1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet. PPv2 support was
29 updated based on HAProxy spec change in May 2014.
31 JH/02 Add ${listextract {number}{list}{success}{fail}}.
33 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
34 Properly escape header and check for NULL return.
36 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
39 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
41 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
42 characters in header names, implemented as a verify condition.
43 Contributed by Michael Fischer v. Mollard.
45 TL/05 Rename SPF condition results err_perm and err_temp to standardized
46 results permerror and temperror. Previous values are deprecated but
47 still accepted. In a future release, err_perm and err_temp will be
48 completely removed, which will be a backward incompatibility if the
49 ACL tests for either of these two old results. Patch contributed by
50 user bes-internal on the mailing list.
52 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
54 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
55 selectors, in both main and reject logs.
57 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
60 JH/07 Add malware type "sock" for talking to simple daemon.
62 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
64 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
65 routers/transports under cutthrough routing.
67 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
68 numbers. Touch up "bool" conditional to keep the same definition.
70 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
72 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
74 JH/12 Expand items in router/transport headers_add or headers_remove lists
75 individually rather than the list as a whole. Bug 1452.
77 Required for reasonable handling of multiple headers_ options when
78 they may be empty; requires that headers_remove items with embedded
79 colons must have them doubled (or the list-separator changed).
81 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
82 view the policy declared in the DMARC record. Currently, $dmarc_status
83 is a combined value of both the record presence and the result of the
86 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
88 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
89 dnslookup router and the smtp transport (applying to the forward
92 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
93 of ldap servers used for a specific lookup. Patch provided by Heiko
96 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
97 New variable $lookup_dnssec_authenticated for observability.
99 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
100 Patch submitted by Lars Timman.
102 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
104 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
105 Requires trusted mode and valid format message id, aborts otherwise.
106 Patch contributed by Heiko Schlichting.
108 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
109 certextract with support for various fields. Bug 1358.
111 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
112 is requested by default, modifiable by smtp transport option
115 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
116 operate on certificate variables to give certificate fingerprints
117 Also new ${sha256:cert_variable}.
119 JH/23 The PRDR feature is moved from being Experimental into the mainline.
121 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
122 Christian Aistleitner.
124 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
126 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
127 file. Patch from Wolfgang Breyha.
129 JH/25 Expand the coverage of the delivery $host and $host_address to
130 client authenticators run in verify callout. Bug 1476.
132 JH/26 Port service names are now accepted for tls_on_connect_ports, to
133 align with daemon_smtp_ports. Bug 72.
135 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
136 support and error reporting did not work properly.
138 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
139 and is readable. Patch from Andrew Colin Kissa.
141 TL/14 Enhance documentation of ${run expansion and how it parses the
142 commandline after expansion, particularly in the case when an
143 unquoted variable expansion results in an empty value.
145 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
147 PP/02 Fix internal collision of T_APL on systems which support RFC3123
148 by renaming away from it. Addresses GH issue 15, reported by
151 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
157 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
159 PP/02 Make -n do something, by making it not do something.
160 When combined with -bP, the name of an option is not output.
162 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
165 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
166 $sender_host_name and config options to manage this, and basic check
169 PP/05 DSCP support for outbound connections and control modifier for inbound.
171 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
172 (Only plugin which currently uses this is kerberos4, which nobody should
173 be using, but we should make it available and other future plugins might
174 conceivably use it, even though it would break NAT; stuff *should* be
175 using channel bindings instead).
177 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
178 name; added for Sendmail compatibility; requires admin caller.
179 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
180 just ignore it); requires trusted caller.
181 Also parse but ignore: -Ac -Am -X<logfile>
184 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
186 TL/02 Add +smtp_confirmation as a default logging option.
188 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
189 Patch by Magnus Holmgren from 2007-02-20.
191 TL/04 Bugzilla 1281 - Spec typo.
192 Bugzilla 1283 - Spec typo.
193 Bugzilla 1290 - Spec grammar fixes.
195 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
197 TL/06 Add Experimental DMARC support using libopendmarc libraries.
199 TL/07 Fix an out of order global option causing a segfault. Reported to dev
200 mailing list by by Dmitry Isaikin.
202 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
204 JH/02 Support "G" suffix to numbers in ${if comparisons.
206 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
208 NM/01 Bugzilla 1197 - Spec typo
209 Bugzilla 1196 - Spec examples corrections
211 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
213 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
214 gnutls_enable_pkcs11, but renamed to more accurately indicate its
217 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
218 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
220 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
221 "acl {{name}{arg}...}", and optional args on acl condition
224 JH/05 Permit multiple router/transport headers_add/remove lines.
226 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
228 JH/07 Avoid using a waiting database for a single-message-only transport.
229 Performance patch from Paul Fisher. Bugzilla 1262.
231 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
234 JH/09 Add $headers_added variable, with content from use of ACL modifier
235 add_header (but not yet added to the message). Bugzilla 199.
237 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
238 Pulled from Bugzilla 817 by Wolfgang Breyha.
240 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
242 (nb: this is the same fix as in Exim 4.80.1)
244 JH/11 Add A= logging on delivery lines, and a client_set_id option on
247 JH/12 Add optional authenticated_sender logging to A= and a log_selector
250 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
252 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
253 advertise SMTP AUTH mechanism to us, instead of a generic
254 protocol violation error. Also, make Exim more robust to bad
255 data from the Dovecot auth socket.
257 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
259 When a queue runner is handling a message, Exim first routes the
260 recipient addresses, during which it prunes them based on the retry
261 hints database. After that it attempts to deliver the message to
262 any remaining recipients. It then updates the hints database using
265 So if a recipient address works intermittently, it can get repeatedly
266 deferred at routing time. The retry hints record remains fresh so the
267 address never reaches the final cutoff time.
269 This is a fairly common occurrence when a user is bumping up against
270 their storage quota. Exim had some logic in its local delivery code
271 to deal with this. However it did not apply to per-recipient defers
272 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
274 This change adds a proper retry rule check during routing so that the
275 final cutoff time is checked against the message's age. We only do
276 this check if there is an address retry record and there is not a
277 domain retry record; this implies that previous attempts to handle
278 the address had the retry_use_local_parts option turned on. We use
279 this as an approximation for the destination being like a local
280 delivery, as in LMTP.
282 I suspect this new check makes the old local delivery cutoff check
283 redundant, but I have not verified this so I left the code in place.
285 TF/02 Correct gecos expansion when From: is a prefix of the username.
287 Test 0254 submits a message to Exim with the header
291 When I ran the test suite under the user fanf2, Exim expanded
292 the header to contain my full name, whereas it should have added
293 a Resent-Sender: header. It erroneously treats any prefix of the
294 username as equal to the username.
296 This change corrects that bug.
298 GF/01 DCC debug and logging tidyup
299 Error conditions log to paniclog rather than rejectlog.
300 Debug lines prefixed by "DCC: " to remove any ambiguity.
302 TF/03 Avoid unnecessary rebuilds of lookup-related code.
304 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
305 Bug spotted by Jeremy Harris; was flawed since initial commit.
306 Would have resulted in OCSP responses post-SNI triggering an Exim
307 NULL dereference and crash.
309 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
311 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
312 Bug detection, analysis and fix by Samuel Thibault.
313 Bugzilla 1331, Debian bug #698092.
315 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
317 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
318 Server implementation by Todd Lyons, client by JH.
319 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
320 config variable "prdr_enable" controls whether the server
321 advertises the facility. If the client requests PRDR a new
322 acl_data_smtp_prdr ACL is called once for each recipient, after
323 the body content is received and before the acl_smtp_data ACL.
324 The client is controlled by bolth of: a hosts_try_prdr option
325 on the smtp transport, and the server advertisement.
326 Default client logging of deliveries and rejections involving
327 PRDR are flagged with the string "PRDR".
329 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
330 fclose(). Diagnosis by Todd Lyons.
332 PP/17 Update configure.default to handle IPv6 localhost better.
333 Patch by Alain Williams (plus minor tweaks).
336 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
337 This is now consistent with GnuTLS, and is now documented: the
338 previous undocumented portable approach to treating the option as
339 unset was to force an expansion failure. That still works, and
340 an empty string is now equivalent.
342 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
343 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
344 not performing validation itself.
346 PP/20 Added force_command boolean option to pipe transport.
347 Patch from Nick Koston, of cPanel Inc.
349 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
352 TF/04 Added udpsend ACL modifer and hexquote expansion operator
354 PP/21 Fix eximon continuous updating with timestamped log-files.
355 Broken in a format-string cleanup in 4.80, missed when I repaired the
356 other false fix of the same issue.
357 Report and fix from Heiko Schlichting.
360 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
361 Report from Prashanth Katuri.
363 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
364 It's SecureTransport, so affects any MacOS clients which use the
365 system-integrated TLS libraries, including email clients.
367 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
368 using a MIME ACL for non-SMTP local injection.
369 Report and assistance in diagnosis by Warren Baker.
371 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
373 JH/16 Fix comparisons for 64b. Bugzilla 1385.
375 TL/09 Add expansion variable $authenticated_fail_id to keep track of
376 last id that failed so it may be referenced in subsequent ACL's.
378 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
381 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
382 ldap library initialization, allowing self-signed CA's to be
383 used. Also properly sets require_cert option later in code by
384 using NULL (global ldap config) instead of ldap handle (per
385 session). Bug diagnosis and testing by alxgomz.
387 TL/12 Enhanced documentation in the ratelimit.pl script provided in
388 the src/util/ subdirectory.
390 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
391 renamed to Transport Post Delivery Action by Jeremy Harris, as
394 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
395 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
396 redis_servers = needs to be configured which will be used by the redis
397 lookup. Patch from Warren Baker, of The Packet Hub.
399 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
401 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
402 hostname or reverse DNS when processing a host list. Used suggestions
403 from multiple comments on this bug.
405 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
407 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
408 Missed a few lines, added it to make the runtest require no keyboard
411 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
412 contains upper case chars. Make router use caseful_local_part.
414 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
415 support when GnuTLS has been built with p11-kit.
421 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
423 This, or similar/improved, will also be change PP/11 of 4.82.
429 PP/01 Handle short writes when writing local log-files.
430 In practice, only affects FreeBSD (8 onwards).
431 Bugzilla 1053, with thanks to Dmitry Isaikin.
433 NM/01 Bugzilla 949 - Documentation tweak
435 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
438 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
440 PP/02 Implemented gsasl authenticator.
442 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
444 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
445 `pkg-config foo` for cflags/libs.
447 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
448 with rest of GSASL and with heimdal_gssapi.
450 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
451 `pkg-config foo` for cflags/libs for the TLS implementation.
453 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
454 properties get this fed in as external SSF. A number of robustness
455 and debugging improvements to the cyrus_sasl authenticator.
457 PP/08 cyrus_sasl server now expands the server_realm option.
459 PP/09 Bugzilla 1214 - Log authentication information in reject log.
460 Patch by Jeremy Harris.
462 PP/10 Added dbmjz lookup type.
464 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
466 PP/12 MAIL args handles TAB as well as SP, for better interop with
467 non-compliant senders.
468 Analysis and variant patch by Todd Lyons.
470 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
471 Bug report from Lars Müller <lars@samba.org> (via SUSE),
472 Patch from Dirk Mueller <dmueller@suse.com>
474 PP/13 tls_peerdn now print-escaped for spool files.
475 Observed some $tls_peerdn in wild which contained \n, which resulted
476 in spool file corruption.
478 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
479 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
480 or write after TLS renegotiation, which otherwise led to messages
483 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
484 as a tracking header (ie: a signed header comes before the signature).
485 Patch from Wolfgang Breyha.
487 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
488 comma-sep list; embedded commas doubled.
490 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
492 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
494 Report and patch from Dmitry Banschikov.
496 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
497 Removed SSL_clear() after SSL_new() which led to protocol negotiation
498 failures. We appear to now support TLS1.1+ with Exim.
500 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
501 lets Exim select keys and certificates based upon TLS SNI from client.
502 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
503 before an outbound SMTP session. New log_selector, +tls_sni.
505 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
506 NULL dereference. Report and patch from Alun Jones.
508 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
509 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
510 Not seeing resolver debug output on NetBSD, but suspect this is a
511 resolver implementation change.
513 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
514 Left warnings. Added "eximon gdb" invocation mode.
516 PP/21 Defaulting "accept_8bitmime" to true, not false.
518 PP/22 Added -bw for inetd wait mode support.
520 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
521 locate the relevant includes and libraries. Made this the default.
523 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
524 Bugzilla 1246, report and most of solution from Tomasz Kusy.
526 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
527 This may cause build issues on older platforms.
529 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
530 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
531 gnutls_require_mac & gnutls_require_protocols (no longer supported).
532 Added SNI support via GnuTLS too.
533 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
535 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
537 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
538 Applied second patch from Janne, implementing suggestion to default
539 multiple-strings-in-record handling to match SPF spec.
541 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
543 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
544 read-only, out of scope).
545 Patch from Wolfgang Breyha, report from Stuart Northfield.
547 PP/29 Fix three issues highlighted by clang analyser static analysis.
548 Only crash-plausible issue would require the Cambridge-specific
549 iplookup router and a misconfiguration.
550 Report from Marcin Mirosław.
552 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
554 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
555 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
556 As part of this, removing so much warning spew let me fix some minor
557 real issues in debug logging.
559 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
560 assignment on my part. Fixed.
562 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
563 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
564 Janne Snabb (who went above and beyond: thank you).
566 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
567 string otherwise requires a connection and a bunch more work and it's
568 relatively easy to get wrong. Should also expose TLS library linkage
571 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
572 64-bit ${eval} (JH/03).
574 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
575 GNU libc to support some of the 64-bit stuff, should not lead to
576 conflicts. Defined before os.h is pulled in, so if a given platform
577 needs to override this, it can.
579 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
580 protection layer was required, which is not implemented.
581 Bugzilla 1254, patch from Wolfgang Breyha.
583 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
584 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
585 tls_dhparam take prime identifiers. Also unbreak combination of
586 OpenSSL+DH_params+TLSSNI.
588 PP/39 Disable SSLv2 by default in OpenSSL support.
594 PP/01 Solaris build fix for Oracle's LDAP libraries.
595 Bugzilla 1109, patch from Stephen Usher.
597 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
599 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
602 TF/02 Fix a couple more cases where we did not log the error message
603 when unlink() failed. See also change 4.74-TF/03.
605 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
606 lock up or crash if it happened to be inside a call to libc when it
607 got a SIGUSR1 from exiwhat.
609 The SIGUSR1 handler appends the current process status to the process
610 log which is later printed by exiwhat. It used to use the general
611 purpose logging code to do this, but several functions it calls are
612 not safe for signals.
614 The new output code in the SIGUSR1 handler is specific to the process
615 log, and simple enough that it's easy to inspect for signal safety.
616 Removing some special cases also simplifies the general logging code.
617 Removing the spurious timestamps from the process log simplifies
620 TF/04 Improved ratelimit ACL condition.
622 The /noupdate option has been deprecated in favour of /readonly which
623 has clearer semantics. The /leaky, /strict, and /readonly update modes
624 are mutually exclusive. The update mode is no longer included in the
625 database key; it just determines when the database is updated. (This
626 means that when you upgrde Exim will forget old rate measurements.)
628 Exim now checks that the per_* options are used with an update mode that
629 makes sense for the current ACL. For example, when Exim is processing a
630 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
631 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
632 must specify per_mail/readonly. If you omit the update mode it defaults to
633 /leaky where that makes sense (as before) or /readonly where required.
635 The /noupdate option is now undocumented but still supported for
636 backwards compatibility. It is equivalent to /readonly except that in
637 ACLs where /readonly is required you may specify /leaky/noupdate or
638 /strict/noupdate which are treated the same as /readonly.
640 A useful new feature is the /count= option. This is a generalization
641 of the per_byte option, so that you can measure the throughput of other
642 aggregate values. For example, the per_byte option is now equivalent
643 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
645 The per_rcpt option has been generalized using the /count= mechanism
646 (though it's more complicated than the per_byte equivalence). When it is
647 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
648 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
649 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
650 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
651 Note that using per_rcpt with a non-readonly update mode in more than
652 one ACL will cause the recipients to be double-counted. (The per_mail
653 and per_byte options don't have this problem.)
655 The handling of very low rates has changed slightly. If the computed rate
656 is less than the event's count (usually one) then this event is the first
657 after a long gap. In this case the rate is set to the same as this event's
658 count, so that the first message of a spam run is counted properly.
660 The major new feature is a mechanism for counting the rate of unique
661 events. The new per_addr option counts the number of different
662 recipients that someone has sent messages to in the last time period. It
663 behaves like per_rcpt if all the recipient addresses are different, but
664 duplicate recipient addresses do not increase the measured rate. Like
665 the /count= option this is a general mechanism, so the per_addr option
666 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
667 example, measure the rate that a client uses different sender addresses
668 with the options per_mail/unique=$sender_address. There are further
669 details in the main documentation.
671 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
673 TF/06 Removed a few PCRE remnants.
675 TF/07 Automatically extract Exim's version number from tags in the git
676 repository when doing development or release builds.
678 PP/02 Raise smtp_cmd_buffer_size to 16kB.
679 Bugzilla 879. Patch from Paul Fisher.
681 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
682 Heavily based on revision 40f9a89a from Simon Arlott's tree.
685 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
687 PP/05 Variable $av_failed, true if the AV scanner deferred.
688 Bugzilla 1078. Patch from John Horne.
690 PP/06 Stop make process more reliably on build failure.
691 Bugzilla 1087. Patch from Heiko Schlittermann.
693 PP/07 Make maildir_use_size_file an _expandable_ boolean.
694 Bugzilla 1089. Patch from Heiko Schlittermann.
696 PP/08 Handle ${run} returning more data than OS pipe buffer size.
697 Bugzilla 1131. Patch from Holger Weiß.
699 PP/09 Handle IPv6 addresses with SPF.
700 Bugzilla 860. Patch from Wolfgang Breyha.
702 PP/10 GnuTLS: support TLS 1.2 & 1.1.
704 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
707 PP/11 match_* no longer expand right-hand-side by default.
708 New compile-time build option, EXPAND_LISTMATCH_RHS.
709 New expansion conditions, "inlist", "inlisti".
711 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
713 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
715 PP/14 fix log_write() format string regression from TF/03.
716 Bugzilla 1152. Patch from Dmitry Isaikin.
722 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
724 PP/02 Harmonised TLS library version reporting; only show if debugging.
725 Layout now matches that introduced for other libraries in 4.74 PP/03.
727 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
729 PP/04 New "dns_use_edns0" global option.
731 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
734 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
735 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
737 TK/01 Updated PolarSSL code to 0.14.2.
738 Bugzilla 1097. Patch from Andreas Metzler.
740 PP/07 Catch divide-by-zero in ${eval:...}.
743 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
746 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
747 format-string attack -- SECURITY: remote arbitrary code execution.
749 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
750 time unintentionally subject to list matching rules, letting the header
751 cause arbitrary Exim lookups (of items which can occur in lists, *not*
752 arbitrary string expansion). This allowed for information disclosure.
754 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
755 INT_MIN/-1 -- value coerced to INT_MAX.
761 NM/01 Workround for PCRE version dependency in version reporting
764 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
765 This fixes portability to compilers other than gcc, notably
766 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
768 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
769 makefiles for portability to HP-UX and POSIX correctness.
771 PP/01 Permit LOOKUP_foo enabling on the make command-line.
772 Also via indented variable definition in the Makefile.
773 (Debugging by Oliver Heesakkers).
775 PP/02 Restore caching of spamd results with expanded spamd_address.
776 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
778 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
779 Improves build reliability. Fix from: Frank Elsner
781 NM/02 Fix wide character breakage in the rfc2047 coding
782 Fixes bug 1064. Patch from Andrey N. Oktyabrski
784 NM/03 Allow underscore in dnslist lookups
785 Fixes bug 1026. Patch from Graeme Fowler
787 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
788 Code patches from Adam Ciarcinski of NetBSD.
790 NM/04 Fixed exiqgrep to cope with mailq missing size issue
793 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
794 is logged, to avoid truncation. Patch from John Horne.
796 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
797 Patch from Jakob Hirsch.
799 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
800 SQL string expansion failure details.
801 Patch from Andrey Oktyabrski.
803 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
804 Patch from Simon Arlott.
806 PP/09 New lookups functionality failed to compile on old gcc which rejects
807 extern declarations in function scope.
808 Patch from Oliver Fleischmann
810 PP/10 Use sig_atomic_t for flags set from signal handlers.
811 Check getgroups() return and improve debugging.
812 Fixed developed for diagnosis in bug 927 (which turned out to be
815 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
816 Patch from Mark Zealey.
818 PP/12 Bugzilla 1056: Improved spamd server selection.
819 Patch from Mark Zealey.
821 PP/13 Bugzilla 1086: Deal with maildir quota file races.
822 Based on patch from Heiko Schlittermann.
824 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
825 Patch from Uwe Doering, sign-off by Michael Haardt.
827 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
828 variable declaration deep within a block. Bug and patch from
831 PP/15 lookups-Makefile IRIX compatibilty coercion.
833 PP/16 Make DISABLE_DKIM build knob functional.
835 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
836 Patch by Simon Arlott
838 TF/03 Fix valgrind.h portability to C89 compilers that do not support
839 variable argument macros. Our copy now differs from upstream.
845 TF/01 Failure to get a lock on a hints database can have serious
846 consequences so log it to the panic log.
848 TF/02 Log LMTP confirmation messages in the same way as SMTP,
849 controlled using the smtp_confirmation log selector.
851 TF/03 Include the error message when we fail to unlink a spool file.
853 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
854 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
855 for maintaining out-of-tree patches for some time.
857 PP/01 Bugzilla 139: Documentation and portability issues.
858 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
859 Handle per-OS dynamic-module compilation flags.
861 PP/02 Let /dev/null have normal permissions.
862 The 4.73 fixes were a little too stringent and complained about the
863 permissions on /dev/null. Exempt it from some checks.
864 Reported by Andreas M. Kirchwitz.
866 PP/03 Report version information for many libraries, including
867 Exim version information for dynamically loaded libraries. Created
868 version.h, now support a version extension string for distributors
869 who patch heavily. Dynamic module ABI change.
871 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
872 privilege escalation vulnerability whereby the Exim run-time user
873 can cause root to append content of the attacker's choosing to
876 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
879 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
880 If dropping privileges for untrusted macros, we disabled normal logging
881 on the basis that it would fail; for the Exim run-time user, this is not
882 the case, and it resulted in successful deliveries going unlogged.
883 Fixed. Reported by Andreas Metzler.
889 PP/01 Date: & Message-Id: revert to normally being appended to a message,
890 only prepend for the Resent-* case. Fixes regression introduced in
891 Exim 4.70 by NM/22 for Bugzilla 607.
893 PP/02 Include check_rfc2047_length in configure.default because we're seeing
894 increasing numbers of administrators be bitten by this.
896 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
898 PP/03 Bugzilla 994: added openssl_options main configuration option.
900 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
902 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
904 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
906 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
907 without a peer certificate, leading to a segfault because of an
908 assumption that peers always have certificates. Be a little more
909 paranoid. Problem reported by Martin Tscholak.
911 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
912 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
913 NB: ClamAV planning to remove STREAM in "middle of 2010".
914 CL also introduces -bmalware, various -d+acl logging additions and
915 more caution in buffer sizes.
917 PP/09 Implemented reverse_ip expansion operator.
919 PP/10 Bugzilla 937: provide a "debug" ACL control.
921 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
923 PP/12 Bugzilla 973: Implement --version.
925 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
927 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
929 PP/15 Bugzilla 816: support multiple condition rules on Routers.
931 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
932 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
933 ignore trailing whitespace.
935 JJ/02 prevent non-panic DKIM error from being sent to paniclog
937 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
940 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
941 Notification from Dr Andrew Aitchison.
943 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
944 ExtendedDetectionInfo response format.
945 Notification from John Horne.
947 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
950 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
951 XSL and documented dependency on system catalogs, with examples of how
954 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
957 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
958 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
959 configuration file which is writeable by the Exim user or group.
961 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
962 of configuration files to cover files specified with the -C option if
963 they are going to be used with root privileges, not just the default
966 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
967 option (effectively making it always true).
969 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
970 files to be used while preserving root privileges.
972 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
973 that rogue child processes cannot use them.
975 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
976 run-time user, instead of root.
978 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
979 Exim run-time user without dropping privileges.
981 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
982 result string, instead of calling string_vformat() twice with the same
985 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
986 for other users. Others should always drop root privileges if they use
987 -C on the command line, even for a whitelisted configure file.
989 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
991 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
997 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
998 $data_path, and $header_path variables; fixed documentation bugs and
1001 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1002 exipick to access non-standard spools, including the "frozen" queue
1005 NM/01 Bugzilla 965: Support mysql stored procedures.
1006 Patch from Alain Williams
1008 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1010 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1011 Patch from Andreas Metzler
1013 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1014 Patch from Kirill Miazine
1016 NM/05 Bugzilla 671: Added umask to procmail example.
1018 JJ/03 installed exipick 20100323.0, fixing doc bug
1020 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1021 directory. Notification and patch from Dan Rosenberg.
1023 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1025 TK/02 Improve log output when DKIM signing operation fails.
1027 MH/01 Treat the transport option dkim_domain as a colon separated
1028 list, not as a single string, and sign the message with each element,
1029 omitting multiple occurences of the same signer.
1031 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1032 Bugzilla 985, 986. Patch by Simon Arlott
1034 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1035 Patch by Simon Arlott
1037 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1038 MBX locking. Notification from Dan Rosenberg.
1044 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1046 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1048 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1050 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1052 NM/04 Bugzilla 915: Flag broken perl installation during build.
1058 TK/01 Added patch by Johannes Berg that expands the main option
1059 "spamd_address" if it starts with a dollar sign.
1061 TK/02 Write list of recipients to X-Envelope-Sender header when building
1062 the mbox-format spool file for content scanning (suggested by Jakob
1065 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1066 (http://www.dcc-servers.net/) support via dccifd. Activated by
1067 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1069 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1070 by Mark Daniel Reidel <mr@df.eu>.
1072 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1073 When building exim an external PCRE library is now needed -
1074 PCRE is a system library on the majority of modern systems.
1075 See entry on PCRE_LIBS in EDITME file.
1077 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1078 conversation. Added nologin parameter to request.
1079 Patch contributed by Kirill Miazine.
1081 TF/01 Do not log submission mode rewrites if they do not change the address.
1083 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1085 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1086 log files in place. Contributed by Roberto Lima.
1088 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1090 TF/03 Bugzilla 615: When checking the local_parts router precondition
1091 after a local_part_suffix or local_part_prefix option, Exim now
1092 does not use the address's named list lookup cache, since this
1093 contains cached lookups for the whole local part.
1095 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1096 Robert Millan. Documentation is in experimental-spec.txt.
1098 TF/04 Bugzilla 668: Fix parallel build (make -j).
1100 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1102 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1103 Patch provided by Jan Srzednicki.
1105 TF/05 Leading white space used to be stripped from $spam_report which
1106 wrecked the formatting. Now it is preserved.
1108 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1109 that they are available at delivery time.
1111 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1113 TF/08 TLS error reporting now respects the incoming_interface and
1114 incoming_port log selectors.
1116 TF/09 Produce a more useful error message if an SMTP transport's hosts
1117 setting expands to an empty string.
1119 NM/06 Bugzilla 744: EXPN did not work under TLS.
1120 Patch provided by Phil Pennock.
1122 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1123 Patch provided by Richard Godbee.
1125 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1126 acl_smtp_notquit, added index entry.
1128 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1129 Patch provided by Eugene Bujak.
1131 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1132 accept(). Patch provided by Maxim Dounin.
1134 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1135 Patch provided by Phil Pennock.
1137 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1139 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1140 Patch provided by Brad "anomie" Jorsch.
1142 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1143 Patch provided by Dean Brooks.
1145 TK/05 Add native DKIM support (does not depend on external libraries).
1147 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1148 Patch provided by Graeme Fowler.
1150 NM/16 Bugzilla 851: Documentation example syntax fix.
1152 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1154 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1157 NM/19 Bugzilla 745: TLS version reporting.
1158 Patch provided by Phil Pennock.
1160 NM/20 Bugzilla 167: bool: condition support.
1161 Patch provided by Phil Pennock.
1163 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1164 clients. Patch provided by Phil Pennock.
1166 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1167 Patch provided by Brad "anomie" Jorsch.
1169 NM/23 Bugzilla 687: Fix misparses in eximstats.
1170 Patch provided by Heiko Schlittermann.
1172 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1173 Patch provided by Heiko Schlittermann.
1175 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1176 plus update to original patch.
1178 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1180 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1181 Patch provided by David Brownlee.
1183 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1185 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1187 NM/30 Bugzilla 888: TLS documentation bugfixes.
1189 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1191 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1192 Unlike the original bugzilla I have changed all shell scripts in src tree.
1194 NM/33 Bugzilla 898: Transport filter timeout fix.
1195 Patch by Todd Rinaldo.
1197 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1198 Patch by Serge Demonchaux.
1200 NM/35 Bugzilla 39: Base64 decode bug fixes.
1201 Patch by Jakob Hirsch.
1203 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1205 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1207 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1209 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1215 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1216 ALT-N's libdkim that I have put here:
1217 http://duncanthrax.net/exim-experimental/
1219 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1220 were called 'true' and it seems that C99 defines that as a reserved
1221 keyword to be used with 'bool' variable types. That means you could
1222 not include C99-style headers which use bools without triggering
1223 build errors in sieve.c.
1225 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1226 as mailq or other aliases. Changed the --help handling significantly
1227 to do whats expected. exim_usage() emits usage/help information.
1229 SC/01 Added the -bylocaldomain option to eximstats.
1231 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1233 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1235 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1241 PH/01 Another patch from the Sieve maintainer.
1243 PH/02 When an IPv6 address is converted to a string for single-key lookup
1244 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1245 dots are used instead of colons so that keys in lsearch files need not
1246 contain colons. This was done some time before quoting was made available
1247 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1248 (notated using the quote facility) so as to distinguish them from IPv4
1249 keys. This meant that lookups for IP addresses in host lists did not work
1250 for iplsearch lookups.
1252 This has been fixed by arranging for IPv6 addresses to be expressed with
1253 colons if the lookup type is iplsearch. This is not incompatible, because
1254 previously such lookups could never work.
1256 The situation is now rather anomolous, since one *can* have colons in
1257 ordinary lsearch keys. However, making the change in all cases is
1258 incompatible and would probably break a number of configurations.
1260 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1263 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1264 conversion specification without a maximum field width, thereby enabling
1265 a rogue spamd server to cause a buffer overflow. While nobody in their
1266 right mind would setup Exim to query an untrusted spamd server, an
1267 attacker that gains access to a server running spamd could potentially
1268 exploit this vulnerability to run arbitrary code as the Exim user.
1270 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1271 $primary_hostname instead of what libspf2 thinks the hosts name is.
1273 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1274 a directory entry by the name of the lookup key. Previously, if a
1275 symlink pointed to a non-existing file or a file in a directory that
1276 Exim lacked permissions to read, a lookup for a key matching that
1277 symlink would fail. Now it is enough that a matching directory entry
1278 exists, symlink or not. (Bugzilla 503.)
1280 PH/03 The body_linecount and body_zerocount variables are now exported in the
1283 PH/04 Added the $dnslist_matched variable.
1285 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1286 This means they are set thereafter only if the connection becomes
1289 PH/06 Added the client_condition to authenticators so that some can be skipped
1290 by clients under certain conditions.
1292 PH/07 The error message for a badly-placed control=no_multiline_responses left
1293 "_responses" off the end of the name.
1295 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1297 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1298 (without spaces) instead of just copying the configuration text.
1300 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1302 PH/11 Added $max_received_linelength.
1304 PH/12 Added +ignore_defer and +include_defer to host lists.
1306 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1307 way in which PCRE > 7.0 is built.
1309 PH/14 Implemented queue_only_load_latch.
1311 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1312 MAIL command. The effect was to mangle the value on 64-bit systems.
1314 PH/16 Another patch from the Sieve maintainer.
1316 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1318 PH/18 If a system quota error occurred while trying to create the file for
1319 a maildir delivery, the message "Mailbox is full" was not appended to the
1320 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1321 only to a quota excession during the actual writing of the file.
1323 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1324 characters?) which causes problems in log lines. The DN values are now
1325 passed through string_printing() before being added to log lines.
1327 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1328 and InterBase are left for another time.)
1330 PH/21 Added message_body_newlines option.
1332 PH/22 Guard against possible overflow in moan_check_errorcopy().
1334 PH/23 POSIX allows open() to be a macro; guard against that.
1336 PH/24 If the recipient of an error message contained an @ in the local part
1337 (suitably quoted, of course), incorrect values were put in $domain and
1338 $local_part during the evaluation of errors_copy.
1344 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1345 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1348 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1349 issue a MAIL command.
1351 PH/02 In an ACL statement such as
1353 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1355 if a client was not listed at all, or was listed with a value other than
1356 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1357 the condition was not true (as it should be), so access was not denied.
1358 The bug was that the ! inversion was incorrectly passed on to the second
1359 item. This has been fixed.
1361 PH/03 Added additional dnslists conditions == and =& which are different from
1362 = and & when the dns lookup returns more than one IP address.
1364 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1365 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1367 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1368 FSYNC, which compiles an option called disable_fsync that allows for
1369 bypassing fsync(). The documentation is heavily laced with warnings.
1371 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1373 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1374 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1375 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1376 including adding "make clean"; (3) Added -fPIC when compiling the test
1377 dynamically loaded module, to get rid of a warning.
1379 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1380 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1381 = 0s. The bug is otherwise harmless.
1383 PH/07 There was a bug in the dovecot authenticator such that the value of
1384 $auth1 could be overwritten, and so not correctly preserved, after a
1385 successful authentication. This usually meant that the value preserved by
1386 the server_setid option was incorrect.
1388 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1390 PH/09 Installed PCRE release 7.0.
1392 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1393 run for batched SMTP input. It is now run at the start of every message
1394 in the batch. While fixing this I discovered that the process information
1395 (output by running exiwhat) was not always getting set for -bs and -bS
1396 input. This is fixed, and it now also says "batched" for BSMTP.
1398 PH/11 Added control=no_pipelining.
1400 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1401 patch, slightly modified), and move the expansion of helo_data till after
1402 the connection is made in the smtp transport (so it can use these
1405 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1407 PH/14 Added log_selector = +pid.
1409 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1411 PH/16 Add ${if forany and ${if forall.
1413 PH/17 Added dsn_from option to vary the From: line in DSNs.
1415 PH/18 Flush SMTP output before performing a callout, unless control =
1416 no_callout_flush is set.
1418 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1419 was true (the default) a successful delivery failed to delete the retry
1420 item, thus causing premature timeout of the address. The bug is now
1423 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1425 PH/21 Long custom messages for fakedefer and fakereject are now split up
1426 into multiline reponses in the same way that messages for "deny" and
1427 other ACL rejections are.
1429 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1430 with slight modification.
1432 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1433 draft, changing the syntax and factoring some duplicate code".
1435 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1436 for deliveries of the second and subsequent messages over the same SMTP
1439 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1440 ${reduce, with only minor "tidies".
1442 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1444 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1445 expansion side effects.
1447 PH/27 When a message times out after an over-quota error from an Exim-imposed
1448 quota, the bounce message says "mailbox is full". This message was not
1449 being given when it was a system quota that was exceeded. It now should
1452 MH/03 Made $recipients available in local_scan(). local_scan() already has
1453 better access to the recipient list through recipients_list[], but
1454 $recipients can be useful in postmaster-provided expansion strings.
1456 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1457 in the case of a MAIL command with additional options following the
1458 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1459 were accidentally chopped off.
1461 PH/29 SMTP synchronization checks are implemented when a command is read -
1462 there is a check that no more input is waiting when there shouldn't be
1463 any. However, for some commands, a delay in an ACL can mean that it is
1464 some time before the response is written. In this time, more input might
1465 arrive, invalidly. So now there are extra checks after an ACL has run for
1466 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1467 pipelining has not been advertised.
1469 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1471 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1472 returned in the SMTP response when smtp_return_error_details was false.
1473 This has been fixed.
1475 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1476 instead of the C I/O that was originally supplied, because problems were
1477 reported on Solaris.
1479 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1480 Exim which did not show up earlier: it was assuming that a call to
1481 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1482 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1483 was a macro that became an assignment, so it seemed to work. This has
1484 changed to a proper function call with a void return, hence the compile
1485 error. Exim's code has been fixed.
1487 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1490 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1492 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1494 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1495 because it turns out to be too big or there is a timeout) there is no
1496 "Completed" line in the log. When some messages of this type were
1497 selected by exigrep, they were listed as "not completed". Others were
1498 picked up by some special patterns. I have improved the selection
1499 criteria to be more general.
1501 PH/38 The host_find_failed option in the manualroute router can now be set
1502 to "ignore", to completely ignore a host whose IP address cannot be
1503 found. If all hosts are ignored, the behaviour is controlled by the new
1504 host_all_ignored option.
1506 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1507 homing or because of multiple MX records with /mx) generated more than
1508 one IP address, and the following item turned out to be the local host,
1509 all the secondary addresses of the first item were incorrectly removed
1510 from the list, along with the local host and any following hosts (which
1511 is what is supposed to happen).
1513 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1514 whoever called Exim into the -H file. In the case of the daemon it was
1515 behaving confusingly. When first started, it used values for whoever
1516 started the daemon, but after a SIGHUP it used the Exim user (because it
1517 calls itself on a restart). I have changed the code so that it now always
1520 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1521 message are rejected with the same error (e.g. no authentication or bad
1522 sender address), and a DATA command is nevertheless sent (as can happen
1523 with PIPELINING or a stupid MUA), the error message that was given to the
1524 RCPT commands is included in the rejection of the DATA command. This is
1525 intended to be helpful for MUAs that show only the final error to their
1528 PH/42 Another patch from the Sieve maintainer.
1530 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1531 Eximstats - Fixed some broken HTML links and added missing column headers
1533 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1534 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1536 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1538 PH/43 Yet another patch from the Sieve maintainer.
1540 PH/44 I found a way to check for a TCP/IP connection going away before sending
1541 the response to the final '.' that terminates a message, but only in the
1542 case where the client has not sent further data following the '.'
1543 (unfortunately, this is allowed). However, in many cases there won't be
1544 any further data because there won't be any more messages to send. A call
1545 to select() can be used: if it shows that the input is "ready", there is
1546 either input waiting, or the socket has been closed. An attempt to read
1547 the next input character can distinguish the two cases. Previously, Exim
1548 would have sent an OK response which the client would never have see.
1549 This could lead to message repetition. This fix should cure that, at
1550 least in a lot of common cases.
1552 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1553 advertised in response to EHLO.
1559 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1560 fixed by 4.65/MH/01 (is this a record?) are fixed:
1562 (i) An empty string was always treated as zero by the numeric comparison
1563 operators. This behaviour has been restored.
1565 (ii) It is documented that the numeric comparison operators always treat
1566 their arguments as decimal numbers. This was broken in that numbers
1567 starting with 0 were being interpreted as octal.
1569 While fixing these problems I realized that there was another issue that
1570 hadn't been noticed. Values of message_size_limit (both the global option
1571 and the transport option) were treated as octal if they started with 0.
1572 The documentation was vague. These values are now always treated as
1573 decimal, and I will make that clear in the documentation.
1579 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1580 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1583 MH/01 Don't check that the operands of numeric comparison operators are
1584 integers when their expansion is in "skipping" mode (fixes bug
1585 introduced by 4.64-PH/07).
1587 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1588 child addresses, Exim now panics and dies. Previously, because the count
1589 is held in a short int, deliveries were likely to be lost. As such a
1590 large number of recipients for a single message is ridiculous
1591 (performance will be very, very poor), I have chosen to impose a limit
1592 rather than extend the field.
1598 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1599 leftover -K file (the existence of which was triggered by #402).
1600 While we were at it, introduced process PID as part of the -K
1601 filename. This should rule out race conditions when creating
1604 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1605 processing considerably. Previous code took too long for large mails,
1606 triggering a timeout which in turn triggers #401.
1608 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1609 in the DK code in transports.c. sendfile() is not really portable,
1610 hence the _LINUX specificness.
1612 TF/01 In the add_headers option to the mail command in an Exim filter,
1613 there was a bug that Exim would claim a syntax error in any
1614 header after the first one which had an odd number of characters
1617 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1618 callout verification, Exim cached a "reject" for the entire domain. This
1619 is correct for most verifications, but it is not correct for a recipient
1620 verification with use_sender or use_postmaster set, because in that case
1621 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1622 case of MAIL FROM:<> rejection from other early rejections (e.g.
1623 rejection of HELO). When verifying a recipient using a non-null MAIL
1624 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1625 Whatever the result of the callout, the value of the domain cache is
1626 left unchanged (for any other kind of callout, getting as far as trying
1627 RCPT means that the domain itself is ok).
1629 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1632 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1633 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1634 session, the daemon ignores ECONNECT errors and logs others; it now
1635 ignores EPIPE as well.
1637 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1638 (quoted-printable decoding).
1640 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1641 later the small subsequent patch to fix an introduced bug.
1643 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1645 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1647 PH/08 An error is now given if message_size_limit is specified negative.
1649 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1650 to be given (somewhat) arbitrary names.
1652 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1655 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1656 miscellaneous code fixes
1658 PH/10 Added the log_reject_target ACL modifier to specify where to log
1661 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1662 hostname. This is wrong, because it relates to the incoming message (and
1663 probably the interface on which it is arriving) and not to the outgoing
1664 callout (which could be using a different interface). This has been
1665 changed to use the value of the helo_data option from the smtp transport
1666 instead - this is what is used when a message is actually being sent. If
1667 there is no remote transport (possible with a router that sets up host
1668 addresses), $smtp_active_hostname is used.
1670 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1671 tweaks were necessary in order to get it to work (see also 21 below):
1672 (a) The code assumed that strncpy() returns a negative number on buffer
1673 overflow, which isn't the case. Replaced with Exim's string_format()
1675 (b) There were several signed/unsigned issues. I just did the minimum
1676 hacking in of casts. There is scope for a larger refactoring.
1677 (c) The code used strcasecmp() which is not a standard C function.
1678 Replaced with Exim's strcmpic() function.
1679 (d) The code set only $1; it now sets $auth1 as well.
1680 (e) A simple test gave the error "authentication client didn't specify
1681 service in request". It would seem that Dovecot has changed its
1682 interface. Fortunately there's a specification; I followed it and
1683 changed what the client sends and it appears to be working now.
1685 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1688 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1689 address A is aliased to B and C, where B exists and C does not. Without
1690 -v the output is "A verified" because verification stops after a
1691 successful redirection if more than one address is generated. However,
1692 with -v the child addresses are also verified. Exim was outputting "A
1693 failed to verify" and then showing the successful verification for C,
1694 with its parentage. It now outputs "B failed to verify", showing B's
1695 parentage before showing the successful verification of C.
1697 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1698 look up a TXT record in a specific list after matching in a combined
1701 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1702 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1703 they consult the DNS. I had assumed they would set it the way they
1704 wanted; and indeed my experiments on Linux seem to show that in some
1705 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1706 To be on the safe side, however, I have now made the interface to
1707 host_find_byname() similar to host_find_bydns(), with an argument
1708 containing the DNS resolver options. The host_find_byname() function now
1709 sets these options at its start, just as host_find_bydns() does. The smtp
1710 transport options dns_qualify_single and dns_search_parents are passed to
1711 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1712 of host_find_byname() use the default settings of RES_DEFNAMES
1713 (qualify_single) but not RES_DNSRCH (search_parents).
1715 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1716 spool_read_header() do less string testing, by means of a preliminary
1717 switch on the second character of optional "-foo" lines. (This is
1718 overdue, caused by the large number of possibilities that now exist.
1719 Originally there were few.) While I was there, I also converted the
1720 str(n)cmp tests so they don't re-test the leading "-" and the first
1721 character, in the hope this might squeeze out yet more improvement.
1723 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1724 flag allowing group syntax was set by the header_syntax check but not
1725 turned off, possible causing trouble later; (2) The flag was not being
1726 set at all for the header_verify test, causing "group"-style headers to
1727 be rejected. I have now set it in this case, and also caused header_
1728 verify to ignore an empty address taken from a group. While doing this, I
1729 came across some other cases where the code for allowing group syntax
1730 while scanning a header line wasn't quite right (mostly, not resetting
1731 the flag correctly in the right place). These bugs could have caused
1732 trouble for malformed header lines. I hope it is now all correct.
1734 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1735 with the "reply" argument non-NULL. The code, however (which originally
1736 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1737 but it didn't always do it. This confused somebody who was copying the
1738 code for some other use. I have removed all the tests.
1740 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1741 feature that was used to support insecure browsers during the U.S. crypto
1742 embargo. It requires special client support, and Exim is probably the
1743 only MTA that supported it -- and would never use it because real RSA is
1744 always available. This code has been removed, because it had the bad
1745 effect of slowing Exim down by computing (never used) parameters for the
1746 RSA_EXPORT functionality.
1748 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1749 authenticator to fail if there's a tab character in the incoming data
1750 (there should never be unless someone is messing about, as it's supposed
1751 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1752 if the connection is using TLS or if the remote IP is the same as the
1753 local IP, and the "valid-client-cert option" if a client certificate has
1756 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1757 authenticators. This can be used for authorization after authentication
1758 succeeds. (In the case of plaintext, it servers for both authentication
1761 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1762 if any retry times were supplied.
1764 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1765 connection, where there is no client IP address to check. In this
1766 situation, the verify now always succeeds.
1768 PH/25 Applied John Jetmore's -Mset patch.
1770 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1772 PH/27 In a string expansion for a processed (not raw) header when multiple
1773 headers of the same name were present, leading whitespace was being
1774 removed from all of them, but trailing whitespace was being removed only
1775 from the last one. Now trailing whitespace is removed from each header
1776 before concatenation. Completely empty headers in a concatenation (as
1777 before) are ignored.
1779 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1780 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1782 PH/29 [Removed. This was a change that I later backed out, and forgot to
1783 correct the ChangeLog entry (that I had efficiently created) before
1784 committing the later change.]
1786 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1787 address errors (4xx response to RCPT) over the same connection as other
1788 messages routed to the same hosts. Such deliveries are always "forced",
1789 so retry times are not inspected. This resulted in far too many retries
1790 for the affected addresses. The effect occurred only when there were more
1791 hosts than the hosts_max_try setting in the smtp transport when it had
1792 the 4xx errors. Those hosts that it had tried were not added to the list
1793 of hosts for which the message was waiting, so if all were tried, there
1794 was no problem. Two fixes have been applied:
1796 (i) If there are any address or message errors in an SMTP delivery, none
1797 of the hosts (tried or untried) are now added to the list of hosts
1798 for which the message is waiting, so the message should not be a
1799 candidate for sending over the same connection that was used for a
1800 successful delivery of some other message. This seems entirely
1801 reasonable: after all the message is NOT "waiting for some host".
1802 This is so "obvious" that I'm not sure why it wasn't done
1803 previously. Hope I haven't missed anything, but it can't do any
1804 harm, as the worst effect is to miss an optimization.
1806 (ii) If, despite (i), such a delivery is accidentally attempted, the
1807 routing retry time is respected, so at least it doesn't keep
1808 hammering the server.
1810 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1811 in ${readsocket because some servers need this prod.
1813 PH/32 Added some extra debug output when updating a wait-xxx database.
1815 PH/33 The hint "could be header name not terminated by colon", which has been
1816 given for certain expansion errors for a long time, was not being given
1817 for the ${if def:h_colon_omitted{... case.
1819 PH/34 The spec says: "With one important exception, whenever a domain list is
1820 being scanned, $domain contains the subject domain." There was at least
1821 one case where this was not true.
1823 PH/35 The error "getsockname() failed: connection reset by peer" was being
1824 written to the panic log as well as the main log, but it isn't really
1825 panic-worthy as it just means the connection died rather early on. I have
1826 removed the panic log writing for the ECONNRESET error when getsockname()
1829 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1830 runs only) independently of the message's sender address. This meant
1831 that, if the 4xx error was in fact related to the sender, a different
1832 message to the same recipient with a different sender could confuse
1833 things. In particualar, this can happen when sending to a greylisting
1834 server, but other circumstances could also provoke similar problems.
1835 I have changed the default so that the retry time for these errors is now
1836 based a combination of the sender and recipient addresses. This change
1837 can be overridden by setting address_retry_include_sender=false in the
1840 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1841 remote server are returned as part of bounce messages. This was not
1842 happening for LMTP over a pipe (the lmtp transport), but now it is the
1843 same for both kinds of LMTP.
1845 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1846 in header lines that were in fact CNAMEs. This is no longer the case.
1848 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1849 and queue runs started by the daemon processed all messages. This has
1850 been fixed so that -R and -S can now usefully be given with -q<time>.
1852 PH/40 Import PCRE release 6.7 (fixes some bugs).
1854 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1856 PH/42 Give an error if -q is specified more than once.
1858 PH/43 Renamed the variables $interface_address and $interface_port as
1859 $received_ip_address and $received_port, to make it clear that these
1860 values apply to message reception, and not to the outgoing interface when
1861 a message is delivered. (The old names remain recognized, of course.)
1863 PH/44 There was no timeout on the connect() call when using a Unix domain
1864 socket in the ${readsocket expansion. There now is.
1866 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1867 be meaningful with "accept".
1869 SC/01 Eximstats V1.43
1870 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1872 SC/02 Eximstats V1.44
1873 Use a glob alias rather than an array ref in the generated
1874 parser. This improves both readability and performance.
1876 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1877 Collect SpamAssassin and rejection statistics.
1878 Don't display local sender or destination tables unless
1879 there is data to show.
1880 Added average volumes into the top table text output.
1882 SC/04 Eximstats V1.46
1883 Collect data on the number of addresses (recipients)
1884 as well as the number of messages.
1886 SC/05 Eximstats V1.47
1887 Added 'Message too big' to the list of mail rejection
1888 reasons (thanks to Marco Gaiarin).
1890 SC/06 Eximstats V1.48
1891 Mainlog lines which have GMT offsets and are too short to
1892 have a flag are now skipped.
1894 SC/07 Eximstats V1.49 (Alain Williams)
1895 Added the -emptyok flag.
1897 SC/08 Eximstats V1.50
1898 Fixes for obtaining the IP address from reject messages.
1900 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1901 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1902 whitesspace changes from 4.64-PH/27
1904 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1907 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1908 are found, allow negative numbers in numeric criteria)
1910 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1912 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1913 to match changes made in 4.64-PH/43
1915 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1917 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1918 "domain" is actually an IP address, and force a failure. This locks out
1919 those revolvers/nameservers that support "A-for-A" lookups, in
1920 contravention of the specifications.
1922 PH/48 When a host name was looked up from an IP address, and the subsequent
1923 forward lookup of the name timed out, the host name was left in
1924 $sender_host_name, contrary to the specification.
1926 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1927 restricted to single-key lookups, Exim was not diagnosing an error if
1928 * or *@ was used with a query-style lookup.
1930 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1932 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1933 long ago, but noone interested enough thought of it. Let's just say that
1934 the "1.1" means that there are some new functions that weren't there at
1935 some point in the past.
1937 PH/51 Error processing for expansion failure of helo_data from an smtp
1938 transport during callout processing was broken.
1940 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1941 tested/used via the -bh/-bhc/-bs options.
1943 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1944 bug, fixed in subsequent PCRE releases).
1946 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1947 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1949 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1955 SC/01 Use a glob alias rather than an array ref in eximstats generated
1956 parser. This improves both readability and performance.
1958 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1959 Don't display local sender or destination tables in eximstats unless
1960 there is data to show.
1961 Added average volumes into the eximstats top table text output.
1963 SC/03 Collect data on the number of addresses (recipients) as well
1964 as the number of messages in eximstats.
1966 TF/01 Correct an error in the documentation for the redirect router. Exim
1967 does (usually) call initgroups() when daemonizing.
1969 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1970 with consistent privilege compared to when running as a daemon.
1972 TF/03 Note in the spec that $authenticated_id is not set for local
1973 submissions from trusted users.
1975 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1976 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1978 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1979 by adding some example configuration directives to the default
1980 configuration file. A little bit of work is required to uncomment the
1981 directives and define how usernames and passwords are checked, but
1982 there is now a framework to start from.
1984 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1985 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1986 without this. I don't know how relevant this is to other LDAP libraries.
1988 PH/02 Add the verb name to the "unknown ACL verb" error.
1990 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1992 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1994 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1995 directory not expanded when it should be if an expanded home directory
1996 was set for the address (which is overridden by the transport).
1998 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2001 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2002 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2003 because it is too late at that time, and has no effect.
2005 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2006 security issue with \' (bugzilla #107). I could not use the
2007 PQescapeStringConn() function, because it needs a PGconn value as one of
2010 PH/08 When testing addresses using -bt, indicate those final addresses that
2011 are duplicates that would not cause an additional delivery. At least one
2012 person was confused, thinking that -bt output corresponded to deliveries.
2013 (Suppressing duplicates isn't a good idea as you lose the information
2014 about possibly different redirections that led to the duplicates.)
2016 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2017 systems where poll() doesn't work, in particular OS X.
2019 PH/10 Added more information to debugging output for retry time not reached.
2021 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2022 operations in malware.c.
2024 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2027 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2028 syslog_duplication set false, log lines that would normally be written
2029 both the the main log and to the reject log were not written to syslog at
2032 PH/14 In the default configuration, change the use of "message" in ACL warn
2033 statements to "add_header".
2035 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2036 not followed by a command (e.g. "seen endif").
2038 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2039 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2042 PH/17 Added extra conditions to the default value of delay_warning_condition
2046 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2047 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2048 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2051 The Auto-Submitted: and various List- headers are standardised, whereas I
2052 don't think Precedence: ever was.
2054 PH/18 Refactored debugging code in route_finduser() to show more information,
2055 in particular, the error code if getpwnam() issues one.
2057 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2058 This is apparently needed in addition to the PH/07 change above to avoid
2059 any possible encoding problems.
2061 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2062 but not after initializing Perl.
2064 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2065 output them only if debugging. By default they are written stderr,
2066 apparently, which is not desirable.
2068 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2071 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2074 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2076 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2077 authenticated or an ident call has been made. Suppress the default
2078 values for $authenticated_id and $authenticated_sender (but permit -oMai
2079 and -oMas) when testing with -bh.
2081 PH/24 Re-jigged the order of the tests in the default configuration so that the
2082 tests for valid domains and recipients precede the DNS black list and CSA
2083 tests, on the grounds that those ones are more expensive.
2085 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2086 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2087 This bug exists in every version of Exim that I still have, right back to
2090 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2091 However, an attempt to turn on case-sensitivity in a regex key by
2092 including (?-i) didn't work because the subject string was already
2093 lowercased, and the effects were non-intuitive. It turns out that a
2094 one-line patch can be used to allow (?-i) to work as expected.
2100 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2101 other effects) broke the use of negated acl sub-conditions.
2103 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2106 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2107 "Deny" causes Exim to reject the incoming connection with a 554 error.
2108 Unfortunately, if there is a major crisis, such as a disk failure,
2109 tcp-wrappers gives "deny", whereas what one would like would be some
2110 kind of temporary error. A kludge has been added to help with this.
2111 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2112 554 error is used if errno is still zero or contains ENOENT (which occurs
2113 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2116 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2118 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2119 errors. Otherwise a message that provokes a temporary error (when other
2120 messages do not) can cause a whole host to time out.
2122 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2123 addresses were routed directly to files or pipes from a redirect router.
2124 File deliveries just didn't batch; pipe deliveries might have suffered
2127 PH/06 A failure to get a lock for a hints database would erroneously always say
2128 "Failed to get write lock", even when it was really a read lock.
2130 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2131 of 0600. This has been changed to use the value of the lockfile_mode
2132 option (which defaults to 0600).
2134 PH/08 Applied small patch from the Sieve maintainer.
2136 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2137 folder from quota calculations, a direct delivery into this folder messed
2138 up the contents of the maildirsize file. This was because the regex was
2139 used only to exclude .Trash (or whatever) when the size of the mailbox
2140 was calculated. There was no check that a delivery was happening into an
2141 excluded directory. This bug has been fixed by ignoring all quota
2142 processing for deliveries into excluded directories.
2144 PH/10 Added the maildirfolder_create_regex option to appendfile.
2150 PH/01 The code for finding all the local interface addresses on a FreeBSD
2151 system running IPv6 was broken. This may well have applied to all BSD
2152 systems, as well as to others that have similar system calls. The broken
2153 code found IPv4 interfaces correctly, but gave incorrect values for the
2154 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2155 that it would not match correctly against @[] and not recognize the IPv6
2158 PH/02 The ipliteral router was not recognizing addresses of the form user@
2159 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2161 PH/03 Added disable_ipv6.
2163 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2164 decoded form, because it is most often used to construct To: headers
2165 lines in autoreplies, and the decoded form may well be syntactically
2166 invalid. However, $reply_address has leading white space removed, and all
2167 newlines turned into spaces so that the autoreply transport does not
2170 PH/05 If group was specified without a user on a router, and no group or user
2171 was specified on a transport, the group from the router was ignored.
2173 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2174 for visible compile-time settings that can be used to change these
2175 numbers, for those that want even more. Backwards compatibility with old
2176 spool files has been maintained. However, going back to a previous Exim
2177 release will lost any variables that are in spool files.
2179 PH/07 Two small changes when running in the test harness: increase delay when
2180 passing a TCP/IP connection to a new process, in case the original
2181 process has to generate a bounce, and remove special handling of
2182 127.0.0.2 (sic), which is no longer necessary.
2184 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2185 be the same on different OS.
2187 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2190 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2191 whether --show-vars was specified or not
2193 JJ/02 exipick: Added support for new ACL variable spool format introduced
2196 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2197 syntactically invalid From: or Reply-to: line, and a filter used this to
2198 generate an autoreply, and therefore failed to obtain an address for the
2199 autoreply, Exim could try to deliver to a non-existent relative file
2200 name, causing unrelated and misleading errors. What now happens is that
2201 it logs this as a hard delivery error, but does not attempt to create a
2204 PH/11 The exinext utility has a -C option for testing purposes, but although
2205 the given file was scanned by exinext itself; it wasn't being passed on
2206 when Exim was called.
2208 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2209 an end-of-file indication when reading a command response.
2211 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2212 compiled. In many other places in Exim, IPv6 addresses are always
2213 recognized, so I have changed this. It also means that IPv4 domain
2214 literals of the form [IPV4:n.n.n.n] are now always recognized.
2216 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2217 used if the router is not running as root, for example, when verifying at
2218 ACL time, or when using -bh. The debugging output from this situation was
2219 non-existent - all you got was a failure to exec. I have made two
2222 (a) Failures to set uid/gid, the current directory, or a process leader
2223 in a subprocess such as that created by queryprogram now generate
2224 suitable debugging ouput when -d is set.
2226 (b) The queryprogram router detects when it is not running as root,
2227 outputs suitable debugging information if -d is set, and then runs
2228 the subprocess without attempting to change uid/gid.
2230 PH/15 Minor change to Makefile for building test_host (undocumented testing
2233 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2234 additional section of a DNS packet that returns MX or SRV records.
2235 Instead, it always explicitly searches for A/AAAA records. This avoids
2236 major problems that occur when a DNS server includes only records of one
2237 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2238 fixed another bug: if SRV records were looked up and the corresponding
2239 address records were *not* found in the additional section, the port
2240 values from the SRV records were lost.
2242 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2243 using the correct key (the original address) when searching the retry
2244 rules in order to find which one to use for generating the retry hint.
2246 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2247 adding the default one. Similarly, if it contains a Reply-To: header, the
2248 errors_reply_to option, if set, is not used.
2250 PH/19 When calculating a retry time, Exim used to measure the "time since
2251 failure" by looking at the "first failed" field in the retry record. Now
2252 it does not use this if it is later than than the arrival time of the
2253 message. Instead it uses the arrival time. This makes for better
2254 behaviour in cases where some deliveries succeed, thus re-setting the
2255 "first failed" field. An example is a quota failure for a huge message
2256 when small messages continue to be delivered. Without this change, the
2257 "time since failure" will always be short, possible causing more frequent
2258 delivery attempts for the huge message than are intended.
2259 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2261 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2262 $1, $2, $3) because the numerical variables can be reset during some
2263 expansion items (e.g. "match"), thereby losing the authentication data.
2265 PH/21 Make -bV show the size of off_t variables so that the test suite can
2266 decide whether to run tests for quotas > 2G.
2268 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2269 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2270 filecount value is greater than 2G or if a quota value is greater than 2G
2271 on a system where the size of off_t is not greater than 4, a panic error
2274 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2275 never match. The debug and -bh output now contains an explicit error
2276 message indicating a malformed IPv4 address or mask.
2278 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2279 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2280 PH/23 above applies.
2282 PH/25 Do not write to syslog when running in the test harness. The only
2283 occasion when this arises is a failure to open the main or panic logs
2284 (for which there is an explicit test).
2286 PH/26 Added the /no_tell option to "control=freeze".
2288 PH/27 If a host name lookup failed very early in a connection, for example, if
2289 the IP address matched host_lookup and the reverse lookup yielded a name
2290 that did not have a forward lookup, an error message of the form "no IP
2291 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2292 could be logged. Now it outputs the IP address instead of "NULL".
2294 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2295 allows the sender and the authenticated sender to be set when
2296 submitting a message from within Exim. Since child_open_exim() is
2297 documented for local_scan(), the new function should be too.
2299 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2300 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2301 results in an empty string is now treated as unset.
2303 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2305 PH/31 Added sender_verify_fail logging option.
2307 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2308 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2309 tidied the source and removed it altogether.
2311 PH/33 When a queue run was abandoned because the load average was too high, a
2312 log line was always written; now it is written only if the queue_run log
2313 selector is set. In addition, the log line for abandonment now contains
2314 information about the queue run such as the pid. This is always present
2315 in "start" and "stop" lines but was omitted from the "abandon" line.
2317 PH/34 Omit spaces between a header name and the colon in the error message that
2318 is given when verify = headers_syntax fails (if there are lots of them,
2319 the message gets confusing).
2321 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2322 names, as there are now some PTR records that contain slashes. This check
2323 is only to protect against broken name servers that fall over on strange
2324 characters, so the fact that it applies to all lookups doesn't matter.
2326 PH/36 Now that the new test suite is complete, we can remove some of the
2327 special code in Exim that was needed for the old test suite. For example,
2328 sorting DNS records because real resolvers return them in an arbitrary
2329 order. The new test suite's fake resolver always returns records in the
2332 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2333 bounces) except when queue_only is set, to avoid logging races between
2334 the different processes.
2336 PH/38 Panic-die if .include specifies a non-absolute path.
2338 PH/39 A tweak to the "H" retry rule from its user.
2340 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2341 a label. They prevented compilation on older perls.
2343 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2344 a warning to be raised on newish perls.
2346 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2347 on queue. Changes to match documented behaviour of showing count of
2348 messages matching specified criteria.
2350 PH/40 Changed the default ident timeout from 30s to 5s.
2352 PH/41 Added support for the use of login_cap features, on those BSD systems
2353 that have them, for controlling the resources used by pipe deliveries.
2355 PH/42 The content-scanning code uses fopen() to create files in which to put
2356 message data. Previously it was not paying any attention to the mode of
2357 the files. Exim runs with umask(0) because the rest of the code creates
2358 files with open(), and sets the required mode explicitly. Thus, these
2359 files were ending up world-writeable. This was not a big issue, because,
2360 being within the spool directory, they were not world-accessible. I have
2361 created a function called modefopen, which takes an additional mode
2362 argument. It sets umask(777), creates the file, chmods it to the required
2363 mode, then resets the umask. All the relevant calls to fopen() in the
2364 content scanning code have been changed to use this function.
2366 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2367 to 24 hours. This avoids potential overflow problems when processing G
2368 and H retry rules. I suspect nobody ever tinkers with this value.
2370 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2372 PH/45 When the plaintext authenticator is running as a client, the server's
2373 challenges are checked to ensure they are valid base64 strings. By
2374 default, the authentication attempt is cancelled if an invalid string is
2375 received. Setting client_ignore_invalid_base64 true ignores these errors.
2376 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2377 they are received. Thus, the responses can be made to depend on the
2378 challenges. If an invalid string is ignored, an empty string is placed in
2381 PH/46 Messages that are created by the autoreply transport now contains a
2382 References: header, in accordance with RFCs 2822 and 3834.
2384 PH/47 Added authenticated_sender_force to the smtp transport.
2386 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2388 PH/49 Installed latest patch from the Sieve maintainer.
2390 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2391 was also set, the appendfile transport was unnecessarily scanning a
2392 directory of message files (e.g. for maildir delivery) to find the count
2393 of files (along with the size), even though it did not need this
2394 information. It now does the scan only if it needs to find either the
2395 size of the count of files.
2397 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2399 PH/52 Two bugs concerned with error handling when the smtp transport is
2402 (i) Exim was not creating retry information for temporary errors given
2403 for individual recipients after the DATA command when the smtp transport
2404 was used in LMTP mode. This meant that they could be retried too
2405 frequently, and not timed out correctly.
2407 (ii) Exim was setting the flag that allows error details to be returned
2408 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2409 recipients that were returned after the DATA command.
2411 PH/53 This is related to PH/52, but is more general: for any failing address,
2412 when detailed error information was permitted to be returned to the
2413 sender, but the error was temporary, then after the final timeout, only
2414 "retry timeout exceeded" was returned. Now it returns the full error as
2415 well as "retry timeout exceeded".
2417 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2418 do this, and (what is worse) MTAs that accept it.
2420 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2421 will now be deprecated.
2423 PH/56 New os.c-cygwin from the Cygwin maintainer.
2425 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2426 formats (previously only available in exim formats via -bpr, -bpru,
2427 and -bpra. Now also available in native and exiqgrep formats)
2429 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2430 with very large, slow to parse queues
2432 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2434 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2436 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2437 responses to authentication challenges, though it was showing the
2438 challenges; (ii) I've removed the CR characters from the debug output for
2441 PH/58 Allow for the insertion of a newline as well as a space when a string
2442 is turned into more than one encoded-word during RFC 2047 encoding. The
2443 Sieve code now uses this.
2445 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2446 data_4xx, lost_connection, tls_required.
2448 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2449 message was being sent as an SMTP response.
2451 PH/61 Add -l and -k options to exicyclog.
2453 PH/62 When verifying, if an address was redirected to one new address, so that
2454 verification continued, and the new address failed or deferred after
2455 having set something in $address_data, the value of $address_data was not
2456 passed back to the ACL. This was different to the case when no
2457 redirection occurred. The value is now passed back in both cases.
2459 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2460 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2461 use login_cap.h, so on its own it isn't the distinguishing feature. The
2462 new name refers directly to the setclassresources() function.
2464 PH/65 Added configuration files for NetBSD3.
2466 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2468 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2469 is preferred over IPv4.
2471 PH/68 The bounce_return_message and bounce_return_body options were not being
2472 honoured for bounces generated during the reception of non-SMTP messages.
2473 In particular, this applied to messages rejected by the ACL. This bug has
2474 been fixed. However, if bounce_return_message is true and bounce_return_
2475 body is false, the headers that are returned for a non-SMTP message
2476 include only those that have been read before the error was detected.
2477 (In the case of an ACL rejection, they have all been read.)
2479 PH/69 The HTML version of the specification is now built in a directory called
2480 spec_html instead of spec.html, because the latter looks like a path with
2481 a MIME-type, and this confuses some software.
2483 PH/70 Catch two compiler warnings in sieve.c.
2485 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2486 function verify_get_ident() calls ip_connect() to connect a socket, but
2487 if the "connect()" function timed out, ip_connect() used to close the
2488 socket. However, verify_get_ident() also closes the socket later, and in
2489 between Exim writes to the log, which may get opened at this point. When
2490 the socket was closed in ip_connect(), the log could get the same file
2491 descriptor number as the socket. This naturally causes chaos. The fix is
2492 not to close the socket in ip_connect(); the socket should be closed by
2493 the function that creates it. There was only one place in the code where
2494 this was missing, in the iplookup router, which I don't think anybody now
2495 uses, but I've fixed it anyway.
2497 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2498 well as to direct DNS lookups. Otherwise the handling of names in host
2499 lists is inconsistent and therefore confusing.
2505 PH/01 Two changes to the default runtime configuration:
2507 (1) Move the checks for relay_from_hosts and authenticated clients from
2508 after to before the (commented out) DNS black list checks.
2510 (2) Add control=submission to the relay_from_hosts and authenticated
2511 clients checks, on the grounds that messages accepted by these
2512 statements are most likely to be submissions.
2514 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2516 (1) Generate an error if the third argument for the ${prvs expansion is
2519 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2522 (3) Reset the variables that are obtained from the first argument of
2523 ${prvscheck and used in the second argument before leaving the code,
2524 because their memory is reclaimed, so using them afterwards may do
2527 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2528 one (it's much easier than Tom thought :-).
2530 (5) Because of (4), we can now allow for the use of $prvscheck_result
2531 inside the third argument.
2533 PH/03 For some reason, the default setting of PATH when running a command from
2534 a pipe transport was just "/usr/bin". I have changed it to
2537 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2538 anything to be listed in the output from -bV.
2540 PH/05 When a filter generated an autoreply, the entire To: header line was
2541 quoted in the delivery log line, like this:
2543 => >A.N.Other <ano@some.domain> <original@ddress> ...
2545 This has been changed so that it extracts the operative address. There
2546 may be more than one such address. If so, they are comma-separated, like
2549 => >ano@some.domain,ona@other.domain <original@ddress> ...
2551 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2552 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2553 not being looked up in the rDNS to get a host name, Exim was showing the
2554 IP address twice in Received: lines, even though the IP addresses were
2555 identical. For example:
2557 Received: from [1.2.3.4] (helo=[1.2.3.4])
2559 However, if the real host name was known, it was omitting the HELO data
2560 if it matched the actual IP address. This has been tidied up so that it
2561 doesn't show the same IP address twice.
2563 PH/07 When both +timestamp and +memory debugging was on, the value given by
2564 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2565 called by the debug printing, thereby overwriting the timestamp buffer.
2566 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2568 PH/08 When the original message was included in an autoreply transport, it
2569 always said "this is a copy of the message, including all the headers",
2570 even if body_only or headers_only was set. It now gives an appropriate
2573 PH/09 Applied a patch from the Sieve maintainer which:
2575 o fixes some comments
2576 o adds the (disabled) notify extension core
2577 o adds some debug output for the result of if/elsif tests
2578 o points to the current vacation draft in the documentation
2579 and documents the missing references header update
2583 o fixes a bug in processing the envelope test (when testing
2584 multiple envelope elements, the last element determinted the
2587 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2588 Electronic Mail") by including:
2590 Auto-submitted: auto-generated
2592 in the messages that it generates (bounce messages and others, such as
2593 warnings). In the case of bounce messages for non-SMTP mesages, there was
2594 also a typo: it was using "Auto_submitted" (underscore instead of
2595 hyphen). Since every message generated by Exim is necessarily in response
2596 to another message, thes have all been changed to:
2598 Auto-Submitted: auto-replied
2600 in accordance with these statements in the RFC:
2602 The auto-replied keyword:
2604 - SHOULD be used on messages sent in direct response to another
2605 message by an automatic process,
2607 - MUST NOT be used on manually-generated messages,
2609 - MAY be used on Delivery Status Notifications (DSNs) and Message
2610 Disposition Notifications (MDNs),
2612 - MUST NOT be used on messages generated by automatic or periodic
2613 processes, except for messages which are automatic responses to
2616 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2617 to the default Received: header definition.
2619 PH/12 Added log selector acl_warn_skipped (default on).
2621 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2622 variables because (a) they are in the wrong storage pool and (b) even if
2623 they were copied, it wouldn't work properly because of the caching.
2625 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2626 checking when decoding. Apparently there are clients that generate
2627 overlong encoded strings. Why am I not surprised?
2629 PH/15 If the first argument of "${if match_address" was not empty, but did not
2630 contain an "@" character, Exim crashed. Now it writes a panic log message
2631 and treats the condition as false.
2633 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2635 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2636 extension "envelope-auth". The code is finished and in agreement with
2637 other implementations, but there is no documentation so far and in fact,
2638 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2639 not changing the active code.
2641 Print executed "if" and "elsif" statements when debugging is used. This
2642 helps a great deal to understand what a filter does.
2644 Document more things not specified clearly in RFC3028. I had all this
2645 sorted out, when out of a sudden new issues came to my mind. Oops."
2647 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2650 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2651 canonical form (as documented). However, after a host name lookup from
2652 the IP address, check_host() was doing a simple string comparison with
2653 addresses acquired from the DNS when checking that the found name did
2654 have the original IP as one of its addresses. Since any found IPv6
2655 addresses are likely to be in abbreviated form, the comparison could
2656 fail. Luckily, there already exists a function for doing the comparison
2657 by converting both addresses to binary, so now that is used instead of
2658 the text comparison.
2660 PH/20 There was another similar case to PH/19, when a complete host name was
2661 given in a host list; looking up its IP address could give an abbreviated
2662 form, whereas the current host's name might or might not be abbreviated.
2663 The same fix has been applied.
2669 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2670 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2673 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2674 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2676 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2678 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2679 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2680 Also, for "Auto-Submitted", treat anything other than "no" as
2681 non-personal, in accordance with RFC 3834. (Previously it treated
2682 anything starting "auto-" as non-personal.)
2684 TF/01 The control=submission/name=... option had a problem with syntax
2685 errors if the name included a slash character. The /name= option
2686 now slurps the rest of the string, so it can include any characters
2687 but it must come last in the list of options (after /sender_retain
2690 PH/05 Some modifications to the interface to the fake nameserver for the new
2698 TK/01 Added the "success_on_redirect" address verification option. See
2699 NewStuff for rationale and an example.
2701 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2703 PH/02 Patch to exigrep to allow it to work on syslog lines.
2705 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2706 fread() to skip over the body file's header line, because in Cygwin the
2707 header line is locked and is inaccessible.
2709 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2710 co-exist for some time) to make it clear that it is the Exim ID that is
2711 referenced, not the Message-ID: header line.
2713 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2714 string_format() function, because snprintf() does not exist on all
2717 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2718 ${stat: expansion item.
2720 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2721 protocol synchronization error", to keep the pedants happy.
2723 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2724 well as for IRIX systems, when gcc is being used. See the host.c source
2727 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2729 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2732 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2733 transport and to the smtp transport in LMTP mode.
2735 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2737 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2738 run a filter in a subprocess. This could lead to confusion in subsequent
2739 lookups in the parent process. There should also be a search_tidyup() at
2740 the end of the subprocess.
2742 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2743 only if the host matched helo_try_verify_hosts, which caused the
2744 verification to occur when the EHLO/HELO command was issued. The ACL just
2745 tested the remembered result. Now, if a previous verification attempt has
2746 not happened, "verify = helo" does it there and then.
2748 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2750 TK/03 Fix log output including CR from clamd.
2752 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2753 exist provoked a memory error which could cause a segfault.
2755 PH/15 Installed PCRE 6.2
2757 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2759 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2760 of the problem. Specifically, suggested +O2 rather than +O1 for the
2763 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2765 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2766 record, the port was not correctly logged when the outgoing_port log
2767 selector was set (it logged the transort's default port).
2769 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2770 fallback_hosts, and "hosts" in the smtp transport.
2772 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2773 host errors such as "Connection refused".
2775 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2776 authentication with radiusclient 0.4.9:
2778 - Error returned from rc_read_config was caught wrongly
2779 - Username/password not passed on to radius server due to wrong length.
2781 The presumption is that some radiusclient API changes for 4.51/PH/17
2782 were not taken care of correctly. The code is still untested by me (my
2783 Linux distribution still has 0.3.2 of radiusclient), but it was
2784 contributed by a Radius user.
2786 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2787 expanding the "port" option of the smtp transport.
2789 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2790 while reading a MIME header. Thanks to Tom Hughes for a patch.
2792 PH/24 Include config.h inside local_scan.h so that configuration settings are
2795 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2796 that in an ACL for RCPT (for example), you can examine exactly what was
2799 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2800 commands, but it was not correctly comparing the address with the actual
2801 client host address. Thus, it would show the EHLO address in Received:
2802 header lines when this was not necessary.
2804 PH/27 Added the % operator to ${eval:}.
2806 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2807 it should be ignoring failures (because with -C, for example, it has lost
2808 privilege). It wasn't ignoring creation failures other than "already
2811 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2814 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2815 because an input line was too long, either on its own, or by virtue of
2816 too many continuations, the temporary file was not being removed, and the
2817 return code was incorrect.
2819 PH/31 Missing "BOOL" in function definition in filtertest.c.
2821 PH/32 Applied Sieve patches from the maintainer.
2823 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2825 PH/33 Added "verify = not_blind".
2827 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2828 Local/Makefile (with some defaults set). These are used in built scripts
2829 such as exicyclog, but they have never been used in the exim_install
2830 script (though there are many overriding facilities there). I have
2831 arranged that the exim_install script now takes note of these two
2834 PH/35 Installed configuration files for Dragonfly.
2836 PH/36 When a locally submitted message by a trusted user did not contain a
2837 From: header, and the sender address was obtained from -f or from an SMTP
2838 MAIL command, and the trusted user did not use -F to supply a sender
2839 name, $originator_name was incorrectly used when constructing a From:
2840 header. Furthermore, $originator_name was used for submission mode
2841 messages from external hosts without From: headers in a similar way,
2842 which is clearly wrong.
2844 PH/37 Added control=suppress_local_fixups.
2846 PH/38 When log_selector = +received_sender was set, and the addition of the
2847 sender made the log line's construction buffer exactly full, or one byte
2848 less than full, an overflow happened when the terminating "\n" was
2851 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2852 when the result of a list match is failure because a DNS lookup failed.
2854 PH/40 RM_COMMAND is now used in the building process.
2856 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2857 the "build-* directories that it finds.
2859 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2860 address in a domain literal was a prefix of an interface address.
2862 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2863 when verifying a sender address, unless rewrite_headers is false.
2865 PH/44 Wrote a long comment about why errors_to addresses are verified as
2866 recipients, not senders.
2868 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2869 the ratelimit ACL was added.
2871 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2873 PH/46 Added extra information about PostgreSQL errors to the error string.
2875 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2876 avoiding the need to install special zones in a real server. This is
2877 backwards compatible; if it can't find the fake resolver, it drops back.
2878 Thus, both old and new test suites can be run.
2880 TF/02 Added util/ratelimit.pl
2882 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2883 clock is set back in time.
2885 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2886 Candler <B.Candler@pobox.com>.
2888 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2889 OK for addresses that are the result of redirections.
2891 PH/48 A number of further additions for the benefit of the new test suite,
2892 including a fake gethostbyname() that interfaces to the fake DNS resolver
2895 TF/06 The fix for widen_domains has also been applied to qualify_single and
2896 search_parents which are the other dnslookup options that can cause
2899 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2902 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2904 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2905 service" if it encountered an MX record with an empty target hostname.
2906 The message is now "an MX or SRV record indicated no SMTP service".
2908 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2909 if the DNS of the sending site is misconfigured. This is quite a
2910 common situation. This change restores the behaviour of treating a
2911 helo verification defer as a failure.
2913 PH/51 If self=fail was set on a router, the bounce message did not include the
2914 actual error message.
2920 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2922 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2923 command itself ended in error, the underlying message about the transport
2924 filter timeout was being overwritten with the pipe command error. Now the
2925 underlying error message should be appended to the second error message.
2927 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2929 PH/02 Reduce the amount of output that "make" produces by default. Full output
2930 can still be requested.
2932 PH/03 The warning log line about a condition test deferring for a "warn" verb
2933 was being output only once per connection, rather than after each
2934 occurrence (because it was using the same function as for successful
2935 "warn" verbs). This seems wrong, so I have changed it.
2937 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2938 it should not have, which might have caused a crash in the right
2939 circumstances, but probably never did.
2941 PH/04 Installed a modified version of Tony Finch's patch to make submission
2942 mode fix the return path as well as the Sender: header line, and to
2943 add a /name= option so that you can make the user's friendly name appear
2946 TF/03 Added the control = fakedefer ACL modifier.
2948 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2949 Mark Lowes for thorough testing.
2951 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2953 TK/03 Merged latest SRS patch from Miles Wilton.
2955 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2956 with the definition in sysexits.h (which is #included earlier).
2957 Fortunately, Exim does not actually use EX_OK. The code used to try to
2958 preserve the sysexits.h value, by assumimg that macro definitions were
2959 scanned for macro replacements. I have been disabused of this notion,
2960 so now the code just undefines EX_OK before #including unistd.h.
2962 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2963 in the smtp transport. When a block could not be written in a single
2964 write() function, the timeout was being re-applied to each part-write.
2965 This seems wrong - if the receiver was accepting one byte at a time it
2966 would take for ever. The timeout is now adjusted when this happens. It
2967 doesn't have to be particularly precise.
2969 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2970 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2972 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2973 without a domain if the check to <postmaster@domain> fails.
2975 SC/01 Eximstats: added -xls and the ability to specify output files
2976 (patch written by Frank Heydlauf).
2978 SC/02 Eximstats: use FileHandles for outputing results.
2980 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2982 SC/04 Eximstats: fixed display of large numbers with -nvr option
2984 SC/05 Eximstats: fixed merging of reports with empty tables.
2986 SC/06 Eximstats: added the -include_original_destination flag
2988 SC/07 Eximstats: removed tabs and trailing whitespace.
2990 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2992 TK/06 MBOX spool code: Add real "From " MBOX separator line
2993 so the .eml file is really in mbox format (even though
2994 most programs do not really care). Patch from Alex Miller.
2996 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2997 The latter is generated from $received_to and is only set if the
2998 message has one envelope recipient. SA can use these headers,
2999 obviously out-of-the-box. Patch from Alex Miller.
3001 PH/08 The ${def test on a variable was returning false if the variable's
3002 value was "0", contrary to what the specification has always said!
3003 The result should be true unless the variable is empty.
3005 PH/09 The syntax error of a character other than { following "${if
3006 def:variable_name" (after optional whitespace) was not being diagnosed.
3007 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3008 accidental colon was present, for example, could give incorrect results.
3010 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3011 result is used (not including appendfile, where other changes are about
3014 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3015 This involved changing a lot of size variables from int to off_t. It
3016 should work with maildirs and everything.
3018 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3019 spamd dying while we are connected to it.
3021 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3024 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3025 function for BDB 4.3.
3027 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3029 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3030 the sources for PCRE have been re-arranged and more files are now
3033 PH/15 The code I had for printing potentially long long variables in PH/11
3034 above was not the best (it lost precision). The length of off_t variables
3035 is now inspected at build time, and an appropriate printing format (%ld
3036 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3037 to be "long long int" or "long int". This is needed for the internal
3038 formatting function string_vformat().
3040 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3041 the configuration file to be ":syslog", then the script "guesses" where
3042 the logs files are, rather than using the compiled in default. In our
3043 case the guess is not the same as the compiled default, so the script
3044 suddenly stopped working when I started to use syslog. The patch checks
3045 to see if log_file_path is "". If so, it attempts to read it from exim
3046 with no configuration file to get the compiled in version, before it
3047 falls back to the previous guessing code."
3049 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3050 implementing BATV in an Exim configuration. See NewStuff for the gory
3053 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3054 Makefile that are specific to HP-UX.
3056 PH/18 If the "use_postmaster" option was set for a recipient callout together
3057 with the "random" option, the postmaster address was used as the MAIL
3058 FROM address for the random test, but not for the subsequent recipient
3059 test. It is now used for both.
3061 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3062 patch removes a few documentation additions to RFC 3028, because the
3063 latest draft now contains them. It adds the new en;ascii-case comparator
3064 and a new error check for 8bit text in MIME parts. Comparator and
3065 require names are now matched exactly. I enabled the subaddress
3066 extension, but it is not well tested yet (read: it works for me)."
3068 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3069 rework some of the code of TK/09 above to avoid the hardwired use of
3070 "%lld" and "long long". Replaced the call to snprintf() with a call to
3073 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3074 records point to non-existent hosts", "retry timeout exceeded", and
3075 "retry time not reached for any host after a long failure period".
3077 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3078 experimental DomainKeys support:
3080 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3081 (2) On an error such as an illegally used "control", the wrong name for
3082 the control was given.
3084 These problems did NOT occur unless DomainKeys support was compiled.
3086 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3088 PH/24 Added ${if match_ip condition.
3090 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3091 databases so that it will be absolutely obvious if a crash occurs in the
3092 DB library. This is a regular occurrence (often caused by mis-matched
3095 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3096 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3097 picked up on a user's system that detects such things. There doesn't seem
3098 to be a gcc warning option for this - only an attribute that has to be
3099 put on the function's prototype. It seems that in Fedora Core 4 they have
3100 set this on a number of new functions. No doubt there will be more in due
3103 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3104 specify a transport. However, if an address that was verified by such a
3105 router was the subject of a callout, Exim crashed because it tried to
3106 read the rcpt_include_affixes from the non-existent transport. Now it
3107 just assumes that the setting of that option is false. This bug was
3108 introduced by 4.51/PH/31.
3110 PH/28 Changed -d+all to exclude +memory, because that information is very
3111 rarely of interest, but it makes the output a lot bigger. People tend to
3112 do -d+all out of habit.
3114 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3115 code in os-type was giving problems when libc.so lives in lib64, like on
3118 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3119 aren't the modern standard, and it seems that some systems' include files
3120 don't always have them. Exim was already checking for some of the newer
3121 ones like T_AAAA, and defining it itself. I've added checks for all the
3122 record types that Exim uses.
3124 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3125 not automatically generating a new one, as it is supposed to. This
3126 prevented TLS from working. If the file did exist, but contained invalid
3127 data, a new version was generated, as expected. It was only the case of a
3128 non-existent file that was broken.
3130 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3131 with a change in libdomainkeys > 0.64.
3133 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3134 from DNS. If the selector record carries the flag, it now has
3135 precedence over the domain-wide flag.
3137 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3139 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3140 the use of a port name that isn't defined in /etc/services) occurred, the
3141 message was deferred as in a normal delivery, and thus remained on the
3142 spool, instead of being failed because of the mua_wrapper setting. This
3143 is now fixed, and I tidied up some of the mua_wrapper messages at the
3146 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3147 the messages in a hash of arrays rather than using individual hashes.
3148 This is a bit cleaner and results in dramatic memory savings, albeit
3149 at a slight CPU cost.
3151 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3152 as requested by Marc Sherman.
3154 SC/10 Eximstats: added histograms for user specified patterns as requested
3157 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3159 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3160 fopen() in the content-scanning modules that did not already have it.
3166 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3167 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3169 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3171 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3173 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3174 to test Sieve filters that use "vacation".
3176 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3177 that changes the way the GnuTLS parameters are stored in the cache file.
3178 The new format can be generated externally. For backward compatibility,
3179 if the data in the cache doesn't make sense, Exim assumes it has read an
3180 old-format file, and it generates new data and writes a new file. This
3181 means that you can't go back to an older release without removing the
3184 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3185 work if there are any delivery delays because "one_time" forces the
3186 parent to be marked "delivered", so its unseen clone is never tried
3187 again. For this reason, Exim now forbids the simultaneous setting of
3190 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3191 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3192 entry if you want to know the details. The fix, however, neglected to
3193 consider the case when local delivery batching is involved. The test for
3194 "previously delivered" was not happening when checking to see if an
3195 address could be batched with a previous (undelivered) one; under
3196 certain circumstances this could lead to multiple deliveries to the same
3199 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3200 in its include files, and this causes problems building Exim.
3202 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3203 header_syntax) but Exim was just ignoring anything given after a slash.
3204 In particular, this caused confusion with an attempt to use "verify =
3205 reverse_host_lookup/defer_ok". An error is now given when options are
3206 supplied for verify items that do not have them. (Maybe reverse_host_
3207 lookup should have a defer_ok option, but that's a different point.)
3209 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3210 defined by RFC 821) to 2048, because there were problems with some AUTH
3211 commands, and RFC 1869 says the size should be increased for extended
3212 SMTP commands that take arguments.
3214 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3217 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3218 "unknown" error; now it says that the functionality isn't in the binary.
3220 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3221 an address' error message when a string expansion fails (syntax or
3222 whatever). Otherwise the password may appear in the log. Following change
3223 PH/42 below, there is no longer a chance of it appearing in a bounce
3226 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3228 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3229 of its addresses was ever tried. (Bugzilla bug #2.)
3231 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3232 the result incorrectly in the debug output. (It correctly added a newline
3233 to what was transported.)
3235 TF/01 Added $received_time.
3237 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3238 commented out examples of how to interface to a virus scanner and to
3239 SpamAssassin. Also added commented examples of av_scanner and
3240 spamd_address settings.
3242 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3243 and controls are allowed in which ACLs. There were a couple of minor
3244 errors. Some of the entries in the conditions table (which is a table of
3245 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3246 negation of where the condition IS allowed.
3248 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3250 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3251 header file does not have a version number, so I've had to invent a new
3252 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3253 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3254 radiusclient), but it was contributed by a Radius user.
3256 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3257 files or directories, for OpenSSL.
3259 PH/19 When an Exim process that is running as root has to create an Exim log
3260 file, it does so in a subprocess that runs as exim:exim so as to get the
3261 ownership right at creation (otherwise, other Exim processes might see
3262 the file with the wrong ownership). There was no test for failure of this
3263 fork() call, which would lead to the process getting stuck as it waited
3264 for a non-existent subprocess. Forks do occasionally fail when resources
3265 run out. I reviewed all the other calls to fork(); they all seem to check
3268 PH/20 When checking for unexpected SMTP input at connect time (before writing
3269 the banner), Exim was not dealing correctly with a non-positive return
3270 from the read() function. If the client had disconnected by this time,
3271 the result was a log entry for a synchronization error with an empty
3272 string after "input=" when read() returned zero. If read() returned -1
3273 (an event I could not check), uninitialized data bytes were printed.
3274 There were reports of junk text (parts of files, etc) appearing after
3277 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3279 PH/22 Added support for macro redefinition, and (re)definition in between
3280 driver and ACL definitions.
3282 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3283 forgetting to use the resulting value; it was using the unexpanded value.
3285 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3286 hadn't been configured. The fix is from Juergen Kreileder, who
3287 understands it better than I do:
3289 "Here's what I see happening with three configured cyrus_sasl
3290 authenticators configured (plain, login, cram-md5):
3292 On startup auth_cyrus_sasl_init() gets called for each of these.
3293 This means three calls to sasl_listmech() without a specified mech_list.
3294 => SASL tests which mechs of all available mechs actually work
3295 => three warnings about OTP not working
3296 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3298 With the patch, sasl_listmech() also gets called three times. But now
3299 SASL's mech_list option is set to the server_mech specified in the the
3300 authenticator. Or in other words, the answer from sasl_listmech()
3301 gets limited to just the mech you're testing for (which is different
3303 => the return list contains just 'plain' or 'login', 'cram-md5' or
3304 nothing depending on the value of ob->server_mech.
3306 I've just tested the patch: Authentication still works fine,
3307 unavailable mechs specified in the exim configuration are still
3308 caught, and the auth.log warnings about OTP are gone."
3310 PH/25 When debugging is enabled, the contents of the command line are added
3311 to the debugging output, even when log_selector=+arguments is not
3314 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3315 answer is "GNU", and only if the return is "GNU/something" is the answer
3318 PH/27 $acl_verify_message is now set immediately after the failure of a
3319 verification in an ACL, and so is available in subsequent modifiers. In
3320 particular, the message can be preserved by coding like this:
3322 warn !verify = sender
3323 set acl_m0 = $acl_verify_message
3325 Previously, $acl_verify_message was set only while expanding "message"
3326 and "log_message" when a very denied access.
3328 PH/28 Modified OS/os.c-Linux with
3330 -#ifndef OS_LOAD_AVERAGE
3331 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3333 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3334 nomenclature these days.)
3336 PH/29 Installed patch from the Sieve maintainer that adds the options
3337 sieve_useraddress and sieve_subaddress to the redirect router.
3339 PH/30 In these circumstances:
3340 . Two addresses routed to the same list of hosts;
3341 . First host does not offer TLS;
3342 . First host accepts first address;
3343 . First host gives temporary error to second address;
3344 . Second host offers TLS and a TLS session is established;
3345 . Second host accepts second address.
3346 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3347 and peerdn, if requested) that were in fact used only for the second
3350 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3351 attention to any local part prefix or suffix that was matched by the
3352 router that accepted the address. It now behaves in the same way as it
3353 does for delivery: the affixes are removed from the local part unless
3354 rcpt_include_affixes is set on the transport.
3356 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3357 timeout during the DATA phase of an incoming message.
3359 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3360 applied a patch sanctioned by the Sieve maintainer.
3362 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3363 the uid or gid is negative. A case of a negative gid caused this to be
3364 noticed. The fix allows for either to be negative.
3366 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3367 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3370 PH/36 The change PH/12 above was broken. Fixed it.
3372 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3373 the grounds that routing the same address twice would always produce the
3374 same answer. This might have been true once, but it is certainly no
3375 longer true now. Routing a child address may depend on the previous
3376 routing that produced that child. Some complicated redirection strategies
3377 went wrong when messages had multiple recipients, and made Exim's
3378 behaviour dependent on the order in which the addresses were given.
3380 I have moved the duplicate checking until after the routing is complete.
3381 Exim scans the addresses that are assigned to local and remote
3382 transports, and removes any duplicates. This means that more work will be
3383 done, as duplicates will always all be routed, but duplicates are
3384 presumably rare, so I don't expect this is of any significance.
3386 For deliveries to pipes, files, and autoreplies, the duplicate checking
3387 still happens during the routing process, since they are not going to be
3390 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3391 It corrects a timeout issue with spamd. This is Ian's comment: "The
3392 background is that sometimes spamd either never reads data from a
3393 connection it has accepted, or it never writes response data. The exiscan
3394 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3395 blindly assumes that writes won't block so it may never time out."
3397 PH/39 Allow G after quota size as well as K and M.
3399 PH/40 The value set for $authenticated_id in an authenticator may not contain
3400 binary zeroes or newlines because the value is written to log lines and
3401 to spool files. There was no check on this. Now the value is run through
3402 the string_printing() function so that such characters are converted to
3403 printable escape sequences.
3405 PH/41 $message_linecount is a new variable that contains the total number of
3406 lines in the message. Compare $body_linecount, which is the count for the
3409 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3410 bounce and delay warning messages, except in certain special cases, which
3413 (a) An SMTP error message from a remote host;
3414 (b) A message specified in a :fail: redirection;
3415 (c) A message specified in a "fail" command in a system filter;
3416 (d) A message specified in a FAIL return from the queryprogram router;
3417 (e) A message specified by the cannot_route_message router option.
3419 In these cases only, Exim does include the error details in bounce and
3420 warning messages. There are also a few cases where bland messages such
3421 as "unrouteable address" or "local delivery error" are given.
3423 PH/43 $value is now also set for the "else" part of a ${run expansion.
3425 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3426 being worked on, but at least Exim now implements the latest version to
3429 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3430 process to complete was treated as a delivery failure, a timeout while
3431 writing the message to the pipe was logged, but erroneously treated as a
3432 successful delivery. Such timeouts include transport filter timeouts. For
3433 consistency with the overall process timeout, these timeouts are now
3434 treated as errors, giving rise to delivery failures by default. However,
3435 there is now a new Boolean option for the pipe transport called
3436 timeout_defer, which, if set TRUE, converts the failures into defers for
3437 both kinds of timeout. A transport filter timeout is now identified in
3440 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3441 systems where "make" and "gmake" are different, calling "gmake" at top
3442 level broke things. I've arranged for the value of $(MAKE) to be passed
3443 from the Makefile to this script so that it can call the same version of
3447 A note about Exim versions 4.44 and 4.50
3448 ----------------------------------------
3450 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3451 changes of various kinds. As a consequence, a big documentation update was
3452 needed. This delayed the release for rather longer than seemed good, especially
3453 in the light of a couple of (minor) security issues. Therefore, the changes
3454 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3455 release. So 4.44 and 4.50 are in effect two different branches that both start
3458 I have left the 4.50 change log unchanged; it contains all the changes since
3459 4.43. The change log for 4.44 is below; many of its items are identical to
3460 those for 4.50. This seems to be the most sensible way to preserve the
3461 historical information.
3467 1. Minor wording change to the doc/README.SIEVE file.
3469 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3470 computation of the current number of files was incorrect.
3472 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3473 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3476 4. Give more explanation in the error message when the command for a transport
3477 filter fails to execute.
3479 5. There are several places where Exim runs a non-Exim command in a
3480 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3481 was being done only for the command run by the queryprogram router. It is
3482 now done for all such subprocesses. The other cases are: ${run, transport
3483 filters, and the commands run by the lmtp and pipe transports.
3485 6. Added CONFIGURE_GROUP build-time option.
3487 7. Some older OS have a limit of 256 on the maximum number of file
3488 descriptors. Exim was using setrlimit() to set 1000 as a large value
3489 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3490 systems. I've change it so that if it can't get 1000, it tries for 256.
3492 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3493 was an oversight, and furthermore, ever since the addition of extra
3494 controls (e.g. 4.43/32), the checks on when to allow different forms of
3495 "control" were broken. There should now be diagnostics for all cases when a
3496 control that does not make sense is encountered.
3498 9. Added the /retain_sender option to "control=submission".
3500 10. $recipients is now available in the predata ACL (oversight).
3502 11. Tidy the search cache before the fork to do a delivery from a message
3503 received from the command line. Otherwise the child will trigger a lookup
3504 failure and thereby defer the delivery if it tries to use (for example) a
3505 cached ldap connection that the parent has called unbind on.
3507 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3508 of $address_data from the recipient verification was clobbered by the
3509 sender verification.
3511 13. The value of address_data from a sender verification is now available in
3512 $sender_address_data in subsequent conditions in the ACL statement.
3514 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3516 15. Added a new option "connect=<time>" to callout options, to set a different
3519 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3520 was its contents. (It was OK if the option was not defined at all.)
3522 17. A "Completed" log line is now written for messages that are removed from
3523 the spool by the -Mrm option.
3525 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3526 information about exactly what failed.
3528 19. Added -dd to debug only the daemon process.
3530 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3531 handles timeouts, both on the server side and network timeouts. Renamed the
3532 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3534 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3535 It is now set to "smtps".
3537 22. $host_address is now set to the target address during the checking of
3538 ignore_target_hosts.
3540 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3541 being passed; this would have caused $sender_host_name to have been used if
3542 matching the list had actually called for a host name (not very likely,
3543 since this list is usually IP addresses). A host name is now passed as
3546 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3547 code that specifies a non-restarting handler (typically sigaction() in
3548 modern systems) in an attempt to fix a rare and obscure crash bug.
3550 25. Narrowed the window for a race in the daemon that could cause it to ignore
3551 SIGCHLD signals. This is not a major problem, because they are used only to
3552 wake it up if nothing else does.
3554 26. A malformed maildirsize file could cause Exim to calculate negative values
3555 for the mailbox size or file count. Odd effects could occur as a result.
3556 The maildirsize information is now recalculated if the size or filecount
3559 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3560 support for a long time. Removed HAVE_SYS_VFS_H.
3562 28. Installed the latest version of exipick from John Jetmore.
3564 29. In an address list, if the pattern was not a regular expression, an empty
3565 subject address (from a bounce message) matched only if the pattern was an
3566 empty string. Non-empty patterns were not even tested. This was the wrong
3567 because it is perfectly reasonable to use an empty address as part of a
3568 database query. An empty address is now tested by patterns that are
3569 lookups. However, all the other forms of pattern expect the subject to
3570 contain a local part and a domain, and therefore, for them, an empty
3571 address still always fails if the pattern is not itself empty.
3573 30. Exim went into a mad DNS loop when attempting to do a callout where the
3574 host was specified on an smtp transport, and looking it up yielded more
3575 than one IP address.
3577 31. Re-factored the code for checking spool and log partition space into a
3578 function that finds that data and another that does the check. The former
3579 is then used to implement four new variables: $spool_space, $log_space,
3580 $spool_inodes, and $log_inodes.
3582 32. The RFC2047 encoding function was originally intended for short strings
3583 such as real names; it was not keeping to the 75-character limit for
3584 encoded words that the RFC imposes. It now respects the limit, and
3585 generates multiple encoded words if necessary. To be on the safe side, I
3586 have increased the buffer size for the ${rfc2047: expansion operator from
3589 33. It is now permitted to omit both strings after an "if" condition; if the
3590 condition is true, the result is "true". As before, when the second string
3591 is omitted, a false condition yields an empty string. This makes it less
3592 cumbersome to write custom ACL and router conditions.
3594 34. Failure to deliver a bounce message always caused it to be frozen, even if
3595 there was an errors_to setting on the router. The errors_to setting is now
3598 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3599 canonical form (fully expanded) before being placed in
3600 $sender_host_address.
3602 36. The table in the code that translates DNS record types into text (T_A to
3603 "A" for instance) was missing entries for NS and CNAME. It is just possible
3604 that this could have caused confusion if both these types were looked up
3605 for the same domain, because the text type is used as part of Exim's
3606 per-process caching. But the chance of anyone hitting this buglet seems
3609 37. The dnsdb lookup has been extended in a number of ways.
3611 (1) There is a new type, "zns", which walks up the domain tree until it
3612 finds some nameserver records. It should be used with care.
3614 (2) There is a new type, "mxh", which is like "mx" except that it returns
3615 just the host names, not the priorities.
3617 (3) It is now possible to give a list of domains (or IP addresses) to be
3618 looked up. The behaviour when one of the lookups defers can be
3619 controlled by a keyword.
3621 (4) It is now possible to specify the separator character for use when
3622 multiple records are returned.
3624 38. The dnslists ACL condition has been extended: it is now possible to supply
3625 a list of IP addresses and/or domains to be looked up in a particular DNS
3628 39. Added log_selector=+queue_time_overall.
3630 40. When running the queue in the test harness, wait just a tad after forking a
3631 delivery process, to get repeatability of debugging output.
3633 41. Include certificate and key file names in error message when GnuTLS fails
3634 to set them up, because the GnuTLS error message doesn't include the name
3635 of the failing file when there is a problem reading it.
3637 42. Allow both -bf and -bF in the same test run.
3639 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3641 44. The "Exiscan patch" is now merged into the mainline Exim source.
3643 45. Sometimes the final signoff response after QUIT could fail to get
3644 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3645 < 0 before doing a fflush(). This bug looks as though it goes back to the
3646 introduction of TLS in release 3.20, but "sometimes" must have been rare
3647 because the tests only now provoked it.
3649 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3650 (this can affect the format of dates).
3652 47. exim_tidydb, when checking for the continued existence of a message for
3653 which it has found a message-specific retry record, was not finding
3654 messages that were in split spool directories. Consequently, it was
3655 deleting retry records that should have stayed in existence.
3657 48. Steve fixed some bugs in eximstats.
3659 49. The SPA authentication driver was not abandoning authentication and moving
3660 on to the next authenticator when an expansion was forced to fail,
3661 contradicting the general specification for all authenticators. Instead it
3662 was generating a temporary error. It now behaves as specified.
3664 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3665 (the order specifies the preference for clients). The order is now AES256,
3666 AES128, 3DES, ARCFOUR128.
3668 51. Small patch to Sieve code - explicitly set From: when generating an
3671 52. Exim crashed if a remote delivery caused a very long error message to be
3672 recorded - for instance if somebody sent an entire SpamAssassin report back
3673 as a large number of 550 error lines. This bug was coincidentally fixed by
3674 increasing the size of one of Exim's internal buffers (big_buffer) that
3675 happened as part of the Exiscan merge. However, to be on the safe side, I
3676 have made the code more robust (and fixed the comments that describe what
3679 53. Now that there can be additional text after "Completed" in log lines (if
3680 the queue_time_overall log selector is set), a one-byte patch to exigrep
3681 was needed to allow it to recognize "Completed" as not the last thing in
3684 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3685 patch that reportedly fixes this has been added. I am not expert enough to
3686 create a test for it. This is what the patch creator wrote:
3688 "I found a little strange behaviour of ldap code when working with
3689 Windows 2003 AD Domain, where users was placed in more than one
3690 Organization Units. When I tried to give exim partial DN, the exit code
3691 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3692 But simultaneously result of request was absolutely normal ldap result,
3693 so I produce this patch..."
3695 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3696 REFERENCE, so I have modified the code to exclude the patch when that macro
3699 55. Some experimental protocols are using DNS PTR records for new purposes. The
3700 keys for these records are domain names, not reversed IP addresses. The
3701 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3702 leaves it alone. Component reversal etc. now happens only for IP addresses.
3705 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3707 57. Double the size of the debug message buffer (to 2048) so that more of very
3708 long debug lines gets shown.
3710 58. The exicyclog utility now does better if the number of log files to keep
3711 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3713 59. Two changes related to the smtp_active_hostname option:
3715 (1) $smtp_active_hostname is now available as a variable.
3716 (2) The default for smtp_banner uses $smtp_active_hostname instead
3717 of $primary_hostname.
3719 60. The host_aton() function is supposed to be passed a string that is known
3720 to be a valid IP address. However, in the case of IPv6 addresses, it was
3721 not checking this. This is a hostage to fortune. Exim now panics and dies
3722 if the condition is not met. A case was found where this could be provoked
3723 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3724 components; fortuitously, this particular loophole had already been fixed
3725 by change 4.50/55 above.
3727 If there are any other similar loopholes, the new check in host_aton()
3728 itself should stop them being exploited. The report I received stated that
3729 data on the command line could provoke the exploit when Exim was running as
3730 exim, but did not say which command line option was involved. All I could
3731 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3732 running as the user.
3735 61. There was a buffer overflow vulnerability in the SPA authentication code
3736 (which came originally from the Samba project). I have added a test to the
3737 spa_base64_to_bits() function which I hope fixes it.
3740 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3741 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3743 63. The daemon start-up calls getloadavg() while still root for those OS that
3744 need the first call to be done as root, but it missed one case: when
3745 deliver_queue_load_max is set with deliver_drop_privilege. This is
3746 necessary for the benefit of the queue runner, because there is no re-exec
3747 when deliver_drop_privilege is set.
3749 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3750 This has been fixed.
3752 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3753 was in use, was not putting the data itself into the right store pool;
3754 consequently, it could be overwritten for a subsequent message in the same
3755 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3758 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3760 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3761 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3762 the function were treating the return as a boolean value, which happened to
3763 work because 0=false and not-0=true, but is not correct code.
3765 68. The host_aton() function was not handling scoped IPv6 addresses (those
3766 with, for example, "%eth0" on the end) correctly.
3768 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3769 negated items (that is, ~something) in unsigned ints. Some compilers
3770 apparently mutter when there is no cast.
3772 70. If an address verification called from an ACL failed, and did not produce a
3773 user-specific message (i.e. there was only a "system" message), nothing was
3774 put in $acl_verify_message. In this situation, it now puts the system
3777 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3778 session; change 4.31/43 added the unwanted input to the log line - except
3779 that it did not do this in the start of session case. It now does.
3781 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3782 This is wrong and can cause the other end to generate a synchronization
3783 error if it is another Exim or anything else that does the synchronization
3784 check. A QUIT command is no longer sent after a timeout.
3786 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3787 during host lookups.
3789 74. The defer_ok option of callout verification was not working if it was used
3790 when verifying addresses in header lines, that is, for this case:
3792 verify = header_sender/callout=defer_ok
3794 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3795 those file descriptors could be used for SMTP connections. If anything
3796 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3797 could be sent to the SMTP client, causing chaos. The daemon now opens
3798 stdin, stdout, and stderr to /dev/null when it puts itself into the
3801 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3802 log by default. The user can override this with suitable Perl magic.
3804 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3805 the log message when discard triggers, was not working for the DATA ACL or
3806 for the non-SMTP ACL.
3808 78. Error message wording change in sieve.c.
3810 79. If smtp_accept_max_per_host was set, the number of connections could be
3811 restricted to fewer than expected, because the daemon was trying to set up
3812 a new connection before checking whether the processes handling previous
3813 connections had finished. The check for completed processes is now done
3814 earlier. On busy systems, this bug wouldn't be noticed because something
3815 else would have woken the daemon, and it would have reaped the completed
3818 80. If a message was submitted locally by a user whose login name contained one
3819 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3820 It caused a spool format error. I have fixed the spool reading code. A
3821 related problem was that the "from" clause in the Received: line became
3822 illegal because of the space(s). It is now covered by ${quote_local_part.
3824 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3827 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3829 83. Patch from Sieve maintainer to fix unterminated string problem in
3830 "vacation" handling.
3832 84. Some minor changes to the Linux configuration files to help with other
3833 OS variants using glibc.
3835 85. One more patch for Sieve to update vacation handling to latest spec.
3838 ----------------------------------------------------
3839 See the note above about the 4.44 and 4.50 releases.
3840 ----------------------------------------------------
3846 1. Change 4.43/35 introduced a bug that caused file counts to be
3847 incorrectly computed when quota_filecount was set in an appendfile
3850 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3851 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3854 3. Give more explanation in the error message when the command for a transport
3855 filter fails to execute.
3857 4. There are several places where Exim runs a non-Exim command in a
3858 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3859 was being done only for the command run by the queryprogram router. It is
3860 now done for all such subprocesses. The other cases are: ${run, transport
3861 filters, and the commands run by the lmtp and pipe transports.
3863 5. Some older OS have a limit of 256 on the maximum number of file
3864 descriptors. Exim was using setrlimit() to set 1000 as a large value
3865 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3866 systems. I've change it so that if it can't get 1000, it tries for 256.
3868 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3869 was an oversight, and furthermore, ever since the addition of extra
3870 controls (e.g. 4.43/32), the checks on when to allow different forms of
3871 "control" were broken. There should now be diagnostics for all cases when a
3872 control that does not make sense is encountered.
3874 7. $recipients is now available in the predata ACL (oversight).
3876 8. Tidy the search cache before the fork to do a delivery from a message
3877 received from the command line. Otherwise the child will trigger a lookup
3878 failure and thereby defer the delivery if it tries to use (for example) a
3879 cached ldap connection that the parent has called unbind on.
3881 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3882 of $address_data from the recipient verification was clobbered by the
3883 sender verification.
3885 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3886 was its contents. (It was OK if the option was not defined at all.)
3888 11. A "Completed" log line is now written for messages that are removed from
3889 the spool by the -Mrm option.
3891 12. $host_address is now set to the target address during the checking of
3892 ignore_target_hosts.
3894 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3895 being passed; this would have caused $sender_host_name to have been used if
3896 matching the list had actually called for a host name (not very likely,
3897 since this list is usually IP addresses). A host name is now passed as
3900 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3901 code that specifies a non-restarting handler (typically sigaction() in
3902 modern systems) in an attempt to fix a rare and obscure crash bug.
3904 15. Narrowed the window for a race in the daemon that could cause it to ignore
3905 SIGCHLD signals. This is not a major problem, because they are used only to
3906 wake it up if nothing else does.
3908 16. A malformed maildirsize file could cause Exim to calculate negative values
3909 for the mailbox size or file count. Odd effects could occur as a result.
3910 The maildirsize information is now recalculated if the size or filecount
3913 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3914 support for a long time. Removed HAVE_SYS_VFS_H.
3916 18. Updated exipick to current release from John Jetmore.
3918 19. Allow an empty sender to be matched against a lookup in an address list.
3919 Previously the only cases considered were a regular expression, or an
3922 20. Exim went into a mad DNS lookup loop when doing a callout where the
3923 host was specified on the transport, if the DNS lookup yielded more than
3926 21. The RFC2047 encoding function was originally intended for short strings
3927 such as real names; it was not keeping to the 75-character limit for
3928 encoded words that the RFC imposes. It now respects the limit, and
3929 generates multiple encoded words if necessary. To be on the safe side, I
3930 have increased the buffer size for the ${rfc2047: expansion operator from
3933 22. Failure to deliver a bounce message always caused it to be frozen, even if
3934 there was an errors_to setting on the router. The errors_to setting is now
3937 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3938 canonical form (fully expanded) before being placed in
3939 $sender_host_address.
3941 24. Updated eximstats to version 1.33
3943 25. Include certificate and key file names in error message when GnuTLS fails
3944 to set them up, because the GnuTLS error message doesn't include the name
3945 of the failing file when there is a problem reading it.
3947 26. Expand error message when OpenSSL has problems setting up cert/key files.
3950 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3951 (this can affect the format of dates).
3953 28. exim_tidydb, when checking for the continued existence of a message for
3954 which it has found a message-specific retry record, was not finding
3955 messages that were in split spool directories. Consequently, it was
3956 deleting retry records that should have stayed in existence.
3958 29. eximstats updated to version 1.35
3959 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3960 1.35 - bugfix such that pie charts by volume are generated correctly
3962 30. The SPA authentication driver was not abandoning authentication and moving
3963 on to the next authenticator when an expansion was forced to fail,
3964 contradicting the general specification for all authenticators. Instead it
3965 was generating a temporary error. It now behaves as specified.
3967 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3968 (the order specifies the preference for clients). The order is now AES256,
3969 AES128, 3DES, ARCFOUR128.
3971 31. Small patch to Sieve code - explicitly set From: when generating an
3974 32. Exim crashed if a remote delivery caused a very long error message to be
3975 recorded - for instance if somebody sent an entire SpamAssassin report back
3976 as a large number of 550 error lines. This bug was coincidentally fixed by
3977 increasing the size of one of Exim's internal buffers (big_buffer) that
3978 happened as part of the Exiscan merge. However, to be on the safe side, I
3979 have made the code more robust (and fixed the comments that describe what
3982 33. Some experimental protocols are using DNS PTR records for new purposes. The
3983 keys for these records are domain names, not reversed IP addresses. The
3984 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3985 leaves it alone. Component reversal etc. now happens only for IP addresses.
3988 34. The host_aton() function is supposed to be passed a string that is known
3989 to be a valid IP address. However, in the case of IPv6 addresses, it was
3990 not checking this. This is a hostage to fortune. Exim now panics and dies
3991 if the condition is not met. A case was found where this could be provoked
3992 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3993 components; fortuitously, this particular loophole had already been fixed
3994 by change 4.50/55 or 4.44/33 above.
3996 If there are any other similar loopholes, the new check in host_aton()
3997 itself should stop them being exploited. The report I received stated that
3998 data on the command line could provoke the exploit when Exim was running as
3999 exim, but did not say which command line option was involved. All I could
4000 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4001 running as the user.
4004 35. There was a buffer overflow vulnerability in the SPA authentication code
4005 (which came originally from the Samba project). I have added a test to the
4006 spa_base64_to_bits() function which I hope fixes it.
4009 36. The daemon start-up calls getloadavg() while still root for those OS that
4010 need the first call to be done as root, but it missed one case: when
4011 deliver_queue_load_max is set with deliver_drop_privilege. This is
4012 necessary for the benefit of the queue runner, because there is no re-exec
4013 when deliver_drop_privilege is set.
4015 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4016 was in use, was not putting the data itself into the right store pool;
4017 consequently, it could be overwritten for a subsequent message in the same
4018 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4021 38. Sometimes the final signoff response after QUIT could fail to get
4022 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4023 < 0 before doing a fflush(). This bug looks as though it goes back to the
4024 introduction of TLS in release 3.20, but "sometimes" must have been rare
4025 because the tests only now provoked it.
4031 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4032 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4033 Now it returns an integer. A number of calls were still expecting a T/F
4034 return. Fortuitously, in all cases, the tests worked in OK situations,
4035 which is the norm. However, things would have gone wrong on any write
4036 failures on the smtp file descriptor. This function is used when sending
4037 messages over SMTP and also when doing verify callouts.
4039 2. When Exim is called to do synchronous delivery of a locally submitted
4040 message (the -odf or -odi options), it no longer closes stderr before doing
4043 3. Implemented the mua_wrapper option.
4045 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4047 5. Implemented the functions header_remove(), header_testname(),
4048 header_add_at_position(), and receive_remove_recipient(), and exported them
4051 6. If an ACL "warn" statement specified the addition of headers, Exim already
4052 inserted X-ACL-Warn: at the start if there was no header name. However, it
4053 was not making this test for the second and subsequent header lines if
4054 there were newlines in the string. This meant that an invalid header could
4055 be inserted if Exim was badly configured.
4057 7. Allow an ACL "warn" statement to add header lines at the start or after all
4058 the Received: headers, as well as at the end.
4060 8. Added the rcpt_4xx retry error code.
4062 9. Added postmaster_mailfrom=xxx to callout verification option.
4064 10. Added mailfrom=xxxx to the callout verification option, for verify=
4067 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4068 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4070 12. Inserted some casts to stop certain compilers warning when using pointer
4071 differences as field lengths or precisions in printf-type calls (mostly
4072 affecting debugging statements).
4074 13. Added optional readline() support for -be (dynamically loaded).
4076 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4077 same clock tick as a message's arrival, so that its received time was the
4078 same as the "first fail" time on the retry record, and that message
4079 remained on the queue past the ultimate address timeout, every queue runner
4080 would try a delivery (because it was past the ultimate address timeout) but
4081 after another failure, the ultimate address timeout, which should have then
4082 bounced the address, did not kick in. This was a "< instead of <=" error;
4083 in most cases the first failure would have been in the next clock tick
4084 after the received time, and all would be well.
4086 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4087 being recognized when the domain list was tested by the match_domain
4088 condition in an expansion string.
4090 16. Added the ${str2b64: operator.
4092 17. Exim was always calling setrlimit() to set a large limit for the number of
4093 processes, without checking whether the existing limit was already
4094 adequate. (It did check for the limit on file descriptors.) Furthermore,
4095 errors from getrlimit() and setrlimit() were being ignored. Now they are
4096 logged to the main and panic logs, but Exim does carry on, to try to do its
4097 job under whatever limits there are.
4099 18. Imported PCRE 5.0.
4101 19. Trivial typo in log message " temporarily refused connection" (the leading
4104 20. If the log selector return_path_on_delivery was set and an address was
4105 redirected to /dev/null, the delivery process crashed because it assumed
4106 that a return path would always be set for a "successful" delivery. In this
4107 case, the whole delivery is bypassed as an optimization, and therefore no
4110 21. Internal re-arrangement: the function for sending a challenge and reading
4111 a response while authentication was assuming a zero-terminated challenge
4112 string. It's now changed to take a pointer and a length, to allow for
4113 binary data in such strings.
4115 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4117 23. Exim was not respecting finduser_retries when seeking the login of the
4118 uid under which it was called; it was always trying 10 times. (The default
4119 setting of finduser_retries is zero.) Also, it was sleeping after the final
4120 failure, which is pointless.
4122 24. Implemented tls_on_connect_ports.
4124 25. Implemented acl_smtp_predata.
4126 26. If the domain in control=submission is set empty, Exim assumes that the
4127 authenticated id is a complete email address when it generates From: or
4128 Sender: header lines.
4130 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4131 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4132 chown and chgrp in /bin and hostname in /usr/bin.
4134 28. Exim was keeping the "process log" file open after each use, just as it
4135 does for the main log. This opens the possibility of it remaining open for
4136 long periods when the USR1 signal hits a daemon. Occasional processlog
4137 errors were reported, that could have been caused by this. Anyway, it seems
4138 much more sensible not to leave this file open at all, so that is what now
4141 29. The long-running daemon process does not normally write to the log once it
4142 has entered its main loop, and it closes the log before doing so. This is
4143 so that log files can straightforwardly be renamed and moved. However,
4144 there are a couple of unusual error situations where the daemon does write
4145 log entries, and I had neglected to close the log afterwards.
4147 30. The text of an SMTP error response that was received during a remote
4148 delivery was being truncated at 512 bytes. This is too short for some of
4149 the long messages that one sometimes sees. I've increased the limit to
4152 31. It is now possible to make retry rules that apply only when a message has a
4153 specific sender, in particular, an empty sender.
4155 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4156 it possible to be selective about when SMTP synchronization is enforced.
4158 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4160 32. Implemented hosts_connection_nolog.
4162 33. Added an ACL for QUIT.
4164 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4167 35. Added mailbox_size and mailbox_filecount to appendfile.
4169 36. Added control = no_multiline_responses to ACLs.
4171 37. There was a bug in the logic of the code that waits for the clock to tick
4172 in the case where the clock went backwards by a substantial amount such
4173 that the microsecond fraction of "now" was more than the microsecond
4174 fraction of "then" (but the whole seconds number was less).
4176 38. Added support for the libradius Radius client library this is found on
4177 FreeBSD (previously only the radiusclient library was supported).
4183 1. When certain lookups returned multiple values in the form name=value, the
4184 quoting of the values was not always being done properly. Specifically:
4185 (a) If the value started with a double quote, but contained no whitespace,
4187 (b) If the value contained whitespace other than a space character (i.e.
4188 tabs or newlines or carriage returns) it was not quoted.
4189 This fix has been applied to the mysql and pgsql lookups by writing a
4190 separate quoting function and calling it from the lookup code. The fix
4191 should probably also be applied to nisplus, ibase and oracle lookups, but
4192 since I cannot test any of those, I have not disturbed their existing code.
4194 2. A hit in the callout cache for a specific address caused a log line with no
4195 reason for rejecting RCPT. Now it says "Previous (cached) callout
4196 verification failure".
4198 3. There was an off-by-one bug in the queryprogram router. An over-long
4199 return line was truncated at 256 instead of 255 characters, thereby
4200 overflowing its buffer with the terminating zero. As well as fixing this, I
4201 have increased the buffer size to 1024 (and made a note to document this).
4203 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4204 when Exim is waiting for an SMTP response from a remote server, Exim
4205 restarts its select() call on the socket, thereby resetting its timeout.
4206 This is not a problem when such interrupts are rare. Somebody set up a cron
4207 job to run exiwhat every 2 minutes, which is less than the normal select()
4208 timeout (5 or 10 minutes). This meant that the select() timeout never
4209 kicked in because it was always reset. I have fixed this by comparing the
4210 time when an interrupt arrives with the time at the start of the first call
4211 to select(). If more time than the timeout has elapsed, the interrupt is
4212 treated as a timeout.
4214 5. Some internal re-factoring in preparation for the addition of Sieve
4215 extensions (by MH). In particular, the "personal" test is moved to a
4216 separate function, and given an option for scanning Cc: and Bcc: (which is
4217 not set for Exim filters).
4219 6. When Exim created an email address using the login of the caller as the
4220 local part (e.g. when creating a From: or Sender: header line), it was not
4221 quoting the local part when it contained special characters such as @.
4223 7. Installed new OpenBSD configuration files.
4225 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4226 try to make them clearer.
4228 9. Callout options, other than the timeout value, were being ignored when
4229 verifying sender addresses in header lines. For example, when using
4231 verify = header_sender/callout=no_cache
4233 the cache was (incorrectly) being used.
4235 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4236 only the Cygwin environment.
4238 11. When return_path_on_delivery was set as a log selector, if different remote
4239 addresses in the same message used different return paths and parallel
4240 remote delivery occurred, the wrong values would sometimes be logged.
4241 (Whenever a remote delivery process finished, the return path value from
4242 the most recently started remote delivery process was logged.)
4244 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4245 lines when AUTH and/or TLS are in use. This is the "received protocol"
4246 field. Exim used to use "asmtp" for authenticated SMTP, without any
4247 indication (in the protocol name) for TLS use. Now it follows the RFC and
4248 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4249 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4250 names appear in log lines as well as in Received: header lines.
4252 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4253 extensions, and comparison tests, and to fix some bugs.
4255 14. Changes to the "personal" filter test:
4257 (1) The test was buggy in that it was just doing the equivalent of
4258 "contains" tests on header lines. For example, if a user's address was
4259 anne@some.where, the "personal" test would incorrectly be true for
4261 To: susanne@some.where
4263 This test is now done by extracting each address from the header in turn,
4264 and checking the entire address. Other tests that are part of "personal"
4265 are now done using regular expressions (for example, to check local parts
4266 of addresses in From: header lines).
4268 (2) The list of non-personal local parts in From: addresses has been
4269 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4270 taken from the Sieve specification recommendations.
4272 (3) If the message contains any header line starting with "List-" it is
4273 treated as non-personal.
4275 (4) The test for "circular" in the Subject: header line has been removed
4276 because it now seems ill-conceived.
4278 15. Minor typos in src/EDITME comments corrected.
4280 16. Installed latest exipick from John Jetmore.
4282 17. If headers_add on a router specified a text string that was too long for
4283 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4284 of string_sprintf() is now avoided.
4286 18. $message_body_size was not set (it was always zero) when running the DATA
4287 ACL and the local_scan() function.
4289 19. For the "mail" command in an Exim filter, no default was being set for
4290 the once_repeat time, causing a random time value to be used if "once" was
4291 specified. (If the value happened to be <= 0, no repeat happened.) The
4292 default is now 0s, meaning "never repeat". The "vacation" command was OK
4293 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4294 (I found it when inspecting the code).
4296 20. There is now an overall timeout for performing a callout verification. It
4297 defaults to 4 times the callout timeout, which applies to individual SMTP
4298 commands during the callout. The overall timeout applies when there is more
4299 than one host that can be tried. The timeout is checked before trying the
4300 next host. This prevents very long delays if there are a large number of
4301 hosts and all are timing out (e.g. when the network connections are timing
4302 out). The value of the overall timeout can be changed by specifying an
4303 additional sub-option for "callout", called "maxwait". For example:
4305 verify = sender/callout=5s,maxwait=20s
4307 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4308 to the end before writing, but this should make it even safer).
4310 22. Exim was forgetting that it had advertised PIPELINING for the second and
4311 subsequent messages on an SMTP connection. It was also not resetting its
4312 memory on STARTTLS and an internal HELO.
4314 23. When Exim logs an SMTP synchronization error within a session, it now
4315 records whether PIPELINING has been advertised or not.
4317 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4318 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4319 rather than long int.
4321 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4323 26. Added the never_mail option to autoreply.
4329 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4330 crash if the getsockname() call failed; this can happen if a connection is
4331 closed very soon after it is established. The problem was simply in the
4332 order in which certain operations were done, causing Exim to try to write
4333 to the SMTP stream before it had set up the file descriptor. The bug has
4334 been fixed by making things happen in the correct order.
4340 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4341 before the connection was closed, thus losing the rejection response.
4343 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4344 some early Solaris releases, but causes trouble in current releases where
4345 socklen_t is defined.
4347 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4350 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4353 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4354 that releasing the top of it at the end releases what was used for sub-
4355 expansions (unless the block got too big). However, discard this block if
4356 the first thing is a variable or header, so that we can use its block when
4357 it is dynamic (useful for very large $message_headers, for example).
4359 6. Lookups now cache *every* query, not just the most recent. A new, separate
4360 store pool is used for this. It can be recovered when all lookup caches are
4361 flushed. Lookups now release memory at the end of their result strings.
4362 This has involved some general refactoring of the lookup sources.
4364 7. Some code has been added to the store_xxx() functions to reduce the amount
4365 of flapping under certain conditions.
4367 8. log_incoming_interface used to affect only the <= reception log lines. Now
4368 it causes the local interface and port to be added to several more SMTP log
4369 lines, for example "SMTP connection from", and rejection lines.
4371 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4373 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4375 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4376 could be overwritten at the end of the current message (or the start of a
4377 new message if it was set in a HELO ACL). The value is now preserved for
4378 the duration of the SMTP connection.
4380 12. If a transport had a headers_rewrite setting, and a matching header line
4381 contained an unqualified address, that address was qualified, even if it
4382 did not match any rewriting rules. The underlying bug was that the values
4383 of the flags that permit the existence of unqualified sender and recipient
4384 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4385 non-local messages, and by -bnq for local messages) were not being
4386 preserved with the message after it was received.
4388 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4389 "next input=" as part of the text comprising the host identity instead of
4390 the correct text. The code was using the same buffer for two different
4391 strings. However, depending on which order the printing function evaluated
4392 its arguments, the bug did not always show up. Under Linux, for example, my
4393 test suite worked just fine.
4395 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4396 doesn't work with some older versions of Perl. It has been changed to "my",
4397 which in any case is probably the better facility to use.
4399 15. A really picky compiler found some instances of statements for creating
4400 error messages that either had too many or two few arguments for the format
4403 16. The size of the buffer for calls to the DNS resolver has been increased
4404 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4405 for addresses that have a lot of PTR records. This alleviates a problem; it
4406 does not fully solve it.
4408 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4409 buffer now truncates the list and logs the incident, which is the same
4410 action as happens when Exim is looking up a host name and its aliases.
4411 Previously in this situation something unpredictable would happen;
4412 sometimes it was "internal error: store_reset failed".
4414 18. If a server dropped the connection unexpectedly when an Exim client was
4415 using GnuTLS and trying to read a response, the client delivery process
4416 crashed while trying to generate an error log message.
4418 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4419 string, for example:
4421 warn message = H1: something\nH2: something
4423 the text was added as a single header line from Exim's point of view
4424 though it ended up OK in the delivered message. However, searching for the
4425 second and subsequent header lines using $h_h2: did not work. This has been
4426 fixed. Similarly, if a system filter added multiple headers in this way,
4427 the routers could not see them.
4429 20. Expanded the error message when iplsearch is called with an invalid key to
4430 suggest using net-iplsearch in a host list.
4432 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4433 ACLs are no longer actually imposed (and a message to that effect is
4436 22. If a "gecos" field in a passwd entry contained escaped characters, in
4437 particular, if it contained a \" sequence, Exim got it wrong when building
4438 a From: or a Sender: header from that name. A second bug also caused
4439 incorrect handling when an unquoted " was present following a character
4440 that needed quoting.
4442 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4443 was not being matched caselessly.
4445 24. Arranged for all hyphens in the exim.8 source to be escaped with
4448 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4449 properly. Recipient callout cache records were still being keyed to include
4450 the sender, even when use_sender was set false. This led to far more
4451 callouts that were necessary. The sender is no longer included in the key
4452 when use_sender is false.
4454 26. Added "control = submission" modifier to ACLs.
4456 27. Added the ${base62d: operator to decode base 62 numbers.
4458 28. dnsdb lookups can now access SRV records.
4460 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4461 the configuration file.
4463 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4464 (-v) mode. This makes the output for a verbose queue run more intelligible.
4466 31. Added a use_postmaster feature to recipient callouts.
4468 32. Added the $body_zerocount variable, containing the number of binary zero
4469 bytes in the message body.
4471 33. The time of last modification of the "new" subdirectory is now used as the
4472 "mailbox time last read" when there is a quota error for a maildir
4475 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4477 35. Added +ignore_unknown as a special item in host lists.
4479 36. Code for decoding IPv6 addresses in host lists is now included, even if
4480 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4481 address was recognized as an IP address, but was then not correctly decoded
4482 into binary, causing unexpected and incorrect effects when compared with
4489 1. Very minor rewording of debugging text in manualroute to say "list of
4490 hosts" instead of "hostlist".
4492 2. If verify=header_syntax was set, and a header line with an unqualified
4493 address (no domain) and a large number of spaces between the end of the
4494 name and the colon was received, the reception process suffered a buffer
4495 overflow, and (when I tested it) crashed. This was caused by some obsolete
4496 code that should have been removed. The fix is to remove it!
4498 3. When running in the test harness, delay a bit after writing a bounce
4499 message to get a bit more predictability in the log output.
4501 4. Added a call to search_tidyup() just before forking a reception process. In
4502 theory, someone could use a lookup in the expansion of smtp_accept_max_
4503 per_host which, without the tidyup, could leave open a database connection.
4505 5. Added the variables $recipient_data and $sender_data which get set from a
4506 lookup success in an ACL "recipients" or "senders" condition, or a router
4507 "senders" option, similar to $domain_data and $local_part_data.
4509 6. Moved the writing of debug_print from before to after the "senders" test
4512 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4513 problems for message scanning, either using a data ACL, or using
4514 local_scan() because the Received: header was not generated till after they
4515 were called (in order to set the time as the time of reception completion).
4516 I have revised the way this works. The header is now generated after the
4517 body is received, but before the ACL or local_scan() are called. After they
4518 are run, the timestamp in the header is updated.
4524 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4525 before starting a queue runner without re-exec. This happened only when
4526 deliver_drop_privilege was set or when the Exim user was set to root. The
4527 effect of the bug was that timeouts during subsequent deliveries caused
4528 crashes instead of being properly handled. The handler is now left at its
4529 default (and expected) setting.
4531 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4532 message, again when deliver_drop_privilege is set or Exim is run as root.
4533 The bug described in (1) was not present in this case, but the tidying up
4534 of the other signals was missing. I have made the two cases consistent.
4536 3. The ignore_target_hosts setting on a manualroute router was being ignored
4537 for hosts that were looked up using the /MX notation.
4539 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4542 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4543 operated on the sender address. After changing the $sender_address to <>
4544 for the sender address verify, Exim was re-instated it as the original
4545 (before rewriting) address, but remembering that it had rewritten it, so it
4546 wasn't rewriting it again. This bug also had the effect of breaking the
4547 sender address verification caching when the sender address was rewritten.
4549 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4550 This has been changed so that if the ip literal address matches
4551 ignore_target_hosts, the router declines.
4553 7. Added expansion conditions match_domain, match_address, and match_local_
4554 part (NOT match_host).
4556 8. The placeholder for the Received: header didn't have a length field set.
4558 9. Added code to Exim itself and to exim_lock to test for a specific race
4559 condition that could lead to file corruption when using MBX delivery. The
4560 issue is with the lockfile that is created in /tmp. If this file is removed
4561 after a process has opened it but before that process has acquired a lock,
4562 there is the potential for a second process to recreate the file and also
4563 acquire a lock. This could lead to two Exim processes writing to the file
4564 at the same time. The added code performs the same test as UW imapd; it
4565 checks after acquiring the lock that its file descriptor still refers to
4566 the same named file.
4568 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4569 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4572 11. Added the smtp_active_hostname option. If used, this will typically be made
4573 to depend on the incoming interface address. Because $interface_address is
4574 not set up until the daemon has forked a reception process, error responses
4575 that can happen earlier (such as "too many connections") no longer contain
4578 12. If an expansion in a condition on a "warn" statement fails because a lookup
4579 defers, the "warn" statement is abandoned, and the next ACL statement is
4580 processed. Previously this caused the whole ACL to be aborted.
4582 13. Added the iplsearch lookup type.
4584 14. Added ident_timeout as a log selector.
4586 15. Added tls_certificate_verified as a log selector.
4588 16. Added a global option tls_require_ciphers (compare the smtp transport
4589 option of the same name). This controls incoming TLS connections.
4591 17. I finally figured out how to make tls_require_ciphers do a similar thing
4592 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4593 before starting the TLS session.
4595 18. Tabs are now shown as \t in -bP output.
4597 19. If the log selector return_path_on_delivery was set, Exim crashed when
4598 bouncing a message because it had too many Received: header lines.
4600 20. If two routers both had headers_remove settings, and the first one included
4601 a superfluous trailing colon, the final name in the first list and the
4602 first name in the second list were incorrectly joined into one item (with a
4603 colon in the middle).
4609 1. Added -C and -D options to the exinext utility, mainly to make it easier
4610 to include in the automated testing, but these could be helpful when
4611 multiple configurations are in use.
4613 2. The exinext utility was not formatting the output nicely when there was
4614 an alternate port involved in the retry record key, nor when there was a
4615 message id as well (for retries that were specific to a specific message
4616 and a specific host). It was also confused by IPv6 addresses, because of
4617 the additional colons they contain. I have fixed the IPv4 problem, and
4618 patched it up to do a reasonable job for IPv6.
4620 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4621 delivery, the log line now contains "pipelined" if PIPELINING was used.
4623 4. An SMTP transport process used to panic and die if the bind() call to set
4624 an explicit outgoing interface failed. This has been changed; it is now
4625 treated in the same way as a connect() failure.
4627 5. A reference to $sender_host_name in the part of a conditional expansion
4628 that was being skipped was still causing a DNS lookup. This no longer
4631 6. The def: expansion condition was not recognizing references to header lines
4632 that used bh_ and bheader_.
4634 7. Added the _cache feature to named lists.
4636 8. The code for checking quota_filecount in the appendfile transport was
4637 allowing one more file than it should have been.
4639 9. For compatibility with Sendmail, the command line option
4647 and sets the incoming protocol and host name (for trusted callers). The
4648 host name and its colon can be omitted when only the protocol is to be set.
4649 Note the Exim already has two private options, -pd and -ps, that refer to
4650 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4651 "s", but I don't think that's a major issue.
4653 10. A number of refactoring changes to the code, none of which should affect
4656 (a) The number of logging options was getting close to filling up the
4657 32-bit word that was used as a bit map. I have split them into two classes:
4658 those that are passed in the argument to log_write(), and those that are
4659 only ever tested independently outside of that function. These are now in
4660 separate 32-bit words, so there is plenty of room for expansion again.
4661 There is no change in the user interface or the logging behaviour.
4663 (b) When building, for example, log lines, the code previously used a
4664 macro that called string_cat() twice, in order to add two strings. This is
4665 not really sufficiently general. Furthermore, there was one instance where
4666 it was actually wrong because one of the argument was used twice, and in
4667 one call a function was used. (As it happened, calling the function twice
4668 did not affect the overall behaviour.) The macro has been replaced by a
4669 function that can join an arbitrary number of extra strings onto a growing
4672 (c) The code for expansion conditions now uses a table and a binary chop
4673 instead of a serial search (which was left over from when there were very
4674 few conditions). Also, it now recognizes conditions like "pam" even when
4675 the relevant support is not compiled in: a suitably worded error message is
4676 given if an attempt is made to use such a condition.
4678 11. Added ${time_interval:xxxxx}.
4680 12. A bug was causing one of the ddress fields not to be passed back correctly
4681 from remote delivery subprocesses. The field in question was not being
4682 subsequently used, so this caused to problems in practice.
4684 13. Added new log selectors queue_time and deliver_time.
4686 14. Might have fixed a bug in maildirsizefile handling that threw up
4687 "unexpected character" debug warnings, and recalculated the data
4688 unnecessarily. In any case, I expanded the warning message to give more
4691 15. Added the message "Restricted characters in address" to the statements in
4692 the default ACL that block characters like @ and % in local parts.
4694 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4695 Three changes have been made:
4697 (a) There was a serious bug; a negative response to MAIL caused the whole
4698 recipient domain to be cached as invalid, thereby blocking all messages
4699 to all local parts at the same domain, from all senders. This bug has
4700 been fixed. The domain is no longer cached after a negative response to
4701 MAIL if the sender used is not empty.
4703 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4706 (c) A new callout option, "use_sender" has been added for people who want
4707 the modified behaviour.
4713 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4716 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4717 indeed breaks things for older releases.
4719 3. Added additional logging to the case where there is a problem reading data
4720 from a filter that is running in a subprocess using a pipe, in order to
4721 try to track down a specific problem.
4723 4. Testing facility fudge: when running in the test harness and attempting
4724 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4725 getting "No route to host". Convert this to a timeout.
4727 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4730 6. Some OS don't have socklen_t but use size_t instead. This affects the
4731 fifth argument of getsockopt() amongst other things. This is now
4732 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4733 can be set for individual OS. I have set it for SunOS5, OSF1, and
4734 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4735 some earlier ones do not.
4737 7. Change 4.30/15 was not doing the test caselessly.
4739 8. The standard form for an IPv6 address literal was being rejected by address
4740 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4741 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4742 this, as well as the form without the "IPv6" on the front (but only when
4743 address literals are enabled, of course).
4745 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4747 10. Exim crashed if a message with an empty sender address specified by -f
4748 encountered a router with an errors_to setting. This could be provoked only
4749 by a command such as
4753 where an empty string was supplied; "<>" did not hit this bug.
4755 11. Installed PCRE release 4.5.
4757 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4758 remained set. It is now erased.
4760 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4761 times from message ids (which are base 36 rather than the normal 62).
4763 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4764 were being counted as actual protocol errors, and logged if the log
4765 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4766 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4767 and DATA following a set of rejected RCPTs do not count as protocol errors.
4768 In other words, Exim assumes they were pipelined, though this may not
4769 actually be the case. Of course, in all cases the client gets an
4770 appropriate error code.
4772 15. If a lookup fails in an ACL condition, a message about the failure may
4773 be available; it is used if testing the ACL cannot continue, because most
4774 such messages specify what the cause of the deferral is. However, some
4775 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4776 that caused an old message to be retained and used if a later statement
4777 caused a defer, replacing the real cause of the deferral.
4779 16. If an IP address had so many PTR records that the DNS lookup buffer
4780 was not large enough to hold them, Exim could crash while trying to process
4781 the truncated data. It now detects and logs this case.
4783 17. Further to 4.21/58, another change has been made: if (and only if) the
4784 first line of a message (the first header line) ends with CRLF, a bare LF
4785 in a subsequent header line has a space inserted after it, so as not to
4786 terminate the header.
4788 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4789 unnecessarily, used atoi() instead of strtol(), and didn't check the
4790 termination when getting file sizes from file names by regex.
4792 19. Completely re-implemented the support for maildirsize files, in the light
4793 of a number of problems with the previous contributed implementation
4794 (4.30/29). In particular:
4796 . If the quota is zero, the maildirsize file is maintained, but no quota is
4799 . If the maildir directory does not exist, it is created before any attempt
4800 to write a maildirsize file.
4802 . The quota value in the file is just a cache; if the quota is changed in
4803 the transport, the new value overrides.
4805 . A regular expression is available for excluding directories from the
4808 20. The autoreply transport checks the characters in options that define the
4809 message's headers; it allows continued headers, but it was checking with
4810 isspace() after an embedded newline instead of explicitly looking for a
4813 21. If all the "regular" hosts to which an address was routed had passed their
4814 expiry times, and had not reached their retry times, the address was
4815 bounced, even if fallback hosts were defined. Now Exim should go on to try
4818 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4819 equivalent code in the SMTP transport. Some hosts send humungous responses
4820 to HELO/EHLO, more than 1024 it seems.
4822 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4823 gives compiler warnings in some environments. I've now done it "properly",
4826 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4827 (because of problems with the built-in one) was declared to return uschar *
4828 instead of char *, causing compiler failure.
4830 25. Fixed a file descriptor leak when processing alias/forward files.
4832 26. Fixed a minor format string issue in dbfn.c.
4834 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4836 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4837 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4840 29. When a sender address is verified, it is cached, to save repeating the test
4841 when there is more than one recipient in a message. However, when the
4842 verification involves a callout, it is possible for different callout
4843 options to be set for different recipients. It is too complicated to keep
4844 track of this in the cache, so now Exim always runs a verification when a
4845 callout is required, relying on the callout cache for the optimization.
4846 The overhead is duplication of the address routing, but this should not be
4849 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4850 to be verified with callout=postmaster, and the main callout worked but the
4851 postmaster check failed, the verification correctly failed. However, if a
4852 subsequent RCPT command asked for sender verification *without* the
4853 postmaster check, incorrect caching caused this verification also to fail,
4856 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4857 it was not caching the DNS options (qualify_single, search_parents) that
4858 were used when the lookup failed. A subsequent lookup with different
4859 options therefore always gave the same answer, though there were cases
4860 where it should not have. (Example: a "domains = !$mx_any" option on a
4861 dnslookup router: the "domains" option is always processed without any
4862 widening, but the router might have qualify_single set.) Now Exim uses the
4863 cached value only when the same options are set.
4865 32. Added John Jetmore's "exipick" utility to the distribution.
4867 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4868 than a timeout (e.g. a certificate is required, and is not provided), an
4869 Exim server now closes the connection immediately. Previously it waited for
4870 the client to close - but if the client is SSL, it seems that they each
4871 wait for each other, leading to a delay before one of them times out.
4873 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4874 maintained 0.8.x compatibility because I don't think many are using it, and
4875 it is clearly obsolete.
4877 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4880 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4881 client certificate was expired. A simple patch fixes this, though I don't
4882 understand the full logic of why the verify callback is called multiple
4885 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4886 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4887 which causes problems with some clients (such as the Certicom SSL Plus
4888 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4889 disables the coutermeasure allowing Eudora to connect."
4891 38. Exim was not checking that a write() to a log file succeeded. This could
4892 lead to Bad Things if a log got too big, in particular if it hit a file
4893 size limit. Exim now panics and dies if it cannot write to a log file, just
4894 as it does if it cannot open a log file.
4896 39. Modified OS/Makefile-Linux so that it now contains
4898 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4900 The two -D definitions ensure that Exim is compiled with large file
4901 support, which makes it possible to handle log files that are bigger than
4904 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4905 instance) a domain was checked against a named list that involved a lookup,
4906 causing $domain_data to be set, then another domain was checked against the
4907 same list, then the first domain was re-checked, the value of $domain_data
4908 after the final check could be wrong. In particular, if the second check
4909 failed, it could be set empty. This bug probably also applied to
4912 41. The strip_trailing_dot option was not being applied to the address given
4913 with the -f command-line option.
4915 42. The code for reading a message's header from the spool was incrementing
4916 $received_count, but never initializing it. This meant that the value was
4917 incorrect (doubled) while delivering a message in the same process in which
4918 it was received. In the most common configuration of Exim, this never
4919 happens - a fresh exec is done - but it can happen when
4920 deliver_drop_privilege is set.
4922 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4923 it now includes up to 150 characters of the unexpected data in the log
4926 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4927 and building data strings. The size of both of these buffers was 10 000
4928 bytes - far larger than anybody would *ever* want, thought I. Needless to
4929 say, somebody hit the limit. I have increased the maximum line length to
4930 20 000 and the maximum data length of concatenated lines to 100 000. I have
4931 also fixed two bugs, because there was no checking on these buffers. Tsk,
4932 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4933 buffer is too small.
4935 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4936 lsearch lookups. Now it does.
4938 46. When parsing a route_list item in a manualroute router, a fixed-length
4939 buffer was used for the list of hosts. I made this 1024 bytes long,
4940 thinking that nobody would ever have a list of hosts that long. Wrong.
4941 Somebody had a whole pile of complicated expansion conditions, and the
4942 string was silently truncated, leading to an expansion error. It turns out
4943 that it is easier to change to an unlimited length (owing to other changes
4944 that have happened since this code was originally written) than to build
4945 structure for giving a limitation error. The length of the item that
4946 expands into the list of hosts is now unlimited.
4948 47. The lsearch lookup could not handle data where the length of text line was
4949 more than 4095 characters. Such lines were truncated, leading to shortened
4950 data being returned. It should now handle lines of any length.
4952 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4953 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4956 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4957 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4958 no longer happens in comment lines. A list of replacements is now placed
4959 at the head of all of the source files, except those whose only change is
4960 to replace PERL_COMMAND in the very first #! line.
4962 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4963 messages on the queue, with a bottom-up merge sort, using code contributed
4964 by Michael Haardt. This should make operations like -bp somewhat faster on
4965 large queues. It won't affect queue runners, except when queue_run_in_order
4968 51. Installed eximstats 1.31 in the distribution.
4970 52. Added support for SRV lookups to the dnslookup router.
4972 53. If an ACL referred to $message_body or $message_body_end, the value was not
4973 reset for any messages that followed in the same SMTP session.
4975 54. The store-handling optimization for building very long strings was not
4976 differentiating between the different store pools. I don't think this
4977 actually made any difference in practice, but I've tidied it.
4979 55. While running the routers to verify a sender address, $sender_address
4980 was still set to the sender address. This is wrong, because when routing to
4981 send a bounce to the sender, it would be empty. Therefore, I have changed
4982 it so that, while verifying a sender address, $sender_address is set to <>.
4983 (There is no change to what happens when verifying a recipient address.)
4985 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4986 target A or AAAA records (if not already returned) without resetting the
4987 qualify_single or search_parents options of the DNS resolver. These are
4988 inappropriate in this case because the targets of MX and SRV records must
4989 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4990 target that, when qualified, matched something in the local domain. These
4991 two options are now turned off when doing these lookups.
4993 57. It seems that at least some releases of Reiserfs (which does not have the
4994 concept of a fixed number of inodes) returns zero and not -1 for the
4995 number of available inodes. This interacted badly with check_spool_inodes,
4996 which assumed that -1 was the "no such thing" setting. What I have done is
4997 to check that the total number of inodes is greater than zero before doing
4998 the test of how many are available.
5000 58. When a "warn" ACL statement has a log_message modifier, the message is
5001 remembered, and not repeated. This is to avoid a lot of repetition when a
5002 message has many recipients that cause the same warning to be written.
5003 Howewer, Exim was preserving the list of already written lines for an
5004 entire SMTP session, which doesn't seem right. The memory is now reset if a
5005 new message is started.
5007 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5008 debugging output unless log_rewrite was also set.
5010 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5011 of (int)(handle) when we know that handle contains (void *)(-1).
5013 61. The Exim daemon panic-logs an error return when it closes the incoming
5014 connection. However "connection reset by peer" seems to be common, and
5015 isn't really an error worthy of noting specially, so that particular error
5018 62. When Exim is trying to find all the local interfaces, it used to panic and
5019 die if the ioctl to get the interface flags failed. However, it seems that
5020 on at least one OS (Solaris 9) it is possible to have an interface that is
5021 included in the list of interfaces, but for which you get a failure error
5022 for this call. This happens when the interface is not "plumbed" into a
5023 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5024 failure of the "get flags" call assumes that the interface is down.
5026 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5027 makes life easier for people who are doing arithmetic on fields extracted
5028 from dates, where you often get leading zeros that should not be
5029 interpreted as octal.
5031 64. Added qualify_domain to the redirect router, to override the global
5034 65. If a pathologically long header line contained very many addresses (the
5035 report of this problem mentioned 10 000) and each of them was rewritten,
5036 Exim could use up a very large amount of memory. (It kept on making new
5037 copies of the header line as it rewrote, and never released the old ones.)
5038 At the expense of a bit more processing, the header rewriting function has
5039 been changed so that it no longer eats memory in this way.
5041 66. The generation of the Received: header has been moved from the time that a
5042 message starts to be received, to the time that it finishes. The timestamp
5043 in the Received: header should now be very close to that of the <= log
5044 line. There are two side-effects of this change:
5046 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5047 logged header lines no longer include the local Received: line, because
5048 it has not yet been created. The same applies to a copy of the message
5049 that is returned to a non-SMTP sender when a message is rejected.
5051 (b) When a filter file is tested using -bf, no additional Received: header
5052 is added to the test message. After some thought, I decided that this
5055 This change does not affect the value of $received_for. It is still set
5056 after address rewriting, but before local_scan() is called.
5058 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5060 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5061 gave an unhelpful panic error message, and a defer error. I have managed to
5062 change this behaviour so that it now rejects any supplied certificate,
5063 which seems right, as the list of acceptable certificates is empty.
5065 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5066 gave an unhelpful defer error. I have not managed to make this reject any
5067 supplied certificates, but the error message it gives is "no certificate
5068 supplied", which is not helpful.
5070 70. exigrep's output now also includes lines that are not associated with any
5071 message, but which match the given pattern. Implemented by a patch from
5072 Martin Sluka, which also tidied up the Perl a bit.
5074 71. Recipient callout verification, like sender verification, was using <> in
5075 the MAIL FROM command. This isn't really the right thing, since the actual
5076 sender may affect whether the remote host accepts the recipient or not. I
5077 have changed it to use the actual sender in the callout; this means that
5078 the cache record is now keyed on a recipient/sender pair, not just the
5079 recipient address. There doesn't seem to be a real danger of callout loops,
5080 since a callout by the remote host to check the sender would use <>.
5081 [SEE ABOVE: changed after hitting problems.]
5083 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5084 temporary errors. However, in the case of such a code being given after
5085 the end of a data transmission (i.e. after ".") Exim was failing to write
5086 a retry record for the message. (Yes, there was some broken host that was
5087 actually sending 8xx at this point.)
5089 73. An unknown lookup type in a host list could cause Exim to panic-die when
5090 the list was checked. (An example that provoked this was putting <; in the
5091 middle of a list instead of at the start.) If this happened during a DATA
5092 ACL check, a -D file could be left lying around. This kind of configuration
5093 error no longer causes Exim to die; instead it causes a defer errror. The
5094 incident is still logged to the main and panic logs.
5096 74. Buglet left over from Exim 3 conversion. The message "too many messages
5097 in one connection" was written to the rejectlog but not the mainlog, except
5098 when address rewriting (yes!) was being logged.
5100 75. Added write_rejectlog option.
5102 76. When a system filter was run not as root (that is, when system_filter_user
5103 was set), the values of the $n variables were not being returned to the
5104 main process; thus, they were not subsequently available in the $sn
5107 77. Added +return_path_on_delivery log selector.
5109 78. A connection timeout was being treated differently from recipients deferred
5110 when testing hosts_max_try with a message that was older than the host's
5111 retry timeout. (The host should not be counted, thus allowing all hosts to
5112 be tried at least once before bouncing.) This may have been the cause of an
5113 occasionally reported bug whereby a message would remain on the queue
5114 longer than the retry timeout, but would be bounced if a delivery was
5115 forced. I say "may" because I never totally pinned down the problem;
5116 setting up timeout/retry tests is difficult. See also the next item.
5118 79. The ultimate address timeout was not being applied to errors that involved
5119 a combination of host plus message (for example, a timeout on a MAIL
5120 command). When an address resolved to a number of possible hosts, and they
5121 were not all tried for each delivery (e.g. because of hosts_max_try), a
5122 message could remain on the queue longer than the retry timeout.
5124 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5127 81. Fixed an obscure SMTP outgoing bug which required at least the following
5128 conditions: (a) there was another message waiting for the same server;
5129 (b) the server returned 5xx to all RCPT commands in the first message so
5130 that the message was not completed; (c) the server dropped the connection
5131 or gave a negative response to the RSET that Exim sends to abort the
5132 transaction. The observed case was a dropped connection after DATA that had
5133 been sent in pipelining mode. That is, the server had advertised PIPELINING
5134 but was not implementing it correctly. The effect of the bug was incorrect
5135 behaviour, such as trying another host, and this could lead to a crash.
5141 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5142 and daemon.c were passed as pointers to ints; they should have been
5143 pointers to socklen_t variables (which are typically unsigned ints).
5145 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5148 3. Fixed a really odd bug that affected only the testing scheme; patching a
5149 certain fixed string in the binary changed the value of another string that
5150 happened to be identical to the end of the original first string.
5152 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5153 name", it returns that address as the IP address. On some operating
5154 systems (e.g. Solaris), it also passes back the IP address string as the
5155 "host name". However, on others (e.g. Linux), it passes back an empty
5156 string. Exim wasn't checking for this, and was changing the host name to an
5157 empty string, assuming it had been canonicized.
5159 5. Although rare, it is permitted to have more than one PTR record for a given
5160 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5161 all the names associated with an address, because they do in Solaris.
5162 However, it seems that they do not in Linux for data that comes from the
5163 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5164 I found this out when I moved to a new Linux workstation and tried to run
5165 the Exim test suite.
5167 To get round this problem I have changed the code so that it now does its
5168 own call to the DNS to look up PTR records when searching for a host name.
5169 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5170 addresses that are only in /etc/hosts are still found.
5172 This behaviour is, however, controlled by an option called host_lookup_
5173 order, which defaults to "bydns:byaddr". If people want to use the other
5174 order, or indeed, just use one or the other means of lookup, they can
5175 specify it in this variable.
5177 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5178 some operating systems, this comes back from gethostbyaddr() as an empty
5179 string, and this is what Exim used to test for. However, it seems that in
5180 other systems, "." is yielded. Exim now tests for this case too.
5182 7. The values of check_spool_space and check_log_space are now held internally
5183 as a number of kilobytes instead of an absolute number of bytes. If a
5184 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5185 kilobyte. This means that much larger values can be stored.
5187 8. Exim monitor: an attempt to get the action menu when not actually pointing
5188 at a message produces an empty menu entitled "No message selected". This
5189 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5190 no entries in it ("Shell widget menu has zero width and/or height"). So I
5191 have added a single, blank menu entry in this case.
5193 9. Added ${quote_local_part.
5195 10. MIME decoding is now applied to the contents of Subject: header lines when
5198 11. Now that a reference to $sender_host_address automatically causes a reverse
5199 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5200 host lookup before query-style lookups in lists that might use this
5201 variable. This has therefore been abolished, and the "net-" prefix is no
5202 longer necessary for query-style lookups.
5204 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5205 have been a typo for LFLAGS, so it has been changed.
5207 13. The install script calls Exim with "-C /dev/null" in order to find the
5208 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5209 to be output. Howeve, since Exim outputs its version number before the
5210 error, it didn't break the script. It just looked ugly. I fixed this by
5211 always allowing "-C /dev/null" if the caller is root.
5213 14. Ignore overlarge ACL variable number when reading spool file - insurance
5214 against a later release with more variables having written the file.
5216 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5217 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5218 this, as well as the form without the "IPv6" on the front.
5220 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5221 OS/Makefile-Darwin file.
5223 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5224 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5226 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5227 variable. However, some OS use ints for inodes. I've added cast to long int
5228 to get rid of the compiler warning.
5230 19. I had forgotten to lock out "/../" in configuration file names when
5231 ALT_CONFIG_PREFIX was set.
5233 20. Routers used for verification do not need to specify transports. However,
5234 if such a router generated a host list, and callout was configured, Exim
5235 crashed, because it could not find a port number from the (non-existent)
5236 transport. It now assumes port 25 in this circumstance.
5238 21. Added the -t option to exigrep.
5240 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5241 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5242 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5243 them is set, LOOKUP_LSEARCH is forced.
5245 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5247 24. Added sender and host information to the "rejected by local_scan()" log
5248 line; previously there was no indication of these.
5250 25. Added .include_if_exists.
5252 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5253 Linux. It turns out that not all file systems support this. Apparently some
5254 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5255 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5256 not supported on the file descriptor, is now ignored when Exim is trying to
5257 sync a directory. This applies only to Linux.
5259 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5261 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5262 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5265 29. Added support for maildirsize files from supplied patch (modified a bit).
5267 30. The use of :fail: followed by an empty string could lead Exim to respond to
5268 sender verification failures with (e.g.):
5270 550 Verification failed for <xxx>
5271 550 Sender verify failed
5273 where the first response line was missing the '-' that indicates it is not
5274 the final line of the response.
5276 31. The loop for finding the name of the user that called Exim had a hardwired
5277 limit of 10; it now uses the value of finduser_retries, which is used for
5278 all other user lookups.
5280 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5283 33. Exim was neglecting to zero errno before one call of strtol() when
5284 expanding a string and expecting an integer value. On some systems this
5285 resulted in spurious "integer overflow" errors. Also, it was casting the
5286 result into an int without checking.
5288 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5289 did not work. The code looks as if it has *never* worked, though it appears
5290 to have been documented since at least releast 1.62. I have made it work.
5292 35. The "timeout_DNS" error in retry rules, also documented since at least
5293 1.62, also never worked. As it isn't clear exactly what this means, and
5294 clearly it isn't a major issue, I have abolished the feature by treating it
5295 as "timeout", and writing a warning to the main and panic logs.
5297 36. The display of retry rules for -brt wasn't always showing the error code
5300 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5301 timeout_connect_A, timeout_connect_MX.
5303 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5304 to the empty sender.
5306 39. The daemon was not analysing the content of -oX till after it had closed
5307 stderr and disconnected from the controlling terminal. This meant that any
5308 syntax errors were only noted on the panic log, and the return code from
5309 the command was 0. By re-arranging the code a little, I've made the
5310 decoding happen first, so such errors now appear on stderr, and the return
5311 code is 1. However, the actual setting up of the sockets still happens in
5312 the disconnected process, so errors there are still only recorded on the
5315 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5316 connections (as happens on some IP stacks) was logged at start up time as
5317 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5318 it from "IPv6 and IPv4", which means that two separate sockets are being
5321 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5322 says whether AF_INET or AF_INET6 was passed as an argument.
5324 42. Exiwhat output was messed up when time zones were included in log
5327 43. Exiwhat now gives more information about the daemon's listening ports,
5328 and whether -tls-on-connect was used.
5330 44. The "port" option of the smtp transport is now expanded.
5332 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5333 silently ignored. Now an error message is written to the main and panic
5336 46. There's a new ACL modifier called "logwrite" which writes to a log file
5337 as soon as it is encountered.
5339 47. Added $local_user_uid and $local_user_gid at routing time.
5341 48. Exim crashed when trying to verify a sender address that was being
5344 49. Exim was recognizing only a space character after ".include". It now also
5345 recognizes a tab character.
5347 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5348 extracting the relevant information from the specification. The man page no
5349 longer contains scrambled data for the -d option, and I've added a section
5350 at the front about calling Exim under different names.
5352 51. Added "extra_headers" argument to the "mail" command in filter files.
5354 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5357 53. Installed eximstats 1.29.
5359 54. Added transport_filter_timeout as a generic transport option.
5361 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5362 Cc: header lines. This was required by RFC 822, but it not required by RFC
5365 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5366 incoming messages that did not have them. Now it does so only if the
5367 message originates locally, that is, if there is no associated remote host
5368 address. When Resent- header lines are present, this applies to the Resent-
5369 lines rather than the non-Resent- lines.
5371 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5372 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5374 58. Messages for configuration errors now include the name of the main
5375 configuration file - useful now that there may be more than one file in a
5376 list (.included file names were always shown).
5378 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5379 for those rare installations that do not start the daemon as root or run it
5380 setuid root. I've cut out the call to initgroups() if the daemon is not
5383 60. The Exim user and group can now be bound into the binary as text strings
5384 that are looked up at the start of Exim's processing.
5386 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5388 62. Added $mailstore_basename variable.
5390 63. Installed patch to sieve.c from Michael Haardt.
5392 64. When Exim failed to open the panic log after failing to open the main log,
5393 the original message it was trying to log was written to stderr and debug
5394 output, but if they were not available (the usual case in production), it
5395 was lost. Now it is written to syslog before the two lines that record the
5396 failures to open the logs.
5398 65. Users' Exim filters run in subprocesses under the user's uid. It is
5399 possible for a "deliver" command or an alias in a "personal" command to
5400 provoke an address rewrite. If logging of address rewriting is configured,
5401 this fails because the process is not running as root or exim. There may be
5402 a better way of dealing with this, but for the moment (because 4.30 needs
5403 to be released), I have disabled address rewrite logging when running a
5404 filter in a non-root, non-exim process.
5410 1. The buildconfig auxiliary program wasn't quoting the value set for
5411 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5412 not defined. This bug was masked in 4.22 by the effect that was fixed in
5415 2. Some messages that were rejected after a message id was allocated were
5416 shown as "incomplete" by exigrep. It no longer does this for messages that
5417 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5419 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5420 have allow_domain_literals set, the ID did not get logged in the <= line.
5421 Domain literals are now always recognized in Message-ID: header lines.
5423 4. The first argument for a ${extract expansion item is the key name or field
5424 number. Leading and trailing spaces in this item were not being ignored,
5425 causing some misleading effects.
5427 5. When deliver_drop_privilege was set, single queue runner processes started
5428 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5429 same command in the process it spins off) were not dropping privilege.
5431 6. When the daemon running as "exim" started a queue runner, it always
5432 re-executed Exim in the spun-off process. This is a waste of effort when
5433 deliver_drop_privilege is set. The new process now just calls the
5434 queue-runner function directly.
5440 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5443 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5444 ignored. Though the use of -oP was forcing the writing of a pid file, it
5445 was always written to the default place.
5447 3. If the message "no IP address found for host xxxx" is generated during
5448 incoming verification, it is now followed by identification of the incoming
5449 connection (so you can more easily find what provoked it).
5451 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5453 5. Added some features to "harden" Exim a bit more against certain attacks:
5455 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5456 be put in Local/Makefile. This is like the never_users runtime option,
5457 but it cannot be overridden. The default setting is "root".
5459 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5460 prefix string with which any file named in a -C command line option
5463 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5464 is retained for -C and -D only if the caller of Exim is root. Without
5465 it, the exim user may also use -C and -D and retain privilege.
5467 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5468 command line option is disabled.
5470 6. Macro names set by the -D option must start with an upper case letter, just
5471 like macro names defined in the configuration file.
5473 7. Added "dereference=" facility to LDAP.
5475 8. Two instances of the typo "uknown" in the source files are fixed.
5477 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5478 the Configure-Makefile script screwed up while processing it.
5480 10. Incorporated PCRE 4.4.
5482 11. The SMTP synchronization check was not operating right at the start of an
5483 SMTP session. For example, it could not catch a HELO sent before the client
5484 waited for the greeting. There is now a check for outstanding input at the
5485 point when the greeting is written. Because of the duplex, asynchronous
5486 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5487 way, but not yet received, when the check is performed.
5489 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5490 on TCP/IP sockets, because this apparently causes some broken clients to
5493 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5494 unchanged) from the Cygwin maintainer.
5496 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5497 was supported instead of testing for mbx. Effectively a typo.
5499 15. The spa authenticator server code was not checking that the input it
5500 received was valid base64.
5502 16. The debug output line for the "set" modifier in ACLs was not showing the
5503 name of the variable that was being set.
5505 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5507 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5508 lookup on its own. Something else was needed to trigger the lookup. For
5509 example, a match in host_lookup or the need for a host name in a host list.
5510 Now, if $sender_host_name is referenced and the host name has not yet been
5511 looked up, a lookup is performed. If the lookup fails, the variable remains
5512 empty, and $host_lookup_failed is set to "1".
5514 19. Added "eqi" as a case-independent comparison operator.
5516 20. The saslauthd authentication condition could segfault if neither service
5517 nor realm was specified.
5519 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5520 error message that was logged was misleading, and incoming SMTP
5521 connections were dropped. The message is now more accurate, and temporary
5522 errors are given to SMTP connections.
5524 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5525 (except RSET) with a 421 error, until QUIT is received. However, it was
5526 failing to send a response to QUIT.
5528 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5529 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5530 ACL is now run after the helo_try_verify_hosts code.
5532 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5533 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5534 case-independent, so other case variants are also recognized). Apparently
5535 some systems use these upper case variants.
5537 25. If more than two messages were waiting for the same host, and a transport
5538 filter was specified for the transport, Exim sent two messages over the
5539 same TCP/IP connection, and then failed with "socket operation on non-
5540 socket" when it tried to send the third.
5542 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5544 27. The extern definition of crypt16() in expand.c was not being excluded when
5545 the OS had its own crypt16() function.
5547 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5548 as a preferred synonym for return_size_limit, both as an option and as an
5551 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5553 30. Changed the default configuration ACL to relax the local part checking rule
5554 for addresses that are not in any local domains. For these addresses,
5555 slashes and pipe symbols are allowed within local parts, but the sequence
5556 /../ is explicitly forbidden.
5558 31. SPA server authentication was not clearing the challenge buffer before
5561 32. log_message in a "warn" ACL statement was writing to the reject log as
5562 well as to the main log, which contradicts the documentation and doesn't
5563 seem right (because no rejection is happening). So I have stopped it.
5565 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5566 However, I am unable to do any testing of this.
5568 34. Fixed an infelicity in the appendfile transport. When checking directories
5569 for a mailbox, to see if any needed to be created, it was accidentally
5570 using path names with one or more superfluous leading slashes; tracing
5571 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5573 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5574 contents are added to the log line that is written for every discarded
5575 recipient. (Previously a log_message setting was ignored.)
5577 36. The ${quote: operator now quotes the string if it is empty.
5579 37. The install script runs exim in order to find its version number. If for
5580 some reason other than non-existence or emptiness, which it checks, it
5581 could not run './exim', it was installing it with an empty version number,
5582 i.e. as "exim-". This error state is now caught, and the installation is
5585 38. An argument was missing from the function that creates an error message
5586 when Exim fails to connect to the socket for saslauthd authentication.
5587 This could cause Exim to crash, or give a corrupted message.
5589 39. Added isip, isip4, and isip6 to ${if conditions.
5591 40. The ACL variables $acl_xx are now saved with the message, and can be
5592 accessed later in routers, transports, and filters.
5594 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5595 strings in the file are not string-expanded.
5597 42. If a MAIL command specified a SIZE value that was too large to fit into an
5598 int variable, the check against message_size_limit failed. Such values are
5599 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5600 day this will have to be increased, but I don't think I want to be around
5601 when emails are that large.
5608 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5609 iconv() is not standard in FreeBSD.
5611 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5612 IPv6 enabled. The observed symptom was a segmentation fault on return from
5613 the function os_common_find_running_interfaces() in src/os.c.
5615 3. In the check_special_case() function in daemon.c I had used "errno" as an
5616 argument name, which causes warnings on some systems. This was basically a
5617 typo, since it was named "eno" in the comments!
5619 4. The code that waits for the clock to tick (at a resolution of some fraction
5620 of a second) so as to ensure message-id uniqueness was always waiting for
5621 at least one whole tick, when it could have waited for less. [This is
5622 almost certainly not relevant at current processor speeds, where it is
5623 unlikely to ever wait at all. But we try to future-proof.]
5625 5. The function that sleeps for a time interval that includes fractions of a
5626 second contained a race. It did not block SIGALRM between setting the
5627 timer, and suspending (a couple of lines later). If the interval was short
5628 and the sigsuspend() was delayed until after it had expired, the suspension
5629 never ended. On busy systems this could lead to processes getting stuck for
5632 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5633 process, before it forks any delivery processes. The open lookup caching
5634 mechanism meant that the open file or database connection was passed into
5635 the delivery process. The problem was that delivery processes always tidy
5636 up cached lookup data. This could cause a problem for the next delivery
5637 process started by the queue runner, because the external queue runner
5638 process does not know about the closure. So the next delivery process
5639 still has data in the lookup cache. In the case of a file lookup, there was
5640 no problem because closing a file descriptor in a subprocess doesn't affect
5641 the parent. However, if the lookup was caching a connection to a database,
5642 the connection was closed, and the second delivery process was likely to
5643 see errors such as "PGSQL: query failed: server closed the connection
5644 unexpectedly". The problem has been fixed by closing all cached lookups
5645 in a queue runner before running a delivery process.
5647 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5648 seem to have the "const" qualifier which it has on other OS. I've
5651 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5652 *of the same type* (client or server) with the same public name that an
5653 error should be diagnosed.
5655 9. When Exim looked up a host name for an IP address, but failed to find the
5656 original IP address when looking up the host name (a safety check), it
5657 output the message "<ip address> does not match any IP for NULL", which was
5658 confusing, to say the least. The bug was that the host name should have
5659 appeared instead of "NULL".
5661 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5662 user that is built into the binary, and the -C or -D options is used, root
5663 privilege is dropped before the configuration file is read. In addition,
5664 logging is switched to stderr instead of the normal log files. If the
5665 configuration then re-defines the Exim user, the unprivileged environment
5666 is probably not what is expected, so Exim logs a panic warning message (but
5669 However, if deliver_drop_privilege is set, the unprivileged state may well
5670 be exactly what is intended, so the warning has been cut out in that case,
5671 and Exim is allowed to try to write to its normal log files.
5677 1. smtp_return_error_details was not giving details for temporary sender
5678 or receiver verification errors.
5680 2. Diagnose a configuration error if two authenticators have the same public
5683 3. Exim used not to create the message log file for a message until the first
5684 delivery attempt. This could be confusing when incoming messages were held
5685 for policy or load reasons. The message log file is now created at the time
5686 the message is received, and an initial "Received" line is written to it.
5688 4. The automatically generated man page for command line options had a minor
5689 bug that caused no ill effects; however, a more serious problem was that
5690 the procedure for building the man page automatically didn't always
5691 operate. Consequently, release 4.20 contains an out-of-date version. This
5692 shouldn't happen again.
5694 5. When building Exim with embedded Perl support, the script that builds the
5695 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5696 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5698 6. The freeze_tell option was not being used for messages that were frozen on
5699 arrival, either by an ACL or by local_scan().
5701 7. Added the smtp_incomplete_transaction log selector.
5703 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5704 was accepting AUTH without a new EHLO.
5706 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5707 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5710 10. Logging of TCP/IP connections (when configured) now happens in the main
5711 daemon process instead of the child process, so that the TCP/IP connection
5712 count is more accurate (but it can never be perfect).
5714 11. The use of "drop" in a nested ACL was not being handled correctly in the
5715 outer ACL. Now, if condition failure induced by the nested "drop" causes
5716 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5717 or "require"), the connection is dropped.
5719 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5720 that yield "discard" can now be used with an "accept" or a "discard" verb,
5721 but an error is generated for any others (because I can't see a useful way
5722 to define what should happen).
5724 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5725 are now processed in the same way as lines in the Exim configuration file.
5726 In particular, continuation lines are supported.
5728 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5730 15. Added -ti meaning -t -i.
5732 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5733 domains, and warn by logging if others are found.
5735 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5736 returned from GIFCONF when Exim is trying to find the list of interfaces on
5737 a host. The code in os.c has been modified to copy each ifreq to an aligned
5738 structure in all cases.
5740 Also, in some cases, the returned ifreq's were being copied to a 'struct
5741 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5742 means the last couple of bytes of an IPv6 address could be chopped if the
5743 ifreq contained only a normal sockaddr (14 bytes storage).
5745 18. Named domain lists were not supported in the hosts_treat_as_local option.
5746 An entry such as +xxxx was not recognized, and was treated as a literal
5749 19. Ensure that header lines added by a DATA ACL are included in the reject log
5750 if the ACL subsequently rejects the message.
5752 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5753 MD5 (which is deprecated).
5755 21. When testing a filter file using -bf, Exim was writing a message when it
5756 took the sender from a "From " line in the message, but it was not doing so
5757 when it took $return_path from a Return-Path: header line. It now does.
5759 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5760 with a valid header line field name (a series of printing characters
5761 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5763 23. Changed "disc" in the source to "disk" to conform to the documentation and
5764 the book and for uniformity.
5766 24. Ignore Sendmail's -Ooption=value command line item.
5768 25. When execve() failed while trying to run a command in a pipe transport,
5769 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5770 could be confused with a return value of 69 from the command itself. This
5771 has been changed to 127, the value the shell returns if it is asked to run
5772 a non-existent command. The wording for the related log line suggests a
5773 non-existent command as the problem.
5775 26. If received_header_text expands to an empty string, do not add a Received:
5776 header line to the message. (Well, it adds a token one on the spool, but
5777 marks it "old" so that it doesn't get used or transmitted.)
5779 27. Installed eximstats 1.28 (addition of -nt option).
5781 28. There was no check for failure on the call to getsockname() in the daemon
5782 code. This can fail if there is a shortage of resources on the system, with
5783 ENOMEM, for example. A temporary error is now given on failure.
5785 29. Contrary to the C standard, it seems that in some environments, the
5786 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5787 program. Exim now does this explicitly; it affects the formatting of
5788 timestamps using strftime().
5790 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5791 complaints. I've now initialized all the variables, to avoid this.
5793 32. Header lines added by a system filter were not being "seen" during
5794 transport-time rewrites.
5796 33. The info_callback() function passed to OpenSSL is set up with type void
5797 (*)(SSL *, int, int), as described somewhere. However, when calling the
5798 function (actually a macro) that sets it up, the type void(*)() is
5799 expected. I've put in a cast to prevent warnings from picky compilers.
5801 34. If a DNS black list lookup found a CNAME record, but there were no A
5802 records associated with the domain it pointed at, Exim crashed.
5804 35. If a DNS black list lookup returned more than one A record, Exim ignored
5805 all but the first. It now scans all returned addresses if a particular IP
5806 value is being sought. In this situation, the contents of the
5807 $dnslist_value variable are a list of all the addresses, separated by a
5810 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5811 to accept a host name and all its aliases if the forward lookup for any of
5812 them yielded the IP address of the incoming connection. Now it accepts only
5813 those names whose forward lookup yields the correct IP address. Any other
5814 names are discarded. This closes a loophole whereby a rogue DNS
5815 administrator could create reverse DNS records to break through a
5816 wildcarded host restriction in an ACL.
5818 37. If a user filter or a system filter that ran in a subprocess used any of
5819 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5820 the wrong values were passed to the pipe command ($thisaddress had the
5821 value of $0, $0 had the value of $1, etc). This bug was introduced by
5822 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5824 38. Improved the line breaking for long SMTP error messages from ACLs.
5825 Previously, if there was no break point between 40 and 75 characters, Exim
5826 left the rest of the message alone. Two changes have been made: (a) I've
5827 reduced the minimum length to 35 characters; (b) if it can't find a break
5828 point between 35 and 75 characters, it looks ahead and uses the first one
5829 that it finds. This may give the occasional overlong line, but at least the
5830 remaining text gets split now.
5832 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5833 file descriptors might be low, and that setting 1000 would always raise it.
5834 It turns out that in some environments, the limit is already over 1000 and
5835 that lowering it causes trouble. So now Exim takes care not to decrease it.
5837 40. When delivering a message, the value of $return_path is set to $sender_
5838 address at the start of routing (routers may change the value). By an
5839 oversight, this default was not being set up when an address was tested by
5840 -bt or -bv, which affected the outcome if any router or filter referred to
5843 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5844 log only when "message" or "log_message" are set. However, if one of the
5845 conditions was an address verification, or a call to a nested ACL, the
5846 messages generated by the underlying test were being passed through. This
5847 no longer happens. The underlying message is available in $acl_verify_
5848 message for both "message" and "log_message" expansions, so it can be
5849 passed through if needed.
5851 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5852 new expansion $bh_ to give the encoded byte string without charset
5853 translation. Translation happens only if iconv() is available; HAVE_ICONV
5854 indicates this at build time. HEADERS_CHARSET gives the charset to
5855 translate to; headers_charset can change it in the configuration, and
5856 "headers charset" can change it in an individual filter file.
5858 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5859 that creates RFC 2047 encoded "words" labels them as that charset instead
5860 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5861 expansion operator; (ii) when Exim creates a From: line for a local
5862 message; (iii) when a header line is rewritten to include a "phrase" part.
5864 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5865 buggy, causing it to skip the first lines of messages whose message ID
5866 ended in 'D'. This would not have bitten before Exim release 4.14, because
5867 message IDs were unlikely to end in 'D' before then. The effect was to have
5868 incorrect size information for certain domains.
5870 45. #include "config.h" was missing at the start of the crypt16.c module. This
5871 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5874 46. If there was a timeout during a "random" callout check, Exim treated it as
5875 a failure of the random address, and carried on sending RSET and the real
5876 address. If the delay was just some slowness somewhere, the response to the
5877 original RCPT would be taken as a response to RSET and so on, causing
5878 mayhem of various kinds.
5880 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5881 when I implemented it. It didn't allow for the fact that some option values
5882 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5883 the right test for positive values.
5885 48. Domain names in DNS records are case-independent. Exim always looks them up
5886 in lower case. Some resolvers return domain names in exactly the case they
5887 appear in the zone file, that is, they may contain uppercase letters. Not
5888 all resolvers do this - some return always lower case. Exim was treating a
5889 change of case by a resolver as a change of domain, similar to a widening
5890 of a domain abbreviation. This triggered its re-routing code and so it was
5891 trying to route what was effectively the same domain again. This normally
5892 caused routing to fail (because the router wouldn't handle the domain
5893 twice). Now Exim checks for this case specially, and just changes the
5894 casing of the domain that it ultimately uses when it transmits the message
5897 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5900 50. If a filter generated a file delivery with a non-absolute name (possible if
5901 no home directory exists for the router), the forbid_file option was not
5904 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5905 the existing equality matching.
5907 52. Exim was using ints instead of ino_t variables in some places where it was
5908 dealing with inode numbers.
5910 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5911 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5912 called TMPDIR, and if it finds it is different, it changes its value.
5914 54. The smtp_printf() function is now made available to local_scan() so
5915 additional output lines can be written before returning. There is also an
5916 smtp_fflush() function to enable the detection of a dropped connection.
5917 The variables smtp_input and smtp_batched_input are exported to
5920 55. Changed the default runtime configuration: the message "Unknown user"
5921 has been removed from the ACL, and instead placed on the localuser router,
5922 using the cannot_route_message feature. This means that any verification
5923 failures that generate their own messages won't get overridden. Similarly,
5924 the "Unrouteable address" message that was in the ACL for unverifiable
5925 relay addresses has also been removed.
5927 56. Added hosts_avoid_esmtp to the smtp transport.
5929 57. The exicyclog script was not checking for the esoteric option
5930 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5931 will work only if exicyclog is run under the appropriate euid.
5933 58. Following a discussion on the list, the rules by which Exim recognises line
5934 endings on incoming messages have been changed. The -dropcr and drop_cr
5935 options are now no-ops, retained only for backwards compatibility. The
5936 following line terminators are recognized: LF CRLF CR. However, special
5937 processing applies to CR:
5939 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5940 nor a local message in the state where . is a terminator.
5942 (ii) If a bare CR is encountered in a header line, an extra space is added
5943 after the line terminator so as not to end the header. The reasoning
5944 behind this is that bare CRs in header lines are most likely either
5945 to be mistakes, or people trying to play silly games.
5947 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5948 was being incorrectly given as 18 bytes larger than it should have been.
5949 This is a VOB (very old bug).
5951 60. This may never have affected anything current, but just in case it has:
5952 When the local host is found other than at the start of a list of hosts,
5953 the local host, those with the same MX, and any that follow, are discarded.
5954 When the list in question was part of a longer list of hosts, the following
5955 hosts (not currently being processed) were also being discarded. This no
5956 longer happens. I'm not sure if this situation could ever has previously
5959 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5962 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5963 header. This is something that is recommended in a new Internet Draft, and
5964 is something that is documented as being done by Sendmail. There are two
5965 possible values. For messages generated by the autoreply transport, Exim
5968 Auto-Submitted: auto-replied
5970 whereas for all other generated messages (e.g. bounces) it adds
5972 Auto-Submitted: auto-generated
5974 63. The "personal" condition in filters now includes a test for the
5975 Auto-Submitted: header. If it contains the string "auto-" the message it
5976 not considered personal.
5978 64. Added rcpt_include_affixes as a generic transport option.
5980 65. Added queue_only_override (default true).
5982 66. Added the syslog_duplication option.
5984 67. If what should have been the first header line of a message consisted of
5985 a space followed by a colon, Exim was mis-interpreting it as a header line.
5986 It isn't of course - it is syntactically invalid and should therefore be
5987 treated as the start of the message body. The misbehaviour could have
5988 caused a number of strange effects, including loss of data in subsequent
5989 header lines, and spool format errors.
5991 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5992 client host had authenticated. This control can now be exercised by an ACL
5993 for more flexibility.
5995 69. By default, callouts do not happen when testing with -bh. There is now a
5996 variant, -bhc, which does actually run the callout code, including
5997 consulting and updating the callout cache.
5999 70. Added support for saslauthd authentication, courtesy of Alexander
6002 71. If statvfs() failed on the spool or log directories while checking their
6003 size for availability, Exim confusingly gave the error "space shortage".
6004 Furthermore, in debugging mode it crashed with a floating point exception.
6005 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6006 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6007 serious problem, Exim now writes to the main and panic logs when this
6008 happens, with details of the failure. It then refuses to accept the
6009 incoming message, giving the message "spool directory problem" or "log
6010 directory problem" with a 421 code for SMTP messages.
6012 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6013 0, 1, and 2 exist, because some OS complain for execs without them (see
6014 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6015 descriptors. However, the code omitted to check that the open succeeded,
6016 causing mysterious errors if for some reason the permissions on /dev/null
6017 got screwed. Now Exim writes a message to the main and panic logs, and
6018 bombs out if it can't open /dev/null.
6020 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6021 interact so that it is all more flexible. It is supposed to remain
6022 backwards compatible. Also added extra_local_interfaces.
6024 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6025 to bomb out with an assertion failure - to the client this appears as a
6026 connection drop. This problem occurs in the part of the code that was taken
6027 from the Samba project. Fortunately, the assertion is in a very simple
6028 function, so I have fixed this by reproducing the function inline in the
6029 one place where it is called, and arranging for authentication to fail
6030 instead of killing the process with assert().
6032 75. The SPA client code was not working when the server requested OEM rather
6033 than Unicode encoding.
6035 76. Added code to make require_files with a specific uid setting more usable in
6036 the case where statting the file as root fails - usually a non-root-mounted
6037 NFS file system. When this happens and the failure is EACCES, Exim now
6038 forks a subprocess and does the per-uid checking as the relevant uid.
6040 77. Added process_log_path.
6042 78. If log_file_path was not explicitly set, a setting of check_log_space or
6043 check_log_inodes was ignored.
6045 79. If a space check for the spool or log partitions fails, the incident is now
6046 logged. Of course, in the latter case the data may get lost...
6048 80. Added the %p formatting code to string_format() so that it can be used to
6049 print addresses in debug_print(). Adjusted all the address printing in the
6050 debugging in store.c to use %p rather than %d.
6052 81. There was a concern that a line of code in smtp_in.c could overflow a
6053 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6054 initially expressed, the concern was not well-founded, because trailing
6055 spaces are removed early. However, if the trailing spaces were followed by
6056 a NULL, they did not get removed, so the overflow was possible. Two fixes
6059 (a) I re-wrote the offending code in a cleaner fashion.
6060 (b) If an incoming SMTP command contains a NULL character, it is rejected
6063 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6064 runs initgroups(), so that if the Exim user is in any additional groups,
6065 they will be used during message reception.
6071 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob