM/23 Bugzilla 687: Fix misparses in eximstats.
1181 Patch provided by Heiko Schlittermann.
1183 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1184 Patch provided by Heiko Schlittermann.
1186 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1187 plus update to original patch.
1189 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1191 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1192 Patch provided by David Brownlee.
1194 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1196 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1198 NM/30 Bugzilla 888: TLS documentation bugfixes.
1200 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1202 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1203 Unlike the original bugzilla I have changed all shell scripts in src tree.
1205 NM/33 Bugzilla 898: Transport filter timeout fix.
1206 Patch by Todd Rinaldo.
1208 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1209 Patch by Serge Demonchaux.
1211 NM/35 Bugzilla 39: Base64 decode bug fixes.
1212 Patch by Jakob Hirsch.
1214 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1216 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1218 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1220 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1226 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1227 ALT-N's libdkim that I have put here:
1228 http://duncanthrax.net/exim-experimental/
1230 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1231 were called 'true' and it seems that C99 defines that as a reserved
1232 keyword to be used with 'bool' variable types. That means you could
1233 not include C99-style headers which use bools without triggering
1234 build errors in sieve.c.
1236 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1237 as mailq or other aliases. Changed the --help handling significantly
1238 to do whats expected. exim_usage() emits usage/help information.
1240 SC/01 Added the -bylocaldomain option to eximstats.
1242 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1244 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1246 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1252 PH/01 Another patch from the Sieve maintainer.
1254 PH/02 When an IPv6 address is converted to a string for single-key lookup
1255 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1256 dots are used instead of colons so that keys in lsearch files need not
1257 contain colons. This was done some time before quoting was made available
1258 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1259 (notated using the quote facility) so as to distinguish them from IPv4
1260 keys. This meant that lookups for IP addresses in host lists did not work
1261 for iplsearch lookups.
1263 This has been fixed by arranging for IPv6 addresses to be expressed with
1264 colons if the lookup type is iplsearch. This is not incompatible, because
1265 previously such lookups could never work.
1267 The situation is now rather anomolous, since one *can* have colons in
1268 ordinary lsearch keys. However, making the change in all cases is
1269 incompatible and would probably break a number of configurations.
1271 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1274 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1275 conversion specification without a maximum field width, thereby enabling
1276 a rogue spamd server to cause a buffer overflow. While nobody in their
1277 right mind would setup Exim to query an untrusted spamd server, an
1278 attacker that gains access to a server running spamd could potentially
1279 exploit this vulnerability to run arbitrary code as the Exim user.
1281 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1282 $primary_hostname instead of what libspf2 thinks the hosts name is.
1284 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1285 a directory entry by the name of the lookup key. Previously, if a
1286 symlink pointed to a non-existing file or a file in a directory that
1287 Exim lacked permissions to read, a lookup for a key matching that
1288 symlink would fail. Now it is enough that a matching directory entry
1289 exists, symlink or not. (Bugzilla 503.)
1291 PH/03 The body_linecount and body_zerocount variables are now exported in the
1294 PH/04 Added the $dnslist_matched variable.
1296 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1297 This means they are set thereafter only if the connection becomes
1300 PH/06 Added the client_condition to authenticators so that some can be skipped
1301 by clients under certain conditions.
1303 PH/07 The error message for a badly-placed control=no_multiline_responses left
1304 "_responses" off the end of the name.
1306 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1308 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1309 (without spaces) instead of just copying the configuration text.
1311 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1313 PH/11 Added $max_received_linelength.
1315 PH/12 Added +ignore_defer and +include_defer to host lists.
1317 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1318 way in which PCRE > 7.0 is built.
1320 PH/14 Implemented queue_only_load_latch.
1322 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1323 MAIL command. The effect was to mangle the value on 64-bit systems.
1325 PH/16 Another patch from the Sieve maintainer.
1327 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1329 PH/18 If a system quota error occurred while trying to create the file for
1330 a maildir delivery, the message "Mailbox is full" was not appended to the
1331 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1332 only to a quota excession during the actual writing of the file.
1334 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1335 characters?) which causes problems in log lines. The DN values are now
1336 passed through string_printing() before being added to log lines.
1338 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1339 and InterBase are left for another time.)
1341 PH/21 Added message_body_newlines option.
1343 PH/22 Guard against possible overflow in moan_check_errorcopy().
1345 PH/23 POSIX allows open() to be a macro; guard against that.
1347 PH/24 If the recipient of an error message contained an @ in the local part
1348 (suitably quoted, of course), incorrect values were put in $domain and
1349 $local_part during the evaluation of errors_copy.
1355 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1356 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1359 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1360 issue a MAIL command.
1362 PH/02 In an ACL statement such as
1364 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1366 if a client was not listed at all, or was listed with a value other than
1367 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1368 the condition was not true (as it should be), so access was not denied.
1369 The bug was that the ! inversion was incorrectly passed on to the second
1370 item. This has been fixed.
1372 PH/03 Added additional dnslists conditions == and =& which are different from
1373 = and & when the dns lookup returns more than one IP address.
1375 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1376 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1378 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1379 FSYNC, which compiles an option called disable_fsync that allows for
1380 bypassing fsync(). The documentation is heavily laced with warnings.
1382 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1384 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1385 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1386 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1387 including adding "make clean"; (3) Added -fPIC when compiling the test
1388 dynamically loaded module, to get rid of a warning.
1390 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1391 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1392 = 0s. The bug is otherwise harmless.
1394 PH/07 There was a bug in the dovecot authenticator such that the value of
1395 $auth1 could be overwritten, and so not correctly preserved, after a
1396 successful authentication. This usually meant that the value preserved by
1397 the server_setid option was incorrect.
1399 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1401 PH/09 Installed PCRE release 7.0.
1403 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1404 run for batched SMTP input. It is now run at the start of every message
1405 in the batch. While fixing this I discovered that the process information
1406 (output by running exiwhat) was not always getting set for -bs and -bS
1407 input. This is fixed, and it now also says "batched" for BSMTP.
1409 PH/11 Added control=no_pipelining.
1411 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1412 patch, slightly modified), and move the expansion of helo_data till after
1413 the connection is made in the smtp transport (so it can use these
1416 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1418 PH/14 Added log_selector = +pid.
1420 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1422 PH/16 Add ${if forany and ${if forall.
1424 PH/17 Added dsn_from option to vary the From: line in DSNs.
1426 PH/18 Flush SMTP output before performing a callout, unless control =
1427 no_callout_flush is set.
1429 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1430 was true (the default) a successful delivery failed to delete the retry
1431 item, thus causing premature timeout of the address. The bug is now
1434 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1436 PH/21 Long custom messages for fakedefer and fakereject are now split up
1437 into multiline reponses in the same way that messages for "deny" and
1438 other ACL rejections are.
1440 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1441 with slight modification.
1443 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1444 draft, changing the syntax and factoring some duplicate code".
1446 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1447 for deliveries of the second and subsequent messages over the same SMTP
1450 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1451 ${reduce, with only minor "tidies".
1453 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1455 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1456 expansion side effects.
1458 PH/27 When a message times out after an over-quota error from an Exim-imposed
1459 quota, the bounce message says "mailbox is full". This message was not
1460 being given when it was a system quota that was exceeded. It now should
1463 MH/03 Made $recipients available in local_scan(). local_scan() already has
1464 better access to the recipient list through recipients_list[], but
1465 $recipients can be useful in postmaster-provided expansion strings.
1467 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1468 in the case of a MAIL command with additional options following the
1469 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1470 were accidentally chopped off.
1472 PH/29 SMTP synchronization checks are implemented when a command is read -
1473 there is a check that no more input is waiting when there shouldn't be
1474 any. However, for some commands, a delay in an ACL can mean that it is
1475 some time before the response is written. In this time, more input might
1476 arrive, invalidly. So now there are extra checks after an ACL has run for
1477 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1478 pipelining has not been advertised.
1480 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1482 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1483 returned in the SMTP response when smtp_return_error_details was false.
1484 This has been fixed.
1486 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1487 instead of the C I/O that was originally supplied, because problems were
1488 reported on Solaris.
1490 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1491 Exim which did not show up earlier: it was assuming that a call to
1492 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1493 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1494 was a macro that became an assignment, so it seemed to work. This has
1495 changed to a proper function call with a void return, hence the compile
1496 error. Exim's code has been fixed.
1498 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1501 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1503 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1505 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1506 because it turns out to be too big or there is a timeout) there is no
1507 "Completed" line in the log. When some messages of this type were
1508 selected by exigrep, they were listed as "not completed". Others were
1509 picked up by some special patterns. I have improved the selection
1510 criteria to be more general.
1512 PH/38 The host_find_failed option in the manualroute router can now be set
1513 to "ignore", to completely ignore a host whose IP address cannot be
1514 found. If all hosts are ignored, the behaviour is controlled by the new
1515 host_all_ignored option.
1517 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1518 homing or because of multiple MX records with /mx) generated more than
1519 one IP address, and the following item turned out to be the local host,
1520 all the secondary addresses of the first item were incorrectly removed
1521 from the list, along with the local host and any following hosts (which
1522 is what is supposed to happen).
1524 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1525 whoever called Exim into the -H file. In the case of the daemon it was
1526 behaving confusingly. When first started, it used values for whoever
1527 started the daemon, but after a SIGHUP it used the Exim user (because it
1528 calls itself on a restart). I have changed the code so that it now always
1531 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1532 message are rejected with the same error (e.g. no authentication or bad
1533 sender address), and a DATA command is nevertheless sent (as can happen
1534 with PIPELINING or a stupid MUA), the error message that was given to the
1535 RCPT commands is included in the rejection of the DATA command. This is
1536 intended to be helpful for MUAs that show only the final error to their
1539 PH/42 Another patch from the Sieve maintainer.
1541 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1542 Eximstats - Fixed some broken HTML links and added missing column headers
1544 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1545 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1547 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1549 PH/43 Yet another patch from the Sieve maintainer.
1551 PH/44 I found a way to check for a TCP/IP connection going away before sending
1552 the response to the final '.' that terminates a message, but only in the
1553 case where the client has not sent further data following the '.'
1554 (unfortunately, this is allowed). However, in many cases there won't be
1555 any further data because there won't be any more messages to send. A call
1556 to select() can be used: if it shows that the input is "ready", there is
1557 either input waiting, or the socket has been closed. An attempt to read
1558 the next input character can distinguish the two cases. Previously, Exim
1559 would have sent an OK response which the client would never have see.
1560 This could lead to message repetition. This fix should cure that, at
1561 least in a lot of common cases.
1563 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1564 advertised in response to EHLO.
1570 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1571 fixed by 4.65/MH/01 (is this a record?) are fixed:
1573 (i) An empty string was always treated as zero by the numeric comparison
1574 operators. This behaviour has been restored.
1576 (ii) It is documented that the numeric comparison operators always treat
1577 their arguments as decimal numbers. This was broken in that numbers
1578 starting with 0 were being interpreted as octal.
1580 While fixing these problems I realized that there was another issue that
1581 hadn't been noticed. Values of message_size_limit (both the global option
1582 and the transport option) were treated as octal if they started with 0.
1583 The documentation was vague. These values are now always treated as
1584 decimal, and I will make that clear in the documentation.
1590 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1591 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1594 MH/01 Don't check that the operands of numeric comparison operators are
1595 integers when their expansion is in "skipping" mode (fixes bug
1596 introduced by 4.64-PH/07).
1598 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1599 child addresses, Exim now panics and dies. Previously, because the count
1600 is held in a short int, deliveries were likely to be lost. As such a
1601 large number of recipients for a single message is ridiculous
1602 (performance will be very, very poor), I have chosen to impose a limit
1603 rather than extend the field.
1609 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1610 leftover -K file (the existence of which was triggered by #402).
1611 While we were at it, introduced process PID as part of the -K
1612 filename. This should rule out race conditions when creating
1615 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1616 processing considerably. Previous code took too long for large mails,
1617 triggering a timeout which in turn triggers #401.
1619 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1620 in the DK code in transports.c. sendfile() is not really portable,
1621 hence the _LINUX specificness.
1623 TF/01 In the add_headers option to the mail command in an Exim filter,
1624 there was a bug that Exim would claim a syntax error in any
1625 header after the first one which had an odd number of characters
1628 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1629 callout verification, Exim cached a "reject" for the entire domain. This
1630 is correct for most verifications, but it is not correct for a recipient
1631 verification with use_sender or use_postmaster set, because in that case
1632 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1633 case of MAIL FROM:<> rejection from other early rejections (e.g.
1634 rejection of HELO). When verifying a recipient using a non-null MAIL
1635 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1636 Whatever the result of the callout, the value of the domain cache is
1637 left unchanged (for any other kind of callout, getting as far as trying
1638 RCPT means that the domain itself is ok).
1640 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1643 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1644 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1645 session, the daemon ignores ECONNECT errors and logs others; it now
1646 ignores EPIPE as well.
1648 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1649 (quoted-printable decoding).
1651 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1652 later the small subsequent patch to fix an introduced bug.
1654 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1656 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1658 PH/08 An error is now given if message_size_limit is specified negative.
1660 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1661 to be given (somewhat) arbitrary names.
1663 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1666 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1667 miscellaneous code fixes
1669 PH/10 Added the log_reject_target ACL modifier to specify where to log
1672 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1673 hostname. This is wrong, because it relates to the incoming message (and
1674 probably the interface on which it is arriving) and not to the outgoing
1675 callout (which could be using a different interface). This has been
1676 changed to use the value of the helo_data option from the smtp transport
1677 instead - this is what is used when a message is actually being sent. If
1678 there is no remote transport (possible with a router that sets up host
1679 addresses), $smtp_active_hostname is used.
1681 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1682 tweaks were necessary in order to get it to work (see also 21 below):
1683 (a) The code assumed that strncpy() returns a negative number on buffer
1684 overflow, which isn't the case. Replaced with Exim's string_format()
1686 (b) There were several signed/unsigned issues. I just did the minimum
1687 hacking in of casts. There is scope for a larger refactoring.
1688 (c) The code used strcasecmp() which is not a standard C function.
1689 Replaced with Exim's strcmpic() function.
1690 (d) The code set only $1; it now sets $auth1 as well.
1691 (e) A simple test gave the error "authentication client didn't specify
1692 service in request". It would seem that Dovecot has changed its
1693 interface. Fortunately there's a specification; I followed it and
1694 changed what the client sends and it appears to be working now.
1696 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1699 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1700 address A is aliased to B and C, where B exists and C does not. Without
1701 -v the output is "A verified" because verification stops after a
1702 successful redirection if more than one address is generated. However,
1703 with -v the child addresses are also verified. Exim was outputting "A
1704 failed to verify" and then showing the successful verification for C,
1705 with its parentage. It now outputs "B failed to verify", showing B's
1706 parentage before showing the successful verification of C.
1708 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1709 look up a TXT record in a specific list after matching in a combined
1712 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1713 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1714 they consult the DNS. I had assumed they would set it the way they
1715 wanted; and indeed my experiments on Linux seem to show that in some
1716 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1717 To be on the safe side, however, I have now made the interface to
1718 host_find_byname() similar to host_find_bydns(), with an argument
1719 containing the DNS resolver options. The host_find_byname() function now
1720 sets these options at its start, just as host_find_bydns() does. The smtp
1721 transport options dns_qualify_single and dns_search_parents are passed to
1722 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1723 of host_find_byname() use the default settings of RES_DEFNAMES
1724 (qualify_single) but not RES_DNSRCH (search_parents).
1726 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1727 spool_read_header() do less string testing, by means of a preliminary
1728 switch on the second character of optional "-foo" lines. (This is
1729 overdue, caused by the large number of possibilities that now exist.
1730 Originally there were few.) While I was there, I also converted the
1731 str(n)cmp tests so they don't re-test the leading "-" and the first
1732 character, in the hope this might squeeze out yet more improvement.
1734 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1735 flag allowing group syntax was set by the header_syntax check but not
1736 turned off, possible causing trouble later; (2) The flag was not being
1737 set at all for the header_verify test, causing "group"-style headers to
1738 be rejected. I have now set it in this case, and also caused header_
1739 verify to ignore an empty address taken from a group. While doing this, I
1740 came across some other cases where the code for allowing group syntax
1741 while scanning a header line wasn't quite right (mostly, not resetting
1742 the flag correctly in the right place). These bugs could have caused
1743 trouble for malformed header lines. I hope it is now all correct.
1745 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1746 with the "reply" argument non-NULL. The code, however (which originally
1747 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1748 but it didn't always do it. This confused somebody who was copying the
1749 code for some other use. I have removed all the tests.
1751 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1752 feature that was used to support insecure browsers during the U.S. crypto
1753 embargo. It requires special client support, and Exim is probably the
1754 only MTA that supported it -- and would never use it because real RSA is
1755 always available. This code has been removed, because it had the bad
1756 effect of slowing Exim down by computing (never used) parameters for the
1757 RSA_EXPORT functionality.
1759 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1760 authenticator to fail if there's a tab character in the incoming data
1761 (there should never be unless someone is messing about, as it's supposed
1762 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1763 if the connection is using TLS or if the remote IP is the same as the
1764 local IP, and the "valid-client-cert option" if a client certificate has
1767 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1768 authenticators. This can be used for authorization after authentication
1769 succeeds. (In the case of plaintext, it servers for both authentication
1772 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1773 if any retry times were supplied.
1775 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1776 connection, where there is no client IP address to check. In this
1777 situation, the verify now always succeeds.
1779 PH/25 Applied John Jetmore's -Mset patch.
1781 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1783 PH/27 In a string expansion for a processed (not raw) header when multiple
1784 headers of the same name were present, leading whitespace was being
1785 removed from all of them, but trailing whitespace was being removed only
1786 from the last one. Now trailing whitespace is removed from each header
1787 before concatenation. Completely empty headers in a concatenation (as
1788 before) are ignored.
1790 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1791 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1793 PH/29 [Removed. This was a change that I later backed out, and forgot to
1794 correct the ChangeLog entry (that I had efficiently created) before
1795 committing the later change.]
1797 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1798 address errors (4xx response to RCPT) over the same connection as other
1799 messages routed to the same hosts. Such deliveries are always "forced",
1800 so retry times are not inspected. This resulted in far too many retries
1801 for the affected addresses. The effect occurred only when there were more
1802 hosts than the hosts_max_try setting in the smtp transport when it had
1803 the 4xx errors. Those hosts that it had tried were not added to the list
1804 of hosts for which the message was waiting, so if all were tried, there
1805 was no problem. Two fixes have been applied:
1807 (i) If there are any address or message errors in an SMTP delivery, none
1808 of the hosts (tried or untried) are now added to the list of hosts
1809 for which the message is waiting, so the message should not be a
1810 candidate for sending over the same connection that was used for a
1811 successful delivery of some other message. This seems entirely
1812 reasonable: after all the message is NOT "waiting for some host".
1813 This is so "obvious" that I'm not sure why it wasn't done
1814 previously. Hope I haven't missed anything, but it can't do any
1815 harm, as the worst effect is to miss an optimization.
1817 (ii) If, despite (i), such a delivery is accidentally attempted, the
1818 routing retry time is respected, so at least it doesn't keep
1819 hammering the server.
1821 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1822 in ${readsocket because some servers need this prod.
1824 PH/32 Added some extra debug output when updating a wait-xxx database.
1826 PH/33 The hint "could be header name not terminated by colon", which has been
1827 given for certain expansion errors for a long time, was not being given
1828 for the ${if def:h_colon_omitted{... case.
1830 PH/34 The spec says: "With one important exception, whenever a domain list is
1831 being scanned, $domain contains the subject domain." There was at least
1832 one case where this was not true.
1834 PH/35 The error "getsockname() failed: connection reset by peer" was being
1835 written to the panic log as well as the main log, but it isn't really
1836 panic-worthy as it just means the connection died rather early on. I have
1837 removed the panic log writing for the ECONNRESET error when getsockname()
1840 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1841 runs only) independently of the message's sender address. This meant
1842 that, if the 4xx error was in fact related to the sender, a different
1843 message to the same recipient with a different sender could confuse
1844 things. In particualar, this can happen when sending to a greylisting
1845 server, but other circumstances could also provoke similar problems.
1846 I have changed the default so that the retry time for these errors is now
1847 based a combination of the sender and recipient addresses. This change
1848 can be overridden by setting address_retry_include_sender=false in the
1851 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1852 remote server are returned as part of bounce messages. This was not
1853 happening for LMTP over a pipe (the lmtp transport), but now it is the
1854 same for both kinds of LMTP.
1856 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1857 in header lines that were in fact CNAMEs. This is no longer the case.
1859 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1860 and queue runs started by the daemon processed all messages. This has
1861 been fixed so that -R and -S can now usefully be given with -q<time>.
1863 PH/40 Import PCRE release 6.7 (fixes some bugs).
1865 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1867 PH/42 Give an error if -q is specified more than once.
1869 PH/43 Renamed the variables $interface_address and $interface_port as
1870 $received_ip_address and $received_port, to make it clear that these
1871 values apply to message reception, and not to the outgoing interface when
1872 a message is delivered. (The old names remain recognized, of course.)
1874 PH/44 There was no timeout on the connect() call when using a Unix domain
1875 socket in the ${readsocket expansion. There now is.
1877 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1878 be meaningful with "accept".
1880 SC/01 Eximstats V1.43
1881 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1883 SC/02 Eximstats V1.44
1884 Use a glob alias rather than an array ref in the generated
1885 parser. This improves both readability and performance.
1887 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1888 Collect SpamAssassin and rejection statistics.
1889 Don't display local sender or destination tables unless
1890 there is data to show.
1891 Added average volumes into the top table text output.
1893 SC/04 Eximstats V1.46
1894 Collect data on the number of addresses (recipients)
1895 as well as the number of messages.
1897 SC/05 Eximstats V1.47
1898 Added 'Message too big' to the list of mail rejection
1899 reasons (thanks to Marco Gaiarin).
1901 SC/06 Eximstats V1.48
1902 Mainlog lines which have GMT offsets and are too short to
1903 have a flag are now skipped.
1905 SC/07 Eximstats V1.49 (Alain Williams)
1906 Added the -emptyok flag.
1908 SC/08 Eximstats V1.50
1909 Fixes for obtaining the IP address from reject messages.
1911 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1912 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1913 whitesspace changes from 4.64-PH/27
1915 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1918 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1919 are found, allow negative numbers in numeric criteria)
1921 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1923 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1924 to match changes made in 4.64-PH/43
1926 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1928 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1929 "domain" is actually an IP address, and force a failure. This locks out
1930 those revolvers/nameservers that support "A-for-A" lookups, in
1931 contravention of the specifications.
1933 PH/48 When a host name was looked up from an IP address, and the subsequent
1934 forward lookup of the name timed out, the host name was left in
1935 $sender_host_name, contrary to the specification.
1937 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1938 restricted to single-key lookups, Exim was not diagnosing an error if
1939 * or *@ was used with a query-style lookup.
1941 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1943 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1944 long ago, but noone interested enough thought of it. Let's just say that
1945 the "1.1" means that there are some new functions that weren't there at
1946 some point in the past.
1948 PH/51 Error processing for expansion failure of helo_data from an smtp
1949 transport during callout processing was broken.
1951 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1952 tested/used via the -bh/-bhc/-bs options.
1954 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1955 bug, fixed in subsequent PCRE releases).
1957 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1958 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1960 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1966 SC/01 Use a glob alias rather than an array ref in eximstats generated
1967 parser. This improves both readability and performance.
1969 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1970 Don't display local sender or destination tables in eximstats unless
1971 there is data to show.
1972 Added average volumes into the eximstats top table text output.
1974 SC/03 Collect data on the number of addresses (recipients) as well
1975 as the number of messages in eximstats.
1977 TF/01 Correct an error in the documentation for the redirect router. Exim
1978 does (usually) call initgroups() when daemonizing.
1980 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1981 with consistent privilege compared to when running as a daemon.
1983 TF/03 Note in the spec that $authenticated_id is not set for local
1984 submissions from trusted users.
1986 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1987 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1989 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1990 by adding some example configuration directives to the default
1991 configuration file. A little bit of work is required to uncomment the
1992 directives and define how usernames and passwords are checked, but
1993 there is now a framework to start from.
1995 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1996 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1997 without this. I don't know how relevant this is to other LDAP libraries.
1999 PH/02 Add the verb name to the "unknown ACL verb" error.
2001 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2003 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2005 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2006 directory not expanded when it should be if an expanded home directory
2007 was set for the address (which is overridden by the transport).
2009 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2012 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2013 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2014 because it is too late at that time, and has no effect.
2016 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2017 security issue with \' (bugzilla #107). I could not use the
2018 PQescapeStringConn() function, because it needs a PGconn value as one of
2021 PH/08 When testing addresses using -bt, indicate those final addresses that
2022 are duplicates that would not cause an additional delivery. At least one
2023 person was confused, thinking that -bt output corresponded to deliveries.
2024 (Suppressing duplicates isn't a good idea as you lose the information
2025 about possibly different redirections that led to the duplicates.)
2027 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2028 systems where poll() doesn't work, in particular OS X.
2030 PH/10 Added more information to debugging output for retry time not reached.
2032 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2033 operations in malware.c.
2035 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2038 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2039 syslog_duplication set false, log lines that would normally be written
2040 both the the main log and to the reject log were not written to syslog at
2043 PH/14 In the default configuration, change the use of "message" in ACL warn
2044 statements to "add_header".
2046 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2047 not followed by a command (e.g. "seen endif").
2049 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2050 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2053 PH/17 Added extra conditions to the default value of delay_warning_condition
2057 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2058 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2059 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2062 The Auto-Submitted: and various List- headers are standardised, whereas I
2063 don't think Precedence: ever was.
2065 PH/18 Refactored debugging code in route_finduser() to show more information,
2066 in particular, the error code if getpwnam() issues one.
2068 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2069 This is apparently needed in addition to the PH/07 change above to avoid
2070 any possible encoding problems.
2072 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2073 but not after initializing Perl.
2075 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2076 output them only if debugging. By default they are written stderr,
2077 apparently, which is not desirable.
2079 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2082 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2085 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2087 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2088 authenticated or an ident call has been made. Suppress the default
2089 values for $authenticated_id and $authenticated_sender (but permit -oMai
2090 and -oMas) when testing with -bh.
2092 PH/24 Re-jigged the order of the tests in the default configuration so that the
2093 tests for valid domains and recipients precede the DNS black list and CSA
2094 tests, on the grounds that those ones are more expensive.
2096 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2097 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2098 This bug exists in every version of Exim that I still have, right back to
2101 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2102 However, an attempt to turn on case-sensitivity in a regex key by
2103 including (?-i) didn't work because the subject string was already
2104 lowercased, and the effects were non-intuitive. It turns out that a
2105 one-line patch can be used to allow (?-i) to work as expected.
2111 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2112 other effects) broke the use of negated acl sub-conditions.
2114 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2117 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2118 "Deny" causes Exim to reject the incoming connection with a 554 error.
2119 Unfortunately, if there is a major crisis, such as a disk failure,
2120 tcp-wrappers gives "deny", whereas what one would like would be some
2121 kind of temporary error. A kludge has been added to help with this.
2122 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2123 554 error is used if errno is still zero or contains ENOENT (which occurs
2124 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2127 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2129 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2130 errors. Otherwise a message that provokes a temporary error (when other
2131 messages do not) can cause a whole host to time out.
2133 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2134 addresses were routed directly to files or pipes from a redirect router.
2135 File deliveries just didn't batch; pipe deliveries might have suffered
2138 PH/06 A failure to get a lock for a hints database would erroneously always say
2139 "Failed to get write lock", even when it was really a read lock.
2141 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2142 of 0600. This has been changed to use the value of the lockfile_mode
2143 option (which defaults to 0600).
2145 PH/08 Applied small patch from the Sieve maintainer.
2147 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2148 folder from quota calculations, a direct delivery into this folder messed
2149 up the contents of the maildirsize file. This was because the regex was
2150 used only to exclude .Trash (or whatever) when the size of the mailbox
2151 was calculated. There was no check that a delivery was happening into an
2152 excluded directory. This bug has been fixed by ignoring all quota
2153 processing for deliveries into excluded directories.
2155 PH/10 Added the maildirfolder_create_regex option to appendfile.
2161 PH/01 The code for finding all the local interface addresses on a FreeBSD
2162 system running IPv6 was broken. This may well have applied to all BSD
2163 systems, as well as to others that have similar system calls. The broken
2164 code found IPv4 interfaces correctly, but gave incorrect values for the
2165 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2166 that it would not match correctly against @[] and not recognize the IPv6
2169 PH/02 The ipliteral router was not recognizing addresses of the form user@
2170 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2172 PH/03 Added disable_ipv6.
2174 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2175 decoded form, because it is most often used to construct To: headers
2176 lines in autoreplies, and the decoded form may well be syntactically
2177 invalid. However, $reply_address has leading white space removed, and all
2178 newlines turned into spaces so that the autoreply transport does not
2181 PH/05 If group was specified without a user on a router, and no group or user
2182 was specified on a transport, the group from the router was ignored.
2184 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2185 for visible compile-time settings that can be used to change these
2186 numbers, for those that want even more. Backwards compatibility with old
2187 spool files has been maintained. However, going back to a previous Exim
2188 release will lost any variables that are in spool files.
2190 PH/07 Two small changes when running in the test harness: increase delay when
2191 passing a TCP/IP connection to a new process, in case the original
2192 process has to generate a bounce, and remove special handling of
2193 127.0.0.2 (sic), which is no longer necessary.
2195 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2196 be the same on different OS.
2198 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2201 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2202 whether --show-vars was specified or not
2204 JJ/02 exipick: Added support for new ACL variable spool format introduced
2207 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2208 syntactically invalid From: or Reply-to: line, and a filter used this to
2209 generate an autoreply, and therefore failed to obtain an address for the
2210 autoreply, Exim could try to deliver to a non-existent relative file
2211 name, causing unrelated and misleading errors. What now happens is that
2212 it logs this as a hard delivery error, but does not attempt to create a
2215 PH/11 The exinext utility has a -C option for testing purposes, but although
2216 the given file was scanned by exinext itself; it wasn't being passed on
2217 when Exim was called.
2219 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2220 an end-of-file indication when reading a command response.
2222 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2223 compiled. In many other places in Exim, IPv6 addresses are always
2224 recognized, so I have changed this. It also means that IPv4 domain
2225 literals of the form [IPV4:n.n.n.n] are now always recognized.
2227 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2228 used if the router is not running as root, for example, when verifying at
2229 ACL time, or when using -bh. The debugging output from this situation was
2230 non-existent - all you got was a failure to exec. I have made two
2233 (a) Failures to set uid/gid, the current directory, or a process leader
2234 in a subprocess such as that created by queryprogram now generate
2235 suitable debugging ouput when -d is set.
2237 (b) The queryprogram router detects when it is not running as root,
2238 outputs suitable debugging information if -d is set, and then runs
2239 the subprocess without attempting to change uid/gid.
2241 PH/15 Minor change to Makefile for building test_host (undocumented testing
2244 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2245 additional section of a DNS packet that returns MX or SRV records.
2246 Instead, it always explicitly searches for A/AAAA records. This avoids
2247 major problems that occur when a DNS server includes only records of one
2248 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2249 fixed another bug: if SRV records were looked up and the corresponding
2250 address records were *not* found in the additional section, the port
2251 values from the SRV records were lost.
2253 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2254 using the correct key (the original address) when searching the retry
2255 rules in order to find which one to use for generating the retry hint.
2257 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2258 adding the default one. Similarly, if it contains a Reply-To: header, the
2259 errors_reply_to option, if set, is not used.
2261 PH/19 When calculating a retry time, Exim used to measure the "time since
2262 failure" by looking at the "first failed" field in the retry record. Now
2263 it does not use this if it is later than than the arrival time of the
2264 message. Instead it uses the arrival time. This makes for better
2265 behaviour in cases where some deliveries succeed, thus re-setting the
2266 "first failed" field. An example is a quota failure for a huge message
2267 when small messages continue to be delivered. Without this change, the
2268 "time since failure" will always be short, possible causing more frequent
2269 delivery attempts for the huge message than are intended.
2270 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2272 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2273 $1, $2, $3) because the numerical variables can be reset during some
2274 expansion items (e.g. "match"), thereby losing the authentication data.
2276 PH/21 Make -bV show the size of off_t variables so that the test suite can
2277 decide whether to run tests for quotas > 2G.
2279 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2280 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2281 filecount value is greater than 2G or if a quota value is greater than 2G
2282 on a system where the size of off_t is not greater than 4, a panic error
2285 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2286 never match. The debug and -bh output now contains an explicit error
2287 message indicating a malformed IPv4 address or mask.
2289 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2290 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2291 PH/23 above applies.
2293 PH/25 Do not write to syslog when running in the test harness. The only
2294 occasion when this arises is a failure to open the main or panic logs
2295 (for which there is an explicit test).
2297 PH/26 Added the /no_tell option to "control=freeze".
2299 PH/27 If a host name lookup failed very early in a connection, for example, if
2300 the IP address matched host_lookup and the reverse lookup yielded a name
2301 that did not have a forward lookup, an error message of the form "no IP
2302 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2303 could be logged. Now it outputs the IP address instead of "NULL".
2305 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2306 allows the sender and the authenticated sender to be set when
2307 submitting a message from within Exim. Since child_open_exim() is
2308 documented for local_scan(), the new function should be too.
2310 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2311 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2312 results in an empty string is now treated as unset.
2314 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2316 PH/31 Added sender_verify_fail logging option.
2318 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2319 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2320 tidied the source and removed it altogether.
2322 PH/33 When a queue run was abandoned because the load average was too high, a
2323 log line was always written; now it is written only if the queue_run log
2324 selector is set. In addition, the log line for abandonment now contains
2325 information about the queue run such as the pid. This is always present
2326 in "start" and "stop" lines but was omitted from the "abandon" line.
2328 PH/34 Omit spaces between a header name and the colon in the error message that
2329 is given when verify = headers_syntax fails (if there are lots of them,
2330 the message gets confusing).
2332 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2333 names, as there are now some PTR records that contain slashes. This check
2334 is only to protect against broken name servers that fall over on strange
2335 characters, so the fact that it applies to all lookups doesn't matter.
2337 PH/36 Now that the new test suite is complete, we can remove some of the
2338 special code in Exim that was needed for the old test suite. For example,
2339 sorting DNS records because real resolvers return them in an arbitrary
2340 order. The new test suite's fake resolver always returns records in the
2343 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2344 bounces) except when queue_only is set, to avoid logging races between
2345 the different processes.
2347 PH/38 Panic-die if .include specifies a non-absolute path.
2349 PH/39 A tweak to the "H" retry rule from its user.
2351 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2352 a label. They prevented compilation on older perls.
2354 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2355 a warning to be raised on newish perls.
2357 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2358 on queue. Changes to match documented behaviour of showing count of
2359 messages matching specified criteria.
2361 PH/40 Changed the default ident timeout from 30s to 5s.
2363 PH/41 Added support for the use of login_cap features, on those BSD systems
2364 that have them, for controlling the resources used by pipe deliveries.
2366 PH/42 The content-scanning code uses fopen() to create files in which to put
2367 message data. Previously it was not paying any attention to the mode of
2368 the files. Exim runs with umask(0) because the rest of the code creates
2369 files with open(), and sets the required mode explicitly. Thus, these
2370 files were ending up world-writeable. This was not a big issue, because,
2371 being within the spool directory, they were not world-accessible. I have
2372 created a function called modefopen, which takes an additional mode
2373 argument. It sets umask(777), creates the file, chmods it to the required
2374 mode, then resets the umask. All the relevant calls to fopen() in the
2375 content scanning code have been changed to use this function.
2377 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2378 to 24 hours. This avoids potential overflow problems when processing G
2379 and H retry rules. I suspect nobody ever tinkers with this value.
2381 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2383 PH/45 When the plaintext authenticator is running as a client, the server's
2384 challenges are checked to ensure they are valid base64 strings. By
2385 default, the authentication attempt is cancelled if an invalid string is
2386 received. Setting client_ignore_invalid_base64 true ignores these errors.
2387 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2388 they are received. Thus, the responses can be made to depend on the
2389 challenges. If an invalid string is ignored, an empty string is placed in
2392 PH/46 Messages that are created by the autoreply transport now contains a
2393 References: header, in accordance with RFCs 2822 and 3834.
2395 PH/47 Added authenticated_sender_force to the smtp transport.
2397 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2399 PH/49 Installed latest patch from the Sieve maintainer.
2401 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2402 was also set, the appendfile transport was unnecessarily scanning a
2403 directory of message files (e.g. for maildir delivery) to find the count
2404 of files (along with the size), even though it did not need this
2405 information. It now does the scan only if it needs to find either the
2406 size of the count of files.
2408 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2410 PH/52 Two bugs concerned with error handling when the smtp transport is
2413 (i) Exim was not creating retry information for temporary errors given
2414 for individual recipients after the DATA command when the smtp transport
2415 was used in LMTP mode. This meant that they could be retried too
2416 frequently, and not timed out correctly.
2418 (ii) Exim was setting the flag that allows error details to be returned
2419 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2420 recipients that were returned after the DATA command.
2422 PH/53 This is related to PH/52, but is more general: for any failing address,
2423 when detailed error information was permitted to be returned to the
2424 sender, but the error was temporary, then after the final timeout, only
2425 "retry timeout exceeded" was returned. Now it returns the full error as
2426 well as "retry timeout exceeded".
2428 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2429 do this, and (what is worse) MTAs that accept it.
2431 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2432 will now be deprecated.
2434 PH/56 New os.c-cygwin from the Cygwin maintainer.
2436 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2437 formats (previously only available in exim formats via -bpr, -bpru,
2438 and -bpra. Now also available in native and exiqgrep formats)
2440 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2441 with very large, slow to parse queues
2443 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2445 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2447 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2448 responses to authentication challenges, though it was showing the
2449 challenges; (ii) I've removed the CR characters from the debug output for
2452 PH/58 Allow for the insertion of a newline as well as a space when a string
2453 is turned into more than one encoded-word during RFC 2047 encoding. The
2454 Sieve code now uses this.
2456 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2457 data_4xx, lost_connection, tls_required.
2459 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2460 message was being sent as an SMTP response.
2462 PH/61 Add -l and -k options to exicyclog.
2464 PH/62 When verifying, if an address was redirected to one new address, so that
2465 verification continued, and the new address failed or deferred after
2466 having set something in $address_data, the value of $address_data was not
2467 passed back to the ACL. This was different to the case when no
2468 redirection occurred. The value is now passed back in both cases.
2470 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2471 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2472 use login_cap.h, so on its own it isn't the distinguishing feature. The
2473 new name refers directly to the setclassresources() function.
2475 PH/65 Added configuration files for NetBSD3.
2477 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2479 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2480 is preferred over IPv4.
2482 PH/68 The bounce_return_message and bounce_return_body options were not being
2483 honoured for bounces generated during the reception of non-SMTP messages.
2484 In particular, this applied to messages rejected by the ACL. This bug has
2485 been fixed. However, if bounce_return_message is true and bounce_return_
2486 body is false, the headers that are returned for a non-SMTP message
2487 include only those that have been read before the error was detected.
2488 (In the case of an ACL rejection, they have all been read.)
2490 PH/69 The HTML version of the specification is now built in a directory called
2491 spec_html instead of spec.html, because the latter looks like a path with
2492 a MIME-type, and this confuses some software.
2494 PH/70 Catch two compiler warnings in sieve.c.
2496 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2497 function verify_get_ident() calls ip_connect() to connect a socket, but
2498 if the "connect()" function timed out, ip_connect() used to close the
2499 socket. However, verify_get_ident() also closes the socket later, and in
2500 between Exim writes to the log, which may get opened at this point. When
2501 the socket was closed in ip_connect(), the log could get the same file
2502 descriptor number as the socket. This naturally causes chaos. The fix is
2503 not to close the socket in ip_connect(); the socket should be closed by
2504 the function that creates it. There was only one place in the code where
2505 this was missing, in the iplookup router, which I don't think anybody now
2506 uses, but I've fixed it anyway.
2508 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2509 well as to direct DNS lookups. Otherwise the handling of names in host
2510 lists is inconsistent and therefore confusing.
2516 PH/01 Two changes to the default runtime configuration:
2518 (1) Move the checks for relay_from_hosts and authenticated clients from
2519 after to before the (commented out) DNS black list checks.
2521 (2) Add control=submission to the relay_from_hosts and authenticated
2522 clients checks, on the grounds that messages accepted by these
2523 statements are most likely to be submissions.
2525 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2527 (1) Generate an error if the third argument for the ${prvs expansion is
2530 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2533 (3) Reset the variables that are obtained from the first argument of
2534 ${prvscheck and used in the second argument before leaving the code,
2535 because their memory is reclaimed, so using them afterwards may do
2538 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2539 one (it's much easier than Tom thought :-).
2541 (5) Because of (4), we can now allow for the use of $prvscheck_result
2542 inside the third argument.
2544 PH/03 For some reason, the default setting of PATH when running a command from
2545 a pipe transport was just "/usr/bin". I have changed it to
2548 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2549 anything to be listed in the output from -bV.
2551 PH/05 When a filter generated an autoreply, the entire To: header line was
2552 quoted in the delivery log line, like this:
2554 => >A.N.Other <ano@some.domain> <original@ddress> ...
2556 This has been changed so that it extracts the operative address. There
2557 may be more than one such address. If so, they are comma-separated, like
2560 => >ano@some.domain,ona@other.domain <original@ddress> ...
2562 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2563 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2564 not being looked up in the rDNS to get a host name, Exim was showing the
2565 IP address twice in Received: lines, even though the IP addresses were
2566 identical. For example:
2568 Received: from [1.2.3.4] (helo=[1.2.3.4])
2570 However, if the real host name was known, it was omitting the HELO data
2571 if it matched the actual IP address. This has been tidied up so that it
2572 doesn't show the same IP address twice.
2574 PH/07 When both +timestamp and +memory debugging was on, the value given by
2575 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2576 called by the debug printing, thereby overwriting the timestamp buffer.
2577 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2579 PH/08 When the original message was included in an autoreply transport, it
2580 always said "this is a copy of the message, including all the headers",
2581 even if body_only or headers_only was set. It now gives an appropriate
2584 PH/09 Applied a patch from the Sieve maintainer which:
2586 o fixes some comments
2587 o adds the (disabled) notify extension core
2588 o adds some debug output for the result of if/elsif tests
2589 o points to the current vacation draft in the documentation
2590 and documents the missing references header update
2594 o fixes a bug in processing the envelope test (when testing
2595 multiple envelope elements, the last element determinted the
2598 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2599 Electronic Mail") by including:
2601 Auto-submitted: auto-generated
2603 in the messages that it generates (bounce messages and others, such as
2604 warnings). In the case of bounce messages for non-SMTP mesages, there was
2605 also a typo: it was using "Auto_submitted" (underscore instead of
2606 hyphen). Since every message generated by Exim is necessarily in response
2607 to another message, thes have all been changed to:
2609 Auto-Submitted: auto-replied
2611 in accordance with these statements in the RFC:
2613 The auto-replied keyword:
2615 - SHOULD be used on messages sent in direct response to another
2616 message by an automatic process,
2618 - MUST NOT be used on manually-generated messages,
2620 - MAY be used on Delivery Status Notifications (DSNs) and Message
2621 Disposition Notifications (MDNs),
2623 - MUST NOT be used on messages generated by automatic or periodic
2624 processes, except for messages which are automatic responses to
2627 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2628 to the default Received: header definition.
2630 PH/12 Added log selector acl_warn_skipped (default on).
2632 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2633 variables because (a) they are in the wrong storage pool and (b) even if
2634 they were copied, it wouldn't work properly because of the caching.
2636 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2637 checking when decoding. Apparently there are clients that generate
2638 overlong encoded strings. Why am I not surprised?
2640 PH/15 If the first argument of "${if match_address" was not empty, but did not
2641 contain an "@" character, Exim crashed. Now it writes a panic log message
2642 and treats the condition as false.
2644 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2646 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2647 extension "envelope-auth". The code is finished and in agreement with
2648 other implementations, but there is no documentation so far and in fact,
2649 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2650 not changing the active code.
2652 Print executed "if" and "elsif" statements when debugging is used. This
2653 helps a great deal to understand what a filter does.
2655 Document more things not specified clearly in RFC3028. I had all this
2656 sorted out, when out of a sudden new issues came to my mind. Oops."
2658 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2661 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2662 canonical form (as documented). However, after a host name lookup from
2663 the IP address, check_host() was doing a simple string comparison with
2664 addresses acquired from the DNS when checking that the found name did
2665 have the original IP as one of its addresses. Since any found IPv6
2666 addresses are likely to be in abbreviated form, the comparison could
2667 fail. Luckily, there already exists a function for doing the comparison
2668 by converting both addresses to binary, so now that is used instead of
2669 the text comparison.
2671 PH/20 There was another similar case to PH/19, when a complete host name was
2672 given in a host list; looking up its IP address could give an abbreviated
2673 form, whereas the current host's name might or might not be abbreviated.
2674 The same fix has been applied.
2680 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2681 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2684 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2685 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2687 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2689 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2690 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2691 Also, for "Auto-Submitted", treat anything other than "no" as
2692 non-personal, in accordance with RFC 3834. (Previously it treated
2693 anything starting "auto-" as non-personal.)
2695 TF/01 The control=submission/name=... option had a problem with syntax
2696 errors if the name included a slash character. The /name= option
2697 now slurps the rest of the string, so it can include any characters
2698 but it must come last in the list of options (after /sender_retain
2701 PH/05 Some modifications to the interface to the fake nameserver for the new
2709 TK/01 Added the "success_on_redirect" address verification option. See
2710 NewStuff for rationale and an example.
2712 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2714 PH/02 Patch to exigrep to allow it to work on syslog lines.
2716 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2717 fread() to skip over the body file's header line, because in Cygwin the
2718 header line is locked and is inaccessible.
2720 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2721 co-exist for some time) to make it clear that it is the Exim ID that is
2722 referenced, not the Message-ID: header line.
2724 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2725 string_format() function, because snprintf() does not exist on all
2728 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2729 ${stat: expansion item.
2731 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2732 protocol synchronization error", to keep the pedants happy.
2734 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2735 well as for IRIX systems, when gcc is being used. See the host.c source
2738 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2740 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2743 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2744 transport and to the smtp transport in LMTP mode.
2746 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2748 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2749 run a filter in a subprocess. This could lead to confusion in subsequent
2750 lookups in the parent process. There should also be a search_tidyup() at
2751 the end of the subprocess.
2753 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2754 only if the host matched helo_try_verify_hosts, which caused the
2755 verification to occur when the EHLO/HELO command was issued. The ACL just
2756 tested the remembered result. Now, if a previous verification attempt has
2757 not happened, "verify = helo" does it there and then.
2759 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2761 TK/03 Fix log output including CR from clamd.
2763 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2764 exist provoked a memory error which could cause a segfault.
2766 PH/15 Installed PCRE 6.2
2768 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2770 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2771 of the problem. Specifically, suggested +O2 rather than +O1 for the
2774 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2776 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2777 record, the port was not correctly logged when the outgoing_port log
2778 selector was set (it logged the transort's default port).
2780 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2781 fallback_hosts, and "hosts" in the smtp transport.
2783 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2784 host errors such as "Connection refused".
2786 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2787 authentication with radiusclient 0.4.9:
2789 - Error returned from rc_read_config was caught wrongly
2790 - Username/password not passed on to radius server due to wrong length.
2792 The presumption is that some radiusclient API changes for 4.51/PH/17
2793 were not taken care of correctly. The code is still untested by me (my
2794 Linux distribution still has 0.3.2 of radiusclient), but it was
2795 contributed by a Radius user.
2797 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2798 expanding the "port" option of the smtp transport.
2800 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2801 while reading a MIME header. Thanks to Tom Hughes for a patch.
2803 PH/24 Include config.h inside local_scan.h so that configuration settings are
2806 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2807 that in an ACL for RCPT (for example), you can examine exactly what was
2810 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2811 commands, but it was not correctly comparing the address with the actual
2812 client host address. Thus, it would show the EHLO address in Received:
2813 header lines when this was not necessary.
2815 PH/27 Added the % operator to ${eval:}.
2817 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2818 it should be ignoring failures (because with -C, for example, it has lost
2819 privilege). It wasn't ignoring creation failures other than "already
2822 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2825 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2826 because an input line was too long, either on its own, or by virtue of
2827 too many continuations, the temporary file was not being removed, and the
2828 return code was incorrect.
2830 PH/31 Missing "BOOL" in function definition in filtertest.c.
2832 PH/32 Applied Sieve patches from the maintainer.
2834 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2836 PH/33 Added "verify = not_blind".
2838 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2839 Local/Makefile (with some defaults set). These are used in built scripts
2840 such as exicyclog, but they have never been used in the exim_install
2841 script (though there are many overriding facilities there). I have
2842 arranged that the exim_install script now takes note of these two
2845 PH/35 Installed configuration files for Dragonfly.
2847 PH/36 When a locally submitted message by a trusted user did not contain a
2848 From: header, and the sender address was obtained from -f or from an SMTP
2849 MAIL command, and the trusted user did not use -F to supply a sender
2850 name, $originator_name was incorrectly used when constructing a From:
2851 header. Furthermore, $originator_name was used for submission mode
2852 messages from external hosts without From: headers in a similar way,
2853 which is clearly wrong.
2855 PH/37 Added control=suppress_local_fixups.
2857 PH/38 When log_selector = +received_sender was set, and the addition of the
2858 sender made the log line's construction buffer exactly full, or one byte
2859 less than full, an overflow happened when the terminating "\n" was
2862 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2863 when the result of a list match is failure because a DNS lookup failed.
2865 PH/40 RM_COMMAND is now used in the building process.
2867 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2868 the "build-* directories that it finds.
2870 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2871 address in a domain literal was a prefix of an interface address.
2873 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2874 when verifying a sender address, unless rewrite_headers is false.
2876 PH/44 Wrote a long comment about why errors_to addresses are verified as
2877 recipients, not senders.
2879 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2880 the ratelimit ACL was added.
2882 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2884 PH/46 Added extra information about PostgreSQL errors to the error string.
2886 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2887 avoiding the need to install special zones in a real server. This is
2888 backwards compatible; if it can't find the fake resolver, it drops back.
2889 Thus, both old and new test suites can be run.
2891 TF/02 Added util/ratelimit.pl
2893 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2894 clock is set back in time.
2896 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2897 Candler <B.Candler@pobox.com>.
2899 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2900 OK for addresses that are the result of redirections.
2902 PH/48 A number of further additions for the benefit of the new test suite,
2903 including a fake gethostbyname() that interfaces to the fake DNS resolver
2906 TF/06 The fix for widen_domains has also been applied to qualify_single and
2907 search_parents which are the other dnslookup options that can cause
2910 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2913 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2915 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2916 service" if it encountered an MX record with an empty target hostname.
2917 The message is now "an MX or SRV record indicated no SMTP service".
2919 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2920 if the DNS of the sending site is misconfigured. This is quite a
2921 common situation. This change restores the behaviour of treating a
2922 helo verification defer as a failure.
2924 PH/51 If self=fail was set on a router, the bounce message did not include the
2925 actual error message.
2931 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2933 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2934 command itself ended in error, the underlying message about the transport
2935 filter timeout was being overwritten with the pipe command error. Now the
2936 underlying error message should be appended to the second error message.
2938 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2940 PH/02 Reduce the amount of output that "make" produces by default. Full output
2941 can still be requested.
2943 PH/03 The warning log line about a condition test deferring for a "warn" verb
2944 was being output only once per connection, rather than after each
2945 occurrence (because it was using the same function as for successful
2946 "warn" verbs). This seems wrong, so I have changed it.
2948 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2949 it should not have, which might have caused a crash in the right
2950 circumstances, but probably never did.
2952 PH/04 Installed a modified version of Tony Finch's patch to make submission
2953 mode fix the return path as well as the Sender: header line, and to
2954 add a /name= option so that you can make the user's friendly name appear
2957 TF/03 Added the control = fakedefer ACL modifier.
2959 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2960 Mark Lowes for thorough testing.
2962 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2964 TK/03 Merged latest SRS patch from Miles Wilton.
2966 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2967 with the definition in sysexits.h (which is #included earlier).
2968 Fortunately, Exim does not actually use EX_OK. The code used to try to
2969 preserve the sysexits.h value, by assumimg that macro definitions were
2970 scanned for macro replacements. I have been disabused of this notion,
2971 so now the code just undefines EX_OK before #including unistd.h.
2973 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2974 in the smtp transport. When a block could not be written in a single
2975 write() function, the timeout was being re-applied to each part-write.
2976 This seems wrong - if the receiver was accepting one byte at a time it
2977 would take for ever. The timeout is now adjusted when this happens. It
2978 doesn't have to be particularly precise.
2980 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2981 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2983 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2984 without a domain if the check to <postmaster@domain> fails.
2986 SC/01 Eximstats: added -xls and the ability to specify output files
2987 (patch written by Frank Heydlauf).
2989 SC/02 Eximstats: use FileHandles for outputing results.
2991 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2993 SC/04 Eximstats: fixed display of large numbers with -nvr option
2995 SC/05 Eximstats: fixed merging of reports with empty tables.
2997 SC/06 Eximstats: added the -include_original_destination flag
2999 SC/07 Eximstats: removed tabs and trailing whitespace.
3001 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3003 TK/06 MBOX spool code: Add real "From " MBOX separator line
3004 so the .eml file is really in mbox format (even though
3005 most programs do not really care). Patch from Alex Miller.
3007 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3008 The latter is generated from $received_to and is only set if the
3009 message has one envelope recipient. SA can use these headers,
3010 obviously out-of-the-box. Patch from Alex Miller.
3012 PH/08 The ${def test on a variable was returning false if the variable's
3013 value was "0", contrary to what the specification has always said!
3014 The result should be true unless the variable is empty.
3016 PH/09 The syntax error of a character other than { following "${if
3017 def:variable_name" (after optional whitespace) was not being diagnosed.
3018 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3019 accidental colon was present, for example, could give incorrect results.
3021 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3022 result is used (not including appendfile, where other changes are about
3025 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3026 This involved changing a lot of size variables from int to off_t. It
3027 should work with maildirs and everything.
3029 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3030 spamd dying while we are connected to it.
3032 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3035 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3036 function for BDB 4.3.
3038 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3040 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3041 the sources for PCRE have been re-arranged and more files are now
3044 PH/15 The code I had for printing potentially long long variables in PH/11
3045 above was not the best (it lost precision). The length of off_t variables
3046 is now inspected at build time, and an appropriate printing format (%ld
3047 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3048 to be "long long int" or "long int". This is needed for the internal
3049 formatting function string_vformat().
3051 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3052 the configuration file to be ":syslog", then the script "guesses" where
3053 the logs files are, rather than using the compiled in default. In our
3054 case the guess is not the same as the compiled default, so the script
3055 suddenly stopped working when I started to use syslog. The patch checks
3056 to see if log_file_path is "". If so, it attempts to read it from exim
3057 with no configuration file to get the compiled in version, before it
3058 falls back to the previous guessing code."
3060 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3061 implementing BATV in an Exim configuration. See NewStuff for the gory
3064 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3065 Makefile that are specific to HP-UX.
3067 PH/18 If the "use_postmaster" option was set for a recipient callout together
3068 with the "random" option, the postmaster address was used as the MAIL
3069 FROM address for the random test, but not for the subsequent recipient
3070 test. It is now used for both.
3072 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3073 patch removes a few documentation additions to RFC 3028, because the
3074 latest draft now contains them. It adds the new en;ascii-case comparator
3075 and a new error check for 8bit text in MIME parts. Comparator and
3076 require names are now matched exactly. I enabled the subaddress
3077 extension, but it is not well tested yet (read: it works for me)."
3079 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3080 rework some of the code of TK/09 above to avoid the hardwired use of
3081 "%lld" and "long long". Replaced the call to snprintf() with a call to
3084 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3085 records point to non-existent hosts", "retry timeout exceeded", and
3086 "retry time not reached for any host after a long failure period".
3088 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3089 experimental DomainKeys support:
3091 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3092 (2) On an error such as an illegally used "control", the wrong name for
3093 the control was given.
3095 These problems did NOT occur unless DomainKeys support was compiled.
3097 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3099 PH/24 Added ${if match_ip condition.
3101 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3102 databases so that it will be absolutely obvious if a crash occurs in the
3103 DB library. This is a regular occurrence (often caused by mis-matched
3106 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3107 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3108 picked up on a user's system that detects such things. There doesn't seem
3109 to be a gcc warning option for this - only an attribute that has to be
3110 put on the function's prototype. It seems that in Fedora Core 4 they have
3111 set this on a number of new functions. No doubt there will be more in due
3114 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3115 specify a transport. However, if an address that was verified by such a
3116 router was the subject of a callout, Exim crashed because it tried to
3117 read the rcpt_include_affixes from the non-existent transport. Now it
3118 just assumes that the setting of that option is false. This bug was
3119 introduced by 4.51/PH/31.
3121 PH/28 Changed -d+all to exclude +memory, because that information is very
3122 rarely of interest, but it makes the output a lot bigger. People tend to
3123 do -d+all out of habit.
3125 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3126 code in os-type was giving problems when libc.so lives in lib64, like on
3129 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3130 aren't the modern standard, and it seems that some systems' include files
3131 don't always have them. Exim was already checking for some of the newer
3132 ones like T_AAAA, and defining it itself. I've added checks for all the
3133 record types that Exim uses.
3135 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3136 not automatically generating a new one, as it is supposed to. This
3137 prevented TLS from working. If the file did exist, but contained invalid
3138 data, a new version was generated, as expected. It was only the case of a
3139 non-existent file that was broken.
3141 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3142 with a change in libdomainkeys > 0.64.
3144 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3145 from DNS. If the selector record carries the flag, it now has
3146 precedence over the domain-wide flag.
3148 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3150 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3151 the use of a port name that isn't defined in /etc/services) occurred, the
3152 message was deferred as in a normal delivery, and thus remained on the
3153 spool, instead of being failed because of the mua_wrapper setting. This
3154 is now fixed, and I tidied up some of the mua_wrapper messages at the
3157 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3158 the messages in a hash of arrays rather than using individual hashes.
3159 This is a bit cleaner and results in dramatic memory savings, albeit
3160 at a slight CPU cost.
3162 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3163 as requested by Marc Sherman.
3165 SC/10 Eximstats: added histograms for user specified patterns as requested
3168 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3170 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3171 fopen() in the content-scanning modules that did not already have it.
3177 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3178 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3180 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3182 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3184 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3185 to test Sieve filters that use "vacation".
3187 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3188 that changes the way the GnuTLS parameters are stored in the cache file.
3189 The new format can be generated externally. For backward compatibility,
3190 if the data in the cache doesn't make sense, Exim assumes it has read an
3191 old-format file, and it generates new data and writes a new file. This
3192 means that you can't go back to an older release without removing the
3195 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3196 work if there are any delivery delays because "one_time" forces the
3197 parent to be marked "delivered", so its unseen clone is never tried
3198 again. For this reason, Exim now forbids the simultaneous setting of
3201 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3202 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3203 entry if you want to know the details. The fix, however, neglected to
3204 consider the case when local delivery batching is involved. The test for
3205 "previously delivered" was not happening when checking to see if an
3206 address could be batched with a previous (undelivered) one; under
3207 certain circumstances this could lead to multiple deliveries to the same
3210 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3211 in its include files, and this causes problems building Exim.
3213 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3214 header_syntax) but Exim was just ignoring anything given after a slash.
3215 In particular, this caused confusion with an attempt to use "verify =
3216 reverse_host_lookup/defer_ok". An error is now given when options are
3217 supplied for verify items that do not have them. (Maybe reverse_host_
3218 lookup should have a defer_ok option, but that's a different point.)
3220 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3221 defined by RFC 821) to 2048, because there were problems with some AUTH
3222 commands, and RFC 1869 says the size should be increased for extended
3223 SMTP commands that take arguments.
3225 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3228 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3229 "unknown" error; now it says that the functionality isn't in the binary.
3231 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3232 an address' error message when a string expansion fails (syntax or
3233 whatever). Otherwise the password may appear in the log. Following change
3234 PH/42 below, there is no longer a chance of it appearing in a bounce
3237 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3239 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3240 of its addresses was ever tried. (Bugzilla bug #2.)
3242 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3243 the result incorrectly in the debug output. (It correctly added a newline
3244 to what was transported.)
3246 TF/01 Added $received_time.
3248 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3249 commented out examples of how to interface to a virus scanner and to
3250 SpamAssassin. Also added commented examples of av_scanner and
3251 spamd_address settings.
3253 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3254 and controls are allowed in which ACLs. There were a couple of minor
3255 errors. Some of the entries in the conditions table (which is a table of
3256 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3257 negation of where the condition IS allowed.
3259 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3261 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3262 header file does not have a version number, so I've had to invent a new
3263 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3264 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3265 radiusclient), but it was contributed by a Radius user.
3267 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3268 files or directories, for OpenSSL.
3270 PH/19 When an Exim process that is running as root has to create an Exim log
3271 file, it does so in a subprocess that runs as exim:exim so as to get the
3272 ownership right at creation (otherwise, other Exim processes might see
3273 the file with the wrong ownership). There was no test for failure of this
3274 fork() call, which would lead to the process getting stuck as it waited
3275 for a non-existent subprocess. Forks do occasionally fail when resources
3276 run out. I reviewed all the other calls to fork(); they all seem to check
3279 PH/20 When checking for unexpected SMTP input at connect time (before writing
3280 the banner), Exim was not dealing correctly with a non-positive return
3281 from the read() function. If the client had disconnected by this time,
3282 the result was a log entry for a synchronization error with an empty
3283 string after "input=" when read() returned zero. If read() returned -1
3284 (an event I could not check), uninitialized data bytes were printed.
3285 There were reports of junk text (parts of files, etc) appearing after
3288 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3290 PH/22 Added support for macro redefinition, and (re)definition in between
3291 driver and ACL definitions.
3293 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3294 forgetting to use the resulting value; it was using the unexpanded value.
3296 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3297 hadn't been configured. The fix is from Juergen Kreileder, who
3298 understands it better than I do:
3300 "Here's what I see happening with three configured cyrus_sasl
3301 authenticators configured (plain, login, cram-md5):
3303 On startup auth_cyrus_sasl_init() gets called for each of these.
3304 This means three calls to sasl_listmech() without a specified mech_list.
3305 => SASL tests which mechs of all available mechs actually work
3306 => three warnings about OTP not working
3307 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3309 With the patch, sasl_listmech() also gets called three times. But now
3310 SASL's mech_list option is set to the server_mech specified in the the
3311 authenticator. Or in other words, the answer from sasl_listmech()
3312 gets limited to just the mech you're testing for (which is different
3314 => the return list contains just 'plain' or 'login', 'cram-md5' or
3315 nothing depending on the value of ob->server_mech.
3317 I've just tested the patch: Authentication still works fine,
3318 unavailable mechs specified in the exim configuration are still
3319 caught, and the auth.log warnings about OTP are gone."
3321 PH/25 When debugging is enabled, the contents of the command line are added
3322 to the debugging output, even when log_selector=+arguments is not
3325 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3326 answer is "GNU", and only if the return is "GNU/something" is the answer
3329 PH/27 $acl_verify_message is now set immediately after the failure of a
3330 verification in an ACL, and so is available in subsequent modifiers. In
3331 particular, the message can be preserved by coding like this:
3333 warn !verify = sender
3334 set acl_m0 = $acl_verify_message
3336 Previously, $acl_verify_message was set only while expanding "message"
3337 and "log_message" when a very denied access.
3339 PH/28 Modified OS/os.c-Linux with
3341 -#ifndef OS_LOAD_AVERAGE
3342 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3344 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3345 nomenclature these days.)
3347 PH/29 Installed patch from the Sieve maintainer that adds the options
3348 sieve_useraddress and sieve_subaddress to the redirect router.
3350 PH/30 In these circumstances:
3351 . Two addresses routed to the same list of hosts;
3352 . First host does not offer TLS;
3353 . First host accepts first address;
3354 . First host gives temporary error to second address;
3355 . Second host offers TLS and a TLS session is established;
3356 . Second host accepts second address.
3357 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3358 and peerdn, if requested) that were in fact used only for the second
3361 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3362 attention to any local part prefix or suffix that was matched by the
3363 router that accepted the address. It now behaves in the same way as it
3364 does for delivery: the affixes are removed from the local part unless
3365 rcpt_include_affixes is set on the transport.
3367 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3368 timeout during the DATA phase of an incoming message.
3370 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3371 applied a patch sanctioned by the Sieve maintainer.
3373 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3374 the uid or gid is negative. A case of a negative gid caused this to be
3375 noticed. The fix allows for either to be negative.
3377 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3378 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3381 PH/36 The change PH/12 above was broken. Fixed it.
3383 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3384 the grounds that routing the same address twice would always produce the
3385 same answer. This might have been true once, but it is certainly no
3386 longer true now. Routing a child address may depend on the previous
3387 routing that produced that child. Some complicated redirection strategies
3388 went wrong when messages had multiple recipients, and made Exim's
3389 behaviour dependent on the order in which the addresses were given.
3391 I have moved the duplicate checking until after the routing is complete.
3392 Exim scans the addresses that are assigned to local and remote
3393 transports, and removes any duplicates. This means that more work will be
3394 done, as duplicates will always all be routed, but duplicates are
3395 presumably rare, so I don't expect this is of any significance.
3397 For deliveries to pipes, files, and autoreplies, the duplicate checking
3398 still happens during the routing process, since they are not going to be
3401 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3402 It corrects a timeout issue with spamd. This is Ian's comment: "The
3403 background is that sometimes spamd either never reads data from a
3404 connection it has accepted, or it never writes response data. The exiscan
3405 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3406 blindly assumes that writes won't block so it may never time out."
3408 PH/39 Allow G after quota size as well as K and M.
3410 PH/40 The value set for $authenticated_id in an authenticator may not contain
3411 binary zeroes or newlines because the value is written to log lines and
3412 to spool files. There was no check on this. Now the value is run through
3413 the string_printing() function so that such characters are converted to
3414 printable escape sequences.
3416 PH/41 $message_linecount is a new variable that contains the total number of
3417 lines in the message. Compare $body_linecount, which is the count for the
3420 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3421 bounce and delay warning messages, except in certain special cases, which
3424 (a) An SMTP error message from a remote host;
3425 (b) A message specified in a :fail: redirection;
3426 (c) A message specified in a "fail" command in a system filter;
3427 (d) A message specified in a FAIL return from the queryprogram router;
3428 (e) A message specified by the cannot_route_message router option.
3430 In these cases only, Exim does include the error details in bounce and
3431 warning messages. There are also a few cases where bland messages such
3432 as "unrouteable address" or "local delivery error" are given.
3434 PH/43 $value is now also set for the "else" part of a ${run expansion.
3436 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3437 being worked on, but at least Exim now implements the latest version to
3440 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3441 process to complete was treated as a delivery failure, a timeout while
3442 writing the message to the pipe was logged, but erroneously treated as a
3443 successful delivery. Such timeouts include transport filter timeouts. For
3444 consistency with the overall process timeout, these timeouts are now
3445 treated as errors, giving rise to delivery failures by default. However,
3446 there is now a new Boolean option for the pipe transport called
3447 timeout_defer, which, if set TRUE, converts the failures into defers for
3448 both kinds of timeout. A transport filter timeout is now identified in
3451 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3452 systems where "make" and "gmake" are different, calling "gmake" at top
3453 level broke things. I've arranged for the value of $(MAKE) to be passed
3454 from the Makefile to this script so that it can call the same version of
3458 A note about Exim versions 4.44 and 4.50
3459 ----------------------------------------
3461 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3462 changes of various kinds. As a consequence, a big documentation update was
3463 needed. This delayed the release for rather longer than seemed good, especially
3464 in the light of a couple of (minor) security issues. Therefore, the changes
3465 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3466 release. So 4.44 and 4.50 are in effect two different branches that both start
3469 I have left the 4.50 change log unchanged; it contains all the changes since
3470 4.43. The change log for 4.44 is below; many of its items are identical to
3471 those for 4.50. This seems to be the most sensible way to preserve the
3472 historical information.
3478 1. Minor wording change to the doc/README.SIEVE file.
3480 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3481 computation of the current number of files was incorrect.
3483 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3484 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3487 4. Give more explanation in the error message when the command for a transport
3488 filter fails to execute.
3490 5. There are several places where Exim runs a non-Exim command in a
3491 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3492 was being done only for the command run by the queryprogram router. It is
3493 now done for all such subprocesses. The other cases are: ${run, transport
3494 filters, and the commands run by the lmtp and pipe transports.
3496 6. Added CONFIGURE_GROUP build-time option.
3498 7. Some older OS have a limit of 256 on the maximum number of file
3499 descriptors. Exim was using setrlimit() to set 1000 as a large value
3500 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3501 systems. I've change it so that if it can't get 1000, it tries for 256.
3503 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3504 was an oversight, and furthermore, ever since the addition of extra
3505 controls (e.g. 4.43/32), the checks on when to allow different forms of
3506 "control" were broken. There should now be diagnostics for all cases when a
3507 control that does not make sense is encountered.
3509 9. Added the /retain_sender option to "control=submission".
3511 10. $recipients is now available in the predata ACL (oversight).
3513 11. Tidy the search cache before the fork to do a delivery from a message
3514 received from the command line. Otherwise the child will trigger a lookup
3515 failure and thereby defer the delivery if it tries to use (for example) a
3516 cached ldap connection that the parent has called unbind on.
3518 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3519 of $address_data from the recipient verification was clobbered by the
3520 sender verification.
3522 13. The value of address_data from a sender verification is now available in
3523 $sender_address_data in subsequent conditions in the ACL statement.
3525 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3527 15. Added a new option "connect=<time>" to callout options, to set a different
3530 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3531 was its contents. (It was OK if the option was not defined at all.)
3533 17. A "Completed" log line is now written for messages that are removed from
3534 the spool by the -Mrm option.
3536 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3537 information about exactly what failed.
3539 19. Added -dd to debug only the daemon process.
3541 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3542 handles timeouts, both on the server side and network timeouts. Renamed the
3543 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3545 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3546 It is now set to "smtps".
3548 22. $host_address is now set to the target address during the checking of
3549 ignore_target_hosts.
3551 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3552 being passed; this would have caused $sender_host_name to have been used if
3553 matching the list had actually called for a host name (not very likely,
3554 since this list is usually IP addresses). A host name is now passed as
3557 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3558 code that specifies a non-restarting handler (typically sigaction() in
3559 modern systems) in an attempt to fix a rare and obscure crash bug.
3561 25. Narrowed the window for a race in the daemon that could cause it to ignore
3562 SIGCHLD signals. This is not a major problem, because they are used only to
3563 wake it up if nothing else does.
3565 26. A malformed maildirsize file could cause Exim to calculate negative values
3566 for the mailbox size or file count. Odd effects could occur as a result.
3567 The maildirsize information is now recalculated if the size or filecount
3570 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3571 support for a long time. Removed HAVE_SYS_VFS_H.
3573 28. Installed the latest version of exipick from John Jetmore.
3575 29. In an address list, if the pattern was not a regular expression, an empty
3576 subject address (from a bounce message) matched only if the pattern was an
3577 empty string. Non-empty patterns were not even tested. This was the wrong
3578 because it is perfectly reasonable to use an empty address as part of a
3579 database query. An empty address is now tested by patterns that are
3580 lookups. However, all the other forms of pattern expect the subject to
3581 contain a local part and a domain, and therefore, for them, an empty
3582 address still always fails if the pattern is not itself empty.
3584 30. Exim went into a mad DNS loop when attempting to do a callout where the
3585 host was specified on an smtp transport, and looking it up yielded more
3586 than one IP address.
3588 31. Re-factored the code for checking spool and log partition space into a
3589 function that finds that data and another that does the check. The former
3590 is then used to implement four new variables: $spool_space, $log_space,
3591 $spool_inodes, and $log_inodes.
3593 32. The RFC2047 encoding function was originally intended for short strings
3594 such as real names; it was not keeping to the 75-character limit for
3595 encoded words that the RFC imposes. It now respects the limit, and
3596 generates multiple encoded words if necessary. To be on the safe side, I
3597 have increased the buffer size for the ${rfc2047: expansion operator from
3600 33. It is now permitted to omit both strings after an "if" condition; if the
3601 condition is true, the result is "true". As before, when the second string
3602 is omitted, a false condition yields an empty string. This makes it less
3603 cumbersome to write custom ACL and router conditions.
3605 34. Failure to deliver a bounce message always caused it to be frozen, even if
3606 there was an errors_to setting on the router. The errors_to setting is now
3609 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3610 canonical form (fully expanded) before being placed in
3611 $sender_host_address.
3613 36. The table in the code that translates DNS record types into text (T_A to
3614 "A" for instance) was missing entries for NS and CNAME. It is just possible
3615 that this could have caused confusion if both these types were looked up
3616 for the same domain, because the text type is used as part of Exim's
3617 per-process caching. But the chance of anyone hitting this buglet seems
3620 37. The dnsdb lookup has been extended in a number of ways.
3622 (1) There is a new type, "zns", which walks up the domain tree until it
3623 finds some nameserver records. It should be used with care.
3625 (2) There is a new type, "mxh", which is like "mx" except that it returns
3626 just the host names, not the priorities.
3628 (3) It is now possible to give a list of domains (or IP addresses) to be
3629 looked up. The behaviour when one of the lookups defers can be
3630 controlled by a keyword.
3632 (4) It is now possible to specify the separator character for use when
3633 multiple records are returned.
3635 38. The dnslists ACL condition has been extended: it is now possible to supply
3636 a list of IP addresses and/or domains to be looked up in a particular DNS
3639 39. Added log_selector=+queue_time_overall.
3641 40. When running the queue in the test harness, wait just a tad after forking a
3642 delivery process, to get repeatability of debugging output.
3644 41. Include certificate and key file names in error message when GnuTLS fails
3645 to set them up, because the GnuTLS error message doesn't include the name
3646 of the failing file when there is a problem reading it.
3648 42. Allow both -bf and -bF in the same test run.
3650 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3652 44. The "Exiscan patch" is now merged into the mainline Exim source.
3654 45. Sometimes the final signoff response after QUIT could fail to get
3655 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3656 < 0 before doing a fflush(). This bug looks as though it goes back to the
3657 introduction of TLS in release 3.20, but "sometimes" must have been rare
3658 because the tests only now provoked it.
3660 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3661 (this can affect the format of dates).
3663 47. exim_tidydb, when checking for the continued existence of a message for
3664 which it has found a message-specific retry record, was not finding
3665 messages that were in split spool directories. Consequently, it was
3666 deleting retry records that should have stayed in existence.
3668 48. Steve fixed some bugs in eximstats.
3670 49. The SPA authentication driver was not abandoning authentication and moving
3671 on to the next authenticator when an expansion was forced to fail,
3672 contradicting the general specification for all authenticators. Instead it
3673 was generating a temporary error. It now behaves as specified.
3675 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3676 (the order specifies the preference for clients). The order is now AES256,
3677 AES128, 3DES, ARCFOUR128.
3679 51. Small patch to Sieve code - explicitly set From: when generating an
3682 52. Exim crashed if a remote delivery caused a very long error message to be
3683 recorded - for instance if somebody sent an entire SpamAssassin report back
3684 as a large number of 550 error lines. This bug was coincidentally fixed by
3685 increasing the size of one of Exim's internal buffers (big_buffer) that
3686 happened as part of the Exiscan merge. However, to be on the safe side, I
3687 have made the code more robust (and fixed the comments that describe what
3690 53. Now that there can be additional text after "Completed" in log lines (if
3691 the queue_time_overall log selector is set), a one-byte patch to exigrep
3692 was needed to allow it to recognize "Completed" as not the last thing in
3695 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3696 patch that reportedly fixes this has been added. I am not expert enough to
3697 create a test for it. This is what the patch creator wrote:
3699 "I found a little strange behaviour of ldap code when working with
3700 Windows 2003 AD Domain, where users was placed in more than one
3701 Organization Units. When I tried to give exim partial DN, the exit code
3702 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3703 But simultaneously result of request was absolutely normal ldap result,
3704 so I produce this patch..."
3706 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3707 REFERENCE, so I have modified the code to exclude the patch when that macro
3710 55. Some experimental protocols are using DNS PTR records for new purposes. The
3711 keys for these records are domain names, not reversed IP addresses. The
3712 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3713 leaves it alone. Component reversal etc. now happens only for IP addresses.
3716 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3718 57. Double the size of the debug message buffer (to 2048) so that more of very
3719 long debug lines gets shown.
3721 58. The exicyclog utility now does better if the number of log files to keep
3722 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3724 59. Two changes related to the smtp_active_hostname option:
3726 (1) $smtp_active_hostname is now available as a variable.
3727 (2) The default for smtp_banner uses $smtp_active_hostname instead
3728 of $primary_hostname.
3730 60. The host_aton() function is supposed to be passed a string that is known
3731 to be a valid IP address. However, in the case of IPv6 addresses, it was
3732 not checking this. This is a hostage to fortune. Exim now panics and dies
3733 if the condition is not met. A case was found where this could be provoked
3734 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3735 components; fortuitously, this particular loophole had already been fixed
3736 by change 4.50/55 above.
3738 If there are any other similar loopholes, the new check in host_aton()
3739 itself should stop them being exploited. The report I received stated that
3740 data on the command line could provoke the exploit when Exim was running as
3741 exim, but did not say which command line option was involved. All I could
3742 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3743 running as the user.
3746 61. There was a buffer overflow vulnerability in the SPA authentication code
3747 (which came originally from the Samba project). I have added a test to the
3748 spa_base64_to_bits() function which I hope fixes it.
3751 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3752 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3754 63. The daemon start-up calls getloadavg() while still root for those OS that
3755 need the first call to be done as root, but it missed one case: when
3756 deliver_queue_load_max is set with deliver_drop_privilege. This is
3757 necessary for the benefit of the queue runner, because there is no re-exec
3758 when deliver_drop_privilege is set.
3760 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3761 This has been fixed.
3763 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3764 was in use, was not putting the data itself into the right store pool;
3765 consequently, it could be overwritten for a subsequent message in the same
3766 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3769 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3771 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3772 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3773 the function were treating the return as a boolean value, which happened to
3774 work because 0=false and not-0=true, but is not correct code.
3776 68. The host_aton() function was not handling scoped IPv6 addresses (those
3777 with, for example, "%eth0" on the end) correctly.
3779 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3780 negated items (that is, ~something) in unsigned ints. Some compilers
3781 apparently mutter when there is no cast.
3783 70. If an address verification called from an ACL failed, and did not produce a
3784 user-specific message (i.e. there was only a "system" message), nothing was
3785 put in $acl_verify_message. In this situation, it now puts the system
3788 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3789 session; change 4.31/43 added the unwanted input to the log line - except
3790 that it did not do this in the start of session case. It now does.
3792 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3793 This is wrong and can cause the other end to generate a synchronization
3794 error if it is another Exim or anything else that does the synchronization
3795 check. A QUIT command is no longer sent after a timeout.
3797 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3798 during host lookups.
3800 74. The defer_ok option of callout verification was not working if it was used
3801 when verifying addresses in header lines, that is, for this case:
3803 verify = header_sender/callout=defer_ok
3805 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3806 those file descriptors could be used for SMTP connections. If anything
3807 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3808 could be sent to the SMTP client, causing chaos. The daemon now opens
3809 stdin, stdout, and stderr to /dev/null when it puts itself into the
3812 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3813 log by default. The user can override this with suitable Perl magic.
3815 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3816 the log message when discard triggers, was not working for the DATA ACL or
3817 for the non-SMTP ACL.
3819 78. Error message wording change in sieve.c.
3821 79. If smtp_accept_max_per_host was set, the number of connections could be
3822 restricted to fewer than expected, because the daemon was trying to set up
3823 a new connection before checking whether the processes handling previous
3824 connections had finished. The check for completed processes is now done
3825 earlier. On busy systems, this bug wouldn't be noticed because something
3826 else would have woken the daemon, and it would have reaped the completed
3829 80. If a message was submitted locally by a user whose login name contained one
3830 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3831 It caused a spool format error. I have fixed the spool reading code. A
3832 related problem was that the "from" clause in the Received: line became
3833 illegal because of the space(s). It is now covered by ${quote_local_part.
3835 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3838 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3840 83. Patch from Sieve maintainer to fix unterminated string problem in
3841 "vacation" handling.
3843 84. Some minor changes to the Linux configuration files to help with other
3844 OS variants using glibc.
3846 85. One more patch for Sieve to update vacation handling to latest spec.
3849 ----------------------------------------------------
3850 See the note above about the 4.44 and 4.50 releases.
3851 ----------------------------------------------------
3857 1. Change 4.43/35 introduced a bug that caused file counts to be
3858 incorrectly computed when quota_filecount was set in an appendfile
3861 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3862 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3865 3. Give more explanation in the error message when the command for a transport
3866 filter fails to execute.
3868 4. There are several places where Exim runs a non-Exim command in a
3869 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3870 was being done only for the command run by the queryprogram router. It is
3871 now done for all such subprocesses. The other cases are: ${run, transport
3872 filters, and the commands run by the lmtp and pipe transports.
3874 5. Some older OS have a limit of 256 on the maximum number of file
3875 descriptors. Exim was using setrlimit() to set 1000 as a large value
3876 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3877 systems. I've change it so that if it can't get 1000, it tries for 256.
3879 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3880 was an oversight, and furthermore, ever since the addition of extra
3881 controls (e.g. 4.43/32), the checks on when to allow different forms of
3882 "control" were broken. There should now be diagnostics for all cases when a
3883 control that does not make sense is encountered.
3885 7. $recipients is now available in the predata ACL (oversight).
3887 8. Tidy the search cache before the fork to do a delivery from a message
3888 received from the command line. Otherwise the child will trigger a lookup
3889 failure and thereby defer the delivery if it tries to use (for example) a
3890 cached ldap connection that the parent has called unbind on.
3892 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3893 of $address_data from the recipient verification was clobbered by the
3894 sender verification.
3896 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3897 was its contents. (It was OK if the option was not defined at all.)
3899 11. A "Completed" log line is now written for messages that are removed from
3900 the spool by the -Mrm option.
3902 12. $host_address is now set to the target address during the checking of
3903 ignore_target_hosts.
3905 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3906 being passed; this would have caused $sender_host_name to have been used if
3907 matching the list had actually called for a host name (not very likely,
3908 since this list is usually IP addresses). A host name is now passed as
3911 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3912 code that specifies a non-restarting handler (typically sigaction() in
3913 modern systems) in an attempt to fix a rare and obscure crash bug.
3915 15. Narrowed the window for a race in the daemon that could cause it to ignore
3916 SIGCHLD signals. This is not a major problem, because they are used only to
3917 wake it up if nothing else does.
3919 16. A malformed maildirsize file could cause Exim to calculate negative values
3920 for the mailbox size or file count. Odd effects could occur as a result.
3921 The maildirsize information is now recalculated if the size or filecount
3924 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3925 support for a long time. Removed HAVE_SYS_VFS_H.
3927 18. Updated exipick to current release from John Jetmore.
3929 19. Allow an empty sender to be matched against a lookup in an address list.
3930 Previously the only cases considered were a regular expression, or an
3933 20. Exim went into a mad DNS lookup loop when doing a callout where the
3934 host was specified on the transport, if the DNS lookup yielded more than
3937 21. The RFC2047 encoding function was originally intended for short strings
3938 such as real names; it was not keeping to the 75-character limit for
3939 encoded words that the RFC imposes. It now respects the limit, and
3940 generates multiple encoded words if necessary. To be on the safe side, I
3941 have increased the buffer size for the ${rfc2047: expansion operator from
3944 22. Failure to deliver a bounce message always caused it to be frozen, even if
3945 there was an errors_to setting on the router. The errors_to setting is now
3948 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3949 canonical form (fully expanded) before being placed in
3950 $sender_host_address.
3952 24. Updated eximstats to version 1.33
3954 25. Include certificate and key file names in error message when GnuTLS fails
3955 to set them up, because the GnuTLS error message doesn't include the name
3956 of the failing file when there is a problem reading it.
3958 26. Expand error message when OpenSSL has problems setting up cert/key files.
3961 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3962 (this can affect the format of dates).
3964 28. exim_tidydb, when checking for the continued existence of a message for
3965 which it has found a message-specific retry record, was not finding
3966 messages that were in split spool directories. Consequently, it was
3967 deleting retry records that should have stayed in existence.
3969 29. eximstats updated to version 1.35
3970 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3971 1.35 - bugfix such that pie charts by volume are generated correctly
3973 30. The SPA authentication driver was not abandoning authentication and moving
3974 on to the next authenticator when an expansion was forced to fail,
3975 contradicting the general specification for all authenticators. Instead it
3976 was generating a temporary error. It now behaves as specified.
3978 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3979 (the order specifies the preference for clients). The order is now AES256,
3980 AES128, 3DES, ARCFOUR128.
3982 31. Small patch to Sieve code - explicitly set From: when generating an
3985 32. Exim crashed if a remote delivery caused a very long error message to be
3986 recorded - for instance if somebody sent an entire SpamAssassin report back
3987 as a large number of 550 error lines. This bug was coincidentally fixed by
3988 increasing the size of one of Exim's internal buffers (big_buffer) that
3989 happened as part of the Exiscan merge. However, to be on the safe side, I
3990 have made the code more robust (and fixed the comments that describe what
3993 33. Some experimental protocols are using DNS PTR records for new purposes. The
3994 keys for these records are domain names, not reversed IP addresses. The
3995 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3996 leaves it alone. Component reversal etc. now happens only for IP addresses.
3999 34. The host_aton() function is supposed to be passed a string that is known
4000 to be a valid IP address. However, in the case of IPv6 addresses, it was
4001 not checking this. This is a hostage to fortune. Exim now panics and dies
4002 if the condition is not met. A case was found where this could be provoked
4003 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4004 components; fortuitously, this particular loophole had already been fixed
4005 by change 4.50/55 or 4.44/33 above.
4007 If there are any other similar loopholes, the new check in host_aton()
4008 itself should stop them being exploited. The report I received stated that
4009 data on the command line could provoke the exploit when Exim was running as
4010 exim, but did not say which command line option was involved. All I could
4011 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4012 running as the user.
4015 35. There was a buffer overflow vulnerability in the SPA authentication code
4016 (which came originally from the Samba project). I have added a test to the
4017 spa_base64_to_bits() function which I hope fixes it.
4020 36. The daemon start-up calls getloadavg() while still root for those OS that
4021 need the first call to be done as root, but it missed one case: when
4022 deliver_queue_load_max is set with deliver_drop_privilege. This is
4023 necessary for the benefit of the queue runner, because there is no re-exec
4024 when deliver_drop_privilege is set.
4026 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4027 was in use, was not putting the data itself into the right store pool;
4028 consequently, it could be overwritten for a subsequent message in the same
4029 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4032 38. Sometimes the final signoff response after QUIT could fail to get
4033 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4034 < 0 before doing a fflush(). This bug looks as though it goes back to the
4035 introduction of TLS in release 3.20, but "sometimes" must have been rare
4036 because the tests only now provoked it.
4042 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4043 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4044 Now it returns an integer. A number of calls were still expecting a T/F
4045 return. Fortuitously, in all cases, the tests worked in OK situations,
4046 which is the norm. However, things would have gone wrong on any write
4047 failures on the smtp file descriptor. This function is used when sending
4048 messages over SMTP and also when doing verify callouts.
4050 2. When Exim is called to do synchronous delivery of a locally submitted
4051 message (the -odf or -odi options), it no longer closes stderr before doing
4054 3. Implemented the mua_wrapper option.
4056 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4058 5. Implemented the functions header_remove(), header_testname(),
4059 header_add_at_position(), and receive_remove_recipient(), and exported them
4062 6. If an ACL "warn" statement specified the addition of headers, Exim already
4063 inserted X-ACL-Warn: at the start if there was no header name. However, it
4064 was not making this test for the second and subsequent header lines if
4065 there were newlines in the string. This meant that an invalid header could
4066 be inserted if Exim was badly configured.
4068 7. Allow an ACL "warn" statement to add header lines at the start or after all
4069 the Received: headers, as well as at the end.
4071 8. Added the rcpt_4xx retry error code.
4073 9. Added postmaster_mailfrom=xxx to callout verification option.
4075 10. Added mailfrom=xxxx to the callout verification option, for verify=
4078 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4079 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4081 12. Inserted some casts to stop certain compilers warning when using pointer
4082 differences as field lengths or precisions in printf-type calls (mostly
4083 affecting debugging statements).
4085 13. Added optional readline() support for -be (dynamically loaded).
4087 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4088 same clock tick as a message's arrival, so that its received time was the
4089 same as the "first fail" time on the retry record, and that message
4090 remained on the queue past the ultimate address timeout, every queue runner
4091 would try a delivery (because it was past the ultimate address timeout) but
4092 after another failure, the ultimate address timeout, which should have then
4093 bounced the address, did not kick in. This was a "< instead of <=" error;
4094 in most cases the first failure would have been in the next clock tick
4095 after the received time, and all would be well.
4097 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4098 being recognized when the domain list was tested by the match_domain
4099 condition in an expansion string.
4101 16. Added the ${str2b64: operator.
4103 17. Exim was always calling setrlimit() to set a large limit for the number of
4104 processes, without checking whether the existing limit was already
4105 adequate. (It did check for the limit on file descriptors.) Furthermore,
4106 errors from getrlimit() and setrlimit() were being ignored. Now they are
4107 logged to the main and panic logs, but Exim does carry on, to try to do its
4108 job under whatever limits there are.
4110 18. Imported PCRE 5.0.
4112 19. Trivial typo in log message " temporarily refused connection" (the leading
4115 20. If the log selector return_path_on_delivery was set and an address was
4116 redirected to /dev/null, the delivery process crashed because it assumed
4117 that a return path would always be set for a "successful" delivery. In this
4118 case, the whole delivery is bypassed as an optimization, and therefore no
4121 21. Internal re-arrangement: the function for sending a challenge and reading
4122 a response while authentication was assuming a zero-terminated challenge
4123 string. It's now changed to take a pointer and a length, to allow for
4124 binary data in such strings.
4126 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4128 23. Exim was not respecting finduser_retries when seeking the login of the
4129 uid under which it was called; it was always trying 10 times. (The default
4130 setting of finduser_retries is zero.) Also, it was sleeping after the final
4131 failure, which is pointless.
4133 24. Implemented tls_on_connect_ports.
4135 25. Implemented acl_smtp_predata.
4137 26. If the domain in control=submission is set empty, Exim assumes that the
4138 authenticated id is a complete email address when it generates From: or
4139 Sender: header lines.
4141 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4142 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4143 chown and chgrp in /bin and hostname in /usr/bin.
4145 28. Exim was keeping the "process log" file open after each use, just as it
4146 does for the main log. This opens the possibility of it remaining open for
4147 long periods when the USR1 signal hits a daemon. Occasional processlog
4148 errors were reported, that could have been caused by this. Anyway, it seems
4149 much more sensible not to leave this file open at all, so that is what now
4152 29. The long-running daemon process does not normally write to the log once it
4153 has entered its main loop, and it closes the log before doing so. This is
4154 so that log files can straightforwardly be renamed and moved. However,
4155 there are a couple of unusual error situations where the daemon does write
4156 log entries, and I had neglected to close the log afterwards.
4158 30. The text of an SMTP error response that was received during a remote
4159 delivery was being truncated at 512 bytes. This is too short for some of
4160 the long messages that one sometimes sees. I've increased the limit to
4163 31. It is now possible to make retry rules that apply only when a message has a
4164 specific sender, in particular, an empty sender.
4166 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4167 it possible to be selective about when SMTP synchronization is enforced.
4169 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4171 32. Implemented hosts_connection_nolog.
4173 33. Added an ACL for QUIT.
4175 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4178 35. Added mailbox_size and mailbox_filecount to appendfile.
4180 36. Added control = no_multiline_responses to ACLs.
4182 37. There was a bug in the logic of the code that waits for the clock to tick
4183 in the case where the clock went backwards by a substantial amount such
4184 that the microsecond fraction of "now" was more than the microsecond
4185 fraction of "then" (but the whole seconds number was less).
4187 38. Added support for the libradius Radius client library this is found on
4188 FreeBSD (previously only the radiusclient library was supported).
4194 1. When certain lookups returned multiple values in the form name=value, the
4195 quoting of the values was not always being done properly. Specifically:
4196 (a) If the value started with a double quote, but contained no whitespace,
4198 (b) If the value contained whitespace other than a space character (i.e.
4199 tabs or newlines or carriage returns) it was not quoted.
4200 This fix has been applied to the mysql and pgsql lookups by writing a
4201 separate quoting function and calling it from the lookup code. The fix
4202 should probably also be applied to nisplus, ibase and oracle lookups, but
4203 since I cannot test any of those, I have not disturbed their existing code.
4205 2. A hit in the callout cache for a specific address caused a log line with no
4206 reason for rejecting RCPT. Now it says "Previous (cached) callout
4207 verification failure".
4209 3. There was an off-by-one bug in the queryprogram router. An over-long
4210 return line was truncated at 256 instead of 255 characters, thereby
4211 overflowing its buffer with the terminating zero. As well as fixing this, I
4212 have increased the buffer size to 1024 (and made a note to document this).
4214 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4215 when Exim is waiting for an SMTP response from a remote server, Exim
4216 restarts its select() call on the socket, thereby resetting its timeout.
4217 This is not a problem when such interrupts are rare. Somebody set up a cron
4218 job to run exiwhat every 2 minutes, which is less than the normal select()
4219 timeout (5 or 10 minutes). This meant that the select() timeout never
4220 kicked in because it was always reset. I have fixed this by comparing the
4221 time when an interrupt arrives with the time at the start of the first call
4222 to select(). If more time than the timeout has elapsed, the interrupt is
4223 treated as a timeout.
4225 5. Some internal re-factoring in preparation for the addition of Sieve
4226 extensions (by MH). In particular, the "personal" test is moved to a
4227 separate function, and given an option for scanning Cc: and Bcc: (which is
4228 not set for Exim filters).
4230 6. When Exim created an email address using the login of the caller as the
4231 local part (e.g. when creating a From: or Sender: header line), it was not
4232 quoting the local part when it contained special characters such as @.
4234 7. Installed new OpenBSD configuration files.
4236 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4237 try to make them clearer.
4239 9. Callout options, other than the timeout value, were being ignored when
4240 verifying sender addresses in header lines. For example, when using
4242 verify = header_sender/callout=no_cache
4244 the cache was (incorrectly) being used.
4246 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4247 only the Cygwin environment.
4249 11. When return_path_on_delivery was set as a log selector, if different remote
4250 addresses in the same message used different return paths and parallel
4251 remote delivery occurred, the wrong values would sometimes be logged.
4252 (Whenever a remote delivery process finished, the return path value from
4253 the most recently started remote delivery process was logged.)
4255 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4256 lines when AUTH and/or TLS are in use. This is the "received protocol"
4257 field. Exim used to use "asmtp" for authenticated SMTP, without any
4258 indication (in the protocol name) for TLS use. Now it follows the RFC and
4259 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4260 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4261 names appear in log lines as well as in Received: header lines.
4263 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4264 extensions, and comparison tests, and to fix some bugs.
4266 14. Changes to the "personal" filter test:
4268 (1) The test was buggy in that it was just doing the equivalent of
4269 "contains" tests on header lines. For example, if a user's address was
4270 anne@some.where, the "personal" test would incorrectly be true for
4272 To: susanne@some.where
4274 This test is now done by extracting each address from the header in turn,
4275 and checking the entire address. Other tests that are part of "personal"
4276 are now done using regular expressions (for example, to check local parts
4277 of addresses in From: header lines).
4279 (2) The list of non-personal local parts in From: addresses has been
4280 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4281 taken from the Sieve specification recommendations.
4283 (3) If the message contains any header line starting with "List-" it is
4284 treated as non-personal.
4286 (4) The test for "circular" in the Subject: header line has been removed
4287 because it now seems ill-conceived.
4289 15. Minor typos in src/EDITME comments corrected.
4291 16. Installed latest exipick from John Jetmore.
4293 17. If headers_add on a router specified a text string that was too long for
4294 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4295 of string_sprintf() is now avoided.
4297 18. $message_body_size was not set (it was always zero) when running the DATA
4298 ACL and the local_scan() function.
4300 19. For the "mail" command in an Exim filter, no default was being set for
4301 the once_repeat time, causing a random time value to be used if "once" was
4302 specified. (If the value happened to be <= 0, no repeat happened.) The
4303 default is now 0s, meaning "never repeat". The "vacation" command was OK
4304 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4305 (I found it when inspecting the code).
4307 20. There is now an overall timeout for performing a callout verification. It
4308 defaults to 4 times the callout timeout, which applies to individual SMTP
4309 commands during the callout. The overall timeout applies when there is more
4310 than one host that can be tried. The timeout is checked before trying the
4311 next host. This prevents very long delays if there are a large number of
4312 hosts and all are timing out (e.g. when the network connections are timing
4313 out). The value of the overall timeout can be changed by specifying an
4314 additional sub-option for "callout", called "maxwait". For example:
4316 verify = sender/callout=5s,maxwait=20s
4318 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4319 to the end before writing, but this should make it even safer).
4321 22. Exim was forgetting that it had advertised PIPELINING for the second and
4322 subsequent messages on an SMTP connection. It was also not resetting its
4323 memory on STARTTLS and an internal HELO.
4325 23. When Exim logs an SMTP synchronization error within a session, it now
4326 records whether PIPELINING has been advertised or not.
4328 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4329 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4330 rather than long int.
4332 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4334 26. Added the never_mail option to autoreply.
4340 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4341 crash if the getsockname() call failed; this can happen if a connection is
4342 closed very soon after it is established. The problem was simply in the
4343 order in which certain operations were done, causing Exim to try to write
4344 to the SMTP stream before it had set up the file descriptor. The bug has
4345 been fixed by making things happen in the correct order.
4351 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4352 before the connection was closed, thus losing the rejection response.
4354 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4355 some early Solaris releases, but causes trouble in current releases where
4356 socklen_t is defined.
4358 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4361 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4364 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4365 that releasing the top of it at the end releases what was used for sub-
4366 expansions (unless the block got too big). However, discard this block if
4367 the first thing is a variable or header, so that we can use its block when
4368 it is dynamic (useful for very large $message_headers, for example).
4370 6. Lookups now cache *every* query, not just the most recent. A new, separate
4371 store pool is used for this. It can be recovered when all lookup caches are
4372 flushed. Lookups now release memory at the end of their result strings.
4373 This has involved some general refactoring of the lookup sources.
4375 7. Some code has been added to the store_xxx() functions to reduce the amount
4376 of flapping under certain conditions.
4378 8. log_incoming_interface used to affect only the <= reception log lines. Now
4379 it causes the local interface and port to be added to several more SMTP log
4380 lines, for example "SMTP connection from", and rejection lines.
4382 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4384 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4386 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4387 could be overwritten at the end of the current message (or the start of a
4388 new message if it was set in a HELO ACL). The value is now preserved for
4389 the duration of the SMTP connection.
4391 12. If a transport had a headers_rewrite setting, and a matching header line
4392 contained an unqualified address, that address was qualified, even if it
4393 did not match any rewriting rules. The underlying bug was that the values
4394 of the flags that permit the existence of unqualified sender and recipient
4395 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4396 non-local messages, and by -bnq for local messages) were not being
4397 preserved with the message after it was received.
4399 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4400 "next input=" as part of the text comprising the host identity instead of
4401 the correct text. The code was using the same buffer for two different
4402 strings. However, depending on which order the printing function evaluated
4403 its arguments, the bug did not always show up. Under Linux, for example, my
4404 test suite worked just fine.
4406 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4407 doesn't work with some older versions of Perl. It has been changed to "my",
4408 which in any case is probably the better facility to use.
4410 15. A really picky compiler found some instances of statements for creating
4411 error messages that either had too many or two few arguments for the format
4414 16. The size of the buffer for calls to the DNS resolver has been increased
4415 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4416 for addresses that have a lot of PTR records. This alleviates a problem; it
4417 does not fully solve it.
4419 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4420 buffer now truncates the list and logs the incident, which is the same
4421 action as happens when Exim is looking up a host name and its aliases.
4422 Previously in this situation something unpredictable would happen;
4423 sometimes it was "internal error: store_reset failed".
4425 18. If a server dropped the connection unexpectedly when an Exim client was
4426 using GnuTLS and trying to read a response, the client delivery process
4427 crashed while trying to generate an error log message.
4429 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4430 string, for example:
4432 warn message = H1: something\nH2: something
4434 the text was added as a single header line from Exim's point of view
4435 though it ended up OK in the delivered message. However, searching for the
4436 second and subsequent header lines using $h_h2: did not work. This has been
4437 fixed. Similarly, if a system filter added multiple headers in this way,
4438 the routers could not see them.
4440 20. Expanded the error message when iplsearch is called with an invalid key to
4441 suggest using net-iplsearch in a host list.
4443 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4444 ACLs are no longer actually imposed (and a message to that effect is
4447 22. If a "gecos" field in a passwd entry contained escaped characters, in
4448 particular, if it contained a \" sequence, Exim got it wrong when building
4449 a From: or a Sender: header from that name. A second bug also caused
4450 incorrect handling when an unquoted " was present following a character
4451 that needed quoting.
4453 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4454 was not being matched caselessly.
4456 24. Arranged for all hyphens in the exim.8 source to be escaped with
4459 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4460 properly. Recipient callout cache records were still being keyed to include
4461 the sender, even when use_sender was set false. This led to far more
4462 callouts that were necessary. The sender is no longer included in the key
4463 when use_sender is false.
4465 26. Added "control = submission" modifier to ACLs.
4467 27. Added the ${base62d: operator to decode base 62 numbers.
4469 28. dnsdb lookups can now access SRV records.
4471 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4472 the configuration file.
4474 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4475 (-v) mode. This makes the output for a verbose queue run more intelligible.
4477 31. Added a use_postmaster feature to recipient callouts.
4479 32. Added the $body_zerocount variable, containing the number of binary zero
4480 bytes in the message body.
4482 33. The time of last modification of the "new" subdirectory is now used as the
4483 "mailbox time last read" when there is a quota error for a maildir
4486 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4488 35. Added +ignore_unknown as a special item in host lists.
4490 36. Code for decoding IPv6 addresses in host lists is now included, even if
4491 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4492 address was recognized as an IP address, but was then not correctly decoded
4493 into binary, causing unexpected and incorrect effects when compared with
4500 1. Very minor rewording of debugging text in manualroute to say "list of
4501 hosts" instead of "hostlist".
4503 2. If verify=header_syntax was set, and a header line with an unqualified
4504 address (no domain) and a large number of spaces between the end of the
4505 name and the colon was received, the reception process suffered a buffer
4506 overflow, and (when I tested it) crashed. This was caused by some obsolete
4507 code that should have been removed. The fix is to remove it!
4509 3. When running in the test harness, delay a bit after writing a bounce
4510 message to get a bit more predictability in the log output.
4512 4. Added a call to search_tidyup() just before forking a reception process. In
4513 theory, someone could use a lookup in the expansion of smtp_accept_max_
4514 per_host which, without the tidyup, could leave open a database connection.
4516 5. Added the variables $recipient_data and $sender_data which get set from a
4517 lookup success in an ACL "recipients" or "senders" condition, or a router
4518 "senders" option, similar to $domain_data and $local_part_data.
4520 6. Moved the writing of debug_print from before to after the "senders" test
4523 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4524 problems for message scanning, either using a data ACL, or using
4525 local_scan() because the Received: header was not generated till after they
4526 were called (in order to set the time as the time of reception completion).
4527 I have revised the way this works. The header is now generated after the
4528 body is received, but before the ACL or local_scan() are called. After they
4529 are run, the timestamp in the header is updated.
4535 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4536 before starting a queue runner without re-exec. This happened only when
4537 deliver_drop_privilege was set or when the Exim user was set to root. The
4538 effect of the bug was that timeouts during subsequent deliveries caused
4539 crashes instead of being properly handled. The handler is now left at its
4540 default (and expected) setting.
4542 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4543 message, again when deliver_drop_privilege is set or Exim is run as root.
4544 The bug described in (1) was not present in this case, but the tidying up
4545 of the other signals was missing. I have made the two cases consistent.
4547 3. The ignore_target_hosts setting on a manualroute router was being ignored
4548 for hosts that were looked up using the /MX notation.
4550 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4553 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4554 operated on the sender address. After changing the $sender_address to <>
4555 for the sender address verify, Exim was re-instated it as the original
4556 (before rewriting) address, but remembering that it had rewritten it, so it
4557 wasn't rewriting it again. This bug also had the effect of breaking the
4558 sender address verification caching when the sender address was rewritten.
4560 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4561 This has been changed so that if the ip literal address matches
4562 ignore_target_hosts, the router declines.
4564 7. Added expansion conditions match_domain, match_address, and match_local_
4565 part (NOT match_host).
4567 8. The placeholder for the Received: header didn't have a length field set.
4569 9. Added code to Exim itself and to exim_lock to test for a specific race
4570 condition that could lead to file corruption when using MBX delivery. The
4571 issue is with the lockfile that is created in /tmp. If this file is removed
4572 after a process has opened it but before that process has acquired a lock,
4573 there is the potential for a second process to recreate the file and also
4574 acquire a lock. This could lead to two Exim processes writing to the file
4575 at the same time. The added code performs the same test as UW imapd; it
4576 checks after acquiring the lock that its file descriptor still refers to
4577 the same named file.
4579 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4580 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4583 11. Added the smtp_active_hostname option. If used, this will typically be made
4584 to depend on the incoming interface address. Because $interface_address is
4585 not set up until the daemon has forked a reception process, error responses
4586 that can happen earlier (such as "too many connections") no longer contain
4589 12. If an expansion in a condition on a "warn" statement fails because a lookup
4590 defers, the "warn" statement is abandoned, and the next ACL statement is
4591 processed. Previously this caused the whole ACL to be aborted.
4593 13. Added the iplsearch lookup type.
4595 14. Added ident_timeout as a log selector.
4597 15. Added tls_certificate_verified as a log selector.
4599 16. Added a global option tls_require_ciphers (compare the smtp transport
4600 option of the same name). This controls incoming TLS connections.
4602 17. I finally figured out how to make tls_require_ciphers do a similar thing
4603 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4604 before starting the TLS session.
4606 18. Tabs are now shown as \t in -bP output.
4608 19. If the log selector return_path_on_delivery was set, Exim crashed when
4609 bouncing a message because it had too many Received: header lines.
4611 20. If two routers both had headers_remove settings, and the first one included
4612 a superfluous trailing colon, the final name in the first list and the
4613 first name in the second list were incorrectly joined into one item (with a
4614 colon in the middle).
4620 1. Added -C and -D options to the exinext utility, mainly to make it easier
4621 to include in the automated testing, but these could be helpful when
4622 multiple configurations are in use.
4624 2. The exinext utility was not formatting the output nicely when there was
4625 an alternate port involved in the retry record key, nor when there was a
4626 message id as well (for retries that were specific to a specific message
4627 and a specific host). It was also confused by IPv6 addresses, because of
4628 the additional colons they contain. I have fixed the IPv4 problem, and
4629 patched it up to do a reasonable job for IPv6.
4631 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4632 delivery, the log line now contains "pipelined" if PIPELINING was used.
4634 4. An SMTP transport process used to panic and die if the bind() call to set
4635 an explicit outgoing interface failed. This has been changed; it is now
4636 treated in the same way as a connect() failure.
4638 5. A reference to $sender_host_name in the part of a conditional expansion
4639 that was being skipped was still causing a DNS lookup. This no longer
4642 6. The def: expansion condition was not recognizing references to header lines
4643 that used bh_ and bheader_.
4645 7. Added the _cache feature to named lists.
4647 8. The code for checking quota_filecount in the appendfile transport was
4648 allowing one more file than it should have been.
4650 9. For compatibility with Sendmail, the command line option
4658 and sets the incoming protocol and host name (for trusted callers). The
4659 host name and its colon can be omitted when only the protocol is to be set.
4660 Note the Exim already has two private options, -pd and -ps, that refer to
4661 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4662 "s", but I don't think that's a major issue.
4664 10. A number of refactoring changes to the code, none of which should affect
4667 (a) The number of logging options was getting close to filling up the
4668 32-bit word that was used as a bit map. I have split them into two classes:
4669 those that are passed in the argument to log_write(), and those that are
4670 only ever tested independently outside of that function. These are now in
4671 separate 32-bit words, so there is plenty of room for expansion again.
4672 There is no change in the user interface or the logging behaviour.
4674 (b) When building, for example, log lines, the code previously used a
4675 macro that called string_cat() twice, in order to add two strings. This is
4676 not really sufficiently general. Furthermore, there was one instance where
4677 it was actually wrong because one of the argument was used twice, and in
4678 one call a function was used. (As it happened, calling the function twice
4679 did not affect the overall behaviour.) The macro has been replaced by a
4680 function that can join an arbitrary number of extra strings onto a growing
4683 (c) The code for expansion conditions now uses a table and a binary chop
4684 instead of a serial search (which was left over from when there were very
4685 few conditions). Also, it now recognizes conditions like "pam" even when
4686 the relevant support is not compiled in: a suitably worded error message is
4687 given if an attempt is made to use such a condition.
4689 11. Added ${time_interval:xxxxx}.
4691 12. A bug was causing one of the ddress fields not to be passed back correctly
4692 from remote delivery subprocesses. The field in question was not being
4693 subsequently used, so this caused to problems in practice.
4695 13. Added new log selectors queue_time and deliver_time.
4697 14. Might have fixed a bug in maildirsizefile handling that threw up
4698 "unexpected character" debug warnings, and recalculated the data
4699 unnecessarily. In any case, I expanded the warning message to give more
4702 15. Added the message "Restricted characters in address" to the statements in
4703 the default ACL that block characters like @ and % in local parts.
4705 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4706 Three changes have been made:
4708 (a) There was a serious bug; a negative response to MAIL caused the whole
4709 recipient domain to be cached as invalid, thereby blocking all messages
4710 to all local parts at the same domain, from all senders. This bug has
4711 been fixed. The domain is no longer cached after a negative response to
4712 MAIL if the sender used is not empty.
4714 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4717 (c) A new callout option, "use_sender" has been added for people who want
4718 the modified behaviour.
4724 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4727 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4728 indeed breaks things for older releases.
4730 3. Added additional logging to the case where there is a problem reading data
4731 from a filter that is running in a subprocess using a pipe, in order to
4732 try to track down a specific problem.
4734 4. Testing facility fudge: when running in the test harness and attempting
4735 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4736 getting "No route to host". Convert this to a timeout.
4738 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4741 6. Some OS don't have socklen_t but use size_t instead. This affects the
4742 fifth argument of getsockopt() amongst other things. This is now
4743 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4744 can be set for individual OS. I have set it for SunOS5, OSF1, and
4745 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4746 some earlier ones do not.
4748 7. Change 4.30/15 was not doing the test caselessly.
4750 8. The standard form for an IPv6 address literal was being rejected by address
4751 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4752 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4753 this, as well as the form without the "IPv6" on the front (but only when
4754 address literals are enabled, of course).
4756 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4758 10. Exim crashed if a message with an empty sender address specified by -f
4759 encountered a router with an errors_to setting. This could be provoked only
4760 by a command such as
4764 where an empty string was supplied; "<>" did not hit this bug.
4766 11. Installed PCRE release 4.5.
4768 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4769 remained set. It is now erased.
4771 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4772 times from message ids (which are base 36 rather than the normal 62).
4774 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4775 were being counted as actual protocol errors, and logged if the log
4776 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4777 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4778 and DATA following a set of rejected RCPTs do not count as protocol errors.
4779 In other words, Exim assumes they were pipelined, though this may not
4780 actually be the case. Of course, in all cases the client gets an
4781 appropriate error code.
4783 15. If a lookup fails in an ACL condition, a message about the failure may
4784 be available; it is used if testing the ACL cannot continue, because most
4785 such messages specify what the cause of the deferral is. However, some
4786 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4787 that caused an old message to be retained and used if a later statement
4788 caused a defer, replacing the real cause of the deferral.
4790 16. If an IP address had so many PTR records that the DNS lookup buffer
4791 was not large enough to hold them, Exim could crash while trying to process
4792 the truncated data. It now detects and logs this case.
4794 17. Further to 4.21/58, another change has been made: if (and only if) the
4795 first line of a message (the first header line) ends with CRLF, a bare LF
4796 in a subsequent header line has a space inserted after it, so as not to
4797 terminate the header.
4799 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4800 unnecessarily, used atoi() instead of strtol(), and didn't check the
4801 termination when getting file sizes from file names by regex.
4803 19. Completely re-implemented the support for maildirsize files, in the light
4804 of a number of problems with the previous contributed implementation
4805 (4.30/29). In particular:
4807 . If the quota is zero, the maildirsize file is maintained, but no quota is
4810 . If the maildir directory does not exist, it is created before any attempt
4811 to write a maildirsize file.
4813 . The quota value in the file is just a cache; if the quota is changed in
4814 the transport, the new value overrides.
4816 . A regular expression is available for excluding directories from the
4819 20. The autoreply transport checks the characters in options that define the
4820 message's headers; it allows continued headers, but it was checking with
4821 isspace() after an embedded newline instead of explicitly looking for a
4824 21. If all the "regular" hosts to which an address was routed had passed their
4825 expiry times, and had not reached their retry times, the address was
4826 bounced, even if fallback hosts were defined. Now Exim should go on to try
4829 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4830 equivalent code in the SMTP transport. Some hosts send humungous responses
4831 to HELO/EHLO, more than 1024 it seems.
4833 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4834 gives compiler warnings in some environments. I've now done it "properly",
4837 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4838 (because of problems with the built-in one) was declared to return uschar *
4839 instead of char *, causing compiler failure.
4841 25. Fixed a file descriptor leak when processing alias/forward files.
4843 26. Fixed a minor format string issue in dbfn.c.
4845 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4847 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4848 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4851 29. When a sender address is verified, it is cached, to save repeating the test
4852 when there is more than one recipient in a message. However, when the
4853 verification involves a callout, it is possible for different callout
4854 options to be set for different recipients. It is too complicated to keep
4855 track of this in the cache, so now Exim always runs a verification when a
4856 callout is required, relying on the callout cache for the optimization.
4857 The overhead is duplication of the address routing, but this should not be
4860 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4861 to be verified with callout=postmaster, and the main callout worked but the
4862 postmaster check failed, the verification correctly failed. However, if a
4863 subsequent RCPT command asked for sender verification *without* the
4864 postmaster check, incorrect caching caused this verification also to fail,
4867 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4868 it was not caching the DNS options (qualify_single, search_parents) that
4869 were used when the lookup failed. A subsequent lookup with different
4870 options therefore always gave the same answer, though there were cases
4871 where it should not have. (Example: a "domains = !$mx_any" option on a
4872 dnslookup router: the "domains" option is always processed without any
4873 widening, but the router might have qualify_single set.) Now Exim uses the
4874 cached value only when the same options are set.
4876 32. Added John Jetmore's "exipick" utility to the distribution.
4878 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4879 than a timeout (e.g. a certificate is required, and is not provided), an
4880 Exim server now closes the connection immediately. Previously it waited for
4881 the client to close - but if the client is SSL, it seems that they each
4882 wait for each other, leading to a delay before one of them times out.
4884 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4885 maintained 0.8.x compatibility because I don't think many are using it, and
4886 it is clearly obsolete.
4888 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4891 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4892 client certificate was expired. A simple patch fixes this, though I don't
4893 understand the full logic of why the verify callback is called multiple
4896 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4897 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4898 which causes problems with some clients (such as the Certicom SSL Plus
4899 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4900 disables the coutermeasure allowing Eudora to connect."
4902 38. Exim was not checking that a write() to a log file succeeded. This could
4903 lead to Bad Things if a log got too big, in particular if it hit a file
4904 size limit. Exim now panics and dies if it cannot write to a log file, just
4905 as it does if it cannot open a log file.
4907 39. Modified OS/Makefile-Linux so that it now contains
4909 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4911 The two -D definitions ensure that Exim is compiled with large file
4912 support, which makes it possible to handle log files that are bigger than
4915 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4916 instance) a domain was checked against a named list that involved a lookup,
4917 causing $domain_data to be set, then another domain was checked against the
4918 same list, then the first domain was re-checked, the value of $domain_data
4919 after the final check could be wrong. In particular, if the second check
4920 failed, it could be set empty. This bug probably also applied to
4923 41. The strip_trailing_dot option was not being applied to the address given
4924 with the -f command-line option.
4926 42. The code for reading a message's header from the spool was incrementing
4927 $received_count, but never initializing it. This meant that the value was
4928 incorrect (doubled) while delivering a message in the same process in which
4929 it was received. In the most common configuration of Exim, this never
4930 happens - a fresh exec is done - but it can happen when
4931 deliver_drop_privilege is set.
4933 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4934 it now includes up to 150 characters of the unexpected data in the log
4937 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4938 and building data strings. The size of both of these buffers was 10 000
4939 bytes - far larger than anybody would *ever* want, thought I. Needless to
4940 say, somebody hit the limit. I have increased the maximum line length to
4941 20 000 and the maximum data length of concatenated lines to 100 000. I have
4942 also fixed two bugs, because there was no checking on these buffers. Tsk,
4943 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4944 buffer is too small.
4946 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4947 lsearch lookups. Now it does.
4949 46. When parsing a route_list item in a manualroute router, a fixed-length
4950 buffer was used for the list of hosts. I made this 1024 bytes long,
4951 thinking that nobody would ever have a list of hosts that long. Wrong.
4952 Somebody had a whole pile of complicated expansion conditions, and the
4953 string was silently truncated, leading to an expansion error. It turns out
4954 that it is easier to change to an unlimited length (owing to other changes
4955 that have happened since this code was originally written) than to build
4956 structure for giving a limitation error. The length of the item that
4957 expands into the list of hosts is now unlimited.
4959 47. The lsearch lookup could not handle data where the length of text line was
4960 more than 4095 characters. Such lines were truncated, leading to shortened
4961 data being returned. It should now handle lines of any length.
4963 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4964 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4967 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4968 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4969 no longer happens in comment lines. A list of replacements is now placed
4970 at the head of all of the source files, except those whose only change is
4971 to replace PERL_COMMAND in the very first #! line.
4973 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4974 messages on the queue, with a bottom-up merge sort, using code contributed
4975 by Michael Haardt. This should make operations like -bp somewhat faster on
4976 large queues. It won't affect queue runners, except when queue_run_in_order
4979 51. Installed eximstats 1.31 in the distribution.
4981 52. Added support for SRV lookups to the dnslookup router.
4983 53. If an ACL referred to $message_body or $message_body_end, the value was not
4984 reset for any messages that followed in the same SMTP session.
4986 54. The store-handling optimization for building very long strings was not
4987 differentiating between the different store pools. I don't think this
4988 actually made any difference in practice, but I've tidied it.
4990 55. While running the routers to verify a sender address, $sender_address
4991 was still set to the sender address. This is wrong, because when routing to
4992 send a bounce to the sender, it would be empty. Therefore, I have changed
4993 it so that, while verifying a sender address, $sender_address is set to <>.
4994 (There is no change to what happens when verifying a recipient address.)
4996 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4997 target A or AAAA records (if not already returned) without resetting the
4998 qualify_single or search_parents options of the DNS resolver. These are
4999 inappropriate in this case because the targets of MX and SRV records must
5000 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5001 target that, when qualified, matched something in the local domain. These
5002 two options are now turned off when doing these lookups.
5004 57. It seems that at least some releases of Reiserfs (which does not have the
5005 concept of a fixed number of inodes) returns zero and not -1 for the
5006 number of available inodes. This interacted badly with check_spool_inodes,
5007 which assumed that -1 was the "no such thing" setting. What I have done is
5008 to check that the total number of inodes is greater than zero before doing
5009 the test of how many are available.
5011 58. When a "warn" ACL statement has a log_message modifier, the message is
5012 remembered, and not repeated. This is to avoid a lot of repetition when a
5013 message has many recipients that cause the same warning to be written.
5014 Howewer, Exim was preserving the list of already written lines for an
5015 entire SMTP session, which doesn't seem right. The memory is now reset if a
5016 new message is started.
5018 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5019 debugging output unless log_rewrite was also set.
5021 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5022 of (int)(handle) when we know that handle contains (void *)(-1).
5024 61. The Exim daemon panic-logs an error return when it closes the incoming
5025 connection. However "connection reset by peer" seems to be common, and
5026 isn't really an error worthy of noting specially, so that particular error
5029 62. When Exim is trying to find all the local interfaces, it used to panic and
5030 die if the ioctl to get the interface flags failed. However, it seems that
5031 on at least one OS (Solaris 9) it is possible to have an interface that is
5032 included in the list of interfaces, but for which you get a failure error
5033 for this call. This happens when the interface is not "plumbed" into a
5034 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5035 failure of the "get flags" call assumes that the interface is down.
5037 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5038 makes life easier for people who are doing arithmetic on fields extracted
5039 from dates, where you often get leading zeros that should not be
5040 interpreted as octal.
5042 64. Added qualify_domain to the redirect router, to override the global
5045 65. If a pathologically long header line contained very many addresses (the
5046 report of this problem mentioned 10 000) and each of them was rewritten,
5047 Exim could use up a very large amount of memory. (It kept on making new
5048 copies of the header line as it rewrote, and never released the old ones.)
5049 At the expense of a bit more processing, the header rewriting function has
5050 been changed so that it no longer eats memory in this way.
5052 66. The generation of the Received: header has been moved from the time that a
5053 message starts to be received, to the time that it finishes. The timestamp
5054 in the Received: header should now be very close to that of the <= log
5055 line. There are two side-effects of this change:
5057 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5058 logged header lines no longer include the local Received: line, because
5059 it has not yet been created. The same applies to a copy of the message
5060 that is returned to a non-SMTP sender when a message is rejected.
5062 (b) When a filter file is tested using -bf, no additional Received: header
5063 is added to the test message. After some thought, I decided that this
5066 This change does not affect the value of $received_for. It is still set
5067 after address rewriting, but before local_scan() is called.
5069 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5071 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5072 gave an unhelpful panic error message, and a defer error. I have managed to
5073 change this behaviour so that it now rejects any supplied certificate,
5074 which seems right, as the list of acceptable certificates is empty.
5076 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5077 gave an unhelpful defer error. I have not managed to make this reject any
5078 supplied certificates, but the error message it gives is "no certificate
5079 supplied", which is not helpful.
5081 70. exigrep's output now also includes lines that are not associated with any
5082 message, but which match the given pattern. Implemented by a patch from
5083 Martin Sluka, which also tidied up the Perl a bit.
5085 71. Recipient callout verification, like sender verification, was using <> in
5086 the MAIL FROM command. This isn't really the right thing, since the actual
5087 sender may affect whether the remote host accepts the recipient or not. I
5088 have changed it to use the actual sender in the callout; this means that
5089 the cache record is now keyed on a recipient/sender pair, not just the
5090 recipient address. There doesn't seem to be a real danger of callout loops,
5091 since a callout by the remote host to check the sender would use <>.
5092 [SEE ABOVE: changed after hitting problems.]
5094 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5095 temporary errors. However, in the case of such a code being given after
5096 the end of a data transmission (i.e. after ".") Exim was failing to write
5097 a retry record for the message. (Yes, there was some broken host that was
5098 actually sending 8xx at this point.)
5100 73. An unknown lookup type in a host list could cause Exim to panic-die when
5101 the list was checked. (An example that provoked this was putting <; in the
5102 middle of a list instead of at the start.) If this happened during a DATA
5103 ACL check, a -D file could be left lying around. This kind of configuration
5104 error no longer causes Exim to die; instead it causes a defer errror. The
5105 incident is still logged to the main and panic logs.
5107 74. Buglet left over from Exim 3 conversion. The message "too many messages
5108 in one connection" was written to the rejectlog but not the mainlog, except
5109 when address rewriting (yes!) was being logged.
5111 75. Added write_rejectlog option.
5113 76. When a system filter was run not as root (that is, when system_filter_user
5114 was set), the values of the $n variables were not being returned to the
5115 main process; thus, they were not subsequently available in the $sn
5118 77. Added +return_path_on_delivery log selector.
5120 78. A connection timeout was being treated differently from recipients deferred
5121 when testing hosts_max_try with a message that was older than the host's
5122 retry timeout. (The host should not be counted, thus allowing all hosts to
5123 be tried at least once before bouncing.) This may have been the cause of an
5124 occasionally reported bug whereby a message would remain on the queue
5125 longer than the retry timeout, but would be bounced if a delivery was
5126 forced. I say "may" because I never totally pinned down the problem;
5127 setting up timeout/retry tests is difficult. See also the next item.
5129 79. The ultimate address timeout was not being applied to errors that involved
5130 a combination of host plus message (for example, a timeout on a MAIL
5131 command). When an address resolved to a number of possible hosts, and they
5132 were not all tried for each delivery (e.g. because of hosts_max_try), a
5133 message could remain on the queue longer than the retry timeout.
5135 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5138 81. Fixed an obscure SMTP outgoing bug which required at least the following
5139 conditions: (a) there was another message waiting for the same server;
5140 (b) the server returned 5xx to all RCPT commands in the first message so
5141 that the message was not completed; (c) the server dropped the connection
5142 or gave a negative response to the RSET that Exim sends to abort the
5143 transaction. The observed case was a dropped connection after DATA that had
5144 been sent in pipelining mode. That is, the server had advertised PIPELINING
5145 but was not implementing it correctly. The effect of the bug was incorrect
5146 behaviour, such as trying another host, and this could lead to a crash.
5152 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5153 and daemon.c were passed as pointers to ints; they should have been
5154 pointers to socklen_t variables (which are typically unsigned ints).
5156 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5159 3. Fixed a really odd bug that affected only the testing scheme; patching a
5160 certain fixed string in the binary changed the value of another string that
5161 happened to be identical to the end of the original first string.
5163 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5164 name", it returns that address as the IP address. On some operating
5165 systems (e.g. Solaris), it also passes back the IP address string as the
5166 "host name". However, on others (e.g. Linux), it passes back an empty
5167 string. Exim wasn't checking for this, and was changing the host name to an
5168 empty string, assuming it had been canonicized.
5170 5. Although rare, it is permitted to have more than one PTR record for a given
5171 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5172 all the names associated with an address, because they do in Solaris.
5173 However, it seems that they do not in Linux for data that comes from the
5174 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5175 I found this out when I moved to a new Linux workstation and tried to run
5176 the Exim test suite.
5178 To get round this problem I have changed the code so that it now does its
5179 own call to the DNS to look up PTR records when searching for a host name.
5180 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5181 addresses that are only in /etc/hosts are still found.
5183 This behaviour is, however, controlled by an option called host_lookup_
5184 order, which defaults to "bydns:byaddr". If people want to use the other
5185 order, or indeed, just use one or the other means of lookup, they can
5186 specify it in this variable.
5188 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5189 some operating systems, this comes back from gethostbyaddr() as an empty
5190 string, and this is what Exim used to test for. However, it seems that in
5191 other systems, "." is yielded. Exim now tests for this case too.
5193 7. The values of check_spool_space and check_log_space are now held internally
5194 as a number of kilobytes instead of an absolute number of bytes. If a
5195 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5196 kilobyte. This means that much larger values can be stored.
5198 8. Exim monitor: an attempt to get the action menu when not actually pointing
5199 at a message produces an empty menu entitled "No message selected". This
5200 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5201 no entries in it ("Shell widget menu has zero width and/or height"). So I
5202 have added a single, blank menu entry in this case.
5204 9. Added ${quote_local_part.
5206 10. MIME decoding is now applied to the contents of Subject: header lines when
5209 11. Now that a reference to $sender_host_address automatically causes a reverse
5210 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5211 host lookup before query-style lookups in lists that might use this
5212 variable. This has therefore been abolished, and the "net-" prefix is no
5213 longer necessary for query-style lookups.
5215 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5216 have been a typo for LFLAGS, so it has been changed.
5218 13. The install script calls Exim with "-C /dev/null" in order to find the
5219 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5220 to be output. Howeve, since Exim outputs its version number before the
5221 error, it didn't break the script. It just looked ugly. I fixed this by
5222 always allowing "-C /dev/null" if the caller is root.
5224 14. Ignore overlarge ACL variable number when reading spool file - insurance
5225 against a later release with more variables having written the file.
5227 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5228 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5229 this, as well as the form without the "IPv6" on the front.
5231 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5232 OS/Makefile-Darwin file.
5234 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5235 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5237 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5238 variable. However, some OS use ints for inodes. I've added cast to long int
5239 to get rid of the compiler warning.
5241 19. I had forgotten to lock out "/../" in configuration file names when
5242 ALT_CONFIG_PREFIX was set.
5244 20. Routers used for verification do not need to specify transports. However,
5245 if such a router generated a host list, and callout was configured, Exim
5246 crashed, because it could not find a port number from the (non-existent)
5247 transport. It now assumes port 25 in this circumstance.
5249 21. Added the -t option to exigrep.
5251 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5252 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5253 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5254 them is set, LOOKUP_LSEARCH is forced.
5256 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5258 24. Added sender and host information to the "rejected by local_scan()" log
5259 line; previously there was no indication of these.
5261 25. Added .include_if_exists.
5263 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5264 Linux. It turns out that not all file systems support this. Apparently some
5265 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5266 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5267 not supported on the file descriptor, is now ignored when Exim is trying to
5268 sync a directory. This applies only to Linux.
5270 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5272 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5273 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5276 29. Added support for maildirsize files from supplied patch (modified a bit).
5278 30. The use of :fail: followed by an empty string could lead Exim to respond to
5279 sender verification failures with (e.g.):
5281 550 Verification failed for <xxx>
5282 550 Sender verify failed
5284 where the first response line was missing the '-' that indicates it is not
5285 the final line of the response.
5287 31. The loop for finding the name of the user that called Exim had a hardwired
5288 limit of 10; it now uses the value of finduser_retries, which is used for
5289 all other user lookups.
5291 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5294 33. Exim was neglecting to zero errno before one call of strtol() when
5295 expanding a string and expecting an integer value. On some systems this
5296 resulted in spurious "integer overflow" errors. Also, it was casting the
5297 result into an int without checking.
5299 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5300 did not work. The code looks as if it has *never* worked, though it appears
5301 to have been documented since at least releast 1.62. I have made it work.
5303 35. The "timeout_DNS" error in retry rules, also documented since at least
5304 1.62, also never worked. As it isn't clear exactly what this means, and
5305 clearly it isn't a major issue, I have abolished the feature by treating it
5306 as "timeout", and writing a warning to the main and panic logs.
5308 36. The display of retry rules for -brt wasn't always showing the error code
5311 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5312 timeout_connect_A, timeout_connect_MX.
5314 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5315 to the empty sender.
5317 39. The daemon was not analysing the content of -oX till after it had closed
5318 stderr and disconnected from the controlling terminal. This meant that any
5319 syntax errors were only noted on the panic log, and the return code from
5320 the command was 0. By re-arranging the code a little, I've made the
5321 decoding happen first, so such errors now appear on stderr, and the return
5322 code is 1. However, the actual setting up of the sockets still happens in
5323 the disconnected process, so errors there are still only recorded on the
5326 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5327 connections (as happens on some IP stacks) was logged at start up time as
5328 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5329 it from "IPv6 and IPv4", which means that two separate sockets are being
5332 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5333 says whether AF_INET or AF_INET6 was passed as an argument.
5335 42. Exiwhat output was messed up when time zones were included in log
5338 43. Exiwhat now gives more information about the daemon's listening ports,
5339 and whether -tls-on-connect was used.
5341 44. The "port" option of the smtp transport is now expanded.
5343 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5344 silently ignored. Now an error message is written to the main and panic
5347 46. There's a new ACL modifier called "logwrite" which writes to a log file
5348 as soon as it is encountered.
5350 47. Added $local_user_uid and $local_user_gid at routing time.
5352 48. Exim crashed when trying to verify a sender address that was being
5355 49. Exim was recognizing only a space character after ".include". It now also
5356 recognizes a tab character.
5358 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5359 extracting the relevant information from the specification. The man page no
5360 longer contains scrambled data for the -d option, and I've added a section
5361 at the front about calling Exim under different names.
5363 51. Added "extra_headers" argument to the "mail" command in filter files.
5365 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5368 53. Installed eximstats 1.29.
5370 54. Added transport_filter_timeout as a generic transport option.
5372 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5373 Cc: header lines. This was required by RFC 822, but it not required by RFC
5376 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5377 incoming messages that did not have them. Now it does so only if the
5378 message originates locally, that is, if there is no associated remote host
5379 address. When Resent- header lines are present, this applies to the Resent-
5380 lines rather than the non-Resent- lines.
5382 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5383 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5385 58. Messages for configuration errors now include the name of the main
5386 configuration file - useful now that there may be more than one file in a
5387 list (.included file names were always shown).
5389 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5390 for those rare installations that do not start the daemon as root or run it
5391 setuid root. I've cut out the call to initgroups() if the daemon is not
5394 60. The Exim user and group can now be bound into the binary as text strings
5395 that are looked up at the start of Exim's processing.
5397 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5399 62. Added $mailstore_basename variable.
5401 63. Installed patch to sieve.c from Michael Haardt.
5403 64. When Exim failed to open the panic log after failing to open the main log,
5404 the original message it was trying to log was written to stderr and debug
5405 output, but if they were not available (the usual case in production), it
5406 was lost. Now it is written to syslog before the two lines that record the
5407 failures to open the logs.
5409 65. Users' Exim filters run in subprocesses under the user's uid. It is
5410 possible for a "deliver" command or an alias in a "personal" command to
5411 provoke an address rewrite. If logging of address rewriting is configured,
5412 this fails because the process is not running as root or exim. There may be
5413 a better way of dealing with this, but for the moment (because 4.30 needs
5414 to be released), I have disabled address rewrite logging when running a
5415 filter in a non-root, non-exim process.
5421 1. The buildconfig auxiliary program wasn't quoting the value set for
5422 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5423 not defined. This bug was masked in 4.22 by the effect that was fixed in
5426 2. Some messages that were rejected after a message id was allocated were
5427 shown as "incomplete" by exigrep. It no longer does this for messages that
5428 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5430 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5431 have allow_domain_literals set, the ID did not get logged in the <= line.
5432 Domain literals are now always recognized in Message-ID: header lines.
5434 4. The first argument for a ${extract expansion item is the key name or field
5435 number. Leading and trailing spaces in this item were not being ignored,
5436 causing some misleading effects.
5438 5. When deliver_drop_privilege was set, single queue runner processes started
5439 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5440 same command in the process it spins off) were not dropping privilege.
5442 6. When the daemon running as "exim" started a queue runner, it always
5443 re-executed Exim in the spun-off process. This is a waste of effort when
5444 deliver_drop_privilege is set. The new process now just calls the
5445 queue-runner function directly.
5451 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5454 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5455 ignored. Though the use of -oP was forcing the writing of a pid file, it
5456 was always written to the default place.
5458 3. If the message "no IP address found for host xxxx" is generated during
5459 incoming verification, it is now followed by identification of the incoming
5460 connection (so you can more easily find what provoked it).
5462 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5464 5. Added some features to "harden" Exim a bit more against certain attacks:
5466 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5467 be put in Local/Makefile. This is like the never_users runtime option,
5468 but it cannot be overridden. The default setting is "root".
5470 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5471 prefix string with which any file named in a -C command line option
5474 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5475 is retained for -C and -D only if the caller of Exim is root. Without
5476 it, the exim user may also use -C and -D and retain privilege.
5478 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5479 command line option is disabled.
5481 6. Macro names set by the -D option must start with an upper case letter, just
5482 like macro names defined in the configuration file.
5484 7. Added "dereference=" facility to LDAP.
5486 8. Two instances of the typo "uknown" in the source files are fixed.
5488 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5489 the Configure-Makefile script screwed up while processing it.
5491 10. Incorporated PCRE 4.4.
5493 11. The SMTP synchronization check was not operating right at the start of an
5494 SMTP session. For example, it could not catch a HELO sent before the client
5495 waited for the greeting. There is now a check for outstanding input at the
5496 point when the greeting is written. Because of the duplex, asynchronous
5497 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5498 way, but not yet received, when the check is performed.
5500 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5501 on TCP/IP sockets, because this apparently causes some broken clients to
5504 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5505 unchanged) from the Cygwin maintainer.
5507 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5508 was supported instead of testing for mbx. Effectively a typo.
5510 15. The spa authenticator server code was not checking that the input it
5511 received was valid base64.
5513 16. The debug output line for the "set" modifier in ACLs was not showing the
5514 name of the variable that was being set.
5516 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5518 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5519 lookup on its own. Something else was needed to trigger the lookup. For
5520 example, a match in host_lookup or the need for a host name in a host list.
5521 Now, if $sender_host_name is referenced and the host name has not yet been
5522 looked up, a lookup is performed. If the lookup fails, the variable remains
5523 empty, and $host_lookup_failed is set to "1".
5525 19. Added "eqi" as a case-independent comparison operator.
5527 20. The saslauthd authentication condition could segfault if neither service
5528 nor realm was specified.
5530 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5531 error message that was logged was misleading, and incoming SMTP
5532 connections were dropped. The message is now more accurate, and temporary
5533 errors are given to SMTP connections.
5535 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5536 (except RSET) with a 421 error, until QUIT is received. However, it was
5537 failing to send a response to QUIT.
5539 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5540 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5541 ACL is now run after the helo_try_verify_hosts code.
5543 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5544 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5545 case-independent, so other case variants are also recognized). Apparently
5546 some systems use these upper case variants.
5548 25. If more than two messages were waiting for the same host, and a transport
5549 filter was specified for the transport, Exim sent two messages over the
5550 same TCP/IP connection, and then failed with "socket operation on non-
5551 socket" when it tried to send the third.
5553 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5555 27. The extern definition of crypt16() in expand.c was not being excluded when
5556 the OS had its own crypt16() function.
5558 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5559 as a preferred synonym for return_size_limit, both as an option and as an
5562 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5564 30. Changed the default configuration ACL to relax the local part checking rule
5565 for addresses that are not in any local domains. For these addresses,
5566 slashes and pipe symbols are allowed within local parts, but the sequence
5567 /../ is explicitly forbidden.
5569 31. SPA server authentication was not clearing the challenge buffer before
5572 32. log_message in a "warn" ACL statement was writing to the reject log as
5573 well as to the main log, which contradicts the documentation and doesn't
5574 seem right (because no rejection is happening). So I have stopped it.
5576 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5577 However, I am unable to do any testing of this.
5579 34. Fixed an infelicity in the appendfile transport. When checking directories
5580 for a mailbox, to see if any needed to be created, it was accidentally
5581 using path names with one or more superfluous leading slashes; tracing
5582 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5584 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5585 contents are added to the log line that is written for every discarded
5586 recipient. (Previously a log_message setting was ignored.)
5588 36. The ${quote: operator now quotes the string if it is empty.
5590 37. The install script runs exim in order to find its version number. If for
5591 some reason other than non-existence or emptiness, which it checks, it
5592 could not run './exim', it was installing it with an empty version number,
5593 i.e. as "exim-". This error state is now caught, and the installation is
5596 38. An argument was missing from the function that creates an error message
5597 when Exim fails to connect to the socket for saslauthd authentication.
5598 This could cause Exim to crash, or give a corrupted message.
5600 39. Added isip, isip4, and isip6 to ${if conditions.
5602 40. The ACL variables $acl_xx are now saved with the message, and can be
5603 accessed later in routers, transports, and filters.
5605 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5606 strings in the file are not string-expanded.
5608 42. If a MAIL command specified a SIZE value that was too large to fit into an
5609 int variable, the check against message_size_limit failed. Such values are
5610 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5611 day this will have to be increased, but I don't think I want to be around
5612 when emails are that large.
5619 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5620 iconv() is not standard in FreeBSD.
5622 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5623 IPv6 enabled. The observed symptom was a segmentation fault on return from
5624 the function os_common_find_running_interfaces() in src/os.c.
5626 3. In the check_special_case() function in daemon.c I had used "errno" as an
5627 argument name, which causes warnings on some systems. This was basically a
5628 typo, since it was named "eno" in the comments!
5630 4. The code that waits for the clock to tick (at a resolution of some fraction
5631 of a second) so as to ensure message-id uniqueness was always waiting for
5632 at least one whole tick, when it could have waited for less. [This is
5633 almost certainly not relevant at current processor speeds, where it is
5634 unlikely to ever wait at all. But we try to future-proof.]
5636 5. The function that sleeps for a time interval that includes fractions of a
5637 second contained a race. It did not block SIGALRM between setting the
5638 timer, and suspending (a couple of lines later). If the interval was short
5639 and the sigsuspend() was delayed until after it had expired, the suspension
5640 never ended. On busy systems this could lead to processes getting stuck for
5643 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5644 process, before it forks any delivery processes. The open lookup caching
5645 mechanism meant that the open file or database connection was passed into
5646 the delivery process. The problem was that delivery processes always tidy
5647 up cached lookup data. This could cause a problem for the next delivery
5648 process started by the queue runner, because the external queue runner
5649 process does not know about the closure. So the next delivery process
5650 still has data in the lookup cache. In the case of a file lookup, there was
5651 no problem because closing a file descriptor in a subprocess doesn't affect
5652 the parent. However, if the lookup was caching a connection to a database,
5653 the connection was closed, and the second delivery process was likely to
5654 see errors such as "PGSQL: query failed: server closed the connection
5655 unexpectedly". The problem has been fixed by closing all cached lookups
5656 in a queue runner before running a delivery process.
5658 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5659 seem to have the "const" qualifier which it has on other OS. I've
5662 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5663 *of the same type* (client or server) with the same public name that an
5664 error should be diagnosed.
5666 9. When Exim looked up a host name for an IP address, but failed to find the
5667 original IP address when looking up the host name (a safety check), it
5668 output the message "<ip address> does not match any IP for NULL", which was
5669 confusing, to say the least. The bug was that the host name should have
5670 appeared instead of "NULL".
5672 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5673 user that is built into the binary, and the -C or -D options is used, root
5674 privilege is dropped before the configuration file is read. In addition,
5675 logging is switched to stderr instead of the normal log files. If the
5676 configuration then re-defines the Exim user, the unprivileged environment
5677 is probably not what is expected, so Exim logs a panic warning message (but
5680 However, if deliver_drop_privilege is set, the unprivileged state may well
5681 be exactly what is intended, so the warning has been cut out in that case,
5682 and Exim is allowed to try to write to its normal log files.
5688 1. smtp_return_error_details was not giving details for temporary sender
5689 or receiver verification errors.
5691 2. Diagnose a configuration error if two authenticators have the same public
5694 3. Exim used not to create the message log file for a message until the first
5695 delivery attempt. This could be confusing when incoming messages were held
5696 for policy or load reasons. The message log file is now created at the time
5697 the message is received, and an initial "Received" line is written to it.
5699 4. The automatically generated man page for command line options had a minor
5700 bug that caused no ill effects; however, a more serious problem was that
5701 the procedure for building the man page automatically didn't always
5702 operate. Consequently, release 4.20 contains an out-of-date version. This
5703 shouldn't happen again.
5705 5. When building Exim with embedded Perl support, the script that builds the
5706 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5707 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5709 6. The freeze_tell option was not being used for messages that were frozen on
5710 arrival, either by an ACL or by local_scan().
5712 7. Added the smtp_incomplete_transaction log selector.
5714 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5715 was accepting AUTH without a new EHLO.
5717 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5718 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5721 10. Logging of TCP/IP connections (when configured) now happens in the main
5722 daemon process instead of the child process, so that the TCP/IP connection
5723 count is more accurate (but it can never be perfect).
5725 11. The use of "drop" in a nested ACL was not being handled correctly in the
5726 outer ACL. Now, if condition failure induced by the nested "drop" causes
5727 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5728 or "require"), the connection is dropped.
5730 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5731 that yield "discard" can now be used with an "accept" or a "discard" verb,
5732 but an error is generated for any others (because I can't see a useful way
5733 to define what should happen).
5735 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5736 are now processed in the same way as lines in the Exim configuration file.
5737 In particular, continuation lines are supported.
5739 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5741 15. Added -ti meaning -t -i.
5743 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5744 domains, and warn by logging if others are found.
5746 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5747 returned from GIFCONF when Exim is trying to find the list of interfaces on
5748 a host. The code in os.c has been modified to copy each ifreq to an aligned
5749 structure in all cases.
5751 Also, in some cases, the returned ifreq's were being copied to a 'struct
5752 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5753 means the last couple of bytes of an IPv6 address could be chopped if the
5754 ifreq contained only a normal sockaddr (14 bytes storage).
5756 18. Named domain lists were not supported in the hosts_treat_as_local option.
5757 An entry such as +xxxx was not recognized, and was treated as a literal
5760 19. Ensure that header lines added by a DATA ACL are included in the reject log
5761 if the ACL subsequently rejects the message.
5763 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5764 MD5 (which is deprecated).
5766 21. When testing a filter file using -bf, Exim was writing a message when it
5767 took the sender from a "From " line in the message, but it was not doing so
5768 when it took $return_path from a Return-Path: header line. It now does.
5770 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5771 with a valid header line field name (a series of printing characters
5772 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5774 23. Changed "disc" in the source to "disk" to conform to the documentation and
5775 the book and for uniformity.
5777 24. Ignore Sendmail's -Ooption=value command line item.
5779 25. When execve() failed while trying to run a command in a pipe transport,
5780 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5781 could be confused with a return value of 69 from the command itself. This
5782 has been changed to 127, the value the shell returns if it is asked to run
5783 a non-existent command. The wording for the related log line suggests a
5784 non-existent command as the problem.
5786 26. If received_header_text expands to an empty string, do not add a Received:
5787 header line to the message. (Well, it adds a token one on the spool, but
5788 marks it "old" so that it doesn't get used or transmitted.)
5790 27. Installed eximstats 1.28 (addition of -nt option).
5792 28. There was no check for failure on the call to getsockname() in the daemon
5793 code. This can fail if there is a shortage of resources on the system, with
5794 ENOMEM, for example. A temporary error is now given on failure.
5796 29. Contrary to the C standard, it seems that in some environments, the
5797 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5798 program. Exim now does this explicitly; it affects the formatting of
5799 timestamps using strftime().
5801 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5802 complaints. I've now initialized all the variables, to avoid this.
5804 32. Header lines added by a system filter were not being "seen" during
5805 transport-time rewrites.
5807 33. The info_callback() function passed to OpenSSL is set up with type void
5808 (*)(SSL *, int, int), as described somewhere. However, when calling the
5809 function (actually a macro) that sets it up, the type void(*)() is
5810 expected. I've put in a cast to prevent warnings from picky compilers.
5812 34. If a DNS black list lookup found a CNAME record, but there were no A
5813 records associated with the domain it pointed at, Exim crashed.
5815 35. If a DNS black list lookup returned more than one A record, Exim ignored
5816 all but the first. It now scans all returned addresses if a particular IP
5817 value is being sought. In this situation, the contents of the
5818 $dnslist_value variable are a list of all the addresses, separated by a
5821 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5822 to accept a host name and all its aliases if the forward lookup for any of
5823 them yielded the IP address of the incoming connection. Now it accepts only
5824 those names whose forward lookup yields the correct IP address. Any other
5825 names are discarded. This closes a loophole whereby a rogue DNS
5826 administrator could create reverse DNS records to break through a
5827 wildcarded host restriction in an ACL.
5829 37. If a user filter or a system filter that ran in a subprocess used any of
5830 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5831 the wrong values were passed to the pipe command ($thisaddress had the
5832 value of $0, $0 had the value of $1, etc). This bug was introduced by
5833 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5835 38. Improved the line breaking for long SMTP error messages from ACLs.
5836 Previously, if there was no break point between 40 and 75 characters, Exim
5837 left the rest of the message alone. Two changes have been made: (a) I've
5838 reduced the minimum length to 35 characters; (b) if it can't find a break
5839 point between 35 and 75 characters, it looks ahead and uses the first one
5840 that it finds. This may give the occasional overlong line, but at least the
5841 remaining text gets split now.
5843 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5844 file descriptors might be low, and that setting 1000 would always raise it.
5845 It turns out that in some environments, the limit is already over 1000 and
5846 that lowering it causes trouble. So now Exim takes care not to decrease it.
5848 40. When delivering a message, the value of $return_path is set to $sender_
5849 address at the start of routing (routers may change the value). By an
5850 oversight, this default was not being set up when an address was tested by
5851 -bt or -bv, which affected the outcome if any router or filter referred to
5854 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5855 log only when "message" or "log_message" are set. However, if one of the
5856 conditions was an address verification, or a call to a nested ACL, the
5857 messages generated by the underlying test were being passed through. This
5858 no longer happens. The underlying message is available in $acl_verify_
5859 message for both "message" and "log_message" expansions, so it can be
5860 passed through if needed.
5862 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5863 new expansion $bh_ to give the encoded byte string without charset
5864 translation. Translation happens only if iconv() is available; HAVE_ICONV
5865 indicates this at build time. HEADERS_CHARSET gives the charset to
5866 translate to; headers_charset can change it in the configuration, and
5867 "headers charset" can change it in an individual filter file.
5869 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5870 that creates RFC 2047 encoded "words" labels them as that charset instead
5871 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5872 expansion operator; (ii) when Exim creates a From: line for a local
5873 message; (iii) when a header line is rewritten to include a "phrase" part.
5875 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5876 buggy, causing it to skip the first lines of messages whose message ID
5877 ended in 'D'. This would not have bitten before Exim release 4.14, because
5878 message IDs were unlikely to end in 'D' before then. The effect was to have
5879 incorrect size information for certain domains.
5881 45. #include "config.h" was missing at the start of the crypt16.c module. This
5882 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5885 46. If there was a timeout during a "random" callout check, Exim treated it as
5886 a failure of the random address, and carried on sending RSET and the real
5887 address. If the delay was just some slowness somewhere, the response to the
5888 original RCPT would be taken as a response to RSET and so on, causing
5889 mayhem of various kinds.
5891 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5892 when I implemented it. It didn't allow for the fact that some option values
5893 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5894 the right test for positive values.
5896 48. Domain names in DNS records are case-independent. Exim always looks them up
5897 in lower case. Some resolvers return domain names in exactly the case they
5898 appear in the zone file, that is, they may contain uppercase letters. Not
5899 all resolvers do this - some return always lower case. Exim was treating a
5900 change of case by a resolver as a change of domain, similar to a widening
5901 of a domain abbreviation. This triggered its re-routing code and so it was
5902 trying to route what was effectively the same domain again. This normally
5903 caused routing to fail (because the router wouldn't handle the domain
5904 twice). Now Exim checks for this case specially, and just changes the
5905 casing of the domain that it ultimately uses when it transmits the message
5908 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5911 50. If a filter generated a file delivery with a non-absolute name (possible if
5912 no home directory exists for the router), the forbid_file option was not
5915 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5916 the existing equality matching.
5918 52. Exim was using ints instead of ino_t variables in some places where it was
5919 dealing with inode numbers.
5921 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5922 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5923 called TMPDIR, and if it finds it is different, it changes its value.
5925 54. The smtp_printf() function is now made available to local_scan() so
5926 additional output lines can be written before returning. There is also an
5927 smtp_fflush() function to enable the detection of a dropped connection.
5928 The variables smtp_input and smtp_batched_input are exported to
5931 55. Changed the default runtime configuration: the message "Unknown user"
5932 has been removed from the ACL, and instead placed on the localuser router,
5933 using the cannot_route_message feature. This means that any verification
5934 failures that generate their own messages won't get overridden. Similarly,
5935 the "Unrouteable address" message that was in the ACL for unverifiable
5936 relay addresses has also been removed.
5938 56. Added hosts_avoid_esmtp to the smtp transport.
5940 57. The exicyclog script was not checking for the esoteric option
5941 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5942 will work only if exicyclog is run under the appropriate euid.
5944 58. Following a discussion on the list, the rules by which Exim recognises line
5945 endings on incoming messages have been changed. The -dropcr and drop_cr
5946 options are now no-ops, retained only for backwards compatibility. The
5947 following line terminators are recognized: LF CRLF CR. However, special
5948 processing applies to CR:
5950 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5951 nor a local message in the state where . is a terminator.
5953 (ii) If a bare CR is encountered in a header line, an extra space is added
5954 after the line terminator so as not to end the header. The reasoning
5955 behind this is that bare CRs in header lines are most likely either
5956 to be mistakes, or people trying to play silly games.
5958 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5959 was being incorrectly given as 18 bytes larger than it should have been.
5960 This is a VOB (very old bug).
5962 60. This may never have affected anything current, but just in case it has:
5963 When the local host is found other than at the start of a list of hosts,
5964 the local host, those with the same MX, and any that follow, are discarded.
5965 When the list in question was part of a longer list of hosts, the following
5966 hosts (not currently being processed) were also being discarded. This no
5967 longer happens. I'm not sure if this situation could ever has previously
5970 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5973 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5974 header. This is something that is recommended in a new Internet Draft, and
5975 is something that is documented as being done by Sendmail. There are two
5976 possible values. For messages generated by the autoreply transport, Exim
5979 Auto-Submitted: auto-replied
5981 whereas for all other generated messages (e.g. bounces) it adds
5983 Auto-Submitted: auto-generated
5985 63. The "personal" condition in filters now includes a test for the
5986 Auto-Submitted: header. If it contains the string "auto-" the message it
5987 not considered personal.
5989 64. Added rcpt_include_affixes as a generic transport option.
5991 65. Added queue_only_override (default true).
5993 66. Added the syslog_duplication option.
5995 67. If what should have been the first header line of a message consisted of
5996 a space followed by a colon, Exim was mis-interpreting it as a header line.
5997 It isn't of course - it is syntactically invalid and should therefore be
5998 treated as the start of the message body. The misbehaviour could have
5999 caused a number of strange effects, including loss of data in subsequent
6000 header lines, and spool format errors.
6002 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6003 client host had authenticated. This control can now be exercised by an ACL
6004 for more flexibility.
6006 69. By default, callouts do not happen when testing with -bh. There is now a
6007 variant, -bhc, which does actually run the callout code, including
6008 consulting and updating the callout cache.
6010 70. Added support for saslauthd authentication, courtesy of Alexander
6013 71. If statvfs() failed on the spool or log directories while checking their
6014 size for availability, Exim confusingly gave the error "space shortage".
6015 Furthermore, in debugging mode it crashed with a floating point exception.
6016 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6017 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6018 serious problem, Exim now writes to the main and panic logs when this
6019 happens, with details of the failure. It then refuses to accept the
6020 incoming message, giving the message "spool directory problem" or "log
6021 directory problem" with a 421 code for SMTP messages.
6023 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6024 0, 1, and 2 exist, because some OS complain for execs without them (see
6025 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6026 descriptors. However, the code omitted to check that the open succeeded,
6027 causing mysterious errors if for some reason the permissions on /dev/null
6028 got screwed. Now Exim writes a message to the main and panic logs, and
6029 bombs out if it can't open /dev/null.
6031 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6032 interact so that it is all more flexible. It is supposed to remain
6033 backwards compatible. Also added extra_local_interfaces.
6035 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6036 to bomb out with an assertion failure - to the client this appears as a
6037 connection drop. This problem occurs in the part of the code that was taken
6038 from the Samba project. Fortunately, the assertion is in a very simple
6039 function, so I have fixed this by reproducing the function inline in the
6040 one place where it is called, and arranging for authentication to fail
6041 instead of killing the process with assert().
6043 75. The SPA client code was not working when the server requested OEM rather
6044 than Unicode encoding.
6046 76. Added code to make require_files with a specific uid setting more usable in
6047 the case where statting the file as root fails - usually a non-root-mounted
6048 NFS file system. When this happens and the failure is EACCES, Exim now
6049 forks a subprocess and does the per-uid checking as the relevant uid.
6051 77. Added process_log_path.
6053 78. If log_file_path was not explicitly set, a setting of check_log_space or
6054 check_log_inodes was ignored.
6056 79. If a space check for the spool or log partitions fails, the incident is now
6057 logged. Of course, in the latter case the data may get lost...
6059 80. Added the %p formatting code to string_format() so that it can be used to
6060 print addresses in debug_print(). Adjusted all the address printing in the
6061 debugging in store.c to use %p rather than %d.
6063 81. There was a concern that a line of code in smtp_in.c could overflow a
6064 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6065 initially expressed, the concern was not well-founded, because trailing
6066 spaces are removed early. However, if the trailing spaces were followed by
6067 a NULL, they did not get removed, so the overflow was possible. Two fixes
6070 (a) I re-wrote the offending code in a cleaner fashion.
6071 (b) If an incoming SMTP command contains a NULL character, it is rejected
6074 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6075 runs initgroups(), so that if the Exim user is in any additional groups,
6076 they will be used during message reception.
6082 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob