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.
153 TL/15 SECURITY: prevent double expansion in math comparison functions
154 (can expand unsanitized data). Not remotely exploitable.
161 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
163 PP/02 Make -n do something, by making it not do something.
164 When combined with -bP, the name of an option is not output.
166 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
169 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
170 $sender_host_name and config options to manage this, and basic check
173 PP/05 DSCP support for outbound connections and control modifier for inbound.
175 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
176 (Only plugin which currently uses this is kerberos4, which nobody should
177 be using, but we should make it available and other future plugins might
178 conceivably use it, even though it would break NAT; stuff *should* be
179 using channel bindings instead).
181 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
182 name; added for Sendmail compatibility; requires admin caller.
183 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
184 just ignore it); requires trusted caller.
185 Also parse but ignore: -Ac -Am -X<logfile>
188 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
190 TL/02 Add +smtp_confirmation as a default logging option.
192 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
193 Patch by Magnus Holmgren from 2007-02-20.
195 TL/04 Bugzilla 1281 - Spec typo.
196 Bugzilla 1283 - Spec typo.
197 Bugzilla 1290 - Spec grammar fixes.
199 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
201 TL/06 Add Experimental DMARC support using libopendmarc libraries.
203 TL/07 Fix an out of order global option causing a segfault. Reported to dev
204 mailing list by by Dmitry Isaikin.
206 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
208 JH/02 Support "G" suffix to numbers in ${if comparisons.
210 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
212 NM/01 Bugzilla 1197 - Spec typo
213 Bugzilla 1196 - Spec examples corrections
215 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
217 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
218 gnutls_enable_pkcs11, but renamed to more accurately indicate its
221 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
222 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
224 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
225 "acl {{name}{arg}...}", and optional args on acl condition
228 JH/05 Permit multiple router/transport headers_add/remove lines.
230 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
232 JH/07 Avoid using a waiting database for a single-message-only transport.
233 Performance patch from Paul Fisher. Bugzilla 1262.
235 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
238 JH/09 Add $headers_added variable, with content from use of ACL modifier
239 add_header (but not yet added to the message). Bugzilla 199.
241 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
242 Pulled from Bugzilla 817 by Wolfgang Breyha.
244 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
246 (nb: this is the same fix as in Exim 4.80.1)
248 JH/11 Add A= logging on delivery lines, and a client_set_id option on
251 JH/12 Add optional authenticated_sender logging to A= and a log_selector
254 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
256 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
257 advertise SMTP AUTH mechanism to us, instead of a generic
258 protocol violation error. Also, make Exim more robust to bad
259 data from the Dovecot auth socket.
261 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
263 When a queue runner is handling a message, Exim first routes the
264 recipient addresses, during which it prunes them based on the retry
265 hints database. After that it attempts to deliver the message to
266 any remaining recipients. It then updates the hints database using
269 So if a recipient address works intermittently, it can get repeatedly
270 deferred at routing time. The retry hints record remains fresh so the
271 address never reaches the final cutoff time.
273 This is a fairly common occurrence when a user is bumping up against
274 their storage quota. Exim had some logic in its local delivery code
275 to deal with this. However it did not apply to per-recipient defers
276 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
278 This change adds a proper retry rule check during routing so that the
279 final cutoff time is checked against the message's age. We only do
280 this check if there is an address retry record and there is not a
281 domain retry record; this implies that previous attempts to handle
282 the address had the retry_use_local_parts option turned on. We use
283 this as an approximation for the destination being like a local
284 delivery, as in LMTP.
286 I suspect this new check makes the old local delivery cutoff check
287 redundant, but I have not verified this so I left the code in place.
289 TF/02 Correct gecos expansion when From: is a prefix of the username.
291 Test 0254 submits a message to Exim with the header
295 When I ran the test suite under the user fanf2, Exim expanded
296 the header to contain my full name, whereas it should have added
297 a Resent-Sender: header. It erroneously treats any prefix of the
298 username as equal to the username.
300 This change corrects that bug.
302 GF/01 DCC debug and logging tidyup
303 Error conditions log to paniclog rather than rejectlog.
304 Debug lines prefixed by "DCC: " to remove any ambiguity.
306 TF/03 Avoid unnecessary rebuilds of lookup-related code.
308 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
309 Bug spotted by Jeremy Harris; was flawed since initial commit.
310 Would have resulted in OCSP responses post-SNI triggering an Exim
311 NULL dereference and crash.
313 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
315 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
316 Bug detection, analysis and fix by Samuel Thibault.
317 Bugzilla 1331, Debian bug #698092.
319 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
321 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
322 Server implementation by Todd Lyons, client by JH.
323 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
324 config variable "prdr_enable" controls whether the server
325 advertises the facility. If the client requests PRDR a new
326 acl_data_smtp_prdr ACL is called once for each recipient, after
327 the body content is received and before the acl_smtp_data ACL.
328 The client is controlled by bolth of: a hosts_try_prdr option
329 on the smtp transport, and the server advertisement.
330 Default client logging of deliveries and rejections involving
331 PRDR are flagged with the string "PRDR".
333 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
334 fclose(). Diagnosis by Todd Lyons.
336 PP/17 Update configure.default to handle IPv6 localhost better.
337 Patch by Alain Williams (plus minor tweaks).
340 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
341 This is now consistent with GnuTLS, and is now documented: the
342 previous undocumented portable approach to treating the option as
343 unset was to force an expansion failure. That still works, and
344 an empty string is now equivalent.
346 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
347 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
348 not performing validation itself.
350 PP/20 Added force_command boolean option to pipe transport.
351 Patch from Nick Koston, of cPanel Inc.
353 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
356 TF/04 Added udpsend ACL modifer and hexquote expansion operator
358 PP/21 Fix eximon continuous updating with timestamped log-files.
359 Broken in a format-string cleanup in 4.80, missed when I repaired the
360 other false fix of the same issue.
361 Report and fix from Heiko Schlichting.
364 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
365 Report from Prashanth Katuri.
367 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
368 It's SecureTransport, so affects any MacOS clients which use the
369 system-integrated TLS libraries, including email clients.
371 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
372 using a MIME ACL for non-SMTP local injection.
373 Report and assistance in diagnosis by Warren Baker.
375 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
377 JH/16 Fix comparisons for 64b. Bugzilla 1385.
379 TL/09 Add expansion variable $authenticated_fail_id to keep track of
380 last id that failed so it may be referenced in subsequent ACL's.
382 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
385 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
386 ldap library initialization, allowing self-signed CA's to be
387 used. Also properly sets require_cert option later in code by
388 using NULL (global ldap config) instead of ldap handle (per
389 session). Bug diagnosis and testing by alxgomz.
391 TL/12 Enhanced documentation in the ratelimit.pl script provided in
392 the src/util/ subdirectory.
394 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
395 renamed to Transport Post Delivery Action by Jeremy Harris, as
398 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
399 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
400 redis_servers = needs to be configured which will be used by the redis
401 lookup. Patch from Warren Baker, of The Packet Hub.
403 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
405 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
406 hostname or reverse DNS when processing a host list. Used suggestions
407 from multiple comments on this bug.
409 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
411 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
412 Missed a few lines, added it to make the runtest require no keyboard
415 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
416 contains upper case chars. Make router use caseful_local_part.
418 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
419 support when GnuTLS has been built with p11-kit.
425 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
427 This, or similar/improved, will also be change PP/11 of 4.82.
433 PP/01 Handle short writes when writing local log-files.
434 In practice, only affects FreeBSD (8 onwards).
435 Bugzilla 1053, with thanks to Dmitry Isaikin.
437 NM/01 Bugzilla 949 - Documentation tweak
439 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
442 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
444 PP/02 Implemented gsasl authenticator.
446 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
448 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
449 `pkg-config foo` for cflags/libs.
451 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
452 with rest of GSASL and with heimdal_gssapi.
454 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
455 `pkg-config foo` for cflags/libs for the TLS implementation.
457 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
458 properties get this fed in as external SSF. A number of robustness
459 and debugging improvements to the cyrus_sasl authenticator.
461 PP/08 cyrus_sasl server now expands the server_realm option.
463 PP/09 Bugzilla 1214 - Log authentication information in reject log.
464 Patch by Jeremy Harris.
466 PP/10 Added dbmjz lookup type.
468 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
470 PP/12 MAIL args handles TAB as well as SP, for better interop with
471 non-compliant senders.
472 Analysis and variant patch by Todd Lyons.
474 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
475 Bug report from Lars Müller <lars@samba.org> (via SUSE),
476 Patch from Dirk Mueller <dmueller@suse.com>
478 PP/13 tls_peerdn now print-escaped for spool files.
479 Observed some $tls_peerdn in wild which contained \n, which resulted
480 in spool file corruption.
482 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
483 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
484 or write after TLS renegotiation, which otherwise led to messages
487 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
488 as a tracking header (ie: a signed header comes before the signature).
489 Patch from Wolfgang Breyha.
491 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
492 comma-sep list; embedded commas doubled.
494 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
496 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
498 Report and patch from Dmitry Banschikov.
500 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
501 Removed SSL_clear() after SSL_new() which led to protocol negotiation
502 failures. We appear to now support TLS1.1+ with Exim.
504 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
505 lets Exim select keys and certificates based upon TLS SNI from client.
506 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
507 before an outbound SMTP session. New log_selector, +tls_sni.
509 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
510 NULL dereference. Report and patch from Alun Jones.
512 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
513 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
514 Not seeing resolver debug output on NetBSD, but suspect this is a
515 resolver implementation change.
517 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
518 Left warnings. Added "eximon gdb" invocation mode.
520 PP/21 Defaulting "accept_8bitmime" to true, not false.
522 PP/22 Added -bw for inetd wait mode support.
524 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
525 locate the relevant includes and libraries. Made this the default.
527 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
528 Bugzilla 1246, report and most of solution from Tomasz Kusy.
530 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
531 This may cause build issues on older platforms.
533 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
534 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
535 gnutls_require_mac & gnutls_require_protocols (no longer supported).
536 Added SNI support via GnuTLS too.
537 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
539 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
541 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
542 Applied second patch from Janne, implementing suggestion to default
543 multiple-strings-in-record handling to match SPF spec.
545 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
547 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
548 read-only, out of scope).
549 Patch from Wolfgang Breyha, report from Stuart Northfield.
551 PP/29 Fix three issues highlighted by clang analyser static analysis.
552 Only crash-plausible issue would require the Cambridge-specific
553 iplookup router and a misconfiguration.
554 Report from Marcin Mirosław.
556 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
558 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
559 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
560 As part of this, removing so much warning spew let me fix some minor
561 real issues in debug logging.
563 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
564 assignment on my part. Fixed.
566 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
567 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
568 Janne Snabb (who went above and beyond: thank you).
570 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
571 string otherwise requires a connection and a bunch more work and it's
572 relatively easy to get wrong. Should also expose TLS library linkage
575 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
576 64-bit ${eval} (JH/03).
578 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
579 GNU libc to support some of the 64-bit stuff, should not lead to
580 conflicts. Defined before os.h is pulled in, so if a given platform
581 needs to override this, it can.
583 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
584 protection layer was required, which is not implemented.
585 Bugzilla 1254, patch from Wolfgang Breyha.
587 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
588 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
589 tls_dhparam take prime identifiers. Also unbreak combination of
590 OpenSSL+DH_params+TLSSNI.
592 PP/39 Disable SSLv2 by default in OpenSSL support.
598 PP/01 Solaris build fix for Oracle's LDAP libraries.
599 Bugzilla 1109, patch from Stephen Usher.
601 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
603 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
606 TF/02 Fix a couple more cases where we did not log the error message
607 when unlink() failed. See also change 4.74-TF/03.
609 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
610 lock up or crash if it happened to be inside a call to libc when it
611 got a SIGUSR1 from exiwhat.
613 The SIGUSR1 handler appends the current process status to the process
614 log which is later printed by exiwhat. It used to use the general
615 purpose logging code to do this, but several functions it calls are
616 not safe for signals.
618 The new output code in the SIGUSR1 handler is specific to the process
619 log, and simple enough that it's easy to inspect for signal safety.
620 Removing some special cases also simplifies the general logging code.
621 Removing the spurious timestamps from the process log simplifies
624 TF/04 Improved ratelimit ACL condition.
626 The /noupdate option has been deprecated in favour of /readonly which
627 has clearer semantics. The /leaky, /strict, and /readonly update modes
628 are mutually exclusive. The update mode is no longer included in the
629 database key; it just determines when the database is updated. (This
630 means that when you upgrde Exim will forget old rate measurements.)
632 Exim now checks that the per_* options are used with an update mode that
633 makes sense for the current ACL. For example, when Exim is processing a
634 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
635 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
636 must specify per_mail/readonly. If you omit the update mode it defaults to
637 /leaky where that makes sense (as before) or /readonly where required.
639 The /noupdate option is now undocumented but still supported for
640 backwards compatibility. It is equivalent to /readonly except that in
641 ACLs where /readonly is required you may specify /leaky/noupdate or
642 /strict/noupdate which are treated the same as /readonly.
644 A useful new feature is the /count= option. This is a generalization
645 of the per_byte option, so that you can measure the throughput of other
646 aggregate values. For example, the per_byte option is now equivalent
647 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
649 The per_rcpt option has been generalized using the /count= mechanism
650 (though it's more complicated than the per_byte equivalence). When it is
651 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
652 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
653 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
654 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
655 Note that using per_rcpt with a non-readonly update mode in more than
656 one ACL will cause the recipients to be double-counted. (The per_mail
657 and per_byte options don't have this problem.)
659 The handling of very low rates has changed slightly. If the computed rate
660 is less than the event's count (usually one) then this event is the first
661 after a long gap. In this case the rate is set to the same as this event's
662 count, so that the first message of a spam run is counted properly.
664 The major new feature is a mechanism for counting the rate of unique
665 events. The new per_addr option counts the number of different
666 recipients that someone has sent messages to in the last time period. It
667 behaves like per_rcpt if all the recipient addresses are different, but
668 duplicate recipient addresses do not increase the measured rate. Like
669 the /count= option this is a general mechanism, so the per_addr option
670 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
671 example, measure the rate that a client uses different sender addresses
672 with the options per_mail/unique=$sender_address. There are further
673 details in the main documentation.
675 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
677 TF/06 Removed a few PCRE remnants.
679 TF/07 Automatically extract Exim's version number from tags in the git
680 repository when doing development or release builds.
682 PP/02 Raise smtp_cmd_buffer_size to 16kB.
683 Bugzilla 879. Patch from Paul Fisher.
685 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
686 Heavily based on revision 40f9a89a from Simon Arlott's tree.
689 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
691 PP/05 Variable $av_failed, true if the AV scanner deferred.
692 Bugzilla 1078. Patch from John Horne.
694 PP/06 Stop make process more reliably on build failure.
695 Bugzilla 1087. Patch from Heiko Schlittermann.
697 PP/07 Make maildir_use_size_file an _expandable_ boolean.
698 Bugzilla 1089. Patch from Heiko Schlittermann.
700 PP/08 Handle ${run} returning more data than OS pipe buffer size.
701 Bugzilla 1131. Patch from Holger Weiß.
703 PP/09 Handle IPv6 addresses with SPF.
704 Bugzilla 860. Patch from Wolfgang Breyha.
706 PP/10 GnuTLS: support TLS 1.2 & 1.1.
708 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
711 PP/11 match_* no longer expand right-hand-side by default.
712 New compile-time build option, EXPAND_LISTMATCH_RHS.
713 New expansion conditions, "inlist", "inlisti".
715 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
717 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
719 PP/14 fix log_write() format string regression from TF/03.
720 Bugzilla 1152. Patch from Dmitry Isaikin.
726 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
728 PP/02 Harmonised TLS library version reporting; only show if debugging.
729 Layout now matches that introduced for other libraries in 4.74 PP/03.
731 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
733 PP/04 New "dns_use_edns0" global option.
735 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
738 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
739 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
741 TK/01 Updated PolarSSL code to 0.14.2.
742 Bugzilla 1097. Patch from Andreas Metzler.
744 PP/07 Catch divide-by-zero in ${eval:...}.
747 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
750 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
751 format-string attack -- SECURITY: remote arbitrary code execution.
753 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
754 time unintentionally subject to list matching rules, letting the header
755 cause arbitrary Exim lookups (of items which can occur in lists, *not*
756 arbitrary string expansion). This allowed for information disclosure.
758 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
759 INT_MIN/-1 -- value coerced to INT_MAX.
765 NM/01 Workround for PCRE version dependency in version reporting
768 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
769 This fixes portability to compilers other than gcc, notably
770 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
772 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
773 makefiles for portability to HP-UX and POSIX correctness.
775 PP/01 Permit LOOKUP_foo enabling on the make command-line.
776 Also via indented variable definition in the Makefile.
777 (Debugging by Oliver Heesakkers).
779 PP/02 Restore caching of spamd results with expanded spamd_address.
780 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
782 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
783 Improves build reliability. Fix from: Frank Elsner
785 NM/02 Fix wide character breakage in the rfc2047 coding
786 Fixes bug 1064. Patch from Andrey N. Oktyabrski
788 NM/03 Allow underscore in dnslist lookups
789 Fixes bug 1026. Patch from Graeme Fowler
791 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
792 Code patches from Adam Ciarcinski of NetBSD.
794 NM/04 Fixed exiqgrep to cope with mailq missing size issue
797 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
798 is logged, to avoid truncation. Patch from John Horne.
800 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
801 Patch from Jakob Hirsch.
803 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
804 SQL string expansion failure details.
805 Patch from Andrey Oktyabrski.
807 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
808 Patch from Simon Arlott.
810 PP/09 New lookups functionality failed to compile on old gcc which rejects
811 extern declarations in function scope.
812 Patch from Oliver Fleischmann
814 PP/10 Use sig_atomic_t for flags set from signal handlers.
815 Check getgroups() return and improve debugging.
816 Fixed developed for diagnosis in bug 927 (which turned out to be
819 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
820 Patch from Mark Zealey.
822 PP/12 Bugzilla 1056: Improved spamd server selection.
823 Patch from Mark Zealey.
825 PP/13 Bugzilla 1086: Deal with maildir quota file races.
826 Based on patch from Heiko Schlittermann.
828 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
829 Patch from Uwe Doering, sign-off by Michael Haardt.
831 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
832 variable declaration deep within a block. Bug and patch from
835 PP/15 lookups-Makefile IRIX compatibilty coercion.
837 PP/16 Make DISABLE_DKIM build knob functional.
839 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
840 Patch by Simon Arlott
842 TF/03 Fix valgrind.h portability to C89 compilers that do not support
843 variable argument macros. Our copy now differs from upstream.
849 TF/01 Failure to get a lock on a hints database can have serious
850 consequences so log it to the panic log.
852 TF/02 Log LMTP confirmation messages in the same way as SMTP,
853 controlled using the smtp_confirmation log selector.
855 TF/03 Include the error message when we fail to unlink a spool file.
857 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
858 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
859 for maintaining out-of-tree patches for some time.
861 PP/01 Bugzilla 139: Documentation and portability issues.
862 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
863 Handle per-OS dynamic-module compilation flags.
865 PP/02 Let /dev/null have normal permissions.
866 The 4.73 fixes were a little too stringent and complained about the
867 permissions on /dev/null. Exempt it from some checks.
868 Reported by Andreas M. Kirchwitz.
870 PP/03 Report version information for many libraries, including
871 Exim version information for dynamically loaded libraries. Created
872 version.h, now support a version extension string for distributors
873 who patch heavily. Dynamic module ABI change.
875 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
876 privilege escalation vulnerability whereby the Exim run-time user
877 can cause root to append content of the attacker's choosing to
880 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
883 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
884 If dropping privileges for untrusted macros, we disabled normal logging
885 on the basis that it would fail; for the Exim run-time user, this is not
886 the case, and it resulted in successful deliveries going unlogged.
887 Fixed. Reported by Andreas Metzler.
893 PP/01 Date: & Message-Id: revert to normally being appended to a message,
894 only prepend for the Resent-* case. Fixes regression introduced in
895 Exim 4.70 by NM/22 for Bugzilla 607.
897 PP/02 Include check_rfc2047_length in configure.default because we're seeing
898 increasing numbers of administrators be bitten by this.
900 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
902 PP/03 Bugzilla 994: added openssl_options main configuration option.
904 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
906 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
908 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
910 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
911 without a peer certificate, leading to a segfault because of an
912 assumption that peers always have certificates. Be a little more
913 paranoid. Problem reported by Martin Tscholak.
915 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
916 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
917 NB: ClamAV planning to remove STREAM in "middle of 2010".
918 CL also introduces -bmalware, various -d+acl logging additions and
919 more caution in buffer sizes.
921 PP/09 Implemented reverse_ip expansion operator.
923 PP/10 Bugzilla 937: provide a "debug" ACL control.
925 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
927 PP/12 Bugzilla 973: Implement --version.
929 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
931 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
933 PP/15 Bugzilla 816: support multiple condition rules on Routers.
935 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
936 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
937 ignore trailing whitespace.
939 JJ/02 prevent non-panic DKIM error from being sent to paniclog
941 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
944 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
945 Notification from Dr Andrew Aitchison.
947 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
948 ExtendedDetectionInfo response format.
949 Notification from John Horne.
951 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
954 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
955 XSL and documented dependency on system catalogs, with examples of how
958 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
961 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
962 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
963 configuration file which is writeable by the Exim user or group.
965 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
966 of configuration files to cover files specified with the -C option if
967 they are going to be used with root privileges, not just the default
970 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
971 option (effectively making it always true).
973 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
974 files to be used while preserving root privileges.
976 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
977 that rogue child processes cannot use them.
979 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
980 run-time user, instead of root.
982 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
983 Exim run-time user without dropping privileges.
985 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
986 result string, instead of calling string_vformat() twice with the same
989 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
990 for other users. Others should always drop root privileges if they use
991 -C on the command line, even for a whitelisted configure file.
993 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
995 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1001 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1002 $data_path, and $header_path variables; fixed documentation bugs and
1005 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1006 exipick to access non-standard spools, including the "frozen" queue
1009 NM/01 Bugzilla 965: Support mysql stored procedures.
1010 Patch from Alain Williams
1012 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1014 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1015 Patch from Andreas Metzler
1017 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1018 Patch from Kirill Miazine
1020 NM/05 Bugzilla 671: Added umask to procmail example.
1022 JJ/03 installed exipick 20100323.0, fixing doc bug
1024 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1025 directory. Notification and patch from Dan Rosenberg.
1027 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1029 TK/02 Improve log output when DKIM signing operation fails.
1031 MH/01 Treat the transport option dkim_domain as a colon separated
1032 list, not as a single string, and sign the message with each element,
1033 omitting multiple occurences of the same signer.
1035 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1036 Bugzilla 985, 986. Patch by Simon Arlott
1038 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1039 Patch by Simon Arlott
1041 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1042 MBX locking. Notification from Dan Rosenberg.
1048 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1050 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1052 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1054 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1056 NM/04 Bugzilla 915: Flag broken perl installation during build.
1062 TK/01 Added patch by Johannes Berg that expands the main option
1063 "spamd_address" if it starts with a dollar sign.
1065 TK/02 Write list of recipients to X-Envelope-Sender header when building
1066 the mbox-format spool file for content scanning (suggested by Jakob
1069 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1070 (http://www.dcc-servers.net/) support via dccifd. Activated by
1071 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1073 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1074 by Mark Daniel Reidel <mr@df.eu>.
1076 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1077 When building exim an external PCRE library is now needed -
1078 PCRE is a system library on the majority of modern systems.
1079 See entry on PCRE_LIBS in EDITME file.
1081 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1082 conversation. Added nologin parameter to request.
1083 Patch contributed by Kirill Miazine.
1085 TF/01 Do not log submission mode rewrites if they do not change the address.
1087 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1089 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1090 log files in place. Contributed by Roberto Lima.
1092 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1094 TF/03 Bugzilla 615: When checking the local_parts router precondition
1095 after a local_part_suffix or local_part_prefix option, Exim now
1096 does not use the address's named list lookup cache, since this
1097 contains cached lookups for the whole local part.
1099 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1100 Robert Millan. Documentation is in experimental-spec.txt.
1102 TF/04 Bugzilla 668: Fix parallel build (make -j).
1104 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1106 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1107 Patch provided by Jan Srzednicki.
1109 TF/05 Leading white space used to be stripped from $spam_report which
1110 wrecked the formatting. Now it is preserved.
1112 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1113 that they are available at delivery time.
1115 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1117 TF/08 TLS error reporting now respects the incoming_interface and
1118 incoming_port log selectors.
1120 TF/09 Produce a more useful error message if an SMTP transport's hosts
1121 setting expands to an empty string.
1123 NM/06 Bugzilla 744: EXPN did not work under TLS.
1124 Patch provided by Phil Pennock.
1126 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1127 Patch provided by Richard Godbee.
1129 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1130 acl_smtp_notquit, added index entry.
1132 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1133 Patch provided by Eugene Bujak.
1135 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1136 accept(). Patch provided by Maxim Dounin.
1138 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1139 Patch provided by Phil Pennock.
1141 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1143 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1144 Patch provided by Brad "anomie" Jorsch.
1146 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1147 Patch provided by Dean Brooks.
1149 TK/05 Add native DKIM support (does not depend on external libraries).
1151 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1152 Patch provided by Graeme Fowler.
1154 NM/16 Bugzilla 851: Documentation example syntax fix.
1156 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1158 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1161 NM/19 Bugzilla 745: TLS version reporting.
1162 Patch provided by Phil Pennock.
1164 NM/20 Bugzilla 167: bool: condition support.
1165 Patch provided by Phil Pennock.
1167 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1168 clients. Patch provided by Phil Pennock.
1170 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1171 Patch provided by Brad "anomie" Jorsch.
1173 NM/23 Bugzilla 687: Fix misparses in eximstats.
1174 Patch provided by Heiko Schlittermann.
1176 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1177 Patch provided by Heiko Schlittermann.
1179 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1180 plus update to original patch.
1182 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1184 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1185 Patch provided by David Brownlee.
1187 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1189 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1191 NM/30 Bugzilla 888: TLS documentation bugfixes.
1193 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1195 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1196 Unlike the original bugzilla I have changed all shell scripts in src tree.
1198 NM/33 Bugzilla 898: Transport filter timeout fix.
1199 Patch by Todd Rinaldo.
1201 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1202 Patch by Serge Demonchaux.
1204 NM/35 Bugzilla 39: Base64 decode bug fixes.
1205 Patch by Jakob Hirsch.
1207 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1209 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1211 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1213 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1219 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1220 ALT-N's libdkim that I have put here:
1221 http://duncanthrax.net/exim-experimental/
1223 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1224 were called 'true' and it seems that C99 defines that as a reserved
1225 keyword to be used with 'bool' variable types. That means you could
1226 not include C99-style headers which use bools without triggering
1227 build errors in sieve.c.
1229 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1230 as mailq or other aliases. Changed the --help handling significantly
1231 to do whats expected. exim_usage() emits usage/help information.
1233 SC/01 Added the -bylocaldomain option to eximstats.
1235 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1237 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1239 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1245 PH/01 Another patch from the Sieve maintainer.
1247 PH/02 When an IPv6 address is converted to a string for single-key lookup
1248 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1249 dots are used instead of colons so that keys in lsearch files need not
1250 contain colons. This was done some time before quoting was made available
1251 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1252 (notated using the quote facility) so as to distinguish them from IPv4
1253 keys. This meant that lookups for IP addresses in host lists did not work
1254 for iplsearch lookups.
1256 This has been fixed by arranging for IPv6 addresses to be expressed with
1257 colons if the lookup type is iplsearch. This is not incompatible, because
1258 previously such lookups could never work.
1260 The situation is now rather anomolous, since one *can* have colons in
1261 ordinary lsearch keys. However, making the change in all cases is
1262 incompatible and would probably break a number of configurations.
1264 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1267 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1268 conversion specification without a maximum field width, thereby enabling
1269 a rogue spamd server to cause a buffer overflow. While nobody in their
1270 right mind would setup Exim to query an untrusted spamd server, an
1271 attacker that gains access to a server running spamd could potentially
1272 exploit this vulnerability to run arbitrary code as the Exim user.
1274 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1275 $primary_hostname instead of what libspf2 thinks the hosts name is.
1277 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1278 a directory entry by the name of the lookup key. Previously, if a
1279 symlink pointed to a non-existing file or a file in a directory that
1280 Exim lacked permissions to read, a lookup for a key matching that
1281 symlink would fail. Now it is enough that a matching directory entry
1282 exists, symlink or not. (Bugzilla 503.)
1284 PH/03 The body_linecount and body_zerocount variables are now exported in the
1287 PH/04 Added the $dnslist_matched variable.
1289 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1290 This means they are set thereafter only if the connection becomes
1293 PH/06 Added the client_condition to authenticators so that some can be skipped
1294 by clients under certain conditions.
1296 PH/07 The error message for a badly-placed control=no_multiline_responses left
1297 "_responses" off the end of the name.
1299 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1301 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1302 (without spaces) instead of just copying the configuration text.
1304 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1306 PH/11 Added $max_received_linelength.
1308 PH/12 Added +ignore_defer and +include_defer to host lists.
1310 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1311 way in which PCRE > 7.0 is built.
1313 PH/14 Implemented queue_only_load_latch.
1315 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1316 MAIL command. The effect was to mangle the value on 64-bit systems.
1318 PH/16 Another patch from the Sieve maintainer.
1320 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1322 PH/18 If a system quota error occurred while trying to create the file for
1323 a maildir delivery, the message "Mailbox is full" was not appended to the
1324 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1325 only to a quota excession during the actual writing of the file.
1327 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1328 characters?) which causes problems in log lines. The DN values are now
1329 passed through string_printing() before being added to log lines.
1331 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1332 and InterBase are left for another time.)
1334 PH/21 Added message_body_newlines option.
1336 PH/22 Guard against possible overflow in moan_check_errorcopy().
1338 PH/23 POSIX allows open() to be a macro; guard against that.
1340 PH/24 If the recipient of an error message contained an @ in the local part
1341 (suitably quoted, of course), incorrect values were put in $domain and
1342 $local_part during the evaluation of errors_copy.
1348 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1349 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1352 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1353 issue a MAIL command.
1355 PH/02 In an ACL statement such as
1357 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1359 if a client was not listed at all, or was listed with a value other than
1360 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1361 the condition was not true (as it should be), so access was not denied.
1362 The bug was that the ! inversion was incorrectly passed on to the second
1363 item. This has been fixed.
1365 PH/03 Added additional dnslists conditions == and =& which are different from
1366 = and & when the dns lookup returns more than one IP address.
1368 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1369 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1371 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1372 FSYNC, which compiles an option called disable_fsync that allows for
1373 bypassing fsync(). The documentation is heavily laced with warnings.
1375 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1377 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1378 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1379 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1380 including adding "make clean"; (3) Added -fPIC when compiling the test
1381 dynamically loaded module, to get rid of a warning.
1383 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1384 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1385 = 0s. The bug is otherwise harmless.
1387 PH/07 There was a bug in the dovecot authenticator such that the value of
1388 $auth1 could be overwritten, and so not correctly preserved, after a
1389 successful authentication. This usually meant that the value preserved by
1390 the server_setid option was incorrect.
1392 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1394 PH/09 Installed PCRE release 7.0.
1396 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1397 run for batched SMTP input. It is now run at the start of every message
1398 in the batch. While fixing this I discovered that the process information
1399 (output by running exiwhat) was not always getting set for -bs and -bS
1400 input. This is fixed, and it now also says "batched" for BSMTP.
1402 PH/11 Added control=no_pipelining.
1404 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1405 patch, slightly modified), and move the expansion of helo_data till after
1406 the connection is made in the smtp transport (so it can use these
1409 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1411 PH/14 Added log_selector = +pid.
1413 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1415 PH/16 Add ${if forany and ${if forall.
1417 PH/17 Added dsn_from option to vary the From: line in DSNs.
1419 PH/18 Flush SMTP output before performing a callout, unless control =
1420 no_callout_flush is set.
1422 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1423 was true (the default) a successful delivery failed to delete the retry
1424 item, thus causing premature timeout of the address. The bug is now
1427 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1429 PH/21 Long custom messages for fakedefer and fakereject are now split up
1430 into multiline reponses in the same way that messages for "deny" and
1431 other ACL rejections are.
1433 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1434 with slight modification.
1436 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1437 draft, changing the syntax and factoring some duplicate code".
1439 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1440 for deliveries of the second and subsequent messages over the same SMTP
1443 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1444 ${reduce, with only minor "tidies".
1446 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1448 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1449 expansion side effects.
1451 PH/27 When a message times out after an over-quota error from an Exim-imposed
1452 quota, the bounce message says "mailbox is full". This message was not
1453 being given when it was a system quota that was exceeded. It now should
1456 MH/03 Made $recipients available in local_scan(). local_scan() already has
1457 better access to the recipient list through recipients_list[], but
1458 $recipients can be useful in postmaster-provided expansion strings.
1460 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1461 in the case of a MAIL command with additional options following the
1462 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1463 were accidentally chopped off.
1465 PH/29 SMTP synchronization checks are implemented when a command is read -
1466 there is a check that no more input is waiting when there shouldn't be
1467 any. However, for some commands, a delay in an ACL can mean that it is
1468 some time before the response is written. In this time, more input might
1469 arrive, invalidly. So now there are extra checks after an ACL has run for
1470 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1471 pipelining has not been advertised.
1473 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1475 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1476 returned in the SMTP response when smtp_return_error_details was false.
1477 This has been fixed.
1479 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1480 instead of the C I/O that was originally supplied, because problems were
1481 reported on Solaris.
1483 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1484 Exim which did not show up earlier: it was assuming that a call to
1485 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1486 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1487 was a macro that became an assignment, so it seemed to work. This has
1488 changed to a proper function call with a void return, hence the compile
1489 error. Exim's code has been fixed.
1491 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1494 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1496 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1498 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1499 because it turns out to be too big or there is a timeout) there is no
1500 "Completed" line in the log. When some messages of this type were
1501 selected by exigrep, they were listed as "not completed". Others were
1502 picked up by some special patterns. I have improved the selection
1503 criteria to be more general.
1505 PH/38 The host_find_failed option in the manualroute router can now be set
1506 to "ignore", to completely ignore a host whose IP address cannot be
1507 found. If all hosts are ignored, the behaviour is controlled by the new
1508 host_all_ignored option.
1510 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1511 homing or because of multiple MX records with /mx) generated more than
1512 one IP address, and the following item turned out to be the local host,
1513 all the secondary addresses of the first item were incorrectly removed
1514 from the list, along with the local host and any following hosts (which
1515 is what is supposed to happen).
1517 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1518 whoever called Exim into the -H file. In the case of the daemon it was
1519 behaving confusingly. When first started, it used values for whoever
1520 started the daemon, but after a SIGHUP it used the Exim user (because it
1521 calls itself on a restart). I have changed the code so that it now always
1524 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1525 message are rejected with the same error (e.g. no authentication or bad
1526 sender address), and a DATA command is nevertheless sent (as can happen
1527 with PIPELINING or a stupid MUA), the error message that was given to the
1528 RCPT commands is included in the rejection of the DATA command. This is
1529 intended to be helpful for MUAs that show only the final error to their
1532 PH/42 Another patch from the Sieve maintainer.
1534 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1535 Eximstats - Fixed some broken HTML links and added missing column headers
1537 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1538 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1540 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1542 PH/43 Yet another patch from the Sieve maintainer.
1544 PH/44 I found a way to check for a TCP/IP connection going away before sending
1545 the response to the final '.' that terminates a message, but only in the
1546 case where the client has not sent further data following the '.'
1547 (unfortunately, this is allowed). However, in many cases there won't be
1548 any further data because there won't be any more messages to send. A call
1549 to select() can be used: if it shows that the input is "ready", there is
1550 either input waiting, or the socket has been closed. An attempt to read
1551 the next input character can distinguish the two cases. Previously, Exim
1552 would have sent an OK response which the client would never have see.
1553 This could lead to message repetition. This fix should cure that, at
1554 least in a lot of common cases.
1556 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1557 advertised in response to EHLO.
1563 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1564 fixed by 4.65/MH/01 (is this a record?) are fixed:
1566 (i) An empty string was always treated as zero by the numeric comparison
1567 operators. This behaviour has been restored.
1569 (ii) It is documented that the numeric comparison operators always treat
1570 their arguments as decimal numbers. This was broken in that numbers
1571 starting with 0 were being interpreted as octal.
1573 While fixing these problems I realized that there was another issue that
1574 hadn't been noticed. Values of message_size_limit (both the global option
1575 and the transport option) were treated as octal if they started with 0.
1576 The documentation was vague. These values are now always treated as
1577 decimal, and I will make that clear in the documentation.
1583 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1584 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1587 MH/01 Don't check that the operands of numeric comparison operators are
1588 integers when their expansion is in "skipping" mode (fixes bug
1589 introduced by 4.64-PH/07).
1591 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1592 child addresses, Exim now panics and dies. Previously, because the count
1593 is held in a short int, deliveries were likely to be lost. As such a
1594 large number of recipients for a single message is ridiculous
1595 (performance will be very, very poor), I have chosen to impose a limit
1596 rather than extend the field.
1602 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1603 leftover -K file (the existence of which was triggered by #402).
1604 While we were at it, introduced process PID as part of the -K
1605 filename. This should rule out race conditions when creating
1608 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1609 processing considerably. Previous code took too long for large mails,
1610 triggering a timeout which in turn triggers #401.
1612 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1613 in the DK code in transports.c. sendfile() is not really portable,
1614 hence the _LINUX specificness.
1616 TF/01 In the add_headers option to the mail command in an Exim filter,
1617 there was a bug that Exim would claim a syntax error in any
1618 header after the first one which had an odd number of characters
1621 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1622 callout verification, Exim cached a "reject" for the entire domain. This
1623 is correct for most verifications, but it is not correct for a recipient
1624 verification with use_sender or use_postmaster set, because in that case
1625 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1626 case of MAIL FROM:<> rejection from other early rejections (e.g.
1627 rejection of HELO). When verifying a recipient using a non-null MAIL
1628 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1629 Whatever the result of the callout, the value of the domain cache is
1630 left unchanged (for any other kind of callout, getting as far as trying
1631 RCPT means that the domain itself is ok).
1633 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1636 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1637 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1638 session, the daemon ignores ECONNECT errors and logs others; it now
1639 ignores EPIPE as well.
1641 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1642 (quoted-printable decoding).
1644 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1645 later the small subsequent patch to fix an introduced bug.
1647 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1649 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1651 PH/08 An error is now given if message_size_limit is specified negative.
1653 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1654 to be given (somewhat) arbitrary names.
1656 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1659 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1660 miscellaneous code fixes
1662 PH/10 Added the log_reject_target ACL modifier to specify where to log
1665 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1666 hostname. This is wrong, because it relates to the incoming message (and
1667 probably the interface on which it is arriving) and not to the outgoing
1668 callout (which could be using a different interface). This has been
1669 changed to use the value of the helo_data option from the smtp transport
1670 instead - this is what is used when a message is actually being sent. If
1671 there is no remote transport (possible with a router that sets up host
1672 addresses), $smtp_active_hostname is used.
1674 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1675 tweaks were necessary in order to get it to work (see also 21 below):
1676 (a) The code assumed that strncpy() returns a negative number on buffer
1677 overflow, which isn't the case. Replaced with Exim's string_format()
1679 (b) There were several signed/unsigned issues. I just did the minimum
1680 hacking in of casts. There is scope for a larger refactoring.
1681 (c) The code used strcasecmp() which is not a standard C function.
1682 Replaced with Exim's strcmpic() function.
1683 (d) The code set only $1; it now sets $auth1 as well.
1684 (e) A simple test gave the error "authentication client didn't specify
1685 service in request". It would seem that Dovecot has changed its
1686 interface. Fortunately there's a specification; I followed it and
1687 changed what the client sends and it appears to be working now.
1689 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1692 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1693 address A is aliased to B and C, where B exists and C does not. Without
1694 -v the output is "A verified" because verification stops after a
1695 successful redirection if more than one address is generated. However,
1696 with -v the child addresses are also verified. Exim was outputting "A
1697 failed to verify" and then showing the successful verification for C,
1698 with its parentage. It now outputs "B failed to verify", showing B's
1699 parentage before showing the successful verification of C.
1701 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1702 look up a TXT record in a specific list after matching in a combined
1705 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1706 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1707 they consult the DNS. I had assumed they would set it the way they
1708 wanted; and indeed my experiments on Linux seem to show that in some
1709 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1710 To be on the safe side, however, I have now made the interface to
1711 host_find_byname() similar to host_find_bydns(), with an argument
1712 containing the DNS resolver options. The host_find_byname() function now
1713 sets these options at its start, just as host_find_bydns() does. The smtp
1714 transport options dns_qualify_single and dns_search_parents are passed to
1715 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1716 of host_find_byname() use the default settings of RES_DEFNAMES
1717 (qualify_single) but not RES_DNSRCH (search_parents).
1719 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1720 spool_read_header() do less string testing, by means of a preliminary
1721 switch on the second character of optional "-foo" lines. (This is
1722 overdue, caused by the large number of possibilities that now exist.
1723 Originally there were few.) While I was there, I also converted the
1724 str(n)cmp tests so they don't re-test the leading "-" and the first
1725 character, in the hope this might squeeze out yet more improvement.
1727 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1728 flag allowing group syntax was set by the header_syntax check but not
1729 turned off, possible causing trouble later; (2) The flag was not being
1730 set at all for the header_verify test, causing "group"-style headers to
1731 be rejected. I have now set it in this case, and also caused header_
1732 verify to ignore an empty address taken from a group. While doing this, I
1733 came across some other cases where the code for allowing group syntax
1734 while scanning a header line wasn't quite right (mostly, not resetting
1735 the flag correctly in the right place). These bugs could have caused
1736 trouble for malformed header lines. I hope it is now all correct.
1738 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1739 with the "reply" argument non-NULL. The code, however (which originally
1740 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1741 but it didn't always do it. This confused somebody who was copying the
1742 code for some other use. I have removed all the tests.
1744 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1745 feature that was used to support insecure browsers during the U.S. crypto
1746 embargo. It requires special client support, and Exim is probably the
1747 only MTA that supported it -- and would never use it because real RSA is
1748 always available. This code has been removed, because it had the bad
1749 effect of slowing Exim down by computing (never used) parameters for the
1750 RSA_EXPORT functionality.
1752 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1753 authenticator to fail if there's a tab character in the incoming data
1754 (there should never be unless someone is messing about, as it's supposed
1755 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1756 if the connection is using TLS or if the remote IP is the same as the
1757 local IP, and the "valid-client-cert option" if a client certificate has
1760 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1761 authenticators. This can be used for authorization after authentication
1762 succeeds. (In the case of plaintext, it servers for both authentication
1765 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1766 if any retry times were supplied.
1768 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1769 connection, where there is no client IP address to check. In this
1770 situation, the verify now always succeeds.
1772 PH/25 Applied John Jetmore's -Mset patch.
1774 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1776 PH/27 In a string expansion for a processed (not raw) header when multiple
1777 headers of the same name were present, leading whitespace was being
1778 removed from all of them, but trailing whitespace was being removed only
1779 from the last one. Now trailing whitespace is removed from each header
1780 before concatenation. Completely empty headers in a concatenation (as
1781 before) are ignored.
1783 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1784 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1786 PH/29 [Removed. This was a change that I later backed out, and forgot to
1787 correct the ChangeLog entry (that I had efficiently created) before
1788 committing the later change.]
1790 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1791 address errors (4xx response to RCPT) over the same connection as other
1792 messages routed to the same hosts. Such deliveries are always "forced",
1793 so retry times are not inspected. This resulted in far too many retries
1794 for the affected addresses. The effect occurred only when there were more
1795 hosts than the hosts_max_try setting in the smtp transport when it had
1796 the 4xx errors. Those hosts that it had tried were not added to the list
1797 of hosts for which the message was waiting, so if all were tried, there
1798 was no problem. Two fixes have been applied:
1800 (i) If there are any address or message errors in an SMTP delivery, none
1801 of the hosts (tried or untried) are now added to the list of hosts
1802 for which the message is waiting, so the message should not be a
1803 candidate for sending over the same connection that was used for a
1804 successful delivery of some other message. This seems entirely
1805 reasonable: after all the message is NOT "waiting for some host".
1806 This is so "obvious" that I'm not sure why it wasn't done
1807 previously. Hope I haven't missed anything, but it can't do any
1808 harm, as the worst effect is to miss an optimization.
1810 (ii) If, despite (i), such a delivery is accidentally attempted, the
1811 routing retry time is respected, so at least it doesn't keep
1812 hammering the server.
1814 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1815 in ${readsocket because some servers need this prod.
1817 PH/32 Added some extra debug output when updating a wait-xxx database.
1819 PH/33 The hint "could be header name not terminated by colon", which has been
1820 given for certain expansion errors for a long time, was not being given
1821 for the ${if def:h_colon_omitted{... case.
1823 PH/34 The spec says: "With one important exception, whenever a domain list is
1824 being scanned, $domain contains the subject domain." There was at least
1825 one case where this was not true.
1827 PH/35 The error "getsockname() failed: connection reset by peer" was being
1828 written to the panic log as well as the main log, but it isn't really
1829 panic-worthy as it just means the connection died rather early on. I have
1830 removed the panic log writing for the ECONNRESET error when getsockname()
1833 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1834 runs only) independently of the message's sender address. This meant
1835 that, if the 4xx error was in fact related to the sender, a different
1836 message to the same recipient with a different sender could confuse
1837 things. In particualar, this can happen when sending to a greylisting
1838 server, but other circumstances could also provoke similar problems.
1839 I have changed the default so that the retry time for these errors is now
1840 based a combination of the sender and recipient addresses. This change
1841 can be overridden by setting address_retry_include_sender=false in the
1844 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1845 remote server are returned as part of bounce messages. This was not
1846 happening for LMTP over a pipe (the lmtp transport), but now it is the
1847 same for both kinds of LMTP.
1849 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1850 in header lines that were in fact CNAMEs. This is no longer the case.
1852 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1853 and queue runs started by the daemon processed all messages. This has
1854 been fixed so that -R and -S can now usefully be given with -q<time>.
1856 PH/40 Import PCRE release 6.7 (fixes some bugs).
1858 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1860 PH/42 Give an error if -q is specified more than once.
1862 PH/43 Renamed the variables $interface_address and $interface_port as
1863 $received_ip_address and $received_port, to make it clear that these
1864 values apply to message reception, and not to the outgoing interface when
1865 a message is delivered. (The old names remain recognized, of course.)
1867 PH/44 There was no timeout on the connect() call when using a Unix domain
1868 socket in the ${readsocket expansion. There now is.
1870 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1871 be meaningful with "accept".
1873 SC/01 Eximstats V1.43
1874 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1876 SC/02 Eximstats V1.44
1877 Use a glob alias rather than an array ref in the generated
1878 parser. This improves both readability and performance.
1880 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1881 Collect SpamAssassin and rejection statistics.
1882 Don't display local sender or destination tables unless
1883 there is data to show.
1884 Added average volumes into the top table text output.
1886 SC/04 Eximstats V1.46
1887 Collect data on the number of addresses (recipients)
1888 as well as the number of messages.
1890 SC/05 Eximstats V1.47
1891 Added 'Message too big' to the list of mail rejection
1892 reasons (thanks to Marco Gaiarin).
1894 SC/06 Eximstats V1.48
1895 Mainlog lines which have GMT offsets and are too short to
1896 have a flag are now skipped.
1898 SC/07 Eximstats V1.49 (Alain Williams)
1899 Added the -emptyok flag.
1901 SC/08 Eximstats V1.50
1902 Fixes for obtaining the IP address from reject messages.
1904 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1905 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1906 whitesspace changes from 4.64-PH/27
1908 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1911 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1912 are found, allow negative numbers in numeric criteria)
1914 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1916 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1917 to match changes made in 4.64-PH/43
1919 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1921 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1922 "domain" is actually an IP address, and force a failure. This locks out
1923 those revolvers/nameservers that support "A-for-A" lookups, in
1924 contravention of the specifications.
1926 PH/48 When a host name was looked up from an IP address, and the subsequent
1927 forward lookup of the name timed out, the host name was left in
1928 $sender_host_name, contrary to the specification.
1930 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1931 restricted to single-key lookups, Exim was not diagnosing an error if
1932 * or *@ was used with a query-style lookup.
1934 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1936 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1937 long ago, but noone interested enough thought of it. Let's just say that
1938 the "1.1" means that there are some new functions that weren't there at
1939 some point in the past.
1941 PH/51 Error processing for expansion failure of helo_data from an smtp
1942 transport during callout processing was broken.
1944 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1945 tested/used via the -bh/-bhc/-bs options.
1947 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1948 bug, fixed in subsequent PCRE releases).
1950 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1951 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1953 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1959 SC/01 Use a glob alias rather than an array ref in eximstats generated
1960 parser. This improves both readability and performance.
1962 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1963 Don't display local sender or destination tables in eximstats unless
1964 there is data to show.
1965 Added average volumes into the eximstats top table text output.
1967 SC/03 Collect data on the number of addresses (recipients) as well
1968 as the number of messages in eximstats.
1970 TF/01 Correct an error in the documentation for the redirect router. Exim
1971 does (usually) call initgroups() when daemonizing.
1973 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1974 with consistent privilege compared to when running as a daemon.
1976 TF/03 Note in the spec that $authenticated_id is not set for local
1977 submissions from trusted users.
1979 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1980 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1982 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1983 by adding some example configuration directives to the default
1984 configuration file. A little bit of work is required to uncomment the
1985 directives and define how usernames and passwords are checked, but
1986 there is now a framework to start from.
1988 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1989 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1990 without this. I don't know how relevant this is to other LDAP libraries.
1992 PH/02 Add the verb name to the "unknown ACL verb" error.
1994 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1996 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1998 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1999 directory not expanded when it should be if an expanded home directory
2000 was set for the address (which is overridden by the transport).
2002 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2005 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2006 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2007 because it is too late at that time, and has no effect.
2009 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2010 security issue with \' (bugzilla #107). I could not use the
2011 PQescapeStringConn() function, because it needs a PGconn value as one of
2014 PH/08 When testing addresses using -bt, indicate those final addresses that
2015 are duplicates that would not cause an additional delivery. At least one
2016 person was confused, thinking that -bt output corresponded to deliveries.
2017 (Suppressing duplicates isn't a good idea as you lose the information
2018 about possibly different redirections that led to the duplicates.)
2020 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2021 systems where poll() doesn't work, in particular OS X.
2023 PH/10 Added more information to debugging output for retry time not reached.
2025 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2026 operations in malware.c.
2028 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2031 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2032 syslog_duplication set false, log lines that would normally be written
2033 both the the main log and to the reject log were not written to syslog at
2036 PH/14 In the default configuration, change the use of "message" in ACL warn
2037 statements to "add_header".
2039 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2040 not followed by a command (e.g. "seen endif").
2042 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2043 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2046 PH/17 Added extra conditions to the default value of delay_warning_condition
2050 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2051 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2052 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2055 The Auto-Submitted: and various List- headers are standardised, whereas I
2056 don't think Precedence: ever was.
2058 PH/18 Refactored debugging code in route_finduser() to show more information,
2059 in particular, the error code if getpwnam() issues one.
2061 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2062 This is apparently needed in addition to the PH/07 change above to avoid
2063 any possible encoding problems.
2065 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2066 but not after initializing Perl.
2068 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2069 output them only if debugging. By default they are written stderr,
2070 apparently, which is not desirable.
2072 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2075 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2078 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2080 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2081 authenticated or an ident call has been made. Suppress the default
2082 values for $authenticated_id and $authenticated_sender (but permit -oMai
2083 and -oMas) when testing with -bh.
2085 PH/24 Re-jigged the order of the tests in the default configuration so that the
2086 tests for valid domains and recipients precede the DNS black list and CSA
2087 tests, on the grounds that those ones are more expensive.
2089 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2090 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2091 This bug exists in every version of Exim that I still have, right back to
2094 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2095 However, an attempt to turn on case-sensitivity in a regex key by
2096 including (?-i) didn't work because the subject string was already
2097 lowercased, and the effects were non-intuitive. It turns out that a
2098 one-line patch can be used to allow (?-i) to work as expected.
2104 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2105 other effects) broke the use of negated acl sub-conditions.
2107 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2110 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2111 "Deny" causes Exim to reject the incoming connection with a 554 error.
2112 Unfortunately, if there is a major crisis, such as a disk failure,
2113 tcp-wrappers gives "deny", whereas what one would like would be some
2114 kind of temporary error. A kludge has been added to help with this.
2115 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2116 554 error is used if errno is still zero or contains ENOENT (which occurs
2117 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2120 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2122 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2123 errors. Otherwise a message that provokes a temporary error (when other
2124 messages do not) can cause a whole host to time out.
2126 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2127 addresses were routed directly to files or pipes from a redirect router.
2128 File deliveries just didn't batch; pipe deliveries might have suffered
2131 PH/06 A failure to get a lock for a hints database would erroneously always say
2132 "Failed to get write lock", even when it was really a read lock.
2134 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2135 of 0600. This has been changed to use the value of the lockfile_mode
2136 option (which defaults to 0600).
2138 PH/08 Applied small patch from the Sieve maintainer.
2140 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2141 folder from quota calculations, a direct delivery into this folder messed
2142 up the contents of the maildirsize file. This was because the regex was
2143 used only to exclude .Trash (or whatever) when the size of the mailbox
2144 was calculated. There was no check that a delivery was happening into an
2145 excluded directory. This bug has been fixed by ignoring all quota
2146 processing for deliveries into excluded directories.
2148 PH/10 Added the maildirfolder_create_regex option to appendfile.
2154 PH/01 The code for finding all the local interface addresses on a FreeBSD
2155 system running IPv6 was broken. This may well have applied to all BSD
2156 systems, as well as to others that have similar system calls. The broken
2157 code found IPv4 interfaces correctly, but gave incorrect values for the
2158 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2159 that it would not match correctly against @[] and not recognize the IPv6
2162 PH/02 The ipliteral router was not recognizing addresses of the form user@
2163 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2165 PH/03 Added disable_ipv6.
2167 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2168 decoded form, because it is most often used to construct To: headers
2169 lines in autoreplies, and the decoded form may well be syntactically
2170 invalid. However, $reply_address has leading white space removed, and all
2171 newlines turned into spaces so that the autoreply transport does not
2174 PH/05 If group was specified without a user on a router, and no group or user
2175 was specified on a transport, the group from the router was ignored.
2177 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2178 for visible compile-time settings that can be used to change these
2179 numbers, for those that want even more. Backwards compatibility with old
2180 spool files has been maintained. However, going back to a previous Exim
2181 release will lost any variables that are in spool files.
2183 PH/07 Two small changes when running in the test harness: increase delay when
2184 passing a TCP/IP connection to a new process, in case the original
2185 process has to generate a bounce, and remove special handling of
2186 127.0.0.2 (sic), which is no longer necessary.
2188 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2189 be the same on different OS.
2191 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2194 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2195 whether --show-vars was specified or not
2197 JJ/02 exipick: Added support for new ACL variable spool format introduced
2200 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2201 syntactically invalid From: or Reply-to: line, and a filter used this to
2202 generate an autoreply, and therefore failed to obtain an address for the
2203 autoreply, Exim could try to deliver to a non-existent relative file
2204 name, causing unrelated and misleading errors. What now happens is that
2205 it logs this as a hard delivery error, but does not attempt to create a
2208 PH/11 The exinext utility has a -C option for testing purposes, but although
2209 the given file was scanned by exinext itself; it wasn't being passed on
2210 when Exim was called.
2212 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2213 an end-of-file indication when reading a command response.
2215 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2216 compiled. In many other places in Exim, IPv6 addresses are always
2217 recognized, so I have changed this. It also means that IPv4 domain
2218 literals of the form [IPV4:n.n.n.n] are now always recognized.
2220 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2221 used if the router is not running as root, for example, when verifying at
2222 ACL time, or when using -bh. The debugging output from this situation was
2223 non-existent - all you got was a failure to exec. I have made two
2226 (a) Failures to set uid/gid, the current directory, or a process leader
2227 in a subprocess such as that created by queryprogram now generate
2228 suitable debugging ouput when -d is set.
2230 (b) The queryprogram router detects when it is not running as root,
2231 outputs suitable debugging information if -d is set, and then runs
2232 the subprocess without attempting to change uid/gid.
2234 PH/15 Minor change to Makefile for building test_host (undocumented testing
2237 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2238 additional section of a DNS packet that returns MX or SRV records.
2239 Instead, it always explicitly searches for A/AAAA records. This avoids
2240 major problems that occur when a DNS server includes only records of one
2241 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2242 fixed another bug: if SRV records were looked up and the corresponding
2243 address records were *not* found in the additional section, the port
2244 values from the SRV records were lost.
2246 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2247 using the correct key (the original address) when searching the retry
2248 rules in order to find which one to use for generating the retry hint.
2250 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2251 adding the default one. Similarly, if it contains a Reply-To: header, the
2252 errors_reply_to option, if set, is not used.
2254 PH/19 When calculating a retry time, Exim used to measure the "time since
2255 failure" by looking at the "first failed" field in the retry record. Now
2256 it does not use this if it is later than than the arrival time of the
2257 message. Instead it uses the arrival time. This makes for better
2258 behaviour in cases where some deliveries succeed, thus re-setting the
2259 "first failed" field. An example is a quota failure for a huge message
2260 when small messages continue to be delivered. Without this change, the
2261 "time since failure" will always be short, possible causing more frequent
2262 delivery attempts for the huge message than are intended.
2263 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2265 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2266 $1, $2, $3) because the numerical variables can be reset during some
2267 expansion items (e.g. "match"), thereby losing the authentication data.
2269 PH/21 Make -bV show the size of off_t variables so that the test suite can
2270 decide whether to run tests for quotas > 2G.
2272 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2273 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2274 filecount value is greater than 2G or if a quota value is greater than 2G
2275 on a system where the size of off_t is not greater than 4, a panic error
2278 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2279 never match. The debug and -bh output now contains an explicit error
2280 message indicating a malformed IPv4 address or mask.
2282 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2283 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2284 PH/23 above applies.
2286 PH/25 Do not write to syslog when running in the test harness. The only
2287 occasion when this arises is a failure to open the main or panic logs
2288 (for which there is an explicit test).
2290 PH/26 Added the /no_tell option to "control=freeze".
2292 PH/27 If a host name lookup failed very early in a connection, for example, if
2293 the IP address matched host_lookup and the reverse lookup yielded a name
2294 that did not have a forward lookup, an error message of the form "no IP
2295 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2296 could be logged. Now it outputs the IP address instead of "NULL".
2298 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2299 allows the sender and the authenticated sender to be set when
2300 submitting a message from within Exim. Since child_open_exim() is
2301 documented for local_scan(), the new function should be too.
2303 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2304 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2305 results in an empty string is now treated as unset.
2307 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2309 PH/31 Added sender_verify_fail logging option.
2311 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2312 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2313 tidied the source and removed it altogether.
2315 PH/33 When a queue run was abandoned because the load average was too high, a
2316 log line was always written; now it is written only if the queue_run log
2317 selector is set. In addition, the log line for abandonment now contains
2318 information about the queue run such as the pid. This is always present
2319 in "start" and "stop" lines but was omitted from the "abandon" line.
2321 PH/34 Omit spaces between a header name and the colon in the error message that
2322 is given when verify = headers_syntax fails (if there are lots of them,
2323 the message gets confusing).
2325 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2326 names, as there are now some PTR records that contain slashes. This check
2327 is only to protect against broken name servers that fall over on strange
2328 characters, so the fact that it applies to all lookups doesn't matter.
2330 PH/36 Now that the new test suite is complete, we can remove some of the
2331 special code in Exim that was needed for the old test suite. For example,
2332 sorting DNS records because real resolvers return them in an arbitrary
2333 order. The new test suite's fake resolver always returns records in the
2336 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2337 bounces) except when queue_only is set, to avoid logging races between
2338 the different processes.
2340 PH/38 Panic-die if .include specifies a non-absolute path.
2342 PH/39 A tweak to the "H" retry rule from its user.
2344 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2345 a label. They prevented compilation on older perls.
2347 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2348 a warning to be raised on newish perls.
2350 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2351 on queue. Changes to match documented behaviour of showing count of
2352 messages matching specified criteria.
2354 PH/40 Changed the default ident timeout from 30s to 5s.
2356 PH/41 Added support for the use of login_cap features, on those BSD systems
2357 that have them, for controlling the resources used by pipe deliveries.
2359 PH/42 The content-scanning code uses fopen() to create files in which to put
2360 message data. Previously it was not paying any attention to the mode of
2361 the files. Exim runs with umask(0) because the rest of the code creates
2362 files with open(), and sets the required mode explicitly. Thus, these
2363 files were ending up world-writeable. This was not a big issue, because,
2364 being within the spool directory, they were not world-accessible. I have
2365 created a function called modefopen, which takes an additional mode
2366 argument. It sets umask(777), creates the file, chmods it to the required
2367 mode, then resets the umask. All the relevant calls to fopen() in the
2368 content scanning code have been changed to use this function.
2370 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2371 to 24 hours. This avoids potential overflow problems when processing G
2372 and H retry rules. I suspect nobody ever tinkers with this value.
2374 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2376 PH/45 When the plaintext authenticator is running as a client, the server's
2377 challenges are checked to ensure they are valid base64 strings. By
2378 default, the authentication attempt is cancelled if an invalid string is
2379 received. Setting client_ignore_invalid_base64 true ignores these errors.
2380 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2381 they are received. Thus, the responses can be made to depend on the
2382 challenges. If an invalid string is ignored, an empty string is placed in
2385 PH/46 Messages that are created by the autoreply transport now contains a
2386 References: header, in accordance with RFCs 2822 and 3834.
2388 PH/47 Added authenticated_sender_force to the smtp transport.
2390 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2392 PH/49 Installed latest patch from the Sieve maintainer.
2394 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2395 was also set, the appendfile transport was unnecessarily scanning a
2396 directory of message files (e.g. for maildir delivery) to find the count
2397 of files (along with the size), even though it did not need this
2398 information. It now does the scan only if it needs to find either the
2399 size of the count of files.
2401 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2403 PH/52 Two bugs concerned with error handling when the smtp transport is
2406 (i) Exim was not creating retry information for temporary errors given
2407 for individual recipients after the DATA command when the smtp transport
2408 was used in LMTP mode. This meant that they could be retried too
2409 frequently, and not timed out correctly.
2411 (ii) Exim was setting the flag that allows error details to be returned
2412 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2413 recipients that were returned after the DATA command.
2415 PH/53 This is related to PH/52, but is more general: for any failing address,
2416 when detailed error information was permitted to be returned to the
2417 sender, but the error was temporary, then after the final timeout, only
2418 "retry timeout exceeded" was returned. Now it returns the full error as
2419 well as "retry timeout exceeded".
2421 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2422 do this, and (what is worse) MTAs that accept it.
2424 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2425 will now be deprecated.
2427 PH/56 New os.c-cygwin from the Cygwin maintainer.
2429 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2430 formats (previously only available in exim formats via -bpr, -bpru,
2431 and -bpra. Now also available in native and exiqgrep formats)
2433 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2434 with very large, slow to parse queues
2436 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2438 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2440 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2441 responses to authentication challenges, though it was showing the
2442 challenges; (ii) I've removed the CR characters from the debug output for
2445 PH/58 Allow for the insertion of a newline as well as a space when a string
2446 is turned into more than one encoded-word during RFC 2047 encoding. The
2447 Sieve code now uses this.
2449 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2450 data_4xx, lost_connection, tls_required.
2452 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2453 message was being sent as an SMTP response.
2455 PH/61 Add -l and -k options to exicyclog.
2457 PH/62 When verifying, if an address was redirected to one new address, so that
2458 verification continued, and the new address failed or deferred after
2459 having set something in $address_data, the value of $address_data was not
2460 passed back to the ACL. This was different to the case when no
2461 redirection occurred. The value is now passed back in both cases.
2463 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2464 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2465 use login_cap.h, so on its own it isn't the distinguishing feature. The
2466 new name refers directly to the setclassresources() function.
2468 PH/65 Added configuration files for NetBSD3.
2470 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2472 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2473 is preferred over IPv4.
2475 PH/68 The bounce_return_message and bounce_return_body options were not being
2476 honoured for bounces generated during the reception of non-SMTP messages.
2477 In particular, this applied to messages rejected by the ACL. This bug has
2478 been fixed. However, if bounce_return_message is true and bounce_return_
2479 body is false, the headers that are returned for a non-SMTP message
2480 include only those that have been read before the error was detected.
2481 (In the case of an ACL rejection, they have all been read.)
2483 PH/69 The HTML version of the specification is now built in a directory called
2484 spec_html instead of spec.html, because the latter looks like a path with
2485 a MIME-type, and this confuses some software.
2487 PH/70 Catch two compiler warnings in sieve.c.
2489 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2490 function verify_get_ident() calls ip_connect() to connect a socket, but
2491 if the "connect()" function timed out, ip_connect() used to close the
2492 socket. However, verify_get_ident() also closes the socket later, and in
2493 between Exim writes to the log, which may get opened at this point. When
2494 the socket was closed in ip_connect(), the log could get the same file
2495 descriptor number as the socket. This naturally causes chaos. The fix is
2496 not to close the socket in ip_connect(); the socket should be closed by
2497 the function that creates it. There was only one place in the code where
2498 this was missing, in the iplookup router, which I don't think anybody now
2499 uses, but I've fixed it anyway.
2501 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2502 well as to direct DNS lookups. Otherwise the handling of names in host
2503 lists is inconsistent and therefore confusing.
2509 PH/01 Two changes to the default runtime configuration:
2511 (1) Move the checks for relay_from_hosts and authenticated clients from
2512 after to before the (commented out) DNS black list checks.
2514 (2) Add control=submission to the relay_from_hosts and authenticated
2515 clients checks, on the grounds that messages accepted by these
2516 statements are most likely to be submissions.
2518 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2520 (1) Generate an error if the third argument for the ${prvs expansion is
2523 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2526 (3) Reset the variables that are obtained from the first argument of
2527 ${prvscheck and used in the second argument before leaving the code,
2528 because their memory is reclaimed, so using them afterwards may do
2531 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2532 one (it's much easier than Tom thought :-).
2534 (5) Because of (4), we can now allow for the use of $prvscheck_result
2535 inside the third argument.
2537 PH/03 For some reason, the default setting of PATH when running a command from
2538 a pipe transport was just "/usr/bin". I have changed it to
2541 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2542 anything to be listed in the output from -bV.
2544 PH/05 When a filter generated an autoreply, the entire To: header line was
2545 quoted in the delivery log line, like this:
2547 => >A.N.Other <ano@some.domain> <original@ddress> ...
2549 This has been changed so that it extracts the operative address. There
2550 may be more than one such address. If so, they are comma-separated, like
2553 => >ano@some.domain,ona@other.domain <original@ddress> ...
2555 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2556 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2557 not being looked up in the rDNS to get a host name, Exim was showing the
2558 IP address twice in Received: lines, even though the IP addresses were
2559 identical. For example:
2561 Received: from [1.2.3.4] (helo=[1.2.3.4])
2563 However, if the real host name was known, it was omitting the HELO data
2564 if it matched the actual IP address. This has been tidied up so that it
2565 doesn't show the same IP address twice.
2567 PH/07 When both +timestamp and +memory debugging was on, the value given by
2568 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2569 called by the debug printing, thereby overwriting the timestamp buffer.
2570 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2572 PH/08 When the original message was included in an autoreply transport, it
2573 always said "this is a copy of the message, including all the headers",
2574 even if body_only or headers_only was set. It now gives an appropriate
2577 PH/09 Applied a patch from the Sieve maintainer which:
2579 o fixes some comments
2580 o adds the (disabled) notify extension core
2581 o adds some debug output for the result of if/elsif tests
2582 o points to the current vacation draft in the documentation
2583 and documents the missing references header update
2587 o fixes a bug in processing the envelope test (when testing
2588 multiple envelope elements, the last element determinted the
2591 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2592 Electronic Mail") by including:
2594 Auto-submitted: auto-generated
2596 in the messages that it generates (bounce messages and others, such as
2597 warnings). In the case of bounce messages for non-SMTP mesages, there was
2598 also a typo: it was using "Auto_submitted" (underscore instead of
2599 hyphen). Since every message generated by Exim is necessarily in response
2600 to another message, thes have all been changed to:
2602 Auto-Submitted: auto-replied
2604 in accordance with these statements in the RFC:
2606 The auto-replied keyword:
2608 - SHOULD be used on messages sent in direct response to another
2609 message by an automatic process,
2611 - MUST NOT be used on manually-generated messages,
2613 - MAY be used on Delivery Status Notifications (DSNs) and Message
2614 Disposition Notifications (MDNs),
2616 - MUST NOT be used on messages generated by automatic or periodic
2617 processes, except for messages which are automatic responses to
2620 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2621 to the default Received: header definition.
2623 PH/12 Added log selector acl_warn_skipped (default on).
2625 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2626 variables because (a) they are in the wrong storage pool and (b) even if
2627 they were copied, it wouldn't work properly because of the caching.
2629 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2630 checking when decoding. Apparently there are clients that generate
2631 overlong encoded strings. Why am I not surprised?
2633 PH/15 If the first argument of "${if match_address" was not empty, but did not
2634 contain an "@" character, Exim crashed. Now it writes a panic log message
2635 and treats the condition as false.
2637 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2639 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2640 extension "envelope-auth". The code is finished and in agreement with
2641 other implementations, but there is no documentation so far and in fact,
2642 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2643 not changing the active code.
2645 Print executed "if" and "elsif" statements when debugging is used. This
2646 helps a great deal to understand what a filter does.
2648 Document more things not specified clearly in RFC3028. I had all this
2649 sorted out, when out of a sudden new issues came to my mind. Oops."
2651 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2654 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2655 canonical form (as documented). However, after a host name lookup from
2656 the IP address, check_host() was doing a simple string comparison with
2657 addresses acquired from the DNS when checking that the found name did
2658 have the original IP as one of its addresses. Since any found IPv6
2659 addresses are likely to be in abbreviated form, the comparison could
2660 fail. Luckily, there already exists a function for doing the comparison
2661 by converting both addresses to binary, so now that is used instead of
2662 the text comparison.
2664 PH/20 There was another similar case to PH/19, when a complete host name was
2665 given in a host list; looking up its IP address could give an abbreviated
2666 form, whereas the current host's name might or might not be abbreviated.
2667 The same fix has been applied.
2673 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2674 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2677 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2678 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2680 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2682 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2683 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2684 Also, for "Auto-Submitted", treat anything other than "no" as
2685 non-personal, in accordance with RFC 3834. (Previously it treated
2686 anything starting "auto-" as non-personal.)
2688 TF/01 The control=submission/name=... option had a problem with syntax
2689 errors if the name included a slash character. The /name= option
2690 now slurps the rest of the string, so it can include any characters
2691 but it must come last in the list of options (after /sender_retain
2694 PH/05 Some modifications to the interface to the fake nameserver for the new
2702 TK/01 Added the "success_on_redirect" address verification option. See
2703 NewStuff for rationale and an example.
2705 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2707 PH/02 Patch to exigrep to allow it to work on syslog lines.
2709 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2710 fread() to skip over the body file's header line, because in Cygwin the
2711 header line is locked and is inaccessible.
2713 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2714 co-exist for some time) to make it clear that it is the Exim ID that is
2715 referenced, not the Message-ID: header line.
2717 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2718 string_format() function, because snprintf() does not exist on all
2721 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2722 ${stat: expansion item.
2724 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2725 protocol synchronization error", to keep the pedants happy.
2727 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2728 well as for IRIX systems, when gcc is being used. See the host.c source
2731 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2733 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2736 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2737 transport and to the smtp transport in LMTP mode.
2739 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2741 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2742 run a filter in a subprocess. This could lead to confusion in subsequent
2743 lookups in the parent process. There should also be a search_tidyup() at
2744 the end of the subprocess.
2746 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2747 only if the host matched helo_try_verify_hosts, which caused the
2748 verification to occur when the EHLO/HELO command was issued. The ACL just
2749 tested the remembered result. Now, if a previous verification attempt has
2750 not happened, "verify = helo" does it there and then.
2752 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2754 TK/03 Fix log output including CR from clamd.
2756 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2757 exist provoked a memory error which could cause a segfault.
2759 PH/15 Installed PCRE 6.2
2761 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2763 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2764 of the problem. Specifically, suggested +O2 rather than +O1 for the
2767 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2769 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2770 record, the port was not correctly logged when the outgoing_port log
2771 selector was set (it logged the transort's default port).
2773 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2774 fallback_hosts, and "hosts" in the smtp transport.
2776 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2777 host errors such as "Connection refused".
2779 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2780 authentication with radiusclient 0.4.9:
2782 - Error returned from rc_read_config was caught wrongly
2783 - Username/password not passed on to radius server due to wrong length.
2785 The presumption is that some radiusclient API changes for 4.51/PH/17
2786 were not taken care of correctly. The code is still untested by me (my
2787 Linux distribution still has 0.3.2 of radiusclient), but it was
2788 contributed by a Radius user.
2790 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2791 expanding the "port" option of the smtp transport.
2793 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2794 while reading a MIME header. Thanks to Tom Hughes for a patch.
2796 PH/24 Include config.h inside local_scan.h so that configuration settings are
2799 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2800 that in an ACL for RCPT (for example), you can examine exactly what was
2803 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2804 commands, but it was not correctly comparing the address with the actual
2805 client host address. Thus, it would show the EHLO address in Received:
2806 header lines when this was not necessary.
2808 PH/27 Added the % operator to ${eval:}.
2810 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2811 it should be ignoring failures (because with -C, for example, it has lost
2812 privilege). It wasn't ignoring creation failures other than "already
2815 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2818 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2819 because an input line was too long, either on its own, or by virtue of
2820 too many continuations, the temporary file was not being removed, and the
2821 return code was incorrect.
2823 PH/31 Missing "BOOL" in function definition in filtertest.c.
2825 PH/32 Applied Sieve patches from the maintainer.
2827 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2829 PH/33 Added "verify = not_blind".
2831 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2832 Local/Makefile (with some defaults set). These are used in built scripts
2833 such as exicyclog, but they have never been used in the exim_install
2834 script (though there are many overriding facilities there). I have
2835 arranged that the exim_install script now takes note of these two
2838 PH/35 Installed configuration files for Dragonfly.
2840 PH/36 When a locally submitted message by a trusted user did not contain a
2841 From: header, and the sender address was obtained from -f or from an SMTP
2842 MAIL command, and the trusted user did not use -F to supply a sender
2843 name, $originator_name was incorrectly used when constructing a From:
2844 header. Furthermore, $originator_name was used for submission mode
2845 messages from external hosts without From: headers in a similar way,
2846 which is clearly wrong.
2848 PH/37 Added control=suppress_local_fixups.
2850 PH/38 When log_selector = +received_sender was set, and the addition of the
2851 sender made the log line's construction buffer exactly full, or one byte
2852 less than full, an overflow happened when the terminating "\n" was
2855 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2856 when the result of a list match is failure because a DNS lookup failed.
2858 PH/40 RM_COMMAND is now used in the building process.
2860 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2861 the "build-* directories that it finds.
2863 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2864 address in a domain literal was a prefix of an interface address.
2866 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2867 when verifying a sender address, unless rewrite_headers is false.
2869 PH/44 Wrote a long comment about why errors_to addresses are verified as
2870 recipients, not senders.
2872 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2873 the ratelimit ACL was added.
2875 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2877 PH/46 Added extra information about PostgreSQL errors to the error string.
2879 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2880 avoiding the need to install special zones in a real server. This is
2881 backwards compatible; if it can't find the fake resolver, it drops back.
2882 Thus, both old and new test suites can be run.
2884 TF/02 Added util/ratelimit.pl
2886 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2887 clock is set back in time.
2889 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2890 Candler <B.Candler@pobox.com>.
2892 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2893 OK for addresses that are the result of redirections.
2895 PH/48 A number of further additions for the benefit of the new test suite,
2896 including a fake gethostbyname() that interfaces to the fake DNS resolver
2899 TF/06 The fix for widen_domains has also been applied to qualify_single and
2900 search_parents which are the other dnslookup options that can cause
2903 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2906 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2908 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2909 service" if it encountered an MX record with an empty target hostname.
2910 The message is now "an MX or SRV record indicated no SMTP service".
2912 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2913 if the DNS of the sending site is misconfigured. This is quite a
2914 common situation. This change restores the behaviour of treating a
2915 helo verification defer as a failure.
2917 PH/51 If self=fail was set on a router, the bounce message did not include the
2918 actual error message.
2924 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2926 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2927 command itself ended in error, the underlying message about the transport
2928 filter timeout was being overwritten with the pipe command error. Now the
2929 underlying error message should be appended to the second error message.
2931 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2933 PH/02 Reduce the amount of output that "make" produces by default. Full output
2934 can still be requested.
2936 PH/03 The warning log line about a condition test deferring for a "warn" verb
2937 was being output only once per connection, rather than after each
2938 occurrence (because it was using the same function as for successful
2939 "warn" verbs). This seems wrong, so I have changed it.
2941 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2942 it should not have, which might have caused a crash in the right
2943 circumstances, but probably never did.
2945 PH/04 Installed a modified version of Tony Finch's patch to make submission
2946 mode fix the return path as well as the Sender: header line, and to
2947 add a /name= option so that you can make the user's friendly name appear
2950 TF/03 Added the control = fakedefer ACL modifier.
2952 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2953 Mark Lowes for thorough testing.
2955 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2957 TK/03 Merged latest SRS patch from Miles Wilton.
2959 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2960 with the definition in sysexits.h (which is #included earlier).
2961 Fortunately, Exim does not actually use EX_OK. The code used to try to
2962 preserve the sysexits.h value, by assumimg that macro definitions were
2963 scanned for macro replacements. I have been disabused of this notion,
2964 so now the code just undefines EX_OK before #including unistd.h.
2966 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2967 in the smtp transport. When a block could not be written in a single
2968 write() function, the timeout was being re-applied to each part-write.
2969 This seems wrong - if the receiver was accepting one byte at a time it
2970 would take for ever. The timeout is now adjusted when this happens. It
2971 doesn't have to be particularly precise.
2973 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2974 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2976 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2977 without a domain if the check to <postmaster@domain> fails.
2979 SC/01 Eximstats: added -xls and the ability to specify output files
2980 (patch written by Frank Heydlauf).
2982 SC/02 Eximstats: use FileHandles for outputing results.
2984 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2986 SC/04 Eximstats: fixed display of large numbers with -nvr option
2988 SC/05 Eximstats: fixed merging of reports with empty tables.
2990 SC/06 Eximstats: added the -include_original_destination flag
2992 SC/07 Eximstats: removed tabs and trailing whitespace.
2994 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2996 TK/06 MBOX spool code: Add real "From " MBOX separator line
2997 so the .eml file is really in mbox format (even though
2998 most programs do not really care). Patch from Alex Miller.
3000 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3001 The latter is generated from $received_to and is only set if the
3002 message has one envelope recipient. SA can use these headers,
3003 obviously out-of-the-box. Patch from Alex Miller.
3005 PH/08 The ${def test on a variable was returning false if the variable's
3006 value was "0", contrary to what the specification has always said!
3007 The result should be true unless the variable is empty.
3009 PH/09 The syntax error of a character other than { following "${if
3010 def:variable_name" (after optional whitespace) was not being diagnosed.
3011 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3012 accidental colon was present, for example, could give incorrect results.
3014 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3015 result is used (not including appendfile, where other changes are about
3018 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3019 This involved changing a lot of size variables from int to off_t. It
3020 should work with maildirs and everything.
3022 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3023 spamd dying while we are connected to it.
3025 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3028 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3029 function for BDB 4.3.
3031 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3033 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3034 the sources for PCRE have been re-arranged and more files are now
3037 PH/15 The code I had for printing potentially long long variables in PH/11
3038 above was not the best (it lost precision). The length of off_t variables
3039 is now inspected at build time, and an appropriate printing format (%ld
3040 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3041 to be "long long int" or "long int". This is needed for the internal
3042 formatting function string_vformat().
3044 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3045 the configuration file to be ":syslog", then the script "guesses" where
3046 the logs files are, rather than using the compiled in default. In our
3047 case the guess is not the same as the compiled default, so the script
3048 suddenly stopped working when I started to use syslog. The patch checks
3049 to see if log_file_path is "". If so, it attempts to read it from exim
3050 with no configuration file to get the compiled in version, before it
3051 falls back to the previous guessing code."
3053 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3054 implementing BATV in an Exim configuration. See NewStuff for the gory
3057 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3058 Makefile that are specific to HP-UX.
3060 PH/18 If the "use_postmaster" option was set for a recipient callout together
3061 with the "random" option, the postmaster address was used as the MAIL
3062 FROM address for the random test, but not for the subsequent recipient
3063 test. It is now used for both.
3065 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3066 patch removes a few documentation additions to RFC 3028, because the
3067 latest draft now contains them. It adds the new en;ascii-case comparator
3068 and a new error check for 8bit text in MIME parts. Comparator and
3069 require names are now matched exactly. I enabled the subaddress
3070 extension, but it is not well tested yet (read: it works for me)."
3072 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3073 rework some of the code of TK/09 above to avoid the hardwired use of
3074 "%lld" and "long long". Replaced the call to snprintf() with a call to
3077 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3078 records point to non-existent hosts", "retry timeout exceeded", and
3079 "retry time not reached for any host after a long failure period".
3081 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3082 experimental DomainKeys support:
3084 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3085 (2) On an error such as an illegally used "control", the wrong name for
3086 the control was given.
3088 These problems did NOT occur unless DomainKeys support was compiled.
3090 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3092 PH/24 Added ${if match_ip condition.
3094 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3095 databases so that it will be absolutely obvious if a crash occurs in the
3096 DB library. This is a regular occurrence (often caused by mis-matched
3099 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3100 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3101 picked up on a user's system that detects such things. There doesn't seem
3102 to be a gcc warning option for this - only an attribute that has to be
3103 put on the function's prototype. It seems that in Fedora Core 4 they have
3104 set this on a number of new functions. No doubt there will be more in due
3107 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3108 specify a transport. However, if an address that was verified by such a
3109 router was the subject of a callout, Exim crashed because it tried to
3110 read the rcpt_include_affixes from the non-existent transport. Now it
3111 just assumes that the setting of that option is false. This bug was
3112 introduced by 4.51/PH/31.
3114 PH/28 Changed -d+all to exclude +memory, because that information is very
3115 rarely of interest, but it makes the output a lot bigger. People tend to
3116 do -d+all out of habit.
3118 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3119 code in os-type was giving problems when libc.so lives in lib64, like on
3122 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3123 aren't the modern standard, and it seems that some systems' include files
3124 don't always have them. Exim was already checking for some of the newer
3125 ones like T_AAAA, and defining it itself. I've added checks for all the
3126 record types that Exim uses.
3128 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3129 not automatically generating a new one, as it is supposed to. This
3130 prevented TLS from working. If the file did exist, but contained invalid
3131 data, a new version was generated, as expected. It was only the case of a
3132 non-existent file that was broken.
3134 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3135 with a change in libdomainkeys > 0.64.
3137 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3138 from DNS. If the selector record carries the flag, it now has
3139 precedence over the domain-wide flag.
3141 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3143 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3144 the use of a port name that isn't defined in /etc/services) occurred, the
3145 message was deferred as in a normal delivery, and thus remained on the
3146 spool, instead of being failed because of the mua_wrapper setting. This
3147 is now fixed, and I tidied up some of the mua_wrapper messages at the
3150 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3151 the messages in a hash of arrays rather than using individual hashes.
3152 This is a bit cleaner and results in dramatic memory savings, albeit
3153 at a slight CPU cost.
3155 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3156 as requested by Marc Sherman.
3158 SC/10 Eximstats: added histograms for user specified patterns as requested
3161 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3163 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3164 fopen() in the content-scanning modules that did not already have it.
3170 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3171 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3173 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3175 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3177 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3178 to test Sieve filters that use "vacation".
3180 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3181 that changes the way the GnuTLS parameters are stored in the cache file.
3182 The new format can be generated externally. For backward compatibility,
3183 if the data in the cache doesn't make sense, Exim assumes it has read an
3184 old-format file, and it generates new data and writes a new file. This
3185 means that you can't go back to an older release without removing the
3188 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3189 work if there are any delivery delays because "one_time" forces the
3190 parent to be marked "delivered", so its unseen clone is never tried
3191 again. For this reason, Exim now forbids the simultaneous setting of
3194 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3195 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3196 entry if you want to know the details. The fix, however, neglected to
3197 consider the case when local delivery batching is involved. The test for
3198 "previously delivered" was not happening when checking to see if an
3199 address could be batched with a previous (undelivered) one; under
3200 certain circumstances this could lead to multiple deliveries to the same
3203 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3204 in its include files, and this causes problems building Exim.
3206 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3207 header_syntax) but Exim was just ignoring anything given after a slash.
3208 In particular, this caused confusion with an attempt to use "verify =
3209 reverse_host_lookup/defer_ok". An error is now given when options are
3210 supplied for verify items that do not have them. (Maybe reverse_host_
3211 lookup should have a defer_ok option, but that's a different point.)
3213 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3214 defined by RFC 821) to 2048, because there were problems with some AUTH
3215 commands, and RFC 1869 says the size should be increased for extended
3216 SMTP commands that take arguments.
3218 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3221 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3222 "unknown" error; now it says that the functionality isn't in the binary.
3224 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3225 an address' error message when a string expansion fails (syntax or
3226 whatever). Otherwise the password may appear in the log. Following change
3227 PH/42 below, there is no longer a chance of it appearing in a bounce
3230 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3232 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3233 of its addresses was ever tried. (Bugzilla bug #2.)
3235 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3236 the result incorrectly in the debug output. (It correctly added a newline
3237 to what was transported.)
3239 TF/01 Added $received_time.
3241 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3242 commented out examples of how to interface to a virus scanner and to
3243 SpamAssassin. Also added commented examples of av_scanner and
3244 spamd_address settings.
3246 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3247 and controls are allowed in which ACLs. There were a couple of minor
3248 errors. Some of the entries in the conditions table (which is a table of
3249 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3250 negation of where the condition IS allowed.
3252 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3254 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3255 header file does not have a version number, so I've had to invent a new
3256 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3257 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3258 radiusclient), but it was contributed by a Radius user.
3260 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3261 files or directories, for OpenSSL.
3263 PH/19 When an Exim process that is running as root has to create an Exim log
3264 file, it does so in a subprocess that runs as exim:exim so as to get the
3265 ownership right at creation (otherwise, other Exim processes might see
3266 the file with the wrong ownership). There was no test for failure of this
3267 fork() call, which would lead to the process getting stuck as it waited
3268 for a non-existent subprocess. Forks do occasionally fail when resources
3269 run out. I reviewed all the other calls to fork(); they all seem to check
3272 PH/20 When checking for unexpected SMTP input at connect time (before writing
3273 the banner), Exim was not dealing correctly with a non-positive return
3274 from the read() function. If the client had disconnected by this time,
3275 the result was a log entry for a synchronization error with an empty
3276 string after "input=" when read() returned zero. If read() returned -1
3277 (an event I could not check), uninitialized data bytes were printed.
3278 There were reports of junk text (parts of files, etc) appearing after
3281 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3283 PH/22 Added support for macro redefinition, and (re)definition in between
3284 driver and ACL definitions.
3286 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3287 forgetting to use the resulting value; it was using the unexpanded value.
3289 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3290 hadn't been configured. The fix is from Juergen Kreileder, who
3291 understands it better than I do:
3293 "Here's what I see happening with three configured cyrus_sasl
3294 authenticators configured (plain, login, cram-md5):
3296 On startup auth_cyrus_sasl_init() gets called for each of these.
3297 This means three calls to sasl_listmech() without a specified mech_list.
3298 => SASL tests which mechs of all available mechs actually work
3299 => three warnings about OTP not working
3300 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3302 With the patch, sasl_listmech() also gets called three times. But now
3303 SASL's mech_list option is set to the server_mech specified in the the
3304 authenticator. Or in other words, the answer from sasl_listmech()
3305 gets limited to just the mech you're testing for (which is different
3307 => the return list contains just 'plain' or 'login', 'cram-md5' or
3308 nothing depending on the value of ob->server_mech.
3310 I've just tested the patch: Authentication still works fine,
3311 unavailable mechs specified in the exim configuration are still
3312 caught, and the auth.log warnings about OTP are gone."
3314 PH/25 When debugging is enabled, the contents of the command line are added
3315 to the debugging output, even when log_selector=+arguments is not
3318 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3319 answer is "GNU", and only if the return is "GNU/something" is the answer
3322 PH/27 $acl_verify_message is now set immediately after the failure of a
3323 verification in an ACL, and so is available in subsequent modifiers. In
3324 particular, the message can be preserved by coding like this:
3326 warn !verify = sender
3327 set acl_m0 = $acl_verify_message
3329 Previously, $acl_verify_message was set only while expanding "message"
3330 and "log_message" when a very denied access.
3332 PH/28 Modified OS/os.c-Linux with
3334 -#ifndef OS_LOAD_AVERAGE
3335 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3337 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3338 nomenclature these days.)
3340 PH/29 Installed patch from the Sieve maintainer that adds the options
3341 sieve_useraddress and sieve_subaddress to the redirect router.
3343 PH/30 In these circumstances:
3344 . Two addresses routed to the same list of hosts;
3345 . First host does not offer TLS;
3346 . First host accepts first address;
3347 . First host gives temporary error to second address;
3348 . Second host offers TLS and a TLS session is established;
3349 . Second host accepts second address.
3350 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3351 and peerdn, if requested) that were in fact used only for the second
3354 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3355 attention to any local part prefix or suffix that was matched by the
3356 router that accepted the address. It now behaves in the same way as it
3357 does for delivery: the affixes are removed from the local part unless
3358 rcpt_include_affixes is set on the transport.
3360 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3361 timeout during the DATA phase of an incoming message.
3363 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3364 applied a patch sanctioned by the Sieve maintainer.
3366 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3367 the uid or gid is negative. A case of a negative gid caused this to be
3368 noticed. The fix allows for either to be negative.
3370 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3371 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3374 PH/36 The change PH/12 above was broken. Fixed it.
3376 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3377 the grounds that routing the same address twice would always produce the
3378 same answer. This might have been true once, but it is certainly no
3379 longer true now. Routing a child address may depend on the previous
3380 routing that produced that child. Some complicated redirection strategies
3381 went wrong when messages had multiple recipients, and made Exim's
3382 behaviour dependent on the order in which the addresses were given.
3384 I have moved the duplicate checking until after the routing is complete.
3385 Exim scans the addresses that are assigned to local and remote
3386 transports, and removes any duplicates. This means that more work will be
3387 done, as duplicates will always all be routed, but duplicates are
3388 presumably rare, so I don't expect this is of any significance.
3390 For deliveries to pipes, files, and autoreplies, the duplicate checking
3391 still happens during the routing process, since they are not going to be
3394 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3395 It corrects a timeout issue with spamd. This is Ian's comment: "The
3396 background is that sometimes spamd either never reads data from a
3397 connection it has accepted, or it never writes response data. The exiscan
3398 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3399 blindly assumes that writes won't block so it may never time out."
3401 PH/39 Allow G after quota size as well as K and M.
3403 PH/40 The value set for $authenticated_id in an authenticator may not contain
3404 binary zeroes or newlines because the value is written to log lines and
3405 to spool files. There was no check on this. Now the value is run through
3406 the string_printing() function so that such characters are converted to
3407 printable escape sequences.
3409 PH/41 $message_linecount is a new variable that contains the total number of
3410 lines in the message. Compare $body_linecount, which is the count for the
3413 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3414 bounce and delay warning messages, except in certain special cases, which
3417 (a) An SMTP error message from a remote host;
3418 (b) A message specified in a :fail: redirection;
3419 (c) A message specified in a "fail" command in a system filter;
3420 (d) A message specified in a FAIL return from the queryprogram router;
3421 (e) A message specified by the cannot_route_message router option.
3423 In these cases only, Exim does include the error details in bounce and
3424 warning messages. There are also a few cases where bland messages such
3425 as "unrouteable address" or "local delivery error" are given.
3427 PH/43 $value is now also set for the "else" part of a ${run expansion.
3429 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3430 being worked on, but at least Exim now implements the latest version to
3433 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3434 process to complete was treated as a delivery failure, a timeout while
3435 writing the message to the pipe was logged, but erroneously treated as a
3436 successful delivery. Such timeouts include transport filter timeouts. For
3437 consistency with the overall process timeout, these timeouts are now
3438 treated as errors, giving rise to delivery failures by default. However,
3439 there is now a new Boolean option for the pipe transport called
3440 timeout_defer, which, if set TRUE, converts the failures into defers for
3441 both kinds of timeout. A transport filter timeout is now identified in
3444 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3445 systems where "make" and "gmake" are different, calling "gmake" at top
3446 level broke things. I've arranged for the value of $(MAKE) to be passed
3447 from the Makefile to this script so that it can call the same version of
3451 A note about Exim versions 4.44 and 4.50
3452 ----------------------------------------
3454 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3455 changes of various kinds. As a consequence, a big documentation update was
3456 needed. This delayed the release for rather longer than seemed good, especially
3457 in the light of a couple of (minor) security issues. Therefore, the changes
3458 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3459 release. So 4.44 and 4.50 are in effect two different branches that both start
3462 I have left the 4.50 change log unchanged; it contains all the changes since
3463 4.43. The change log for 4.44 is below; many of its items are identical to
3464 those for 4.50. This seems to be the most sensible way to preserve the
3465 historical information.
3471 1. Minor wording change to the doc/README.SIEVE file.
3473 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3474 computation of the current number of files was incorrect.
3476 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3477 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3480 4. Give more explanation in the error message when the command for a transport
3481 filter fails to execute.
3483 5. There are several places where Exim runs a non-Exim command in a
3484 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3485 was being done only for the command run by the queryprogram router. It is
3486 now done for all such subprocesses. The other cases are: ${run, transport
3487 filters, and the commands run by the lmtp and pipe transports.
3489 6. Added CONFIGURE_GROUP build-time option.
3491 7. Some older OS have a limit of 256 on the maximum number of file
3492 descriptors. Exim was using setrlimit() to set 1000 as a large value
3493 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3494 systems. I've change it so that if it can't get 1000, it tries for 256.
3496 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3497 was an oversight, and furthermore, ever since the addition of extra
3498 controls (e.g. 4.43/32), the checks on when to allow different forms of
3499 "control" were broken. There should now be diagnostics for all cases when a
3500 control that does not make sense is encountered.
3502 9. Added the /retain_sender option to "control=submission".
3504 10. $recipients is now available in the predata ACL (oversight).
3506 11. Tidy the search cache before the fork to do a delivery from a message
3507 received from the command line. Otherwise the child will trigger a lookup
3508 failure and thereby defer the delivery if it tries to use (for example) a
3509 cached ldap connection that the parent has called unbind on.
3511 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3512 of $address_data from the recipient verification was clobbered by the
3513 sender verification.
3515 13. The value of address_data from a sender verification is now available in
3516 $sender_address_data in subsequent conditions in the ACL statement.
3518 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3520 15. Added a new option "connect=<time>" to callout options, to set a different
3523 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3524 was its contents. (It was OK if the option was not defined at all.)
3526 17. A "Completed" log line is now written for messages that are removed from
3527 the spool by the -Mrm option.
3529 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3530 information about exactly what failed.
3532 19. Added -dd to debug only the daemon process.
3534 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3535 handles timeouts, both on the server side and network timeouts. Renamed the
3536 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3538 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3539 It is now set to "smtps".
3541 22. $host_address is now set to the target address during the checking of
3542 ignore_target_hosts.
3544 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3545 being passed; this would have caused $sender_host_name to have been used if
3546 matching the list had actually called for a host name (not very likely,
3547 since this list is usually IP addresses). A host name is now passed as
3550 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3551 code that specifies a non-restarting handler (typically sigaction() in
3552 modern systems) in an attempt to fix a rare and obscure crash bug.
3554 25. Narrowed the window for a race in the daemon that could cause it to ignore
3555 SIGCHLD signals. This is not a major problem, because they are used only to
3556 wake it up if nothing else does.
3558 26. A malformed maildirsize file could cause Exim to calculate negative values
3559 for the mailbox size or file count. Odd effects could occur as a result.
3560 The maildirsize information is now recalculated if the size or filecount
3563 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3564 support for a long time. Removed HAVE_SYS_VFS_H.
3566 28. Installed the latest version of exipick from John Jetmore.
3568 29. In an address list, if the pattern was not a regular expression, an empty
3569 subject address (from a bounce message) matched only if the pattern was an
3570 empty string. Non-empty patterns were not even tested. This was the wrong
3571 because it is perfectly reasonable to use an empty address as part of a
3572 database query. An empty address is now tested by patterns that are
3573 lookups. However, all the other forms of pattern expect the subject to
3574 contain a local part and a domain, and therefore, for them, an empty
3575 address still always fails if the pattern is not itself empty.
3577 30. Exim went into a mad DNS loop when attempting to do a callout where the
3578 host was specified on an smtp transport, and looking it up yielded more
3579 than one IP address.
3581 31. Re-factored the code for checking spool and log partition space into a
3582 function that finds that data and another that does the check. The former
3583 is then used to implement four new variables: $spool_space, $log_space,
3584 $spool_inodes, and $log_inodes.
3586 32. The RFC2047 encoding function was originally intended for short strings
3587 such as real names; it was not keeping to the 75-character limit for
3588 encoded words that the RFC imposes. It now respects the limit, and
3589 generates multiple encoded words if necessary. To be on the safe side, I
3590 have increased the buffer size for the ${rfc2047: expansion operator from
3593 33. It is now permitted to omit both strings after an "if" condition; if the
3594 condition is true, the result is "true". As before, when the second string
3595 is omitted, a false condition yields an empty string. This makes it less
3596 cumbersome to write custom ACL and router conditions.
3598 34. Failure to deliver a bounce message always caused it to be frozen, even if
3599 there was an errors_to setting on the router. The errors_to setting is now
3602 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3603 canonical form (fully expanded) before being placed in
3604 $sender_host_address.
3606 36. The table in the code that translates DNS record types into text (T_A to
3607 "A" for instance) was missing entries for NS and CNAME. It is just possible
3608 that this could have caused confusion if both these types were looked up
3609 for the same domain, because the text type is used as part of Exim's
3610 per-process caching. But the chance of anyone hitting this buglet seems
3613 37. The dnsdb lookup has been extended in a number of ways.
3615 (1) There is a new type, "zns", which walks up the domain tree until it
3616 finds some nameserver records. It should be used with care.
3618 (2) There is a new type, "mxh", which is like "mx" except that it returns
3619 just the host names, not the priorities.
3621 (3) It is now possible to give a list of domains (or IP addresses) to be
3622 looked up. The behaviour when one of the lookups defers can be
3623 controlled by a keyword.
3625 (4) It is now possible to specify the separator character for use when
3626 multiple records are returned.
3628 38. The dnslists ACL condition has been extended: it is now possible to supply
3629 a list of IP addresses and/or domains to be looked up in a particular DNS
3632 39. Added log_selector=+queue_time_overall.
3634 40. When running the queue in the test harness, wait just a tad after forking a
3635 delivery process, to get repeatability of debugging output.
3637 41. Include certificate and key file names in error message when GnuTLS fails
3638 to set them up, because the GnuTLS error message doesn't include the name
3639 of the failing file when there is a problem reading it.
3641 42. Allow both -bf and -bF in the same test run.
3643 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3645 44. The "Exiscan patch" is now merged into the mainline Exim source.
3647 45. Sometimes the final signoff response after QUIT could fail to get
3648 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3649 < 0 before doing a fflush(). This bug looks as though it goes back to the
3650 introduction of TLS in release 3.20, but "sometimes" must have been rare
3651 because the tests only now provoked it.
3653 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3654 (this can affect the format of dates).
3656 47. exim_tidydb, when checking for the continued existence of a message for
3657 which it has found a message-specific retry record, was not finding
3658 messages that were in split spool directories. Consequently, it was
3659 deleting retry records that should have stayed in existence.
3661 48. Steve fixed some bugs in eximstats.
3663 49. The SPA authentication driver was not abandoning authentication and moving
3664 on to the next authenticator when an expansion was forced to fail,
3665 contradicting the general specification for all authenticators. Instead it
3666 was generating a temporary error. It now behaves as specified.
3668 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3669 (the order specifies the preference for clients). The order is now AES256,
3670 AES128, 3DES, ARCFOUR128.
3672 51. Small patch to Sieve code - explicitly set From: when generating an
3675 52. Exim crashed if a remote delivery caused a very long error message to be
3676 recorded - for instance if somebody sent an entire SpamAssassin report back
3677 as a large number of 550 error lines. This bug was coincidentally fixed by
3678 increasing the size of one of Exim's internal buffers (big_buffer) that
3679 happened as part of the Exiscan merge. However, to be on the safe side, I
3680 have made the code more robust (and fixed the comments that describe what
3683 53. Now that there can be additional text after "Completed" in log lines (if
3684 the queue_time_overall log selector is set), a one-byte patch to exigrep
3685 was needed to allow it to recognize "Completed" as not the last thing in
3688 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3689 patch that reportedly fixes this has been added. I am not expert enough to
3690 create a test for it. This is what the patch creator wrote:
3692 "I found a little strange behaviour of ldap code when working with
3693 Windows 2003 AD Domain, where users was placed in more than one
3694 Organization Units. When I tried to give exim partial DN, the exit code
3695 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3696 But simultaneously result of request was absolutely normal ldap result,
3697 so I produce this patch..."
3699 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3700 REFERENCE, so I have modified the code to exclude the patch when that macro
3703 55. Some experimental protocols are using DNS PTR records for new purposes. The
3704 keys for these records are domain names, not reversed IP addresses. The
3705 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3706 leaves it alone. Component reversal etc. now happens only for IP addresses.
3709 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3711 57. Double the size of the debug message buffer (to 2048) so that more of very
3712 long debug lines gets shown.
3714 58. The exicyclog utility now does better if the number of log files to keep
3715 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3717 59. Two changes related to the smtp_active_hostname option:
3719 (1) $smtp_active_hostname is now available as a variable.
3720 (2) The default for smtp_banner uses $smtp_active_hostname instead
3721 of $primary_hostname.
3723 60. The host_aton() function is supposed to be passed a string that is known
3724 to be a valid IP address. However, in the case of IPv6 addresses, it was
3725 not checking this. This is a hostage to fortune. Exim now panics and dies
3726 if the condition is not met. A case was found where this could be provoked
3727 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3728 components; fortuitously, this particular loophole had already been fixed
3729 by change 4.50/55 above.
3731 If there are any other similar loopholes, the new check in host_aton()
3732 itself should stop them being exploited. The report I received stated that
3733 data on the command line could provoke the exploit when Exim was running as
3734 exim, but did not say which command line option was involved. All I could
3735 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3736 running as the user.
3739 61. There was a buffer overflow vulnerability in the SPA authentication code
3740 (which came originally from the Samba project). I have added a test to the
3741 spa_base64_to_bits() function which I hope fixes it.
3744 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3745 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3747 63. The daemon start-up calls getloadavg() while still root for those OS that
3748 need the first call to be done as root, but it missed one case: when
3749 deliver_queue_load_max is set with deliver_drop_privilege. This is
3750 necessary for the benefit of the queue runner, because there is no re-exec
3751 when deliver_drop_privilege is set.
3753 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3754 This has been fixed.
3756 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3757 was in use, was not putting the data itself into the right store pool;
3758 consequently, it could be overwritten for a subsequent message in the same
3759 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3762 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3764 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3765 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3766 the function were treating the return as a boolean value, which happened to
3767 work because 0=false and not-0=true, but is not correct code.
3769 68. The host_aton() function was not handling scoped IPv6 addresses (those
3770 with, for example, "%eth0" on the end) correctly.
3772 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3773 negated items (that is, ~something) in unsigned ints. Some compilers
3774 apparently mutter when there is no cast.
3776 70. If an address verification called from an ACL failed, and did not produce a
3777 user-specific message (i.e. there was only a "system" message), nothing was
3778 put in $acl_verify_message. In this situation, it now puts the system
3781 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3782 session; change 4.31/43 added the unwanted input to the log line - except
3783 that it did not do this in the start of session case. It now does.
3785 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3786 This is wrong and can cause the other end to generate a synchronization
3787 error if it is another Exim or anything else that does the synchronization
3788 check. A QUIT command is no longer sent after a timeout.
3790 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3791 during host lookups.
3793 74. The defer_ok option of callout verification was not working if it was used
3794 when verifying addresses in header lines, that is, for this case:
3796 verify = header_sender/callout=defer_ok
3798 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3799 those file descriptors could be used for SMTP connections. If anything
3800 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3801 could be sent to the SMTP client, causing chaos. The daemon now opens
3802 stdin, stdout, and stderr to /dev/null when it puts itself into the
3805 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3806 log by default. The user can override this with suitable Perl magic.
3808 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3809 the log message when discard triggers, was not working for the DATA ACL or
3810 for the non-SMTP ACL.
3812 78. Error message wording change in sieve.c.
3814 79. If smtp_accept_max_per_host was set, the number of connections could be
3815 restricted to fewer than expected, because the daemon was trying to set up
3816 a new connection before checking whether the processes handling previous
3817 connections had finished. The check for completed processes is now done
3818 earlier. On busy systems, this bug wouldn't be noticed because something
3819 else would have woken the daemon, and it would have reaped the completed
3822 80. If a message was submitted locally by a user whose login name contained one
3823 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3824 It caused a spool format error. I have fixed the spool reading code. A
3825 related problem was that the "from" clause in the Received: line became
3826 illegal because of the space(s). It is now covered by ${quote_local_part.
3828 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3831 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3833 83. Patch from Sieve maintainer to fix unterminated string problem in
3834 "vacation" handling.
3836 84. Some minor changes to the Linux configuration files to help with other
3837 OS variants using glibc.
3839 85. One more patch for Sieve to update vacation handling to latest spec.
3842 ----------------------------------------------------
3843 See the note above about the 4.44 and 4.50 releases.
3844 ----------------------------------------------------
3850 1. Change 4.43/35 introduced a bug that caused file counts to be
3851 incorrectly computed when quota_filecount was set in an appendfile
3854 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3855 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3858 3. Give more explanation in the error message when the command for a transport
3859 filter fails to execute.
3861 4. There are several places where Exim runs a non-Exim command in a
3862 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3863 was being done only for the command run by the queryprogram router. It is
3864 now done for all such subprocesses. The other cases are: ${run, transport
3865 filters, and the commands run by the lmtp and pipe transports.
3867 5. Some older OS have a limit of 256 on the maximum number of file
3868 descriptors. Exim was using setrlimit() to set 1000 as a large value
3869 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3870 systems. I've change it so that if it can't get 1000, it tries for 256.
3872 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3873 was an oversight, and furthermore, ever since the addition of extra
3874 controls (e.g. 4.43/32), the checks on when to allow different forms of
3875 "control" were broken. There should now be diagnostics for all cases when a
3876 control that does not make sense is encountered.
3878 7. $recipients is now available in the predata ACL (oversight).
3880 8. Tidy the search cache before the fork to do a delivery from a message
3881 received from the command line. Otherwise the child will trigger a lookup
3882 failure and thereby defer the delivery if it tries to use (for example) a
3883 cached ldap connection that the parent has called unbind on.
3885 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3886 of $address_data from the recipient verification was clobbered by the
3887 sender verification.
3889 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3890 was its contents. (It was OK if the option was not defined at all.)
3892 11. A "Completed" log line is now written for messages that are removed from
3893 the spool by the -Mrm option.
3895 12. $host_address is now set to the target address during the checking of
3896 ignore_target_hosts.
3898 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3899 being passed; this would have caused $sender_host_name to have been used if
3900 matching the list had actually called for a host name (not very likely,
3901 since this list is usually IP addresses). A host name is now passed as
3904 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3905 code that specifies a non-restarting handler (typically sigaction() in
3906 modern systems) in an attempt to fix a rare and obscure crash bug.
3908 15. Narrowed the window for a race in the daemon that could cause it to ignore
3909 SIGCHLD signals. This is not a major problem, because they are used only to
3910 wake it up if nothing else does.
3912 16. A malformed maildirsize file could cause Exim to calculate negative values
3913 for the mailbox size or file count. Odd effects could occur as a result.
3914 The maildirsize information is now recalculated if the size or filecount
3917 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3918 support for a long time. Removed HAVE_SYS_VFS_H.
3920 18. Updated exipick to current release from John Jetmore.
3922 19. Allow an empty sender to be matched against a lookup in an address list.
3923 Previously the only cases considered were a regular expression, or an
3926 20. Exim went into a mad DNS lookup loop when doing a callout where the
3927 host was specified on the transport, if the DNS lookup yielded more than
3930 21. The RFC2047 encoding function was originally intended for short strings
3931 such as real names; it was not keeping to the 75-character limit for
3932 encoded words that the RFC imposes. It now respects the limit, and
3933 generates multiple encoded words if necessary. To be on the safe side, I
3934 have increased the buffer size for the ${rfc2047: expansion operator from
3937 22. Failure to deliver a bounce message always caused it to be frozen, even if
3938 there was an errors_to setting on the router. The errors_to setting is now
3941 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3942 canonical form (fully expanded) before being placed in
3943 $sender_host_address.
3945 24. Updated eximstats to version 1.33
3947 25. Include certificate and key file names in error message when GnuTLS fails
3948 to set them up, because the GnuTLS error message doesn't include the name
3949 of the failing file when there is a problem reading it.
3951 26. Expand error message when OpenSSL has problems setting up cert/key files.
3954 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3955 (this can affect the format of dates).
3957 28. exim_tidydb, when checking for the continued existence of a message for
3958 which it has found a message-specific retry record, was not finding
3959 messages that were in split spool directories. Consequently, it was
3960 deleting retry records that should have stayed in existence.
3962 29. eximstats updated to version 1.35
3963 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3964 1.35 - bugfix such that pie charts by volume are generated correctly
3966 30. The SPA authentication driver was not abandoning authentication and moving
3967 on to the next authenticator when an expansion was forced to fail,
3968 contradicting the general specification for all authenticators. Instead it
3969 was generating a temporary error. It now behaves as specified.
3971 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3972 (the order specifies the preference for clients). The order is now AES256,
3973 AES128, 3DES, ARCFOUR128.
3975 31. Small patch to Sieve code - explicitly set From: when generating an
3978 32. Exim crashed if a remote delivery caused a very long error message to be
3979 recorded - for instance if somebody sent an entire SpamAssassin report back
3980 as a large number of 550 error lines. This bug was coincidentally fixed by
3981 increasing the size of one of Exim's internal buffers (big_buffer) that
3982 happened as part of the Exiscan merge. However, to be on the safe side, I
3983 have made the code more robust (and fixed the comments that describe what
3986 33. Some experimental protocols are using DNS PTR records for new purposes. The
3987 keys for these records are domain names, not reversed IP addresses. The
3988 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3989 leaves it alone. Component reversal etc. now happens only for IP addresses.
3992 34. The host_aton() function is supposed to be passed a string that is known
3993 to be a valid IP address. However, in the case of IPv6 addresses, it was
3994 not checking this. This is a hostage to fortune. Exim now panics and dies
3995 if the condition is not met. A case was found where this could be provoked
3996 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3997 components; fortuitously, this particular loophole had already been fixed
3998 by change 4.50/55 or 4.44/33 above.
4000 If there are any other similar loopholes, the new check in host_aton()
4001 itself should stop them being exploited. The report I received stated that
4002 data on the command line could provoke the exploit when Exim was running as
4003 exim, but did not say which command line option was involved. All I could
4004 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4005 running as the user.
4008 35. There was a buffer overflow vulnerability in the SPA authentication code
4009 (which came originally from the Samba project). I have added a test to the
4010 spa_base64_to_bits() function which I hope fixes it.
4013 36. The daemon start-up calls getloadavg() while still root for those OS that
4014 need the first call to be done as root, but it missed one case: when
4015 deliver_queue_load_max is set with deliver_drop_privilege. This is
4016 necessary for the benefit of the queue runner, because there is no re-exec
4017 when deliver_drop_privilege is set.
4019 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4020 was in use, was not putting the data itself into the right store pool;
4021 consequently, it could be overwritten for a subsequent message in the same
4022 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4025 38. Sometimes the final signoff response after QUIT could fail to get
4026 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4027 < 0 before doing a fflush(). This bug looks as though it goes back to the
4028 introduction of TLS in release 3.20, but "sometimes" must have been rare
4029 because the tests only now provoked it.
4035 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4036 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4037 Now it returns an integer. A number of calls were still expecting a T/F
4038 return. Fortuitously, in all cases, the tests worked in OK situations,
4039 which is the norm. However, things would have gone wrong on any write
4040 failures on the smtp file descriptor. This function is used when sending
4041 messages over SMTP and also when doing verify callouts.
4043 2. When Exim is called to do synchronous delivery of a locally submitted
4044 message (the -odf or -odi options), it no longer closes stderr before doing
4047 3. Implemented the mua_wrapper option.
4049 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4051 5. Implemented the functions header_remove(), header_testname(),
4052 header_add_at_position(), and receive_remove_recipient(), and exported them
4055 6. If an ACL "warn" statement specified the addition of headers, Exim already
4056 inserted X-ACL-Warn: at the start if there was no header name. However, it
4057 was not making this test for the second and subsequent header lines if
4058 there were newlines in the string. This meant that an invalid header could
4059 be inserted if Exim was badly configured.
4061 7. Allow an ACL "warn" statement to add header lines at the start or after all
4062 the Received: headers, as well as at the end.
4064 8. Added the rcpt_4xx retry error code.
4066 9. Added postmaster_mailfrom=xxx to callout verification option.
4068 10. Added mailfrom=xxxx to the callout verification option, for verify=
4071 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4072 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4074 12. Inserted some casts to stop certain compilers warning when using pointer
4075 differences as field lengths or precisions in printf-type calls (mostly
4076 affecting debugging statements).
4078 13. Added optional readline() support for -be (dynamically loaded).
4080 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4081 same clock tick as a message's arrival, so that its received time was the
4082 same as the "first fail" time on the retry record, and that message
4083 remained on the queue past the ultimate address timeout, every queue runner
4084 would try a delivery (because it was past the ultimate address timeout) but
4085 after another failure, the ultimate address timeout, which should have then
4086 bounced the address, did not kick in. This was a "< instead of <=" error;
4087 in most cases the first failure would have been in the next clock tick
4088 after the received time, and all would be well.
4090 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4091 being recognized when the domain list was tested by the match_domain
4092 condition in an expansion string.
4094 16. Added the ${str2b64: operator.
4096 17. Exim was always calling setrlimit() to set a large limit for the number of
4097 processes, without checking whether the existing limit was already
4098 adequate. (It did check for the limit on file descriptors.) Furthermore,
4099 errors from getrlimit() and setrlimit() were being ignored. Now they are
4100 logged to the main and panic logs, but Exim does carry on, to try to do its
4101 job under whatever limits there are.
4103 18. Imported PCRE 5.0.
4105 19. Trivial typo in log message " temporarily refused connection" (the leading
4108 20. If the log selector return_path_on_delivery was set and an address was
4109 redirected to /dev/null, the delivery process crashed because it assumed
4110 that a return path would always be set for a "successful" delivery. In this
4111 case, the whole delivery is bypassed as an optimization, and therefore no
4114 21. Internal re-arrangement: the function for sending a challenge and reading
4115 a response while authentication was assuming a zero-terminated challenge
4116 string. It's now changed to take a pointer and a length, to allow for
4117 binary data in such strings.
4119 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4121 23. Exim was not respecting finduser_retries when seeking the login of the
4122 uid under which it was called; it was always trying 10 times. (The default
4123 setting of finduser_retries is zero.) Also, it was sleeping after the final
4124 failure, which is pointless.
4126 24. Implemented tls_on_connect_ports.
4128 25. Implemented acl_smtp_predata.
4130 26. If the domain in control=submission is set empty, Exim assumes that the
4131 authenticated id is a complete email address when it generates From: or
4132 Sender: header lines.
4134 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4135 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4136 chown and chgrp in /bin and hostname in /usr/bin.
4138 28. Exim was keeping the "process log" file open after each use, just as it
4139 does for the main log. This opens the possibility of it remaining open for
4140 long periods when the USR1 signal hits a daemon. Occasional processlog
4141 errors were reported, that could have been caused by this. Anyway, it seems
4142 much more sensible not to leave this file open at all, so that is what now
4145 29. The long-running daemon process does not normally write to the log once it
4146 has entered its main loop, and it closes the log before doing so. This is
4147 so that log files can straightforwardly be renamed and moved. However,
4148 there are a couple of unusual error situations where the daemon does write
4149 log entries, and I had neglected to close the log afterwards.
4151 30. The text of an SMTP error response that was received during a remote
4152 delivery was being truncated at 512 bytes. This is too short for some of
4153 the long messages that one sometimes sees. I've increased the limit to
4156 31. It is now possible to make retry rules that apply only when a message has a
4157 specific sender, in particular, an empty sender.
4159 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4160 it possible to be selective about when SMTP synchronization is enforced.
4162 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4164 32. Implemented hosts_connection_nolog.
4166 33. Added an ACL for QUIT.
4168 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4171 35. Added mailbox_size and mailbox_filecount to appendfile.
4173 36. Added control = no_multiline_responses to ACLs.
4175 37. There was a bug in the logic of the code that waits for the clock to tick
4176 in the case where the clock went backwards by a substantial amount such
4177 that the microsecond fraction of "now" was more than the microsecond
4178 fraction of "then" (but the whole seconds number was less).
4180 38. Added support for the libradius Radius client library this is found on
4181 FreeBSD (previously only the radiusclient library was supported).
4187 1. When certain lookups returned multiple values in the form name=value, the
4188 quoting of the values was not always being done properly. Specifically:
4189 (a) If the value started with a double quote, but contained no whitespace,
4191 (b) If the value contained whitespace other than a space character (i.e.
4192 tabs or newlines or carriage returns) it was not quoted.
4193 This fix has been applied to the mysql and pgsql lookups by writing a
4194 separate quoting function and calling it from the lookup code. The fix
4195 should probably also be applied to nisplus, ibase and oracle lookups, but
4196 since I cannot test any of those, I have not disturbed their existing code.
4198 2. A hit in the callout cache for a specific address caused a log line with no
4199 reason for rejecting RCPT. Now it says "Previous (cached) callout
4200 verification failure".
4202 3. There was an off-by-one bug in the queryprogram router. An over-long
4203 return line was truncated at 256 instead of 255 characters, thereby
4204 overflowing its buffer with the terminating zero. As well as fixing this, I
4205 have increased the buffer size to 1024 (and made a note to document this).
4207 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4208 when Exim is waiting for an SMTP response from a remote server, Exim
4209 restarts its select() call on the socket, thereby resetting its timeout.
4210 This is not a problem when such interrupts are rare. Somebody set up a cron
4211 job to run exiwhat every 2 minutes, which is less than the normal select()
4212 timeout (5 or 10 minutes). This meant that the select() timeout never
4213 kicked in because it was always reset. I have fixed this by comparing the
4214 time when an interrupt arrives with the time at the start of the first call
4215 to select(). If more time than the timeout has elapsed, the interrupt is
4216 treated as a timeout.
4218 5. Some internal re-factoring in preparation for the addition of Sieve
4219 extensions (by MH). In particular, the "personal" test is moved to a
4220 separate function, and given an option for scanning Cc: and Bcc: (which is
4221 not set for Exim filters).
4223 6. When Exim created an email address using the login of the caller as the
4224 local part (e.g. when creating a From: or Sender: header line), it was not
4225 quoting the local part when it contained special characters such as @.
4227 7. Installed new OpenBSD configuration files.
4229 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4230 try to make them clearer.
4232 9. Callout options, other than the timeout value, were being ignored when
4233 verifying sender addresses in header lines. For example, when using
4235 verify = header_sender/callout=no_cache
4237 the cache was (incorrectly) being used.
4239 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4240 only the Cygwin environment.
4242 11. When return_path_on_delivery was set as a log selector, if different remote
4243 addresses in the same message used different return paths and parallel
4244 remote delivery occurred, the wrong values would sometimes be logged.
4245 (Whenever a remote delivery process finished, the return path value from
4246 the most recently started remote delivery process was logged.)
4248 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4249 lines when AUTH and/or TLS are in use. This is the "received protocol"
4250 field. Exim used to use "asmtp" for authenticated SMTP, without any
4251 indication (in the protocol name) for TLS use. Now it follows the RFC and
4252 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4253 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4254 names appear in log lines as well as in Received: header lines.
4256 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4257 extensions, and comparison tests, and to fix some bugs.
4259 14. Changes to the "personal" filter test:
4261 (1) The test was buggy in that it was just doing the equivalent of
4262 "contains" tests on header lines. For example, if a user's address was
4263 anne@some.where, the "personal" test would incorrectly be true for
4265 To: susanne@some.where
4267 This test is now done by extracting each address from the header in turn,
4268 and checking the entire address. Other tests that are part of "personal"
4269 are now done using regular expressions (for example, to check local parts
4270 of addresses in From: header lines).
4272 (2) The list of non-personal local parts in From: addresses has been
4273 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4274 taken from the Sieve specification recommendations.
4276 (3) If the message contains any header line starting with "List-" it is
4277 treated as non-personal.
4279 (4) The test for "circular" in the Subject: header line has been removed
4280 because it now seems ill-conceived.
4282 15. Minor typos in src/EDITME comments corrected.
4284 16. Installed latest exipick from John Jetmore.
4286 17. If headers_add on a router specified a text string that was too long for
4287 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4288 of string_sprintf() is now avoided.
4290 18. $message_body_size was not set (it was always zero) when running the DATA
4291 ACL and the local_scan() function.
4293 19. For the "mail" command in an Exim filter, no default was being set for
4294 the once_repeat time, causing a random time value to be used if "once" was
4295 specified. (If the value happened to be <= 0, no repeat happened.) The
4296 default is now 0s, meaning "never repeat". The "vacation" command was OK
4297 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4298 (I found it when inspecting the code).
4300 20. There is now an overall timeout for performing a callout verification. It
4301 defaults to 4 times the callout timeout, which applies to individual SMTP
4302 commands during the callout. The overall timeout applies when there is more
4303 than one host that can be tried. The timeout is checked before trying the
4304 next host. This prevents very long delays if there are a large number of
4305 hosts and all are timing out (e.g. when the network connections are timing
4306 out). The value of the overall timeout can be changed by specifying an
4307 additional sub-option for "callout", called "maxwait". For example:
4309 verify = sender/callout=5s,maxwait=20s
4311 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4312 to the end before writing, but this should make it even safer).
4314 22. Exim was forgetting that it had advertised PIPELINING for the second and
4315 subsequent messages on an SMTP connection. It was also not resetting its
4316 memory on STARTTLS and an internal HELO.
4318 23. When Exim logs an SMTP synchronization error within a session, it now
4319 records whether PIPELINING has been advertised or not.
4321 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4322 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4323 rather than long int.
4325 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4327 26. Added the never_mail option to autoreply.
4333 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4334 crash if the getsockname() call failed; this can happen if a connection is
4335 closed very soon after it is established. The problem was simply in the
4336 order in which certain operations were done, causing Exim to try to write
4337 to the SMTP stream before it had set up the file descriptor. The bug has
4338 been fixed by making things happen in the correct order.
4344 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4345 before the connection was closed, thus losing the rejection response.
4347 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4348 some early Solaris releases, but causes trouble in current releases where
4349 socklen_t is defined.
4351 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4354 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4357 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4358 that releasing the top of it at the end releases what was used for sub-
4359 expansions (unless the block got too big). However, discard this block if
4360 the first thing is a variable or header, so that we can use its block when
4361 it is dynamic (useful for very large $message_headers, for example).
4363 6. Lookups now cache *every* query, not just the most recent. A new, separate
4364 store pool is used for this. It can be recovered when all lookup caches are
4365 flushed. Lookups now release memory at the end of their result strings.
4366 This has involved some general refactoring of the lookup sources.
4368 7. Some code has been added to the store_xxx() functions to reduce the amount
4369 of flapping under certain conditions.
4371 8. log_incoming_interface used to affect only the <= reception log lines. Now
4372 it causes the local interface and port to be added to several more SMTP log
4373 lines, for example "SMTP connection from", and rejection lines.
4375 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4377 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4379 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4380 could be overwritten at the end of the current message (or the start of a
4381 new message if it was set in a HELO ACL). The value is now preserved for
4382 the duration of the SMTP connection.
4384 12. If a transport had a headers_rewrite setting, and a matching header line
4385 contained an unqualified address, that address was qualified, even if it
4386 did not match any rewriting rules. The underlying bug was that the values
4387 of the flags that permit the existence of unqualified sender and recipient
4388 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4389 non-local messages, and by -bnq for local messages) were not being
4390 preserved with the message after it was received.
4392 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4393 "next input=" as part of the text comprising the host identity instead of
4394 the correct text. The code was using the same buffer for two different
4395 strings. However, depending on which order the printing function evaluated
4396 its arguments, the bug did not always show up. Under Linux, for example, my
4397 test suite worked just fine.
4399 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4400 doesn't work with some older versions of Perl. It has been changed to "my",
4401 which in any case is probably the better facility to use.
4403 15. A really picky compiler found some instances of statements for creating
4404 error messages that either had too many or two few arguments for the format
4407 16. The size of the buffer for calls to the DNS resolver has been increased
4408 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4409 for addresses that have a lot of PTR records. This alleviates a problem; it
4410 does not fully solve it.
4412 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4413 buffer now truncates the list and logs the incident, which is the same
4414 action as happens when Exim is looking up a host name and its aliases.
4415 Previously in this situation something unpredictable would happen;
4416 sometimes it was "internal error: store_reset failed".
4418 18. If a server dropped the connection unexpectedly when an Exim client was
4419 using GnuTLS and trying to read a response, the client delivery process
4420 crashed while trying to generate an error log message.
4422 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4423 string, for example:
4425 warn message = H1: something\nH2: something
4427 the text was added as a single header line from Exim's point of view
4428 though it ended up OK in the delivered message. However, searching for the
4429 second and subsequent header lines using $h_h2: did not work. This has been
4430 fixed. Similarly, if a system filter added multiple headers in this way,
4431 the routers could not see them.
4433 20. Expanded the error message when iplsearch is called with an invalid key to
4434 suggest using net-iplsearch in a host list.
4436 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4437 ACLs are no longer actually imposed (and a message to that effect is
4440 22. If a "gecos" field in a passwd entry contained escaped characters, in
4441 particular, if it contained a \" sequence, Exim got it wrong when building
4442 a From: or a Sender: header from that name. A second bug also caused
4443 incorrect handling when an unquoted " was present following a character
4444 that needed quoting.
4446 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4447 was not being matched caselessly.
4449 24. Arranged for all hyphens in the exim.8 source to be escaped with
4452 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4453 properly. Recipient callout cache records were still being keyed to include
4454 the sender, even when use_sender was set false. This led to far more
4455 callouts that were necessary. The sender is no longer included in the key
4456 when use_sender is false.
4458 26. Added "control = submission" modifier to ACLs.
4460 27. Added the ${base62d: operator to decode base 62 numbers.
4462 28. dnsdb lookups can now access SRV records.
4464 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4465 the configuration file.
4467 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4468 (-v) mode. This makes the output for a verbose queue run more intelligible.
4470 31. Added a use_postmaster feature to recipient callouts.
4472 32. Added the $body_zerocount variable, containing the number of binary zero
4473 bytes in the message body.
4475 33. The time of last modification of the "new" subdirectory is now used as the
4476 "mailbox time last read" when there is a quota error for a maildir
4479 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4481 35. Added +ignore_unknown as a special item in host lists.
4483 36. Code for decoding IPv6 addresses in host lists is now included, even if
4484 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4485 address was recognized as an IP address, but was then not correctly decoded
4486 into binary, causing unexpected and incorrect effects when compared with
4493 1. Very minor rewording of debugging text in manualroute to say "list of
4494 hosts" instead of "hostlist".
4496 2. If verify=header_syntax was set, and a header line with an unqualified
4497 address (no domain) and a large number of spaces between the end of the
4498 name and the colon was received, the reception process suffered a buffer
4499 overflow, and (when I tested it) crashed. This was caused by some obsolete
4500 code that should have been removed. The fix is to remove it!
4502 3. When running in the test harness, delay a bit after writing a bounce
4503 message to get a bit more predictability in the log output.
4505 4. Added a call to search_tidyup() just before forking a reception process. In
4506 theory, someone could use a lookup in the expansion of smtp_accept_max_
4507 per_host which, without the tidyup, could leave open a database connection.
4509 5. Added the variables $recipient_data and $sender_data which get set from a
4510 lookup success in an ACL "recipients" or "senders" condition, or a router
4511 "senders" option, similar to $domain_data and $local_part_data.
4513 6. Moved the writing of debug_print from before to after the "senders" test
4516 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4517 problems for message scanning, either using a data ACL, or using
4518 local_scan() because the Received: header was not generated till after they
4519 were called (in order to set the time as the time of reception completion).
4520 I have revised the way this works. The header is now generated after the
4521 body is received, but before the ACL or local_scan() are called. After they
4522 are run, the timestamp in the header is updated.
4528 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4529 before starting a queue runner without re-exec. This happened only when
4530 deliver_drop_privilege was set or when the Exim user was set to root. The
4531 effect of the bug was that timeouts during subsequent deliveries caused
4532 crashes instead of being properly handled. The handler is now left at its
4533 default (and expected) setting.
4535 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4536 message, again when deliver_drop_privilege is set or Exim is run as root.
4537 The bug described in (1) was not present in this case, but the tidying up
4538 of the other signals was missing. I have made the two cases consistent.
4540 3. The ignore_target_hosts setting on a manualroute router was being ignored
4541 for hosts that were looked up using the /MX notation.
4543 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4546 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4547 operated on the sender address. After changing the $sender_address to <>
4548 for the sender address verify, Exim was re-instated it as the original
4549 (before rewriting) address, but remembering that it had rewritten it, so it
4550 wasn't rewriting it again. This bug also had the effect of breaking the
4551 sender address verification caching when the sender address was rewritten.
4553 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4554 This has been changed so that if the ip literal address matches
4555 ignore_target_hosts, the router declines.
4557 7. Added expansion conditions match_domain, match_address, and match_local_
4558 part (NOT match_host).
4560 8. The placeholder for the Received: header didn't have a length field set.
4562 9. Added code to Exim itself and to exim_lock to test for a specific race
4563 condition that could lead to file corruption when using MBX delivery. The
4564 issue is with the lockfile that is created in /tmp. If this file is removed
4565 after a process has opened it but before that process has acquired a lock,
4566 there is the potential for a second process to recreate the file and also
4567 acquire a lock. This could lead to two Exim processes writing to the file
4568 at the same time. The added code performs the same test as UW imapd; it
4569 checks after acquiring the lock that its file descriptor still refers to
4570 the same named file.
4572 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4573 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4576 11. Added the smtp_active_hostname option. If used, this will typically be made
4577 to depend on the incoming interface address. Because $interface_address is
4578 not set up until the daemon has forked a reception process, error responses
4579 that can happen earlier (such as "too many connections") no longer contain
4582 12. If an expansion in a condition on a "warn" statement fails because a lookup
4583 defers, the "warn" statement is abandoned, and the next ACL statement is
4584 processed. Previously this caused the whole ACL to be aborted.
4586 13. Added the iplsearch lookup type.
4588 14. Added ident_timeout as a log selector.
4590 15. Added tls_certificate_verified as a log selector.
4592 16. Added a global option tls_require_ciphers (compare the smtp transport
4593 option of the same name). This controls incoming TLS connections.
4595 17. I finally figured out how to make tls_require_ciphers do a similar thing
4596 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4597 before starting the TLS session.
4599 18. Tabs are now shown as \t in -bP output.
4601 19. If the log selector return_path_on_delivery was set, Exim crashed when
4602 bouncing a message because it had too many Received: header lines.
4604 20. If two routers both had headers_remove settings, and the first one included
4605 a superfluous trailing colon, the final name in the first list and the
4606 first name in the second list were incorrectly joined into one item (with a
4607 colon in the middle).
4613 1. Added -C and -D options to the exinext utility, mainly to make it easier
4614 to include in the automated testing, but these could be helpful when
4615 multiple configurations are in use.
4617 2. The exinext utility was not formatting the output nicely when there was
4618 an alternate port involved in the retry record key, nor when there was a
4619 message id as well (for retries that were specific to a specific message
4620 and a specific host). It was also confused by IPv6 addresses, because of
4621 the additional colons they contain. I have fixed the IPv4 problem, and
4622 patched it up to do a reasonable job for IPv6.
4624 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4625 delivery, the log line now contains "pipelined" if PIPELINING was used.
4627 4. An SMTP transport process used to panic and die if the bind() call to set
4628 an explicit outgoing interface failed. This has been changed; it is now
4629 treated in the same way as a connect() failure.
4631 5. A reference to $sender_host_name in the part of a conditional expansion
4632 that was being skipped was still causing a DNS lookup. This no longer
4635 6. The def: expansion condition was not recognizing references to header lines
4636 that used bh_ and bheader_.
4638 7. Added the _cache feature to named lists.
4640 8. The code for checking quota_filecount in the appendfile transport was
4641 allowing one more file than it should have been.
4643 9. For compatibility with Sendmail, the command line option
4651 and sets the incoming protocol and host name (for trusted callers). The
4652 host name and its colon can be omitted when only the protocol is to be set.
4653 Note the Exim already has two private options, -pd and -ps, that refer to
4654 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4655 "s", but I don't think that's a major issue.
4657 10. A number of refactoring changes to the code, none of which should affect
4660 (a) The number of logging options was getting close to filling up the
4661 32-bit word that was used as a bit map. I have split them into two classes:
4662 those that are passed in the argument to log_write(), and those that are
4663 only ever tested independently outside of that function. These are now in
4664 separate 32-bit words, so there is plenty of room for expansion again.
4665 There is no change in the user interface or the logging behaviour.
4667 (b) When building, for example, log lines, the code previously used a
4668 macro that called string_cat() twice, in order to add two strings. This is
4669 not really sufficiently general. Furthermore, there was one instance where
4670 it was actually wrong because one of the argument was used twice, and in
4671 one call a function was used. (As it happened, calling the function twice
4672 did not affect the overall behaviour.) The macro has been replaced by a
4673 function that can join an arbitrary number of extra strings onto a growing
4676 (c) The code for expansion conditions now uses a table and a binary chop
4677 instead of a serial search (which was left over from when there were very
4678 few conditions). Also, it now recognizes conditions like "pam" even when
4679 the relevant support is not compiled in: a suitably worded error message is
4680 given if an attempt is made to use such a condition.
4682 11. Added ${time_interval:xxxxx}.
4684 12. A bug was causing one of the ddress fields not to be passed back correctly
4685 from remote delivery subprocesses. The field in question was not being
4686 subsequently used, so this caused to problems in practice.
4688 13. Added new log selectors queue_time and deliver_time.
4690 14. Might have fixed a bug in maildirsizefile handling that threw up
4691 "unexpected character" debug warnings, and recalculated the data
4692 unnecessarily. In any case, I expanded the warning message to give more
4695 15. Added the message "Restricted characters in address" to the statements in
4696 the default ACL that block characters like @ and % in local parts.
4698 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4699 Three changes have been made:
4701 (a) There was a serious bug; a negative response to MAIL caused the whole
4702 recipient domain to be cached as invalid, thereby blocking all messages
4703 to all local parts at the same domain, from all senders. This bug has
4704 been fixed. The domain is no longer cached after a negative response to
4705 MAIL if the sender used is not empty.
4707 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4710 (c) A new callout option, "use_sender" has been added for people who want
4711 the modified behaviour.
4717 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4720 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4721 indeed breaks things for older releases.
4723 3. Added additional logging to the case where there is a problem reading data
4724 from a filter that is running in a subprocess using a pipe, in order to
4725 try to track down a specific problem.
4727 4. Testing facility fudge: when running in the test harness and attempting
4728 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4729 getting "No route to host". Convert this to a timeout.
4731 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4734 6. Some OS don't have socklen_t but use size_t instead. This affects the
4735 fifth argument of getsockopt() amongst other things. This is now
4736 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4737 can be set for individual OS. I have set it for SunOS5, OSF1, and
4738 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4739 some earlier ones do not.
4741 7. Change 4.30/15 was not doing the test caselessly.
4743 8. The standard form for an IPv6 address literal was being rejected by address
4744 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4745 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4746 this, as well as the form without the "IPv6" on the front (but only when
4747 address literals are enabled, of course).
4749 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4751 10. Exim crashed if a message with an empty sender address specified by -f
4752 encountered a router with an errors_to setting. This could be provoked only
4753 by a command such as
4757 where an empty string was supplied; "<>" did not hit this bug.
4759 11. Installed PCRE release 4.5.
4761 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4762 remained set. It is now erased.
4764 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4765 times from message ids (which are base 36 rather than the normal 62).
4767 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4768 were being counted as actual protocol errors, and logged if the log
4769 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4770 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4771 and DATA following a set of rejected RCPTs do not count as protocol errors.
4772 In other words, Exim assumes they were pipelined, though this may not
4773 actually be the case. Of course, in all cases the client gets an
4774 appropriate error code.
4776 15. If a lookup fails in an ACL condition, a message about the failure may
4777 be available; it is used if testing the ACL cannot continue, because most
4778 such messages specify what the cause of the deferral is. However, some
4779 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4780 that caused an old message to be retained and used if a later statement
4781 caused a defer, replacing the real cause of the deferral.
4783 16. If an IP address had so many PTR records that the DNS lookup buffer
4784 was not large enough to hold them, Exim could crash while trying to process
4785 the truncated data. It now detects and logs this case.
4787 17. Further to 4.21/58, another change has been made: if (and only if) the
4788 first line of a message (the first header line) ends with CRLF, a bare LF
4789 in a subsequent header line has a space inserted after it, so as not to
4790 terminate the header.
4792 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4793 unnecessarily, used atoi() instead of strtol(), and didn't check the
4794 termination when getting file sizes from file names by regex.
4796 19. Completely re-implemented the support for maildirsize files, in the light
4797 of a number of problems with the previous contributed implementation
4798 (4.30/29). In particular:
4800 . If the quota is zero, the maildirsize file is maintained, but no quota is
4803 . If the maildir directory does not exist, it is created before any attempt
4804 to write a maildirsize file.
4806 . The quota value in the file is just a cache; if the quota is changed in
4807 the transport, the new value overrides.
4809 . A regular expression is available for excluding directories from the
4812 20. The autoreply transport checks the characters in options that define the
4813 message's headers; it allows continued headers, but it was checking with
4814 isspace() after an embedded newline instead of explicitly looking for a
4817 21. If all the "regular" hosts to which an address was routed had passed their
4818 expiry times, and had not reached their retry times, the address was
4819 bounced, even if fallback hosts were defined. Now Exim should go on to try
4822 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4823 equivalent code in the SMTP transport. Some hosts send humungous responses
4824 to HELO/EHLO, more than 1024 it seems.
4826 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4827 gives compiler warnings in some environments. I've now done it "properly",
4830 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4831 (because of problems with the built-in one) was declared to return uschar *
4832 instead of char *, causing compiler failure.
4834 25. Fixed a file descriptor leak when processing alias/forward files.
4836 26. Fixed a minor format string issue in dbfn.c.
4838 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4840 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4841 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4844 29. When a sender address is verified, it is cached, to save repeating the test
4845 when there is more than one recipient in a message. However, when the
4846 verification involves a callout, it is possible for different callout
4847 options to be set for different recipients. It is too complicated to keep
4848 track of this in the cache, so now Exim always runs a verification when a
4849 callout is required, relying on the callout cache for the optimization.
4850 The overhead is duplication of the address routing, but this should not be
4853 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4854 to be verified with callout=postmaster, and the main callout worked but the
4855 postmaster check failed, the verification correctly failed. However, if a
4856 subsequent RCPT command asked for sender verification *without* the
4857 postmaster check, incorrect caching caused this verification also to fail,
4860 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4861 it was not caching the DNS options (qualify_single, search_parents) that
4862 were used when the lookup failed. A subsequent lookup with different
4863 options therefore always gave the same answer, though there were cases
4864 where it should not have. (Example: a "domains = !$mx_any" option on a
4865 dnslookup router: the "domains" option is always processed without any
4866 widening, but the router might have qualify_single set.) Now Exim uses the
4867 cached value only when the same options are set.
4869 32. Added John Jetmore's "exipick" utility to the distribution.
4871 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4872 than a timeout (e.g. a certificate is required, and is not provided), an
4873 Exim server now closes the connection immediately. Previously it waited for
4874 the client to close - but if the client is SSL, it seems that they each
4875 wait for each other, leading to a delay before one of them times out.
4877 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4878 maintained 0.8.x compatibility because I don't think many are using it, and
4879 it is clearly obsolete.
4881 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4884 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4885 client certificate was expired. A simple patch fixes this, though I don't
4886 understand the full logic of why the verify callback is called multiple
4889 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4890 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4891 which causes problems with some clients (such as the Certicom SSL Plus
4892 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4893 disables the coutermeasure allowing Eudora to connect."
4895 38. Exim was not checking that a write() to a log file succeeded. This could
4896 lead to Bad Things if a log got too big, in particular if it hit a file
4897 size limit. Exim now panics and dies if it cannot write to a log file, just
4898 as it does if it cannot open a log file.
4900 39. Modified OS/Makefile-Linux so that it now contains
4902 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4904 The two -D definitions ensure that Exim is compiled with large file
4905 support, which makes it possible to handle log files that are bigger than
4908 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4909 instance) a domain was checked against a named list that involved a lookup,
4910 causing $domain_data to be set, then another domain was checked against the
4911 same list, then the first domain was re-checked, the value of $domain_data
4912 after the final check could be wrong. In particular, if the second check
4913 failed, it could be set empty. This bug probably also applied to
4916 41. The strip_trailing_dot option was not being applied to the address given
4917 with the -f command-line option.
4919 42. The code for reading a message's header from the spool was incrementing
4920 $received_count, but never initializing it. This meant that the value was
4921 incorrect (doubled) while delivering a message in the same process in which
4922 it was received. In the most common configuration of Exim, this never
4923 happens - a fresh exec is done - but it can happen when
4924 deliver_drop_privilege is set.
4926 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4927 it now includes up to 150 characters of the unexpected data in the log
4930 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4931 and building data strings. The size of both of these buffers was 10 000
4932 bytes - far larger than anybody would *ever* want, thought I. Needless to
4933 say, somebody hit the limit. I have increased the maximum line length to
4934 20 000 and the maximum data length of concatenated lines to 100 000. I have
4935 also fixed two bugs, because there was no checking on these buffers. Tsk,
4936 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4937 buffer is too small.
4939 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4940 lsearch lookups. Now it does.
4942 46. When parsing a route_list item in a manualroute router, a fixed-length
4943 buffer was used for the list of hosts. I made this 1024 bytes long,
4944 thinking that nobody would ever have a list of hosts that long. Wrong.
4945 Somebody had a whole pile of complicated expansion conditions, and the
4946 string was silently truncated, leading to an expansion error. It turns out
4947 that it is easier to change to an unlimited length (owing to other changes
4948 that have happened since this code was originally written) than to build
4949 structure for giving a limitation error. The length of the item that
4950 expands into the list of hosts is now unlimited.
4952 47. The lsearch lookup could not handle data where the length of text line was
4953 more than 4095 characters. Such lines were truncated, leading to shortened
4954 data being returned. It should now handle lines of any length.
4956 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4957 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4960 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4961 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4962 no longer happens in comment lines. A list of replacements is now placed
4963 at the head of all of the source files, except those whose only change is
4964 to replace PERL_COMMAND in the very first #! line.
4966 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4967 messages on the queue, with a bottom-up merge sort, using code contributed
4968 by Michael Haardt. This should make operations like -bp somewhat faster on
4969 large queues. It won't affect queue runners, except when queue_run_in_order
4972 51. Installed eximstats 1.31 in the distribution.
4974 52. Added support for SRV lookups to the dnslookup router.
4976 53. If an ACL referred to $message_body or $message_body_end, the value was not
4977 reset for any messages that followed in the same SMTP session.
4979 54. The store-handling optimization for building very long strings was not
4980 differentiating between the different store pools. I don't think this
4981 actually made any difference in practice, but I've tidied it.
4983 55. While running the routers to verify a sender address, $sender_address
4984 was still set to the sender address. This is wrong, because when routing to
4985 send a bounce to the sender, it would be empty. Therefore, I have changed
4986 it so that, while verifying a sender address, $sender_address is set to <>.
4987 (There is no change to what happens when verifying a recipient address.)
4989 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4990 target A or AAAA records (if not already returned) without resetting the
4991 qualify_single or search_parents options of the DNS resolver. These are
4992 inappropriate in this case because the targets of MX and SRV records must
4993 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4994 target that, when qualified, matched something in the local domain. These
4995 two options are now turned off when doing these lookups.
4997 57. It seems that at least some releases of Reiserfs (which does not have the
4998 concept of a fixed number of inodes) returns zero and not -1 for the
4999 number of available inodes. This interacted badly with check_spool_inodes,
5000 which assumed that -1 was the "no such thing" setting. What I have done is
5001 to check that the total number of inodes is greater than zero before doing
5002 the test of how many are available.
5004 58. When a "warn" ACL statement has a log_message modifier, the message is
5005 remembered, and not repeated. This is to avoid a lot of repetition when a
5006 message has many recipients that cause the same warning to be written.
5007 Howewer, Exim was preserving the list of already written lines for an
5008 entire SMTP session, which doesn't seem right. The memory is now reset if a
5009 new message is started.
5011 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5012 debugging output unless log_rewrite was also set.
5014 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5015 of (int)(handle) when we know that handle contains (void *)(-1).
5017 61. The Exim daemon panic-logs an error return when it closes the incoming
5018 connection. However "connection reset by peer" seems to be common, and
5019 isn't really an error worthy of noting specially, so that particular error
5022 62. When Exim is trying to find all the local interfaces, it used to panic and
5023 die if the ioctl to get the interface flags failed. However, it seems that
5024 on at least one OS (Solaris 9) it is possible to have an interface that is
5025 included in the list of interfaces, but for which you get a failure error
5026 for this call. This happens when the interface is not "plumbed" into a
5027 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5028 failure of the "get flags" call assumes that the interface is down.
5030 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5031 makes life easier for people who are doing arithmetic on fields extracted
5032 from dates, where you often get leading zeros that should not be
5033 interpreted as octal.
5035 64. Added qualify_domain to the redirect router, to override the global
5038 65. If a pathologically long header line contained very many addresses (the
5039 report of this problem mentioned 10 000) and each of them was rewritten,
5040 Exim could use up a very large amount of memory. (It kept on making new
5041 copies of the header line as it rewrote, and never released the old ones.)
5042 At the expense of a bit more processing, the header rewriting function has
5043 been changed so that it no longer eats memory in this way.
5045 66. The generation of the Received: header has been moved from the time that a
5046 message starts to be received, to the time that it finishes. The timestamp
5047 in the Received: header should now be very close to that of the <= log
5048 line. There are two side-effects of this change:
5050 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5051 logged header lines no longer include the local Received: line, because
5052 it has not yet been created. The same applies to a copy of the message
5053 that is returned to a non-SMTP sender when a message is rejected.
5055 (b) When a filter file is tested using -bf, no additional Received: header
5056 is added to the test message. After some thought, I decided that this
5059 This change does not affect the value of $received_for. It is still set
5060 after address rewriting, but before local_scan() is called.
5062 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5064 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5065 gave an unhelpful panic error message, and a defer error. I have managed to
5066 change this behaviour so that it now rejects any supplied certificate,
5067 which seems right, as the list of acceptable certificates is empty.
5069 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5070 gave an unhelpful defer error. I have not managed to make this reject any
5071 supplied certificates, but the error message it gives is "no certificate
5072 supplied", which is not helpful.
5074 70. exigrep's output now also includes lines that are not associated with any
5075 message, but which match the given pattern. Implemented by a patch from
5076 Martin Sluka, which also tidied up the Perl a bit.
5078 71. Recipient callout verification, like sender verification, was using <> in
5079 the MAIL FROM command. This isn't really the right thing, since the actual
5080 sender may affect whether the remote host accepts the recipient or not. I
5081 have changed it to use the actual sender in the callout; this means that
5082 the cache record is now keyed on a recipient/sender pair, not just the
5083 recipient address. There doesn't seem to be a real danger of callout loops,
5084 since a callout by the remote host to check the sender would use <>.
5085 [SEE ABOVE: changed after hitting problems.]
5087 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5088 temporary errors. However, in the case of such a code being given after
5089 the end of a data transmission (i.e. after ".") Exim was failing to write
5090 a retry record for the message. (Yes, there was some broken host that was
5091 actually sending 8xx at this point.)
5093 73. An unknown lookup type in a host list could cause Exim to panic-die when
5094 the list was checked. (An example that provoked this was putting <; in the
5095 middle of a list instead of at the start.) If this happened during a DATA
5096 ACL check, a -D file could be left lying around. This kind of configuration
5097 error no longer causes Exim to die; instead it causes a defer errror. The
5098 incident is still logged to the main and panic logs.
5100 74. Buglet left over from Exim 3 conversion. The message "too many messages
5101 in one connection" was written to the rejectlog but not the mainlog, except
5102 when address rewriting (yes!) was being logged.
5104 75. Added write_rejectlog option.
5106 76. When a system filter was run not as root (that is, when system_filter_user
5107 was set), the values of the $n variables were not being returned to the
5108 main process; thus, they were not subsequently available in the $sn
5111 77. Added +return_path_on_delivery log selector.
5113 78. A connection timeout was being treated differently from recipients deferred
5114 when testing hosts_max_try with a message that was older than the host's
5115 retry timeout. (The host should not be counted, thus allowing all hosts to
5116 be tried at least once before bouncing.) This may have been the cause of an
5117 occasionally reported bug whereby a message would remain on the queue
5118 longer than the retry timeout, but would be bounced if a delivery was
5119 forced. I say "may" because I never totally pinned down the problem;
5120 setting up timeout/retry tests is difficult. See also the next item.
5122 79. The ultimate address timeout was not being applied to errors that involved
5123 a combination of host plus message (for example, a timeout on a MAIL
5124 command). When an address resolved to a number of possible hosts, and they
5125 were not all tried for each delivery (e.g. because of hosts_max_try), a
5126 message could remain on the queue longer than the retry timeout.
5128 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5131 81. Fixed an obscure SMTP outgoing bug which required at least the following
5132 conditions: (a) there was another message waiting for the same server;
5133 (b) the server returned 5xx to all RCPT commands in the first message so
5134 that the message was not completed; (c) the server dropped the connection
5135 or gave a negative response to the RSET that Exim sends to abort the
5136 transaction. The observed case was a dropped connection after DATA that had
5137 been sent in pipelining mode. That is, the server had advertised PIPELINING
5138 but was not implementing it correctly. The effect of the bug was incorrect
5139 behaviour, such as trying another host, and this could lead to a crash.
5145 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5146 and daemon.c were passed as pointers to ints; they should have been
5147 pointers to socklen_t variables (which are typically unsigned ints).
5149 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5152 3. Fixed a really odd bug that affected only the testing scheme; patching a
5153 certain fixed string in the binary changed the value of another string that
5154 happened to be identical to the end of the original first string.
5156 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5157 name", it returns that address as the IP address. On some operating
5158 systems (e.g. Solaris), it also passes back the IP address string as the
5159 "host name". However, on others (e.g. Linux), it passes back an empty
5160 string. Exim wasn't checking for this, and was changing the host name to an
5161 empty string, assuming it had been canonicized.
5163 5. Although rare, it is permitted to have more than one PTR record for a given
5164 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5165 all the names associated with an address, because they do in Solaris.
5166 However, it seems that they do not in Linux for data that comes from the
5167 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5168 I found this out when I moved to a new Linux workstation and tried to run
5169 the Exim test suite.
5171 To get round this problem I have changed the code so that it now does its
5172 own call to the DNS to look up PTR records when searching for a host name.
5173 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5174 addresses that are only in /etc/hosts are still found.
5176 This behaviour is, however, controlled by an option called host_lookup_
5177 order, which defaults to "bydns:byaddr". If people want to use the other
5178 order, or indeed, just use one or the other means of lookup, they can
5179 specify it in this variable.
5181 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5182 some operating systems, this comes back from gethostbyaddr() as an empty
5183 string, and this is what Exim used to test for. However, it seems that in
5184 other systems, "." is yielded. Exim now tests for this case too.
5186 7. The values of check_spool_space and check_log_space are now held internally
5187 as a number of kilobytes instead of an absolute number of bytes. If a
5188 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5189 kilobyte. This means that much larger values can be stored.
5191 8. Exim monitor: an attempt to get the action menu when not actually pointing
5192 at a message produces an empty menu entitled "No message selected". This
5193 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5194 no entries in it ("Shell widget menu has zero width and/or height"). So I
5195 have added a single, blank menu entry in this case.
5197 9. Added ${quote_local_part.
5199 10. MIME decoding is now applied to the contents of Subject: header lines when
5202 11. Now that a reference to $sender_host_address automatically causes a reverse
5203 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5204 host lookup before query-style lookups in lists that might use this
5205 variable. This has therefore been abolished, and the "net-" prefix is no
5206 longer necessary for query-style lookups.
5208 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5209 have been a typo for LFLAGS, so it has been changed.
5211 13. The install script calls Exim with "-C /dev/null" in order to find the
5212 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5213 to be output. Howeve, since Exim outputs its version number before the
5214 error, it didn't break the script. It just looked ugly. I fixed this by
5215 always allowing "-C /dev/null" if the caller is root.
5217 14. Ignore overlarge ACL variable number when reading spool file - insurance
5218 against a later release with more variables having written the file.
5220 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5221 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5222 this, as well as the form without the "IPv6" on the front.
5224 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5225 OS/Makefile-Darwin file.
5227 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5228 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5230 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5231 variable. However, some OS use ints for inodes. I've added cast to long int
5232 to get rid of the compiler warning.
5234 19. I had forgotten to lock out "/../" in configuration file names when
5235 ALT_CONFIG_PREFIX was set.
5237 20. Routers used for verification do not need to specify transports. However,
5238 if such a router generated a host list, and callout was configured, Exim
5239 crashed, because it could not find a port number from the (non-existent)
5240 transport. It now assumes port 25 in this circumstance.
5242 21. Added the -t option to exigrep.
5244 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5245 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5246 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5247 them is set, LOOKUP_LSEARCH is forced.
5249 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5251 24. Added sender and host information to the "rejected by local_scan()" log
5252 line; previously there was no indication of these.
5254 25. Added .include_if_exists.
5256 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5257 Linux. It turns out that not all file systems support this. Apparently some
5258 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5259 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5260 not supported on the file descriptor, is now ignored when Exim is trying to
5261 sync a directory. This applies only to Linux.
5263 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5265 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5266 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5269 29. Added support for maildirsize files from supplied patch (modified a bit).
5271 30. The use of :fail: followed by an empty string could lead Exim to respond to
5272 sender verification failures with (e.g.):
5274 550 Verification failed for <xxx>
5275 550 Sender verify failed
5277 where the first response line was missing the '-' that indicates it is not
5278 the final line of the response.
5280 31. The loop for finding the name of the user that called Exim had a hardwired
5281 limit of 10; it now uses the value of finduser_retries, which is used for
5282 all other user lookups.
5284 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5287 33. Exim was neglecting to zero errno before one call of strtol() when
5288 expanding a string and expecting an integer value. On some systems this
5289 resulted in spurious "integer overflow" errors. Also, it was casting the
5290 result into an int without checking.
5292 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5293 did not work. The code looks as if it has *never* worked, though it appears
5294 to have been documented since at least releast 1.62. I have made it work.
5296 35. The "timeout_DNS" error in retry rules, also documented since at least
5297 1.62, also never worked. As it isn't clear exactly what this means, and
5298 clearly it isn't a major issue, I have abolished the feature by treating it
5299 as "timeout", and writing a warning to the main and panic logs.
5301 36. The display of retry rules for -brt wasn't always showing the error code
5304 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5305 timeout_connect_A, timeout_connect_MX.
5307 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5308 to the empty sender.
5310 39. The daemon was not analysing the content of -oX till after it had closed
5311 stderr and disconnected from the controlling terminal. This meant that any
5312 syntax errors were only noted on the panic log, and the return code from
5313 the command was 0. By re-arranging the code a little, I've made the
5314 decoding happen first, so such errors now appear on stderr, and the return
5315 code is 1. However, the actual setting up of the sockets still happens in
5316 the disconnected process, so errors there are still only recorded on the
5319 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5320 connections (as happens on some IP stacks) was logged at start up time as
5321 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5322 it from "IPv6 and IPv4", which means that two separate sockets are being
5325 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5326 says whether AF_INET or AF_INET6 was passed as an argument.
5328 42. Exiwhat output was messed up when time zones were included in log
5331 43. Exiwhat now gives more information about the daemon's listening ports,
5332 and whether -tls-on-connect was used.
5334 44. The "port" option of the smtp transport is now expanded.
5336 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5337 silently ignored. Now an error message is written to the main and panic
5340 46. There's a new ACL modifier called "logwrite" which writes to a log file
5341 as soon as it is encountered.
5343 47. Added $local_user_uid and $local_user_gid at routing time.
5345 48. Exim crashed when trying to verify a sender address that was being
5348 49. Exim was recognizing only a space character after ".include". It now also
5349 recognizes a tab character.
5351 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5352 extracting the relevant information from the specification. The man page no
5353 longer contains scrambled data for the -d option, and I've added a section
5354 at the front about calling Exim under different names.
5356 51. Added "extra_headers" argument to the "mail" command in filter files.
5358 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5361 53. Installed eximstats 1.29.
5363 54. Added transport_filter_timeout as a generic transport option.
5365 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5366 Cc: header lines. This was required by RFC 822, but it not required by RFC
5369 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5370 incoming messages that did not have them. Now it does so only if the
5371 message originates locally, that is, if there is no associated remote host
5372 address. When Resent- header lines are present, this applies to the Resent-
5373 lines rather than the non-Resent- lines.
5375 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5376 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5378 58. Messages for configuration errors now include the name of the main
5379 configuration file - useful now that there may be more than one file in a
5380 list (.included file names were always shown).
5382 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5383 for those rare installations that do not start the daemon as root or run it
5384 setuid root. I've cut out the call to initgroups() if the daemon is not
5387 60. The Exim user and group can now be bound into the binary as text strings
5388 that are looked up at the start of Exim's processing.
5390 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5392 62. Added $mailstore_basename variable.
5394 63. Installed patch to sieve.c from Michael Haardt.
5396 64. When Exim failed to open the panic log after failing to open the main log,
5397 the original message it was trying to log was written to stderr and debug
5398 output, but if they were not available (the usual case in production), it
5399 was lost. Now it is written to syslog before the two lines that record the
5400 failures to open the logs.
5402 65. Users' Exim filters run in subprocesses under the user's uid. It is
5403 possible for a "deliver" command or an alias in a "personal" command to
5404 provoke an address rewrite. If logging of address rewriting is configured,
5405 this fails because the process is not running as root or exim. There may be
5406 a better way of dealing with this, but for the moment (because 4.30 needs
5407 to be released), I have disabled address rewrite logging when running a
5408 filter in a non-root, non-exim process.
5414 1. The buildconfig auxiliary program wasn't quoting the value set for
5415 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5416 not defined. This bug was masked in 4.22 by the effect that was fixed in
5419 2. Some messages that were rejected after a message id was allocated were
5420 shown as "incomplete" by exigrep. It no longer does this for messages that
5421 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5423 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5424 have allow_domain_literals set, the ID did not get logged in the <= line.
5425 Domain literals are now always recognized in Message-ID: header lines.
5427 4. The first argument for a ${extract expansion item is the key name or field
5428 number. Leading and trailing spaces in this item were not being ignored,
5429 causing some misleading effects.
5431 5. When deliver_drop_privilege was set, single queue runner processes started
5432 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5433 same command in the process it spins off) were not dropping privilege.
5435 6. When the daemon running as "exim" started a queue runner, it always
5436 re-executed Exim in the spun-off process. This is a waste of effort when
5437 deliver_drop_privilege is set. The new process now just calls the
5438 queue-runner function directly.
5444 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5447 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5448 ignored. Though the use of -oP was forcing the writing of a pid file, it
5449 was always written to the default place.
5451 3. If the message "no IP address found for host xxxx" is generated during
5452 incoming verification, it is now followed by identification of the incoming
5453 connection (so you can more easily find what provoked it).
5455 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5457 5. Added some features to "harden" Exim a bit more against certain attacks:
5459 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5460 be put in Local/Makefile. This is like the never_users runtime option,
5461 but it cannot be overridden. The default setting is "root".
5463 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5464 prefix string with which any file named in a -C command line option
5467 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5468 is retained for -C and -D only if the caller of Exim is root. Without
5469 it, the exim user may also use -C and -D and retain privilege.
5471 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5472 command line option is disabled.
5474 6. Macro names set by the -D option must start with an upper case letter, just
5475 like macro names defined in the configuration file.
5477 7. Added "dereference=" facility to LDAP.
5479 8. Two instances of the typo "uknown" in the source files are fixed.
5481 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5482 the Configure-Makefile script screwed up while processing it.
5484 10. Incorporated PCRE 4.4.
5486 11. The SMTP synchronization check was not operating right at the start of an
5487 SMTP session. For example, it could not catch a HELO sent before the client
5488 waited for the greeting. There is now a check for outstanding input at the
5489 point when the greeting is written. Because of the duplex, asynchronous
5490 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5491 way, but not yet received, when the check is performed.
5493 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5494 on TCP/IP sockets, because this apparently causes some broken clients to
5497 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5498 unchanged) from the Cygwin maintainer.
5500 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5501 was supported instead of testing for mbx. Effectively a typo.
5503 15. The spa authenticator server code was not checking that the input it
5504 received was valid base64.
5506 16. The debug output line for the "set" modifier in ACLs was not showing the
5507 name of the variable that was being set.
5509 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5511 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5512 lookup on its own. Something else was needed to trigger the lookup. For
5513 example, a match in host_lookup or the need for a host name in a host list.
5514 Now, if $sender_host_name is referenced and the host name has not yet been
5515 looked up, a lookup is performed. If the lookup fails, the variable remains
5516 empty, and $host_lookup_failed is set to "1".
5518 19. Added "eqi" as a case-independent comparison operator.
5520 20. The saslauthd authentication condition could segfault if neither service
5521 nor realm was specified.
5523 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5524 error message that was logged was misleading, and incoming SMTP
5525 connections were dropped. The message is now more accurate, and temporary
5526 errors are given to SMTP connections.
5528 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5529 (except RSET) with a 421 error, until QUIT is received. However, it was
5530 failing to send a response to QUIT.
5532 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5533 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5534 ACL is now run after the helo_try_verify_hosts code.
5536 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5537 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5538 case-independent, so other case variants are also recognized). Apparently
5539 some systems use these upper case variants.
5541 25. If more than two messages were waiting for the same host, and a transport
5542 filter was specified for the transport, Exim sent two messages over the
5543 same TCP/IP connection, and then failed with "socket operation on non-
5544 socket" when it tried to send the third.
5546 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5548 27. The extern definition of crypt16() in expand.c was not being excluded when
5549 the OS had its own crypt16() function.
5551 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5552 as a preferred synonym for return_size_limit, both as an option and as an
5555 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5557 30. Changed the default configuration ACL to relax the local part checking rule
5558 for addresses that are not in any local domains. For these addresses,
5559 slashes and pipe symbols are allowed within local parts, but the sequence
5560 /../ is explicitly forbidden.
5562 31. SPA server authentication was not clearing the challenge buffer before
5565 32. log_message in a "warn" ACL statement was writing to the reject log as
5566 well as to the main log, which contradicts the documentation and doesn't
5567 seem right (because no rejection is happening). So I have stopped it.
5569 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5570 However, I am unable to do any testing of this.
5572 34. Fixed an infelicity in the appendfile transport. When checking directories
5573 for a mailbox, to see if any needed to be created, it was accidentally
5574 using path names with one or more superfluous leading slashes; tracing
5575 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5577 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5578 contents are added to the log line that is written for every discarded
5579 recipient. (Previously a log_message setting was ignored.)
5581 36. The ${quote: operator now quotes the string if it is empty.
5583 37. The install script runs exim in order to find its version number. If for
5584 some reason other than non-existence or emptiness, which it checks, it
5585 could not run './exim', it was installing it with an empty version number,
5586 i.e. as "exim-". This error state is now caught, and the installation is
5589 38. An argument was missing from the function that creates an error message
5590 when Exim fails to connect to the socket for saslauthd authentication.
5591 This could cause Exim to crash, or give a corrupted message.
5593 39. Added isip, isip4, and isip6 to ${if conditions.
5595 40. The ACL variables $acl_xx are now saved with the message, and can be
5596 accessed later in routers, transports, and filters.
5598 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5599 strings in the file are not string-expanded.
5601 42. If a MAIL command specified a SIZE value that was too large to fit into an
5602 int variable, the check against message_size_limit failed. Such values are
5603 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5604 day this will have to be increased, but I don't think I want to be around
5605 when emails are that large.
5612 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5613 iconv() is not standard in FreeBSD.
5615 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5616 IPv6 enabled. The observed symptom was a segmentation fault on return from
5617 the function os_common_find_running_interfaces() in src/os.c.
5619 3. In the check_special_case() function in daemon.c I had used "errno" as an
5620 argument name, which causes warnings on some systems. This was basically a
5621 typo, since it was named "eno" in the comments!
5623 4. The code that waits for the clock to tick (at a resolution of some fraction
5624 of a second) so as to ensure message-id uniqueness was always waiting for
5625 at least one whole tick, when it could have waited for less. [This is
5626 almost certainly not relevant at current processor speeds, where it is
5627 unlikely to ever wait at all. But we try to future-proof.]
5629 5. The function that sleeps for a time interval that includes fractions of a
5630 second contained a race. It did not block SIGALRM between setting the
5631 timer, and suspending (a couple of lines later). If the interval was short
5632 and the sigsuspend() was delayed until after it had expired, the suspension
5633 never ended. On busy systems this could lead to processes getting stuck for
5636 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5637 process, before it forks any delivery processes. The open lookup caching
5638 mechanism meant that the open file or database connection was passed into
5639 the delivery process. The problem was that delivery processes always tidy
5640 up cached lookup data. This could cause a problem for the next delivery
5641 process started by the queue runner, because the external queue runner
5642 process does not know about the closure. So the next delivery process
5643 still has data in the lookup cache. In the case of a file lookup, there was
5644 no problem because closing a file descriptor in a subprocess doesn't affect
5645 the parent. However, if the lookup was caching a connection to a database,
5646 the connection was closed, and the second delivery process was likely to
5647 see errors such as "PGSQL: query failed: server closed the connection
5648 unexpectedly". The problem has been fixed by closing all cached lookups
5649 in a queue runner before running a delivery process.
5651 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5652 seem to have the "const" qualifier which it has on other OS. I've
5655 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5656 *of the same type* (client or server) with the same public name that an
5657 error should be diagnosed.
5659 9. When Exim looked up a host name for an IP address, but failed to find the
5660 original IP address when looking up the host name (a safety check), it
5661 output the message "<ip address> does not match any IP for NULL", which was
5662 confusing, to say the least. The bug was that the host name should have
5663 appeared instead of "NULL".
5665 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5666 user that is built into the binary, and the -C or -D options is used, root
5667 privilege is dropped before the configuration file is read. In addition,
5668 logging is switched to stderr instead of the normal log files. If the
5669 configuration then re-defines the Exim user, the unprivileged environment
5670 is probably not what is expected, so Exim logs a panic warning message (but
5673 However, if deliver_drop_privilege is set, the unprivileged state may well
5674 be exactly what is intended, so the warning has been cut out in that case,
5675 and Exim is allowed to try to write to its normal log files.
5681 1. smtp_return_error_details was not giving details for temporary sender
5682 or receiver verification errors.
5684 2. Diagnose a configuration error if two authenticators have the same public
5687 3. Exim used not to create the message log file for a message until the first
5688 delivery attempt. This could be confusing when incoming messages were held
5689 for policy or load reasons. The message log file is now created at the time
5690 the message is received, and an initial "Received" line is written to it.
5692 4. The automatically generated man page for command line options had a minor
5693 bug that caused no ill effects; however, a more serious problem was that
5694 the procedure for building the man page automatically didn't always
5695 operate. Consequently, release 4.20 contains an out-of-date version. This
5696 shouldn't happen again.
5698 5. When building Exim with embedded Perl support, the script that builds the
5699 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5700 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5702 6. The freeze_tell option was not being used for messages that were frozen on
5703 arrival, either by an ACL or by local_scan().
5705 7. Added the smtp_incomplete_transaction log selector.
5707 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5708 was accepting AUTH without a new EHLO.
5710 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5711 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5714 10. Logging of TCP/IP connections (when configured) now happens in the main
5715 daemon process instead of the child process, so that the TCP/IP connection
5716 count is more accurate (but it can never be perfect).
5718 11. The use of "drop" in a nested ACL was not being handled correctly in the
5719 outer ACL. Now, if condition failure induced by the nested "drop" causes
5720 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5721 or "require"), the connection is dropped.
5723 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5724 that yield "discard" can now be used with an "accept" or a "discard" verb,
5725 but an error is generated for any others (because I can't see a useful way
5726 to define what should happen).
5728 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5729 are now processed in the same way as lines in the Exim configuration file.
5730 In particular, continuation lines are supported.
5732 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5734 15. Added -ti meaning -t -i.
5736 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5737 domains, and warn by logging if others are found.
5739 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5740 returned from GIFCONF when Exim is trying to find the list of interfaces on
5741 a host. The code in os.c has been modified to copy each ifreq to an aligned
5742 structure in all cases.
5744 Also, in some cases, the returned ifreq's were being copied to a 'struct
5745 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5746 means the last couple of bytes of an IPv6 address could be chopped if the
5747 ifreq contained only a normal sockaddr (14 bytes storage).
5749 18. Named domain lists were not supported in the hosts_treat_as_local option.
5750 An entry such as +xxxx was not recognized, and was treated as a literal
5753 19. Ensure that header lines added by a DATA ACL are included in the reject log
5754 if the ACL subsequently rejects the message.
5756 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5757 MD5 (which is deprecated).
5759 21. When testing a filter file using -bf, Exim was writing a message when it
5760 took the sender from a "From " line in the message, but it was not doing so
5761 when it took $return_path from a Return-Path: header line. It now does.
5763 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5764 with a valid header line field name (a series of printing characters
5765 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5767 23. Changed "disc" in the source to "disk" to conform to the documentation and
5768 the book and for uniformity.
5770 24. Ignore Sendmail's -Ooption=value command line item.
5772 25. When execve() failed while trying to run a command in a pipe transport,
5773 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5774 could be confused with a return value of 69 from the command itself. This
5775 has been changed to 127, the value the shell returns if it is asked to run
5776 a non-existent command. The wording for the related log line suggests a
5777 non-existent command as the problem.
5779 26. If received_header_text expands to an empty string, do not add a Received:
5780 header line to the message. (Well, it adds a token one on the spool, but
5781 marks it "old" so that it doesn't get used or transmitted.)
5783 27. Installed eximstats 1.28 (addition of -nt option).
5785 28. There was no check for failure on the call to getsockname() in the daemon
5786 code. This can fail if there is a shortage of resources on the system, with
5787 ENOMEM, for example. A temporary error is now given on failure.
5789 29. Contrary to the C standard, it seems that in some environments, the
5790 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5791 program. Exim now does this explicitly; it affects the formatting of
5792 timestamps using strftime().
5794 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5795 complaints. I've now initialized all the variables, to avoid this.
5797 32. Header lines added by a system filter were not being "seen" during
5798 transport-time rewrites.
5800 33. The info_callback() function passed to OpenSSL is set up with type void
5801 (*)(SSL *, int, int), as described somewhere. However, when calling the
5802 function (actually a macro) that sets it up, the type void(*)() is
5803 expected. I've put in a cast to prevent warnings from picky compilers.
5805 34. If a DNS black list lookup found a CNAME record, but there were no A
5806 records associated with the domain it pointed at, Exim crashed.
5808 35. If a DNS black list lookup returned more than one A record, Exim ignored
5809 all but the first. It now scans all returned addresses if a particular IP
5810 value is being sought. In this situation, the contents of the
5811 $dnslist_value variable are a list of all the addresses, separated by a
5814 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5815 to accept a host name and all its aliases if the forward lookup for any of
5816 them yielded the IP address of the incoming connection. Now it accepts only
5817 those names whose forward lookup yields the correct IP address. Any other
5818 names are discarded. This closes a loophole whereby a rogue DNS
5819 administrator could create reverse DNS records to break through a
5820 wildcarded host restriction in an ACL.
5822 37. If a user filter or a system filter that ran in a subprocess used any of
5823 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5824 the wrong values were passed to the pipe command ($thisaddress had the
5825 value of $0, $0 had the value of $1, etc). This bug was introduced by
5826 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5828 38. Improved the line breaking for long SMTP error messages from ACLs.
5829 Previously, if there was no break point between 40 and 75 characters, Exim
5830 left the rest of the message alone. Two changes have been made: (a) I've
5831 reduced the minimum length to 35 characters; (b) if it can't find a break
5832 point between 35 and 75 characters, it looks ahead and uses the first one
5833 that it finds. This may give the occasional overlong line, but at least the
5834 remaining text gets split now.
5836 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5837 file descriptors might be low, and that setting 1000 would always raise it.
5838 It turns out that in some environments, the limit is already over 1000 and
5839 that lowering it causes trouble. So now Exim takes care not to decrease it.
5841 40. When delivering a message, the value of $return_path is set to $sender_
5842 address at the start of routing (routers may change the value). By an
5843 oversight, this default was not being set up when an address was tested by
5844 -bt or -bv, which affected the outcome if any router or filter referred to
5847 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5848 log only when "message" or "log_message" are set. However, if one of the
5849 conditions was an address verification, or a call to a nested ACL, the
5850 messages generated by the underlying test were being passed through. This
5851 no longer happens. The underlying message is available in $acl_verify_
5852 message for both "message" and "log_message" expansions, so it can be
5853 passed through if needed.
5855 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5856 new expansion $bh_ to give the encoded byte string without charset
5857 translation. Translation happens only if iconv() is available; HAVE_ICONV
5858 indicates this at build time. HEADERS_CHARSET gives the charset to
5859 translate to; headers_charset can change it in the configuration, and
5860 "headers charset" can change it in an individual filter file.
5862 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5863 that creates RFC 2047 encoded "words" labels them as that charset instead
5864 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5865 expansion operator; (ii) when Exim creates a From: line for a local
5866 message; (iii) when a header line is rewritten to include a "phrase" part.
5868 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5869 buggy, causing it to skip the first lines of messages whose message ID
5870 ended in 'D'. This would not have bitten before Exim release 4.14, because
5871 message IDs were unlikely to end in 'D' before then. The effect was to have
5872 incorrect size information for certain domains.
5874 45. #include "config.h" was missing at the start of the crypt16.c module. This
5875 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5878 46. If there was a timeout during a "random" callout check, Exim treated it as
5879 a failure of the random address, and carried on sending RSET and the real
5880 address. If the delay was just some slowness somewhere, the response to the
5881 original RCPT would be taken as a response to RSET and so on, causing
5882 mayhem of various kinds.
5884 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5885 when I implemented it. It didn't allow for the fact that some option values
5886 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5887 the right test for positive values.
5889 48. Domain names in DNS records are case-independent. Exim always looks them up
5890 in lower case. Some resolvers return domain names in exactly the case they
5891 appear in the zone file, that is, they may contain uppercase letters. Not
5892 all resolvers do this - some return always lower case. Exim was treating a
5893 change of case by a resolver as a change of domain, similar to a widening
5894 of a domain abbreviation. This triggered its re-routing code and so it was
5895 trying to route what was effectively the same domain again. This normally
5896 caused routing to fail (because the router wouldn't handle the domain
5897 twice). Now Exim checks for this case specially, and just changes the
5898 casing of the domain that it ultimately uses when it transmits the message
5901 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5904 50. If a filter generated a file delivery with a non-absolute name (possible if
5905 no home directory exists for the router), the forbid_file option was not
5908 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5909 the existing equality matching.
5911 52. Exim was using ints instead of ino_t variables in some places where it was
5912 dealing with inode numbers.
5914 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5915 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5916 called TMPDIR, and if it finds it is different, it changes its value.
5918 54. The smtp_printf() function is now made available to local_scan() so
5919 additional output lines can be written before returning. There is also an
5920 smtp_fflush() function to enable the detection of a dropped connection.
5921 The variables smtp_input and smtp_batched_input are exported to
5924 55. Changed the default runtime configuration: the message "Unknown user"
5925 has been removed from the ACL, and instead placed on the localuser router,
5926 using the cannot_route_message feature. This means that any verification
5927 failures that generate their own messages won't get overridden. Similarly,
5928 the "Unrouteable address" message that was in the ACL for unverifiable
5929 relay addresses has also been removed.
5931 56. Added hosts_avoid_esmtp to the smtp transport.
5933 57. The exicyclog script was not checking for the esoteric option
5934 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5935 will work only if exicyclog is run under the appropriate euid.
5937 58. Following a discussion on the list, the rules by which Exim recognises line
5938 endings on incoming messages have been changed. The -dropcr and drop_cr
5939 options are now no-ops, retained only for backwards compatibility. The
5940 following line terminators are recognized: LF CRLF CR. However, special
5941 processing applies to CR:
5943 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5944 nor a local message in the state where . is a terminator.
5946 (ii) If a bare CR is encountered in a header line, an extra space is added
5947 after the line terminator so as not to end the header. The reasoning
5948 behind this is that bare CRs in header lines are most likely either
5949 to be mistakes, or people trying to play silly games.
5951 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5952 was being incorrectly given as 18 bytes larger than it should have been.
5953 This is a VOB (very old bug).
5955 60. This may never have affected anything current, but just in case it has:
5956 When the local host is found other than at the start of a list of hosts,
5957 the local host, those with the same MX, and any that follow, are discarded.
5958 When the list in question was part of a longer list of hosts, the following
5959 hosts (not currently being processed) were also being discarded. This no
5960 longer happens. I'm not sure if this situation could ever has previously
5963 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5966 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5967 header. This is something that is recommended in a new Internet Draft, and
5968 is something that is documented as being done by Sendmail. There are two
5969 possible values. For messages generated by the autoreply transport, Exim
5972 Auto-Submitted: auto-replied
5974 whereas for all other generated messages (e.g. bounces) it adds
5976 Auto-Submitted: auto-generated
5978 63. The "personal" condition in filters now includes a test for the
5979 Auto-Submitted: header. If it contains the string "auto-" the message it
5980 not considered personal.
5982 64. Added rcpt_include_affixes as a generic transport option.
5984 65. Added queue_only_override (default true).
5986 66. Added the syslog_duplication option.
5988 67. If what should have been the first header line of a message consisted of
5989 a space followed by a colon, Exim was mis-interpreting it as a header line.
5990 It isn't of course - it is syntactically invalid and should therefore be
5991 treated as the start of the message body. The misbehaviour could have
5992 caused a number of strange effects, including loss of data in subsequent
5993 header lines, and spool format errors.
5995 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5996 client host had authenticated. This control can now be exercised by an ACL
5997 for more flexibility.
5999 69. By default, callouts do not happen when testing with -bh. There is now a
6000 variant, -bhc, which does actually run the callout code, including
6001 consulting and updating the callout cache.
6003 70. Added support for saslauthd authentication, courtesy of Alexander
6006 71. If statvfs() failed on the spool or log directories while checking their
6007 size for availability, Exim confusingly gave the error "space shortage".
6008 Furthermore, in debugging mode it crashed with a floating point exception.
6009 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6010 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6011 serious problem, Exim now writes to the main and panic logs when this
6012 happens, with details of the failure. It then refuses to accept the
6013 incoming message, giving the message "spool directory problem" or "log
6014 directory problem" with a 421 code for SMTP messages.
6016 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6017 0, 1, and 2 exist, because some OS complain for execs without them (see
6018 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6019 descriptors. However, the code omitted to check that the open succeeded,
6020 causing mysterious errors if for some reason the permissions on /dev/null
6021 got screwed. Now Exim writes a message to the main and panic logs, and
6022 bombs out if it can't open /dev/null.
6024 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6025 interact so that it is all more flexible. It is supposed to remain
6026 backwards compatible. Also added extra_local_interfaces.
6028 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6029 to bomb out with an assertion failure - to the client this appears as a
6030 connection drop. This problem occurs in the part of the code that was taken
6031 from the Samba project. Fortunately, the assertion is in a very simple
6032 function, so I have fixed this by reproducing the function inline in the
6033 one place where it is called, and arranging for authentication to fail
6034 instead of killing the process with assert().
6036 75. The SPA client code was not working when the server requested OEM rather
6037 than Unicode encoding.
6039 76. Added code to make require_files with a specific uid setting more usable in
6040 the case where statting the file as root fails - usually a non-root-mounted
6041 NFS file system. When this happens and the failure is EACCES, Exim now
6042 forks a subprocess and does the per-uid checking as the relevant uid.
6044 77. Added process_log_path.
6046 78. If log_file_path was not explicitly set, a setting of check_log_space or
6047 check_log_inodes was ignored.
6049 79. If a space check for the spool or log partitions fails, the incident is now
6050 logged. Of course, in the latter case the data may get lost...
6052 80. Added the %p formatting code to string_format() so that it can be used to
6053 print addresses in debug_print(). Adjusted all the address printing in the
6054 debugging in store.c to use %p rather than %d.
6056 81. There was a concern that a line of code in smtp_in.c could overflow a
6057 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6058 initially expressed, the concern was not well-founded, because trailing
6059 spaces are removed early. However, if the trailing spaces were followed by
6060 a NULL, they did not get removed, so the overflow was possible. Two fixes
6063 (a) I re-wrote the offending code in a cleaner fashion.
6064 (b) If an incoming SMTP command contains a NULL character, it is rejected
6067 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6068 runs initgroups(), so that if the Exim user is in any additional groups,
6069 they will be used during message reception.
6075 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob