1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
18 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
19 checkers that were complaining about end of non-void function with no
22 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
23 This was a regression intruduced in 4.83 by another bugfix.
25 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
27 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
28 EXPERIMENTAL_DNS is enabled. Fix from Wolfgang Breyha.
34 TF/01 Correctly close the server side of TLS when forking for delivery.
36 When a message was received over SMTP with TLS, Exim failed to clear up
37 the incoming connection properly after forking off the child process to
38 deliver the message. In some situations the subsequent outgoing
39 delivery connection happened to have the same fd number as the incoming
40 connection previously had. Exim would try to use TLS and fail, logging
41 a "Bad file descriptor" error.
43 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
44 utilities have not been installed.
46 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
47 temporary space as the ACL may create new global variables.
49 TL/01 LDAP support uses per connection or global context settings, depending
50 upon the detected version of the libraries at build time.
52 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
53 to extract and use the src ip:port in logging and expansions as if it
54 were a direct connection from the outside internet. PPv2 support was
55 updated based on HAProxy spec change in May 2014.
57 JH/02 Add ${listextract {number}{list}{success}{fail}}.
59 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
60 Properly escape header and check for NULL return.
62 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
65 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
67 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
68 characters in header names, implemented as a verify condition.
69 Contributed by Michael Fischer v. Mollard.
71 TL/05 Rename SPF condition results err_perm and err_temp to standardized
72 results permerror and temperror. Previous values are deprecated but
73 still accepted. In a future release, err_perm and err_temp will be
74 completely removed, which will be a backward incompatibility if the
75 ACL tests for either of these two old results. Patch contributed by
76 user bes-internal on the mailing list.
78 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
80 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
81 selectors, in both main and reject logs.
83 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
86 JH/07 Add malware type "sock" for talking to simple daemon.
88 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
90 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
91 routers/transports under cutthrough routing.
93 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
94 numbers. Touch up "bool" conditional to keep the same definition.
96 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
98 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
100 JH/12 Expand items in router/transport headers_add or headers_remove lists
101 individually rather than the list as a whole. Bug 1452.
103 Required for reasonable handling of multiple headers_ options when
104 they may be empty; requires that headers_remove items with embedded
105 colons must have them doubled (or the list-separator changed).
107 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
108 view the policy declared in the DMARC record. Currently, $dmarc_status
109 is a combined value of both the record presence and the result of the
112 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
114 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
115 dnslookup router and the smtp transport (applying to the forward
118 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
119 of ldap servers used for a specific lookup. Patch provided by Heiko
122 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
123 New variable $lookup_dnssec_authenticated for observability.
125 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
126 Patch submitted by Lars Timman.
128 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
130 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
131 Requires trusted mode and valid format message id, aborts otherwise.
132 Patch contributed by Heiko Schlichting.
134 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
135 certextract with support for various fields. Bug 1358.
137 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
138 is requested by default, modifiable by smtp transport option
141 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
142 operate on certificate variables to give certificate fingerprints
143 Also new ${sha256:cert_variable}.
145 JH/23 The PRDR feature is moved from being Experimental into the mainline.
147 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
148 Christian Aistleitner.
150 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
152 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
153 file. Patch from Wolfgang Breyha.
155 JH/25 Expand the coverage of the delivery $host and $host_address to
156 client authenticators run in verify callout. Bug 1476.
158 JH/26 Port service names are now accepted for tls_on_connect_ports, to
159 align with daemon_smtp_ports. Bug 72.
161 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
162 support and error reporting did not work properly.
164 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
165 and is readable. Patch from Andrew Colin Kissa.
167 TL/14 Enhance documentation of ${run expansion and how it parses the
168 commandline after expansion, particularly in the case when an
169 unquoted variable expansion results in an empty value.
171 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
173 PP/02 Fix internal collision of T_APL on systems which support RFC3123
174 by renaming away from it. Addresses GH issue 15, reported by
177 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
179 TL/15 SECURITY: prevent double expansion in math comparison functions
180 (can expand unsanitized data). Not remotely exploitable.
187 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
189 PP/02 Make -n do something, by making it not do something.
190 When combined with -bP, the name of an option is not output.
192 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
195 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
196 $sender_host_name and config options to manage this, and basic check
199 PP/05 DSCP support for outbound connections and control modifier for inbound.
201 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
202 (Only plugin which currently uses this is kerberos4, which nobody should
203 be using, but we should make it available and other future plugins might
204 conceivably use it, even though it would break NAT; stuff *should* be
205 using channel bindings instead).
207 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
208 name; added for Sendmail compatibility; requires admin caller.
209 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
210 just ignore it); requires trusted caller.
211 Also parse but ignore: -Ac -Am -X<logfile>
214 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
216 TL/02 Add +smtp_confirmation as a default logging option.
218 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
219 Patch by Magnus Holmgren from 2007-02-20.
221 TL/04 Bugzilla 1281 - Spec typo.
222 Bugzilla 1283 - Spec typo.
223 Bugzilla 1290 - Spec grammar fixes.
225 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
227 TL/06 Add Experimental DMARC support using libopendmarc libraries.
229 TL/07 Fix an out of order global option causing a segfault. Reported to dev
230 mailing list by by Dmitry Isaikin.
232 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
234 JH/02 Support "G" suffix to numbers in ${if comparisons.
236 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
238 NM/01 Bugzilla 1197 - Spec typo
239 Bugzilla 1196 - Spec examples corrections
241 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
243 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
244 gnutls_enable_pkcs11, but renamed to more accurately indicate its
247 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
248 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
250 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
251 "acl {{name}{arg}...}", and optional args on acl condition
254 JH/05 Permit multiple router/transport headers_add/remove lines.
256 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
258 JH/07 Avoid using a waiting database for a single-message-only transport.
259 Performance patch from Paul Fisher. Bugzilla 1262.
261 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
264 JH/09 Add $headers_added variable, with content from use of ACL modifier
265 add_header (but not yet added to the message). Bugzilla 199.
267 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
268 Pulled from Bugzilla 817 by Wolfgang Breyha.
270 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
272 (nb: this is the same fix as in Exim 4.80.1)
274 JH/11 Add A= logging on delivery lines, and a client_set_id option on
277 JH/12 Add optional authenticated_sender logging to A= and a log_selector
280 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
282 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
283 advertise SMTP AUTH mechanism to us, instead of a generic
284 protocol violation error. Also, make Exim more robust to bad
285 data from the Dovecot auth socket.
287 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
289 When a queue runner is handling a message, Exim first routes the
290 recipient addresses, during which it prunes them based on the retry
291 hints database. After that it attempts to deliver the message to
292 any remaining recipients. It then updates the hints database using
295 So if a recipient address works intermittently, it can get repeatedly
296 deferred at routing time. The retry hints record remains fresh so the
297 address never reaches the final cutoff time.
299 This is a fairly common occurrence when a user is bumping up against
300 their storage quota. Exim had some logic in its local delivery code
301 to deal with this. However it did not apply to per-recipient defers
302 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
304 This change adds a proper retry rule check during routing so that the
305 final cutoff time is checked against the message's age. We only do
306 this check if there is an address retry record and there is not a
307 domain retry record; this implies that previous attempts to handle
308 the address had the retry_use_local_parts option turned on. We use
309 this as an approximation for the destination being like a local
310 delivery, as in LMTP.
312 I suspect this new check makes the old local delivery cutoff check
313 redundant, but I have not verified this so I left the code in place.
315 TF/02 Correct gecos expansion when From: is a prefix of the username.
317 Test 0254 submits a message to Exim with the header
321 When I ran the test suite under the user fanf2, Exim expanded
322 the header to contain my full name, whereas it should have added
323 a Resent-Sender: header. It erroneously treats any prefix of the
324 username as equal to the username.
326 This change corrects that bug.
328 GF/01 DCC debug and logging tidyup
329 Error conditions log to paniclog rather than rejectlog.
330 Debug lines prefixed by "DCC: " to remove any ambiguity.
332 TF/03 Avoid unnecessary rebuilds of lookup-related code.
334 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
335 Bug spotted by Jeremy Harris; was flawed since initial commit.
336 Would have resulted in OCSP responses post-SNI triggering an Exim
337 NULL dereference and crash.
339 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
341 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
342 Bug detection, analysis and fix by Samuel Thibault.
343 Bugzilla 1331, Debian bug #698092.
345 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
347 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
348 Server implementation by Todd Lyons, client by JH.
349 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
350 config variable "prdr_enable" controls whether the server
351 advertises the facility. If the client requests PRDR a new
352 acl_data_smtp_prdr ACL is called once for each recipient, after
353 the body content is received and before the acl_smtp_data ACL.
354 The client is controlled by bolth of: a hosts_try_prdr option
355 on the smtp transport, and the server advertisement.
356 Default client logging of deliveries and rejections involving
357 PRDR are flagged with the string "PRDR".
359 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
360 fclose(). Diagnosis by Todd Lyons.
362 PP/17 Update configure.default to handle IPv6 localhost better.
363 Patch by Alain Williams (plus minor tweaks).
366 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
367 This is now consistent with GnuTLS, and is now documented: the
368 previous undocumented portable approach to treating the option as
369 unset was to force an expansion failure. That still works, and
370 an empty string is now equivalent.
372 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
373 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
374 not performing validation itself.
376 PP/20 Added force_command boolean option to pipe transport.
377 Patch from Nick Koston, of cPanel Inc.
379 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
382 TF/04 Added udpsend ACL modifer and hexquote expansion operator
384 PP/21 Fix eximon continuous updating with timestamped log-files.
385 Broken in a format-string cleanup in 4.80, missed when I repaired the
386 other false fix of the same issue.
387 Report and fix from Heiko Schlichting.
390 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
391 Report from Prashanth Katuri.
393 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
394 It's SecureTransport, so affects any MacOS clients which use the
395 system-integrated TLS libraries, including email clients.
397 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
398 using a MIME ACL for non-SMTP local injection.
399 Report and assistance in diagnosis by Warren Baker.
401 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
403 JH/16 Fix comparisons for 64b. Bugzilla 1385.
405 TL/09 Add expansion variable $authenticated_fail_id to keep track of
406 last id that failed so it may be referenced in subsequent ACL's.
408 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
411 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
412 ldap library initialization, allowing self-signed CA's to be
413 used. Also properly sets require_cert option later in code by
414 using NULL (global ldap config) instead of ldap handle (per
415 session). Bug diagnosis and testing by alxgomz.
417 TL/12 Enhanced documentation in the ratelimit.pl script provided in
418 the src/util/ subdirectory.
420 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
421 renamed to Transport Post Delivery Action by Jeremy Harris, as
424 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
425 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
426 redis_servers = needs to be configured which will be used by the redis
427 lookup. Patch from Warren Baker, of The Packet Hub.
429 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
431 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
432 hostname or reverse DNS when processing a host list. Used suggestions
433 from multiple comments on this bug.
435 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
437 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
438 Missed a few lines, added it to make the runtest require no keyboard
441 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
442 contains upper case chars. Make router use caseful_local_part.
444 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
445 support when GnuTLS has been built with p11-kit.
451 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
453 This, or similar/improved, will also be change PP/11 of 4.82.
459 PP/01 Handle short writes when writing local log-files.
460 In practice, only affects FreeBSD (8 onwards).
461 Bugzilla 1053, with thanks to Dmitry Isaikin.
463 NM/01 Bugzilla 949 - Documentation tweak
465 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
468 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
470 PP/02 Implemented gsasl authenticator.
472 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
474 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
475 `pkg-config foo` for cflags/libs.
477 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
478 with rest of GSASL and with heimdal_gssapi.
480 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
481 `pkg-config foo` for cflags/libs for the TLS implementation.
483 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
484 properties get this fed in as external SSF. A number of robustness
485 and debugging improvements to the cyrus_sasl authenticator.
487 PP/08 cyrus_sasl server now expands the server_realm option.
489 PP/09 Bugzilla 1214 - Log authentication information in reject log.
490 Patch by Jeremy Harris.
492 PP/10 Added dbmjz lookup type.
494 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
496 PP/12 MAIL args handles TAB as well as SP, for better interop with
497 non-compliant senders.
498 Analysis and variant patch by Todd Lyons.
500 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
501 Bug report from Lars Müller <lars@samba.org> (via SUSE),
502 Patch from Dirk Mueller <dmueller@suse.com>
504 PP/13 tls_peerdn now print-escaped for spool files.
505 Observed some $tls_peerdn in wild which contained \n, which resulted
506 in spool file corruption.
508 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
509 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
510 or write after TLS renegotiation, which otherwise led to messages
513 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
514 as a tracking header (ie: a signed header comes before the signature).
515 Patch from Wolfgang Breyha.
517 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
518 comma-sep list; embedded commas doubled.
520 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
522 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
524 Report and patch from Dmitry Banschikov.
526 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
527 Removed SSL_clear() after SSL_new() which led to protocol negotiation
528 failures. We appear to now support TLS1.1+ with Exim.
530 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
531 lets Exim select keys and certificates based upon TLS SNI from client.
532 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
533 before an outbound SMTP session. New log_selector, +tls_sni.
535 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
536 NULL dereference. Report and patch from Alun Jones.
538 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
539 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
540 Not seeing resolver debug output on NetBSD, but suspect this is a
541 resolver implementation change.
543 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
544 Left warnings. Added "eximon gdb" invocation mode.
546 PP/21 Defaulting "accept_8bitmime" to true, not false.
548 PP/22 Added -bw for inetd wait mode support.
550 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
551 locate the relevant includes and libraries. Made this the default.
553 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
554 Bugzilla 1246, report and most of solution from Tomasz Kusy.
556 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
557 This may cause build issues on older platforms.
559 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
560 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
561 gnutls_require_mac & gnutls_require_protocols (no longer supported).
562 Added SNI support via GnuTLS too.
563 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
565 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
567 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
568 Applied second patch from Janne, implementing suggestion to default
569 multiple-strings-in-record handling to match SPF spec.
571 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
573 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
574 read-only, out of scope).
575 Patch from Wolfgang Breyha, report from Stuart Northfield.
577 PP/29 Fix three issues highlighted by clang analyser static analysis.
578 Only crash-plausible issue would require the Cambridge-specific
579 iplookup router and a misconfiguration.
580 Report from Marcin Mirosław.
582 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
584 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
585 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
586 As part of this, removing so much warning spew let me fix some minor
587 real issues in debug logging.
589 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
590 assignment on my part. Fixed.
592 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
593 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
594 Janne Snabb (who went above and beyond: thank you).
596 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
597 string otherwise requires a connection and a bunch more work and it's
598 relatively easy to get wrong. Should also expose TLS library linkage
601 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
602 64-bit ${eval} (JH/03).
604 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
605 GNU libc to support some of the 64-bit stuff, should not lead to
606 conflicts. Defined before os.h is pulled in, so if a given platform
607 needs to override this, it can.
609 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
610 protection layer was required, which is not implemented.
611 Bugzilla 1254, patch from Wolfgang Breyha.
613 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
614 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
615 tls_dhparam take prime identifiers. Also unbreak combination of
616 OpenSSL+DH_params+TLSSNI.
618 PP/39 Disable SSLv2 by default in OpenSSL support.
624 PP/01 Solaris build fix for Oracle's LDAP libraries.
625 Bugzilla 1109, patch from Stephen Usher.
627 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
629 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
632 TF/02 Fix a couple more cases where we did not log the error message
633 when unlink() failed. See also change 4.74-TF/03.
635 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
636 lock up or crash if it happened to be inside a call to libc when it
637 got a SIGUSR1 from exiwhat.
639 The SIGUSR1 handler appends the current process status to the process
640 log which is later printed by exiwhat. It used to use the general
641 purpose logging code to do this, but several functions it calls are
642 not safe for signals.
644 The new output code in the SIGUSR1 handler is specific to the process
645 log, and simple enough that it's easy to inspect for signal safety.
646 Removing some special cases also simplifies the general logging code.
647 Removing the spurious timestamps from the process log simplifies
650 TF/04 Improved ratelimit ACL condition.
652 The /noupdate option has been deprecated in favour of /readonly which
653 has clearer semantics. The /leaky, /strict, and /readonly update modes
654 are mutually exclusive. The update mode is no longer included in the
655 database key; it just determines when the database is updated. (This
656 means that when you upgrde Exim will forget old rate measurements.)
658 Exim now checks that the per_* options are used with an update mode that
659 makes sense for the current ACL. For example, when Exim is processing a
660 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
661 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
662 must specify per_mail/readonly. If you omit the update mode it defaults to
663 /leaky where that makes sense (as before) or /readonly where required.
665 The /noupdate option is now undocumented but still supported for
666 backwards compatibility. It is equivalent to /readonly except that in
667 ACLs where /readonly is required you may specify /leaky/noupdate or
668 /strict/noupdate which are treated the same as /readonly.
670 A useful new feature is the /count= option. This is a generalization
671 of the per_byte option, so that you can measure the throughput of other
672 aggregate values. For example, the per_byte option is now equivalent
673 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
675 The per_rcpt option has been generalized using the /count= mechanism
676 (though it's more complicated than the per_byte equivalence). When it is
677 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
678 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
679 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
680 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
681 Note that using per_rcpt with a non-readonly update mode in more than
682 one ACL will cause the recipients to be double-counted. (The per_mail
683 and per_byte options don't have this problem.)
685 The handling of very low rates has changed slightly. If the computed rate
686 is less than the event's count (usually one) then this event is the first
687 after a long gap. In this case the rate is set to the same as this event's
688 count, so that the first message of a spam run is counted properly.
690 The major new feature is a mechanism for counting the rate of unique
691 events. The new per_addr option counts the number of different
692 recipients that someone has sent messages to in the last time period. It
693 behaves like per_rcpt if all the recipient addresses are different, but
694 duplicate recipient addresses do not increase the measured rate. Like
695 the /count= option this is a general mechanism, so the per_addr option
696 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
697 example, measure the rate that a client uses different sender addresses
698 with the options per_mail/unique=$sender_address. There are further
699 details in the main documentation.
701 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
703 TF/06 Removed a few PCRE remnants.
705 TF/07 Automatically extract Exim's version number from tags in the git
706 repository when doing development or release builds.
708 PP/02 Raise smtp_cmd_buffer_size to 16kB.
709 Bugzilla 879. Patch from Paul Fisher.
711 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
712 Heavily based on revision 40f9a89a from Simon Arlott's tree.
715 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
717 PP/05 Variable $av_failed, true if the AV scanner deferred.
718 Bugzilla 1078. Patch from John Horne.
720 PP/06 Stop make process more reliably on build failure.
721 Bugzilla 1087. Patch from Heiko Schlittermann.
723 PP/07 Make maildir_use_size_file an _expandable_ boolean.
724 Bugzilla 1089. Patch from Heiko Schlittermann.
726 PP/08 Handle ${run} returning more data than OS pipe buffer size.
727 Bugzilla 1131. Patch from Holger Weiß.
729 PP/09 Handle IPv6 addresses with SPF.
730 Bugzilla 860. Patch from Wolfgang Breyha.
732 PP/10 GnuTLS: support TLS 1.2 & 1.1.
734 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
737 PP/11 match_* no longer expand right-hand-side by default.
738 New compile-time build option, EXPAND_LISTMATCH_RHS.
739 New expansion conditions, "inlist", "inlisti".
741 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
743 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
745 PP/14 fix log_write() format string regression from TF/03.
746 Bugzilla 1152. Patch from Dmitry Isaikin.
752 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
754 PP/02 Harmonised TLS library version reporting; only show if debugging.
755 Layout now matches that introduced for other libraries in 4.74 PP/03.
757 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
759 PP/04 New "dns_use_edns0" global option.
761 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
764 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
765 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
767 TK/01 Updated PolarSSL code to 0.14.2.
768 Bugzilla 1097. Patch from Andreas Metzler.
770 PP/07 Catch divide-by-zero in ${eval:...}.
773 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
776 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
777 format-string attack -- SECURITY: remote arbitrary code execution.
779 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
780 time unintentionally subject to list matching rules, letting the header
781 cause arbitrary Exim lookups (of items which can occur in lists, *not*
782 arbitrary string expansion). This allowed for information disclosure.
784 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
785 INT_MIN/-1 -- value coerced to INT_MAX.
791 NM/01 Workround for PCRE version dependency in version reporting
794 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
795 This fixes portability to compilers other than gcc, notably
796 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
798 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
799 makefiles for portability to HP-UX and POSIX correctness.
801 PP/01 Permit LOOKUP_foo enabling on the make command-line.
802 Also via indented variable definition in the Makefile.
803 (Debugging by Oliver Heesakkers).
805 PP/02 Restore caching of spamd results with expanded spamd_address.
806 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
808 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
809 Improves build reliability. Fix from: Frank Elsner
811 NM/02 Fix wide character breakage in the rfc2047 coding
812 Fixes bug 1064. Patch from Andrey N. Oktyabrski
814 NM/03 Allow underscore in dnslist lookups
815 Fixes bug 1026. Patch from Graeme Fowler
817 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
818 Code patches from Adam Ciarcinski of NetBSD.
820 NM/04 Fixed exiqgrep to cope with mailq missing size issue
823 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
824 is logged, to avoid truncation. Patch from John Horne.
826 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
827 Patch from Jakob Hirsch.
829 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
830 SQL string expansion failure details.
831 Patch from Andrey Oktyabrski.
833 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
834 Patch from Simon Arlott.
836 PP/09 New lookups functionality failed to compile on old gcc which rejects
837 extern declarations in function scope.
838 Patch from Oliver Fleischmann
840 PP/10 Use sig_atomic_t for flags set from signal handlers.
841 Check getgroups() return and improve debugging.
842 Fixed developed for diagnosis in bug 927 (which turned out to be
845 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
846 Patch from Mark Zealey.
848 PP/12 Bugzilla 1056: Improved spamd server selection.
849 Patch from Mark Zealey.
851 PP/13 Bugzilla 1086: Deal with maildir quota file races.
852 Based on patch from Heiko Schlittermann.
854 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
855 Patch from Uwe Doering, sign-off by Michael Haardt.
857 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
858 variable declaration deep within a block. Bug and patch from
861 PP/15 lookups-Makefile IRIX compatibilty coercion.
863 PP/16 Make DISABLE_DKIM build knob functional.
865 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
866 Patch by Simon Arlott
868 TF/03 Fix valgrind.h portability to C89 compilers that do not support
869 variable argument macros. Our copy now differs from upstream.
875 TF/01 Failure to get a lock on a hints database can have serious
876 consequences so log it to the panic log.
878 TF/02 Log LMTP confirmation messages in the same way as SMTP,
879 controlled using the smtp_confirmation log selector.
881 TF/03 Include the error message when we fail to unlink a spool file.
883 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
884 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
885 for maintaining out-of-tree patches for some time.
887 PP/01 Bugzilla 139: Documentation and portability issues.
888 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
889 Handle per-OS dynamic-module compilation flags.
891 PP/02 Let /dev/null have normal permissions.
892 The 4.73 fixes were a little too stringent and complained about the
893 permissions on /dev/null. Exempt it from some checks.
894 Reported by Andreas M. Kirchwitz.
896 PP/03 Report version information for many libraries, including
897 Exim version information for dynamically loaded libraries. Created
898 version.h, now support a version extension string for distributors
899 who patch heavily. Dynamic module ABI change.
901 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
902 privilege escalation vulnerability whereby the Exim run-time user
903 can cause root to append content of the attacker's choosing to
906 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
909 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
910 If dropping privileges for untrusted macros, we disabled normal logging
911 on the basis that it would fail; for the Exim run-time user, this is not
912 the case, and it resulted in successful deliveries going unlogged.
913 Fixed. Reported by Andreas Metzler.
919 PP/01 Date: & Message-Id: revert to normally being appended to a message,
920 only prepend for the Resent-* case. Fixes regression introduced in
921 Exim 4.70 by NM/22 for Bugzilla 607.
923 PP/02 Include check_rfc2047_length in configure.default because we're seeing
924 increasing numbers of administrators be bitten by this.
926 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
928 PP/03 Bugzilla 994: added openssl_options main configuration option.
930 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
932 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
934 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
936 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
937 without a peer certificate, leading to a segfault because of an
938 assumption that peers always have certificates. Be a little more
939 paranoid. Problem reported by Martin Tscholak.
941 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
942 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
943 NB: ClamAV planning to remove STREAM in "middle of 2010".
944 CL also introduces -bmalware, various -d+acl logging additions and
945 more caution in buffer sizes.
947 PP/09 Implemented reverse_ip expansion operator.
949 PP/10 Bugzilla 937: provide a "debug" ACL control.
951 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
953 PP/12 Bugzilla 973: Implement --version.
955 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
957 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
959 PP/15 Bugzilla 816: support multiple condition rules on Routers.
961 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
962 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
963 ignore trailing whitespace.
965 JJ/02 prevent non-panic DKIM error from being sent to paniclog
967 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
970 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
971 Notification from Dr Andrew Aitchison.
973 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
974 ExtendedDetectionInfo response format.
975 Notification from John Horne.
977 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
980 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
981 XSL and documented dependency on system catalogs, with examples of how
984 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
987 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
988 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
989 configuration file which is writeable by the Exim user or group.
991 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
992 of configuration files to cover files specified with the -C option if
993 they are going to be used with root privileges, not just the default
996 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
997 option (effectively making it always true).
999 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1000 files to be used while preserving root privileges.
1002 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1003 that rogue child processes cannot use them.
1005 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1006 run-time user, instead of root.
1008 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1009 Exim run-time user without dropping privileges.
1011 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1012 result string, instead of calling string_vformat() twice with the same
1015 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1016 for other users. Others should always drop root privileges if they use
1017 -C on the command line, even for a whitelisted configure file.
1019 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1021 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1027 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1028 $data_path, and $header_path variables; fixed documentation bugs and
1031 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1032 exipick to access non-standard spools, including the "frozen" queue
1035 NM/01 Bugzilla 965: Support mysql stored procedures.
1036 Patch from Alain Williams
1038 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1040 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1041 Patch from Andreas Metzler
1043 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1044 Patch from Kirill Miazine
1046 NM/05 Bugzilla 671: Added umask to procmail example.
1048 JJ/03 installed exipick 20100323.0, fixing doc bug
1050 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1051 directory. Notification and patch from Dan Rosenberg.
1053 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1055 TK/02 Improve log output when DKIM signing operation fails.
1057 MH/01 Treat the transport option dkim_domain as a colon separated
1058 list, not as a single string, and sign the message with each element,
1059 omitting multiple occurences of the same signer.
1061 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1062 Bugzilla 985, 986. Patch by Simon Arlott
1064 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1065 Patch by Simon Arlott
1067 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1068 MBX locking. Notification from Dan Rosenberg.
1074 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1076 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1078 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1080 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1082 NM/04 Bugzilla 915: Flag broken perl installation during build.
1088 TK/01 Added patch by Johannes Berg that expands the main option
1089 "spamd_address" if it starts with a dollar sign.
1091 TK/02 Write list of recipients to X-Envelope-Sender header when building
1092 the mbox-format spool file for content scanning (suggested by Jakob
1095 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1096 (http://www.dcc-servers.net/) support via dccifd. Activated by
1097 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1099 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1100 by Mark Daniel Reidel <mr@df.eu>.
1102 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1103 When building exim an external PCRE library is now needed -
1104 PCRE is a system library on the majority of modern systems.
1105 See entry on PCRE_LIBS in EDITME file.
1107 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1108 conversation. Added nologin parameter to request.
1109 Patch contributed by Kirill Miazine.
1111 TF/01 Do not log submission mode rewrites if they do not change the address.
1113 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1115 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1116 log files in place. Contributed by Roberto Lima.
1118 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1120 TF/03 Bugzilla 615: When checking the local_parts router precondition
1121 after a local_part_suffix or local_part_prefix option, Exim now
1122 does not use the address's named list lookup cache, since this
1123 contains cached lookups for the whole local part.
1125 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1126 Robert Millan. Documentation is in experimental-spec.txt.
1128 TF/04 Bugzilla 668: Fix parallel build (make -j).
1130 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1132 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1133 Patch provided by Jan Srzednicki.
1135 TF/05 Leading white space used to be stripped from $spam_report which
1136 wrecked the formatting. Now it is preserved.
1138 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1139 that they are available at delivery time.
1141 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1143 TF/08 TLS error reporting now respects the incoming_interface and
1144 incoming_port log selectors.
1146 TF/09 Produce a more useful error message if an SMTP transport's hosts
1147 setting expands to an empty string.
1149 NM/06 Bugzilla 744: EXPN did not work under TLS.
1150 Patch provided by Phil Pennock.
1152 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1153 Patch provided by Richard Godbee.
1155 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1156 acl_smtp_notquit, added index entry.
1158 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1159 Patch provided by Eugene Bujak.
1161 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1162 accept(). Patch provided by Maxim Dounin.
1164 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1165 Patch provided by Phil Pennock.
1167 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1169 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1170 Patch provided by Brad "anomie" Jorsch.
1172 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1173 Patch provided by Dean Brooks.
1175 TK/05 Add native DKIM support (does not depend on external libraries).
1177 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1178 Patch provided by Graeme Fowler.
1180 NM/16 Bugzilla 851: Documentation example syntax fix.
1182 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1184 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1187 NM/19 Bugzilla 745: TLS version reporting.
1188 Patch provided by Phil Pennock.
1190 NM/20 Bugzilla 167: bool: condition support.
1191 Patch provided by Phil Pennock.
1193 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1194 clients. Patch provided by Phil Pennock.
1196 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1197 Patch provided by Brad "anomie" Jorsch.
1199 NM/23 Bugzilla 687: Fix misparses in eximstats.
1200 Patch provided by Heiko Schlittermann.
1202 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1203 Patch provided by Heiko Schlittermann.
1205 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1206 plus update to original patch.
1208 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1210 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1211 Patch provided by David Brownlee.
1213 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1215 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1217 NM/30 Bugzilla 888: TLS documentation bugfixes.
1219 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1221 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1222 Unlike the original bugzilla I have changed all shell scripts in src tree.
1224 NM/33 Bugzilla 898: Transport filter timeout fix.
1225 Patch by Todd Rinaldo.
1227 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1228 Patch by Serge Demonchaux.
1230 NM/35 Bugzilla 39: Base64 decode bug fixes.
1231 Patch by Jakob Hirsch.
1233 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1235 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1237 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1239 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1245 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1246 ALT-N's libdkim that I have put here:
1247 http://duncanthrax.net/exim-experimental/
1249 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1250 were called 'true' and it seems that C99 defines that as a reserved
1251 keyword to be used with 'bool' variable types. That means you could
1252 not include C99-style headers which use bools without triggering
1253 build errors in sieve.c.
1255 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1256 as mailq or other aliases. Changed the --help handling significantly
1257 to do whats expected. exim_usage() emits usage/help information.
1259 SC/01 Added the -bylocaldomain option to eximstats.
1261 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1263 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1265 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1271 PH/01 Another patch from the Sieve maintainer.
1273 PH/02 When an IPv6 address is converted to a string for single-key lookup
1274 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1275 dots are used instead of colons so that keys in lsearch files need not
1276 contain colons. This was done some time before quoting was made available
1277 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1278 (notated using the quote facility) so as to distinguish them from IPv4
1279 keys. This meant that lookups for IP addresses in host lists did not work
1280 for iplsearch lookups.
1282 This has been fixed by arranging for IPv6 addresses to be expressed with
1283 colons if the lookup type is iplsearch. This is not incompatible, because
1284 previously such lookups could never work.
1286 The situation is now rather anomolous, since one *can* have colons in
1287 ordinary lsearch keys. However, making the change in all cases is
1288 incompatible and would probably break a number of configurations.
1290 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1293 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1294 conversion specification without a maximum field width, thereby enabling
1295 a rogue spamd server to cause a buffer overflow. While nobody in their
1296 right mind would setup Exim to query an untrusted spamd server, an
1297 attacker that gains access to a server running spamd could potentially
1298 exploit this vulnerability to run arbitrary code as the Exim user.
1300 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1301 $primary_hostname instead of what libspf2 thinks the hosts name is.
1303 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1304 a directory entry by the name of the lookup key. Previously, if a
1305 symlink pointed to a non-existing file or a file in a directory that
1306 Exim lacked permissions to read, a lookup for a key matching that
1307 symlink would fail. Now it is enough that a matching directory entry
1308 exists, symlink or not. (Bugzilla 503.)
1310 PH/03 The body_linecount and body_zerocount variables are now exported in the
1313 PH/04 Added the $dnslist_matched variable.
1315 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1316 This means they are set thereafter only if the connection becomes
1319 PH/06 Added the client_condition to authenticators so that some can be skipped
1320 by clients under certain conditions.
1322 PH/07 The error message for a badly-placed control=no_multiline_responses left
1323 "_responses" off the end of the name.
1325 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1327 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1328 (without spaces) instead of just copying the configuration text.
1330 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1332 PH/11 Added $max_received_linelength.
1334 PH/12 Added +ignore_defer and +include_defer to host lists.
1336 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1337 way in which PCRE > 7.0 is built.
1339 PH/14 Implemented queue_only_load_latch.
1341 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1342 MAIL command. The effect was to mangle the value on 64-bit systems.
1344 PH/16 Another patch from the Sieve maintainer.
1346 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1348 PH/18 If a system quota error occurred while trying to create the file for
1349 a maildir delivery, the message "Mailbox is full" was not appended to the
1350 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1351 only to a quota excession during the actual writing of the file.
1353 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1354 characters?) which causes problems in log lines. The DN values are now
1355 passed through string_printing() before being added to log lines.
1357 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1358 and InterBase are left for another time.)
1360 PH/21 Added message_body_newlines option.
1362 PH/22 Guard against possible overflow in moan_check_errorcopy().
1364 PH/23 POSIX allows open() to be a macro; guard against that.
1366 PH/24 If the recipient of an error message contained an @ in the local part
1367 (suitably quoted, of course), incorrect values were put in $domain and
1368 $local_part during the evaluation of errors_copy.
1374 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1375 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1378 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1379 issue a MAIL command.
1381 PH/02 In an ACL statement such as
1383 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1385 if a client was not listed at all, or was listed with a value other than
1386 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1387 the condition was not true (as it should be), so access was not denied.
1388 The bug was that the ! inversion was incorrectly passed on to the second
1389 item. This has been fixed.
1391 PH/03 Added additional dnslists conditions == and =& which are different from
1392 = and & when the dns lookup returns more than one IP address.
1394 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1395 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1397 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1398 FSYNC, which compiles an option called disable_fsync that allows for
1399 bypassing fsync(). The documentation is heavily laced with warnings.
1401 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1403 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1404 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1405 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1406 including adding "make clean"; (3) Added -fPIC when compiling the test
1407 dynamically loaded module, to get rid of a warning.
1409 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1410 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1411 = 0s. The bug is otherwise harmless.
1413 PH/07 There was a bug in the dovecot authenticator such that the value of
1414 $auth1 could be overwritten, and so not correctly preserved, after a
1415 successful authentication. This usually meant that the value preserved by
1416 the server_setid option was incorrect.
1418 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1420 PH/09 Installed PCRE release 7.0.
1422 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1423 run for batched SMTP input. It is now run at the start of every message
1424 in the batch. While fixing this I discovered that the process information
1425 (output by running exiwhat) was not always getting set for -bs and -bS
1426 input. This is fixed, and it now also says "batched" for BSMTP.
1428 PH/11 Added control=no_pipelining.
1430 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1431 patch, slightly modified), and move the expansion of helo_data till after
1432 the connection is made in the smtp transport (so it can use these
1435 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1437 PH/14 Added log_selector = +pid.
1439 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1441 PH/16 Add ${if forany and ${if forall.
1443 PH/17 Added dsn_from option to vary the From: line in DSNs.
1445 PH/18 Flush SMTP output before performing a callout, unless control =
1446 no_callout_flush is set.
1448 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1449 was true (the default) a successful delivery failed to delete the retry
1450 item, thus causing premature timeout of the address. The bug is now
1453 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1455 PH/21 Long custom messages for fakedefer and fakereject are now split up
1456 into multiline reponses in the same way that messages for "deny" and
1457 other ACL rejections are.
1459 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1460 with slight modification.
1462 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1463 draft, changing the syntax and factoring some duplicate code".
1465 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1466 for deliveries of the second and subsequent messages over the same SMTP
1469 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1470 ${reduce, with only minor "tidies".
1472 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1474 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1475 expansion side effects.
1477 PH/27 When a message times out after an over-quota error from an Exim-imposed
1478 quota, the bounce message says "mailbox is full". This message was not
1479 being given when it was a system quota that was exceeded. It now should
1482 MH/03 Made $recipients available in local_scan(). local_scan() already has
1483 better access to the recipient list through recipients_list[], but
1484 $recipients can be useful in postmaster-provided expansion strings.
1486 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1487 in the case of a MAIL command with additional options following the
1488 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1489 were accidentally chopped off.
1491 PH/29 SMTP synchronization checks are implemented when a command is read -
1492 there is a check that no more input is waiting when there shouldn't be
1493 any. However, for some commands, a delay in an ACL can mean that it is
1494 some time before the response is written. In this time, more input might
1495 arrive, invalidly. So now there are extra checks after an ACL has run for
1496 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1497 pipelining has not been advertised.
1499 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1501 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1502 returned in the SMTP response when smtp_return_error_details was false.
1503 This has been fixed.
1505 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1506 instead of the C I/O that was originally supplied, because problems were
1507 reported on Solaris.
1509 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1510 Exim which did not show up earlier: it was assuming that a call to
1511 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1512 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1513 was a macro that became an assignment, so it seemed to work. This has
1514 changed to a proper function call with a void return, hence the compile
1515 error. Exim's code has been fixed.
1517 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1520 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1522 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1524 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1525 because it turns out to be too big or there is a timeout) there is no
1526 "Completed" line in the log. When some messages of this type were
1527 selected by exigrep, they were listed as "not completed". Others were
1528 picked up by some special patterns. I have improved the selection
1529 criteria to be more general.
1531 PH/38 The host_find_failed option in the manualroute router can now be set
1532 to "ignore", to completely ignore a host whose IP address cannot be
1533 found. If all hosts are ignored, the behaviour is controlled by the new
1534 host_all_ignored option.
1536 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1537 homing or because of multiple MX records with /mx) generated more than
1538 one IP address, and the following item turned out to be the local host,
1539 all the secondary addresses of the first item were incorrectly removed
1540 from the list, along with the local host and any following hosts (which
1541 is what is supposed to happen).
1543 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1544 whoever called Exim into the -H file. In the case of the daemon it was
1545 behaving confusingly. When first started, it used values for whoever
1546 started the daemon, but after a SIGHUP it used the Exim user (because it
1547 calls itself on a restart). I have changed the code so that it now always
1550 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1551 message are rejected with the same error (e.g. no authentication or bad
1552 sender address), and a DATA command is nevertheless sent (as can happen
1553 with PIPELINING or a stupid MUA), the error message that was given to the
1554 RCPT commands is included in the rejection of the DATA command. This is
1555 intended to be helpful for MUAs that show only the final error to their
1558 PH/42 Another patch from the Sieve maintainer.
1560 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1561 Eximstats - Fixed some broken HTML links and added missing column headers
1563 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1564 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1566 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1568 PH/43 Yet another patch from the Sieve maintainer.
1570 PH/44 I found a way to check for a TCP/IP connection going away before sending
1571 the response to the final '.' that terminates a message, but only in the
1572 case where the client has not sent further data following the '.'
1573 (unfortunately, this is allowed). However, in many cases there won't be
1574 any further data because there won't be any more messages to send. A call
1575 to select() can be used: if it shows that the input is "ready", there is
1576 either input waiting, or the socket has been closed. An attempt to read
1577 the next input character can distinguish the two cases. Previously, Exim
1578 would have sent an OK response which the client would never have see.
1579 This could lead to message repetition. This fix should cure that, at
1580 least in a lot of common cases.
1582 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1583 advertised in response to EHLO.
1589 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1590 fixed by 4.65/MH/01 (is this a record?) are fixed:
1592 (i) An empty string was always treated as zero by the numeric comparison
1593 operators. This behaviour has been restored.
1595 (ii) It is documented that the numeric comparison operators always treat
1596 their arguments as decimal numbers. This was broken in that numbers
1597 starting with 0 were being interpreted as octal.
1599 While fixing these problems I realized that there was another issue that
1600 hadn't been noticed. Values of message_size_limit (both the global option
1601 and the transport option) were treated as octal if they started with 0.
1602 The documentation was vague. These values are now always treated as
1603 decimal, and I will make that clear in the documentation.
1609 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1610 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1613 MH/01 Don't check that the operands of numeric comparison operators are
1614 integers when their expansion is in "skipping" mode (fixes bug
1615 introduced by 4.64-PH/07).
1617 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1618 child addresses, Exim now panics and dies. Previously, because the count
1619 is held in a short int, deliveries were likely to be lost. As such a
1620 large number of recipients for a single message is ridiculous
1621 (performance will be very, very poor), I have chosen to impose a limit
1622 rather than extend the field.
1628 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1629 leftover -K file (the existence of which was triggered by #402).
1630 While we were at it, introduced process PID as part of the -K
1631 filename. This should rule out race conditions when creating
1634 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1635 processing considerably. Previous code took too long for large mails,
1636 triggering a timeout which in turn triggers #401.
1638 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1639 in the DK code in transports.c. sendfile() is not really portable,
1640 hence the _LINUX specificness.
1642 TF/01 In the add_headers option to the mail command in an Exim filter,
1643 there was a bug that Exim would claim a syntax error in any
1644 header after the first one which had an odd number of characters
1647 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1648 callout verification, Exim cached a "reject" for the entire domain. This
1649 is correct for most verifications, but it is not correct for a recipient
1650 verification with use_sender or use_postmaster set, because in that case
1651 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1652 case of MAIL FROM:<> rejection from other early rejections (e.g.
1653 rejection of HELO). When verifying a recipient using a non-null MAIL
1654 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1655 Whatever the result of the callout, the value of the domain cache is
1656 left unchanged (for any other kind of callout, getting as far as trying
1657 RCPT means that the domain itself is ok).
1659 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1662 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1663 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1664 session, the daemon ignores ECONNECT errors and logs others; it now
1665 ignores EPIPE as well.
1667 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1668 (quoted-printable decoding).
1670 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1671 later the small subsequent patch to fix an introduced bug.
1673 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1675 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1677 PH/08 An error is now given if message_size_limit is specified negative.
1679 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1680 to be given (somewhat) arbitrary names.
1682 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1685 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1686 miscellaneous code fixes
1688 PH/10 Added the log_reject_target ACL modifier to specify where to log
1691 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1692 hostname. This is wrong, because it relates to the incoming message (and
1693 probably the interface on which it is arriving) and not to the outgoing
1694 callout (which could be using a different interface). This has been
1695 changed to use the value of the helo_data option from the smtp transport
1696 instead - this is what is used when a message is actually being sent. If
1697 there is no remote transport (possible with a router that sets up host
1698 addresses), $smtp_active_hostname is used.
1700 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1701 tweaks were necessary in order to get it to work (see also 21 below):
1702 (a) The code assumed that strncpy() returns a negative number on buffer
1703 overflow, which isn't the case. Replaced with Exim's string_format()
1705 (b) There were several signed/unsigned issues. I just did the minimum
1706 hacking in of casts. There is scope for a larger refactoring.
1707 (c) The code used strcasecmp() which is not a standard C function.
1708 Replaced with Exim's strcmpic() function.
1709 (d) The code set only $1; it now sets $auth1 as well.
1710 (e) A simple test gave the error "authentication client didn't specify
1711 service in request". It would seem that Dovecot has changed its
1712 interface. Fortunately there's a specification; I followed it and
1713 changed what the client sends and it appears to be working now.
1715 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1718 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1719 address A is aliased to B and C, where B exists and C does not. Without
1720 -v the output is "A verified" because verification stops after a
1721 successful redirection if more than one address is generated. However,
1722 with -v the child addresses are also verified. Exim was outputting "A
1723 failed to verify" and then showing the successful verification for C,
1724 with its parentage. It now outputs "B failed to verify", showing B's
1725 parentage before showing the successful verification of C.
1727 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1728 look up a TXT record in a specific list after matching in a combined
1731 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1732 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1733 they consult the DNS. I had assumed they would set it the way they
1734 wanted; and indeed my experiments on Linux seem to show that in some
1735 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1736 To be on the safe side, however, I have now made the interface to
1737 host_find_byname() similar to host_find_bydns(), with an argument
1738 containing the DNS resolver options. The host_find_byname() function now
1739 sets these options at its start, just as host_find_bydns() does. The smtp
1740 transport options dns_qualify_single and dns_search_parents are passed to
1741 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1742 of host_find_byname() use the default settings of RES_DEFNAMES
1743 (qualify_single) but not RES_DNSRCH (search_parents).
1745 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1746 spool_read_header() do less string testing, by means of a preliminary
1747 switch on the second character of optional "-foo" lines. (This is
1748 overdue, caused by the large number of possibilities that now exist.
1749 Originally there were few.) While I was there, I also converted the
1750 str(n)cmp tests so they don't re-test the leading "-" and the first
1751 character, in the hope this might squeeze out yet more improvement.
1753 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1754 flag allowing group syntax was set by the header_syntax check but not
1755 turned off, possible causing trouble later; (2) The flag was not being
1756 set at all for the header_verify test, causing "group"-style headers to
1757 be rejected. I have now set it in this case, and also caused header_
1758 verify to ignore an empty address taken from a group. While doing this, I
1759 came across some other cases where the code for allowing group syntax
1760 while scanning a header line wasn't quite right (mostly, not resetting
1761 the flag correctly in the right place). These bugs could have caused
1762 trouble for malformed header lines. I hope it is now all correct.
1764 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1765 with the "reply" argument non-NULL. The code, however (which originally
1766 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1767 but it didn't always do it. This confused somebody who was copying the
1768 code for some other use. I have removed all the tests.
1770 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1771 feature that was used to support insecure browsers during the U.S. crypto
1772 embargo. It requires special client support, and Exim is probably the
1773 only MTA that supported it -- and would never use it because real RSA is
1774 always available. This code has been removed, because it had the bad
1775 effect of slowing Exim down by computing (never used) parameters for the
1776 RSA_EXPORT functionality.
1778 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1779 authenticator to fail if there's a tab character in the incoming data
1780 (there should never be unless someone is messing about, as it's supposed
1781 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1782 if the connection is using TLS or if the remote IP is the same as the
1783 local IP, and the "valid-client-cert option" if a client certificate has
1786 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1787 authenticators. This can be used for authorization after authentication
1788 succeeds. (In the case of plaintext, it servers for both authentication
1791 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1792 if any retry times were supplied.
1794 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1795 connection, where there is no client IP address to check. In this
1796 situation, the verify now always succeeds.
1798 PH/25 Applied John Jetmore's -Mset patch.
1800 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1802 PH/27 In a string expansion for a processed (not raw) header when multiple
1803 headers of the same name were present, leading whitespace was being
1804 removed from all of them, but trailing whitespace was being removed only
1805 from the last one. Now trailing whitespace is removed from each header
1806 before concatenation. Completely empty headers in a concatenation (as
1807 before) are ignored.
1809 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1810 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1812 PH/29 [Removed. This was a change that I later backed out, and forgot to
1813 correct the ChangeLog entry (that I had efficiently created) before
1814 committing the later change.]
1816 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1817 address errors (4xx response to RCPT) over the same connection as other
1818 messages routed to the same hosts. Such deliveries are always "forced",
1819 so retry times are not inspected. This resulted in far too many retries
1820 for the affected addresses. The effect occurred only when there were more
1821 hosts than the hosts_max_try setting in the smtp transport when it had
1822 the 4xx errors. Those hosts that it had tried were not added to the list
1823 of hosts for which the message was waiting, so if all were tried, there
1824 was no problem. Two fixes have been applied:
1826 (i) If there are any address or message errors in an SMTP delivery, none
1827 of the hosts (tried or untried) are now added to the list of hosts
1828 for which the message is waiting, so the message should not be a
1829 candidate for sending over the same connection that was used for a
1830 successful delivery of some other message. This seems entirely
1831 reasonable: after all the message is NOT "waiting for some host".
1832 This is so "obvious" that I'm not sure why it wasn't done
1833 previously. Hope I haven't missed anything, but it can't do any
1834 harm, as the worst effect is to miss an optimization.
1836 (ii) If, despite (i), such a delivery is accidentally attempted, the
1837 routing retry time is respected, so at least it doesn't keep
1838 hammering the server.
1840 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1841 in ${readsocket because some servers need this prod.
1843 PH/32 Added some extra debug output when updating a wait-xxx database.
1845 PH/33 The hint "could be header name not terminated by colon", which has been
1846 given for certain expansion errors for a long time, was not being given
1847 for the ${if def:h_colon_omitted{... case.
1849 PH/34 The spec says: "With one important exception, whenever a domain list is
1850 being scanned, $domain contains the subject domain." There was at least
1851 one case where this was not true.
1853 PH/35 The error "getsockname() failed: connection reset by peer" was being
1854 written to the panic log as well as the main log, but it isn't really
1855 panic-worthy as it just means the connection died rather early on. I have
1856 removed the panic log writing for the ECONNRESET error when getsockname()
1859 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1860 runs only) independently of the message's sender address. This meant
1861 that, if the 4xx error was in fact related to the sender, a different
1862 message to the same recipient with a different sender could confuse
1863 things. In particualar, this can happen when sending to a greylisting
1864 server, but other circumstances could also provoke similar problems.
1865 I have changed the default so that the retry time for these errors is now
1866 based a combination of the sender and recipient addresses. This change
1867 can be overridden by setting address_retry_include_sender=false in the
1870 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1871 remote server are returned as part of bounce messages. This was not
1872 happening for LMTP over a pipe (the lmtp transport), but now it is the
1873 same for both kinds of LMTP.
1875 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1876 in header lines that were in fact CNAMEs. This is no longer the case.
1878 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1879 and queue runs started by the daemon processed all messages. This has
1880 been fixed so that -R and -S can now usefully be given with -q<time>.
1882 PH/40 Import PCRE release 6.7 (fixes some bugs).
1884 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1886 PH/42 Give an error if -q is specified more than once.
1888 PH/43 Renamed the variables $interface_address and $interface_port as
1889 $received_ip_address and $received_port, to make it clear that these
1890 values apply to message reception, and not to the outgoing interface when
1891 a message is delivered. (The old names remain recognized, of course.)
1893 PH/44 There was no timeout on the connect() call when using a Unix domain
1894 socket in the ${readsocket expansion. There now is.
1896 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1897 be meaningful with "accept".
1899 SC/01 Eximstats V1.43
1900 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1902 SC/02 Eximstats V1.44
1903 Use a glob alias rather than an array ref in the generated
1904 parser. This improves both readability and performance.
1906 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1907 Collect SpamAssassin and rejection statistics.
1908 Don't display local sender or destination tables unless
1909 there is data to show.
1910 Added average volumes into the top table text output.
1912 SC/04 Eximstats V1.46
1913 Collect data on the number of addresses (recipients)
1914 as well as the number of messages.
1916 SC/05 Eximstats V1.47
1917 Added 'Message too big' to the list of mail rejection
1918 reasons (thanks to Marco Gaiarin).
1920 SC/06 Eximstats V1.48
1921 Mainlog lines which have GMT offsets and are too short to
1922 have a flag are now skipped.
1924 SC/07 Eximstats V1.49 (Alain Williams)
1925 Added the -emptyok flag.
1927 SC/08 Eximstats V1.50
1928 Fixes for obtaining the IP address from reject messages.
1930 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1931 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1932 whitesspace changes from 4.64-PH/27
1934 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1937 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1938 are found, allow negative numbers in numeric criteria)
1940 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1942 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1943 to match changes made in 4.64-PH/43
1945 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1947 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1948 "domain" is actually an IP address, and force a failure. This locks out
1949 those revolvers/nameservers that support "A-for-A" lookups, in
1950 contravention of the specifications.
1952 PH/48 When a host name was looked up from an IP address, and the subsequent
1953 forward lookup of the name timed out, the host name was left in
1954 $sender_host_name, contrary to the specification.
1956 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1957 restricted to single-key lookups, Exim was not diagnosing an error if
1958 * or *@ was used with a query-style lookup.
1960 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1962 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1963 long ago, but noone interested enough thought of it. Let's just say that
1964 the "1.1" means that there are some new functions that weren't there at
1965 some point in the past.
1967 PH/51 Error processing for expansion failure of helo_data from an smtp
1968 transport during callout processing was broken.
1970 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1971 tested/used via the -bh/-bhc/-bs options.
1973 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1974 bug, fixed in subsequent PCRE releases).
1976 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1977 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1979 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1985 SC/01 Use a glob alias rather than an array ref in eximstats generated
1986 parser. This improves both readability and performance.
1988 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1989 Don't display local sender or destination tables in eximstats unless
1990 there is data to show.
1991 Added average volumes into the eximstats top table text output.
1993 SC/03 Collect data on the number of addresses (recipients) as well
1994 as the number of messages in eximstats.
1996 TF/01 Correct an error in the documentation for the redirect router. Exim
1997 does (usually) call initgroups() when daemonizing.
1999 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2000 with consistent privilege compared to when running as a daemon.
2002 TF/03 Note in the spec that $authenticated_id is not set for local
2003 submissions from trusted users.
2005 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2006 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2008 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2009 by adding some example configuration directives to the default
2010 configuration file. A little bit of work is required to uncomment the
2011 directives and define how usernames and passwords are checked, but
2012 there is now a framework to start from.
2014 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2015 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2016 without this. I don't know how relevant this is to other LDAP libraries.
2018 PH/02 Add the verb name to the "unknown ACL verb" error.
2020 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2022 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2024 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2025 directory not expanded when it should be if an expanded home directory
2026 was set for the address (which is overridden by the transport).
2028 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2031 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2032 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2033 because it is too late at that time, and has no effect.
2035 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2036 security issue with \' (bugzilla #107). I could not use the
2037 PQescapeStringConn() function, because it needs a PGconn value as one of
2040 PH/08 When testing addresses using -bt, indicate those final addresses that
2041 are duplicates that would not cause an additional delivery. At least one
2042 person was confused, thinking that -bt output corresponded to deliveries.
2043 (Suppressing duplicates isn't a good idea as you lose the information
2044 about possibly different redirections that led to the duplicates.)
2046 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2047 systems where poll() doesn't work, in particular OS X.
2049 PH/10 Added more information to debugging output for retry time not reached.
2051 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2052 operations in malware.c.
2054 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2057 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2058 syslog_duplication set false, log lines that would normally be written
2059 both the the main log and to the reject log were not written to syslog at
2062 PH/14 In the default configuration, change the use of "message" in ACL warn
2063 statements to "add_header".
2065 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2066 not followed by a command (e.g. "seen endif").
2068 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2069 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2072 PH/17 Added extra conditions to the default value of delay_warning_condition
2076 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2077 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2078 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2081 The Auto-Submitted: and various List- headers are standardised, whereas I
2082 don't think Precedence: ever was.
2084 PH/18 Refactored debugging code in route_finduser() to show more information,
2085 in particular, the error code if getpwnam() issues one.
2087 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2088 This is apparently needed in addition to the PH/07 change above to avoid
2089 any possible encoding problems.
2091 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2092 but not after initializing Perl.
2094 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2095 output them only if debugging. By default they are written stderr,
2096 apparently, which is not desirable.
2098 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2101 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2104 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2106 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2107 authenticated or an ident call has been made. Suppress the default
2108 values for $authenticated_id and $authenticated_sender (but permit -oMai
2109 and -oMas) when testing with -bh.
2111 PH/24 Re-jigged the order of the tests in the default configuration so that the
2112 tests for valid domains and recipients precede the DNS black list and CSA
2113 tests, on the grounds that those ones are more expensive.
2115 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2116 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2117 This bug exists in every version of Exim that I still have, right back to
2120 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2121 However, an attempt to turn on case-sensitivity in a regex key by
2122 including (?-i) didn't work because the subject string was already
2123 lowercased, and the effects were non-intuitive. It turns out that a
2124 one-line patch can be used to allow (?-i) to work as expected.
2130 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2131 other effects) broke the use of negated acl sub-conditions.
2133 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2136 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2137 "Deny" causes Exim to reject the incoming connection with a 554 error.
2138 Unfortunately, if there is a major crisis, such as a disk failure,
2139 tcp-wrappers gives "deny", whereas what one would like would be some
2140 kind of temporary error. A kludge has been added to help with this.
2141 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2142 554 error is used if errno is still zero or contains ENOENT (which occurs
2143 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2146 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2148 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2149 errors. Otherwise a message that provokes a temporary error (when other
2150 messages do not) can cause a whole host to time out.
2152 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2153 addresses were routed directly to files or pipes from a redirect router.
2154 File deliveries just didn't batch; pipe deliveries might have suffered
2157 PH/06 A failure to get a lock for a hints database would erroneously always say
2158 "Failed to get write lock", even when it was really a read lock.
2160 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2161 of 0600. This has been changed to use the value of the lockfile_mode
2162 option (which defaults to 0600).
2164 PH/08 Applied small patch from the Sieve maintainer.
2166 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2167 folder from quota calculations, a direct delivery into this folder messed
2168 up the contents of the maildirsize file. This was because the regex was
2169 used only to exclude .Trash (or whatever) when the size of the mailbox
2170 was calculated. There was no check that a delivery was happening into an
2171 excluded directory. This bug has been fixed by ignoring all quota
2172 processing for deliveries into excluded directories.
2174 PH/10 Added the maildirfolder_create_regex option to appendfile.
2180 PH/01 The code for finding all the local interface addresses on a FreeBSD
2181 system running IPv6 was broken. This may well have applied to all BSD
2182 systems, as well as to others that have similar system calls. The broken
2183 code found IPv4 interfaces correctly, but gave incorrect values for the
2184 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2185 that it would not match correctly against @[] and not recognize the IPv6
2188 PH/02 The ipliteral router was not recognizing addresses of the form user@
2189 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2191 PH/03 Added disable_ipv6.
2193 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2194 decoded form, because it is most often used to construct To: headers
2195 lines in autoreplies, and the decoded form may well be syntactically
2196 invalid. However, $reply_address has leading white space removed, and all
2197 newlines turned into spaces so that the autoreply transport does not
2200 PH/05 If group was specified without a user on a router, and no group or user
2201 was specified on a transport, the group from the router was ignored.
2203 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2204 for visible compile-time settings that can be used to change these
2205 numbers, for those that want even more. Backwards compatibility with old
2206 spool files has been maintained. However, going back to a previous Exim
2207 release will lost any variables that are in spool files.
2209 PH/07 Two small changes when running in the test harness: increase delay when
2210 passing a TCP/IP connection to a new process, in case the original
2211 process has to generate a bounce, and remove special handling of
2212 127.0.0.2 (sic), which is no longer necessary.
2214 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2215 be the same on different OS.
2217 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2220 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2221 whether --show-vars was specified or not
2223 JJ/02 exipick: Added support for new ACL variable spool format introduced
2226 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2227 syntactically invalid From: or Reply-to: line, and a filter used this to
2228 generate an autoreply, and therefore failed to obtain an address for the
2229 autoreply, Exim could try to deliver to a non-existent relative file
2230 name, causing unrelated and misleading errors. What now happens is that
2231 it logs this as a hard delivery error, but does not attempt to create a
2234 PH/11 The exinext utility has a -C option for testing purposes, but although
2235 the given file was scanned by exinext itself; it wasn't being passed on
2236 when Exim was called.
2238 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2239 an end-of-file indication when reading a command response.
2241 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2242 compiled. In many other places in Exim, IPv6 addresses are always
2243 recognized, so I have changed this. It also means that IPv4 domain
2244 literals of the form [IPV4:n.n.n.n] are now always recognized.
2246 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2247 used if the router is not running as root, for example, when verifying at
2248 ACL time, or when using -bh. The debugging output from this situation was
2249 non-existent - all you got was a failure to exec. I have made two
2252 (a) Failures to set uid/gid, the current directory, or a process leader
2253 in a subprocess such as that created by queryprogram now generate
2254 suitable debugging ouput when -d is set.
2256 (b) The queryprogram router detects when it is not running as root,
2257 outputs suitable debugging information if -d is set, and then runs
2258 the subprocess without attempting to change uid/gid.
2260 PH/15 Minor change to Makefile for building test_host (undocumented testing
2263 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2264 additional section of a DNS packet that returns MX or SRV records.
2265 Instead, it always explicitly searches for A/AAAA records. This avoids
2266 major problems that occur when a DNS server includes only records of one
2267 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2268 fixed another bug: if SRV records were looked up and the corresponding
2269 address records were *not* found in the additional section, the port
2270 values from the SRV records were lost.
2272 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2273 using the correct key (the original address) when searching the retry
2274 rules in order to find which one to use for generating the retry hint.
2276 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2277 adding the default one. Similarly, if it contains a Reply-To: header, the
2278 errors_reply_to option, if set, is not used.
2280 PH/19 When calculating a retry time, Exim used to measure the "time since
2281 failure" by looking at the "first failed" field in the retry record. Now
2282 it does not use this if it is later than than the arrival time of the
2283 message. Instead it uses the arrival time. This makes for better
2284 behaviour in cases where some deliveries succeed, thus re-setting the
2285 "first failed" field. An example is a quota failure for a huge message
2286 when small messages continue to be delivered. Without this change, the
2287 "time since failure" will always be short, possible causing more frequent
2288 delivery attempts for the huge message than are intended.
2289 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2291 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2292 $1, $2, $3) because the numerical variables can be reset during some
2293 expansion items (e.g. "match"), thereby losing the authentication data.
2295 PH/21 Make -bV show the size of off_t variables so that the test suite can
2296 decide whether to run tests for quotas > 2G.
2298 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2299 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2300 filecount value is greater than 2G or if a quota value is greater than 2G
2301 on a system where the size of off_t is not greater than 4, a panic error
2304 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2305 never match. The debug and -bh output now contains an explicit error
2306 message indicating a malformed IPv4 address or mask.
2308 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2309 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2310 PH/23 above applies.
2312 PH/25 Do not write to syslog when running in the test harness. The only
2313 occasion when this arises is a failure to open the main or panic logs
2314 (for which there is an explicit test).
2316 PH/26 Added the /no_tell option to "control=freeze".
2318 PH/27 If a host name lookup failed very early in a connection, for example, if
2319 the IP address matched host_lookup and the reverse lookup yielded a name
2320 that did not have a forward lookup, an error message of the form "no IP
2321 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2322 could be logged. Now it outputs the IP address instead of "NULL".
2324 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2325 allows the sender and the authenticated sender to be set when
2326 submitting a message from within Exim. Since child_open_exim() is
2327 documented for local_scan(), the new function should be too.
2329 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2330 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2331 results in an empty string is now treated as unset.
2333 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2335 PH/31 Added sender_verify_fail logging option.
2337 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2338 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2339 tidied the source and removed it altogether.
2341 PH/33 When a queue run was abandoned because the load average was too high, a
2342 log line was always written; now it is written only if the queue_run log
2343 selector is set. In addition, the log line for abandonment now contains
2344 information about the queue run such as the pid. This is always present
2345 in "start" and "stop" lines but was omitted from the "abandon" line.
2347 PH/34 Omit spaces between a header name and the colon in the error message that
2348 is given when verify = headers_syntax fails (if there are lots of them,
2349 the message gets confusing).
2351 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2352 names, as there are now some PTR records that contain slashes. This check
2353 is only to protect against broken name servers that fall over on strange
2354 characters, so the fact that it applies to all lookups doesn't matter.
2356 PH/36 Now that the new test suite is complete, we can remove some of the
2357 special code in Exim that was needed for the old test suite. For example,
2358 sorting DNS records because real resolvers return them in an arbitrary
2359 order. The new test suite's fake resolver always returns records in the
2362 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2363 bounces) except when queue_only is set, to avoid logging races between
2364 the different processes.
2366 PH/38 Panic-die if .include specifies a non-absolute path.
2368 PH/39 A tweak to the "H" retry rule from its user.
2370 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2371 a label. They prevented compilation on older perls.
2373 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2374 a warning to be raised on newish perls.
2376 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2377 on queue. Changes to match documented behaviour of showing count of
2378 messages matching specified criteria.
2380 PH/40 Changed the default ident timeout from 30s to 5s.
2382 PH/41 Added support for the use of login_cap features, on those BSD systems
2383 that have them, for controlling the resources used by pipe deliveries.
2385 PH/42 The content-scanning code uses fopen() to create files in which to put
2386 message data. Previously it was not paying any attention to the mode of
2387 the files. Exim runs with umask(0) because the rest of the code creates
2388 files with open(), and sets the required mode explicitly. Thus, these
2389 files were ending up world-writeable. This was not a big issue, because,
2390 being within the spool directory, they were not world-accessible. I have
2391 created a function called modefopen, which takes an additional mode
2392 argument. It sets umask(777), creates the file, chmods it to the required
2393 mode, then resets the umask. All the relevant calls to fopen() in the
2394 content scanning code have been changed to use this function.
2396 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2397 to 24 hours. This avoids potential overflow problems when processing G
2398 and H retry rules. I suspect nobody ever tinkers with this value.
2400 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2402 PH/45 When the plaintext authenticator is running as a client, the server's
2403 challenges are checked to ensure they are valid base64 strings. By
2404 default, the authentication attempt is cancelled if an invalid string is
2405 received. Setting client_ignore_invalid_base64 true ignores these errors.
2406 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2407 they are received. Thus, the responses can be made to depend on the
2408 challenges. If an invalid string is ignored, an empty string is placed in
2411 PH/46 Messages that are created by the autoreply transport now contains a
2412 References: header, in accordance with RFCs 2822 and 3834.
2414 PH/47 Added authenticated_sender_force to the smtp transport.
2416 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2418 PH/49 Installed latest patch from the Sieve maintainer.
2420 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2421 was also set, the appendfile transport was unnecessarily scanning a
2422 directory of message files (e.g. for maildir delivery) to find the count
2423 of files (along with the size), even though it did not need this
2424 information. It now does the scan only if it needs to find either the
2425 size of the count of files.
2427 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2429 PH/52 Two bugs concerned with error handling when the smtp transport is
2432 (i) Exim was not creating retry information for temporary errors given
2433 for individual recipients after the DATA command when the smtp transport
2434 was used in LMTP mode. This meant that they could be retried too
2435 frequently, and not timed out correctly.
2437 (ii) Exim was setting the flag that allows error details to be returned
2438 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2439 recipients that were returned after the DATA command.
2441 PH/53 This is related to PH/52, but is more general: for any failing address,
2442 when detailed error information was permitted to be returned to the
2443 sender, but the error was temporary, then after the final timeout, only
2444 "retry timeout exceeded" was returned. Now it returns the full error as
2445 well as "retry timeout exceeded".
2447 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2448 do this, and (what is worse) MTAs that accept it.
2450 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2451 will now be deprecated.
2453 PH/56 New os.c-cygwin from the Cygwin maintainer.
2455 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2456 formats (previously only available in exim formats via -bpr, -bpru,
2457 and -bpra. Now also available in native and exiqgrep formats)
2459 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2460 with very large, slow to parse queues
2462 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2464 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2466 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2467 responses to authentication challenges, though it was showing the
2468 challenges; (ii) I've removed the CR characters from the debug output for
2471 PH/58 Allow for the insertion of a newline as well as a space when a string
2472 is turned into more than one encoded-word during RFC 2047 encoding. The
2473 Sieve code now uses this.
2475 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2476 data_4xx, lost_connection, tls_required.
2478 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2479 message was being sent as an SMTP response.
2481 PH/61 Add -l and -k options to exicyclog.
2483 PH/62 When verifying, if an address was redirected to one new address, so that
2484 verification continued, and the new address failed or deferred after
2485 having set something in $address_data, the value of $address_data was not
2486 passed back to the ACL. This was different to the case when no
2487 redirection occurred. The value is now passed back in both cases.
2489 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2490 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2491 use login_cap.h, so on its own it isn't the distinguishing feature. The
2492 new name refers directly to the setclassresources() function.
2494 PH/65 Added configuration files for NetBSD3.
2496 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2498 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2499 is preferred over IPv4.
2501 PH/68 The bounce_return_message and bounce_return_body options were not being
2502 honoured for bounces generated during the reception of non-SMTP messages.
2503 In particular, this applied to messages rejected by the ACL. This bug has
2504 been fixed. However, if bounce_return_message is true and bounce_return_
2505 body is false, the headers that are returned for a non-SMTP message
2506 include only those that have been read before the error was detected.
2507 (In the case of an ACL rejection, they have all been read.)
2509 PH/69 The HTML version of the specification is now built in a directory called
2510 spec_html instead of spec.html, because the latter looks like a path with
2511 a MIME-type, and this confuses some software.
2513 PH/70 Catch two compiler warnings in sieve.c.
2515 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2516 function verify_get_ident() calls ip_connect() to connect a socket, but
2517 if the "connect()" function timed out, ip_connect() used to close the
2518 socket. However, verify_get_ident() also closes the socket later, and in
2519 between Exim writes to the log, which may get opened at this point. When
2520 the socket was closed in ip_connect(), the log could get the same file
2521 descriptor number as the socket. This naturally causes chaos. The fix is
2522 not to close the socket in ip_connect(); the socket should be closed by
2523 the function that creates it. There was only one place in the code where
2524 this was missing, in the iplookup router, which I don't think anybody now
2525 uses, but I've fixed it anyway.
2527 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2528 well as to direct DNS lookups. Otherwise the handling of names in host
2529 lists is inconsistent and therefore confusing.
2535 PH/01 Two changes to the default runtime configuration:
2537 (1) Move the checks for relay_from_hosts and authenticated clients from
2538 after to before the (commented out) DNS black list checks.
2540 (2) Add control=submission to the relay_from_hosts and authenticated
2541 clients checks, on the grounds that messages accepted by these
2542 statements are most likely to be submissions.
2544 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2546 (1) Generate an error if the third argument for the ${prvs expansion is
2549 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2552 (3) Reset the variables that are obtained from the first argument of
2553 ${prvscheck and used in the second argument before leaving the code,
2554 because their memory is reclaimed, so using them afterwards may do
2557 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2558 one (it's much easier than Tom thought :-).
2560 (5) Because of (4), we can now allow for the use of $prvscheck_result
2561 inside the third argument.
2563 PH/03 For some reason, the default setting of PATH when running a command from
2564 a pipe transport was just "/usr/bin". I have changed it to
2567 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2568 anything to be listed in the output from -bV.
2570 PH/05 When a filter generated an autoreply, the entire To: header line was
2571 quoted in the delivery log line, like this:
2573 => >A.N.Other <ano@some.domain> <original@ddress> ...
2575 This has been changed so that it extracts the operative address. There
2576 may be more than one such address. If so, they are comma-separated, like
2579 => >ano@some.domain,ona@other.domain <original@ddress> ...
2581 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2582 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2583 not being looked up in the rDNS to get a host name, Exim was showing the
2584 IP address twice in Received: lines, even though the IP addresses were
2585 identical. For example:
2587 Received: from [1.2.3.4] (helo=[1.2.3.4])
2589 However, if the real host name was known, it was omitting the HELO data
2590 if it matched the actual IP address. This has been tidied up so that it
2591 doesn't show the same IP address twice.
2593 PH/07 When both +timestamp and +memory debugging was on, the value given by
2594 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2595 called by the debug printing, thereby overwriting the timestamp buffer.
2596 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2598 PH/08 When the original message was included in an autoreply transport, it
2599 always said "this is a copy of the message, including all the headers",
2600 even if body_only or headers_only was set. It now gives an appropriate
2603 PH/09 Applied a patch from the Sieve maintainer which:
2605 o fixes some comments
2606 o adds the (disabled) notify extension core
2607 o adds some debug output for the result of if/elsif tests
2608 o points to the current vacation draft in the documentation
2609 and documents the missing references header update
2613 o fixes a bug in processing the envelope test (when testing
2614 multiple envelope elements, the last element determinted the
2617 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2618 Electronic Mail") by including:
2620 Auto-submitted: auto-generated
2622 in the messages that it generates (bounce messages and others, such as
2623 warnings). In the case of bounce messages for non-SMTP mesages, there was
2624 also a typo: it was using "Auto_submitted" (underscore instead of
2625 hyphen). Since every message generated by Exim is necessarily in response
2626 to another message, thes have all been changed to:
2628 Auto-Submitted: auto-replied
2630 in accordance with these statements in the RFC:
2632 The auto-replied keyword:
2634 - SHOULD be used on messages sent in direct response to another
2635 message by an automatic process,
2637 - MUST NOT be used on manually-generated messages,
2639 - MAY be used on Delivery Status Notifications (DSNs) and Message
2640 Disposition Notifications (MDNs),
2642 - MUST NOT be used on messages generated by automatic or periodic
2643 processes, except for messages which are automatic responses to
2646 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2647 to the default Received: header definition.
2649 PH/12 Added log selector acl_warn_skipped (default on).
2651 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2652 variables because (a) they are in the wrong storage pool and (b) even if
2653 they were copied, it wouldn't work properly because of the caching.
2655 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2656 checking when decoding. Apparently there are clients that generate
2657 overlong encoded strings. Why am I not surprised?
2659 PH/15 If the first argument of "${if match_address" was not empty, but did not
2660 contain an "@" character, Exim crashed. Now it writes a panic log message
2661 and treats the condition as false.
2663 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2665 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2666 extension "envelope-auth". The code is finished and in agreement with
2667 other implementations, but there is no documentation so far and in fact,
2668 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2669 not changing the active code.
2671 Print executed "if" and "elsif" statements when debugging is used. This
2672 helps a great deal to understand what a filter does.
2674 Document more things not specified clearly in RFC3028. I had all this
2675 sorted out, when out of a sudden new issues came to my mind. Oops."
2677 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2680 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2681 canonical form (as documented). However, after a host name lookup from
2682 the IP address, check_host() was doing a simple string comparison with
2683 addresses acquired from the DNS when checking that the found name did
2684 have the original IP as one of its addresses. Since any found IPv6
2685 addresses are likely to be in abbreviated form, the comparison could
2686 fail. Luckily, there already exists a function for doing the comparison
2687 by converting both addresses to binary, so now that is used instead of
2688 the text comparison.
2690 PH/20 There was another similar case to PH/19, when a complete host name was
2691 given in a host list; looking up its IP address could give an abbreviated
2692 form, whereas the current host's name might or might not be abbreviated.
2693 The same fix has been applied.
2699 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2700 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2703 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2704 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2706 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2708 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2709 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2710 Also, for "Auto-Submitted", treat anything other than "no" as
2711 non-personal, in accordance with RFC 3834. (Previously it treated
2712 anything starting "auto-" as non-personal.)
2714 TF/01 The control=submission/name=... option had a problem with syntax
2715 errors if the name included a slash character. The /name= option
2716 now slurps the rest of the string, so it can include any characters
2717 but it must come last in the list of options (after /sender_retain
2720 PH/05 Some modifications to the interface to the fake nameserver for the new
2728 TK/01 Added the "success_on_redirect" address verification option. See
2729 NewStuff for rationale and an example.
2731 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2733 PH/02 Patch to exigrep to allow it to work on syslog lines.
2735 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2736 fread() to skip over the body file's header line, because in Cygwin the
2737 header line is locked and is inaccessible.
2739 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2740 co-exist for some time) to make it clear that it is the Exim ID that is
2741 referenced, not the Message-ID: header line.
2743 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2744 string_format() function, because snprintf() does not exist on all
2747 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2748 ${stat: expansion item.
2750 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2751 protocol synchronization error", to keep the pedants happy.
2753 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2754 well as for IRIX systems, when gcc is being used. See the host.c source
2757 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2759 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2762 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2763 transport and to the smtp transport in LMTP mode.
2765 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2767 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2768 run a filter in a subprocess. This could lead to confusion in subsequent
2769 lookups in the parent process. There should also be a search_tidyup() at
2770 the end of the subprocess.
2772 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2773 only if the host matched helo_try_verify_hosts, which caused the
2774 verification to occur when the EHLO/HELO command was issued. The ACL just
2775 tested the remembered result. Now, if a previous verification attempt has
2776 not happened, "verify = helo" does it there and then.
2778 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2780 TK/03 Fix log output including CR from clamd.
2782 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2783 exist provoked a memory error which could cause a segfault.
2785 PH/15 Installed PCRE 6.2
2787 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2789 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2790 of the problem. Specifically, suggested +O2 rather than +O1 for the
2793 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2795 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2796 record, the port was not correctly logged when the outgoing_port log
2797 selector was set (it logged the transort's default port).
2799 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2800 fallback_hosts, and "hosts" in the smtp transport.
2802 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2803 host errors such as "Connection refused".
2805 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2806 authentication with radiusclient 0.4.9:
2808 - Error returned from rc_read_config was caught wrongly
2809 - Username/password not passed on to radius server due to wrong length.
2811 The presumption is that some radiusclient API changes for 4.51/PH/17
2812 were not taken care of correctly. The code is still untested by me (my
2813 Linux distribution still has 0.3.2 of radiusclient), but it was
2814 contributed by a Radius user.
2816 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2817 expanding the "port" option of the smtp transport.
2819 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2820 while reading a MIME header. Thanks to Tom Hughes for a patch.
2822 PH/24 Include config.h inside local_scan.h so that configuration settings are
2825 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2826 that in an ACL for RCPT (for example), you can examine exactly what was
2829 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2830 commands, but it was not correctly comparing the address with the actual
2831 client host address. Thus, it would show the EHLO address in Received:
2832 header lines when this was not necessary.
2834 PH/27 Added the % operator to ${eval:}.
2836 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2837 it should be ignoring failures (because with -C, for example, it has lost
2838 privilege). It wasn't ignoring creation failures other than "already
2841 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2844 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2845 because an input line was too long, either on its own, or by virtue of
2846 too many continuations, the temporary file was not being removed, and the
2847 return code was incorrect.
2849 PH/31 Missing "BOOL" in function definition in filtertest.c.
2851 PH/32 Applied Sieve patches from the maintainer.
2853 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2855 PH/33 Added "verify = not_blind".
2857 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2858 Local/Makefile (with some defaults set). These are used in built scripts
2859 such as exicyclog, but they have never been used in the exim_install
2860 script (though there are many overriding facilities there). I have
2861 arranged that the exim_install script now takes note of these two
2864 PH/35 Installed configuration files for Dragonfly.
2866 PH/36 When a locally submitted message by a trusted user did not contain a
2867 From: header, and the sender address was obtained from -f or from an SMTP
2868 MAIL command, and the trusted user did not use -F to supply a sender
2869 name, $originator_name was incorrectly used when constructing a From:
2870 header. Furthermore, $originator_name was used for submission mode
2871 messages from external hosts without From: headers in a similar way,
2872 which is clearly wrong.
2874 PH/37 Added control=suppress_local_fixups.
2876 PH/38 When log_selector = +received_sender was set, and the addition of the
2877 sender made the log line's construction buffer exactly full, or one byte
2878 less than full, an overflow happened when the terminating "\n" was
2881 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2882 when the result of a list match is failure because a DNS lookup failed.
2884 PH/40 RM_COMMAND is now used in the building process.
2886 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2887 the "build-* directories that it finds.
2889 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2890 address in a domain literal was a prefix of an interface address.
2892 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2893 when verifying a sender address, unless rewrite_headers is false.
2895 PH/44 Wrote a long comment about why errors_to addresses are verified as
2896 recipients, not senders.
2898 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2899 the ratelimit ACL was added.
2901 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2903 PH/46 Added extra information about PostgreSQL errors to the error string.
2905 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2906 avoiding the need to install special zones in a real server. This is
2907 backwards compatible; if it can't find the fake resolver, it drops back.
2908 Thus, both old and new test suites can be run.
2910 TF/02 Added util/ratelimit.pl
2912 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2913 clock is set back in time.
2915 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2916 Candler <B.Candler@pobox.com>.
2918 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2919 OK for addresses that are the result of redirections.
2921 PH/48 A number of further additions for the benefit of the new test suite,
2922 including a fake gethostbyname() that interfaces to the fake DNS resolver
2925 TF/06 The fix for widen_domains has also been applied to qualify_single and
2926 search_parents which are the other dnslookup options that can cause
2929 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2932 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2934 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2935 service" if it encountered an MX record with an empty target hostname.
2936 The message is now "an MX or SRV record indicated no SMTP service".
2938 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2939 if the DNS of the sending site is misconfigured. This is quite a
2940 common situation. This change restores the behaviour of treating a
2941 helo verification defer as a failure.
2943 PH/51 If self=fail was set on a router, the bounce message did not include the
2944 actual error message.
2950 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2952 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2953 command itself ended in error, the underlying message about the transport
2954 filter timeout was being overwritten with the pipe command error. Now the
2955 underlying error message should be appended to the second error message.
2957 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2959 PH/02 Reduce the amount of output that "make" produces by default. Full output
2960 can still be requested.
2962 PH/03 The warning log line about a condition test deferring for a "warn" verb
2963 was being output only once per connection, rather than after each
2964 occurrence (because it was using the same function as for successful
2965 "warn" verbs). This seems wrong, so I have changed it.
2967 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2968 it should not have, which might have caused a crash in the right
2969 circumstances, but probably never did.
2971 PH/04 Installed a modified version of Tony Finch's patch to make submission
2972 mode fix the return path as well as the Sender: header line, and to
2973 add a /name= option so that you can make the user's friendly name appear
2976 TF/03 Added the control = fakedefer ACL modifier.
2978 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2979 Mark Lowes for thorough testing.
2981 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2983 TK/03 Merged latest SRS patch from Miles Wilton.
2985 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2986 with the definition in sysexits.h (which is #included earlier).
2987 Fortunately, Exim does not actually use EX_OK. The code used to try to
2988 preserve the sysexits.h value, by assumimg that macro definitions were
2989 scanned for macro replacements. I have been disabused of this notion,
2990 so now the code just undefines EX_OK before #including unistd.h.
2992 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2993 in the smtp transport. When a block could not be written in a single
2994 write() function, the timeout was being re-applied to each part-write.
2995 This seems wrong - if the receiver was accepting one byte at a time it
2996 would take for ever. The timeout is now adjusted when this happens. It
2997 doesn't have to be particularly precise.
2999 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3000 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3002 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3003 without a domain if the check to <postmaster@domain> fails.
3005 SC/01 Eximstats: added -xls and the ability to specify output files
3006 (patch written by Frank Heydlauf).
3008 SC/02 Eximstats: use FileHandles for outputing results.
3010 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3012 SC/04 Eximstats: fixed display of large numbers with -nvr option
3014 SC/05 Eximstats: fixed merging of reports with empty tables.
3016 SC/06 Eximstats: added the -include_original_destination flag
3018 SC/07 Eximstats: removed tabs and trailing whitespace.
3020 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3022 TK/06 MBOX spool code: Add real "From " MBOX separator line
3023 so the .eml file is really in mbox format (even though
3024 most programs do not really care). Patch from Alex Miller.
3026 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3027 The latter is generated from $received_to and is only set if the
3028 message has one envelope recipient. SA can use these headers,
3029 obviously out-of-the-box. Patch from Alex Miller.
3031 PH/08 The ${def test on a variable was returning false if the variable's
3032 value was "0", contrary to what the specification has always said!
3033 The result should be true unless the variable is empty.
3035 PH/09 The syntax error of a character other than { following "${if
3036 def:variable_name" (after optional whitespace) was not being diagnosed.
3037 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3038 accidental colon was present, for example, could give incorrect results.
3040 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3041 result is used (not including appendfile, where other changes are about
3044 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3045 This involved changing a lot of size variables from int to off_t. It
3046 should work with maildirs and everything.
3048 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3049 spamd dying while we are connected to it.
3051 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3054 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3055 function for BDB 4.3.
3057 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3059 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3060 the sources for PCRE have been re-arranged and more files are now
3063 PH/15 The code I had for printing potentially long long variables in PH/11
3064 above was not the best (it lost precision). The length of off_t variables
3065 is now inspected at build time, and an appropriate printing format (%ld
3066 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3067 to be "long long int" or "long int". This is needed for the internal
3068 formatting function string_vformat().
3070 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3071 the configuration file to be ":syslog", then the script "guesses" where
3072 the logs files are, rather than using the compiled in default. In our
3073 case the guess is not the same as the compiled default, so the script
3074 suddenly stopped working when I started to use syslog. The patch checks
3075 to see if log_file_path is "". If so, it attempts to read it from exim
3076 with no configuration file to get the compiled in version, before it
3077 falls back to the previous guessing code."
3079 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3080 implementing BATV in an Exim configuration. See NewStuff for the gory
3083 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3084 Makefile that are specific to HP-UX.
3086 PH/18 If the "use_postmaster" option was set for a recipient callout together
3087 with the "random" option, the postmaster address was used as the MAIL
3088 FROM address for the random test, but not for the subsequent recipient
3089 test. It is now used for both.
3091 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3092 patch removes a few documentation additions to RFC 3028, because the
3093 latest draft now contains them. It adds the new en;ascii-case comparator
3094 and a new error check for 8bit text in MIME parts. Comparator and
3095 require names are now matched exactly. I enabled the subaddress
3096 extension, but it is not well tested yet (read: it works for me)."
3098 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3099 rework some of the code of TK/09 above to avoid the hardwired use of
3100 "%lld" and "long long". Replaced the call to snprintf() with a call to
3103 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3104 records point to non-existent hosts", "retry timeout exceeded", and
3105 "retry time not reached for any host after a long failure period".
3107 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3108 experimental DomainKeys support:
3110 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3111 (2) On an error such as an illegally used "control", the wrong name for
3112 the control was given.
3114 These problems did NOT occur unless DomainKeys support was compiled.
3116 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3118 PH/24 Added ${if match_ip condition.
3120 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3121 databases so that it will be absolutely obvious if a crash occurs in the
3122 DB library. This is a regular occurrence (often caused by mis-matched
3125 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3126 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3127 picked up on a user's system that detects such things. There doesn't seem
3128 to be a gcc warning option for this - only an attribute that has to be
3129 put on the function's prototype. It seems that in Fedora Core 4 they have
3130 set this on a number of new functions. No doubt there will be more in due
3133 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3134 specify a transport. However, if an address that was verified by such a
3135 router was the subject of a callout, Exim crashed because it tried to
3136 read the rcpt_include_affixes from the non-existent transport. Now it
3137 just assumes that the setting of that option is false. This bug was
3138 introduced by 4.51/PH/31.
3140 PH/28 Changed -d+all to exclude +memory, because that information is very
3141 rarely of interest, but it makes the output a lot bigger. People tend to
3142 do -d+all out of habit.
3144 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3145 code in os-type was giving problems when libc.so lives in lib64, like on
3148 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3149 aren't the modern standard, and it seems that some systems' include files
3150 don't always have them. Exim was already checking for some of the newer
3151 ones like T_AAAA, and defining it itself. I've added checks for all the
3152 record types that Exim uses.
3154 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3155 not automatically generating a new one, as it is supposed to. This
3156 prevented TLS from working. If the file did exist, but contained invalid
3157 data, a new version was generated, as expected. It was only the case of a
3158 non-existent file that was broken.
3160 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3161 with a change in libdomainkeys > 0.64.
3163 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3164 from DNS. If the selector record carries the flag, it now has
3165 precedence over the domain-wide flag.
3167 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3169 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3170 the use of a port name that isn't defined in /etc/services) occurred, the
3171 message was deferred as in a normal delivery, and thus remained on the
3172 spool, instead of being failed because of the mua_wrapper setting. This
3173 is now fixed, and I tidied up some of the mua_wrapper messages at the
3176 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3177 the messages in a hash of arrays rather than using individual hashes.
3178 This is a bit cleaner and results in dramatic memory savings, albeit
3179 at a slight CPU cost.
3181 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3182 as requested by Marc Sherman.
3184 SC/10 Eximstats: added histograms for user specified patterns as requested
3187 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3189 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3190 fopen() in the content-scanning modules that did not already have it.
3196 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3197 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3199 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3201 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3203 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3204 to test Sieve filters that use "vacation".
3206 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3207 that changes the way the GnuTLS parameters are stored in the cache file.
3208 The new format can be generated externally. For backward compatibility,
3209 if the data in the cache doesn't make sense, Exim assumes it has read an
3210 old-format file, and it generates new data and writes a new file. This
3211 means that you can't go back to an older release without removing the
3214 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3215 work if there are any delivery delays because "one_time" forces the
3216 parent to be marked "delivered", so its unseen clone is never tried
3217 again. For this reason, Exim now forbids the simultaneous setting of
3220 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3221 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3222 entry if you want to know the details. The fix, however, neglected to
3223 consider the case when local delivery batching is involved. The test for
3224 "previously delivered" was not happening when checking to see if an
3225 address could be batched with a previous (undelivered) one; under
3226 certain circumstances this could lead to multiple deliveries to the same
3229 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3230 in its include files, and this causes problems building Exim.
3232 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3233 header_syntax) but Exim was just ignoring anything given after a slash.
3234 In particular, this caused confusion with an attempt to use "verify =
3235 reverse_host_lookup/defer_ok". An error is now given when options are
3236 supplied for verify items that do not have them. (Maybe reverse_host_
3237 lookup should have a defer_ok option, but that's a different point.)
3239 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3240 defined by RFC 821) to 2048, because there were problems with some AUTH
3241 commands, and RFC 1869 says the size should be increased for extended
3242 SMTP commands that take arguments.
3244 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3247 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3248 "unknown" error; now it says that the functionality isn't in the binary.
3250 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3251 an address' error message when a string expansion fails (syntax or
3252 whatever). Otherwise the password may appear in the log. Following change
3253 PH/42 below, there is no longer a chance of it appearing in a bounce
3256 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3258 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3259 of its addresses was ever tried. (Bugzilla bug #2.)
3261 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3262 the result incorrectly in the debug output. (It correctly added a newline
3263 to what was transported.)
3265 TF/01 Added $received_time.
3267 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3268 commented out examples of how to interface to a virus scanner and to
3269 SpamAssassin. Also added commented examples of av_scanner and
3270 spamd_address settings.
3272 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3273 and controls are allowed in which ACLs. There were a couple of minor
3274 errors. Some of the entries in the conditions table (which is a table of
3275 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3276 negation of where the condition IS allowed.
3278 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3280 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3281 header file does not have a version number, so I've had to invent a new
3282 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3283 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3284 radiusclient), but it was contributed by a Radius user.
3286 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3287 files or directories, for OpenSSL.
3289 PH/19 When an Exim process that is running as root has to create an Exim log
3290 file, it does so in a subprocess that runs as exim:exim so as to get the
3291 ownership right at creation (otherwise, other Exim processes might see
3292 the file with the wrong ownership). There was no test for failure of this
3293 fork() call, which would lead to the process getting stuck as it waited
3294 for a non-existent subprocess. Forks do occasionally fail when resources
3295 run out. I reviewed all the other calls to fork(); they all seem to check
3298 PH/20 When checking for unexpected SMTP input at connect time (before writing
3299 the banner), Exim was not dealing correctly with a non-positive return
3300 from the read() function. If the client had disconnected by this time,
3301 the result was a log entry for a synchronization error with an empty
3302 string after "input=" when read() returned zero. If read() returned -1
3303 (an event I could not check), uninitialized data bytes were printed.
3304 There were reports of junk text (parts of files, etc) appearing after
3307 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3309 PH/22 Added support for macro redefinition, and (re)definition in between
3310 driver and ACL definitions.
3312 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3313 forgetting to use the resulting value; it was using the unexpanded value.
3315 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3316 hadn't been configured. The fix is from Juergen Kreileder, who
3317 understands it better than I do:
3319 "Here's what I see happening with three configured cyrus_sasl
3320 authenticators configured (plain, login, cram-md5):
3322 On startup auth_cyrus_sasl_init() gets called for each of these.
3323 This means three calls to sasl_listmech() without a specified mech_list.
3324 => SASL tests which mechs of all available mechs actually work
3325 => three warnings about OTP not working
3326 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3328 With the patch, sasl_listmech() also gets called three times. But now
3329 SASL's mech_list option is set to the server_mech specified in the the
3330 authenticator. Or in other words, the answer from sasl_listmech()
3331 gets limited to just the mech you're testing for (which is different
3333 => the return list contains just 'plain' or 'login', 'cram-md5' or
3334 nothing depending on the value of ob->server_mech.
3336 I've just tested the patch: Authentication still works fine,
3337 unavailable mechs specified in the exim configuration are still
3338 caught, and the auth.log warnings about OTP are gone."
3340 PH/25 When debugging is enabled, the contents of the command line are added
3341 to the debugging output, even when log_selector=+arguments is not
3344 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3345 answer is "GNU", and only if the return is "GNU/something" is the answer
3348 PH/27 $acl_verify_message is now set immediately after the failure of a
3349 verification in an ACL, and so is available in subsequent modifiers. In
3350 particular, the message can be preserved by coding like this:
3352 warn !verify = sender
3353 set acl_m0 = $acl_verify_message
3355 Previously, $acl_verify_message was set only while expanding "message"
3356 and "log_message" when a very denied access.
3358 PH/28 Modified OS/os.c-Linux with
3360 -#ifndef OS_LOAD_AVERAGE
3361 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3363 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3364 nomenclature these days.)
3366 PH/29 Installed patch from the Sieve maintainer that adds the options
3367 sieve_useraddress and sieve_subaddress to the redirect router.
3369 PH/30 In these circumstances:
3370 . Two addresses routed to the same list of hosts;
3371 . First host does not offer TLS;
3372 . First host accepts first address;
3373 . First host gives temporary error to second address;
3374 . Second host offers TLS and a TLS session is established;
3375 . Second host accepts second address.
3376 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3377 and peerdn, if requested) that were in fact used only for the second
3380 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3381 attention to any local part prefix or suffix that was matched by the
3382 router that accepted the address. It now behaves in the same way as it
3383 does for delivery: the affixes are removed from the local part unless
3384 rcpt_include_affixes is set on the transport.
3386 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3387 timeout during the DATA phase of an incoming message.
3389 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3390 applied a patch sanctioned by the Sieve maintainer.
3392 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3393 the uid or gid is negative. A case of a negative gid caused this to be
3394 noticed. The fix allows for either to be negative.
3396 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3397 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3400 PH/36 The change PH/12 above was broken. Fixed it.
3402 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3403 the grounds that routing the same address twice would always produce the
3404 same answer. This might have been true once, but it is certainly no
3405 longer true now. Routing a child address may depend on the previous
3406 routing that produced that child. Some complicated redirection strategies
3407 went wrong when messages had multiple recipients, and made Exim's
3408 behaviour dependent on the order in which the addresses were given.
3410 I have moved the duplicate checking until after the routing is complete.
3411 Exim scans the addresses that are assigned to local and remote
3412 transports, and removes any duplicates. This means that more work will be
3413 done, as duplicates will always all be routed, but duplicates are
3414 presumably rare, so I don't expect this is of any significance.
3416 For deliveries to pipes, files, and autoreplies, the duplicate checking
3417 still happens during the routing process, since they are not going to be
3420 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3421 It corrects a timeout issue with spamd. This is Ian's comment: "The
3422 background is that sometimes spamd either never reads data from a
3423 connection it has accepted, or it never writes response data. The exiscan
3424 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3425 blindly assumes that writes won't block so it may never time out."
3427 PH/39 Allow G after quota size as well as K and M.
3429 PH/40 The value set for $authenticated_id in an authenticator may not contain
3430 binary zeroes or newlines because the value is written to log lines and
3431 to spool files. There was no check on this. Now the value is run through
3432 the string_printing() function so that such characters are converted to
3433 printable escape sequences.
3435 PH/41 $message_linecount is a new variable that contains the total number of
3436 lines in the message. Compare $body_linecount, which is the count for the
3439 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3440 bounce and delay warning messages, except in certain special cases, which
3443 (a) An SMTP error message from a remote host;
3444 (b) A message specified in a :fail: redirection;
3445 (c) A message specified in a "fail" command in a system filter;
3446 (d) A message specified in a FAIL return from the queryprogram router;
3447 (e) A message specified by the cannot_route_message router option.
3449 In these cases only, Exim does include the error details in bounce and
3450 warning messages. There are also a few cases where bland messages such
3451 as "unrouteable address" or "local delivery error" are given.
3453 PH/43 $value is now also set for the "else" part of a ${run expansion.
3455 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3456 being worked on, but at least Exim now implements the latest version to
3459 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3460 process to complete was treated as a delivery failure, a timeout while
3461 writing the message to the pipe was logged, but erroneously treated as a
3462 successful delivery. Such timeouts include transport filter timeouts. For
3463 consistency with the overall process timeout, these timeouts are now
3464 treated as errors, giving rise to delivery failures by default. However,
3465 there is now a new Boolean option for the pipe transport called
3466 timeout_defer, which, if set TRUE, converts the failures into defers for
3467 both kinds of timeout. A transport filter timeout is now identified in
3470 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3471 systems where "make" and "gmake" are different, calling "gmake" at top
3472 level broke things. I've arranged for the value of $(MAKE) to be passed
3473 from the Makefile to this script so that it can call the same version of
3477 A note about Exim versions 4.44 and 4.50
3478 ----------------------------------------
3480 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3481 changes of various kinds. As a consequence, a big documentation update was
3482 needed. This delayed the release for rather longer than seemed good, especially
3483 in the light of a couple of (minor) security issues. Therefore, the changes
3484 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3485 release. So 4.44 and 4.50 are in effect two different branches that both start
3488 I have left the 4.50 change log unchanged; it contains all the changes since
3489 4.43. The change log for 4.44 is below; many of its items are identical to
3490 those for 4.50. This seems to be the most sensible way to preserve the
3491 historical information.
3497 1. Minor wording change to the doc/README.SIEVE file.
3499 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3500 computation of the current number of files was incorrect.
3502 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3503 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3506 4. Give more explanation in the error message when the command for a transport
3507 filter fails to execute.
3509 5. There are several places where Exim runs a non-Exim command in a
3510 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3511 was being done only for the command run by the queryprogram router. It is
3512 now done for all such subprocesses. The other cases are: ${run, transport
3513 filters, and the commands run by the lmtp and pipe transports.
3515 6. Added CONFIGURE_GROUP build-time option.
3517 7. Some older OS have a limit of 256 on the maximum number of file
3518 descriptors. Exim was using setrlimit() to set 1000 as a large value
3519 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3520 systems. I've change it so that if it can't get 1000, it tries for 256.
3522 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3523 was an oversight, and furthermore, ever since the addition of extra
3524 controls (e.g. 4.43/32), the checks on when to allow different forms of
3525 "control" were broken. There should now be diagnostics for all cases when a
3526 control that does not make sense is encountered.
3528 9. Added the /retain_sender option to "control=submission".
3530 10. $recipients is now available in the predata ACL (oversight).
3532 11. Tidy the search cache before the fork to do a delivery from a message
3533 received from the command line. Otherwise the child will trigger a lookup
3534 failure and thereby defer the delivery if it tries to use (for example) a
3535 cached ldap connection that the parent has called unbind on.
3537 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3538 of $address_data from the recipient verification was clobbered by the
3539 sender verification.
3541 13. The value of address_data from a sender verification is now available in
3542 $sender_address_data in subsequent conditions in the ACL statement.
3544 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3546 15. Added a new option "connect=<time>" to callout options, to set a different
3549 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3550 was its contents. (It was OK if the option was not defined at all.)
3552 17. A "Completed" log line is now written for messages that are removed from
3553 the spool by the -Mrm option.
3555 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3556 information about exactly what failed.
3558 19. Added -dd to debug only the daemon process.
3560 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3561 handles timeouts, both on the server side and network timeouts. Renamed the
3562 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3564 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3565 It is now set to "smtps".
3567 22. $host_address is now set to the target address during the checking of
3568 ignore_target_hosts.
3570 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3571 being passed; this would have caused $sender_host_name to have been used if
3572 matching the list had actually called for a host name (not very likely,
3573 since this list is usually IP addresses). A host name is now passed as
3576 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3577 code that specifies a non-restarting handler (typically sigaction() in
3578 modern systems) in an attempt to fix a rare and obscure crash bug.
3580 25. Narrowed the window for a race in the daemon that could cause it to ignore
3581 SIGCHLD signals. This is not a major problem, because they are used only to
3582 wake it up if nothing else does.
3584 26. A malformed maildirsize file could cause Exim to calculate negative values
3585 for the mailbox size or file count. Odd effects could occur as a result.
3586 The maildirsize information is now recalculated if the size or filecount
3589 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3590 support for a long time. Removed HAVE_SYS_VFS_H.
3592 28. Installed the latest version of exipick from John Jetmore.
3594 29. In an address list, if the pattern was not a regular expression, an empty
3595 subject address (from a bounce message) matched only if the pattern was an
3596 empty string. Non-empty patterns were not even tested. This was the wrong
3597 because it is perfectly reasonable to use an empty address as part of a
3598 database query. An empty address is now tested by patterns that are
3599 lookups. However, all the other forms of pattern expect the subject to
3600 contain a local part and a domain, and therefore, for them, an empty
3601 address still always fails if the pattern is not itself empty.
3603 30. Exim went into a mad DNS loop when attempting to do a callout where the
3604 host was specified on an smtp transport, and looking it up yielded more
3605 than one IP address.
3607 31. Re-factored the code for checking spool and log partition space into a
3608 function that finds that data and another that does the check. The former
3609 is then used to implement four new variables: $spool_space, $log_space,
3610 $spool_inodes, and $log_inodes.
3612 32. The RFC2047 encoding function was originally intended for short strings
3613 such as real names; it was not keeping to the 75-character limit for
3614 encoded words that the RFC imposes. It now respects the limit, and
3615 generates multiple encoded words if necessary. To be on the safe side, I
3616 have increased the buffer size for the ${rfc2047: expansion operator from
3619 33. It is now permitted to omit both strings after an "if" condition; if the
3620 condition is true, the result is "true". As before, when the second string
3621 is omitted, a false condition yields an empty string. This makes it less
3622 cumbersome to write custom ACL and router conditions.
3624 34. Failure to deliver a bounce message always caused it to be frozen, even if
3625 there was an errors_to setting on the router. The errors_to setting is now
3628 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3629 canonical form (fully expanded) before being placed in
3630 $sender_host_address.
3632 36. The table in the code that translates DNS record types into text (T_A to
3633 "A" for instance) was missing entries for NS and CNAME. It is just possible
3634 that this could have caused confusion if both these types were looked up
3635 for the same domain, because the text type is used as part of Exim's
3636 per-process caching. But the chance of anyone hitting this buglet seems
3639 37. The dnsdb lookup has been extended in a number of ways.
3641 (1) There is a new type, "zns", which walks up the domain tree until it
3642 finds some nameserver records. It should be used with care.
3644 (2) There is a new type, "mxh", which is like "mx" except that it returns
3645 just the host names, not the priorities.
3647 (3) It is now possible to give a list of domains (or IP addresses) to be
3648 looked up. The behaviour when one of the lookups defers can be
3649 controlled by a keyword.
3651 (4) It is now possible to specify the separator character for use when
3652 multiple records are returned.
3654 38. The dnslists ACL condition has been extended: it is now possible to supply
3655 a list of IP addresses and/or domains to be looked up in a particular DNS
3658 39. Added log_selector=+queue_time_overall.
3660 40. When running the queue in the test harness, wait just a tad after forking a
3661 delivery process, to get repeatability of debugging output.
3663 41. Include certificate and key file names in error message when GnuTLS fails
3664 to set them up, because the GnuTLS error message doesn't include the name
3665 of the failing file when there is a problem reading it.
3667 42. Allow both -bf and -bF in the same test run.
3669 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3671 44. The "Exiscan patch" is now merged into the mainline Exim source.
3673 45. Sometimes the final signoff response after QUIT could fail to get
3674 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3675 < 0 before doing a fflush(). This bug looks as though it goes back to the
3676 introduction of TLS in release 3.20, but "sometimes" must have been rare
3677 because the tests only now provoked it.
3679 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3680 (this can affect the format of dates).
3682 47. exim_tidydb, when checking for the continued existence of a message for
3683 which it has found a message-specific retry record, was not finding
3684 messages that were in split spool directories. Consequently, it was
3685 deleting retry records that should have stayed in existence.
3687 48. Steve fixed some bugs in eximstats.
3689 49. The SPA authentication driver was not abandoning authentication and moving
3690 on to the next authenticator when an expansion was forced to fail,
3691 contradicting the general specification for all authenticators. Instead it
3692 was generating a temporary error. It now behaves as specified.
3694 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3695 (the order specifies the preference for clients). The order is now AES256,
3696 AES128, 3DES, ARCFOUR128.
3698 51. Small patch to Sieve code - explicitly set From: when generating an
3701 52. Exim crashed if a remote delivery caused a very long error message to be
3702 recorded - for instance if somebody sent an entire SpamAssassin report back
3703 as a large number of 550 error lines. This bug was coincidentally fixed by
3704 increasing the size of one of Exim's internal buffers (big_buffer) that
3705 happened as part of the Exiscan merge. However, to be on the safe side, I
3706 have made the code more robust (and fixed the comments that describe what
3709 53. Now that there can be additional text after "Completed" in log lines (if
3710 the queue_time_overall log selector is set), a one-byte patch to exigrep
3711 was needed to allow it to recognize "Completed" as not the last thing in
3714 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3715 patch that reportedly fixes this has been added. I am not expert enough to
3716 create a test for it. This is what the patch creator wrote:
3718 "I found a little strange behaviour of ldap code when working with
3719 Windows 2003 AD Domain, where users was placed in more than one
3720 Organization Units. When I tried to give exim partial DN, the exit code
3721 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3722 But simultaneously result of request was absolutely normal ldap result,
3723 so I produce this patch..."
3725 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3726 REFERENCE, so I have modified the code to exclude the patch when that macro
3729 55. Some experimental protocols are using DNS PTR records for new purposes. The
3730 keys for these records are domain names, not reversed IP addresses. The
3731 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3732 leaves it alone. Component reversal etc. now happens only for IP addresses.
3735 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3737 57. Double the size of the debug message buffer (to 2048) so that more of very
3738 long debug lines gets shown.
3740 58. The exicyclog utility now does better if the number of log files to keep
3741 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3743 59. Two changes related to the smtp_active_hostname option:
3745 (1) $smtp_active_hostname is now available as a variable.
3746 (2) The default for smtp_banner uses $smtp_active_hostname instead
3747 of $primary_hostname.
3749 60. The host_aton() function is supposed to be passed a string that is known
3750 to be a valid IP address. However, in the case of IPv6 addresses, it was
3751 not checking this. This is a hostage to fortune. Exim now panics and dies
3752 if the condition is not met. A case was found where this could be provoked
3753 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3754 components; fortuitously, this particular loophole had already been fixed
3755 by change 4.50/55 above.
3757 If there are any other similar loopholes, the new check in host_aton()
3758 itself should stop them being exploited. The report I received stated that
3759 data on the command line could provoke the exploit when Exim was running as
3760 exim, but did not say which command line option was involved. All I could
3761 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3762 running as the user.
3765 61. There was a buffer overflow vulnerability in the SPA authentication code
3766 (which came originally from the Samba project). I have added a test to the
3767 spa_base64_to_bits() function which I hope fixes it.
3770 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3771 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3773 63. The daemon start-up calls getloadavg() while still root for those OS that
3774 need the first call to be done as root, but it missed one case: when
3775 deliver_queue_load_max is set with deliver_drop_privilege. This is
3776 necessary for the benefit of the queue runner, because there is no re-exec
3777 when deliver_drop_privilege is set.
3779 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3780 This has been fixed.
3782 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3783 was in use, was not putting the data itself into the right store pool;
3784 consequently, it could be overwritten for a subsequent message in the same
3785 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3788 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3790 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3791 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3792 the function were treating the return as a boolean value, which happened to
3793 work because 0=false and not-0=true, but is not correct code.
3795 68. The host_aton() function was not handling scoped IPv6 addresses (those
3796 with, for example, "%eth0" on the end) correctly.
3798 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3799 negated items (that is, ~something) in unsigned ints. Some compilers
3800 apparently mutter when there is no cast.
3802 70. If an address verification called from an ACL failed, and did not produce a
3803 user-specific message (i.e. there was only a "system" message), nothing was
3804 put in $acl_verify_message. In this situation, it now puts the system
3807 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3808 session; change 4.31/43 added the unwanted input to the log line - except
3809 that it did not do this in the start of session case. It now does.
3811 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3812 This is wrong and can cause the other end to generate a synchronization
3813 error if it is another Exim or anything else that does the synchronization
3814 check. A QUIT command is no longer sent after a timeout.
3816 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3817 during host lookups.
3819 74. The defer_ok option of callout verification was not working if it was used
3820 when verifying addresses in header lines, that is, for this case:
3822 verify = header_sender/callout=defer_ok
3824 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3825 those file descriptors could be used for SMTP connections. If anything
3826 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3827 could be sent to the SMTP client, causing chaos. The daemon now opens
3828 stdin, stdout, and stderr to /dev/null when it puts itself into the
3831 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3832 log by default. The user can override this with suitable Perl magic.
3834 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3835 the log message when discard triggers, was not working for the DATA ACL or
3836 for the non-SMTP ACL.
3838 78. Error message wording change in sieve.c.
3840 79. If smtp_accept_max_per_host was set, the number of connections could be
3841 restricted to fewer than expected, because the daemon was trying to set up
3842 a new connection before checking whether the processes handling previous
3843 connections had finished. The check for completed processes is now done
3844 earlier. On busy systems, this bug wouldn't be noticed because something
3845 else would have woken the daemon, and it would have reaped the completed
3848 80. If a message was submitted locally by a user whose login name contained one
3849 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3850 It caused a spool format error. I have fixed the spool reading code. A
3851 related problem was that the "from" clause in the Received: line became
3852 illegal because of the space(s). It is now covered by ${quote_local_part.
3854 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3857 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3859 83. Patch from Sieve maintainer to fix unterminated string problem in
3860 "vacation" handling.
3862 84. Some minor changes to the Linux configuration files to help with other
3863 OS variants using glibc.
3865 85. One more patch for Sieve to update vacation handling to latest spec.
3868 ----------------------------------------------------
3869 See the note above about the 4.44 and 4.50 releases.
3870 ----------------------------------------------------
3876 1. Change 4.43/35 introduced a bug that caused file counts to be
3877 incorrectly computed when quota_filecount was set in an appendfile
3880 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3881 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3884 3. Give more explanation in the error message when the command for a transport
3885 filter fails to execute.
3887 4. There are several places where Exim runs a non-Exim command in a
3888 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3889 was being done only for the command run by the queryprogram router. It is
3890 now done for all such subprocesses. The other cases are: ${run, transport
3891 filters, and the commands run by the lmtp and pipe transports.
3893 5. Some older OS have a limit of 256 on the maximum number of file
3894 descriptors. Exim was using setrlimit() to set 1000 as a large value
3895 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3896 systems. I've change it so that if it can't get 1000, it tries for 256.
3898 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3899 was an oversight, and furthermore, ever since the addition of extra
3900 controls (e.g. 4.43/32), the checks on when to allow different forms of
3901 "control" were broken. There should now be diagnostics for all cases when a
3902 control that does not make sense is encountered.
3904 7. $recipients is now available in the predata ACL (oversight).
3906 8. Tidy the search cache before the fork to do a delivery from a message
3907 received from the command line. Otherwise the child will trigger a lookup
3908 failure and thereby defer the delivery if it tries to use (for example) a
3909 cached ldap connection that the parent has called unbind on.
3911 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3912 of $address_data from the recipient verification was clobbered by the
3913 sender verification.
3915 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3916 was its contents. (It was OK if the option was not defined at all.)
3918 11. A "Completed" log line is now written for messages that are removed from
3919 the spool by the -Mrm option.
3921 12. $host_address is now set to the target address during the checking of
3922 ignore_target_hosts.
3924 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3925 being passed; this would have caused $sender_host_name to have been used if
3926 matching the list had actually called for a host name (not very likely,
3927 since this list is usually IP addresses). A host name is now passed as
3930 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3931 code that specifies a non-restarting handler (typically sigaction() in
3932 modern systems) in an attempt to fix a rare and obscure crash bug.
3934 15. Narrowed the window for a race in the daemon that could cause it to ignore
3935 SIGCHLD signals. This is not a major problem, because they are used only to
3936 wake it up if nothing else does.
3938 16. A malformed maildirsize file could cause Exim to calculate negative values
3939 for the mailbox size or file count. Odd effects could occur as a result.
3940 The maildirsize information is now recalculated if the size or filecount
3943 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3944 support for a long time. Removed HAVE_SYS_VFS_H.
3946 18. Updated exipick to current release from John Jetmore.
3948 19. Allow an empty sender to be matched against a lookup in an address list.
3949 Previously the only cases considered were a regular expression, or an
3952 20. Exim went into a mad DNS lookup loop when doing a callout where the
3953 host was specified on the transport, if the DNS lookup yielded more than
3956 21. The RFC2047 encoding function was originally intended for short strings
3957 such as real names; it was not keeping to the 75-character limit for
3958 encoded words that the RFC imposes. It now respects the limit, and
3959 generates multiple encoded words if necessary. To be on the safe side, I
3960 have increased the buffer size for the ${rfc2047: expansion operator from
3963 22. Failure to deliver a bounce message always caused it to be frozen, even if
3964 there was an errors_to setting on the router. The errors_to setting is now
3967 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3968 canonical form (fully expanded) before being placed in
3969 $sender_host_address.
3971 24. Updated eximstats to version 1.33
3973 25. Include certificate and key file names in error message when GnuTLS fails
3974 to set them up, because the GnuTLS error message doesn't include the name
3975 of the failing file when there is a problem reading it.
3977 26. Expand error message when OpenSSL has problems setting up cert/key files.
3980 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3981 (this can affect the format of dates).
3983 28. exim_tidydb, when checking for the continued existence of a message for
3984 which it has found a message-specific retry record, was not finding
3985 messages that were in split spool directories. Consequently, it was
3986 deleting retry records that should have stayed in existence.
3988 29. eximstats updated to version 1.35
3989 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3990 1.35 - bugfix such that pie charts by volume are generated correctly
3992 30. The SPA authentication driver was not abandoning authentication and moving
3993 on to the next authenticator when an expansion was forced to fail,
3994 contradicting the general specification for all authenticators. Instead it
3995 was generating a temporary error. It now behaves as specified.
3997 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3998 (the order specifies the preference for clients). The order is now AES256,
3999 AES128, 3DES, ARCFOUR128.
4001 31. Small patch to Sieve code - explicitly set From: when generating an
4004 32. Exim crashed if a remote delivery caused a very long error message to be
4005 recorded - for instance if somebody sent an entire SpamAssassin report back
4006 as a large number of 550 error lines. This bug was coincidentally fixed by
4007 increasing the size of one of Exim's internal buffers (big_buffer) that
4008 happened as part of the Exiscan merge. However, to be on the safe side, I
4009 have made the code more robust (and fixed the comments that describe what
4012 33. Some experimental protocols are using DNS PTR records for new purposes. The
4013 keys for these records are domain names, not reversed IP addresses. The
4014 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4015 leaves it alone. Component reversal etc. now happens only for IP addresses.
4018 34. The host_aton() function is supposed to be passed a string that is known
4019 to be a valid IP address. However, in the case of IPv6 addresses, it was
4020 not checking this. This is a hostage to fortune. Exim now panics and dies
4021 if the condition is not met. A case was found where this could be provoked
4022 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4023 components; fortuitously, this particular loophole had already been fixed
4024 by change 4.50/55 or 4.44/33 above.
4026 If there are any other similar loopholes, the new check in host_aton()
4027 itself should stop them being exploited. The report I received stated that
4028 data on the command line could provoke the exploit when Exim was running as
4029 exim, but did not say which command line option was involved. All I could
4030 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4031 running as the user.
4034 35. There was a buffer overflow vulnerability in the SPA authentication code
4035 (which came originally from the Samba project). I have added a test to the
4036 spa_base64_to_bits() function which I hope fixes it.
4039 36. The daemon start-up calls getloadavg() while still root for those OS that
4040 need the first call to be done as root, but it missed one case: when
4041 deliver_queue_load_max is set with deliver_drop_privilege. This is
4042 necessary for the benefit of the queue runner, because there is no re-exec
4043 when deliver_drop_privilege is set.
4045 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4046 was in use, was not putting the data itself into the right store pool;
4047 consequently, it could be overwritten for a subsequent message in the same
4048 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4051 38. Sometimes the final signoff response after QUIT could fail to get
4052 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4053 < 0 before doing a fflush(). This bug looks as though it goes back to the
4054 introduction of TLS in release 3.20, but "sometimes" must have been rare
4055 because the tests only now provoked it.
4061 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4062 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4063 Now it returns an integer. A number of calls were still expecting a T/F
4064 return. Fortuitously, in all cases, the tests worked in OK situations,
4065 which is the norm. However, things would have gone wrong on any write
4066 failures on the smtp file descriptor. This function is used when sending
4067 messages over SMTP and also when doing verify callouts.
4069 2. When Exim is called to do synchronous delivery of a locally submitted
4070 message (the -odf or -odi options), it no longer closes stderr before doing
4073 3. Implemented the mua_wrapper option.
4075 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4077 5. Implemented the functions header_remove(), header_testname(),
4078 header_add_at_position(), and receive_remove_recipient(), and exported them
4081 6. If an ACL "warn" statement specified the addition of headers, Exim already
4082 inserted X-ACL-Warn: at the start if there was no header name. However, it
4083 was not making this test for the second and subsequent header lines if
4084 there were newlines in the string. This meant that an invalid header could
4085 be inserted if Exim was badly configured.
4087 7. Allow an ACL "warn" statement to add header lines at the start or after all
4088 the Received: headers, as well as at the end.
4090 8. Added the rcpt_4xx retry error code.
4092 9. Added postmaster_mailfrom=xxx to callout verification option.
4094 10. Added mailfrom=xxxx to the callout verification option, for verify=
4097 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4098 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4100 12. Inserted some casts to stop certain compilers warning when using pointer
4101 differences as field lengths or precisions in printf-type calls (mostly
4102 affecting debugging statements).
4104 13. Added optional readline() support for -be (dynamically loaded).
4106 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4107 same clock tick as a message's arrival, so that its received time was the
4108 same as the "first fail" time on the retry record, and that message
4109 remained on the queue past the ultimate address timeout, every queue runner
4110 would try a delivery (because it was past the ultimate address timeout) but
4111 after another failure, the ultimate address timeout, which should have then
4112 bounced the address, did not kick in. This was a "< instead of <=" error;
4113 in most cases the first failure would have been in the next clock tick
4114 after the received time, and all would be well.
4116 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4117 being recognized when the domain list was tested by the match_domain
4118 condition in an expansion string.
4120 16. Added the ${str2b64: operator.
4122 17. Exim was always calling setrlimit() to set a large limit for the number of
4123 processes, without checking whether the existing limit was already
4124 adequate. (It did check for the limit on file descriptors.) Furthermore,
4125 errors from getrlimit() and setrlimit() were being ignored. Now they are
4126 logged to the main and panic logs, but Exim does carry on, to try to do its
4127 job under whatever limits there are.
4129 18. Imported PCRE 5.0.
4131 19. Trivial typo in log message " temporarily refused connection" (the leading
4134 20. If the log selector return_path_on_delivery was set and an address was
4135 redirected to /dev/null, the delivery process crashed because it assumed
4136 that a return path would always be set for a "successful" delivery. In this
4137 case, the whole delivery is bypassed as an optimization, and therefore no
4140 21. Internal re-arrangement: the function for sending a challenge and reading
4141 a response while authentication was assuming a zero-terminated challenge
4142 string. It's now changed to take a pointer and a length, to allow for
4143 binary data in such strings.
4145 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4147 23. Exim was not respecting finduser_retries when seeking the login of the
4148 uid under which it was called; it was always trying 10 times. (The default
4149 setting of finduser_retries is zero.) Also, it was sleeping after the final
4150 failure, which is pointless.
4152 24. Implemented tls_on_connect_ports.
4154 25. Implemented acl_smtp_predata.
4156 26. If the domain in control=submission is set empty, Exim assumes that the
4157 authenticated id is a complete email address when it generates From: or
4158 Sender: header lines.
4160 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4161 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4162 chown and chgrp in /bin and hostname in /usr/bin.
4164 28. Exim was keeping the "process log" file open after each use, just as it
4165 does for the main log. This opens the possibility of it remaining open for
4166 long periods when the USR1 signal hits a daemon. Occasional processlog
4167 errors were reported, that could have been caused by this. Anyway, it seems
4168 much more sensible not to leave this file open at all, so that is what now
4171 29. The long-running daemon process does not normally write to the log once it
4172 has entered its main loop, and it closes the log before doing so. This is
4173 so that log files can straightforwardly be renamed and moved. However,
4174 there are a couple of unusual error situations where the daemon does write
4175 log entries, and I had neglected to close the log afterwards.
4177 30. The text of an SMTP error response that was received during a remote
4178 delivery was being truncated at 512 bytes. This is too short for some of
4179 the long messages that one sometimes sees. I've increased the limit to
4182 31. It is now possible to make retry rules that apply only when a message has a
4183 specific sender, in particular, an empty sender.
4185 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4186 it possible to be selective about when SMTP synchronization is enforced.
4188 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4190 32. Implemented hosts_connection_nolog.
4192 33. Added an ACL for QUIT.
4194 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4197 35. Added mailbox_size and mailbox_filecount to appendfile.
4199 36. Added control = no_multiline_responses to ACLs.
4201 37. There was a bug in the logic of the code that waits for the clock to tick
4202 in the case where the clock went backwards by a substantial amount such
4203 that the microsecond fraction of "now" was more than the microsecond
4204 fraction of "then" (but the whole seconds number was less).
4206 38. Added support for the libradius Radius client library this is found on
4207 FreeBSD (previously only the radiusclient library was supported).
4213 1. When certain lookups returned multiple values in the form name=value, the
4214 quoting of the values was not always being done properly. Specifically:
4215 (a) If the value started with a double quote, but contained no whitespace,
4217 (b) If the value contained whitespace other than a space character (i.e.
4218 tabs or newlines or carriage returns) it was not quoted.
4219 This fix has been applied to the mysql and pgsql lookups by writing a
4220 separate quoting function and calling it from the lookup code. The fix
4221 should probably also be applied to nisplus, ibase and oracle lookups, but
4222 since I cannot test any of those, I have not disturbed their existing code.
4224 2. A hit in the callout cache for a specific address caused a log line with no
4225 reason for rejecting RCPT. Now it says "Previous (cached) callout
4226 verification failure".
4228 3. There was an off-by-one bug in the queryprogram router. An over-long
4229 return line was truncated at 256 instead of 255 characters, thereby
4230 overflowing its buffer with the terminating zero. As well as fixing this, I
4231 have increased the buffer size to 1024 (and made a note to document this).
4233 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4234 when Exim is waiting for an SMTP response from a remote server, Exim
4235 restarts its select() call on the socket, thereby resetting its timeout.
4236 This is not a problem when such interrupts are rare. Somebody set up a cron
4237 job to run exiwhat every 2 minutes, which is less than the normal select()
4238 timeout (5 or 10 minutes). This meant that the select() timeout never
4239 kicked in because it was always reset. I have fixed this by comparing the
4240 time when an interrupt arrives with the time at the start of the first call
4241 to select(). If more time than the timeout has elapsed, the interrupt is
4242 treated as a timeout.
4244 5. Some internal re-factoring in preparation for the addition of Sieve
4245 extensions (by MH). In particular, the "personal" test is moved to a
4246 separate function, and given an option for scanning Cc: and Bcc: (which is
4247 not set for Exim filters).
4249 6. When Exim created an email address using the login of the caller as the
4250 local part (e.g. when creating a From: or Sender: header line), it was not
4251 quoting the local part when it contained special characters such as @.
4253 7. Installed new OpenBSD configuration files.
4255 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4256 try to make them clearer.
4258 9. Callout options, other than the timeout value, were being ignored when
4259 verifying sender addresses in header lines. For example, when using
4261 verify = header_sender/callout=no_cache
4263 the cache was (incorrectly) being used.
4265 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4266 only the Cygwin environment.
4268 11. When return_path_on_delivery was set as a log selector, if different remote
4269 addresses in the same message used different return paths and parallel
4270 remote delivery occurred, the wrong values would sometimes be logged.
4271 (Whenever a remote delivery process finished, the return path value from
4272 the most recently started remote delivery process was logged.)
4274 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4275 lines when AUTH and/or TLS are in use. This is the "received protocol"
4276 field. Exim used to use "asmtp" for authenticated SMTP, without any
4277 indication (in the protocol name) for TLS use. Now it follows the RFC and
4278 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4279 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4280 names appear in log lines as well as in Received: header lines.
4282 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4283 extensions, and comparison tests, and to fix some bugs.
4285 14. Changes to the "personal" filter test:
4287 (1) The test was buggy in that it was just doing the equivalent of
4288 "contains" tests on header lines. For example, if a user's address was
4289 anne@some.where, the "personal" test would incorrectly be true for
4291 To: susanne@some.where
4293 This test is now done by extracting each address from the header in turn,
4294 and checking the entire address. Other tests that are part of "personal"
4295 are now done using regular expressions (for example, to check local parts
4296 of addresses in From: header lines).
4298 (2) The list of non-personal local parts in From: addresses has been
4299 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4300 taken from the Sieve specification recommendations.
4302 (3) If the message contains any header line starting with "List-" it is
4303 treated as non-personal.
4305 (4) The test for "circular" in the Subject: header line has been removed
4306 because it now seems ill-conceived.
4308 15. Minor typos in src/EDITME comments corrected.
4310 16. Installed latest exipick from John Jetmore.
4312 17. If headers_add on a router specified a text string that was too long for
4313 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4314 of string_sprintf() is now avoided.
4316 18. $message_body_size was not set (it was always zero) when running the DATA
4317 ACL and the local_scan() function.
4319 19. For the "mail" command in an Exim filter, no default was being set for
4320 the once_repeat time, causing a random time value to be used if "once" was
4321 specified. (If the value happened to be <= 0, no repeat happened.) The
4322 default is now 0s, meaning "never repeat". The "vacation" command was OK
4323 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4324 (I found it when inspecting the code).
4326 20. There is now an overall timeout for performing a callout verification. It
4327 defaults to 4 times the callout timeout, which applies to individual SMTP
4328 commands during the callout. The overall timeout applies when there is more
4329 than one host that can be tried. The timeout is checked before trying the
4330 next host. This prevents very long delays if there are a large number of
4331 hosts and all are timing out (e.g. when the network connections are timing
4332 out). The value of the overall timeout can be changed by specifying an
4333 additional sub-option for "callout", called "maxwait". For example:
4335 verify = sender/callout=5s,maxwait=20s
4337 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4338 to the end before writing, but this should make it even safer).
4340 22. Exim was forgetting that it had advertised PIPELINING for the second and
4341 subsequent messages on an SMTP connection. It was also not resetting its
4342 memory on STARTTLS and an internal HELO.
4344 23. When Exim logs an SMTP synchronization error within a session, it now
4345 records whether PIPELINING has been advertised or not.
4347 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4348 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4349 rather than long int.
4351 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4353 26. Added the never_mail option to autoreply.
4359 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4360 crash if the getsockname() call failed; this can happen if a connection is
4361 closed very soon after it is established. The problem was simply in the
4362 order in which certain operations were done, causing Exim to try to write
4363 to the SMTP stream before it had set up the file descriptor. The bug has
4364 been fixed by making things happen in the correct order.
4370 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4371 before the connection was closed, thus losing the rejection response.
4373 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4374 some early Solaris releases, but causes trouble in current releases where
4375 socklen_t is defined.
4377 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4380 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4383 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4384 that releasing the top of it at the end releases what was used for sub-
4385 expansions (unless the block got too big). However, discard this block if
4386 the first thing is a variable or header, so that we can use its block when
4387 it is dynamic (useful for very large $message_headers, for example).
4389 6. Lookups now cache *every* query, not just the most recent. A new, separate
4390 store pool is used for this. It can be recovered when all lookup caches are
4391 flushed. Lookups now release memory at the end of their result strings.
4392 This has involved some general refactoring of the lookup sources.
4394 7. Some code has been added to the store_xxx() functions to reduce the amount
4395 of flapping under certain conditions.
4397 8. log_incoming_interface used to affect only the <= reception log lines. Now
4398 it causes the local interface and port to be added to several more SMTP log
4399 lines, for example "SMTP connection from", and rejection lines.
4401 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4403 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4405 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4406 could be overwritten at the end of the current message (or the start of a
4407 new message if it was set in a HELO ACL). The value is now preserved for
4408 the duration of the SMTP connection.
4410 12. If a transport had a headers_rewrite setting, and a matching header line
4411 contained an unqualified address, that address was qualified, even if it
4412 did not match any rewriting rules. The underlying bug was that the values
4413 of the flags that permit the existence of unqualified sender and recipient
4414 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4415 non-local messages, and by -bnq for local messages) were not being
4416 preserved with the message after it was received.
4418 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4419 "next input=" as part of the text comprising the host identity instead of
4420 the correct text. The code was using the same buffer for two different
4421 strings. However, depending on which order the printing function evaluated
4422 its arguments, the bug did not always show up. Under Linux, for example, my
4423 test suite worked just fine.
4425 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4426 doesn't work with some older versions of Perl. It has been changed to "my",
4427 which in any case is probably the better facility to use.
4429 15. A really picky compiler found some instances of statements for creating
4430 error messages that either had too many or two few arguments for the format
4433 16. The size of the buffer for calls to the DNS resolver has been increased
4434 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4435 for addresses that have a lot of PTR records. This alleviates a problem; it
4436 does not fully solve it.
4438 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4439 buffer now truncates the list and logs the incident, which is the same
4440 action as happens when Exim is looking up a host name and its aliases.
4441 Previously in this situation something unpredictable would happen;
4442 sometimes it was "internal error: store_reset failed".
4444 18. If a server dropped the connection unexpectedly when an Exim client was
4445 using GnuTLS and trying to read a response, the client delivery process
4446 crashed while trying to generate an error log message.
4448 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4449 string, for example:
4451 warn message = H1: something\nH2: something
4453 the text was added as a single header line from Exim's point of view
4454 though it ended up OK in the delivered message. However, searching for the
4455 second and subsequent header lines using $h_h2: did not work. This has been
4456 fixed. Similarly, if a system filter added multiple headers in this way,
4457 the routers could not see them.
4459 20. Expanded the error message when iplsearch is called with an invalid key to
4460 suggest using net-iplsearch in a host list.
4462 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4463 ACLs are no longer actually imposed (and a message to that effect is
4466 22. If a "gecos" field in a passwd entry contained escaped characters, in
4467 particular, if it contained a \" sequence, Exim got it wrong when building
4468 a From: or a Sender: header from that name. A second bug also caused
4469 incorrect handling when an unquoted " was present following a character
4470 that needed quoting.
4472 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4473 was not being matched caselessly.
4475 24. Arranged for all hyphens in the exim.8 source to be escaped with
4478 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4479 properly. Recipient callout cache records were still being keyed to include
4480 the sender, even when use_sender was set false. This led to far more
4481 callouts that were necessary. The sender is no longer included in the key
4482 when use_sender is false.
4484 26. Added "control = submission" modifier to ACLs.
4486 27. Added the ${base62d: operator to decode base 62 numbers.
4488 28. dnsdb lookups can now access SRV records.
4490 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4491 the configuration file.
4493 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4494 (-v) mode. This makes the output for a verbose queue run more intelligible.
4496 31. Added a use_postmaster feature to recipient callouts.
4498 32. Added the $body_zerocount variable, containing the number of binary zero
4499 bytes in the message body.
4501 33. The time of last modification of the "new" subdirectory is now used as the
4502 "mailbox time last read" when there is a quota error for a maildir
4505 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4507 35. Added +ignore_unknown as a special item in host lists.
4509 36. Code for decoding IPv6 addresses in host lists is now included, even if
4510 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4511 address was recognized as an IP address, but was then not correctly decoded
4512 into binary, causing unexpected and incorrect effects when compared with
4519 1. Very minor rewording of debugging text in manualroute to say "list of
4520 hosts" instead of "hostlist".
4522 2. If verify=header_syntax was set, and a header line with an unqualified
4523 address (no domain) and a large number of spaces between the end of the
4524 name and the colon was received, the reception process suffered a buffer
4525 overflow, and (when I tested it) crashed. This was caused by some obsolete
4526 code that should have been removed. The fix is to remove it!
4528 3. When running in the test harness, delay a bit after writing a bounce
4529 message to get a bit more predictability in the log output.
4531 4. Added a call to search_tidyup() just before forking a reception process. In
4532 theory, someone could use a lookup in the expansion of smtp_accept_max_
4533 per_host which, without the tidyup, could leave open a database connection.
4535 5. Added the variables $recipient_data and $sender_data which get set from a
4536 lookup success in an ACL "recipients" or "senders" condition, or a router
4537 "senders" option, similar to $domain_data and $local_part_data.
4539 6. Moved the writing of debug_print from before to after the "senders" test
4542 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4543 problems for message scanning, either using a data ACL, or using
4544 local_scan() because the Received: header was not generated till after they
4545 were called (in order to set the time as the time of reception completion).
4546 I have revised the way this works. The header is now generated after the
4547 body is received, but before the ACL or local_scan() are called. After they
4548 are run, the timestamp in the header is updated.
4554 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4555 before starting a queue runner without re-exec. This happened only when
4556 deliver_drop_privilege was set or when the Exim user was set to root. The
4557 effect of the bug was that timeouts during subsequent deliveries caused
4558 crashes instead of being properly handled. The handler is now left at its
4559 default (and expected) setting.
4561 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4562 message, again when deliver_drop_privilege is set or Exim is run as root.
4563 The bug described in (1) was not present in this case, but the tidying up
4564 of the other signals was missing. I have made the two cases consistent.
4566 3. The ignore_target_hosts setting on a manualroute router was being ignored
4567 for hosts that were looked up using the /MX notation.
4569 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4572 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4573 operated on the sender address. After changing the $sender_address to <>
4574 for the sender address verify, Exim was re-instated it as the original
4575 (before rewriting) address, but remembering that it had rewritten it, so it
4576 wasn't rewriting it again. This bug also had the effect of breaking the
4577 sender address verification caching when the sender address was rewritten.
4579 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4580 This has been changed so that if the ip literal address matches
4581 ignore_target_hosts, the router declines.
4583 7. Added expansion conditions match_domain, match_address, and match_local_
4584 part (NOT match_host).
4586 8. The placeholder for the Received: header didn't have a length field set.
4588 9. Added code to Exim itself and to exim_lock to test for a specific race
4589 condition that could lead to file corruption when using MBX delivery. The
4590 issue is with the lockfile that is created in /tmp. If this file is removed
4591 after a process has opened it but before that process has acquired a lock,
4592 there is the potential for a second process to recreate the file and also
4593 acquire a lock. This could lead to two Exim processes writing to the file
4594 at the same time. The added code performs the same test as UW imapd; it
4595 checks after acquiring the lock that its file descriptor still refers to
4596 the same named file.
4598 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4599 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4602 11. Added the smtp_active_hostname option. If used, this will typically be made
4603 to depend on the incoming interface address. Because $interface_address is
4604 not set up until the daemon has forked a reception process, error responses
4605 that can happen earlier (such as "too many connections") no longer contain
4608 12. If an expansion in a condition on a "warn" statement fails because a lookup
4609 defers, the "warn" statement is abandoned, and the next ACL statement is
4610 processed. Previously this caused the whole ACL to be aborted.
4612 13. Added the iplsearch lookup type.
4614 14. Added ident_timeout as a log selector.
4616 15. Added tls_certificate_verified as a log selector.
4618 16. Added a global option tls_require_ciphers (compare the smtp transport
4619 option of the same name). This controls incoming TLS connections.
4621 17. I finally figured out how to make tls_require_ciphers do a similar thing
4622 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4623 before starting the TLS session.
4625 18. Tabs are now shown as \t in -bP output.
4627 19. If the log selector return_path_on_delivery was set, Exim crashed when
4628 bouncing a message because it had too many Received: header lines.
4630 20. If two routers both had headers_remove settings, and the first one included
4631 a superfluous trailing colon, the final name in the first list and the
4632 first name in the second list were incorrectly joined into one item (with a
4633 colon in the middle).
4639 1. Added -C and -D options to the exinext utility, mainly to make it easier
4640 to include in the automated testing, but these could be helpful when
4641 multiple configurations are in use.
4643 2. The exinext utility was not formatting the output nicely when there was
4644 an alternate port involved in the retry record key, nor when there was a
4645 message id as well (for retries that were specific to a specific message
4646 and a specific host). It was also confused by IPv6 addresses, because of
4647 the additional colons they contain. I have fixed the IPv4 problem, and
4648 patched it up to do a reasonable job for IPv6.
4650 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4651 delivery, the log line now contains "pipelined" if PIPELINING was used.
4653 4. An SMTP transport process used to panic and die if the bind() call to set
4654 an explicit outgoing interface failed. This has been changed; it is now
4655 treated in the same way as a connect() failure.
4657 5. A reference to $sender_host_name in the part of a conditional expansion
4658 that was being skipped was still causing a DNS lookup. This no longer
4661 6. The def: expansion condition was not recognizing references to header lines
4662 that used bh_ and bheader_.
4664 7. Added the _cache feature to named lists.
4666 8. The code for checking quota_filecount in the appendfile transport was
4667 allowing one more file than it should have been.
4669 9. For compatibility with Sendmail, the command line option
4677 and sets the incoming protocol and host name (for trusted callers). The
4678 host name and its colon can be omitted when only the protocol is to be set.
4679 Note the Exim already has two private options, -pd and -ps, that refer to
4680 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4681 "s", but I don't think that's a major issue.
4683 10. A number of refactoring changes to the code, none of which should affect
4686 (a) The number of logging options was getting close to filling up the
4687 32-bit word that was used as a bit map. I have split them into two classes:
4688 those that are passed in the argument to log_write(), and those that are
4689 only ever tested independently outside of that function. These are now in
4690 separate 32-bit words, so there is plenty of room for expansion again.
4691 There is no change in the user interface or the logging behaviour.
4693 (b) When building, for example, log lines, the code previously used a
4694 macro that called string_cat() twice, in order to add two strings. This is
4695 not really sufficiently general. Furthermore, there was one instance where
4696 it was actually wrong because one of the argument was used twice, and in
4697 one call a function was used. (As it happened, calling the function twice
4698 did not affect the overall behaviour.) The macro has been replaced by a
4699 function that can join an arbitrary number of extra strings onto a growing
4702 (c) The code for expansion conditions now uses a table and a binary chop
4703 instead of a serial search (which was left over from when there were very
4704 few conditions). Also, it now recognizes conditions like "pam" even when
4705 the relevant support is not compiled in: a suitably worded error message is
4706 given if an attempt is made to use such a condition.
4708 11. Added ${time_interval:xxxxx}.
4710 12. A bug was causing one of the ddress fields not to be passed back correctly
4711 from remote delivery subprocesses. The field in question was not being
4712 subsequently used, so this caused to problems in practice.
4714 13. Added new log selectors queue_time and deliver_time.
4716 14. Might have fixed a bug in maildirsizefile handling that threw up
4717 "unexpected character" debug warnings, and recalculated the data
4718 unnecessarily. In any case, I expanded the warning message to give more
4721 15. Added the message "Restricted characters in address" to the statements in
4722 the default ACL that block characters like @ and % in local parts.
4724 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4725 Three changes have been made:
4727 (a) There was a serious bug; a negative response to MAIL caused the whole
4728 recipient domain to be cached as invalid, thereby blocking all messages
4729 to all local parts at the same domain, from all senders. This bug has
4730 been fixed. The domain is no longer cached after a negative response to
4731 MAIL if the sender used is not empty.
4733 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4736 (c) A new callout option, "use_sender" has been added for people who want
4737 the modified behaviour.
4743 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4746 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4747 indeed breaks things for older releases.
4749 3. Added additional logging to the case where there is a problem reading data
4750 from a filter that is running in a subprocess using a pipe, in order to
4751 try to track down a specific problem.
4753 4. Testing facility fudge: when running in the test harness and attempting
4754 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4755 getting "No route to host". Convert this to a timeout.
4757 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4760 6. Some OS don't have socklen_t but use size_t instead. This affects the
4761 fifth argument of getsockopt() amongst other things. This is now
4762 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4763 can be set for individual OS. I have set it for SunOS5, OSF1, and
4764 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4765 some earlier ones do not.
4767 7. Change 4.30/15 was not doing the test caselessly.
4769 8. The standard form for an IPv6 address literal was being rejected by address
4770 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4771 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4772 this, as well as the form without the "IPv6" on the front (but only when
4773 address literals are enabled, of course).
4775 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4777 10. Exim crashed if a message with an empty sender address specified by -f
4778 encountered a router with an errors_to setting. This could be provoked only
4779 by a command such as
4783 where an empty string was supplied; "<>" did not hit this bug.
4785 11. Installed PCRE release 4.5.
4787 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4788 remained set. It is now erased.
4790 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4791 times from message ids (which are base 36 rather than the normal 62).
4793 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4794 were being counted as actual protocol errors, and logged if the log
4795 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4796 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4797 and DATA following a set of rejected RCPTs do not count as protocol errors.
4798 In other words, Exim assumes they were pipelined, though this may not
4799 actually be the case. Of course, in all cases the client gets an
4800 appropriate error code.
4802 15. If a lookup fails in an ACL condition, a message about the failure may
4803 be available; it is used if testing the ACL cannot continue, because most
4804 such messages specify what the cause of the deferral is. However, some
4805 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4806 that caused an old message to be retained and used if a later statement
4807 caused a defer, replacing the real cause of the deferral.
4809 16. If an IP address had so many PTR records that the DNS lookup buffer
4810 was not large enough to hold them, Exim could crash while trying to process
4811 the truncated data. It now detects and logs this case.
4813 17. Further to 4.21/58, another change has been made: if (and only if) the
4814 first line of a message (the first header line) ends with CRLF, a bare LF
4815 in a subsequent header line has a space inserted after it, so as not to
4816 terminate the header.
4818 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4819 unnecessarily, used atoi() instead of strtol(), and didn't check the
4820 termination when getting file sizes from file names by regex.
4822 19. Completely re-implemented the support for maildirsize files, in the light
4823 of a number of problems with the previous contributed implementation
4824 (4.30/29). In particular:
4826 . If the quota is zero, the maildirsize file is maintained, but no quota is
4829 . If the maildir directory does not exist, it is created before any attempt
4830 to write a maildirsize file.
4832 . The quota value in the file is just a cache; if the quota is changed in
4833 the transport, the new value overrides.
4835 . A regular expression is available for excluding directories from the
4838 20. The autoreply transport checks the characters in options that define the
4839 message's headers; it allows continued headers, but it was checking with
4840 isspace() after an embedded newline instead of explicitly looking for a
4843 21. If all the "regular" hosts to which an address was routed had passed their
4844 expiry times, and had not reached their retry times, the address was
4845 bounced, even if fallback hosts were defined. Now Exim should go on to try
4848 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4849 equivalent code in the SMTP transport. Some hosts send humungous responses
4850 to HELO/EHLO, more than 1024 it seems.
4852 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4853 gives compiler warnings in some environments. I've now done it "properly",
4856 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4857 (because of problems with the built-in one) was declared to return uschar *
4858 instead of char *, causing compiler failure.
4860 25. Fixed a file descriptor leak when processing alias/forward files.
4862 26. Fixed a minor format string issue in dbfn.c.
4864 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4866 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4867 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4870 29. When a sender address is verified, it is cached, to save repeating the test
4871 when there is more than one recipient in a message. However, when the
4872 verification involves a callout, it is possible for different callout
4873 options to be set for different recipients. It is too complicated to keep
4874 track of this in the cache, so now Exim always runs a verification when a
4875 callout is required, relying on the callout cache for the optimization.
4876 The overhead is duplication of the address routing, but this should not be
4879 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4880 to be verified with callout=postmaster, and the main callout worked but the
4881 postmaster check failed, the verification correctly failed. However, if a
4882 subsequent RCPT command asked for sender verification *without* the
4883 postmaster check, incorrect caching caused this verification also to fail,
4886 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4887 it was not caching the DNS options (qualify_single, search_parents) that
4888 were used when the lookup failed. A subsequent lookup with different
4889 options therefore always gave the same answer, though there were cases
4890 where it should not have. (Example: a "domains = !$mx_any" option on a
4891 dnslookup router: the "domains" option is always processed without any
4892 widening, but the router might have qualify_single set.) Now Exim uses the
4893 cached value only when the same options are set.
4895 32. Added John Jetmore's "exipick" utility to the distribution.
4897 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4898 than a timeout (e.g. a certificate is required, and is not provided), an
4899 Exim server now closes the connection immediately. Previously it waited for
4900 the client to close - but if the client is SSL, it seems that they each
4901 wait for each other, leading to a delay before one of them times out.
4903 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4904 maintained 0.8.x compatibility because I don't think many are using it, and
4905 it is clearly obsolete.
4907 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4910 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4911 client certificate was expired. A simple patch fixes this, though I don't
4912 understand the full logic of why the verify callback is called multiple
4915 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4916 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4917 which causes problems with some clients (such as the Certicom SSL Plus
4918 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4919 disables the coutermeasure allowing Eudora to connect."
4921 38. Exim was not checking that a write() to a log file succeeded. This could
4922 lead to Bad Things if a log got too big, in particular if it hit a file
4923 size limit. Exim now panics and dies if it cannot write to a log file, just
4924 as it does if it cannot open a log file.
4926 39. Modified OS/Makefile-Linux so that it now contains
4928 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4930 The two -D definitions ensure that Exim is compiled with large file
4931 support, which makes it possible to handle log files that are bigger than
4934 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4935 instance) a domain was checked against a named list that involved a lookup,
4936 causing $domain_data to be set, then another domain was checked against the
4937 same list, then the first domain was re-checked, the value of $domain_data
4938 after the final check could be wrong. In particular, if the second check
4939 failed, it could be set empty. This bug probably also applied to
4942 41. The strip_trailing_dot option was not being applied to the address given
4943 with the -f command-line option.
4945 42. The code for reading a message's header from the spool was incrementing
4946 $received_count, but never initializing it. This meant that the value was
4947 incorrect (doubled) while delivering a message in the same process in which
4948 it was received. In the most common configuration of Exim, this never
4949 happens - a fresh exec is done - but it can happen when
4950 deliver_drop_privilege is set.
4952 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4953 it now includes up to 150 characters of the unexpected data in the log
4956 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4957 and building data strings. The size of both of these buffers was 10 000
4958 bytes - far larger than anybody would *ever* want, thought I. Needless to
4959 say, somebody hit the limit. I have increased the maximum line length to
4960 20 000 and the maximum data length of concatenated lines to 100 000. I have
4961 also fixed two bugs, because there was no checking on these buffers. Tsk,
4962 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4963 buffer is too small.
4965 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4966 lsearch lookups. Now it does.
4968 46. When parsing a route_list item in a manualroute router, a fixed-length
4969 buffer was used for the list of hosts. I made this 1024 bytes long,
4970 thinking that nobody would ever have a list of hosts that long. Wrong.
4971 Somebody had a whole pile of complicated expansion conditions, and the
4972 string was silently truncated, leading to an expansion error. It turns out
4973 that it is easier to change to an unlimited length (owing to other changes
4974 that have happened since this code was originally written) than to build
4975 structure for giving a limitation error. The length of the item that
4976 expands into the list of hosts is now unlimited.
4978 47. The lsearch lookup could not handle data where the length of text line was
4979 more than 4095 characters. Such lines were truncated, leading to shortened
4980 data being returned. It should now handle lines of any length.
4982 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4983 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4986 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4987 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4988 no longer happens in comment lines. A list of replacements is now placed
4989 at the head of all of the source files, except those whose only change is
4990 to replace PERL_COMMAND in the very first #! line.
4992 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4993 messages on the queue, with a bottom-up merge sort, using code contributed
4994 by Michael Haardt. This should make operations like -bp somewhat faster on
4995 large queues. It won't affect queue runners, except when queue_run_in_order
4998 51. Installed eximstats 1.31 in the distribution.
5000 52. Added support for SRV lookups to the dnslookup router.
5002 53. If an ACL referred to $message_body or $message_body_end, the value was not
5003 reset for any messages that followed in the same SMTP session.
5005 54. The store-handling optimization for building very long strings was not
5006 differentiating between the different store pools. I don't think this
5007 actually made any difference in practice, but I've tidied it.
5009 55. While running the routers to verify a sender address, $sender_address
5010 was still set to the sender address. This is wrong, because when routing to
5011 send a bounce to the sender, it would be empty. Therefore, I have changed
5012 it so that, while verifying a sender address, $sender_address is set to <>.
5013 (There is no change to what happens when verifying a recipient address.)
5015 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5016 target A or AAAA records (if not already returned) without resetting the
5017 qualify_single or search_parents options of the DNS resolver. These are
5018 inappropriate in this case because the targets of MX and SRV records must
5019 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5020 target that, when qualified, matched something in the local domain. These
5021 two options are now turned off when doing these lookups.
5023 57. It seems that at least some releases of Reiserfs (which does not have the
5024 concept of a fixed number of inodes) returns zero and not -1 for the
5025 number of available inodes. This interacted badly with check_spool_inodes,
5026 which assumed that -1 was the "no such thing" setting. What I have done is
5027 to check that the total number of inodes is greater than zero before doing
5028 the test of how many are available.
5030 58. When a "warn" ACL statement has a log_message modifier, the message is
5031 remembered, and not repeated. This is to avoid a lot of repetition when a
5032 message has many recipients that cause the same warning to be written.
5033 Howewer, Exim was preserving the list of already written lines for an
5034 entire SMTP session, which doesn't seem right. The memory is now reset if a
5035 new message is started.
5037 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5038 debugging output unless log_rewrite was also set.
5040 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5041 of (int)(handle) when we know that handle contains (void *)(-1).
5043 61. The Exim daemon panic-logs an error return when it closes the incoming
5044 connection. However "connection reset by peer" seems to be common, and
5045 isn't really an error worthy of noting specially, so that particular error
5048 62. When Exim is trying to find all the local interfaces, it used to panic and
5049 die if the ioctl to get the interface flags failed. However, it seems that
5050 on at least one OS (Solaris 9) it is possible to have an interface that is
5051 included in the list of interfaces, but for which you get a failure error
5052 for this call. This happens when the interface is not "plumbed" into a
5053 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5054 failure of the "get flags" call assumes that the interface is down.
5056 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5057 makes life easier for people who are doing arithmetic on fields extracted
5058 from dates, where you often get leading zeros that should not be
5059 interpreted as octal.
5061 64. Added qualify_domain to the redirect router, to override the global
5064 65. If a pathologically long header line contained very many addresses (the
5065 report of this problem mentioned 10 000) and each of them was rewritten,
5066 Exim could use up a very large amount of memory. (It kept on making new
5067 copies of the header line as it rewrote, and never released the old ones.)
5068 At the expense of a bit more processing, the header rewriting function has
5069 been changed so that it no longer eats memory in this way.
5071 66. The generation of the Received: header has been moved from the time that a
5072 message starts to be received, to the time that it finishes. The timestamp
5073 in the Received: header should now be very close to that of the <= log
5074 line. There are two side-effects of this change:
5076 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5077 logged header lines no longer include the local Received: line, because
5078 it has not yet been created. The same applies to a copy of the message
5079 that is returned to a non-SMTP sender when a message is rejected.
5081 (b) When a filter file is tested using -bf, no additional Received: header
5082 is added to the test message. After some thought, I decided that this
5085 This change does not affect the value of $received_for. It is still set
5086 after address rewriting, but before local_scan() is called.
5088 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5090 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5091 gave an unhelpful panic error message, and a defer error. I have managed to
5092 change this behaviour so that it now rejects any supplied certificate,
5093 which seems right, as the list of acceptable certificates is empty.
5095 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5096 gave an unhelpful defer error. I have not managed to make this reject any
5097 supplied certificates, but the error message it gives is "no certificate
5098 supplied", which is not helpful.
5100 70. exigrep's output now also includes lines that are not associated with any
5101 message, but which match the given pattern. Implemented by a patch from
5102 Martin Sluka, which also tidied up the Perl a bit.
5104 71. Recipient callout verification, like sender verification, was using <> in
5105 the MAIL FROM command. This isn't really the right thing, since the actual
5106 sender may affect whether the remote host accepts the recipient or not. I
5107 have changed it to use the actual sender in the callout; this means that
5108 the cache record is now keyed on a recipient/sender pair, not just the
5109 recipient address. There doesn't seem to be a real danger of callout loops,
5110 since a callout by the remote host to check the sender would use <>.
5111 [SEE ABOVE: changed after hitting problems.]
5113 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5114 temporary errors. However, in the case of such a code being given after
5115 the end of a data transmission (i.e. after ".") Exim was failing to write
5116 a retry record for the message. (Yes, there was some broken host that was
5117 actually sending 8xx at this point.)
5119 73. An unknown lookup type in a host list could cause Exim to panic-die when
5120 the list was checked. (An example that provoked this was putting <; in the
5121 middle of a list instead of at the start.) If this happened during a DATA
5122 ACL check, a -D file could be left lying around. This kind of configuration
5123 error no longer causes Exim to die; instead it causes a defer errror. The
5124 incident is still logged to the main and panic logs.
5126 74. Buglet left over from Exim 3 conversion. The message "too many messages
5127 in one connection" was written to the rejectlog but not the mainlog, except
5128 when address rewriting (yes!) was being logged.
5130 75. Added write_rejectlog option.
5132 76. When a system filter was run not as root (that is, when system_filter_user
5133 was set), the values of the $n variables were not being returned to the
5134 main process; thus, they were not subsequently available in the $sn
5137 77. Added +return_path_on_delivery log selector.
5139 78. A connection timeout was being treated differently from recipients deferred
5140 when testing hosts_max_try with a message that was older than the host's
5141 retry timeout. (The host should not be counted, thus allowing all hosts to
5142 be tried at least once before bouncing.) This may have been the cause of an
5143 occasionally reported bug whereby a message would remain on the queue
5144 longer than the retry timeout, but would be bounced if a delivery was
5145 forced. I say "may" because I never totally pinned down the problem;
5146 setting up timeout/retry tests is difficult. See also the next item.
5148 79. The ultimate address timeout was not being applied to errors that involved
5149 a combination of host plus message (for example, a timeout on a MAIL
5150 command). When an address resolved to a number of possible hosts, and they
5151 were not all tried for each delivery (e.g. because of hosts_max_try), a
5152 message could remain on the queue longer than the retry timeout.
5154 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5157 81. Fixed an obscure SMTP outgoing bug which required at least the following
5158 conditions: (a) there was another message waiting for the same server;
5159 (b) the server returned 5xx to all RCPT commands in the first message so
5160 that the message was not completed; (c) the server dropped the connection
5161 or gave a negative response to the RSET that Exim sends to abort the
5162 transaction. The observed case was a dropped connection after DATA that had
5163 been sent in pipelining mode. That is, the server had advertised PIPELINING
5164 but was not implementing it correctly. The effect of the bug was incorrect
5165 behaviour, such as trying another host, and this could lead to a crash.
5171 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5172 and daemon.c were passed as pointers to ints; they should have been
5173 pointers to socklen_t variables (which are typically unsigned ints).
5175 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5178 3. Fixed a really odd bug that affected only the testing scheme; patching a
5179 certain fixed string in the binary changed the value of another string that
5180 happened to be identical to the end of the original first string.
5182 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5183 name", it returns that address as the IP address. On some operating
5184 systems (e.g. Solaris), it also passes back the IP address string as the
5185 "host name". However, on others (e.g. Linux), it passes back an empty
5186 string. Exim wasn't checking for this, and was changing the host name to an
5187 empty string, assuming it had been canonicized.
5189 5. Although rare, it is permitted to have more than one PTR record for a given
5190 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5191 all the names associated with an address, because they do in Solaris.
5192 However, it seems that they do not in Linux for data that comes from the
5193 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5194 I found this out when I moved to a new Linux workstation and tried to run
5195 the Exim test suite.
5197 To get round this problem I have changed the code so that it now does its
5198 own call to the DNS to look up PTR records when searching for a host name.
5199 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5200 addresses that are only in /etc/hosts are still found.
5202 This behaviour is, however, controlled by an option called host_lookup_
5203 order, which defaults to "bydns:byaddr". If people want to use the other
5204 order, or indeed, just use one or the other means of lookup, they can
5205 specify it in this variable.
5207 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5208 some operating systems, this comes back from gethostbyaddr() as an empty
5209 string, and this is what Exim used to test for. However, it seems that in
5210 other systems, "." is yielded. Exim now tests for this case too.
5212 7. The values of check_spool_space and check_log_space are now held internally
5213 as a number of kilobytes instead of an absolute number of bytes. If a
5214 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5215 kilobyte. This means that much larger values can be stored.
5217 8. Exim monitor: an attempt to get the action menu when not actually pointing
5218 at a message produces an empty menu entitled "No message selected". This
5219 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5220 no entries in it ("Shell widget menu has zero width and/or height"). So I
5221 have added a single, blank menu entry in this case.
5223 9. Added ${quote_local_part.
5225 10. MIME decoding is now applied to the contents of Subject: header lines when
5228 11. Now that a reference to $sender_host_address automatically causes a reverse
5229 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5230 host lookup before query-style lookups in lists that might use this
5231 variable. This has therefore been abolished, and the "net-" prefix is no
5232 longer necessary for query-style lookups.
5234 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5235 have been a typo for LFLAGS, so it has been changed.
5237 13. The install script calls Exim with "-C /dev/null" in order to find the
5238 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5239 to be output. Howeve, since Exim outputs its version number before the
5240 error, it didn't break the script. It just looked ugly. I fixed this by
5241 always allowing "-C /dev/null" if the caller is root.
5243 14. Ignore overlarge ACL variable number when reading spool file - insurance
5244 against a later release with more variables having written the file.
5246 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5247 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5248 this, as well as the form without the "IPv6" on the front.
5250 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5251 OS/Makefile-Darwin file.
5253 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5254 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5256 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5257 variable. However, some OS use ints for inodes. I've added cast to long int
5258 to get rid of the compiler warning.
5260 19. I had forgotten to lock out "/../" in configuration file names when
5261 ALT_CONFIG_PREFIX was set.
5263 20. Routers used for verification do not need to specify transports. However,
5264 if such a router generated a host list, and callout was configured, Exim
5265 crashed, because it could not find a port number from the (non-existent)
5266 transport. It now assumes port 25 in this circumstance.
5268 21. Added the -t option to exigrep.
5270 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5271 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5272 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5273 them is set, LOOKUP_LSEARCH is forced.
5275 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5277 24. Added sender and host information to the "rejected by local_scan()" log
5278 line; previously there was no indication of these.
5280 25. Added .include_if_exists.
5282 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5283 Linux. It turns out that not all file systems support this. Apparently some
5284 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5285 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5286 not supported on the file descriptor, is now ignored when Exim is trying to
5287 sync a directory. This applies only to Linux.
5289 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5291 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5292 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5295 29. Added support for maildirsize files from supplied patch (modified a bit).
5297 30. The use of :fail: followed by an empty string could lead Exim to respond to
5298 sender verification failures with (e.g.):
5300 550 Verification failed for <xxx>
5301 550 Sender verify failed
5303 where the first response line was missing the '-' that indicates it is not
5304 the final line of the response.
5306 31. The loop for finding the name of the user that called Exim had a hardwired
5307 limit of 10; it now uses the value of finduser_retries, which is used for
5308 all other user lookups.
5310 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5313 33. Exim was neglecting to zero errno before one call of strtol() when
5314 expanding a string and expecting an integer value. On some systems this
5315 resulted in spurious "integer overflow" errors. Also, it was casting the
5316 result into an int without checking.
5318 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5319 did not work. The code looks as if it has *never* worked, though it appears
5320 to have been documented since at least releast 1.62. I have made it work.
5322 35. The "timeout_DNS" error in retry rules, also documented since at least
5323 1.62, also never worked. As it isn't clear exactly what this means, and
5324 clearly it isn't a major issue, I have abolished the feature by treating it
5325 as "timeout", and writing a warning to the main and panic logs.
5327 36. The display of retry rules for -brt wasn't always showing the error code
5330 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5331 timeout_connect_A, timeout_connect_MX.
5333 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5334 to the empty sender.
5336 39. The daemon was not analysing the content of -oX till after it had closed
5337 stderr and disconnected from the controlling terminal. This meant that any
5338 syntax errors were only noted on the panic log, and the return code from
5339 the command was 0. By re-arranging the code a little, I've made the
5340 decoding happen first, so such errors now appear on stderr, and the return
5341 code is 1. However, the actual setting up of the sockets still happens in
5342 the disconnected process, so errors there are still only recorded on the
5345 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5346 connections (as happens on some IP stacks) was logged at start up time as
5347 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5348 it from "IPv6 and IPv4", which means that two separate sockets are being
5351 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5352 says whether AF_INET or AF_INET6 was passed as an argument.
5354 42. Exiwhat output was messed up when time zones were included in log
5357 43. Exiwhat now gives more information about the daemon's listening ports,
5358 and whether -tls-on-connect was used.
5360 44. The "port" option of the smtp transport is now expanded.
5362 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5363 silently ignored. Now an error message is written to the main and panic
5366 46. There's a new ACL modifier called "logwrite" which writes to a log file
5367 as soon as it is encountered.
5369 47. Added $local_user_uid and $local_user_gid at routing time.
5371 48. Exim crashed when trying to verify a sender address that was being
5374 49. Exim was recognizing only a space character after ".include". It now also
5375 recognizes a tab character.
5377 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5378 extracting the relevant information from the specification. The man page no
5379 longer contains scrambled data for the -d option, and I've added a section
5380 at the front about calling Exim under different names.
5382 51. Added "extra_headers" argument to the "mail" command in filter files.
5384 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5387 53. Installed eximstats 1.29.
5389 54. Added transport_filter_timeout as a generic transport option.
5391 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5392 Cc: header lines. This was required by RFC 822, but it not required by RFC
5395 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5396 incoming messages that did not have them. Now it does so only if the
5397 message originates locally, that is, if there is no associated remote host
5398 address. When Resent- header lines are present, this applies to the Resent-
5399 lines rather than the non-Resent- lines.
5401 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5402 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5404 58. Messages for configuration errors now include the name of the main
5405 configuration file - useful now that there may be more than one file in a
5406 list (.included file names were always shown).
5408 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5409 for those rare installations that do not start the daemon as root or run it
5410 setuid root. I've cut out the call to initgroups() if the daemon is not
5413 60. The Exim user and group can now be bound into the binary as text strings
5414 that are looked up at the start of Exim's processing.
5416 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5418 62. Added $mailstore_basename variable.
5420 63. Installed patch to sieve.c from Michael Haardt.
5422 64. When Exim failed to open the panic log after failing to open the main log,
5423 the original message it was trying to log was written to stderr and debug
5424 output, but if they were not available (the usual case in production), it
5425 was lost. Now it is written to syslog before the two lines that record the
5426 failures to open the logs.
5428 65. Users' Exim filters run in subprocesses under the user's uid. It is
5429 possible for a "deliver" command or an alias in a "personal" command to
5430 provoke an address rewrite. If logging of address rewriting is configured,
5431 this fails because the process is not running as root or exim. There may be
5432 a better way of dealing with this, but for the moment (because 4.30 needs
5433 to be released), I have disabled address rewrite logging when running a
5434 filter in a non-root, non-exim process.
5440 1. The buildconfig auxiliary program wasn't quoting the value set for
5441 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5442 not defined. This bug was masked in 4.22 by the effect that was fixed in
5445 2. Some messages that were rejected after a message id was allocated were
5446 shown as "incomplete" by exigrep. It no longer does this for messages that
5447 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5449 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5450 have allow_domain_literals set, the ID did not get logged in the <= line.
5451 Domain literals are now always recognized in Message-ID: header lines.
5453 4. The first argument for a ${extract expansion item is the key name or field
5454 number. Leading and trailing spaces in this item were not being ignored,
5455 causing some misleading effects.
5457 5. When deliver_drop_privilege was set, single queue runner processes started
5458 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5459 same command in the process it spins off) were not dropping privilege.
5461 6. When the daemon running as "exim" started a queue runner, it always
5462 re-executed Exim in the spun-off process. This is a waste of effort when
5463 deliver_drop_privilege is set. The new process now just calls the
5464 queue-runner function directly.
5470 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5473 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5474 ignored. Though the use of -oP was forcing the writing of a pid file, it
5475 was always written to the default place.
5477 3. If the message "no IP address found for host xxxx" is generated during
5478 incoming verification, it is now followed by identification of the incoming
5479 connection (so you can more easily find what provoked it).
5481 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5483 5. Added some features to "harden" Exim a bit more against certain attacks:
5485 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5486 be put in Local/Makefile. This is like the never_users runtime option,
5487 but it cannot be overridden. The default setting is "root".
5489 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5490 prefix string with which any file named in a -C command line option
5493 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5494 is retained for -C and -D only if the caller of Exim is root. Without
5495 it, the exim user may also use -C and -D and retain privilege.
5497 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5498 command line option is disabled.
5500 6. Macro names set by the -D option must start with an upper case letter, just
5501 like macro names defined in the configuration file.
5503 7. Added "dereference=" facility to LDAP.
5505 8. Two instances of the typo "uknown" in the source files are fixed.
5507 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5508 the Configure-Makefile script screwed up while processing it.
5510 10. Incorporated PCRE 4.4.
5512 11. The SMTP synchronization check was not operating right at the start of an
5513 SMTP session. For example, it could not catch a HELO sent before the client
5514 waited for the greeting. There is now a check for outstanding input at the
5515 point when the greeting is written. Because of the duplex, asynchronous
5516 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5517 way, but not yet received, when the check is performed.
5519 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5520 on TCP/IP sockets, because this apparently causes some broken clients to
5523 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5524 unchanged) from the Cygwin maintainer.
5526 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5527 was supported instead of testing for mbx. Effectively a typo.
5529 15. The spa authenticator server code was not checking that the input it
5530 received was valid base64.
5532 16. The debug output line for the "set" modifier in ACLs was not showing the
5533 name of the variable that was being set.
5535 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5537 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5538 lookup on its own. Something else was needed to trigger the lookup. For
5539 example, a match in host_lookup or the need for a host name in a host list.
5540 Now, if $sender_host_name is referenced and the host name has not yet been
5541 looked up, a lookup is performed. If the lookup fails, the variable remains
5542 empty, and $host_lookup_failed is set to "1".
5544 19. Added "eqi" as a case-independent comparison operator.
5546 20. The saslauthd authentication condition could segfault if neither service
5547 nor realm was specified.
5549 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5550 error message that was logged was misleading, and incoming SMTP
5551 connections were dropped. The message is now more accurate, and temporary
5552 errors are given to SMTP connections.
5554 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5555 (except RSET) with a 421 error, until QUIT is received. However, it was
5556 failing to send a response to QUIT.
5558 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5559 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5560 ACL is now run after the helo_try_verify_hosts code.
5562 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5563 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5564 case-independent, so other case variants are also recognized). Apparently
5565 some systems use these upper case variants.
5567 25. If more than two messages were waiting for the same host, and a transport
5568 filter was specified for the transport, Exim sent two messages over the
5569 same TCP/IP connection, and then failed with "socket operation on non-
5570 socket" when it tried to send the third.
5572 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5574 27. The extern definition of crypt16() in expand.c was not being excluded when
5575 the OS had its own crypt16() function.
5577 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5578 as a preferred synonym for return_size_limit, both as an option and as an
5581 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5583 30. Changed the default configuration ACL to relax the local part checking rule
5584 for addresses that are not in any local domains. For these addresses,
5585 slashes and pipe symbols are allowed within local parts, but the sequence
5586 /../ is explicitly forbidden.
5588 31. SPA server authentication was not clearing the challenge buffer before
5591 32. log_message in a "warn" ACL statement was writing to the reject log as
5592 well as to the main log, which contradicts the documentation and doesn't
5593 seem right (because no rejection is happening). So I have stopped it.
5595 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5596 However, I am unable to do any testing of this.
5598 34. Fixed an infelicity in the appendfile transport. When checking directories
5599 for a mailbox, to see if any needed to be created, it was accidentally
5600 using path names with one or more superfluous leading slashes; tracing
5601 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5603 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5604 contents are added to the log line that is written for every discarded
5605 recipient. (Previously a log_message setting was ignored.)
5607 36. The ${quote: operator now quotes the string if it is empty.
5609 37. The install script runs exim in order to find its version number. If for
5610 some reason other than non-existence or emptiness, which it checks, it
5611 could not run './exim', it was installing it with an empty version number,
5612 i.e. as "exim-". This error state is now caught, and the installation is
5615 38. An argument was missing from the function that creates an error message
5616 when Exim fails to connect to the socket for saslauthd authentication.
5617 This could cause Exim to crash, or give a corrupted message.
5619 39. Added isip, isip4, and isip6 to ${if conditions.
5621 40. The ACL variables $acl_xx are now saved with the message, and can be
5622 accessed later in routers, transports, and filters.
5624 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5625 strings in the file are not string-expanded.
5627 42. If a MAIL command specified a SIZE value that was too large to fit into an
5628 int variable, the check against message_size_limit failed. Such values are
5629 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5630 day this will have to be increased, but I don't think I want to be around
5631 when emails are that large.
5638 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5639 iconv() is not standard in FreeBSD.
5641 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5642 IPv6 enabled. The observed symptom was a segmentation fault on return from
5643 the function os_common_find_running_interfaces() in src/os.c.
5645 3. In the check_special_case() function in daemon.c I had used "errno" as an
5646 argument name, which causes warnings on some systems. This was basically a
5647 typo, since it was named "eno" in the comments!
5649 4. The code that waits for the clock to tick (at a resolution of some fraction
5650 of a second) so as to ensure message-id uniqueness was always waiting for
5651 at least one whole tick, when it could have waited for less. [This is
5652 almost certainly not relevant at current processor speeds, where it is
5653 unlikely to ever wait at all. But we try to future-proof.]
5655 5. The function that sleeps for a time interval that includes fractions of a
5656 second contained a race. It did not block SIGALRM between setting the
5657 timer, and suspending (a couple of lines later). If the interval was short
5658 and the sigsuspend() was delayed until after it had expired, the suspension
5659 never ended. On busy systems this could lead to processes getting stuck for
5662 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5663 process, before it forks any delivery processes. The open lookup caching
5664 mechanism meant that the open file or database connection was passed into
5665 the delivery process. The problem was that delivery processes always tidy
5666 up cached lookup data. This could cause a problem for the next delivery
5667 process started by the queue runner, because the external queue runner
5668 process does not know about the closure. So the next delivery process
5669 still has data in the lookup cache. In the case of a file lookup, there was
5670 no problem because closing a file descriptor in a subprocess doesn't affect
5671 the parent. However, if the lookup was caching a connection to a database,
5672 the connection was closed, and the second delivery process was likely to
5673 see errors such as "PGSQL: query failed: server closed the connection
5674 unexpectedly". The problem has been fixed by closing all cached lookups
5675 in a queue runner before running a delivery process.
5677 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5678 seem to have the "const" qualifier which it has on other OS. I've
5681 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5682 *of the same type* (client or server) with the same public name that an
5683 error should be diagnosed.
5685 9. When Exim looked up a host name for an IP address, but failed to find the
5686 original IP address when looking up the host name (a safety check), it
5687 output the message "<ip address> does not match any IP for NULL", which was
5688 confusing, to say the least. The bug was that the host name should have
5689 appeared instead of "NULL".
5691 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5692 user that is built into the binary, and the -C or -D options is used, root
5693 privilege is dropped before the configuration file is read. In addition,
5694 logging is switched to stderr instead of the normal log files. If the
5695 configuration then re-defines the Exim user, the unprivileged environment
5696 is probably not what is expected, so Exim logs a panic warning message (but
5699 However, if deliver_drop_privilege is set, the unprivileged state may well
5700 be exactly what is intended, so the warning has been cut out in that case,
5701 and Exim is allowed to try to write to its normal log files.
5707 1. smtp_return_error_details was not giving details for temporary sender
5708 or receiver verification errors.
5710 2. Diagnose a configuration error if two authenticators have the same public
5713 3. Exim used not to create the message log file for a message until the first
5714 delivery attempt. This could be confusing when incoming messages were held
5715 for policy or load reasons. The message log file is now created at the time
5716 the message is received, and an initial "Received" line is written to it.
5718 4. The automatically generated man page for command line options had a minor
5719 bug that caused no ill effects; however, a more serious problem was that
5720 the procedure for building the man page automatically didn't always
5721 operate. Consequently, release 4.20 contains an out-of-date version. This
5722 shouldn't happen again.
5724 5. When building Exim with embedded Perl support, the script that builds the
5725 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5726 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5728 6. The freeze_tell option was not being used for messages that were frozen on
5729 arrival, either by an ACL or by local_scan().
5731 7. Added the smtp_incomplete_transaction log selector.
5733 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5734 was accepting AUTH without a new EHLO.
5736 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5737 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5740 10. Logging of TCP/IP connections (when configured) now happens in the main
5741 daemon process instead of the child process, so that the TCP/IP connection
5742 count is more accurate (but it can never be perfect).
5744 11. The use of "drop" in a nested ACL was not being handled correctly in the
5745 outer ACL. Now, if condition failure induced by the nested "drop" causes
5746 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5747 or "require"), the connection is dropped.
5749 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5750 that yield "discard" can now be used with an "accept" or a "discard" verb,
5751 but an error is generated for any others (because I can't see a useful way
5752 to define what should happen).
5754 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5755 are now processed in the same way as lines in the Exim configuration file.
5756 In particular, continuation lines are supported.
5758 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5760 15. Added -ti meaning -t -i.
5762 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5763 domains, and warn by logging if others are found.
5765 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5766 returned from GIFCONF when Exim is trying to find the list of interfaces on
5767 a host. The code in os.c has been modified to copy each ifreq to an aligned
5768 structure in all cases.
5770 Also, in some cases, the returned ifreq's were being copied to a 'struct
5771 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5772 means the last couple of bytes of an IPv6 address could be chopped if the
5773 ifreq contained only a normal sockaddr (14 bytes storage).
5775 18. Named domain lists were not supported in the hosts_treat_as_local option.
5776 An entry such as +xxxx was not recognized, and was treated as a literal
5779 19. Ensure that header lines added by a DATA ACL are included in the reject log
5780 if the ACL subsequently rejects the message.
5782 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5783 MD5 (which is deprecated).
5785 21. When testing a filter file using -bf, Exim was writing a message when it
5786 took the sender from a "From " line in the message, but it was not doing so
5787 when it took $return_path from a Return-Path: header line. It now does.
5789 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5790 with a valid header line field name (a series of printing characters
5791 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5793 23. Changed "disc" in the source to "disk" to conform to the documentation and
5794 the book and for uniformity.
5796 24. Ignore Sendmail's -Ooption=value command line item.
5798 25. When execve() failed while trying to run a command in a pipe transport,
5799 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5800 could be confused with a return value of 69 from the command itself. This
5801 has been changed to 127, the value the shell returns if it is asked to run
5802 a non-existent command. The wording for the related log line suggests a
5803 non-existent command as the problem.
5805 26. If received_header_text expands to an empty string, do not add a Received:
5806 header line to the message. (Well, it adds a token one on the spool, but
5807 marks it "old" so that it doesn't get used or transmitted.)
5809 27. Installed eximstats 1.28 (addition of -nt option).
5811 28. There was no check for failure on the call to getsockname() in the daemon
5812 code. This can fail if there is a shortage of resources on the system, with
5813 ENOMEM, for example. A temporary error is now given on failure.
5815 29. Contrary to the C standard, it seems that in some environments, the
5816 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5817 program. Exim now does this explicitly; it affects the formatting of
5818 timestamps using strftime().
5820 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5821 complaints. I've now initialized all the variables, to avoid this.
5823 32. Header lines added by a system filter were not being "seen" during
5824 transport-time rewrites.
5826 33. The info_callback() function passed to OpenSSL is set up with type void
5827 (*)(SSL *, int, int), as described somewhere. However, when calling the
5828 function (actually a macro) that sets it up, the type void(*)() is
5829 expected. I've put in a cast to prevent warnings from picky compilers.
5831 34. If a DNS black list lookup found a CNAME record, but there were no A
5832 records associated with the domain it pointed at, Exim crashed.
5834 35. If a DNS black list lookup returned more than one A record, Exim ignored
5835 all but the first. It now scans all returned addresses if a particular IP
5836 value is being sought. In this situation, the contents of the
5837 $dnslist_value variable are a list of all the addresses, separated by a
5840 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5841 to accept a host name and all its aliases if the forward lookup for any of
5842 them yielded the IP address of the incoming connection. Now it accepts only
5843 those names whose forward lookup yields the correct IP address. Any other
5844 names are discarded. This closes a loophole whereby a rogue DNS
5845 administrator could create reverse DNS records to break through a
5846 wildcarded host restriction in an ACL.
5848 37. If a user filter or a system filter that ran in a subprocess used any of
5849 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5850 the wrong values were passed to the pipe command ($thisaddress had the
5851 value of $0, $0 had the value of $1, etc). This bug was introduced by
5852 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5854 38. Improved the line breaking for long SMTP error messages from ACLs.
5855 Previously, if there was no break point between 40 and 75 characters, Exim
5856 left the rest of the message alone. Two changes have been made: (a) I've
5857 reduced the minimum length to 35 characters; (b) if it can't find a break
5858 point between 35 and 75 characters, it looks ahead and uses the first one
5859 that it finds. This may give the occasional overlong line, but at least the
5860 remaining text gets split now.
5862 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5863 file descriptors might be low, and that setting 1000 would always raise it.
5864 It turns out that in some environments, the limit is already over 1000 and
5865 that lowering it causes trouble. So now Exim takes care not to decrease it.
5867 40. When delivering a message, the value of $return_path is set to $sender_
5868 address at the start of routing (routers may change the value). By an
5869 oversight, this default was not being set up when an address was tested by
5870 -bt or -bv, which affected the outcome if any router or filter referred to
5873 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5874 log only when "message" or "log_message" are set. However, if one of the
5875 conditions was an address verification, or a call to a nested ACL, the
5876 messages generated by the underlying test were being passed through. This
5877 no longer happens. The underlying message is available in $acl_verify_
5878 message for both "message" and "log_message" expansions, so it can be
5879 passed through if needed.
5881 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5882 new expansion $bh_ to give the encoded byte string without charset
5883 translation. Translation happens only if iconv() is available; HAVE_ICONV
5884 indicates this at build time. HEADERS_CHARSET gives the charset to
5885 translate to; headers_charset can change it in the configuration, and
5886 "headers charset" can change it in an individual filter file.
5888 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5889 that creates RFC 2047 encoded "words" labels them as that charset instead
5890 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5891 expansion operator; (ii) when Exim creates a From: line for a local
5892 message; (iii) when a header line is rewritten to include a "phrase" part.
5894 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5895 buggy, causing it to skip the first lines of messages whose message ID
5896 ended in 'D'. This would not have bitten before Exim release 4.14, because
5897 message IDs were unlikely to end in 'D' before then. The effect was to have
5898 incorrect size information for certain domains.
5900 45. #include "config.h" was missing at the start of the crypt16.c module. This
5901 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5904 46. If there was a timeout during a "random" callout check, Exim treated it as
5905 a failure of the random address, and carried on sending RSET and the real
5906 address. If the delay was just some slowness somewhere, the response to the
5907 original RCPT would be taken as a response to RSET and so on, causing
5908 mayhem of various kinds.
5910 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5911 when I implemented it. It didn't allow for the fact that some option values
5912 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5913 the right test for positive values.
5915 48. Domain names in DNS records are case-independent. Exim always looks them up
5916 in lower case. Some resolvers return domain names in exactly the case they
5917 appear in the zone file, that is, they may contain uppercase letters. Not
5918 all resolvers do this - some return always lower case. Exim was treating a
5919 change of case by a resolver as a change of domain, similar to a widening
5920 of a domain abbreviation. This triggered its re-routing code and so it was
5921 trying to route what was effectively the same domain again. This normally
5922 caused routing to fail (because the router wouldn't handle the domain
5923 twice). Now Exim checks for this case specially, and just changes the
5924 casing of the domain that it ultimately uses when it transmits the message
5927 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5930 50. If a filter generated a file delivery with a non-absolute name (possible if
5931 no home directory exists for the router), the forbid_file option was not
5934 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5935 the existing equality matching.
5937 52. Exim was using ints instead of ino_t variables in some places where it was
5938 dealing with inode numbers.
5940 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5941 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5942 called TMPDIR, and if it finds it is different, it changes its value.
5944 54. The smtp_printf() function is now made available to local_scan() so
5945 additional output lines can be written before returning. There is also an
5946 smtp_fflush() function to enable the detection of a dropped connection.
5947 The variables smtp_input and smtp_batched_input are exported to
5950 55. Changed the default runtime configuration: the message "Unknown user"
5951 has been removed from the ACL, and instead placed on the localuser router,
5952 using the cannot_route_message feature. This means that any verification
5953 failures that generate their own messages won't get overridden. Similarly,
5954 the "Unrouteable address" message that was in the ACL for unverifiable
5955 relay addresses has also been removed.
5957 56. Added hosts_avoid_esmtp to the smtp transport.
5959 57. The exicyclog script was not checking for the esoteric option
5960 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5961 will work only if exicyclog is run under the appropriate euid.
5963 58. Following a discussion on the list, the rules by which Exim recognises line
5964 endings on incoming messages have been changed. The -dropcr and drop_cr
5965 options are now no-ops, retained only for backwards compatibility. The
5966 following line terminators are recognized: LF CRLF CR. However, special
5967 processing applies to CR:
5969 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5970 nor a local message in the state where . is a terminator.
5972 (ii) If a bare CR is encountered in a header line, an extra space is added
5973 after the line terminator so as not to end the header. The reasoning
5974 behind this is that bare CRs in header lines are most likely either
5975 to be mistakes, or people trying to play silly games.
5977 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5978 was being incorrectly given as 18 bytes larger than it should have been.
5979 This is a VOB (very old bug).
5981 60. This may never have affected anything current, but just in case it has:
5982 When the local host is found other than at the start of a list of hosts,
5983 the local host, those with the same MX, and any that follow, are discarded.
5984 When the list in question was part of a longer list of hosts, the following
5985 hosts (not currently being processed) were also being discarded. This no
5986 longer happens. I'm not sure if this situation could ever has previously
5989 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5992 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5993 header. This is something that is recommended in a new Internet Draft, and
5994 is something that is documented as being done by Sendmail. There are two
5995 possible values. For messages generated by the autoreply transport, Exim
5998 Auto-Submitted: auto-replied
6000 whereas for all other generated messages (e.g. bounces) it adds
6002 Auto-Submitted: auto-generated
6004 63. The "personal" condition in filters now includes a test for the
6005 Auto-Submitted: header. If it contains the string "auto-" the message it
6006 not considered personal.
6008 64. Added rcpt_include_affixes as a generic transport option.
6010 65. Added queue_only_override (default true).
6012 66. Added the syslog_duplication option.
6014 67. If what should have been the first header line of a message consisted of
6015 a space followed by a colon, Exim was mis-interpreting it as a header line.
6016 It isn't of course - it is syntactically invalid and should therefore be
6017 treated as the start of the message body. The misbehaviour could have
6018 caused a number of strange effects, including loss of data in subsequent
6019 header lines, and spool format errors.
6021 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6022 client host had authenticated. This control can now be exercised by an ACL
6023 for more flexibility.
6025 69. By default, callouts do not happen when testing with -bh. There is now a
6026 variant, -bhc, which does actually run the callout code, including
6027 consulting and updating the callout cache.
6029 70. Added support for saslauthd authentication, courtesy of Alexander
6032 71. If statvfs() failed on the spool or log directories while checking their
6033 size for availability, Exim confusingly gave the error "space shortage".
6034 Furthermore, in debugging mode it crashed with a floating point exception.
6035 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6036 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6037 serious problem, Exim now writes to the main and panic logs when this
6038 happens, with details of the failure. It then refuses to accept the
6039 incoming message, giving the message "spool directory problem" or "log
6040 directory problem" with a 421 code for SMTP messages.
6042 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6043 0, 1, and 2 exist, because some OS complain for execs without them (see
6044 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6045 descriptors. However, the code omitted to check that the open succeeded,
6046 causing mysterious errors if for some reason the permissions on /dev/null
6047 got screwed. Now Exim writes a message to the main and panic logs, and
6048 bombs out if it can't open /dev/null.
6050 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6051 interact so that it is all more flexible. It is supposed to remain
6052 backwards compatible. Also added extra_local_interfaces.
6054 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6055 to bomb out with an assertion failure - to the client this appears as a
6056 connection drop. This problem occurs in the part of the code that was taken
6057 from the Samba project. Fortunately, the assertion is in a very simple
6058 function, so I have fixed this by reproducing the function inline in the
6059 one place where it is called, and arranging for authentication to fail
6060 instead of killing the process with assert().
6062 75. The SPA client code was not working when the server requested OEM rather
6063 than Unicode encoding.
6065 76. Added code to make require_files with a specific uid setting more usable in
6066 the case where statting the file as root fails - usually a non-root-mounted
6067 NFS file system. When this happens and the failure is EACCES, Exim now
6068 forks a subprocess and does the per-uid checking as the relevant uid.
6070 77. Added process_log_path.
6072 78. If log_file_path was not explicitly set, a setting of check_log_space or
6073 check_log_inodes was ignored.
6075 79. If a space check for the spool or log partitions fails, the incident is now
6076 logged. Of course, in the latter case the data may get lost...
6078 80. Added the %p formatting code to string_format() so that it can be used to
6079 print addresses in debug_print(). Adjusted all the address printing in the
6080 debugging in store.c to use %p rather than %d.
6082 81. There was a concern that a line of code in smtp_in.c could overflow a
6083 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6084 initially expressed, the concern was not well-founded, because trailing
6085 spaces are removed early. However, if the trailing spaces were followed by
6086 a NULL, they did not get removed, so the overflow was possible. Two fixes
6089 (a) I re-wrote the offending code in a cleaner fashion.
6090 (b) If an incoming SMTP command contains a NULL character, it is rejected
6093 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6094 runs initgroups(), so that if the Exim user is in any additional groups,
6095 they will be used during message reception.
6101 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob