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
11 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
12 This was a regression intruduced in 4.83 by another bugfix.
14 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
16 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
17 EXPERIMENTAL_DNS is enabled. Fix from Wolfgang Breyha.
23 TF/01 Correctly close the server side of TLS when forking for delivery.
25 When a message was received over SMTP with TLS, Exim failed to clear up
26 the incoming connection properly after forking off the child process to
27 deliver the message. In some situations the subsequent outgoing
28 delivery connection happened to have the same fd number as the incoming
29 connection previously had. Exim would try to use TLS and fail, logging
30 a "Bad file descriptor" error.
32 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
33 utilities have not been installed.
35 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
36 temporary space as the ACL may create new global variables.
38 TL/01 LDAP support uses per connection or global context settings, depending
39 upon the detected version of the libraries at build time.
41 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
42 to extract and use the src ip:port in logging and expansions as if it
43 were a direct connection from the outside internet. PPv2 support was
44 updated based on HAProxy spec change in May 2014.
46 JH/02 Add ${listextract {number}{list}{success}{fail}}.
48 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
49 Properly escape header and check for NULL return.
51 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
54 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
56 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
57 characters in header names, implemented as a verify condition.
58 Contributed by Michael Fischer v. Mollard.
60 TL/05 Rename SPF condition results err_perm and err_temp to standardized
61 results permerror and temperror. Previous values are deprecated but
62 still accepted. In a future release, err_perm and err_temp will be
63 completely removed, which will be a backward incompatibility if the
64 ACL tests for either of these two old results. Patch contributed by
65 user bes-internal on the mailing list.
67 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
69 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
70 selectors, in both main and reject logs.
72 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
75 JH/07 Add malware type "sock" for talking to simple daemon.
77 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
79 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
80 routers/transports under cutthrough routing.
82 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
83 numbers. Touch up "bool" conditional to keep the same definition.
85 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
87 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
89 JH/12 Expand items in router/transport headers_add or headers_remove lists
90 individually rather than the list as a whole. Bug 1452.
92 Required for reasonable handling of multiple headers_ options when
93 they may be empty; requires that headers_remove items with embedded
94 colons must have them doubled (or the list-separator changed).
96 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
97 view the policy declared in the DMARC record. Currently, $dmarc_status
98 is a combined value of both the record presence and the result of the
101 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
103 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
104 dnslookup router and the smtp transport (applying to the forward
107 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
108 of ldap servers used for a specific lookup. Patch provided by Heiko
111 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
112 New variable $lookup_dnssec_authenticated for observability.
114 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
115 Patch submitted by Lars Timman.
117 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
119 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
120 Requires trusted mode and valid format message id, aborts otherwise.
121 Patch contributed by Heiko Schlichting.
123 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
124 certextract with support for various fields. Bug 1358.
126 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
127 is requested by default, modifiable by smtp transport option
130 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
131 operate on certificate variables to give certificate fingerprints
132 Also new ${sha256:cert_variable}.
134 JH/23 The PRDR feature is moved from being Experimental into the mainline.
136 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
137 Christian Aistleitner.
139 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
141 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
142 file. Patch from Wolfgang Breyha.
144 JH/25 Expand the coverage of the delivery $host and $host_address to
145 client authenticators run in verify callout. Bug 1476.
147 JH/26 Port service names are now accepted for tls_on_connect_ports, to
148 align with daemon_smtp_ports. Bug 72.
150 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
151 support and error reporting did not work properly.
153 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
154 and is readable. Patch from Andrew Colin Kissa.
156 TL/14 Enhance documentation of ${run expansion and how it parses the
157 commandline after expansion, particularly in the case when an
158 unquoted variable expansion results in an empty value.
160 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
162 PP/02 Fix internal collision of T_APL on systems which support RFC3123
163 by renaming away from it. Addresses GH issue 15, reported by
166 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
168 TL/15 SECURITY: prevent double expansion in math comparison functions
169 (can expand unsanitized data). Not remotely exploitable.
176 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
178 PP/02 Make -n do something, by making it not do something.
179 When combined with -bP, the name of an option is not output.
181 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
184 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
185 $sender_host_name and config options to manage this, and basic check
188 PP/05 DSCP support for outbound connections and control modifier for inbound.
190 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
191 (Only plugin which currently uses this is kerberos4, which nobody should
192 be using, but we should make it available and other future plugins might
193 conceivably use it, even though it would break NAT; stuff *should* be
194 using channel bindings instead).
196 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
197 name; added for Sendmail compatibility; requires admin caller.
198 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
199 just ignore it); requires trusted caller.
200 Also parse but ignore: -Ac -Am -X<logfile>
203 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
205 TL/02 Add +smtp_confirmation as a default logging option.
207 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
208 Patch by Magnus Holmgren from 2007-02-20.
210 TL/04 Bugzilla 1281 - Spec typo.
211 Bugzilla 1283 - Spec typo.
212 Bugzilla 1290 - Spec grammar fixes.
214 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
216 TL/06 Add Experimental DMARC support using libopendmarc libraries.
218 TL/07 Fix an out of order global option causing a segfault. Reported to dev
219 mailing list by by Dmitry Isaikin.
221 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
223 JH/02 Support "G" suffix to numbers in ${if comparisons.
225 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
227 NM/01 Bugzilla 1197 - Spec typo
228 Bugzilla 1196 - Spec examples corrections
230 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
232 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
233 gnutls_enable_pkcs11, but renamed to more accurately indicate its
236 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
237 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
239 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
240 "acl {{name}{arg}...}", and optional args on acl condition
243 JH/05 Permit multiple router/transport headers_add/remove lines.
245 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
247 JH/07 Avoid using a waiting database for a single-message-only transport.
248 Performance patch from Paul Fisher. Bugzilla 1262.
250 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
253 JH/09 Add $headers_added variable, with content from use of ACL modifier
254 add_header (but not yet added to the message). Bugzilla 199.
256 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
257 Pulled from Bugzilla 817 by Wolfgang Breyha.
259 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
261 (nb: this is the same fix as in Exim 4.80.1)
263 JH/11 Add A= logging on delivery lines, and a client_set_id option on
266 JH/12 Add optional authenticated_sender logging to A= and a log_selector
269 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
271 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
272 advertise SMTP AUTH mechanism to us, instead of a generic
273 protocol violation error. Also, make Exim more robust to bad
274 data from the Dovecot auth socket.
276 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
278 When a queue runner is handling a message, Exim first routes the
279 recipient addresses, during which it prunes them based on the retry
280 hints database. After that it attempts to deliver the message to
281 any remaining recipients. It then updates the hints database using
284 So if a recipient address works intermittently, it can get repeatedly
285 deferred at routing time. The retry hints record remains fresh so the
286 address never reaches the final cutoff time.
288 This is a fairly common occurrence when a user is bumping up against
289 their storage quota. Exim had some logic in its local delivery code
290 to deal with this. However it did not apply to per-recipient defers
291 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
293 This change adds a proper retry rule check during routing so that the
294 final cutoff time is checked against the message's age. We only do
295 this check if there is an address retry record and there is not a
296 domain retry record; this implies that previous attempts to handle
297 the address had the retry_use_local_parts option turned on. We use
298 this as an approximation for the destination being like a local
299 delivery, as in LMTP.
301 I suspect this new check makes the old local delivery cutoff check
302 redundant, but I have not verified this so I left the code in place.
304 TF/02 Correct gecos expansion when From: is a prefix of the username.
306 Test 0254 submits a message to Exim with the header
310 When I ran the test suite under the user fanf2, Exim expanded
311 the header to contain my full name, whereas it should have added
312 a Resent-Sender: header. It erroneously treats any prefix of the
313 username as equal to the username.
315 This change corrects that bug.
317 GF/01 DCC debug and logging tidyup
318 Error conditions log to paniclog rather than rejectlog.
319 Debug lines prefixed by "DCC: " to remove any ambiguity.
321 TF/03 Avoid unnecessary rebuilds of lookup-related code.
323 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
324 Bug spotted by Jeremy Harris; was flawed since initial commit.
325 Would have resulted in OCSP responses post-SNI triggering an Exim
326 NULL dereference and crash.
328 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
330 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
331 Bug detection, analysis and fix by Samuel Thibault.
332 Bugzilla 1331, Debian bug #698092.
334 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
336 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
337 Server implementation by Todd Lyons, client by JH.
338 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
339 config variable "prdr_enable" controls whether the server
340 advertises the facility. If the client requests PRDR a new
341 acl_data_smtp_prdr ACL is called once for each recipient, after
342 the body content is received and before the acl_smtp_data ACL.
343 The client is controlled by bolth of: a hosts_try_prdr option
344 on the smtp transport, and the server advertisement.
345 Default client logging of deliveries and rejections involving
346 PRDR are flagged with the string "PRDR".
348 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
349 fclose(). Diagnosis by Todd Lyons.
351 PP/17 Update configure.default to handle IPv6 localhost better.
352 Patch by Alain Williams (plus minor tweaks).
355 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
356 This is now consistent with GnuTLS, and is now documented: the
357 previous undocumented portable approach to treating the option as
358 unset was to force an expansion failure. That still works, and
359 an empty string is now equivalent.
361 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
362 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
363 not performing validation itself.
365 PP/20 Added force_command boolean option to pipe transport.
366 Patch from Nick Koston, of cPanel Inc.
368 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
371 TF/04 Added udpsend ACL modifer and hexquote expansion operator
373 PP/21 Fix eximon continuous updating with timestamped log-files.
374 Broken in a format-string cleanup in 4.80, missed when I repaired the
375 other false fix of the same issue.
376 Report and fix from Heiko Schlichting.
379 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
380 Report from Prashanth Katuri.
382 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
383 It's SecureTransport, so affects any MacOS clients which use the
384 system-integrated TLS libraries, including email clients.
386 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
387 using a MIME ACL for non-SMTP local injection.
388 Report and assistance in diagnosis by Warren Baker.
390 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
392 JH/16 Fix comparisons for 64b. Bugzilla 1385.
394 TL/09 Add expansion variable $authenticated_fail_id to keep track of
395 last id that failed so it may be referenced in subsequent ACL's.
397 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
400 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
401 ldap library initialization, allowing self-signed CA's to be
402 used. Also properly sets require_cert option later in code by
403 using NULL (global ldap config) instead of ldap handle (per
404 session). Bug diagnosis and testing by alxgomz.
406 TL/12 Enhanced documentation in the ratelimit.pl script provided in
407 the src/util/ subdirectory.
409 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
410 renamed to Transport Post Delivery Action by Jeremy Harris, as
413 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
414 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
415 redis_servers = needs to be configured which will be used by the redis
416 lookup. Patch from Warren Baker, of The Packet Hub.
418 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
420 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
421 hostname or reverse DNS when processing a host list. Used suggestions
422 from multiple comments on this bug.
424 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
426 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
427 Missed a few lines, added it to make the runtest require no keyboard
430 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
431 contains upper case chars. Make router use caseful_local_part.
433 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
434 support when GnuTLS has been built with p11-kit.
440 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
442 This, or similar/improved, will also be change PP/11 of 4.82.
448 PP/01 Handle short writes when writing local log-files.
449 In practice, only affects FreeBSD (8 onwards).
450 Bugzilla 1053, with thanks to Dmitry Isaikin.
452 NM/01 Bugzilla 949 - Documentation tweak
454 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
457 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
459 PP/02 Implemented gsasl authenticator.
461 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
463 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
464 `pkg-config foo` for cflags/libs.
466 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
467 with rest of GSASL and with heimdal_gssapi.
469 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
470 `pkg-config foo` for cflags/libs for the TLS implementation.
472 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
473 properties get this fed in as external SSF. A number of robustness
474 and debugging improvements to the cyrus_sasl authenticator.
476 PP/08 cyrus_sasl server now expands the server_realm option.
478 PP/09 Bugzilla 1214 - Log authentication information in reject log.
479 Patch by Jeremy Harris.
481 PP/10 Added dbmjz lookup type.
483 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
485 PP/12 MAIL args handles TAB as well as SP, for better interop with
486 non-compliant senders.
487 Analysis and variant patch by Todd Lyons.
489 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
490 Bug report from Lars Müller <lars@samba.org> (via SUSE),
491 Patch from Dirk Mueller <dmueller@suse.com>
493 PP/13 tls_peerdn now print-escaped for spool files.
494 Observed some $tls_peerdn in wild which contained \n, which resulted
495 in spool file corruption.
497 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
498 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
499 or write after TLS renegotiation, which otherwise led to messages
502 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
503 as a tracking header (ie: a signed header comes before the signature).
504 Patch from Wolfgang Breyha.
506 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
507 comma-sep list; embedded commas doubled.
509 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
511 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
513 Report and patch from Dmitry Banschikov.
515 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
516 Removed SSL_clear() after SSL_new() which led to protocol negotiation
517 failures. We appear to now support TLS1.1+ with Exim.
519 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
520 lets Exim select keys and certificates based upon TLS SNI from client.
521 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
522 before an outbound SMTP session. New log_selector, +tls_sni.
524 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
525 NULL dereference. Report and patch from Alun Jones.
527 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
528 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
529 Not seeing resolver debug output on NetBSD, but suspect this is a
530 resolver implementation change.
532 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
533 Left warnings. Added "eximon gdb" invocation mode.
535 PP/21 Defaulting "accept_8bitmime" to true, not false.
537 PP/22 Added -bw for inetd wait mode support.
539 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
540 locate the relevant includes and libraries. Made this the default.
542 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
543 Bugzilla 1246, report and most of solution from Tomasz Kusy.
545 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
546 This may cause build issues on older platforms.
548 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
549 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
550 gnutls_require_mac & gnutls_require_protocols (no longer supported).
551 Added SNI support via GnuTLS too.
552 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
554 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
556 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
557 Applied second patch from Janne, implementing suggestion to default
558 multiple-strings-in-record handling to match SPF spec.
560 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
562 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
563 read-only, out of scope).
564 Patch from Wolfgang Breyha, report from Stuart Northfield.
566 PP/29 Fix three issues highlighted by clang analyser static analysis.
567 Only crash-plausible issue would require the Cambridge-specific
568 iplookup router and a misconfiguration.
569 Report from Marcin Mirosław.
571 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
573 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
574 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
575 As part of this, removing so much warning spew let me fix some minor
576 real issues in debug logging.
578 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
579 assignment on my part. Fixed.
581 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
582 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
583 Janne Snabb (who went above and beyond: thank you).
585 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
586 string otherwise requires a connection and a bunch more work and it's
587 relatively easy to get wrong. Should also expose TLS library linkage
590 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
591 64-bit ${eval} (JH/03).
593 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
594 GNU libc to support some of the 64-bit stuff, should not lead to
595 conflicts. Defined before os.h is pulled in, so if a given platform
596 needs to override this, it can.
598 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
599 protection layer was required, which is not implemented.
600 Bugzilla 1254, patch from Wolfgang Breyha.
602 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
603 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
604 tls_dhparam take prime identifiers. Also unbreak combination of
605 OpenSSL+DH_params+TLSSNI.
607 PP/39 Disable SSLv2 by default in OpenSSL support.
613 PP/01 Solaris build fix for Oracle's LDAP libraries.
614 Bugzilla 1109, patch from Stephen Usher.
616 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
618 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
621 TF/02 Fix a couple more cases where we did not log the error message
622 when unlink() failed. See also change 4.74-TF/03.
624 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
625 lock up or crash if it happened to be inside a call to libc when it
626 got a SIGUSR1 from exiwhat.
628 The SIGUSR1 handler appends the current process status to the process
629 log which is later printed by exiwhat. It used to use the general
630 purpose logging code to do this, but several functions it calls are
631 not safe for signals.
633 The new output code in the SIGUSR1 handler is specific to the process
634 log, and simple enough that it's easy to inspect for signal safety.
635 Removing some special cases also simplifies the general logging code.
636 Removing the spurious timestamps from the process log simplifies
639 TF/04 Improved ratelimit ACL condition.
641 The /noupdate option has been deprecated in favour of /readonly which
642 has clearer semantics. The /leaky, /strict, and /readonly update modes
643 are mutually exclusive. The update mode is no longer included in the
644 database key; it just determines when the database is updated. (This
645 means that when you upgrde Exim will forget old rate measurements.)
647 Exim now checks that the per_* options are used with an update mode that
648 makes sense for the current ACL. For example, when Exim is processing a
649 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
650 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
651 must specify per_mail/readonly. If you omit the update mode it defaults to
652 /leaky where that makes sense (as before) or /readonly where required.
654 The /noupdate option is now undocumented but still supported for
655 backwards compatibility. It is equivalent to /readonly except that in
656 ACLs where /readonly is required you may specify /leaky/noupdate or
657 /strict/noupdate which are treated the same as /readonly.
659 A useful new feature is the /count= option. This is a generalization
660 of the per_byte option, so that you can measure the throughput of other
661 aggregate values. For example, the per_byte option is now equivalent
662 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
664 The per_rcpt option has been generalized using the /count= mechanism
665 (though it's more complicated than the per_byte equivalence). When it is
666 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
667 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
668 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
669 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
670 Note that using per_rcpt with a non-readonly update mode in more than
671 one ACL will cause the recipients to be double-counted. (The per_mail
672 and per_byte options don't have this problem.)
674 The handling of very low rates has changed slightly. If the computed rate
675 is less than the event's count (usually one) then this event is the first
676 after a long gap. In this case the rate is set to the same as this event's
677 count, so that the first message of a spam run is counted properly.
679 The major new feature is a mechanism for counting the rate of unique
680 events. The new per_addr option counts the number of different
681 recipients that someone has sent messages to in the last time period. It
682 behaves like per_rcpt if all the recipient addresses are different, but
683 duplicate recipient addresses do not increase the measured rate. Like
684 the /count= option this is a general mechanism, so the per_addr option
685 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
686 example, measure the rate that a client uses different sender addresses
687 with the options per_mail/unique=$sender_address. There are further
688 details in the main documentation.
690 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
692 TF/06 Removed a few PCRE remnants.
694 TF/07 Automatically extract Exim's version number from tags in the git
695 repository when doing development or release builds.
697 PP/02 Raise smtp_cmd_buffer_size to 16kB.
698 Bugzilla 879. Patch from Paul Fisher.
700 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
701 Heavily based on revision 40f9a89a from Simon Arlott's tree.
704 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
706 PP/05 Variable $av_failed, true if the AV scanner deferred.
707 Bugzilla 1078. Patch from John Horne.
709 PP/06 Stop make process more reliably on build failure.
710 Bugzilla 1087. Patch from Heiko Schlittermann.
712 PP/07 Make maildir_use_size_file an _expandable_ boolean.
713 Bugzilla 1089. Patch from Heiko Schlittermann.
715 PP/08 Handle ${run} returning more data than OS pipe buffer size.
716 Bugzilla 1131. Patch from Holger Weiß.
718 PP/09 Handle IPv6 addresses with SPF.
719 Bugzilla 860. Patch from Wolfgang Breyha.
721 PP/10 GnuTLS: support TLS 1.2 & 1.1.
723 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
726 PP/11 match_* no longer expand right-hand-side by default.
727 New compile-time build option, EXPAND_LISTMATCH_RHS.
728 New expansion conditions, "inlist", "inlisti".
730 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
732 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
734 PP/14 fix log_write() format string regression from TF/03.
735 Bugzilla 1152. Patch from Dmitry Isaikin.
741 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
743 PP/02 Harmonised TLS library version reporting; only show if debugging.
744 Layout now matches that introduced for other libraries in 4.74 PP/03.
746 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
748 PP/04 New "dns_use_edns0" global option.
750 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
753 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
754 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
756 TK/01 Updated PolarSSL code to 0.14.2.
757 Bugzilla 1097. Patch from Andreas Metzler.
759 PP/07 Catch divide-by-zero in ${eval:...}.
762 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
765 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
766 format-string attack -- SECURITY: remote arbitrary code execution.
768 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
769 time unintentionally subject to list matching rules, letting the header
770 cause arbitrary Exim lookups (of items which can occur in lists, *not*
771 arbitrary string expansion). This allowed for information disclosure.
773 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
774 INT_MIN/-1 -- value coerced to INT_MAX.
780 NM/01 Workround for PCRE version dependency in version reporting
783 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
784 This fixes portability to compilers other than gcc, notably
785 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
787 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
788 makefiles for portability to HP-UX and POSIX correctness.
790 PP/01 Permit LOOKUP_foo enabling on the make command-line.
791 Also via indented variable definition in the Makefile.
792 (Debugging by Oliver Heesakkers).
794 PP/02 Restore caching of spamd results with expanded spamd_address.
795 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
797 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
798 Improves build reliability. Fix from: Frank Elsner
800 NM/02 Fix wide character breakage in the rfc2047 coding
801 Fixes bug 1064. Patch from Andrey N. Oktyabrski
803 NM/03 Allow underscore in dnslist lookups
804 Fixes bug 1026. Patch from Graeme Fowler
806 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
807 Code patches from Adam Ciarcinski of NetBSD.
809 NM/04 Fixed exiqgrep to cope with mailq missing size issue
812 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
813 is logged, to avoid truncation. Patch from John Horne.
815 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
816 Patch from Jakob Hirsch.
818 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
819 SQL string expansion failure details.
820 Patch from Andrey Oktyabrski.
822 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
823 Patch from Simon Arlott.
825 PP/09 New lookups functionality failed to compile on old gcc which rejects
826 extern declarations in function scope.
827 Patch from Oliver Fleischmann
829 PP/10 Use sig_atomic_t for flags set from signal handlers.
830 Check getgroups() return and improve debugging.
831 Fixed developed for diagnosis in bug 927 (which turned out to be
834 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
835 Patch from Mark Zealey.
837 PP/12 Bugzilla 1056: Improved spamd server selection.
838 Patch from Mark Zealey.
840 PP/13 Bugzilla 1086: Deal with maildir quota file races.
841 Based on patch from Heiko Schlittermann.
843 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
844 Patch from Uwe Doering, sign-off by Michael Haardt.
846 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
847 variable declaration deep within a block. Bug and patch from
850 PP/15 lookups-Makefile IRIX compatibilty coercion.
852 PP/16 Make DISABLE_DKIM build knob functional.
854 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
855 Patch by Simon Arlott
857 TF/03 Fix valgrind.h portability to C89 compilers that do not support
858 variable argument macros. Our copy now differs from upstream.
864 TF/01 Failure to get a lock on a hints database can have serious
865 consequences so log it to the panic log.
867 TF/02 Log LMTP confirmation messages in the same way as SMTP,
868 controlled using the smtp_confirmation log selector.
870 TF/03 Include the error message when we fail to unlink a spool file.
872 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
873 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
874 for maintaining out-of-tree patches for some time.
876 PP/01 Bugzilla 139: Documentation and portability issues.
877 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
878 Handle per-OS dynamic-module compilation flags.
880 PP/02 Let /dev/null have normal permissions.
881 The 4.73 fixes were a little too stringent and complained about the
882 permissions on /dev/null. Exempt it from some checks.
883 Reported by Andreas M. Kirchwitz.
885 PP/03 Report version information for many libraries, including
886 Exim version information for dynamically loaded libraries. Created
887 version.h, now support a version extension string for distributors
888 who patch heavily. Dynamic module ABI change.
890 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
891 privilege escalation vulnerability whereby the Exim run-time user
892 can cause root to append content of the attacker's choosing to
895 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
898 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
899 If dropping privileges for untrusted macros, we disabled normal logging
900 on the basis that it would fail; for the Exim run-time user, this is not
901 the case, and it resulted in successful deliveries going unlogged.
902 Fixed. Reported by Andreas Metzler.
908 PP/01 Date: & Message-Id: revert to normally being appended to a message,
909 only prepend for the Resent-* case. Fixes regression introduced in
910 Exim 4.70 by NM/22 for Bugzilla 607.
912 PP/02 Include check_rfc2047_length in configure.default because we're seeing
913 increasing numbers of administrators be bitten by this.
915 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
917 PP/03 Bugzilla 994: added openssl_options main configuration option.
919 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
921 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
923 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
925 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
926 without a peer certificate, leading to a segfault because of an
927 assumption that peers always have certificates. Be a little more
928 paranoid. Problem reported by Martin Tscholak.
930 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
931 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
932 NB: ClamAV planning to remove STREAM in "middle of 2010".
933 CL also introduces -bmalware, various -d+acl logging additions and
934 more caution in buffer sizes.
936 PP/09 Implemented reverse_ip expansion operator.
938 PP/10 Bugzilla 937: provide a "debug" ACL control.
940 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
942 PP/12 Bugzilla 973: Implement --version.
944 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
946 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
948 PP/15 Bugzilla 816: support multiple condition rules on Routers.
950 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
951 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
952 ignore trailing whitespace.
954 JJ/02 prevent non-panic DKIM error from being sent to paniclog
956 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
959 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
960 Notification from Dr Andrew Aitchison.
962 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
963 ExtendedDetectionInfo response format.
964 Notification from John Horne.
966 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
969 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
970 XSL and documented dependency on system catalogs, with examples of how
973 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
976 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
977 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
978 configuration file which is writeable by the Exim user or group.
980 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
981 of configuration files to cover files specified with the -C option if
982 they are going to be used with root privileges, not just the default
985 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
986 option (effectively making it always true).
988 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
989 files to be used while preserving root privileges.
991 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
992 that rogue child processes cannot use them.
994 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
995 run-time user, instead of root.
997 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
998 Exim run-time user without dropping privileges.
1000 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1001 result string, instead of calling string_vformat() twice with the same
1004 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1005 for other users. Others should always drop root privileges if they use
1006 -C on the command line, even for a whitelisted configure file.
1008 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1010 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1016 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1017 $data_path, and $header_path variables; fixed documentation bugs and
1020 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1021 exipick to access non-standard spools, including the "frozen" queue
1024 NM/01 Bugzilla 965: Support mysql stored procedures.
1025 Patch from Alain Williams
1027 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1029 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1030 Patch from Andreas Metzler
1032 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1033 Patch from Kirill Miazine
1035 NM/05 Bugzilla 671: Added umask to procmail example.
1037 JJ/03 installed exipick 20100323.0, fixing doc bug
1039 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1040 directory. Notification and patch from Dan Rosenberg.
1042 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1044 TK/02 Improve log output when DKIM signing operation fails.
1046 MH/01 Treat the transport option dkim_domain as a colon separated
1047 list, not as a single string, and sign the message with each element,
1048 omitting multiple occurences of the same signer.
1050 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1051 Bugzilla 985, 986. Patch by Simon Arlott
1053 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1054 Patch by Simon Arlott
1056 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1057 MBX locking. Notification from Dan Rosenberg.
1063 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1065 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1067 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1069 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1071 NM/04 Bugzilla 915: Flag broken perl installation during build.
1077 TK/01 Added patch by Johannes Berg that expands the main option
1078 "spamd_address" if it starts with a dollar sign.
1080 TK/02 Write list of recipients to X-Envelope-Sender header when building
1081 the mbox-format spool file for content scanning (suggested by Jakob
1084 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1085 (http://www.dcc-servers.net/) support via dccifd. Activated by
1086 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1088 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1089 by Mark Daniel Reidel <mr@df.eu>.
1091 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1092 When building exim an external PCRE library is now needed -
1093 PCRE is a system library on the majority of modern systems.
1094 See entry on PCRE_LIBS in EDITME file.
1096 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1097 conversation. Added nologin parameter to request.
1098 Patch contributed by Kirill Miazine.
1100 TF/01 Do not log submission mode rewrites if they do not change the address.
1102 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1104 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1105 log files in place. Contributed by Roberto Lima.
1107 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1109 TF/03 Bugzilla 615: When checking the local_parts router precondition
1110 after a local_part_suffix or local_part_prefix option, Exim now
1111 does not use the address's named list lookup cache, since this
1112 contains cached lookups for the whole local part.
1114 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1115 Robert Millan. Documentation is in experimental-spec.txt.
1117 TF/04 Bugzilla 668: Fix parallel build (make -j).
1119 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1121 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1122 Patch provided by Jan Srzednicki.
1124 TF/05 Leading white space used to be stripped from $spam_report which
1125 wrecked the formatting. Now it is preserved.
1127 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1128 that they are available at delivery time.
1130 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1132 TF/08 TLS error reporting now respects the incoming_interface and
1133 incoming_port log selectors.
1135 TF/09 Produce a more useful error message if an SMTP transport's hosts
1136 setting expands to an empty string.
1138 NM/06 Bugzilla 744: EXPN did not work under TLS.
1139 Patch provided by Phil Pennock.
1141 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1142 Patch provided by Richard Godbee.
1144 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1145 acl_smtp_notquit, added index entry.
1147 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1148 Patch provided by Eugene Bujak.
1150 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1151 accept(). Patch provided by Maxim Dounin.
1153 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1154 Patch provided by Phil Pennock.
1156 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1158 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1159 Patch provided by Brad "anomie" Jorsch.
1161 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1162 Patch provided by Dean Brooks.
1164 TK/05 Add native DKIM support (does not depend on external libraries).
1166 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1167 Patch provided by Graeme Fowler.
1169 NM/16 Bugzilla 851: Documentation example syntax fix.
1171 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1173 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1176 NM/19 Bugzilla 745: TLS version reporting.
1177 Patch provided by Phil Pennock.
1179 NM/20 Bugzilla 167: bool: condition support.
1180 Patch provided by Phil Pennock.
1182 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1183 clients. Patch provided by Phil Pennock.
1185 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1186 Patch provided by Brad "anomie" Jorsch.
1188 NM/23 Bugzilla 687: Fix misparses in eximstats.
1189 Patch provided by Heiko Schlittermann.
1191 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1192 Patch provided by Heiko Schlittermann.
1194 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1195 plus update to original patch.
1197 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1199 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1200 Patch provided by David Brownlee.
1202 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1204 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1206 NM/30 Bugzilla 888: TLS documentation bugfixes.
1208 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1210 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1211 Unlike the original bugzilla I have changed all shell scripts in src tree.
1213 NM/33 Bugzilla 898: Transport filter timeout fix.
1214 Patch by Todd Rinaldo.
1216 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1217 Patch by Serge Demonchaux.
1219 NM/35 Bugzilla 39: Base64 decode bug fixes.
1220 Patch by Jakob Hirsch.
1222 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1224 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1226 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1228 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1234 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1235 ALT-N's libdkim that I have put here:
1236 http://duncanthrax.net/exim-experimental/
1238 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1239 were called 'true' and it seems that C99 defines that as a reserved
1240 keyword to be used with 'bool' variable types. That means you could
1241 not include C99-style headers which use bools without triggering
1242 build errors in sieve.c.
1244 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1245 as mailq or other aliases. Changed the --help handling significantly
1246 to do whats expected. exim_usage() emits usage/help information.
1248 SC/01 Added the -bylocaldomain option to eximstats.
1250 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1252 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1254 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1260 PH/01 Another patch from the Sieve maintainer.
1262 PH/02 When an IPv6 address is converted to a string for single-key lookup
1263 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1264 dots are used instead of colons so that keys in lsearch files need not
1265 contain colons. This was done some time before quoting was made available
1266 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1267 (notated using the quote facility) so as to distinguish them from IPv4
1268 keys. This meant that lookups for IP addresses in host lists did not work
1269 for iplsearch lookups.
1271 This has been fixed by arranging for IPv6 addresses to be expressed with
1272 colons if the lookup type is iplsearch. This is not incompatible, because
1273 previously such lookups could never work.
1275 The situation is now rather anomolous, since one *can* have colons in
1276 ordinary lsearch keys. However, making the change in all cases is
1277 incompatible and would probably break a number of configurations.
1279 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1282 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1283 conversion specification without a maximum field width, thereby enabling
1284 a rogue spamd server to cause a buffer overflow. While nobody in their
1285 right mind would setup Exim to query an untrusted spamd server, an
1286 attacker that gains access to a server running spamd could potentially
1287 exploit this vulnerability to run arbitrary code as the Exim user.
1289 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1290 $primary_hostname instead of what libspf2 thinks the hosts name is.
1292 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1293 a directory entry by the name of the lookup key. Previously, if a
1294 symlink pointed to a non-existing file or a file in a directory that
1295 Exim lacked permissions to read, a lookup for a key matching that
1296 symlink would fail. Now it is enough that a matching directory entry
1297 exists, symlink or not. (Bugzilla 503.)
1299 PH/03 The body_linecount and body_zerocount variables are now exported in the
1302 PH/04 Added the $dnslist_matched variable.
1304 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1305 This means they are set thereafter only if the connection becomes
1308 PH/06 Added the client_condition to authenticators so that some can be skipped
1309 by clients under certain conditions.
1311 PH/07 The error message for a badly-placed control=no_multiline_responses left
1312 "_responses" off the end of the name.
1314 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1316 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1317 (without spaces) instead of just copying the configuration text.
1319 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1321 PH/11 Added $max_received_linelength.
1323 PH/12 Added +ignore_defer and +include_defer to host lists.
1325 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1326 way in which PCRE > 7.0 is built.
1328 PH/14 Implemented queue_only_load_latch.
1330 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1331 MAIL command. The effect was to mangle the value on 64-bit systems.
1333 PH/16 Another patch from the Sieve maintainer.
1335 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1337 PH/18 If a system quota error occurred while trying to create the file for
1338 a maildir delivery, the message "Mailbox is full" was not appended to the
1339 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1340 only to a quota excession during the actual writing of the file.
1342 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1343 characters?) which causes problems in log lines. The DN values are now
1344 passed through string_printing() before being added to log lines.
1346 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1347 and InterBase are left for another time.)
1349 PH/21 Added message_body_newlines option.
1351 PH/22 Guard against possible overflow in moan_check_errorcopy().
1353 PH/23 POSIX allows open() to be a macro; guard against that.
1355 PH/24 If the recipient of an error message contained an @ in the local part
1356 (suitably quoted, of course), incorrect values were put in $domain and
1357 $local_part during the evaluation of errors_copy.
1363 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1364 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1367 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1368 issue a MAIL command.
1370 PH/02 In an ACL statement such as
1372 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1374 if a client was not listed at all, or was listed with a value other than
1375 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1376 the condition was not true (as it should be), so access was not denied.
1377 The bug was that the ! inversion was incorrectly passed on to the second
1378 item. This has been fixed.
1380 PH/03 Added additional dnslists conditions == and =& which are different from
1381 = and & when the dns lookup returns more than one IP address.
1383 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1384 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1386 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1387 FSYNC, which compiles an option called disable_fsync that allows for
1388 bypassing fsync(). The documentation is heavily laced with warnings.
1390 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1392 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1393 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1394 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1395 including adding "make clean"; (3) Added -fPIC when compiling the test
1396 dynamically loaded module, to get rid of a warning.
1398 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1399 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1400 = 0s. The bug is otherwise harmless.
1402 PH/07 There was a bug in the dovecot authenticator such that the value of
1403 $auth1 could be overwritten, and so not correctly preserved, after a
1404 successful authentication. This usually meant that the value preserved by
1405 the server_setid option was incorrect.
1407 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1409 PH/09 Installed PCRE release 7.0.
1411 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1412 run for batched SMTP input. It is now run at the start of every message
1413 in the batch. While fixing this I discovered that the process information
1414 (output by running exiwhat) was not always getting set for -bs and -bS
1415 input. This is fixed, and it now also says "batched" for BSMTP.
1417 PH/11 Added control=no_pipelining.
1419 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1420 patch, slightly modified), and move the expansion of helo_data till after
1421 the connection is made in the smtp transport (so it can use these
1424 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1426 PH/14 Added log_selector = +pid.
1428 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1430 PH/16 Add ${if forany and ${if forall.
1432 PH/17 Added dsn_from option to vary the From: line in DSNs.
1434 PH/18 Flush SMTP output before performing a callout, unless control =
1435 no_callout_flush is set.
1437 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1438 was true (the default) a successful delivery failed to delete the retry
1439 item, thus causing premature timeout of the address. The bug is now
1442 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1444 PH/21 Long custom messages for fakedefer and fakereject are now split up
1445 into multiline reponses in the same way that messages for "deny" and
1446 other ACL rejections are.
1448 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1449 with slight modification.
1451 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1452 draft, changing the syntax and factoring some duplicate code".
1454 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1455 for deliveries of the second and subsequent messages over the same SMTP
1458 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1459 ${reduce, with only minor "tidies".
1461 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1463 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1464 expansion side effects.
1466 PH/27 When a message times out after an over-quota error from an Exim-imposed
1467 quota, the bounce message says "mailbox is full". This message was not
1468 being given when it was a system quota that was exceeded. It now should
1471 MH/03 Made $recipients available in local_scan(). local_scan() already has
1472 better access to the recipient list through recipients_list[], but
1473 $recipients can be useful in postmaster-provided expansion strings.
1475 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1476 in the case of a MAIL command with additional options following the
1477 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1478 were accidentally chopped off.
1480 PH/29 SMTP synchronization checks are implemented when a command is read -
1481 there is a check that no more input is waiting when there shouldn't be
1482 any. However, for some commands, a delay in an ACL can mean that it is
1483 some time before the response is written. In this time, more input might
1484 arrive, invalidly. So now there are extra checks after an ACL has run for
1485 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1486 pipelining has not been advertised.
1488 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1490 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1491 returned in the SMTP response when smtp_return_error_details was false.
1492 This has been fixed.
1494 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1495 instead of the C I/O that was originally supplied, because problems were
1496 reported on Solaris.
1498 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1499 Exim which did not show up earlier: it was assuming that a call to
1500 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1501 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1502 was a macro that became an assignment, so it seemed to work. This has
1503 changed to a proper function call with a void return, hence the compile
1504 error. Exim's code has been fixed.
1506 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1509 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1511 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1513 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1514 because it turns out to be too big or there is a timeout) there is no
1515 "Completed" line in the log. When some messages of this type were
1516 selected by exigrep, they were listed as "not completed". Others were
1517 picked up by some special patterns. I have improved the selection
1518 criteria to be more general.
1520 PH/38 The host_find_failed option in the manualroute router can now be set
1521 to "ignore", to completely ignore a host whose IP address cannot be
1522 found. If all hosts are ignored, the behaviour is controlled by the new
1523 host_all_ignored option.
1525 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1526 homing or because of multiple MX records with /mx) generated more than
1527 one IP address, and the following item turned out to be the local host,
1528 all the secondary addresses of the first item were incorrectly removed
1529 from the list, along with the local host and any following hosts (which
1530 is what is supposed to happen).
1532 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1533 whoever called Exim into the -H file. In the case of the daemon it was
1534 behaving confusingly. When first started, it used values for whoever
1535 started the daemon, but after a SIGHUP it used the Exim user (because it
1536 calls itself on a restart). I have changed the code so that it now always
1539 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1540 message are rejected with the same error (e.g. no authentication or bad
1541 sender address), and a DATA command is nevertheless sent (as can happen
1542 with PIPELINING or a stupid MUA), the error message that was given to the
1543 RCPT commands is included in the rejection of the DATA command. This is
1544 intended to be helpful for MUAs that show only the final error to their
1547 PH/42 Another patch from the Sieve maintainer.
1549 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1550 Eximstats - Fixed some broken HTML links and added missing column headers
1552 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1553 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1555 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1557 PH/43 Yet another patch from the Sieve maintainer.
1559 PH/44 I found a way to check for a TCP/IP connection going away before sending
1560 the response to the final '.' that terminates a message, but only in the
1561 case where the client has not sent further data following the '.'
1562 (unfortunately, this is allowed). However, in many cases there won't be
1563 any further data because there won't be any more messages to send. A call
1564 to select() can be used: if it shows that the input is "ready", there is
1565 either input waiting, or the socket has been closed. An attempt to read
1566 the next input character can distinguish the two cases. Previously, Exim
1567 would have sent an OK response which the client would never have see.
1568 This could lead to message repetition. This fix should cure that, at
1569 least in a lot of common cases.
1571 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1572 advertised in response to EHLO.
1578 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1579 fixed by 4.65/MH/01 (is this a record?) are fixed:
1581 (i) An empty string was always treated as zero by the numeric comparison
1582 operators. This behaviour has been restored.
1584 (ii) It is documented that the numeric comparison operators always treat
1585 their arguments as decimal numbers. This was broken in that numbers
1586 starting with 0 were being interpreted as octal.
1588 While fixing these problems I realized that there was another issue that
1589 hadn't been noticed. Values of message_size_limit (both the global option
1590 and the transport option) were treated as octal if they started with 0.
1591 The documentation was vague. These values are now always treated as
1592 decimal, and I will make that clear in the documentation.
1598 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1599 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1602 MH/01 Don't check that the operands of numeric comparison operators are
1603 integers when their expansion is in "skipping" mode (fixes bug
1604 introduced by 4.64-PH/07).
1606 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1607 child addresses, Exim now panics and dies. Previously, because the count
1608 is held in a short int, deliveries were likely to be lost. As such a
1609 large number of recipients for a single message is ridiculous
1610 (performance will be very, very poor), I have chosen to impose a limit
1611 rather than extend the field.
1617 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1618 leftover -K file (the existence of which was triggered by #402).
1619 While we were at it, introduced process PID as part of the -K
1620 filename. This should rule out race conditions when creating
1623 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1624 processing considerably. Previous code took too long for large mails,
1625 triggering a timeout which in turn triggers #401.
1627 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1628 in the DK code in transports.c. sendfile() is not really portable,
1629 hence the _LINUX specificness.
1631 TF/01 In the add_headers option to the mail command in an Exim filter,
1632 there was a bug that Exim would claim a syntax error in any
1633 header after the first one which had an odd number of characters
1636 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1637 callout verification, Exim cached a "reject" for the entire domain. This
1638 is correct for most verifications, but it is not correct for a recipient
1639 verification with use_sender or use_postmaster set, because in that case
1640 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1641 case of MAIL FROM:<> rejection from other early rejections (e.g.
1642 rejection of HELO). When verifying a recipient using a non-null MAIL
1643 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1644 Whatever the result of the callout, the value of the domain cache is
1645 left unchanged (for any other kind of callout, getting as far as trying
1646 RCPT means that the domain itself is ok).
1648 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1651 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1652 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1653 session, the daemon ignores ECONNECT errors and logs others; it now
1654 ignores EPIPE as well.
1656 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1657 (quoted-printable decoding).
1659 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1660 later the small subsequent patch to fix an introduced bug.
1662 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1664 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1666 PH/08 An error is now given if message_size_limit is specified negative.
1668 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1669 to be given (somewhat) arbitrary names.
1671 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1674 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1675 miscellaneous code fixes
1677 PH/10 Added the log_reject_target ACL modifier to specify where to log
1680 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1681 hostname. This is wrong, because it relates to the incoming message (and
1682 probably the interface on which it is arriving) and not to the outgoing
1683 callout (which could be using a different interface). This has been
1684 changed to use the value of the helo_data option from the smtp transport
1685 instead - this is what is used when a message is actually being sent. If
1686 there is no remote transport (possible with a router that sets up host
1687 addresses), $smtp_active_hostname is used.
1689 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1690 tweaks were necessary in order to get it to work (see also 21 below):
1691 (a) The code assumed that strncpy() returns a negative number on buffer
1692 overflow, which isn't the case. Replaced with Exim's string_format()
1694 (b) There were several signed/unsigned issues. I just did the minimum
1695 hacking in of casts. There is scope for a larger refactoring.
1696 (c) The code used strcasecmp() which is not a standard C function.
1697 Replaced with Exim's strcmpic() function.
1698 (d) The code set only $1; it now sets $auth1 as well.
1699 (e) A simple test gave the error "authentication client didn't specify
1700 service in request". It would seem that Dovecot has changed its
1701 interface. Fortunately there's a specification; I followed it and
1702 changed what the client sends and it appears to be working now.
1704 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1707 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1708 address A is aliased to B and C, where B exists and C does not. Without
1709 -v the output is "A verified" because verification stops after a
1710 successful redirection if more than one address is generated. However,
1711 with -v the child addresses are also verified. Exim was outputting "A
1712 failed to verify" and then showing the successful verification for C,
1713 with its parentage. It now outputs "B failed to verify", showing B's
1714 parentage before showing the successful verification of C.
1716 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1717 look up a TXT record in a specific list after matching in a combined
1720 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1721 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1722 they consult the DNS. I had assumed they would set it the way they
1723 wanted; and indeed my experiments on Linux seem to show that in some
1724 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1725 To be on the safe side, however, I have now made the interface to
1726 host_find_byname() similar to host_find_bydns(), with an argument
1727 containing the DNS resolver options. The host_find_byname() function now
1728 sets these options at its start, just as host_find_bydns() does. The smtp
1729 transport options dns_qualify_single and dns_search_parents are passed to
1730 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1731 of host_find_byname() use the default settings of RES_DEFNAMES
1732 (qualify_single) but not RES_DNSRCH (search_parents).
1734 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1735 spool_read_header() do less string testing, by means of a preliminary
1736 switch on the second character of optional "-foo" lines. (This is
1737 overdue, caused by the large number of possibilities that now exist.
1738 Originally there were few.) While I was there, I also converted the
1739 str(n)cmp tests so they don't re-test the leading "-" and the first
1740 character, in the hope this might squeeze out yet more improvement.
1742 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1743 flag allowing group syntax was set by the header_syntax check but not
1744 turned off, possible causing trouble later; (2) The flag was not being
1745 set at all for the header_verify test, causing "group"-style headers to
1746 be rejected. I have now set it in this case, and also caused header_
1747 verify to ignore an empty address taken from a group. While doing this, I
1748 came across some other cases where the code for allowing group syntax
1749 while scanning a header line wasn't quite right (mostly, not resetting
1750 the flag correctly in the right place). These bugs could have caused
1751 trouble for malformed header lines. I hope it is now all correct.
1753 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1754 with the "reply" argument non-NULL. The code, however (which originally
1755 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1756 but it didn't always do it. This confused somebody who was copying the
1757 code for some other use. I have removed all the tests.
1759 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1760 feature that was used to support insecure browsers during the U.S. crypto
1761 embargo. It requires special client support, and Exim is probably the
1762 only MTA that supported it -- and would never use it because real RSA is
1763 always available. This code has been removed, because it had the bad
1764 effect of slowing Exim down by computing (never used) parameters for the
1765 RSA_EXPORT functionality.
1767 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1768 authenticator to fail if there's a tab character in the incoming data
1769 (there should never be unless someone is messing about, as it's supposed
1770 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1771 if the connection is using TLS or if the remote IP is the same as the
1772 local IP, and the "valid-client-cert option" if a client certificate has
1775 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1776 authenticators. This can be used for authorization after authentication
1777 succeeds. (In the case of plaintext, it servers for both authentication
1780 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1781 if any retry times were supplied.
1783 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1784 connection, where there is no client IP address to check. In this
1785 situation, the verify now always succeeds.
1787 PH/25 Applied John Jetmore's -Mset patch.
1789 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1791 PH/27 In a string expansion for a processed (not raw) header when multiple
1792 headers of the same name were present, leading whitespace was being
1793 removed from all of them, but trailing whitespace was being removed only
1794 from the last one. Now trailing whitespace is removed from each header
1795 before concatenation. Completely empty headers in a concatenation (as
1796 before) are ignored.
1798 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1799 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1801 PH/29 [Removed. This was a change that I later backed out, and forgot to
1802 correct the ChangeLog entry (that I had efficiently created) before
1803 committing the later change.]
1805 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1806 address errors (4xx response to RCPT) over the same connection as other
1807 messages routed to the same hosts. Such deliveries are always "forced",
1808 so retry times are not inspected. This resulted in far too many retries
1809 for the affected addresses. The effect occurred only when there were more
1810 hosts than the hosts_max_try setting in the smtp transport when it had
1811 the 4xx errors. Those hosts that it had tried were not added to the list
1812 of hosts for which the message was waiting, so if all were tried, there
1813 was no problem. Two fixes have been applied:
1815 (i) If there are any address or message errors in an SMTP delivery, none
1816 of the hosts (tried or untried) are now added to the list of hosts
1817 for which the message is waiting, so the message should not be a
1818 candidate for sending over the same connection that was used for a
1819 successful delivery of some other message. This seems entirely
1820 reasonable: after all the message is NOT "waiting for some host".
1821 This is so "obvious" that I'm not sure why it wasn't done
1822 previously. Hope I haven't missed anything, but it can't do any
1823 harm, as the worst effect is to miss an optimization.
1825 (ii) If, despite (i), such a delivery is accidentally attempted, the
1826 routing retry time is respected, so at least it doesn't keep
1827 hammering the server.
1829 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1830 in ${readsocket because some servers need this prod.
1832 PH/32 Added some extra debug output when updating a wait-xxx database.
1834 PH/33 The hint "could be header name not terminated by colon", which has been
1835 given for certain expansion errors for a long time, was not being given
1836 for the ${if def:h_colon_omitted{... case.
1838 PH/34 The spec says: "With one important exception, whenever a domain list is
1839 being scanned, $domain contains the subject domain." There was at least
1840 one case where this was not true.
1842 PH/35 The error "getsockname() failed: connection reset by peer" was being
1843 written to the panic log as well as the main log, but it isn't really
1844 panic-worthy as it just means the connection died rather early on. I have
1845 removed the panic log writing for the ECONNRESET error when getsockname()
1848 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1849 runs only) independently of the message's sender address. This meant
1850 that, if the 4xx error was in fact related to the sender, a different
1851 message to the same recipient with a different sender could confuse
1852 things. In particualar, this can happen when sending to a greylisting
1853 server, but other circumstances could also provoke similar problems.
1854 I have changed the default so that the retry time for these errors is now
1855 based a combination of the sender and recipient addresses. This change
1856 can be overridden by setting address_retry_include_sender=false in the
1859 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1860 remote server are returned as part of bounce messages. This was not
1861 happening for LMTP over a pipe (the lmtp transport), but now it is the
1862 same for both kinds of LMTP.
1864 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1865 in header lines that were in fact CNAMEs. This is no longer the case.
1867 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1868 and queue runs started by the daemon processed all messages. This has
1869 been fixed so that -R and -S can now usefully be given with -q<time>.
1871 PH/40 Import PCRE release 6.7 (fixes some bugs).
1873 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1875 PH/42 Give an error if -q is specified more than once.
1877 PH/43 Renamed the variables $interface_address and $interface_port as
1878 $received_ip_address and $received_port, to make it clear that these
1879 values apply to message reception, and not to the outgoing interface when
1880 a message is delivered. (The old names remain recognized, of course.)
1882 PH/44 There was no timeout on the connect() call when using a Unix domain
1883 socket in the ${readsocket expansion. There now is.
1885 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1886 be meaningful with "accept".
1888 SC/01 Eximstats V1.43
1889 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1891 SC/02 Eximstats V1.44
1892 Use a glob alias rather than an array ref in the generated
1893 parser. This improves both readability and performance.
1895 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1896 Collect SpamAssassin and rejection statistics.
1897 Don't display local sender or destination tables unless
1898 there is data to show.
1899 Added average volumes into the top table text output.
1901 SC/04 Eximstats V1.46
1902 Collect data on the number of addresses (recipients)
1903 as well as the number of messages.
1905 SC/05 Eximstats V1.47
1906 Added 'Message too big' to the list of mail rejection
1907 reasons (thanks to Marco Gaiarin).
1909 SC/06 Eximstats V1.48
1910 Mainlog lines which have GMT offsets and are too short to
1911 have a flag are now skipped.
1913 SC/07 Eximstats V1.49 (Alain Williams)
1914 Added the -emptyok flag.
1916 SC/08 Eximstats V1.50
1917 Fixes for obtaining the IP address from reject messages.
1919 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1920 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1921 whitesspace changes from 4.64-PH/27
1923 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1926 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1927 are found, allow negative numbers in numeric criteria)
1929 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1931 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1932 to match changes made in 4.64-PH/43
1934 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1936 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1937 "domain" is actually an IP address, and force a failure. This locks out
1938 those revolvers/nameservers that support "A-for-A" lookups, in
1939 contravention of the specifications.
1941 PH/48 When a host name was looked up from an IP address, and the subsequent
1942 forward lookup of the name timed out, the host name was left in
1943 $sender_host_name, contrary to the specification.
1945 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1946 restricted to single-key lookups, Exim was not diagnosing an error if
1947 * or *@ was used with a query-style lookup.
1949 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1951 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1952 long ago, but noone interested enough thought of it. Let's just say that
1953 the "1.1" means that there are some new functions that weren't there at
1954 some point in the past.
1956 PH/51 Error processing for expansion failure of helo_data from an smtp
1957 transport during callout processing was broken.
1959 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1960 tested/used via the -bh/-bhc/-bs options.
1962 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1963 bug, fixed in subsequent PCRE releases).
1965 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1966 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1968 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1974 SC/01 Use a glob alias rather than an array ref in eximstats generated
1975 parser. This improves both readability and performance.
1977 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1978 Don't display local sender or destination tables in eximstats unless
1979 there is data to show.
1980 Added average volumes into the eximstats top table text output.
1982 SC/03 Collect data on the number of addresses (recipients) as well
1983 as the number of messages in eximstats.
1985 TF/01 Correct an error in the documentation for the redirect router. Exim
1986 does (usually) call initgroups() when daemonizing.
1988 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1989 with consistent privilege compared to when running as a daemon.
1991 TF/03 Note in the spec that $authenticated_id is not set for local
1992 submissions from trusted users.
1994 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1995 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1997 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1998 by adding some example configuration directives to the default
1999 configuration file. A little bit of work is required to uncomment the
2000 directives and define how usernames and passwords are checked, but
2001 there is now a framework to start from.
2003 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2004 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2005 without this. I don't know how relevant this is to other LDAP libraries.
2007 PH/02 Add the verb name to the "unknown ACL verb" error.
2009 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2011 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2013 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2014 directory not expanded when it should be if an expanded home directory
2015 was set for the address (which is overridden by the transport).
2017 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2020 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2021 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2022 because it is too late at that time, and has no effect.
2024 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2025 security issue with \' (bugzilla #107). I could not use the
2026 PQescapeStringConn() function, because it needs a PGconn value as one of
2029 PH/08 When testing addresses using -bt, indicate those final addresses that
2030 are duplicates that would not cause an additional delivery. At least one
2031 person was confused, thinking that -bt output corresponded to deliveries.
2032 (Suppressing duplicates isn't a good idea as you lose the information
2033 about possibly different redirections that led to the duplicates.)
2035 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2036 systems where poll() doesn't work, in particular OS X.
2038 PH/10 Added more information to debugging output for retry time not reached.
2040 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2041 operations in malware.c.
2043 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2046 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2047 syslog_duplication set false, log lines that would normally be written
2048 both the the main log and to the reject log were not written to syslog at
2051 PH/14 In the default configuration, change the use of "message" in ACL warn
2052 statements to "add_header".
2054 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2055 not followed by a command (e.g. "seen endif").
2057 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2058 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2061 PH/17 Added extra conditions to the default value of delay_warning_condition
2065 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2066 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2067 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2070 The Auto-Submitted: and various List- headers are standardised, whereas I
2071 don't think Precedence: ever was.
2073 PH/18 Refactored debugging code in route_finduser() to show more information,
2074 in particular, the error code if getpwnam() issues one.
2076 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2077 This is apparently needed in addition to the PH/07 change above to avoid
2078 any possible encoding problems.
2080 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2081 but not after initializing Perl.
2083 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2084 output them only if debugging. By default they are written stderr,
2085 apparently, which is not desirable.
2087 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2090 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2093 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2095 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2096 authenticated or an ident call has been made. Suppress the default
2097 values for $authenticated_id and $authenticated_sender (but permit -oMai
2098 and -oMas) when testing with -bh.
2100 PH/24 Re-jigged the order of the tests in the default configuration so that the
2101 tests for valid domains and recipients precede the DNS black list and CSA
2102 tests, on the grounds that those ones are more expensive.
2104 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2105 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2106 This bug exists in every version of Exim that I still have, right back to
2109 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2110 However, an attempt to turn on case-sensitivity in a regex key by
2111 including (?-i) didn't work because the subject string was already
2112 lowercased, and the effects were non-intuitive. It turns out that a
2113 one-line patch can be used to allow (?-i) to work as expected.
2119 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2120 other effects) broke the use of negated acl sub-conditions.
2122 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2125 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2126 "Deny" causes Exim to reject the incoming connection with a 554 error.
2127 Unfortunately, if there is a major crisis, such as a disk failure,
2128 tcp-wrappers gives "deny", whereas what one would like would be some
2129 kind of temporary error. A kludge has been added to help with this.
2130 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2131 554 error is used if errno is still zero or contains ENOENT (which occurs
2132 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2135 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2137 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2138 errors. Otherwise a message that provokes a temporary error (when other
2139 messages do not) can cause a whole host to time out.
2141 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2142 addresses were routed directly to files or pipes from a redirect router.
2143 File deliveries just didn't batch; pipe deliveries might have suffered
2146 PH/06 A failure to get a lock for a hints database would erroneously always say
2147 "Failed to get write lock", even when it was really a read lock.
2149 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2150 of 0600. This has been changed to use the value of the lockfile_mode
2151 option (which defaults to 0600).
2153 PH/08 Applied small patch from the Sieve maintainer.
2155 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2156 folder from quota calculations, a direct delivery into this folder messed
2157 up the contents of the maildirsize file. This was because the regex was
2158 used only to exclude .Trash (or whatever) when the size of the mailbox
2159 was calculated. There was no check that a delivery was happening into an
2160 excluded directory. This bug has been fixed by ignoring all quota
2161 processing for deliveries into excluded directories.
2163 PH/10 Added the maildirfolder_create_regex option to appendfile.
2169 PH/01 The code for finding all the local interface addresses on a FreeBSD
2170 system running IPv6 was broken. This may well have applied to all BSD
2171 systems, as well as to others that have similar system calls. The broken
2172 code found IPv4 interfaces correctly, but gave incorrect values for the
2173 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2174 that it would not match correctly against @[] and not recognize the IPv6
2177 PH/02 The ipliteral router was not recognizing addresses of the form user@
2178 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2180 PH/03 Added disable_ipv6.
2182 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2183 decoded form, because it is most often used to construct To: headers
2184 lines in autoreplies, and the decoded form may well be syntactically
2185 invalid. However, $reply_address has leading white space removed, and all
2186 newlines turned into spaces so that the autoreply transport does not
2189 PH/05 If group was specified without a user on a router, and no group or user
2190 was specified on a transport, the group from the router was ignored.
2192 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2193 for visible compile-time settings that can be used to change these
2194 numbers, for those that want even more. Backwards compatibility with old
2195 spool files has been maintained. However, going back to a previous Exim
2196 release will lost any variables that are in spool files.
2198 PH/07 Two small changes when running in the test harness: increase delay when
2199 passing a TCP/IP connection to a new process, in case the original
2200 process has to generate a bounce, and remove special handling of
2201 127.0.0.2 (sic), which is no longer necessary.
2203 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2204 be the same on different OS.
2206 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2209 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2210 whether --show-vars was specified or not
2212 JJ/02 exipick: Added support for new ACL variable spool format introduced
2215 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2216 syntactically invalid From: or Reply-to: line, and a filter used this to
2217 generate an autoreply, and therefore failed to obtain an address for the
2218 autoreply, Exim could try to deliver to a non-existent relative file
2219 name, causing unrelated and misleading errors. What now happens is that
2220 it logs this as a hard delivery error, but does not attempt to create a
2223 PH/11 The exinext utility has a -C option for testing purposes, but although
2224 the given file was scanned by exinext itself; it wasn't being passed on
2225 when Exim was called.
2227 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2228 an end-of-file indication when reading a command response.
2230 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2231 compiled. In many other places in Exim, IPv6 addresses are always
2232 recognized, so I have changed this. It also means that IPv4 domain
2233 literals of the form [IPV4:n.n.n.n] are now always recognized.
2235 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2236 used if the router is not running as root, for example, when verifying at
2237 ACL time, or when using -bh. The debugging output from this situation was
2238 non-existent - all you got was a failure to exec. I have made two
2241 (a) Failures to set uid/gid, the current directory, or a process leader
2242 in a subprocess such as that created by queryprogram now generate
2243 suitable debugging ouput when -d is set.
2245 (b) The queryprogram router detects when it is not running as root,
2246 outputs suitable debugging information if -d is set, and then runs
2247 the subprocess without attempting to change uid/gid.
2249 PH/15 Minor change to Makefile for building test_host (undocumented testing
2252 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2253 additional section of a DNS packet that returns MX or SRV records.
2254 Instead, it always explicitly searches for A/AAAA records. This avoids
2255 major problems that occur when a DNS server includes only records of one
2256 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2257 fixed another bug: if SRV records were looked up and the corresponding
2258 address records were *not* found in the additional section, the port
2259 values from the SRV records were lost.
2261 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2262 using the correct key (the original address) when searching the retry
2263 rules in order to find which one to use for generating the retry hint.
2265 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2266 adding the default one. Similarly, if it contains a Reply-To: header, the
2267 errors_reply_to option, if set, is not used.
2269 PH/19 When calculating a retry time, Exim used to measure the "time since
2270 failure" by looking at the "first failed" field in the retry record. Now
2271 it does not use this if it is later than than the arrival time of the
2272 message. Instead it uses the arrival time. This makes for better
2273 behaviour in cases where some deliveries succeed, thus re-setting the
2274 "first failed" field. An example is a quota failure for a huge message
2275 when small messages continue to be delivered. Without this change, the
2276 "time since failure" will always be short, possible causing more frequent
2277 delivery attempts for the huge message than are intended.
2278 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2280 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2281 $1, $2, $3) because the numerical variables can be reset during some
2282 expansion items (e.g. "match"), thereby losing the authentication data.
2284 PH/21 Make -bV show the size of off_t variables so that the test suite can
2285 decide whether to run tests for quotas > 2G.
2287 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2288 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2289 filecount value is greater than 2G or if a quota value is greater than 2G
2290 on a system where the size of off_t is not greater than 4, a panic error
2293 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2294 never match. The debug and -bh output now contains an explicit error
2295 message indicating a malformed IPv4 address or mask.
2297 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2298 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2299 PH/23 above applies.
2301 PH/25 Do not write to syslog when running in the test harness. The only
2302 occasion when this arises is a failure to open the main or panic logs
2303 (for which there is an explicit test).
2305 PH/26 Added the /no_tell option to "control=freeze".
2307 PH/27 If a host name lookup failed very early in a connection, for example, if
2308 the IP address matched host_lookup and the reverse lookup yielded a name
2309 that did not have a forward lookup, an error message of the form "no IP
2310 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2311 could be logged. Now it outputs the IP address instead of "NULL".
2313 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2314 allows the sender and the authenticated sender to be set when
2315 submitting a message from within Exim. Since child_open_exim() is
2316 documented for local_scan(), the new function should be too.
2318 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2319 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2320 results in an empty string is now treated as unset.
2322 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2324 PH/31 Added sender_verify_fail logging option.
2326 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2327 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2328 tidied the source and removed it altogether.
2330 PH/33 When a queue run was abandoned because the load average was too high, a
2331 log line was always written; now it is written only if the queue_run log
2332 selector is set. In addition, the log line for abandonment now contains
2333 information about the queue run such as the pid. This is always present
2334 in "start" and "stop" lines but was omitted from the "abandon" line.
2336 PH/34 Omit spaces between a header name and the colon in the error message that
2337 is given when verify = headers_syntax fails (if there are lots of them,
2338 the message gets confusing).
2340 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2341 names, as there are now some PTR records that contain slashes. This check
2342 is only to protect against broken name servers that fall over on strange
2343 characters, so the fact that it applies to all lookups doesn't matter.
2345 PH/36 Now that the new test suite is complete, we can remove some of the
2346 special code in Exim that was needed for the old test suite. For example,
2347 sorting DNS records because real resolvers return them in an arbitrary
2348 order. The new test suite's fake resolver always returns records in the
2351 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2352 bounces) except when queue_only is set, to avoid logging races between
2353 the different processes.
2355 PH/38 Panic-die if .include specifies a non-absolute path.
2357 PH/39 A tweak to the "H" retry rule from its user.
2359 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2360 a label. They prevented compilation on older perls.
2362 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2363 a warning to be raised on newish perls.
2365 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2366 on queue. Changes to match documented behaviour of showing count of
2367 messages matching specified criteria.
2369 PH/40 Changed the default ident timeout from 30s to 5s.
2371 PH/41 Added support for the use of login_cap features, on those BSD systems
2372 that have them, for controlling the resources used by pipe deliveries.
2374 PH/42 The content-scanning code uses fopen() to create files in which to put
2375 message data. Previously it was not paying any attention to the mode of
2376 the files. Exim runs with umask(0) because the rest of the code creates
2377 files with open(), and sets the required mode explicitly. Thus, these
2378 files were ending up world-writeable. This was not a big issue, because,
2379 being within the spool directory, they were not world-accessible. I have
2380 created a function called modefopen, which takes an additional mode
2381 argument. It sets umask(777), creates the file, chmods it to the required
2382 mode, then resets the umask. All the relevant calls to fopen() in the
2383 content scanning code have been changed to use this function.
2385 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2386 to 24 hours. This avoids potential overflow problems when processing G
2387 and H retry rules. I suspect nobody ever tinkers with this value.
2389 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2391 PH/45 When the plaintext authenticator is running as a client, the server's
2392 challenges are checked to ensure they are valid base64 strings. By
2393 default, the authentication attempt is cancelled if an invalid string is
2394 received. Setting client_ignore_invalid_base64 true ignores these errors.
2395 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2396 they are received. Thus, the responses can be made to depend on the
2397 challenges. If an invalid string is ignored, an empty string is placed in
2400 PH/46 Messages that are created by the autoreply transport now contains a
2401 References: header, in accordance with RFCs 2822 and 3834.
2403 PH/47 Added authenticated_sender_force to the smtp transport.
2405 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2407 PH/49 Installed latest patch from the Sieve maintainer.
2409 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2410 was also set, the appendfile transport was unnecessarily scanning a
2411 directory of message files (e.g. for maildir delivery) to find the count
2412 of files (along with the size), even though it did not need this
2413 information. It now does the scan only if it needs to find either the
2414 size of the count of files.
2416 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2418 PH/52 Two bugs concerned with error handling when the smtp transport is
2421 (i) Exim was not creating retry information for temporary errors given
2422 for individual recipients after the DATA command when the smtp transport
2423 was used in LMTP mode. This meant that they could be retried too
2424 frequently, and not timed out correctly.
2426 (ii) Exim was setting the flag that allows error details to be returned
2427 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2428 recipients that were returned after the DATA command.
2430 PH/53 This is related to PH/52, but is more general: for any failing address,
2431 when detailed error information was permitted to be returned to the
2432 sender, but the error was temporary, then after the final timeout, only
2433 "retry timeout exceeded" was returned. Now it returns the full error as
2434 well as "retry timeout exceeded".
2436 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2437 do this, and (what is worse) MTAs that accept it.
2439 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2440 will now be deprecated.
2442 PH/56 New os.c-cygwin from the Cygwin maintainer.
2444 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2445 formats (previously only available in exim formats via -bpr, -bpru,
2446 and -bpra. Now also available in native and exiqgrep formats)
2448 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2449 with very large, slow to parse queues
2451 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2453 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2455 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2456 responses to authentication challenges, though it was showing the
2457 challenges; (ii) I've removed the CR characters from the debug output for
2460 PH/58 Allow for the insertion of a newline as well as a space when a string
2461 is turned into more than one encoded-word during RFC 2047 encoding. The
2462 Sieve code now uses this.
2464 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2465 data_4xx, lost_connection, tls_required.
2467 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2468 message was being sent as an SMTP response.
2470 PH/61 Add -l and -k options to exicyclog.
2472 PH/62 When verifying, if an address was redirected to one new address, so that
2473 verification continued, and the new address failed or deferred after
2474 having set something in $address_data, the value of $address_data was not
2475 passed back to the ACL. This was different to the case when no
2476 redirection occurred. The value is now passed back in both cases.
2478 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2479 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2480 use login_cap.h, so on its own it isn't the distinguishing feature. The
2481 new name refers directly to the setclassresources() function.
2483 PH/65 Added configuration files for NetBSD3.
2485 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2487 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2488 is preferred over IPv4.
2490 PH/68 The bounce_return_message and bounce_return_body options were not being
2491 honoured for bounces generated during the reception of non-SMTP messages.
2492 In particular, this applied to messages rejected by the ACL. This bug has
2493 been fixed. However, if bounce_return_message is true and bounce_return_
2494 body is false, the headers that are returned for a non-SMTP message
2495 include only those that have been read before the error was detected.
2496 (In the case of an ACL rejection, they have all been read.)
2498 PH/69 The HTML version of the specification is now built in a directory called
2499 spec_html instead of spec.html, because the latter looks like a path with
2500 a MIME-type, and this confuses some software.
2502 PH/70 Catch two compiler warnings in sieve.c.
2504 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2505 function verify_get_ident() calls ip_connect() to connect a socket, but
2506 if the "connect()" function timed out, ip_connect() used to close the
2507 socket. However, verify_get_ident() also closes the socket later, and in
2508 between Exim writes to the log, which may get opened at this point. When
2509 the socket was closed in ip_connect(), the log could get the same file
2510 descriptor number as the socket. This naturally causes chaos. The fix is
2511 not to close the socket in ip_connect(); the socket should be closed by
2512 the function that creates it. There was only one place in the code where
2513 this was missing, in the iplookup router, which I don't think anybody now
2514 uses, but I've fixed it anyway.
2516 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2517 well as to direct DNS lookups. Otherwise the handling of names in host
2518 lists is inconsistent and therefore confusing.
2524 PH/01 Two changes to the default runtime configuration:
2526 (1) Move the checks for relay_from_hosts and authenticated clients from
2527 after to before the (commented out) DNS black list checks.
2529 (2) Add control=submission to the relay_from_hosts and authenticated
2530 clients checks, on the grounds that messages accepted by these
2531 statements are most likely to be submissions.
2533 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2535 (1) Generate an error if the third argument for the ${prvs expansion is
2538 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2541 (3) Reset the variables that are obtained from the first argument of
2542 ${prvscheck and used in the second argument before leaving the code,
2543 because their memory is reclaimed, so using them afterwards may do
2546 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2547 one (it's much easier than Tom thought :-).
2549 (5) Because of (4), we can now allow for the use of $prvscheck_result
2550 inside the third argument.
2552 PH/03 For some reason, the default setting of PATH when running a command from
2553 a pipe transport was just "/usr/bin". I have changed it to
2556 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2557 anything to be listed in the output from -bV.
2559 PH/05 When a filter generated an autoreply, the entire To: header line was
2560 quoted in the delivery log line, like this:
2562 => >A.N.Other <ano@some.domain> <original@ddress> ...
2564 This has been changed so that it extracts the operative address. There
2565 may be more than one such address. If so, they are comma-separated, like
2568 => >ano@some.domain,ona@other.domain <original@ddress> ...
2570 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2571 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2572 not being looked up in the rDNS to get a host name, Exim was showing the
2573 IP address twice in Received: lines, even though the IP addresses were
2574 identical. For example:
2576 Received: from [1.2.3.4] (helo=[1.2.3.4])
2578 However, if the real host name was known, it was omitting the HELO data
2579 if it matched the actual IP address. This has been tidied up so that it
2580 doesn't show the same IP address twice.
2582 PH/07 When both +timestamp and +memory debugging was on, the value given by
2583 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2584 called by the debug printing, thereby overwriting the timestamp buffer.
2585 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2587 PH/08 When the original message was included in an autoreply transport, it
2588 always said "this is a copy of the message, including all the headers",
2589 even if body_only or headers_only was set. It now gives an appropriate
2592 PH/09 Applied a patch from the Sieve maintainer which:
2594 o fixes some comments
2595 o adds the (disabled) notify extension core
2596 o adds some debug output for the result of if/elsif tests
2597 o points to the current vacation draft in the documentation
2598 and documents the missing references header update
2602 o fixes a bug in processing the envelope test (when testing
2603 multiple envelope elements, the last element determinted the
2606 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2607 Electronic Mail") by including:
2609 Auto-submitted: auto-generated
2611 in the messages that it generates (bounce messages and others, such as
2612 warnings). In the case of bounce messages for non-SMTP mesages, there was
2613 also a typo: it was using "Auto_submitted" (underscore instead of
2614 hyphen). Since every message generated by Exim is necessarily in response
2615 to another message, thes have all been changed to:
2617 Auto-Submitted: auto-replied
2619 in accordance with these statements in the RFC:
2621 The auto-replied keyword:
2623 - SHOULD be used on messages sent in direct response to another
2624 message by an automatic process,
2626 - MUST NOT be used on manually-generated messages,
2628 - MAY be used on Delivery Status Notifications (DSNs) and Message
2629 Disposition Notifications (MDNs),
2631 - MUST NOT be used on messages generated by automatic or periodic
2632 processes, except for messages which are automatic responses to
2635 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2636 to the default Received: header definition.
2638 PH/12 Added log selector acl_warn_skipped (default on).
2640 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2641 variables because (a) they are in the wrong storage pool and (b) even if
2642 they were copied, it wouldn't work properly because of the caching.
2644 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2645 checking when decoding. Apparently there are clients that generate
2646 overlong encoded strings. Why am I not surprised?
2648 PH/15 If the first argument of "${if match_address" was not empty, but did not
2649 contain an "@" character, Exim crashed. Now it writes a panic log message
2650 and treats the condition as false.
2652 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2654 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2655 extension "envelope-auth". The code is finished and in agreement with
2656 other implementations, but there is no documentation so far and in fact,
2657 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2658 not changing the active code.
2660 Print executed "if" and "elsif" statements when debugging is used. This
2661 helps a great deal to understand what a filter does.
2663 Document more things not specified clearly in RFC3028. I had all this
2664 sorted out, when out of a sudden new issues came to my mind. Oops."
2666 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2669 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2670 canonical form (as documented). However, after a host name lookup from
2671 the IP address, check_host() was doing a simple string comparison with
2672 addresses acquired from the DNS when checking that the found name did
2673 have the original IP as one of its addresses. Since any found IPv6
2674 addresses are likely to be in abbreviated form, the comparison could
2675 fail. Luckily, there already exists a function for doing the comparison
2676 by converting both addresses to binary, so now that is used instead of
2677 the text comparison.
2679 PH/20 There was another similar case to PH/19, when a complete host name was
2680 given in a host list; looking up its IP address could give an abbreviated
2681 form, whereas the current host's name might or might not be abbreviated.
2682 The same fix has been applied.
2688 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2689 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2692 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2693 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2695 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2697 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2698 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2699 Also, for "Auto-Submitted", treat anything other than "no" as
2700 non-personal, in accordance with RFC 3834. (Previously it treated
2701 anything starting "auto-" as non-personal.)
2703 TF/01 The control=submission/name=... option had a problem with syntax
2704 errors if the name included a slash character. The /name= option
2705 now slurps the rest of the string, so it can include any characters
2706 but it must come last in the list of options (after /sender_retain
2709 PH/05 Some modifications to the interface to the fake nameserver for the new
2717 TK/01 Added the "success_on_redirect" address verification option. See
2718 NewStuff for rationale and an example.
2720 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2722 PH/02 Patch to exigrep to allow it to work on syslog lines.
2724 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2725 fread() to skip over the body file's header line, because in Cygwin the
2726 header line is locked and is inaccessible.
2728 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2729 co-exist for some time) to make it clear that it is the Exim ID that is
2730 referenced, not the Message-ID: header line.
2732 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2733 string_format() function, because snprintf() does not exist on all
2736 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2737 ${stat: expansion item.
2739 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2740 protocol synchronization error", to keep the pedants happy.
2742 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2743 well as for IRIX systems, when gcc is being used. See the host.c source
2746 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2748 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2751 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2752 transport and to the smtp transport in LMTP mode.
2754 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2756 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2757 run a filter in a subprocess. This could lead to confusion in subsequent
2758 lookups in the parent process. There should also be a search_tidyup() at
2759 the end of the subprocess.
2761 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2762 only if the host matched helo_try_verify_hosts, which caused the
2763 verification to occur when the EHLO/HELO command was issued. The ACL just
2764 tested the remembered result. Now, if a previous verification attempt has
2765 not happened, "verify = helo" does it there and then.
2767 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2769 TK/03 Fix log output including CR from clamd.
2771 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2772 exist provoked a memory error which could cause a segfault.
2774 PH/15 Installed PCRE 6.2
2776 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2778 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2779 of the problem. Specifically, suggested +O2 rather than +O1 for the
2782 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2784 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2785 record, the port was not correctly logged when the outgoing_port log
2786 selector was set (it logged the transort's default port).
2788 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2789 fallback_hosts, and "hosts" in the smtp transport.
2791 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2792 host errors such as "Connection refused".
2794 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2795 authentication with radiusclient 0.4.9:
2797 - Error returned from rc_read_config was caught wrongly
2798 - Username/password not passed on to radius server due to wrong length.
2800 The presumption is that some radiusclient API changes for 4.51/PH/17
2801 were not taken care of correctly. The code is still untested by me (my
2802 Linux distribution still has 0.3.2 of radiusclient), but it was
2803 contributed by a Radius user.
2805 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2806 expanding the "port" option of the smtp transport.
2808 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2809 while reading a MIME header. Thanks to Tom Hughes for a patch.
2811 PH/24 Include config.h inside local_scan.h so that configuration settings are
2814 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2815 that in an ACL for RCPT (for example), you can examine exactly what was
2818 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2819 commands, but it was not correctly comparing the address with the actual
2820 client host address. Thus, it would show the EHLO address in Received:
2821 header lines when this was not necessary.
2823 PH/27 Added the % operator to ${eval:}.
2825 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2826 it should be ignoring failures (because with -C, for example, it has lost
2827 privilege). It wasn't ignoring creation failures other than "already
2830 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2833 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2834 because an input line was too long, either on its own, or by virtue of
2835 too many continuations, the temporary file was not being removed, and the
2836 return code was incorrect.
2838 PH/31 Missing "BOOL" in function definition in filtertest.c.
2840 PH/32 Applied Sieve patches from the maintainer.
2842 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2844 PH/33 Added "verify = not_blind".
2846 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2847 Local/Makefile (with some defaults set). These are used in built scripts
2848 such as exicyclog, but they have never been used in the exim_install
2849 script (though there are many overriding facilities there). I have
2850 arranged that the exim_install script now takes note of these two
2853 PH/35 Installed configuration files for Dragonfly.
2855 PH/36 When a locally submitted message by a trusted user did not contain a
2856 From: header, and the sender address was obtained from -f or from an SMTP
2857 MAIL command, and the trusted user did not use -F to supply a sender
2858 name, $originator_name was incorrectly used when constructing a From:
2859 header. Furthermore, $originator_name was used for submission mode
2860 messages from external hosts without From: headers in a similar way,
2861 which is clearly wrong.
2863 PH/37 Added control=suppress_local_fixups.
2865 PH/38 When log_selector = +received_sender was set, and the addition of the
2866 sender made the log line's construction buffer exactly full, or one byte
2867 less than full, an overflow happened when the terminating "\n" was
2870 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2871 when the result of a list match is failure because a DNS lookup failed.
2873 PH/40 RM_COMMAND is now used in the building process.
2875 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2876 the "build-* directories that it finds.
2878 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2879 address in a domain literal was a prefix of an interface address.
2881 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2882 when verifying a sender address, unless rewrite_headers is false.
2884 PH/44 Wrote a long comment about why errors_to addresses are verified as
2885 recipients, not senders.
2887 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2888 the ratelimit ACL was added.
2890 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2892 PH/46 Added extra information about PostgreSQL errors to the error string.
2894 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2895 avoiding the need to install special zones in a real server. This is
2896 backwards compatible; if it can't find the fake resolver, it drops back.
2897 Thus, both old and new test suites can be run.
2899 TF/02 Added util/ratelimit.pl
2901 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2902 clock is set back in time.
2904 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2905 Candler <B.Candler@pobox.com>.
2907 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2908 OK for addresses that are the result of redirections.
2910 PH/48 A number of further additions for the benefit of the new test suite,
2911 including a fake gethostbyname() that interfaces to the fake DNS resolver
2914 TF/06 The fix for widen_domains has also been applied to qualify_single and
2915 search_parents which are the other dnslookup options that can cause
2918 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2921 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2923 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2924 service" if it encountered an MX record with an empty target hostname.
2925 The message is now "an MX or SRV record indicated no SMTP service".
2927 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2928 if the DNS of the sending site is misconfigured. This is quite a
2929 common situation. This change restores the behaviour of treating a
2930 helo verification defer as a failure.
2932 PH/51 If self=fail was set on a router, the bounce message did not include the
2933 actual error message.
2939 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2941 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2942 command itself ended in error, the underlying message about the transport
2943 filter timeout was being overwritten with the pipe command error. Now the
2944 underlying error message should be appended to the second error message.
2946 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2948 PH/02 Reduce the amount of output that "make" produces by default. Full output
2949 can still be requested.
2951 PH/03 The warning log line about a condition test deferring for a "warn" verb
2952 was being output only once per connection, rather than after each
2953 occurrence (because it was using the same function as for successful
2954 "warn" verbs). This seems wrong, so I have changed it.
2956 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2957 it should not have, which might have caused a crash in the right
2958 circumstances, but probably never did.
2960 PH/04 Installed a modified version of Tony Finch's patch to make submission
2961 mode fix the return path as well as the Sender: header line, and to
2962 add a /name= option so that you can make the user's friendly name appear
2965 TF/03 Added the control = fakedefer ACL modifier.
2967 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2968 Mark Lowes for thorough testing.
2970 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2972 TK/03 Merged latest SRS patch from Miles Wilton.
2974 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2975 with the definition in sysexits.h (which is #included earlier).
2976 Fortunately, Exim does not actually use EX_OK. The code used to try to
2977 preserve the sysexits.h value, by assumimg that macro definitions were
2978 scanned for macro replacements. I have been disabused of this notion,
2979 so now the code just undefines EX_OK before #including unistd.h.
2981 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2982 in the smtp transport. When a block could not be written in a single
2983 write() function, the timeout was being re-applied to each part-write.
2984 This seems wrong - if the receiver was accepting one byte at a time it
2985 would take for ever. The timeout is now adjusted when this happens. It
2986 doesn't have to be particularly precise.
2988 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2989 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2991 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2992 without a domain if the check to <postmaster@domain> fails.
2994 SC/01 Eximstats: added -xls and the ability to specify output files
2995 (patch written by Frank Heydlauf).
2997 SC/02 Eximstats: use FileHandles for outputing results.
2999 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3001 SC/04 Eximstats: fixed display of large numbers with -nvr option
3003 SC/05 Eximstats: fixed merging of reports with empty tables.
3005 SC/06 Eximstats: added the -include_original_destination flag
3007 SC/07 Eximstats: removed tabs and trailing whitespace.
3009 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3011 TK/06 MBOX spool code: Add real "From " MBOX separator line
3012 so the .eml file is really in mbox format (even though
3013 most programs do not really care). Patch from Alex Miller.
3015 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3016 The latter is generated from $received_to and is only set if the
3017 message has one envelope recipient. SA can use these headers,
3018 obviously out-of-the-box. Patch from Alex Miller.
3020 PH/08 The ${def test on a variable was returning false if the variable's
3021 value was "0", contrary to what the specification has always said!
3022 The result should be true unless the variable is empty.
3024 PH/09 The syntax error of a character other than { following "${if
3025 def:variable_name" (after optional whitespace) was not being diagnosed.
3026 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3027 accidental colon was present, for example, could give incorrect results.
3029 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3030 result is used (not including appendfile, where other changes are about
3033 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3034 This involved changing a lot of size variables from int to off_t. It
3035 should work with maildirs and everything.
3037 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3038 spamd dying while we are connected to it.
3040 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3043 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3044 function for BDB 4.3.
3046 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3048 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3049 the sources for PCRE have been re-arranged and more files are now
3052 PH/15 The code I had for printing potentially long long variables in PH/11
3053 above was not the best (it lost precision). The length of off_t variables
3054 is now inspected at build time, and an appropriate printing format (%ld
3055 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3056 to be "long long int" or "long int". This is needed for the internal
3057 formatting function string_vformat().
3059 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3060 the configuration file to be ":syslog", then the script "guesses" where
3061 the logs files are, rather than using the compiled in default. In our
3062 case the guess is not the same as the compiled default, so the script
3063 suddenly stopped working when I started to use syslog. The patch checks
3064 to see if log_file_path is "". If so, it attempts to read it from exim
3065 with no configuration file to get the compiled in version, before it
3066 falls back to the previous guessing code."
3068 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3069 implementing BATV in an Exim configuration. See NewStuff for the gory
3072 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3073 Makefile that are specific to HP-UX.
3075 PH/18 If the "use_postmaster" option was set for a recipient callout together
3076 with the "random" option, the postmaster address was used as the MAIL
3077 FROM address for the random test, but not for the subsequent recipient
3078 test. It is now used for both.
3080 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3081 patch removes a few documentation additions to RFC 3028, because the
3082 latest draft now contains them. It adds the new en;ascii-case comparator
3083 and a new error check for 8bit text in MIME parts. Comparator and
3084 require names are now matched exactly. I enabled the subaddress
3085 extension, but it is not well tested yet (read: it works for me)."
3087 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3088 rework some of the code of TK/09 above to avoid the hardwired use of
3089 "%lld" and "long long". Replaced the call to snprintf() with a call to
3092 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3093 records point to non-existent hosts", "retry timeout exceeded", and
3094 "retry time not reached for any host after a long failure period".
3096 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3097 experimental DomainKeys support:
3099 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3100 (2) On an error such as an illegally used "control", the wrong name for
3101 the control was given.
3103 These problems did NOT occur unless DomainKeys support was compiled.
3105 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3107 PH/24 Added ${if match_ip condition.
3109 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3110 databases so that it will be absolutely obvious if a crash occurs in the
3111 DB library. This is a regular occurrence (often caused by mis-matched
3114 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3115 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3116 picked up on a user's system that detects such things. There doesn't seem
3117 to be a gcc warning option for this - only an attribute that has to be
3118 put on the function's prototype. It seems that in Fedora Core 4 they have
3119 set this on a number of new functions. No doubt there will be more in due
3122 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3123 specify a transport. However, if an address that was verified by such a
3124 router was the subject of a callout, Exim crashed because it tried to
3125 read the rcpt_include_affixes from the non-existent transport. Now it
3126 just assumes that the setting of that option is false. This bug was
3127 introduced by 4.51/PH/31.
3129 PH/28 Changed -d+all to exclude +memory, because that information is very
3130 rarely of interest, but it makes the output a lot bigger. People tend to
3131 do -d+all out of habit.
3133 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3134 code in os-type was giving problems when libc.so lives in lib64, like on
3137 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3138 aren't the modern standard, and it seems that some systems' include files
3139 don't always have them. Exim was already checking for some of the newer
3140 ones like T_AAAA, and defining it itself. I've added checks for all the
3141 record types that Exim uses.
3143 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3144 not automatically generating a new one, as it is supposed to. This
3145 prevented TLS from working. If the file did exist, but contained invalid
3146 data, a new version was generated, as expected. It was only the case of a
3147 non-existent file that was broken.
3149 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3150 with a change in libdomainkeys > 0.64.
3152 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3153 from DNS. If the selector record carries the flag, it now has
3154 precedence over the domain-wide flag.
3156 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3158 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3159 the use of a port name that isn't defined in /etc/services) occurred, the
3160 message was deferred as in a normal delivery, and thus remained on the
3161 spool, instead of being failed because of the mua_wrapper setting. This
3162 is now fixed, and I tidied up some of the mua_wrapper messages at the
3165 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3166 the messages in a hash of arrays rather than using individual hashes.
3167 This is a bit cleaner and results in dramatic memory savings, albeit
3168 at a slight CPU cost.
3170 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3171 as requested by Marc Sherman.
3173 SC/10 Eximstats: added histograms for user specified patterns as requested
3176 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3178 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3179 fopen() in the content-scanning modules that did not already have it.
3185 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3186 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3188 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3190 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3192 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3193 to test Sieve filters that use "vacation".
3195 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3196 that changes the way the GnuTLS parameters are stored in the cache file.
3197 The new format can be generated externally. For backward compatibility,
3198 if the data in the cache doesn't make sense, Exim assumes it has read an
3199 old-format file, and it generates new data and writes a new file. This
3200 means that you can't go back to an older release without removing the
3203 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3204 work if there are any delivery delays because "one_time" forces the
3205 parent to be marked "delivered", so its unseen clone is never tried
3206 again. For this reason, Exim now forbids the simultaneous setting of
3209 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3210 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3211 entry if you want to know the details. The fix, however, neglected to
3212 consider the case when local delivery batching is involved. The test for
3213 "previously delivered" was not happening when checking to see if an
3214 address could be batched with a previous (undelivered) one; under
3215 certain circumstances this could lead to multiple deliveries to the same
3218 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3219 in its include files, and this causes problems building Exim.
3221 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3222 header_syntax) but Exim was just ignoring anything given after a slash.
3223 In particular, this caused confusion with an attempt to use "verify =
3224 reverse_host_lookup/defer_ok". An error is now given when options are
3225 supplied for verify items that do not have them. (Maybe reverse_host_
3226 lookup should have a defer_ok option, but that's a different point.)
3228 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3229 defined by RFC 821) to 2048, because there were problems with some AUTH
3230 commands, and RFC 1869 says the size should be increased for extended
3231 SMTP commands that take arguments.
3233 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3236 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3237 "unknown" error; now it says that the functionality isn't in the binary.
3239 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3240 an address' error message when a string expansion fails (syntax or
3241 whatever). Otherwise the password may appear in the log. Following change
3242 PH/42 below, there is no longer a chance of it appearing in a bounce
3245 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3247 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3248 of its addresses was ever tried. (Bugzilla bug #2.)
3250 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3251 the result incorrectly in the debug output. (It correctly added a newline
3252 to what was transported.)
3254 TF/01 Added $received_time.
3256 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3257 commented out examples of how to interface to a virus scanner and to
3258 SpamAssassin. Also added commented examples of av_scanner and
3259 spamd_address settings.
3261 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3262 and controls are allowed in which ACLs. There were a couple of minor
3263 errors. Some of the entries in the conditions table (which is a table of
3264 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3265 negation of where the condition IS allowed.
3267 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3269 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3270 header file does not have a version number, so I've had to invent a new
3271 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3272 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3273 radiusclient), but it was contributed by a Radius user.
3275 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3276 files or directories, for OpenSSL.
3278 PH/19 When an Exim process that is running as root has to create an Exim log
3279 file, it does so in a subprocess that runs as exim:exim so as to get the
3280 ownership right at creation (otherwise, other Exim processes might see
3281 the file with the wrong ownership). There was no test for failure of this
3282 fork() call, which would lead to the process getting stuck as it waited
3283 for a non-existent subprocess. Forks do occasionally fail when resources
3284 run out. I reviewed all the other calls to fork(); they all seem to check
3287 PH/20 When checking for unexpected SMTP input at connect time (before writing
3288 the banner), Exim was not dealing correctly with a non-positive return
3289 from the read() function. If the client had disconnected by this time,
3290 the result was a log entry for a synchronization error with an empty
3291 string after "input=" when read() returned zero. If read() returned -1
3292 (an event I could not check), uninitialized data bytes were printed.
3293 There were reports of junk text (parts of files, etc) appearing after
3296 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3298 PH/22 Added support for macro redefinition, and (re)definition in between
3299 driver and ACL definitions.
3301 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3302 forgetting to use the resulting value; it was using the unexpanded value.
3304 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3305 hadn't been configured. The fix is from Juergen Kreileder, who
3306 understands it better than I do:
3308 "Here's what I see happening with three configured cyrus_sasl
3309 authenticators configured (plain, login, cram-md5):
3311 On startup auth_cyrus_sasl_init() gets called for each of these.
3312 This means three calls to sasl_listmech() without a specified mech_list.
3313 => SASL tests which mechs of all available mechs actually work
3314 => three warnings about OTP not working
3315 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3317 With the patch, sasl_listmech() also gets called three times. But now
3318 SASL's mech_list option is set to the server_mech specified in the the
3319 authenticator. Or in other words, the answer from sasl_listmech()
3320 gets limited to just the mech you're testing for (which is different
3322 => the return list contains just 'plain' or 'login', 'cram-md5' or
3323 nothing depending on the value of ob->server_mech.
3325 I've just tested the patch: Authentication still works fine,
3326 unavailable mechs specified in the exim configuration are still
3327 caught, and the auth.log warnings about OTP are gone."
3329 PH/25 When debugging is enabled, the contents of the command line are added
3330 to the debugging output, even when log_selector=+arguments is not
3333 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3334 answer is "GNU", and only if the return is "GNU/something" is the answer
3337 PH/27 $acl_verify_message is now set immediately after the failure of a
3338 verification in an ACL, and so is available in subsequent modifiers. In
3339 particular, the message can be preserved by coding like this:
3341 warn !verify = sender
3342 set acl_m0 = $acl_verify_message
3344 Previously, $acl_verify_message was set only while expanding "message"
3345 and "log_message" when a very denied access.
3347 PH/28 Modified OS/os.c-Linux with
3349 -#ifndef OS_LOAD_AVERAGE
3350 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3352 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3353 nomenclature these days.)
3355 PH/29 Installed patch from the Sieve maintainer that adds the options
3356 sieve_useraddress and sieve_subaddress to the redirect router.
3358 PH/30 In these circumstances:
3359 . Two addresses routed to the same list of hosts;
3360 . First host does not offer TLS;
3361 . First host accepts first address;
3362 . First host gives temporary error to second address;
3363 . Second host offers TLS and a TLS session is established;
3364 . Second host accepts second address.
3365 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3366 and peerdn, if requested) that were in fact used only for the second
3369 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3370 attention to any local part prefix or suffix that was matched by the
3371 router that accepted the address. It now behaves in the same way as it
3372 does for delivery: the affixes are removed from the local part unless
3373 rcpt_include_affixes is set on the transport.
3375 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3376 timeout during the DATA phase of an incoming message.
3378 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3379 applied a patch sanctioned by the Sieve maintainer.
3381 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3382 the uid or gid is negative. A case of a negative gid caused this to be
3383 noticed. The fix allows for either to be negative.
3385 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3386 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3389 PH/36 The change PH/12 above was broken. Fixed it.
3391 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3392 the grounds that routing the same address twice would always produce the
3393 same answer. This might have been true once, but it is certainly no
3394 longer true now. Routing a child address may depend on the previous
3395 routing that produced that child. Some complicated redirection strategies
3396 went wrong when messages had multiple recipients, and made Exim's
3397 behaviour dependent on the order in which the addresses were given.
3399 I have moved the duplicate checking until after the routing is complete.
3400 Exim scans the addresses that are assigned to local and remote
3401 transports, and removes any duplicates. This means that more work will be
3402 done, as duplicates will always all be routed, but duplicates are
3403 presumably rare, so I don't expect this is of any significance.
3405 For deliveries to pipes, files, and autoreplies, the duplicate checking
3406 still happens during the routing process, since they are not going to be
3409 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3410 It corrects a timeout issue with spamd. This is Ian's comment: "The
3411 background is that sometimes spamd either never reads data from a
3412 connection it has accepted, or it never writes response data. The exiscan
3413 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3414 blindly assumes that writes won't block so it may never time out."
3416 PH/39 Allow G after quota size as well as K and M.
3418 PH/40 The value set for $authenticated_id in an authenticator may not contain
3419 binary zeroes or newlines because the value is written to log lines and
3420 to spool files. There was no check on this. Now the value is run through
3421 the string_printing() function so that such characters are converted to
3422 printable escape sequences.
3424 PH/41 $message_linecount is a new variable that contains the total number of
3425 lines in the message. Compare $body_linecount, which is the count for the
3428 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3429 bounce and delay warning messages, except in certain special cases, which
3432 (a) An SMTP error message from a remote host;
3433 (b) A message specified in a :fail: redirection;
3434 (c) A message specified in a "fail" command in a system filter;
3435 (d) A message specified in a FAIL return from the queryprogram router;
3436 (e) A message specified by the cannot_route_message router option.
3438 In these cases only, Exim does include the error details in bounce and
3439 warning messages. There are also a few cases where bland messages such
3440 as "unrouteable address" or "local delivery error" are given.
3442 PH/43 $value is now also set for the "else" part of a ${run expansion.
3444 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3445 being worked on, but at least Exim now implements the latest version to
3448 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3449 process to complete was treated as a delivery failure, a timeout while
3450 writing the message to the pipe was logged, but erroneously treated as a
3451 successful delivery. Such timeouts include transport filter timeouts. For
3452 consistency with the overall process timeout, these timeouts are now
3453 treated as errors, giving rise to delivery failures by default. However,
3454 there is now a new Boolean option for the pipe transport called
3455 timeout_defer, which, if set TRUE, converts the failures into defers for
3456 both kinds of timeout. A transport filter timeout is now identified in
3459 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3460 systems where "make" and "gmake" are different, calling "gmake" at top
3461 level broke things. I've arranged for the value of $(MAKE) to be passed
3462 from the Makefile to this script so that it can call the same version of
3466 A note about Exim versions 4.44 and 4.50
3467 ----------------------------------------
3469 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3470 changes of various kinds. As a consequence, a big documentation update was
3471 needed. This delayed the release for rather longer than seemed good, especially
3472 in the light of a couple of (minor) security issues. Therefore, the changes
3473 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3474 release. So 4.44 and 4.50 are in effect two different branches that both start
3477 I have left the 4.50 change log unchanged; it contains all the changes since
3478 4.43. The change log for 4.44 is below; many of its items are identical to
3479 those for 4.50. This seems to be the most sensible way to preserve the
3480 historical information.
3486 1. Minor wording change to the doc/README.SIEVE file.
3488 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3489 computation of the current number of files was incorrect.
3491 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3492 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3495 4. Give more explanation in the error message when the command for a transport
3496 filter fails to execute.
3498 5. There are several places where Exim runs a non-Exim command in a
3499 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3500 was being done only for the command run by the queryprogram router. It is
3501 now done for all such subprocesses. The other cases are: ${run, transport
3502 filters, and the commands run by the lmtp and pipe transports.
3504 6. Added CONFIGURE_GROUP build-time option.
3506 7. Some older OS have a limit of 256 on the maximum number of file
3507 descriptors. Exim was using setrlimit() to set 1000 as a large value
3508 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3509 systems. I've change it so that if it can't get 1000, it tries for 256.
3511 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3512 was an oversight, and furthermore, ever since the addition of extra
3513 controls (e.g. 4.43/32), the checks on when to allow different forms of
3514 "control" were broken. There should now be diagnostics for all cases when a
3515 control that does not make sense is encountered.
3517 9. Added the /retain_sender option to "control=submission".
3519 10. $recipients is now available in the predata ACL (oversight).
3521 11. Tidy the search cache before the fork to do a delivery from a message
3522 received from the command line. Otherwise the child will trigger a lookup
3523 failure and thereby defer the delivery if it tries to use (for example) a
3524 cached ldap connection that the parent has called unbind on.
3526 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3527 of $address_data from the recipient verification was clobbered by the
3528 sender verification.
3530 13. The value of address_data from a sender verification is now available in
3531 $sender_address_data in subsequent conditions in the ACL statement.
3533 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3535 15. Added a new option "connect=<time>" to callout options, to set a different
3538 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3539 was its contents. (It was OK if the option was not defined at all.)
3541 17. A "Completed" log line is now written for messages that are removed from
3542 the spool by the -Mrm option.
3544 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3545 information about exactly what failed.
3547 19. Added -dd to debug only the daemon process.
3549 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3550 handles timeouts, both on the server side and network timeouts. Renamed the
3551 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3553 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3554 It is now set to "smtps".
3556 22. $host_address is now set to the target address during the checking of
3557 ignore_target_hosts.
3559 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3560 being passed; this would have caused $sender_host_name to have been used if
3561 matching the list had actually called for a host name (not very likely,
3562 since this list is usually IP addresses). A host name is now passed as
3565 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3566 code that specifies a non-restarting handler (typically sigaction() in
3567 modern systems) in an attempt to fix a rare and obscure crash bug.
3569 25. Narrowed the window for a race in the daemon that could cause it to ignore
3570 SIGCHLD signals. This is not a major problem, because they are used only to
3571 wake it up if nothing else does.
3573 26. A malformed maildirsize file could cause Exim to calculate negative values
3574 for the mailbox size or file count. Odd effects could occur as a result.
3575 The maildirsize information is now recalculated if the size or filecount
3578 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3579 support for a long time. Removed HAVE_SYS_VFS_H.
3581 28. Installed the latest version of exipick from John Jetmore.
3583 29. In an address list, if the pattern was not a regular expression, an empty
3584 subject address (from a bounce message) matched only if the pattern was an
3585 empty string. Non-empty patterns were not even tested. This was the wrong
3586 because it is perfectly reasonable to use an empty address as part of a
3587 database query. An empty address is now tested by patterns that are
3588 lookups. However, all the other forms of pattern expect the subject to
3589 contain a local part and a domain, and therefore, for them, an empty
3590 address still always fails if the pattern is not itself empty.
3592 30. Exim went into a mad DNS loop when attempting to do a callout where the
3593 host was specified on an smtp transport, and looking it up yielded more
3594 than one IP address.
3596 31. Re-factored the code for checking spool and log partition space into a
3597 function that finds that data and another that does the check. The former
3598 is then used to implement four new variables: $spool_space, $log_space,
3599 $spool_inodes, and $log_inodes.
3601 32. The RFC2047 encoding function was originally intended for short strings
3602 such as real names; it was not keeping to the 75-character limit for
3603 encoded words that the RFC imposes. It now respects the limit, and
3604 generates multiple encoded words if necessary. To be on the safe side, I
3605 have increased the buffer size for the ${rfc2047: expansion operator from
3608 33. It is now permitted to omit both strings after an "if" condition; if the
3609 condition is true, the result is "true". As before, when the second string
3610 is omitted, a false condition yields an empty string. This makes it less
3611 cumbersome to write custom ACL and router conditions.
3613 34. Failure to deliver a bounce message always caused it to be frozen, even if
3614 there was an errors_to setting on the router. The errors_to setting is now
3617 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3618 canonical form (fully expanded) before being placed in
3619 $sender_host_address.
3621 36. The table in the code that translates DNS record types into text (T_A to
3622 "A" for instance) was missing entries for NS and CNAME. It is just possible
3623 that this could have caused confusion if both these types were looked up
3624 for the same domain, because the text type is used as part of Exim's
3625 per-process caching. But the chance of anyone hitting this buglet seems
3628 37. The dnsdb lookup has been extended in a number of ways.
3630 (1) There is a new type, "zns", which walks up the domain tree until it
3631 finds some nameserver records. It should be used with care.
3633 (2) There is a new type, "mxh", which is like "mx" except that it returns
3634 just the host names, not the priorities.
3636 (3) It is now possible to give a list of domains (or IP addresses) to be
3637 looked up. The behaviour when one of the lookups defers can be
3638 controlled by a keyword.
3640 (4) It is now possible to specify the separator character for use when
3641 multiple records are returned.
3643 38. The dnslists ACL condition has been extended: it is now possible to supply
3644 a list of IP addresses and/or domains to be looked up in a particular DNS
3647 39. Added log_selector=+queue_time_overall.
3649 40. When running the queue in the test harness, wait just a tad after forking a
3650 delivery process, to get repeatability of debugging output.
3652 41. Include certificate and key file names in error message when GnuTLS fails
3653 to set them up, because the GnuTLS error message doesn't include the name
3654 of the failing file when there is a problem reading it.
3656 42. Allow both -bf and -bF in the same test run.
3658 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3660 44. The "Exiscan patch" is now merged into the mainline Exim source.
3662 45. Sometimes the final signoff response after QUIT could fail to get
3663 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3664 < 0 before doing a fflush(). This bug looks as though it goes back to the
3665 introduction of TLS in release 3.20, but "sometimes" must have been rare
3666 because the tests only now provoked it.
3668 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3669 (this can affect the format of dates).
3671 47. exim_tidydb, when checking for the continued existence of a message for
3672 which it has found a message-specific retry record, was not finding
3673 messages that were in split spool directories. Consequently, it was
3674 deleting retry records that should have stayed in existence.
3676 48. Steve fixed some bugs in eximstats.
3678 49. The SPA authentication driver was not abandoning authentication and moving
3679 on to the next authenticator when an expansion was forced to fail,
3680 contradicting the general specification for all authenticators. Instead it
3681 was generating a temporary error. It now behaves as specified.
3683 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3684 (the order specifies the preference for clients). The order is now AES256,
3685 AES128, 3DES, ARCFOUR128.
3687 51. Small patch to Sieve code - explicitly set From: when generating an
3690 52. Exim crashed if a remote delivery caused a very long error message to be
3691 recorded - for instance if somebody sent an entire SpamAssassin report back
3692 as a large number of 550 error lines. This bug was coincidentally fixed by
3693 increasing the size of one of Exim's internal buffers (big_buffer) that
3694 happened as part of the Exiscan merge. However, to be on the safe side, I
3695 have made the code more robust (and fixed the comments that describe what
3698 53. Now that there can be additional text after "Completed" in log lines (if
3699 the queue_time_overall log selector is set), a one-byte patch to exigrep
3700 was needed to allow it to recognize "Completed" as not the last thing in
3703 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3704 patch that reportedly fixes this has been added. I am not expert enough to
3705 create a test for it. This is what the patch creator wrote:
3707 "I found a little strange behaviour of ldap code when working with
3708 Windows 2003 AD Domain, where users was placed in more than one
3709 Organization Units. When I tried to give exim partial DN, the exit code
3710 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3711 But simultaneously result of request was absolutely normal ldap result,
3712 so I produce this patch..."
3714 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3715 REFERENCE, so I have modified the code to exclude the patch when that macro
3718 55. Some experimental protocols are using DNS PTR records for new purposes. The
3719 keys for these records are domain names, not reversed IP addresses. The
3720 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3721 leaves it alone. Component reversal etc. now happens only for IP addresses.
3724 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3726 57. Double the size of the debug message buffer (to 2048) so that more of very
3727 long debug lines gets shown.
3729 58. The exicyclog utility now does better if the number of log files to keep
3730 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3732 59. Two changes related to the smtp_active_hostname option:
3734 (1) $smtp_active_hostname is now available as a variable.
3735 (2) The default for smtp_banner uses $smtp_active_hostname instead
3736 of $primary_hostname.
3738 60. The host_aton() function is supposed to be passed a string that is known
3739 to be a valid IP address. However, in the case of IPv6 addresses, it was
3740 not checking this. This is a hostage to fortune. Exim now panics and dies
3741 if the condition is not met. A case was found where this could be provoked
3742 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3743 components; fortuitously, this particular loophole had already been fixed
3744 by change 4.50/55 above.
3746 If there are any other similar loopholes, the new check in host_aton()
3747 itself should stop them being exploited. The report I received stated that
3748 data on the command line could provoke the exploit when Exim was running as
3749 exim, but did not say which command line option was involved. All I could
3750 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3751 running as the user.
3754 61. There was a buffer overflow vulnerability in the SPA authentication code
3755 (which came originally from the Samba project). I have added a test to the
3756 spa_base64_to_bits() function which I hope fixes it.
3759 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3760 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3762 63. The daemon start-up calls getloadavg() while still root for those OS that
3763 need the first call to be done as root, but it missed one case: when
3764 deliver_queue_load_max is set with deliver_drop_privilege. This is
3765 necessary for the benefit of the queue runner, because there is no re-exec
3766 when deliver_drop_privilege is set.
3768 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3769 This has been fixed.
3771 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3772 was in use, was not putting the data itself into the right store pool;
3773 consequently, it could be overwritten for a subsequent message in the same
3774 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3777 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3779 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3780 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3781 the function were treating the return as a boolean value, which happened to
3782 work because 0=false and not-0=true, but is not correct code.
3784 68. The host_aton() function was not handling scoped IPv6 addresses (those
3785 with, for example, "%eth0" on the end) correctly.
3787 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3788 negated items (that is, ~something) in unsigned ints. Some compilers
3789 apparently mutter when there is no cast.
3791 70. If an address verification called from an ACL failed, and did not produce a
3792 user-specific message (i.e. there was only a "system" message), nothing was
3793 put in $acl_verify_message. In this situation, it now puts the system
3796 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3797 session; change 4.31/43 added the unwanted input to the log line - except
3798 that it did not do this in the start of session case. It now does.
3800 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3801 This is wrong and can cause the other end to generate a synchronization
3802 error if it is another Exim or anything else that does the synchronization
3803 check. A QUIT command is no longer sent after a timeout.
3805 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3806 during host lookups.
3808 74. The defer_ok option of callout verification was not working if it was used
3809 when verifying addresses in header lines, that is, for this case:
3811 verify = header_sender/callout=defer_ok
3813 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3814 those file descriptors could be used for SMTP connections. If anything
3815 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3816 could be sent to the SMTP client, causing chaos. The daemon now opens
3817 stdin, stdout, and stderr to /dev/null when it puts itself into the
3820 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3821 log by default. The user can override this with suitable Perl magic.
3823 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3824 the log message when discard triggers, was not working for the DATA ACL or
3825 for the non-SMTP ACL.
3827 78. Error message wording change in sieve.c.
3829 79. If smtp_accept_max_per_host was set, the number of connections could be
3830 restricted to fewer than expected, because the daemon was trying to set up
3831 a new connection before checking whether the processes handling previous
3832 connections had finished. The check for completed processes is now done
3833 earlier. On busy systems, this bug wouldn't be noticed because something
3834 else would have woken the daemon, and it would have reaped the completed
3837 80. If a message was submitted locally by a user whose login name contained one
3838 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3839 It caused a spool format error. I have fixed the spool reading code. A
3840 related problem was that the "from" clause in the Received: line became
3841 illegal because of the space(s). It is now covered by ${quote_local_part.
3843 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3846 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3848 83. Patch from Sieve maintainer to fix unterminated string problem in
3849 "vacation" handling.
3851 84. Some minor changes to the Linux configuration files to help with other
3852 OS variants using glibc.
3854 85. One more patch for Sieve to update vacation handling to latest spec.
3857 ----------------------------------------------------
3858 See the note above about the 4.44 and 4.50 releases.
3859 ----------------------------------------------------
3865 1. Change 4.43/35 introduced a bug that caused file counts to be
3866 incorrectly computed when quota_filecount was set in an appendfile
3869 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3870 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3873 3. Give more explanation in the error message when the command for a transport
3874 filter fails to execute.
3876 4. There are several places where Exim runs a non-Exim command in a
3877 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3878 was being done only for the command run by the queryprogram router. It is
3879 now done for all such subprocesses. The other cases are: ${run, transport
3880 filters, and the commands run by the lmtp and pipe transports.
3882 5. Some older OS have a limit of 256 on the maximum number of file
3883 descriptors. Exim was using setrlimit() to set 1000 as a large value
3884 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3885 systems. I've change it so that if it can't get 1000, it tries for 256.
3887 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3888 was an oversight, and furthermore, ever since the addition of extra
3889 controls (e.g. 4.43/32), the checks on when to allow different forms of
3890 "control" were broken. There should now be diagnostics for all cases when a
3891 control that does not make sense is encountered.
3893 7. $recipients is now available in the predata ACL (oversight).
3895 8. Tidy the search cache before the fork to do a delivery from a message
3896 received from the command line. Otherwise the child will trigger a lookup
3897 failure and thereby defer the delivery if it tries to use (for example) a
3898 cached ldap connection that the parent has called unbind on.
3900 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3901 of $address_data from the recipient verification was clobbered by the
3902 sender verification.
3904 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3905 was its contents. (It was OK if the option was not defined at all.)
3907 11. A "Completed" log line is now written for messages that are removed from
3908 the spool by the -Mrm option.
3910 12. $host_address is now set to the target address during the checking of
3911 ignore_target_hosts.
3913 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3914 being passed; this would have caused $sender_host_name to have been used if
3915 matching the list had actually called for a host name (not very likely,
3916 since this list is usually IP addresses). A host name is now passed as
3919 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3920 code that specifies a non-restarting handler (typically sigaction() in
3921 modern systems) in an attempt to fix a rare and obscure crash bug.
3923 15. Narrowed the window for a race in the daemon that could cause it to ignore
3924 SIGCHLD signals. This is not a major problem, because they are used only to
3925 wake it up if nothing else does.
3927 16. A malformed maildirsize file could cause Exim to calculate negative values
3928 for the mailbox size or file count. Odd effects could occur as a result.
3929 The maildirsize information is now recalculated if the size or filecount
3932 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3933 support for a long time. Removed HAVE_SYS_VFS_H.
3935 18. Updated exipick to current release from John Jetmore.
3937 19. Allow an empty sender to be matched against a lookup in an address list.
3938 Previously the only cases considered were a regular expression, or an
3941 20. Exim went into a mad DNS lookup loop when doing a callout where the
3942 host was specified on the transport, if the DNS lookup yielded more than
3945 21. The RFC2047 encoding function was originally intended for short strings
3946 such as real names; it was not keeping to the 75-character limit for
3947 encoded words that the RFC imposes. It now respects the limit, and
3948 generates multiple encoded words if necessary. To be on the safe side, I
3949 have increased the buffer size for the ${rfc2047: expansion operator from
3952 22. Failure to deliver a bounce message always caused it to be frozen, even if
3953 there was an errors_to setting on the router. The errors_to setting is now
3956 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3957 canonical form (fully expanded) before being placed in
3958 $sender_host_address.
3960 24. Updated eximstats to version 1.33
3962 25. Include certificate and key file names in error message when GnuTLS fails
3963 to set them up, because the GnuTLS error message doesn't include the name
3964 of the failing file when there is a problem reading it.
3966 26. Expand error message when OpenSSL has problems setting up cert/key files.
3969 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3970 (this can affect the format of dates).
3972 28. exim_tidydb, when checking for the continued existence of a message for
3973 which it has found a message-specific retry record, was not finding
3974 messages that were in split spool directories. Consequently, it was
3975 deleting retry records that should have stayed in existence.
3977 29. eximstats updated to version 1.35
3978 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3979 1.35 - bugfix such that pie charts by volume are generated correctly
3981 30. The SPA authentication driver was not abandoning authentication and moving
3982 on to the next authenticator when an expansion was forced to fail,
3983 contradicting the general specification for all authenticators. Instead it
3984 was generating a temporary error. It now behaves as specified.
3986 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3987 (the order specifies the preference for clients). The order is now AES256,
3988 AES128, 3DES, ARCFOUR128.
3990 31. Small patch to Sieve code - explicitly set From: when generating an
3993 32. Exim crashed if a remote delivery caused a very long error message to be
3994 recorded - for instance if somebody sent an entire SpamAssassin report back
3995 as a large number of 550 error lines. This bug was coincidentally fixed by
3996 increasing the size of one of Exim's internal buffers (big_buffer) that
3997 happened as part of the Exiscan merge. However, to be on the safe side, I
3998 have made the code more robust (and fixed the comments that describe what
4001 33. Some experimental protocols are using DNS PTR records for new purposes. The
4002 keys for these records are domain names, not reversed IP addresses. The
4003 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4004 leaves it alone. Component reversal etc. now happens only for IP addresses.
4007 34. The host_aton() function is supposed to be passed a string that is known
4008 to be a valid IP address. However, in the case of IPv6 addresses, it was
4009 not checking this. This is a hostage to fortune. Exim now panics and dies
4010 if the condition is not met. A case was found where this could be provoked
4011 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4012 components; fortuitously, this particular loophole had already been fixed
4013 by change 4.50/55 or 4.44/33 above.
4015 If there are any other similar loopholes, the new check in host_aton()
4016 itself should stop them being exploited. The report I received stated that
4017 data on the command line could provoke the exploit when Exim was running as
4018 exim, but did not say which command line option was involved. All I could
4019 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4020 running as the user.
4023 35. There was a buffer overflow vulnerability in the SPA authentication code
4024 (which came originally from the Samba project). I have added a test to the
4025 spa_base64_to_bits() function which I hope fixes it.
4028 36. The daemon start-up calls getloadavg() while still root for those OS that
4029 need the first call to be done as root, but it missed one case: when
4030 deliver_queue_load_max is set with deliver_drop_privilege. This is
4031 necessary for the benefit of the queue runner, because there is no re-exec
4032 when deliver_drop_privilege is set.
4034 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4035 was in use, was not putting the data itself into the right store pool;
4036 consequently, it could be overwritten for a subsequent message in the same
4037 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4040 38. Sometimes the final signoff response after QUIT could fail to get
4041 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4042 < 0 before doing a fflush(). This bug looks as though it goes back to the
4043 introduction of TLS in release 3.20, but "sometimes" must have been rare
4044 because the tests only now provoked it.
4050 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4051 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4052 Now it returns an integer. A number of calls were still expecting a T/F
4053 return. Fortuitously, in all cases, the tests worked in OK situations,
4054 which is the norm. However, things would have gone wrong on any write
4055 failures on the smtp file descriptor. This function is used when sending
4056 messages over SMTP and also when doing verify callouts.
4058 2. When Exim is called to do synchronous delivery of a locally submitted
4059 message (the -odf or -odi options), it no longer closes stderr before doing
4062 3. Implemented the mua_wrapper option.
4064 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4066 5. Implemented the functions header_remove(), header_testname(),
4067 header_add_at_position(), and receive_remove_recipient(), and exported them
4070 6. If an ACL "warn" statement specified the addition of headers, Exim already
4071 inserted X-ACL-Warn: at the start if there was no header name. However, it
4072 was not making this test for the second and subsequent header lines if
4073 there were newlines in the string. This meant that an invalid header could
4074 be inserted if Exim was badly configured.
4076 7. Allow an ACL "warn" statement to add header lines at the start or after all
4077 the Received: headers, as well as at the end.
4079 8. Added the rcpt_4xx retry error code.
4081 9. Added postmaster_mailfrom=xxx to callout verification option.
4083 10. Added mailfrom=xxxx to the callout verification option, for verify=
4086 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4087 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4089 12. Inserted some casts to stop certain compilers warning when using pointer
4090 differences as field lengths or precisions in printf-type calls (mostly
4091 affecting debugging statements).
4093 13. Added optional readline() support for -be (dynamically loaded).
4095 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4096 same clock tick as a message's arrival, so that its received time was the
4097 same as the "first fail" time on the retry record, and that message
4098 remained on the queue past the ultimate address timeout, every queue runner
4099 would try a delivery (because it was past the ultimate address timeout) but
4100 after another failure, the ultimate address timeout, which should have then
4101 bounced the address, did not kick in. This was a "< instead of <=" error;
4102 in most cases the first failure would have been in the next clock tick
4103 after the received time, and all would be well.
4105 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4106 being recognized when the domain list was tested by the match_domain
4107 condition in an expansion string.
4109 16. Added the ${str2b64: operator.
4111 17. Exim was always calling setrlimit() to set a large limit for the number of
4112 processes, without checking whether the existing limit was already
4113 adequate. (It did check for the limit on file descriptors.) Furthermore,
4114 errors from getrlimit() and setrlimit() were being ignored. Now they are
4115 logged to the main and panic logs, but Exim does carry on, to try to do its
4116 job under whatever limits there are.
4118 18. Imported PCRE 5.0.
4120 19. Trivial typo in log message " temporarily refused connection" (the leading
4123 20. If the log selector return_path_on_delivery was set and an address was
4124 redirected to /dev/null, the delivery process crashed because it assumed
4125 that a return path would always be set for a "successful" delivery. In this
4126 case, the whole delivery is bypassed as an optimization, and therefore no
4129 21. Internal re-arrangement: the function for sending a challenge and reading
4130 a response while authentication was assuming a zero-terminated challenge
4131 string. It's now changed to take a pointer and a length, to allow for
4132 binary data in such strings.
4134 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4136 23. Exim was not respecting finduser_retries when seeking the login of the
4137 uid under which it was called; it was always trying 10 times. (The default
4138 setting of finduser_retries is zero.) Also, it was sleeping after the final
4139 failure, which is pointless.
4141 24. Implemented tls_on_connect_ports.
4143 25. Implemented acl_smtp_predata.
4145 26. If the domain in control=submission is set empty, Exim assumes that the
4146 authenticated id is a complete email address when it generates From: or
4147 Sender: header lines.
4149 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4150 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4151 chown and chgrp in /bin and hostname in /usr/bin.
4153 28. Exim was keeping the "process log" file open after each use, just as it
4154 does for the main log. This opens the possibility of it remaining open for
4155 long periods when the USR1 signal hits a daemon. Occasional processlog
4156 errors were reported, that could have been caused by this. Anyway, it seems
4157 much more sensible not to leave this file open at all, so that is what now
4160 29. The long-running daemon process does not normally write to the log once it
4161 has entered its main loop, and it closes the log before doing so. This is
4162 so that log files can straightforwardly be renamed and moved. However,
4163 there are a couple of unusual error situations where the daemon does write
4164 log entries, and I had neglected to close the log afterwards.
4166 30. The text of an SMTP error response that was received during a remote
4167 delivery was being truncated at 512 bytes. This is too short for some of
4168 the long messages that one sometimes sees. I've increased the limit to
4171 31. It is now possible to make retry rules that apply only when a message has a
4172 specific sender, in particular, an empty sender.
4174 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4175 it possible to be selective about when SMTP synchronization is enforced.
4177 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4179 32. Implemented hosts_connection_nolog.
4181 33. Added an ACL for QUIT.
4183 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4186 35. Added mailbox_size and mailbox_filecount to appendfile.
4188 36. Added control = no_multiline_responses to ACLs.
4190 37. There was a bug in the logic of the code that waits for the clock to tick
4191 in the case where the clock went backwards by a substantial amount such
4192 that the microsecond fraction of "now" was more than the microsecond
4193 fraction of "then" (but the whole seconds number was less).
4195 38. Added support for the libradius Radius client library this is found on
4196 FreeBSD (previously only the radiusclient library was supported).
4202 1. When certain lookups returned multiple values in the form name=value, the
4203 quoting of the values was not always being done properly. Specifically:
4204 (a) If the value started with a double quote, but contained no whitespace,
4206 (b) If the value contained whitespace other than a space character (i.e.
4207 tabs or newlines or carriage returns) it was not quoted.
4208 This fix has been applied to the mysql and pgsql lookups by writing a
4209 separate quoting function and calling it from the lookup code. The fix
4210 should probably also be applied to nisplus, ibase and oracle lookups, but
4211 since I cannot test any of those, I have not disturbed their existing code.
4213 2. A hit in the callout cache for a specific address caused a log line with no
4214 reason for rejecting RCPT. Now it says "Previous (cached) callout
4215 verification failure".
4217 3. There was an off-by-one bug in the queryprogram router. An over-long
4218 return line was truncated at 256 instead of 255 characters, thereby
4219 overflowing its buffer with the terminating zero. As well as fixing this, I
4220 have increased the buffer size to 1024 (and made a note to document this).
4222 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4223 when Exim is waiting for an SMTP response from a remote server, Exim
4224 restarts its select() call on the socket, thereby resetting its timeout.
4225 This is not a problem when such interrupts are rare. Somebody set up a cron
4226 job to run exiwhat every 2 minutes, which is less than the normal select()
4227 timeout (5 or 10 minutes). This meant that the select() timeout never
4228 kicked in because it was always reset. I have fixed this by comparing the
4229 time when an interrupt arrives with the time at the start of the first call
4230 to select(). If more time than the timeout has elapsed, the interrupt is
4231 treated as a timeout.
4233 5. Some internal re-factoring in preparation for the addition of Sieve
4234 extensions (by MH). In particular, the "personal" test is moved to a
4235 separate function, and given an option for scanning Cc: and Bcc: (which is
4236 not set for Exim filters).
4238 6. When Exim created an email address using the login of the caller as the
4239 local part (e.g. when creating a From: or Sender: header line), it was not
4240 quoting the local part when it contained special characters such as @.
4242 7. Installed new OpenBSD configuration files.
4244 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4245 try to make them clearer.
4247 9. Callout options, other than the timeout value, were being ignored when
4248 verifying sender addresses in header lines. For example, when using
4250 verify = header_sender/callout=no_cache
4252 the cache was (incorrectly) being used.
4254 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4255 only the Cygwin environment.
4257 11. When return_path_on_delivery was set as a log selector, if different remote
4258 addresses in the same message used different return paths and parallel
4259 remote delivery occurred, the wrong values would sometimes be logged.
4260 (Whenever a remote delivery process finished, the return path value from
4261 the most recently started remote delivery process was logged.)
4263 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4264 lines when AUTH and/or TLS are in use. This is the "received protocol"
4265 field. Exim used to use "asmtp" for authenticated SMTP, without any
4266 indication (in the protocol name) for TLS use. Now it follows the RFC and
4267 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4268 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4269 names appear in log lines as well as in Received: header lines.
4271 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4272 extensions, and comparison tests, and to fix some bugs.
4274 14. Changes to the "personal" filter test:
4276 (1) The test was buggy in that it was just doing the equivalent of
4277 "contains" tests on header lines. For example, if a user's address was
4278 anne@some.where, the "personal" test would incorrectly be true for
4280 To: susanne@some.where
4282 This test is now done by extracting each address from the header in turn,
4283 and checking the entire address. Other tests that are part of "personal"
4284 are now done using regular expressions (for example, to check local parts
4285 of addresses in From: header lines).
4287 (2) The list of non-personal local parts in From: addresses has been
4288 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4289 taken from the Sieve specification recommendations.
4291 (3) If the message contains any header line starting with "List-" it is
4292 treated as non-personal.
4294 (4) The test for "circular" in the Subject: header line has been removed
4295 because it now seems ill-conceived.
4297 15. Minor typos in src/EDITME comments corrected.
4299 16. Installed latest exipick from John Jetmore.
4301 17. If headers_add on a router specified a text string that was too long for
4302 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4303 of string_sprintf() is now avoided.
4305 18. $message_body_size was not set (it was always zero) when running the DATA
4306 ACL and the local_scan() function.
4308 19. For the "mail" command in an Exim filter, no default was being set for
4309 the once_repeat time, causing a random time value to be used if "once" was
4310 specified. (If the value happened to be <= 0, no repeat happened.) The
4311 default is now 0s, meaning "never repeat". The "vacation" command was OK
4312 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4313 (I found it when inspecting the code).
4315 20. There is now an overall timeout for performing a callout verification. It
4316 defaults to 4 times the callout timeout, which applies to individual SMTP
4317 commands during the callout. The overall timeout applies when there is more
4318 than one host that can be tried. The timeout is checked before trying the
4319 next host. This prevents very long delays if there are a large number of
4320 hosts and all are timing out (e.g. when the network connections are timing
4321 out). The value of the overall timeout can be changed by specifying an
4322 additional sub-option for "callout", called "maxwait". For example:
4324 verify = sender/callout=5s,maxwait=20s
4326 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4327 to the end before writing, but this should make it even safer).
4329 22. Exim was forgetting that it had advertised PIPELINING for the second and
4330 subsequent messages on an SMTP connection. It was also not resetting its
4331 memory on STARTTLS and an internal HELO.
4333 23. When Exim logs an SMTP synchronization error within a session, it now
4334 records whether PIPELINING has been advertised or not.
4336 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4337 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4338 rather than long int.
4340 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4342 26. Added the never_mail option to autoreply.
4348 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4349 crash if the getsockname() call failed; this can happen if a connection is
4350 closed very soon after it is established. The problem was simply in the
4351 order in which certain operations were done, causing Exim to try to write
4352 to the SMTP stream before it had set up the file descriptor. The bug has
4353 been fixed by making things happen in the correct order.
4359 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4360 before the connection was closed, thus losing the rejection response.
4362 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4363 some early Solaris releases, but causes trouble in current releases where
4364 socklen_t is defined.
4366 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4369 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4372 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4373 that releasing the top of it at the end releases what was used for sub-
4374 expansions (unless the block got too big). However, discard this block if
4375 the first thing is a variable or header, so that we can use its block when
4376 it is dynamic (useful for very large $message_headers, for example).
4378 6. Lookups now cache *every* query, not just the most recent. A new, separate
4379 store pool is used for this. It can be recovered when all lookup caches are
4380 flushed. Lookups now release memory at the end of their result strings.
4381 This has involved some general refactoring of the lookup sources.
4383 7. Some code has been added to the store_xxx() functions to reduce the amount
4384 of flapping under certain conditions.
4386 8. log_incoming_interface used to affect only the <= reception log lines. Now
4387 it causes the local interface and port to be added to several more SMTP log
4388 lines, for example "SMTP connection from", and rejection lines.
4390 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4392 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4394 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4395 could be overwritten at the end of the current message (or the start of a
4396 new message if it was set in a HELO ACL). The value is now preserved for
4397 the duration of the SMTP connection.
4399 12. If a transport had a headers_rewrite setting, and a matching header line
4400 contained an unqualified address, that address was qualified, even if it
4401 did not match any rewriting rules. The underlying bug was that the values
4402 of the flags that permit the existence of unqualified sender and recipient
4403 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4404 non-local messages, and by -bnq for local messages) were not being
4405 preserved with the message after it was received.
4407 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4408 "next input=" as part of the text comprising the host identity instead of
4409 the correct text. The code was using the same buffer for two different
4410 strings. However, depending on which order the printing function evaluated
4411 its arguments, the bug did not always show up. Under Linux, for example, my
4412 test suite worked just fine.
4414 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4415 doesn't work with some older versions of Perl. It has been changed to "my",
4416 which in any case is probably the better facility to use.
4418 15. A really picky compiler found some instances of statements for creating
4419 error messages that either had too many or two few arguments for the format
4422 16. The size of the buffer for calls to the DNS resolver has been increased
4423 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4424 for addresses that have a lot of PTR records. This alleviates a problem; it
4425 does not fully solve it.
4427 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4428 buffer now truncates the list and logs the incident, which is the same
4429 action as happens when Exim is looking up a host name and its aliases.
4430 Previously in this situation something unpredictable would happen;
4431 sometimes it was "internal error: store_reset failed".
4433 18. If a server dropped the connection unexpectedly when an Exim client was
4434 using GnuTLS and trying to read a response, the client delivery process
4435 crashed while trying to generate an error log message.
4437 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4438 string, for example:
4440 warn message = H1: something\nH2: something
4442 the text was added as a single header line from Exim's point of view
4443 though it ended up OK in the delivered message. However, searching for the
4444 second and subsequent header lines using $h_h2: did not work. This has been
4445 fixed. Similarly, if a system filter added multiple headers in this way,
4446 the routers could not see them.
4448 20. Expanded the error message when iplsearch is called with an invalid key to
4449 suggest using net-iplsearch in a host list.
4451 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4452 ACLs are no longer actually imposed (and a message to that effect is
4455 22. If a "gecos" field in a passwd entry contained escaped characters, in
4456 particular, if it contained a \" sequence, Exim got it wrong when building
4457 a From: or a Sender: header from that name. A second bug also caused
4458 incorrect handling when an unquoted " was present following a character
4459 that needed quoting.
4461 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4462 was not being matched caselessly.
4464 24. Arranged for all hyphens in the exim.8 source to be escaped with
4467 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4468 properly. Recipient callout cache records were still being keyed to include
4469 the sender, even when use_sender was set false. This led to far more
4470 callouts that were necessary. The sender is no longer included in the key
4471 when use_sender is false.
4473 26. Added "control = submission" modifier to ACLs.
4475 27. Added the ${base62d: operator to decode base 62 numbers.
4477 28. dnsdb lookups can now access SRV records.
4479 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4480 the configuration file.
4482 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4483 (-v) mode. This makes the output for a verbose queue run more intelligible.
4485 31. Added a use_postmaster feature to recipient callouts.
4487 32. Added the $body_zerocount variable, containing the number of binary zero
4488 bytes in the message body.
4490 33. The time of last modification of the "new" subdirectory is now used as the
4491 "mailbox time last read" when there is a quota error for a maildir
4494 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4496 35. Added +ignore_unknown as a special item in host lists.
4498 36. Code for decoding IPv6 addresses in host lists is now included, even if
4499 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4500 address was recognized as an IP address, but was then not correctly decoded
4501 into binary, causing unexpected and incorrect effects when compared with
4508 1. Very minor rewording of debugging text in manualroute to say "list of
4509 hosts" instead of "hostlist".
4511 2. If verify=header_syntax was set, and a header line with an unqualified
4512 address (no domain) and a large number of spaces between the end of the
4513 name and the colon was received, the reception process suffered a buffer
4514 overflow, and (when I tested it) crashed. This was caused by some obsolete
4515 code that should have been removed. The fix is to remove it!
4517 3. When running in the test harness, delay a bit after writing a bounce
4518 message to get a bit more predictability in the log output.
4520 4. Added a call to search_tidyup() just before forking a reception process. In
4521 theory, someone could use a lookup in the expansion of smtp_accept_max_
4522 per_host which, without the tidyup, could leave open a database connection.
4524 5. Added the variables $recipient_data and $sender_data which get set from a
4525 lookup success in an ACL "recipients" or "senders" condition, or a router
4526 "senders" option, similar to $domain_data and $local_part_data.
4528 6. Moved the writing of debug_print from before to after the "senders" test
4531 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4532 problems for message scanning, either using a data ACL, or using
4533 local_scan() because the Received: header was not generated till after they
4534 were called (in order to set the time as the time of reception completion).
4535 I have revised the way this works. The header is now generated after the
4536 body is received, but before the ACL or local_scan() are called. After they
4537 are run, the timestamp in the header is updated.
4543 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4544 before starting a queue runner without re-exec. This happened only when
4545 deliver_drop_privilege was set or when the Exim user was set to root. The
4546 effect of the bug was that timeouts during subsequent deliveries caused
4547 crashes instead of being properly handled. The handler is now left at its
4548 default (and expected) setting.
4550 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4551 message, again when deliver_drop_privilege is set or Exim is run as root.
4552 The bug described in (1) was not present in this case, but the tidying up
4553 of the other signals was missing. I have made the two cases consistent.
4555 3. The ignore_target_hosts setting on a manualroute router was being ignored
4556 for hosts that were looked up using the /MX notation.
4558 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4561 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4562 operated on the sender address. After changing the $sender_address to <>
4563 for the sender address verify, Exim was re-instated it as the original
4564 (before rewriting) address, but remembering that it had rewritten it, so it
4565 wasn't rewriting it again. This bug also had the effect of breaking the
4566 sender address verification caching when the sender address was rewritten.
4568 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4569 This has been changed so that if the ip literal address matches
4570 ignore_target_hosts, the router declines.
4572 7. Added expansion conditions match_domain, match_address, and match_local_
4573 part (NOT match_host).
4575 8. The placeholder for the Received: header didn't have a length field set.
4577 9. Added code to Exim itself and to exim_lock to test for a specific race
4578 condition that could lead to file corruption when using MBX delivery. The
4579 issue is with the lockfile that is created in /tmp. If this file is removed
4580 after a process has opened it but before that process has acquired a lock,
4581 there is the potential for a second process to recreate the file and also
4582 acquire a lock. This could lead to two Exim processes writing to the file
4583 at the same time. The added code performs the same test as UW imapd; it
4584 checks after acquiring the lock that its file descriptor still refers to
4585 the same named file.
4587 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4588 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4591 11. Added the smtp_active_hostname option. If used, this will typically be made
4592 to depend on the incoming interface address. Because $interface_address is
4593 not set up until the daemon has forked a reception process, error responses
4594 that can happen earlier (such as "too many connections") no longer contain
4597 12. If an expansion in a condition on a "warn" statement fails because a lookup
4598 defers, the "warn" statement is abandoned, and the next ACL statement is
4599 processed. Previously this caused the whole ACL to be aborted.
4601 13. Added the iplsearch lookup type.
4603 14. Added ident_timeout as a log selector.
4605 15. Added tls_certificate_verified as a log selector.
4607 16. Added a global option tls_require_ciphers (compare the smtp transport
4608 option of the same name). This controls incoming TLS connections.
4610 17. I finally figured out how to make tls_require_ciphers do a similar thing
4611 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4612 before starting the TLS session.
4614 18. Tabs are now shown as \t in -bP output.
4616 19. If the log selector return_path_on_delivery was set, Exim crashed when
4617 bouncing a message because it had too many Received: header lines.
4619 20. If two routers both had headers_remove settings, and the first one included
4620 a superfluous trailing colon, the final name in the first list and the
4621 first name in the second list were incorrectly joined into one item (with a
4622 colon in the middle).
4628 1. Added -C and -D options to the exinext utility, mainly to make it easier
4629 to include in the automated testing, but these could be helpful when
4630 multiple configurations are in use.
4632 2. The exinext utility was not formatting the output nicely when there was
4633 an alternate port involved in the retry record key, nor when there was a
4634 message id as well (for retries that were specific to a specific message
4635 and a specific host). It was also confused by IPv6 addresses, because of
4636 the additional colons they contain. I have fixed the IPv4 problem, and
4637 patched it up to do a reasonable job for IPv6.
4639 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4640 delivery, the log line now contains "pipelined" if PIPELINING was used.
4642 4. An SMTP transport process used to panic and die if the bind() call to set
4643 an explicit outgoing interface failed. This has been changed; it is now
4644 treated in the same way as a connect() failure.
4646 5. A reference to $sender_host_name in the part of a conditional expansion
4647 that was being skipped was still causing a DNS lookup. This no longer
4650 6. The def: expansion condition was not recognizing references to header lines
4651 that used bh_ and bheader_.
4653 7. Added the _cache feature to named lists.
4655 8. The code for checking quota_filecount in the appendfile transport was
4656 allowing one more file than it should have been.
4658 9. For compatibility with Sendmail, the command line option
4666 and sets the incoming protocol and host name (for trusted callers). The
4667 host name and its colon can be omitted when only the protocol is to be set.
4668 Note the Exim already has two private options, -pd and -ps, that refer to
4669 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4670 "s", but I don't think that's a major issue.
4672 10. A number of refactoring changes to the code, none of which should affect
4675 (a) The number of logging options was getting close to filling up the
4676 32-bit word that was used as a bit map. I have split them into two classes:
4677 those that are passed in the argument to log_write(), and those that are
4678 only ever tested independently outside of that function. These are now in
4679 separate 32-bit words, so there is plenty of room for expansion again.
4680 There is no change in the user interface or the logging behaviour.
4682 (b) When building, for example, log lines, the code previously used a
4683 macro that called string_cat() twice, in order to add two strings. This is
4684 not really sufficiently general. Furthermore, there was one instance where
4685 it was actually wrong because one of the argument was used twice, and in
4686 one call a function was used. (As it happened, calling the function twice
4687 did not affect the overall behaviour.) The macro has been replaced by a
4688 function that can join an arbitrary number of extra strings onto a growing
4691 (c) The code for expansion conditions now uses a table and a binary chop
4692 instead of a serial search (which was left over from when there were very
4693 few conditions). Also, it now recognizes conditions like "pam" even when
4694 the relevant support is not compiled in: a suitably worded error message is
4695 given if an attempt is made to use such a condition.
4697 11. Added ${time_interval:xxxxx}.
4699 12. A bug was causing one of the ddress fields not to be passed back correctly
4700 from remote delivery subprocesses. The field in question was not being
4701 subsequently used, so this caused to problems in practice.
4703 13. Added new log selectors queue_time and deliver_time.
4705 14. Might have fixed a bug in maildirsizefile handling that threw up
4706 "unexpected character" debug warnings, and recalculated the data
4707 unnecessarily. In any case, I expanded the warning message to give more
4710 15. Added the message "Restricted characters in address" to the statements in
4711 the default ACL that block characters like @ and % in local parts.
4713 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4714 Three changes have been made:
4716 (a) There was a serious bug; a negative response to MAIL caused the whole
4717 recipient domain to be cached as invalid, thereby blocking all messages
4718 to all local parts at the same domain, from all senders. This bug has
4719 been fixed. The domain is no longer cached after a negative response to
4720 MAIL if the sender used is not empty.
4722 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4725 (c) A new callout option, "use_sender" has been added for people who want
4726 the modified behaviour.
4732 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4735 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4736 indeed breaks things for older releases.
4738 3. Added additional logging to the case where there is a problem reading data
4739 from a filter that is running in a subprocess using a pipe, in order to
4740 try to track down a specific problem.
4742 4. Testing facility fudge: when running in the test harness and attempting
4743 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4744 getting "No route to host". Convert this to a timeout.
4746 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4749 6. Some OS don't have socklen_t but use size_t instead. This affects the
4750 fifth argument of getsockopt() amongst other things. This is now
4751 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4752 can be set for individual OS. I have set it for SunOS5, OSF1, and
4753 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4754 some earlier ones do not.
4756 7. Change 4.30/15 was not doing the test caselessly.
4758 8. The standard form for an IPv6 address literal was being rejected by address
4759 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4760 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4761 this, as well as the form without the "IPv6" on the front (but only when
4762 address literals are enabled, of course).
4764 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4766 10. Exim crashed if a message with an empty sender address specified by -f
4767 encountered a router with an errors_to setting. This could be provoked only
4768 by a command such as
4772 where an empty string was supplied; "<>" did not hit this bug.
4774 11. Installed PCRE release 4.5.
4776 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4777 remained set. It is now erased.
4779 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4780 times from message ids (which are base 36 rather than the normal 62).
4782 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4783 were being counted as actual protocol errors, and logged if the log
4784 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4785 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4786 and DATA following a set of rejected RCPTs do not count as protocol errors.
4787 In other words, Exim assumes they were pipelined, though this may not
4788 actually be the case. Of course, in all cases the client gets an
4789 appropriate error code.
4791 15. If a lookup fails in an ACL condition, a message about the failure may
4792 be available; it is used if testing the ACL cannot continue, because most
4793 such messages specify what the cause of the deferral is. However, some
4794 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4795 that caused an old message to be retained and used if a later statement
4796 caused a defer, replacing the real cause of the deferral.
4798 16. If an IP address had so many PTR records that the DNS lookup buffer
4799 was not large enough to hold them, Exim could crash while trying to process
4800 the truncated data. It now detects and logs this case.
4802 17. Further to 4.21/58, another change has been made: if (and only if) the
4803 first line of a message (the first header line) ends with CRLF, a bare LF
4804 in a subsequent header line has a space inserted after it, so as not to
4805 terminate the header.
4807 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4808 unnecessarily, used atoi() instead of strtol(), and didn't check the
4809 termination when getting file sizes from file names by regex.
4811 19. Completely re-implemented the support for maildirsize files, in the light
4812 of a number of problems with the previous contributed implementation
4813 (4.30/29). In particular:
4815 . If the quota is zero, the maildirsize file is maintained, but no quota is
4818 . If the maildir directory does not exist, it is created before any attempt
4819 to write a maildirsize file.
4821 . The quota value in the file is just a cache; if the quota is changed in
4822 the transport, the new value overrides.
4824 . A regular expression is available for excluding directories from the
4827 20. The autoreply transport checks the characters in options that define the
4828 message's headers; it allows continued headers, but it was checking with
4829 isspace() after an embedded newline instead of explicitly looking for a
4832 21. If all the "regular" hosts to which an address was routed had passed their
4833 expiry times, and had not reached their retry times, the address was
4834 bounced, even if fallback hosts were defined. Now Exim should go on to try
4837 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4838 equivalent code in the SMTP transport. Some hosts send humungous responses
4839 to HELO/EHLO, more than 1024 it seems.
4841 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4842 gives compiler warnings in some environments. I've now done it "properly",
4845 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4846 (because of problems with the built-in one) was declared to return uschar *
4847 instead of char *, causing compiler failure.
4849 25. Fixed a file descriptor leak when processing alias/forward files.
4851 26. Fixed a minor format string issue in dbfn.c.
4853 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4855 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4856 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4859 29. When a sender address is verified, it is cached, to save repeating the test
4860 when there is more than one recipient in a message. However, when the
4861 verification involves a callout, it is possible for different callout
4862 options to be set for different recipients. It is too complicated to keep
4863 track of this in the cache, so now Exim always runs a verification when a
4864 callout is required, relying on the callout cache for the optimization.
4865 The overhead is duplication of the address routing, but this should not be
4868 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4869 to be verified with callout=postmaster, and the main callout worked but the
4870 postmaster check failed, the verification correctly failed. However, if a
4871 subsequent RCPT command asked for sender verification *without* the
4872 postmaster check, incorrect caching caused this verification also to fail,
4875 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4876 it was not caching the DNS options (qualify_single, search_parents) that
4877 were used when the lookup failed. A subsequent lookup with different
4878 options therefore always gave the same answer, though there were cases
4879 where it should not have. (Example: a "domains = !$mx_any" option on a
4880 dnslookup router: the "domains" option is always processed without any
4881 widening, but the router might have qualify_single set.) Now Exim uses the
4882 cached value only when the same options are set.
4884 32. Added John Jetmore's "exipick" utility to the distribution.
4886 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4887 than a timeout (e.g. a certificate is required, and is not provided), an
4888 Exim server now closes the connection immediately. Previously it waited for
4889 the client to close - but if the client is SSL, it seems that they each
4890 wait for each other, leading to a delay before one of them times out.
4892 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4893 maintained 0.8.x compatibility because I don't think many are using it, and
4894 it is clearly obsolete.
4896 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4899 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4900 client certificate was expired. A simple patch fixes this, though I don't
4901 understand the full logic of why the verify callback is called multiple
4904 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4905 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4906 which causes problems with some clients (such as the Certicom SSL Plus
4907 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4908 disables the coutermeasure allowing Eudora to connect."
4910 38. Exim was not checking that a write() to a log file succeeded. This could
4911 lead to Bad Things if a log got too big, in particular if it hit a file
4912 size limit. Exim now panics and dies if it cannot write to a log file, just
4913 as it does if it cannot open a log file.
4915 39. Modified OS/Makefile-Linux so that it now contains
4917 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4919 The two -D definitions ensure that Exim is compiled with large file
4920 support, which makes it possible to handle log files that are bigger than
4923 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4924 instance) a domain was checked against a named list that involved a lookup,
4925 causing $domain_data to be set, then another domain was checked against the
4926 same list, then the first domain was re-checked, the value of $domain_data
4927 after the final check could be wrong. In particular, if the second check
4928 failed, it could be set empty. This bug probably also applied to
4931 41. The strip_trailing_dot option was not being applied to the address given
4932 with the -f command-line option.
4934 42. The code for reading a message's header from the spool was incrementing
4935 $received_count, but never initializing it. This meant that the value was
4936 incorrect (doubled) while delivering a message in the same process in which
4937 it was received. In the most common configuration of Exim, this never
4938 happens - a fresh exec is done - but it can happen when
4939 deliver_drop_privilege is set.
4941 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4942 it now includes up to 150 characters of the unexpected data in the log
4945 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4946 and building data strings. The size of both of these buffers was 10 000
4947 bytes - far larger than anybody would *ever* want, thought I. Needless to
4948 say, somebody hit the limit. I have increased the maximum line length to
4949 20 000 and the maximum data length of concatenated lines to 100 000. I have
4950 also fixed two bugs, because there was no checking on these buffers. Tsk,
4951 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4952 buffer is too small.
4954 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4955 lsearch lookups. Now it does.
4957 46. When parsing a route_list item in a manualroute router, a fixed-length
4958 buffer was used for the list of hosts. I made this 1024 bytes long,
4959 thinking that nobody would ever have a list of hosts that long. Wrong.
4960 Somebody had a whole pile of complicated expansion conditions, and the
4961 string was silently truncated, leading to an expansion error. It turns out
4962 that it is easier to change to an unlimited length (owing to other changes
4963 that have happened since this code was originally written) than to build
4964 structure for giving a limitation error. The length of the item that
4965 expands into the list of hosts is now unlimited.
4967 47. The lsearch lookup could not handle data where the length of text line was
4968 more than 4095 characters. Such lines were truncated, leading to shortened
4969 data being returned. It should now handle lines of any length.
4971 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4972 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4975 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4976 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4977 no longer happens in comment lines. A list of replacements is now placed
4978 at the head of all of the source files, except those whose only change is
4979 to replace PERL_COMMAND in the very first #! line.
4981 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4982 messages on the queue, with a bottom-up merge sort, using code contributed
4983 by Michael Haardt. This should make operations like -bp somewhat faster on
4984 large queues. It won't affect queue runners, except when queue_run_in_order
4987 51. Installed eximstats 1.31 in the distribution.
4989 52. Added support for SRV lookups to the dnslookup router.
4991 53. If an ACL referred to $message_body or $message_body_end, the value was not
4992 reset for any messages that followed in the same SMTP session.
4994 54. The store-handling optimization for building very long strings was not
4995 differentiating between the different store pools. I don't think this
4996 actually made any difference in practice, but I've tidied it.
4998 55. While running the routers to verify a sender address, $sender_address
4999 was still set to the sender address. This is wrong, because when routing to
5000 send a bounce to the sender, it would be empty. Therefore, I have changed
5001 it so that, while verifying a sender address, $sender_address is set to <>.
5002 (There is no change to what happens when verifying a recipient address.)
5004 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5005 target A or AAAA records (if not already returned) without resetting the
5006 qualify_single or search_parents options of the DNS resolver. These are
5007 inappropriate in this case because the targets of MX and SRV records must
5008 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5009 target that, when qualified, matched something in the local domain. These
5010 two options are now turned off when doing these lookups.
5012 57. It seems that at least some releases of Reiserfs (which does not have the
5013 concept of a fixed number of inodes) returns zero and not -1 for the
5014 number of available inodes. This interacted badly with check_spool_inodes,
5015 which assumed that -1 was the "no such thing" setting. What I have done is
5016 to check that the total number of inodes is greater than zero before doing
5017 the test of how many are available.
5019 58. When a "warn" ACL statement has a log_message modifier, the message is
5020 remembered, and not repeated. This is to avoid a lot of repetition when a
5021 message has many recipients that cause the same warning to be written.
5022 Howewer, Exim was preserving the list of already written lines for an
5023 entire SMTP session, which doesn't seem right. The memory is now reset if a
5024 new message is started.
5026 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5027 debugging output unless log_rewrite was also set.
5029 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5030 of (int)(handle) when we know that handle contains (void *)(-1).
5032 61. The Exim daemon panic-logs an error return when it closes the incoming
5033 connection. However "connection reset by peer" seems to be common, and
5034 isn't really an error worthy of noting specially, so that particular error
5037 62. When Exim is trying to find all the local interfaces, it used to panic and
5038 die if the ioctl to get the interface flags failed. However, it seems that
5039 on at least one OS (Solaris 9) it is possible to have an interface that is
5040 included in the list of interfaces, but for which you get a failure error
5041 for this call. This happens when the interface is not "plumbed" into a
5042 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5043 failure of the "get flags" call assumes that the interface is down.
5045 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5046 makes life easier for people who are doing arithmetic on fields extracted
5047 from dates, where you often get leading zeros that should not be
5048 interpreted as octal.
5050 64. Added qualify_domain to the redirect router, to override the global
5053 65. If a pathologically long header line contained very many addresses (the
5054 report of this problem mentioned 10 000) and each of them was rewritten,
5055 Exim could use up a very large amount of memory. (It kept on making new
5056 copies of the header line as it rewrote, and never released the old ones.)
5057 At the expense of a bit more processing, the header rewriting function has
5058 been changed so that it no longer eats memory in this way.
5060 66. The generation of the Received: header has been moved from the time that a
5061 message starts to be received, to the time that it finishes. The timestamp
5062 in the Received: header should now be very close to that of the <= log
5063 line. There are two side-effects of this change:
5065 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5066 logged header lines no longer include the local Received: line, because
5067 it has not yet been created. The same applies to a copy of the message
5068 that is returned to a non-SMTP sender when a message is rejected.
5070 (b) When a filter file is tested using -bf, no additional Received: header
5071 is added to the test message. After some thought, I decided that this
5074 This change does not affect the value of $received_for. It is still set
5075 after address rewriting, but before local_scan() is called.
5077 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5079 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5080 gave an unhelpful panic error message, and a defer error. I have managed to
5081 change this behaviour so that it now rejects any supplied certificate,
5082 which seems right, as the list of acceptable certificates is empty.
5084 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5085 gave an unhelpful defer error. I have not managed to make this reject any
5086 supplied certificates, but the error message it gives is "no certificate
5087 supplied", which is not helpful.
5089 70. exigrep's output now also includes lines that are not associated with any
5090 message, but which match the given pattern. Implemented by a patch from
5091 Martin Sluka, which also tidied up the Perl a bit.
5093 71. Recipient callout verification, like sender verification, was using <> in
5094 the MAIL FROM command. This isn't really the right thing, since the actual
5095 sender may affect whether the remote host accepts the recipient or not. I
5096 have changed it to use the actual sender in the callout; this means that
5097 the cache record is now keyed on a recipient/sender pair, not just the
5098 recipient address. There doesn't seem to be a real danger of callout loops,
5099 since a callout by the remote host to check the sender would use <>.
5100 [SEE ABOVE: changed after hitting problems.]
5102 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5103 temporary errors. However, in the case of such a code being given after
5104 the end of a data transmission (i.e. after ".") Exim was failing to write
5105 a retry record for the message. (Yes, there was some broken host that was
5106 actually sending 8xx at this point.)
5108 73. An unknown lookup type in a host list could cause Exim to panic-die when
5109 the list was checked. (An example that provoked this was putting <; in the
5110 middle of a list instead of at the start.) If this happened during a DATA
5111 ACL check, a -D file could be left lying around. This kind of configuration
5112 error no longer causes Exim to die; instead it causes a defer errror. The
5113 incident is still logged to the main and panic logs.
5115 74. Buglet left over from Exim 3 conversion. The message "too many messages
5116 in one connection" was written to the rejectlog but not the mainlog, except
5117 when address rewriting (yes!) was being logged.
5119 75. Added write_rejectlog option.
5121 76. When a system filter was run not as root (that is, when system_filter_user
5122 was set), the values of the $n variables were not being returned to the
5123 main process; thus, they were not subsequently available in the $sn
5126 77. Added +return_path_on_delivery log selector.
5128 78. A connection timeout was being treated differently from recipients deferred
5129 when testing hosts_max_try with a message that was older than the host's
5130 retry timeout. (The host should not be counted, thus allowing all hosts to
5131 be tried at least once before bouncing.) This may have been the cause of an
5132 occasionally reported bug whereby a message would remain on the queue
5133 longer than the retry timeout, but would be bounced if a delivery was
5134 forced. I say "may" because I never totally pinned down the problem;
5135 setting up timeout/retry tests is difficult. See also the next item.
5137 79. The ultimate address timeout was not being applied to errors that involved
5138 a combination of host plus message (for example, a timeout on a MAIL
5139 command). When an address resolved to a number of possible hosts, and they
5140 were not all tried for each delivery (e.g. because of hosts_max_try), a
5141 message could remain on the queue longer than the retry timeout.
5143 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5146 81. Fixed an obscure SMTP outgoing bug which required at least the following
5147 conditions: (a) there was another message waiting for the same server;
5148 (b) the server returned 5xx to all RCPT commands in the first message so
5149 that the message was not completed; (c) the server dropped the connection
5150 or gave a negative response to the RSET that Exim sends to abort the
5151 transaction. The observed case was a dropped connection after DATA that had
5152 been sent in pipelining mode. That is, the server had advertised PIPELINING
5153 but was not implementing it correctly. The effect of the bug was incorrect
5154 behaviour, such as trying another host, and this could lead to a crash.
5160 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5161 and daemon.c were passed as pointers to ints; they should have been
5162 pointers to socklen_t variables (which are typically unsigned ints).
5164 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5167 3. Fixed a really odd bug that affected only the testing scheme; patching a
5168 certain fixed string in the binary changed the value of another string that
5169 happened to be identical to the end of the original first string.
5171 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5172 name", it returns that address as the IP address. On some operating
5173 systems (e.g. Solaris), it also passes back the IP address string as the
5174 "host name". However, on others (e.g. Linux), it passes back an empty
5175 string. Exim wasn't checking for this, and was changing the host name to an
5176 empty string, assuming it had been canonicized.
5178 5. Although rare, it is permitted to have more than one PTR record for a given
5179 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5180 all the names associated with an address, because they do in Solaris.
5181 However, it seems that they do not in Linux for data that comes from the
5182 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5183 I found this out when I moved to a new Linux workstation and tried to run
5184 the Exim test suite.
5186 To get round this problem I have changed the code so that it now does its
5187 own call to the DNS to look up PTR records when searching for a host name.
5188 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5189 addresses that are only in /etc/hosts are still found.
5191 This behaviour is, however, controlled by an option called host_lookup_
5192 order, which defaults to "bydns:byaddr". If people want to use the other
5193 order, or indeed, just use one or the other means of lookup, they can
5194 specify it in this variable.
5196 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5197 some operating systems, this comes back from gethostbyaddr() as an empty
5198 string, and this is what Exim used to test for. However, it seems that in
5199 other systems, "." is yielded. Exim now tests for this case too.
5201 7. The values of check_spool_space and check_log_space are now held internally
5202 as a number of kilobytes instead of an absolute number of bytes. If a
5203 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5204 kilobyte. This means that much larger values can be stored.
5206 8. Exim monitor: an attempt to get the action menu when not actually pointing
5207 at a message produces an empty menu entitled "No message selected". This
5208 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5209 no entries in it ("Shell widget menu has zero width and/or height"). So I
5210 have added a single, blank menu entry in this case.
5212 9. Added ${quote_local_part.
5214 10. MIME decoding is now applied to the contents of Subject: header lines when
5217 11. Now that a reference to $sender_host_address automatically causes a reverse
5218 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5219 host lookup before query-style lookups in lists that might use this
5220 variable. This has therefore been abolished, and the "net-" prefix is no
5221 longer necessary for query-style lookups.
5223 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5224 have been a typo for LFLAGS, so it has been changed.
5226 13. The install script calls Exim with "-C /dev/null" in order to find the
5227 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5228 to be output. Howeve, since Exim outputs its version number before the
5229 error, it didn't break the script. It just looked ugly. I fixed this by
5230 always allowing "-C /dev/null" if the caller is root.
5232 14. Ignore overlarge ACL variable number when reading spool file - insurance
5233 against a later release with more variables having written the file.
5235 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5236 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5237 this, as well as the form without the "IPv6" on the front.
5239 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5240 OS/Makefile-Darwin file.
5242 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5243 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5245 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5246 variable. However, some OS use ints for inodes. I've added cast to long int
5247 to get rid of the compiler warning.
5249 19. I had forgotten to lock out "/../" in configuration file names when
5250 ALT_CONFIG_PREFIX was set.
5252 20. Routers used for verification do not need to specify transports. However,
5253 if such a router generated a host list, and callout was configured, Exim
5254 crashed, because it could not find a port number from the (non-existent)
5255 transport. It now assumes port 25 in this circumstance.
5257 21. Added the -t option to exigrep.
5259 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5260 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5261 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5262 them is set, LOOKUP_LSEARCH is forced.
5264 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5266 24. Added sender and host information to the "rejected by local_scan()" log
5267 line; previously there was no indication of these.
5269 25. Added .include_if_exists.
5271 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5272 Linux. It turns out that not all file systems support this. Apparently some
5273 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5274 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5275 not supported on the file descriptor, is now ignored when Exim is trying to
5276 sync a directory. This applies only to Linux.
5278 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5280 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5281 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5284 29. Added support for maildirsize files from supplied patch (modified a bit).
5286 30. The use of :fail: followed by an empty string could lead Exim to respond to
5287 sender verification failures with (e.g.):
5289 550 Verification failed for <xxx>
5290 550 Sender verify failed
5292 where the first response line was missing the '-' that indicates it is not
5293 the final line of the response.
5295 31. The loop for finding the name of the user that called Exim had a hardwired
5296 limit of 10; it now uses the value of finduser_retries, which is used for
5297 all other user lookups.
5299 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5302 33. Exim was neglecting to zero errno before one call of strtol() when
5303 expanding a string and expecting an integer value. On some systems this
5304 resulted in spurious "integer overflow" errors. Also, it was casting the
5305 result into an int without checking.
5307 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5308 did not work. The code looks as if it has *never* worked, though it appears
5309 to have been documented since at least releast 1.62. I have made it work.
5311 35. The "timeout_DNS" error in retry rules, also documented since at least
5312 1.62, also never worked. As it isn't clear exactly what this means, and
5313 clearly it isn't a major issue, I have abolished the feature by treating it
5314 as "timeout", and writing a warning to the main and panic logs.
5316 36. The display of retry rules for -brt wasn't always showing the error code
5319 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5320 timeout_connect_A, timeout_connect_MX.
5322 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5323 to the empty sender.
5325 39. The daemon was not analysing the content of -oX till after it had closed
5326 stderr and disconnected from the controlling terminal. This meant that any
5327 syntax errors were only noted on the panic log, and the return code from
5328 the command was 0. By re-arranging the code a little, I've made the
5329 decoding happen first, so such errors now appear on stderr, and the return
5330 code is 1. However, the actual setting up of the sockets still happens in
5331 the disconnected process, so errors there are still only recorded on the
5334 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5335 connections (as happens on some IP stacks) was logged at start up time as
5336 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5337 it from "IPv6 and IPv4", which means that two separate sockets are being
5340 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5341 says whether AF_INET or AF_INET6 was passed as an argument.
5343 42. Exiwhat output was messed up when time zones were included in log
5346 43. Exiwhat now gives more information about the daemon's listening ports,
5347 and whether -tls-on-connect was used.
5349 44. The "port" option of the smtp transport is now expanded.
5351 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5352 silently ignored. Now an error message is written to the main and panic
5355 46. There's a new ACL modifier called "logwrite" which writes to a log file
5356 as soon as it is encountered.
5358 47. Added $local_user_uid and $local_user_gid at routing time.
5360 48. Exim crashed when trying to verify a sender address that was being
5363 49. Exim was recognizing only a space character after ".include". It now also
5364 recognizes a tab character.
5366 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5367 extracting the relevant information from the specification. The man page no
5368 longer contains scrambled data for the -d option, and I've added a section
5369 at the front about calling Exim under different names.
5371 51. Added "extra_headers" argument to the "mail" command in filter files.
5373 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5376 53. Installed eximstats 1.29.
5378 54. Added transport_filter_timeout as a generic transport option.
5380 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5381 Cc: header lines. This was required by RFC 822, but it not required by RFC
5384 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5385 incoming messages that did not have them. Now it does so only if the
5386 message originates locally, that is, if there is no associated remote host
5387 address. When Resent- header lines are present, this applies to the Resent-
5388 lines rather than the non-Resent- lines.
5390 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5391 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5393 58. Messages for configuration errors now include the name of the main
5394 configuration file - useful now that there may be more than one file in a
5395 list (.included file names were always shown).
5397 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5398 for those rare installations that do not start the daemon as root or run it
5399 setuid root. I've cut out the call to initgroups() if the daemon is not
5402 60. The Exim user and group can now be bound into the binary as text strings
5403 that are looked up at the start of Exim's processing.
5405 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5407 62. Added $mailstore_basename variable.
5409 63. Installed patch to sieve.c from Michael Haardt.
5411 64. When Exim failed to open the panic log after failing to open the main log,
5412 the original message it was trying to log was written to stderr and debug
5413 output, but if they were not available (the usual case in production), it
5414 was lost. Now it is written to syslog before the two lines that record the
5415 failures to open the logs.
5417 65. Users' Exim filters run in subprocesses under the user's uid. It is
5418 possible for a "deliver" command or an alias in a "personal" command to
5419 provoke an address rewrite. If logging of address rewriting is configured,
5420 this fails because the process is not running as root or exim. There may be
5421 a better way of dealing with this, but for the moment (because 4.30 needs
5422 to be released), I have disabled address rewrite logging when running a
5423 filter in a non-root, non-exim process.
5429 1. The buildconfig auxiliary program wasn't quoting the value set for
5430 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5431 not defined. This bug was masked in 4.22 by the effect that was fixed in
5434 2. Some messages that were rejected after a message id was allocated were
5435 shown as "incomplete" by exigrep. It no longer does this for messages that
5436 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5438 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5439 have allow_domain_literals set, the ID did not get logged in the <= line.
5440 Domain literals are now always recognized in Message-ID: header lines.
5442 4. The first argument for a ${extract expansion item is the key name or field
5443 number. Leading and trailing spaces in this item were not being ignored,
5444 causing some misleading effects.
5446 5. When deliver_drop_privilege was set, single queue runner processes started
5447 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5448 same command in the process it spins off) were not dropping privilege.
5450 6. When the daemon running as "exim" started a queue runner, it always
5451 re-executed Exim in the spun-off process. This is a waste of effort when
5452 deliver_drop_privilege is set. The new process now just calls the
5453 queue-runner function directly.
5459 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5462 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5463 ignored. Though the use of -oP was forcing the writing of a pid file, it
5464 was always written to the default place.
5466 3. If the message "no IP address found for host xxxx" is generated during
5467 incoming verification, it is now followed by identification of the incoming
5468 connection (so you can more easily find what provoked it).
5470 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5472 5. Added some features to "harden" Exim a bit more against certain attacks:
5474 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5475 be put in Local/Makefile. This is like the never_users runtime option,
5476 but it cannot be overridden. The default setting is "root".
5478 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5479 prefix string with which any file named in a -C command line option
5482 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5483 is retained for -C and -D only if the caller of Exim is root. Without
5484 it, the exim user may also use -C and -D and retain privilege.
5486 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5487 command line option is disabled.
5489 6. Macro names set by the -D option must start with an upper case letter, just
5490 like macro names defined in the configuration file.
5492 7. Added "dereference=" facility to LDAP.
5494 8. Two instances of the typo "uknown" in the source files are fixed.
5496 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5497 the Configure-Makefile script screwed up while processing it.
5499 10. Incorporated PCRE 4.4.
5501 11. The SMTP synchronization check was not operating right at the start of an
5502 SMTP session. For example, it could not catch a HELO sent before the client
5503 waited for the greeting. There is now a check for outstanding input at the
5504 point when the greeting is written. Because of the duplex, asynchronous
5505 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5506 way, but not yet received, when the check is performed.
5508 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5509 on TCP/IP sockets, because this apparently causes some broken clients to
5512 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5513 unchanged) from the Cygwin maintainer.
5515 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5516 was supported instead of testing for mbx. Effectively a typo.
5518 15. The spa authenticator server code was not checking that the input it
5519 received was valid base64.
5521 16. The debug output line for the "set" modifier in ACLs was not showing the
5522 name of the variable that was being set.
5524 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5526 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5527 lookup on its own. Something else was needed to trigger the lookup. For
5528 example, a match in host_lookup or the need for a host name in a host list.
5529 Now, if $sender_host_name is referenced and the host name has not yet been
5530 looked up, a lookup is performed. If the lookup fails, the variable remains
5531 empty, and $host_lookup_failed is set to "1".
5533 19. Added "eqi" as a case-independent comparison operator.
5535 20. The saslauthd authentication condition could segfault if neither service
5536 nor realm was specified.
5538 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5539 error message that was logged was misleading, and incoming SMTP
5540 connections were dropped. The message is now more accurate, and temporary
5541 errors are given to SMTP connections.
5543 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5544 (except RSET) with a 421 error, until QUIT is received. However, it was
5545 failing to send a response to QUIT.
5547 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5548 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5549 ACL is now run after the helo_try_verify_hosts code.
5551 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5552 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5553 case-independent, so other case variants are also recognized). Apparently
5554 some systems use these upper case variants.
5556 25. If more than two messages were waiting for the same host, and a transport
5557 filter was specified for the transport, Exim sent two messages over the
5558 same TCP/IP connection, and then failed with "socket operation on non-
5559 socket" when it tried to send the third.
5561 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5563 27. The extern definition of crypt16() in expand.c was not being excluded when
5564 the OS had its own crypt16() function.
5566 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5567 as a preferred synonym for return_size_limit, both as an option and as an
5570 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5572 30. Changed the default configuration ACL to relax the local part checking rule
5573 for addresses that are not in any local domains. For these addresses,
5574 slashes and pipe symbols are allowed within local parts, but the sequence
5575 /../ is explicitly forbidden.
5577 31. SPA server authentication was not clearing the challenge buffer before
5580 32. log_message in a "warn" ACL statement was writing to the reject log as
5581 well as to the main log, which contradicts the documentation and doesn't
5582 seem right (because no rejection is happening). So I have stopped it.
5584 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5585 However, I am unable to do any testing of this.
5587 34. Fixed an infelicity in the appendfile transport. When checking directories
5588 for a mailbox, to see if any needed to be created, it was accidentally
5589 using path names with one or more superfluous leading slashes; tracing
5590 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5592 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5593 contents are added to the log line that is written for every discarded
5594 recipient. (Previously a log_message setting was ignored.)
5596 36. The ${quote: operator now quotes the string if it is empty.
5598 37. The install script runs exim in order to find its version number. If for
5599 some reason other than non-existence or emptiness, which it checks, it
5600 could not run './exim', it was installing it with an empty version number,
5601 i.e. as "exim-". This error state is now caught, and the installation is
5604 38. An argument was missing from the function that creates an error message
5605 when Exim fails to connect to the socket for saslauthd authentication.
5606 This could cause Exim to crash, or give a corrupted message.
5608 39. Added isip, isip4, and isip6 to ${if conditions.
5610 40. The ACL variables $acl_xx are now saved with the message, and can be
5611 accessed later in routers, transports, and filters.
5613 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5614 strings in the file are not string-expanded.
5616 42. If a MAIL command specified a SIZE value that was too large to fit into an
5617 int variable, the check against message_size_limit failed. Such values are
5618 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5619 day this will have to be increased, but I don't think I want to be around
5620 when emails are that large.
5627 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5628 iconv() is not standard in FreeBSD.
5630 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5631 IPv6 enabled. The observed symptom was a segmentation fault on return from
5632 the function os_common_find_running_interfaces() in src/os.c.
5634 3. In the check_special_case() function in daemon.c I had used "errno" as an
5635 argument name, which causes warnings on some systems. This was basically a
5636 typo, since it was named "eno" in the comments!
5638 4. The code that waits for the clock to tick (at a resolution of some fraction
5639 of a second) so as to ensure message-id uniqueness was always waiting for
5640 at least one whole tick, when it could have waited for less. [This is
5641 almost certainly not relevant at current processor speeds, where it is
5642 unlikely to ever wait at all. But we try to future-proof.]
5644 5. The function that sleeps for a time interval that includes fractions of a
5645 second contained a race. It did not block SIGALRM between setting the
5646 timer, and suspending (a couple of lines later). If the interval was short
5647 and the sigsuspend() was delayed until after it had expired, the suspension
5648 never ended. On busy systems this could lead to processes getting stuck for
5651 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5652 process, before it forks any delivery processes. The open lookup caching
5653 mechanism meant that the open file or database connection was passed into
5654 the delivery process. The problem was that delivery processes always tidy
5655 up cached lookup data. This could cause a problem for the next delivery
5656 process started by the queue runner, because the external queue runner
5657 process does not know about the closure. So the next delivery process
5658 still has data in the lookup cache. In the case of a file lookup, there was
5659 no problem because closing a file descriptor in a subprocess doesn't affect
5660 the parent. However, if the lookup was caching a connection to a database,
5661 the connection was closed, and the second delivery process was likely to
5662 see errors such as "PGSQL: query failed: server closed the connection
5663 unexpectedly". The problem has been fixed by closing all cached lookups
5664 in a queue runner before running a delivery process.
5666 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5667 seem to have the "const" qualifier which it has on other OS. I've
5670 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5671 *of the same type* (client or server) with the same public name that an
5672 error should be diagnosed.
5674 9. When Exim looked up a host name for an IP address, but failed to find the
5675 original IP address when looking up the host name (a safety check), it
5676 output the message "<ip address> does not match any IP for NULL", which was
5677 confusing, to say the least. The bug was that the host name should have
5678 appeared instead of "NULL".
5680 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5681 user that is built into the binary, and the -C or -D options is used, root
5682 privilege is dropped before the configuration file is read. In addition,
5683 logging is switched to stderr instead of the normal log files. If the
5684 configuration then re-defines the Exim user, the unprivileged environment
5685 is probably not what is expected, so Exim logs a panic warning message (but
5688 However, if deliver_drop_privilege is set, the unprivileged state may well
5689 be exactly what is intended, so the warning has been cut out in that case,
5690 and Exim is allowed to try to write to its normal log files.
5696 1. smtp_return_error_details was not giving details for temporary sender
5697 or receiver verification errors.
5699 2. Diagnose a configuration error if two authenticators have the same public
5702 3. Exim used not to create the message log file for a message until the first
5703 delivery attempt. This could be confusing when incoming messages were held
5704 for policy or load reasons. The message log file is now created at the time
5705 the message is received, and an initial "Received" line is written to it.
5707 4. The automatically generated man page for command line options had a minor
5708 bug that caused no ill effects; however, a more serious problem was that
5709 the procedure for building the man page automatically didn't always
5710 operate. Consequently, release 4.20 contains an out-of-date version. This
5711 shouldn't happen again.
5713 5. When building Exim with embedded Perl support, the script that builds the
5714 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5715 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5717 6. The freeze_tell option was not being used for messages that were frozen on
5718 arrival, either by an ACL or by local_scan().
5720 7. Added the smtp_incomplete_transaction log selector.
5722 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5723 was accepting AUTH without a new EHLO.
5725 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5726 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5729 10. Logging of TCP/IP connections (when configured) now happens in the main
5730 daemon process instead of the child process, so that the TCP/IP connection
5731 count is more accurate (but it can never be perfect).
5733 11. The use of "drop" in a nested ACL was not being handled correctly in the
5734 outer ACL. Now, if condition failure induced by the nested "drop" causes
5735 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5736 or "require"), the connection is dropped.
5738 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5739 that yield "discard" can now be used with an "accept" or a "discard" verb,
5740 but an error is generated for any others (because I can't see a useful way
5741 to define what should happen).
5743 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5744 are now processed in the same way as lines in the Exim configuration file.
5745 In particular, continuation lines are supported.
5747 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5749 15. Added -ti meaning -t -i.
5751 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5752 domains, and warn by logging if others are found.
5754 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5755 returned from GIFCONF when Exim is trying to find the list of interfaces on
5756 a host. The code in os.c has been modified to copy each ifreq to an aligned
5757 structure in all cases.
5759 Also, in some cases, the returned ifreq's were being copied to a 'struct
5760 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5761 means the last couple of bytes of an IPv6 address could be chopped if the
5762 ifreq contained only a normal sockaddr (14 bytes storage).
5764 18. Named domain lists were not supported in the hosts_treat_as_local option.
5765 An entry such as +xxxx was not recognized, and was treated as a literal
5768 19. Ensure that header lines added by a DATA ACL are included in the reject log
5769 if the ACL subsequently rejects the message.
5771 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5772 MD5 (which is deprecated).
5774 21. When testing a filter file using -bf, Exim was writing a message when it
5775 took the sender from a "From " line in the message, but it was not doing so
5776 when it took $return_path from a Return-Path: header line. It now does.
5778 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5779 with a valid header line field name (a series of printing characters
5780 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5782 23. Changed "disc" in the source to "disk" to conform to the documentation and
5783 the book and for uniformity.
5785 24. Ignore Sendmail's -Ooption=value command line item.
5787 25. When execve() failed while trying to run a command in a pipe transport,
5788 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5789 could be confused with a return value of 69 from the command itself. This
5790 has been changed to 127, the value the shell returns if it is asked to run
5791 a non-existent command. The wording for the related log line suggests a
5792 non-existent command as the problem.
5794 26. If received_header_text expands to an empty string, do not add a Received:
5795 header line to the message. (Well, it adds a token one on the spool, but
5796 marks it "old" so that it doesn't get used or transmitted.)
5798 27. Installed eximstats 1.28 (addition of -nt option).
5800 28. There was no check for failure on the call to getsockname() in the daemon
5801 code. This can fail if there is a shortage of resources on the system, with
5802 ENOMEM, for example. A temporary error is now given on failure.
5804 29. Contrary to the C standard, it seems that in some environments, the
5805 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5806 program. Exim now does this explicitly; it affects the formatting of
5807 timestamps using strftime().
5809 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5810 complaints. I've now initialized all the variables, to avoid this.
5812 32. Header lines added by a system filter were not being "seen" during
5813 transport-time rewrites.
5815 33. The info_callback() function passed to OpenSSL is set up with type void
5816 (*)(SSL *, int, int), as described somewhere. However, when calling the
5817 function (actually a macro) that sets it up, the type void(*)() is
5818 expected. I've put in a cast to prevent warnings from picky compilers.
5820 34. If a DNS black list lookup found a CNAME record, but there were no A
5821 records associated with the domain it pointed at, Exim crashed.
5823 35. If a DNS black list lookup returned more than one A record, Exim ignored
5824 all but the first. It now scans all returned addresses if a particular IP
5825 value is being sought. In this situation, the contents of the
5826 $dnslist_value variable are a list of all the addresses, separated by a
5829 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5830 to accept a host name and all its aliases if the forward lookup for any of
5831 them yielded the IP address of the incoming connection. Now it accepts only
5832 those names whose forward lookup yields the correct IP address. Any other
5833 names are discarded. This closes a loophole whereby a rogue DNS
5834 administrator could create reverse DNS records to break through a
5835 wildcarded host restriction in an ACL.
5837 37. If a user filter or a system filter that ran in a subprocess used any of
5838 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5839 the wrong values were passed to the pipe command ($thisaddress had the
5840 value of $0, $0 had the value of $1, etc). This bug was introduced by
5841 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5843 38. Improved the line breaking for long SMTP error messages from ACLs.
5844 Previously, if there was no break point between 40 and 75 characters, Exim
5845 left the rest of the message alone. Two changes have been made: (a) I've
5846 reduced the minimum length to 35 characters; (b) if it can't find a break
5847 point between 35 and 75 characters, it looks ahead and uses the first one
5848 that it finds. This may give the occasional overlong line, but at least the
5849 remaining text gets split now.
5851 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5852 file descriptors might be low, and that setting 1000 would always raise it.
5853 It turns out that in some environments, the limit is already over 1000 and
5854 that lowering it causes trouble. So now Exim takes care not to decrease it.
5856 40. When delivering a message, the value of $return_path is set to $sender_
5857 address at the start of routing (routers may change the value). By an
5858 oversight, this default was not being set up when an address was tested by
5859 -bt or -bv, which affected the outcome if any router or filter referred to
5862 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5863 log only when "message" or "log_message" are set. However, if one of the
5864 conditions was an address verification, or a call to a nested ACL, the
5865 messages generated by the underlying test were being passed through. This
5866 no longer happens. The underlying message is available in $acl_verify_
5867 message for both "message" and "log_message" expansions, so it can be
5868 passed through if needed.
5870 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5871 new expansion $bh_ to give the encoded byte string without charset
5872 translation. Translation happens only if iconv() is available; HAVE_ICONV
5873 indicates this at build time. HEADERS_CHARSET gives the charset to
5874 translate to; headers_charset can change it in the configuration, and
5875 "headers charset" can change it in an individual filter file.
5877 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5878 that creates RFC 2047 encoded "words" labels them as that charset instead
5879 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5880 expansion operator; (ii) when Exim creates a From: line for a local
5881 message; (iii) when a header line is rewritten to include a "phrase" part.
5883 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5884 buggy, causing it to skip the first lines of messages whose message ID
5885 ended in 'D'. This would not have bitten before Exim release 4.14, because
5886 message IDs were unlikely to end in 'D' before then. The effect was to have
5887 incorrect size information for certain domains.
5889 45. #include "config.h" was missing at the start of the crypt16.c module. This
5890 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5893 46. If there was a timeout during a "random" callout check, Exim treated it as
5894 a failure of the random address, and carried on sending RSET and the real
5895 address. If the delay was just some slowness somewhere, the response to the
5896 original RCPT would be taken as a response to RSET and so on, causing
5897 mayhem of various kinds.
5899 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5900 when I implemented it. It didn't allow for the fact that some option values
5901 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5902 the right test for positive values.
5904 48. Domain names in DNS records are case-independent. Exim always looks them up
5905 in lower case. Some resolvers return domain names in exactly the case they
5906 appear in the zone file, that is, they may contain uppercase letters. Not
5907 all resolvers do this - some return always lower case. Exim was treating a
5908 change of case by a resolver as a change of domain, similar to a widening
5909 of a domain abbreviation. This triggered its re-routing code and so it was
5910 trying to route what was effectively the same domain again. This normally
5911 caused routing to fail (because the router wouldn't handle the domain
5912 twice). Now Exim checks for this case specially, and just changes the
5913 casing of the domain that it ultimately uses when it transmits the message
5916 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5919 50. If a filter generated a file delivery with a non-absolute name (possible if
5920 no home directory exists for the router), the forbid_file option was not
5923 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5924 the existing equality matching.
5926 52. Exim was using ints instead of ino_t variables in some places where it was
5927 dealing with inode numbers.
5929 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5930 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5931 called TMPDIR, and if it finds it is different, it changes its value.
5933 54. The smtp_printf() function is now made available to local_scan() so
5934 additional output lines can be written before returning. There is also an
5935 smtp_fflush() function to enable the detection of a dropped connection.
5936 The variables smtp_input and smtp_batched_input are exported to
5939 55. Changed the default runtime configuration: the message "Unknown user"
5940 has been removed from the ACL, and instead placed on the localuser router,
5941 using the cannot_route_message feature. This means that any verification
5942 failures that generate their own messages won't get overridden. Similarly,
5943 the "Unrouteable address" message that was in the ACL for unverifiable
5944 relay addresses has also been removed.
5946 56. Added hosts_avoid_esmtp to the smtp transport.
5948 57. The exicyclog script was not checking for the esoteric option
5949 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5950 will work only if exicyclog is run under the appropriate euid.
5952 58. Following a discussion on the list, the rules by which Exim recognises line
5953 endings on incoming messages have been changed. The -dropcr and drop_cr
5954 options are now no-ops, retained only for backwards compatibility. The
5955 following line terminators are recognized: LF CRLF CR. However, special
5956 processing applies to CR:
5958 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5959 nor a local message in the state where . is a terminator.
5961 (ii) If a bare CR is encountered in a header line, an extra space is added
5962 after the line terminator so as not to end the header. The reasoning
5963 behind this is that bare CRs in header lines are most likely either
5964 to be mistakes, or people trying to play silly games.
5966 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5967 was being incorrectly given as 18 bytes larger than it should have been.
5968 This is a VOB (very old bug).
5970 60. This may never have affected anything current, but just in case it has:
5971 When the local host is found other than at the start of a list of hosts,
5972 the local host, those with the same MX, and any that follow, are discarded.
5973 When the list in question was part of a longer list of hosts, the following
5974 hosts (not currently being processed) were also being discarded. This no
5975 longer happens. I'm not sure if this situation could ever has previously
5978 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5981 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5982 header. This is something that is recommended in a new Internet Draft, and
5983 is something that is documented as being done by Sendmail. There are two
5984 possible values. For messages generated by the autoreply transport, Exim
5987 Auto-Submitted: auto-replied
5989 whereas for all other generated messages (e.g. bounces) it adds
5991 Auto-Submitted: auto-generated
5993 63. The "personal" condition in filters now includes a test for the
5994 Auto-Submitted: header. If it contains the string "auto-" the message it
5995 not considered personal.
5997 64. Added rcpt_include_affixes as a generic transport option.
5999 65. Added queue_only_override (default true).
6001 66. Added the syslog_duplication option.
6003 67. If what should have been the first header line of a message consisted of
6004 a space followed by a colon, Exim was mis-interpreting it as a header line.
6005 It isn't of course - it is syntactically invalid and should therefore be
6006 treated as the start of the message body. The misbehaviour could have
6007 caused a number of strange effects, including loss of data in subsequent
6008 header lines, and spool format errors.
6010 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6011 client host had authenticated. This control can now be exercised by an ACL
6012 for more flexibility.
6014 69. By default, callouts do not happen when testing with -bh. There is now a
6015 variant, -bhc, which does actually run the callout code, including
6016 consulting and updating the callout cache.
6018 70. Added support for saslauthd authentication, courtesy of Alexander
6021 71. If statvfs() failed on the spool or log directories while checking their
6022 size for availability, Exim confusingly gave the error "space shortage".
6023 Furthermore, in debugging mode it crashed with a floating point exception.
6024 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6025 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6026 serious problem, Exim now writes to the main and panic logs when this
6027 happens, with details of the failure. It then refuses to accept the
6028 incoming message, giving the message "spool directory problem" or "log
6029 directory problem" with a 421 code for SMTP messages.
6031 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6032 0, 1, and 2 exist, because some OS complain for execs without them (see
6033 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6034 descriptors. However, the code omitted to check that the open succeeded,
6035 causing mysterious errors if for some reason the permissions on /dev/null
6036 got screwed. Now Exim writes a message to the main and panic logs, and
6037 bombs out if it can't open /dev/null.
6039 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6040 interact so that it is all more flexible. It is supposed to remain
6041 backwards compatible. Also added extra_local_interfaces.
6043 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6044 to bomb out with an assertion failure - to the client this appears as a
6045 connection drop. This problem occurs in the part of the code that was taken
6046 from the Samba project. Fortunately, the assertion is in a very simple
6047 function, so I have fixed this by reproducing the function inline in the
6048 one place where it is called, and arranging for authentication to fail
6049 instead of killing the process with assert().
6051 75. The SPA client code was not working when the server requested OEM rather
6052 than Unicode encoding.
6054 76. Added code to make require_files with a specific uid setting more usable in
6055 the case where statting the file as root fails - usually a non-root-mounted
6056 NFS file system. When this happens and the failure is EACCES, Exim now
6057 forks a subprocess and does the per-uid checking as the relevant uid.
6059 77. Added process_log_path.
6061 78. If log_file_path was not explicitly set, a setting of check_log_space or
6062 check_log_inodes was ignored.
6064 79. If a space check for the spool or log partitions fails, the incident is now
6065 logged. Of course, in the latter case the data may get lost...
6067 80. Added the %p formatting code to string_format() so that it can be used to
6068 print addresses in debug_print(). Adjusted all the address printing in the
6069 debugging in store.c to use %p rather than %d.
6071 81. There was a concern that a line of code in smtp_in.c could overflow a
6072 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6073 initially expressed, the concern was not well-founded, because trailing
6074 spaces are removed early. However, if the trailing spaces were followed by
6075 a NULL, they did not get removed, so the overflow was possible. Two fixes
6078 (a) I re-wrote the offending code in a cleaner fashion.
6079 (b) If an incoming SMTP command contains a NULL character, it is rejected
6082 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6083 runs initgroups(), so that if the Exim user is in any additional groups,
6084 they will be used during message reception.
6090 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob