1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
23 TL/03 Bugzilla 1518: Clarify "condition" processing in routers; that
24 syntax errors in an expansion can be treated as a string instead of
25 logging or causing an error, due to the internal use of bool_lax
26 instead of bool when processing it.
28 JH/02 Add EXPERIMENTAL_DANE, allowing for using the DNS as trust-anchor for
29 server certificates when making smtp deliveries.
31 JH/03 Support secondary-separator specifier for MX, SRV, TLSA lookups.
33 JH/04 Add ${sort {list}{condition}{extractor}} expansion item.
35 TL/04 Bugzilla 1216: Add -M (related messages) option to exigrep.
37 TL/05 GitHub Issue 18: Adjust logic testing for true/false in redis lookups.
38 Merged patch from Sebastian Wiedenroth.
42 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
43 checkers that were complaining about end of non-void function with no
46 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
47 This was a regression intruduced in 4.83 by another bugfix.
49 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
51 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
52 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
58 TF/01 Correctly close the server side of TLS when forking for delivery.
60 When a message was received over SMTP with TLS, Exim failed to clear up
61 the incoming connection properly after forking off the child process to
62 deliver the message. In some situations the subsequent outgoing
63 delivery connection happened to have the same fd number as the incoming
64 connection previously had. Exim would try to use TLS and fail, logging
65 a "Bad file descriptor" error.
67 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
68 utilities have not been installed.
70 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
71 temporary space as the ACL may create new global variables.
73 TL/01 LDAP support uses per connection or global context settings, depending
74 upon the detected version of the libraries at build time.
76 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
77 to extract and use the src ip:port in logging and expansions as if it
78 were a direct connection from the outside internet. PPv2 support was
79 updated based on HAProxy spec change in May 2014.
81 JH/02 Add ${listextract {number}{list}{success}{fail}}.
83 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
84 Properly escape header and check for NULL return.
86 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
89 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
91 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
92 characters in header names, implemented as a verify condition.
93 Contributed by Michael Fischer v. Mollard.
95 TL/05 Rename SPF condition results err_perm and err_temp to standardized
96 results permerror and temperror. Previous values are deprecated but
97 still accepted. In a future release, err_perm and err_temp will be
98 completely removed, which will be a backward incompatibility if the
99 ACL tests for either of these two old results. Patch contributed by
100 user bes-internal on the mailing list.
102 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
104 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
105 selectors, in both main and reject logs.
107 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
110 JH/07 Add malware type "sock" for talking to simple daemon.
112 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
114 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
115 routers/transports under cutthrough routing.
117 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
118 numbers. Touch up "bool" conditional to keep the same definition.
120 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
122 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
124 JH/12 Expand items in router/transport headers_add or headers_remove lists
125 individually rather than the list as a whole. Bug 1452.
127 Required for reasonable handling of multiple headers_ options when
128 they may be empty; requires that headers_remove items with embedded
129 colons must have them doubled (or the list-separator changed).
131 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
132 view the policy declared in the DMARC record. Currently, $dmarc_status
133 is a combined value of both the record presence and the result of the
136 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
138 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
139 dnslookup router and the smtp transport (applying to the forward
142 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
143 of ldap servers used for a specific lookup. Patch provided by Heiko
146 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
147 New variable $lookup_dnssec_authenticated for observability.
149 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
150 Patch submitted by Lars Timman.
152 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
154 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
155 Requires trusted mode and valid format message id, aborts otherwise.
156 Patch contributed by Heiko Schlichting.
158 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
159 certextract with support for various fields. Bug 1358.
161 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
162 is requested by default, modifiable by smtp transport option
165 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
166 operate on certificate variables to give certificate fingerprints
167 Also new ${sha256:cert_variable}.
169 JH/23 The PRDR feature is moved from being Experimental into the mainline.
171 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
172 Christian Aistleitner.
174 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
176 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
177 file. Patch from Wolfgang Breyha.
179 JH/25 Expand the coverage of the delivery $host and $host_address to
180 client authenticators run in verify callout. Bug 1476.
182 JH/26 Port service names are now accepted for tls_on_connect_ports, to
183 align with daemon_smtp_ports. Bug 72.
185 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
186 support and error reporting did not work properly.
188 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
189 and is readable. Patch from Andrew Colin Kissa.
191 TL/14 Enhance documentation of ${run expansion and how it parses the
192 commandline after expansion, particularly in the case when an
193 unquoted variable expansion results in an empty value.
195 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
197 PP/02 Fix internal collision of T_APL on systems which support RFC3123
198 by renaming away from it. Addresses GH issue 15, reported by
201 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
203 TL/15 SECURITY: prevent double expansion in math comparison functions
204 (can expand unsanitized data). Not remotely exploitable.
211 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
213 PP/02 Make -n do something, by making it not do something.
214 When combined with -bP, the name of an option is not output.
216 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
219 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
220 $sender_host_name and config options to manage this, and basic check
223 PP/05 DSCP support for outbound connections and control modifier for inbound.
225 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
226 (Only plugin which currently uses this is kerberos4, which nobody should
227 be using, but we should make it available and other future plugins might
228 conceivably use it, even though it would break NAT; stuff *should* be
229 using channel bindings instead).
231 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
232 name; added for Sendmail compatibility; requires admin caller.
233 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
234 just ignore it); requires trusted caller.
235 Also parse but ignore: -Ac -Am -X<logfile>
238 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
240 TL/02 Add +smtp_confirmation as a default logging option.
242 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
243 Patch by Magnus Holmgren from 2007-02-20.
245 TL/04 Bugzilla 1281 - Spec typo.
246 Bugzilla 1283 - Spec typo.
247 Bugzilla 1290 - Spec grammar fixes.
249 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
251 TL/06 Add Experimental DMARC support using libopendmarc libraries.
253 TL/07 Fix an out of order global option causing a segfault. Reported to dev
254 mailing list by by Dmitry Isaikin.
256 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
258 JH/02 Support "G" suffix to numbers in ${if comparisons.
260 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
262 NM/01 Bugzilla 1197 - Spec typo
263 Bugzilla 1196 - Spec examples corrections
265 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
267 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
268 gnutls_enable_pkcs11, but renamed to more accurately indicate its
271 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
272 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
274 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
275 "acl {{name}{arg}...}", and optional args on acl condition
278 JH/05 Permit multiple router/transport headers_add/remove lines.
280 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
282 JH/07 Avoid using a waiting database for a single-message-only transport.
283 Performance patch from Paul Fisher. Bugzilla 1262.
285 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
288 JH/09 Add $headers_added variable, with content from use of ACL modifier
289 add_header (but not yet added to the message). Bugzilla 199.
291 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
292 Pulled from Bugzilla 817 by Wolfgang Breyha.
294 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
296 (nb: this is the same fix as in Exim 4.80.1)
298 JH/11 Add A= logging on delivery lines, and a client_set_id option on
301 JH/12 Add optional authenticated_sender logging to A= and a log_selector
304 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
306 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
307 advertise SMTP AUTH mechanism to us, instead of a generic
308 protocol violation error. Also, make Exim more robust to bad
309 data from the Dovecot auth socket.
311 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
313 When a queue runner is handling a message, Exim first routes the
314 recipient addresses, during which it prunes them based on the retry
315 hints database. After that it attempts to deliver the message to
316 any remaining recipients. It then updates the hints database using
319 So if a recipient address works intermittently, it can get repeatedly
320 deferred at routing time. The retry hints record remains fresh so the
321 address never reaches the final cutoff time.
323 This is a fairly common occurrence when a user is bumping up against
324 their storage quota. Exim had some logic in its local delivery code
325 to deal with this. However it did not apply to per-recipient defers
326 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
328 This change adds a proper retry rule check during routing so that the
329 final cutoff time is checked against the message's age. We only do
330 this check if there is an address retry record and there is not a
331 domain retry record; this implies that previous attempts to handle
332 the address had the retry_use_local_parts option turned on. We use
333 this as an approximation for the destination being like a local
334 delivery, as in LMTP.
336 I suspect this new check makes the old local delivery cutoff check
337 redundant, but I have not verified this so I left the code in place.
339 TF/02 Correct gecos expansion when From: is a prefix of the username.
341 Test 0254 submits a message to Exim with the header
345 When I ran the test suite under the user fanf2, Exim expanded
346 the header to contain my full name, whereas it should have added
347 a Resent-Sender: header. It erroneously treats any prefix of the
348 username as equal to the username.
350 This change corrects that bug.
352 GF/01 DCC debug and logging tidyup
353 Error conditions log to paniclog rather than rejectlog.
354 Debug lines prefixed by "DCC: " to remove any ambiguity.
356 TF/03 Avoid unnecessary rebuilds of lookup-related code.
358 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
359 Bug spotted by Jeremy Harris; was flawed since initial commit.
360 Would have resulted in OCSP responses post-SNI triggering an Exim
361 NULL dereference and crash.
363 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
365 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
366 Bug detection, analysis and fix by Samuel Thibault.
367 Bugzilla 1331, Debian bug #698092.
369 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
371 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
372 Server implementation by Todd Lyons, client by JH.
373 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
374 config variable "prdr_enable" controls whether the server
375 advertises the facility. If the client requests PRDR a new
376 acl_data_smtp_prdr ACL is called once for each recipient, after
377 the body content is received and before the acl_smtp_data ACL.
378 The client is controlled by bolth of: a hosts_try_prdr option
379 on the smtp transport, and the server advertisement.
380 Default client logging of deliveries and rejections involving
381 PRDR are flagged with the string "PRDR".
383 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
384 fclose(). Diagnosis by Todd Lyons.
386 PP/17 Update configure.default to handle IPv6 localhost better.
387 Patch by Alain Williams (plus minor tweaks).
390 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
391 This is now consistent with GnuTLS, and is now documented: the
392 previous undocumented portable approach to treating the option as
393 unset was to force an expansion failure. That still works, and
394 an empty string is now equivalent.
396 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
397 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
398 not performing validation itself.
400 PP/20 Added force_command boolean option to pipe transport.
401 Patch from Nick Koston, of cPanel Inc.
403 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
406 TF/04 Added udpsend ACL modifer and hexquote expansion operator
408 PP/21 Fix eximon continuous updating with timestamped log-files.
409 Broken in a format-string cleanup in 4.80, missed when I repaired the
410 other false fix of the same issue.
411 Report and fix from Heiko Schlichting.
414 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
415 Report from Prashanth Katuri.
417 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
418 It's SecureTransport, so affects any MacOS clients which use the
419 system-integrated TLS libraries, including email clients.
421 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
422 using a MIME ACL for non-SMTP local injection.
423 Report and assistance in diagnosis by Warren Baker.
425 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
427 JH/16 Fix comparisons for 64b. Bugzilla 1385.
429 TL/09 Add expansion variable $authenticated_fail_id to keep track of
430 last id that failed so it may be referenced in subsequent ACL's.
432 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
435 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
436 ldap library initialization, allowing self-signed CA's to be
437 used. Also properly sets require_cert option later in code by
438 using NULL (global ldap config) instead of ldap handle (per
439 session). Bug diagnosis and testing by alxgomz.
441 TL/12 Enhanced documentation in the ratelimit.pl script provided in
442 the src/util/ subdirectory.
444 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
445 renamed to Transport Post Delivery Action by Jeremy Harris, as
448 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
449 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
450 redis_servers = needs to be configured which will be used by the redis
451 lookup. Patch from Warren Baker, of The Packet Hub.
453 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
455 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
456 hostname or reverse DNS when processing a host list. Used suggestions
457 from multiple comments on this bug.
459 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
461 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
462 Missed a few lines, added it to make the runtest require no keyboard
465 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
466 contains upper case chars. Make router use caseful_local_part.
468 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
469 support when GnuTLS has been built with p11-kit.
475 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
477 This, or similar/improved, will also be change PP/11 of 4.82.
483 PP/01 Handle short writes when writing local log-files.
484 In practice, only affects FreeBSD (8 onwards).
485 Bugzilla 1053, with thanks to Dmitry Isaikin.
487 NM/01 Bugzilla 949 - Documentation tweak
489 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
492 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
494 PP/02 Implemented gsasl authenticator.
496 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
498 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
499 `pkg-config foo` for cflags/libs.
501 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
502 with rest of GSASL and with heimdal_gssapi.
504 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
505 `pkg-config foo` for cflags/libs for the TLS implementation.
507 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
508 properties get this fed in as external SSF. A number of robustness
509 and debugging improvements to the cyrus_sasl authenticator.
511 PP/08 cyrus_sasl server now expands the server_realm option.
513 PP/09 Bugzilla 1214 - Log authentication information in reject log.
514 Patch by Jeremy Harris.
516 PP/10 Added dbmjz lookup type.
518 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
520 PP/12 MAIL args handles TAB as well as SP, for better interop with
521 non-compliant senders.
522 Analysis and variant patch by Todd Lyons.
524 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
525 Bug report from Lars Müller <lars@samba.org> (via SUSE),
526 Patch from Dirk Mueller <dmueller@suse.com>
528 PP/13 tls_peerdn now print-escaped for spool files.
529 Observed some $tls_peerdn in wild which contained \n, which resulted
530 in spool file corruption.
532 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
533 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
534 or write after TLS renegotiation, which otherwise led to messages
537 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
538 as a tracking header (ie: a signed header comes before the signature).
539 Patch from Wolfgang Breyha.
541 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
542 comma-sep list; embedded commas doubled.
544 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
546 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
548 Report and patch from Dmitry Banschikov.
550 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
551 Removed SSL_clear() after SSL_new() which led to protocol negotiation
552 failures. We appear to now support TLS1.1+ with Exim.
554 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
555 lets Exim select keys and certificates based upon TLS SNI from client.
556 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
557 before an outbound SMTP session. New log_selector, +tls_sni.
559 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
560 NULL dereference. Report and patch from Alun Jones.
562 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
563 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
564 Not seeing resolver debug output on NetBSD, but suspect this is a
565 resolver implementation change.
567 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
568 Left warnings. Added "eximon gdb" invocation mode.
570 PP/21 Defaulting "accept_8bitmime" to true, not false.
572 PP/22 Added -bw for inetd wait mode support.
574 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
575 locate the relevant includes and libraries. Made this the default.
577 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
578 Bugzilla 1246, report and most of solution from Tomasz Kusy.
580 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
581 This may cause build issues on older platforms.
583 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
584 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
585 gnutls_require_mac & gnutls_require_protocols (no longer supported).
586 Added SNI support via GnuTLS too.
587 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
589 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
591 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
592 Applied second patch from Janne, implementing suggestion to default
593 multiple-strings-in-record handling to match SPF spec.
595 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
597 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
598 read-only, out of scope).
599 Patch from Wolfgang Breyha, report from Stuart Northfield.
601 PP/29 Fix three issues highlighted by clang analyser static analysis.
602 Only crash-plausible issue would require the Cambridge-specific
603 iplookup router and a misconfiguration.
604 Report from Marcin Mirosław.
606 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
608 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
609 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
610 As part of this, removing so much warning spew let me fix some minor
611 real issues in debug logging.
613 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
614 assignment on my part. Fixed.
616 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
617 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
618 Janne Snabb (who went above and beyond: thank you).
620 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
621 string otherwise requires a connection and a bunch more work and it's
622 relatively easy to get wrong. Should also expose TLS library linkage
625 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
626 64-bit ${eval} (JH/03).
628 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
629 GNU libc to support some of the 64-bit stuff, should not lead to
630 conflicts. Defined before os.h is pulled in, so if a given platform
631 needs to override this, it can.
633 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
634 protection layer was required, which is not implemented.
635 Bugzilla 1254, patch from Wolfgang Breyha.
637 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
638 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
639 tls_dhparam take prime identifiers. Also unbreak combination of
640 OpenSSL+DH_params+TLSSNI.
642 PP/39 Disable SSLv2 by default in OpenSSL support.
648 PP/01 Solaris build fix for Oracle's LDAP libraries.
649 Bugzilla 1109, patch from Stephen Usher.
651 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
653 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
656 TF/02 Fix a couple more cases where we did not log the error message
657 when unlink() failed. See also change 4.74-TF/03.
659 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
660 lock up or crash if it happened to be inside a call to libc when it
661 got a SIGUSR1 from exiwhat.
663 The SIGUSR1 handler appends the current process status to the process
664 log which is later printed by exiwhat. It used to use the general
665 purpose logging code to do this, but several functions it calls are
666 not safe for signals.
668 The new output code in the SIGUSR1 handler is specific to the process
669 log, and simple enough that it's easy to inspect for signal safety.
670 Removing some special cases also simplifies the general logging code.
671 Removing the spurious timestamps from the process log simplifies
674 TF/04 Improved ratelimit ACL condition.
676 The /noupdate option has been deprecated in favour of /readonly which
677 has clearer semantics. The /leaky, /strict, and /readonly update modes
678 are mutually exclusive. The update mode is no longer included in the
679 database key; it just determines when the database is updated. (This
680 means that when you upgrde Exim will forget old rate measurements.)
682 Exim now checks that the per_* options are used with an update mode that
683 makes sense for the current ACL. For example, when Exim is processing a
684 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
685 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
686 must specify per_mail/readonly. If you omit the update mode it defaults to
687 /leaky where that makes sense (as before) or /readonly where required.
689 The /noupdate option is now undocumented but still supported for
690 backwards compatibility. It is equivalent to /readonly except that in
691 ACLs where /readonly is required you may specify /leaky/noupdate or
692 /strict/noupdate which are treated the same as /readonly.
694 A useful new feature is the /count= option. This is a generalization
695 of the per_byte option, so that you can measure the throughput of other
696 aggregate values. For example, the per_byte option is now equivalent
697 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
699 The per_rcpt option has been generalized using the /count= mechanism
700 (though it's more complicated than the per_byte equivalence). When it is
701 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
702 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
703 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
704 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
705 Note that using per_rcpt with a non-readonly update mode in more than
706 one ACL will cause the recipients to be double-counted. (The per_mail
707 and per_byte options don't have this problem.)
709 The handling of very low rates has changed slightly. If the computed rate
710 is less than the event's count (usually one) then this event is the first
711 after a long gap. In this case the rate is set to the same as this event's
712 count, so that the first message of a spam run is counted properly.
714 The major new feature is a mechanism for counting the rate of unique
715 events. The new per_addr option counts the number of different
716 recipients that someone has sent messages to in the last time period. It
717 behaves like per_rcpt if all the recipient addresses are different, but
718 duplicate recipient addresses do not increase the measured rate. Like
719 the /count= option this is a general mechanism, so the per_addr option
720 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
721 example, measure the rate that a client uses different sender addresses
722 with the options per_mail/unique=$sender_address. There are further
723 details in the main documentation.
725 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
727 TF/06 Removed a few PCRE remnants.
729 TF/07 Automatically extract Exim's version number from tags in the git
730 repository when doing development or release builds.
732 PP/02 Raise smtp_cmd_buffer_size to 16kB.
733 Bugzilla 879. Patch from Paul Fisher.
735 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
736 Heavily based on revision 40f9a89a from Simon Arlott's tree.
739 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
741 PP/05 Variable $av_failed, true if the AV scanner deferred.
742 Bugzilla 1078. Patch from John Horne.
744 PP/06 Stop make process more reliably on build failure.
745 Bugzilla 1087. Patch from Heiko Schlittermann.
747 PP/07 Make maildir_use_size_file an _expandable_ boolean.
748 Bugzilla 1089. Patch from Heiko Schlittermann.
750 PP/08 Handle ${run} returning more data than OS pipe buffer size.
751 Bugzilla 1131. Patch from Holger Weiß.
753 PP/09 Handle IPv6 addresses with SPF.
754 Bugzilla 860. Patch from Wolfgang Breyha.
756 PP/10 GnuTLS: support TLS 1.2 & 1.1.
758 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
761 PP/11 match_* no longer expand right-hand-side by default.
762 New compile-time build option, EXPAND_LISTMATCH_RHS.
763 New expansion conditions, "inlist", "inlisti".
765 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
767 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
769 PP/14 fix log_write() format string regression from TF/03.
770 Bugzilla 1152. Patch from Dmitry Isaikin.
776 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
778 PP/02 Harmonised TLS library version reporting; only show if debugging.
779 Layout now matches that introduced for other libraries in 4.74 PP/03.
781 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
783 PP/04 New "dns_use_edns0" global option.
785 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
788 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
789 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
791 TK/01 Updated PolarSSL code to 0.14.2.
792 Bugzilla 1097. Patch from Andreas Metzler.
794 PP/07 Catch divide-by-zero in ${eval:...}.
797 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
800 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
801 format-string attack -- SECURITY: remote arbitrary code execution.
803 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
804 time unintentionally subject to list matching rules, letting the header
805 cause arbitrary Exim lookups (of items which can occur in lists, *not*
806 arbitrary string expansion). This allowed for information disclosure.
808 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
809 INT_MIN/-1 -- value coerced to INT_MAX.
815 NM/01 Workround for PCRE version dependency in version reporting
818 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
819 This fixes portability to compilers other than gcc, notably
820 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
822 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
823 makefiles for portability to HP-UX and POSIX correctness.
825 PP/01 Permit LOOKUP_foo enabling on the make command-line.
826 Also via indented variable definition in the Makefile.
827 (Debugging by Oliver Heesakkers).
829 PP/02 Restore caching of spamd results with expanded spamd_address.
830 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
832 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
833 Improves build reliability. Fix from: Frank Elsner
835 NM/02 Fix wide character breakage in the rfc2047 coding
836 Fixes bug 1064. Patch from Andrey N. Oktyabrski
838 NM/03 Allow underscore in dnslist lookups
839 Fixes bug 1026. Patch from Graeme Fowler
841 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
842 Code patches from Adam Ciarcinski of NetBSD.
844 NM/04 Fixed exiqgrep to cope with mailq missing size issue
847 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
848 is logged, to avoid truncation. Patch from John Horne.
850 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
851 Patch from Jakob Hirsch.
853 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
854 SQL string expansion failure details.
855 Patch from Andrey Oktyabrski.
857 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
858 Patch from Simon Arlott.
860 PP/09 New lookups functionality failed to compile on old gcc which rejects
861 extern declarations in function scope.
862 Patch from Oliver Fleischmann
864 PP/10 Use sig_atomic_t for flags set from signal handlers.
865 Check getgroups() return and improve debugging.
866 Fixed developed for diagnosis in bug 927 (which turned out to be
869 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
870 Patch from Mark Zealey.
872 PP/12 Bugzilla 1056: Improved spamd server selection.
873 Patch from Mark Zealey.
875 PP/13 Bugzilla 1086: Deal with maildir quota file races.
876 Based on patch from Heiko Schlittermann.
878 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
879 Patch from Uwe Doering, sign-off by Michael Haardt.
881 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
882 variable declaration deep within a block. Bug and patch from
885 PP/15 lookups-Makefile IRIX compatibilty coercion.
887 PP/16 Make DISABLE_DKIM build knob functional.
889 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
890 Patch by Simon Arlott
892 TF/03 Fix valgrind.h portability to C89 compilers that do not support
893 variable argument macros. Our copy now differs from upstream.
899 TF/01 Failure to get a lock on a hints database can have serious
900 consequences so log it to the panic log.
902 TF/02 Log LMTP confirmation messages in the same way as SMTP,
903 controlled using the smtp_confirmation log selector.
905 TF/03 Include the error message when we fail to unlink a spool file.
907 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
908 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
909 for maintaining out-of-tree patches for some time.
911 PP/01 Bugzilla 139: Documentation and portability issues.
912 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
913 Handle per-OS dynamic-module compilation flags.
915 PP/02 Let /dev/null have normal permissions.
916 The 4.73 fixes were a little too stringent and complained about the
917 permissions on /dev/null. Exempt it from some checks.
918 Reported by Andreas M. Kirchwitz.
920 PP/03 Report version information for many libraries, including
921 Exim version information for dynamically loaded libraries. Created
922 version.h, now support a version extension string for distributors
923 who patch heavily. Dynamic module ABI change.
925 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
926 privilege escalation vulnerability whereby the Exim run-time user
927 can cause root to append content of the attacker's choosing to
930 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
933 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
934 If dropping privileges for untrusted macros, we disabled normal logging
935 on the basis that it would fail; for the Exim run-time user, this is not
936 the case, and it resulted in successful deliveries going unlogged.
937 Fixed. Reported by Andreas Metzler.
943 PP/01 Date: & Message-Id: revert to normally being appended to a message,
944 only prepend for the Resent-* case. Fixes regression introduced in
945 Exim 4.70 by NM/22 for Bugzilla 607.
947 PP/02 Include check_rfc2047_length in configure.default because we're seeing
948 increasing numbers of administrators be bitten by this.
950 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
952 PP/03 Bugzilla 994: added openssl_options main configuration option.
954 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
956 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
958 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
960 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
961 without a peer certificate, leading to a segfault because of an
962 assumption that peers always have certificates. Be a little more
963 paranoid. Problem reported by Martin Tscholak.
965 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
966 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
967 NB: ClamAV planning to remove STREAM in "middle of 2010".
968 CL also introduces -bmalware, various -d+acl logging additions and
969 more caution in buffer sizes.
971 PP/09 Implemented reverse_ip expansion operator.
973 PP/10 Bugzilla 937: provide a "debug" ACL control.
975 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
977 PP/12 Bugzilla 973: Implement --version.
979 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
981 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
983 PP/15 Bugzilla 816: support multiple condition rules on Routers.
985 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
986 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
987 ignore trailing whitespace.
989 JJ/02 prevent non-panic DKIM error from being sent to paniclog
991 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
994 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
995 Notification from Dr Andrew Aitchison.
997 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
998 ExtendedDetectionInfo response format.
999 Notification from John Horne.
1001 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
1004 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
1005 XSL and documented dependency on system catalogs, with examples of how
1008 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
1011 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
1012 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
1013 configuration file which is writeable by the Exim user or group.
1015 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1016 of configuration files to cover files specified with the -C option if
1017 they are going to be used with root privileges, not just the default
1020 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1021 option (effectively making it always true).
1023 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1024 files to be used while preserving root privileges.
1026 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1027 that rogue child processes cannot use them.
1029 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1030 run-time user, instead of root.
1032 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1033 Exim run-time user without dropping privileges.
1035 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1036 result string, instead of calling string_vformat() twice with the same
1039 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1040 for other users. Others should always drop root privileges if they use
1041 -C on the command line, even for a whitelisted configure file.
1043 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1045 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1051 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1052 $data_path, and $header_path variables; fixed documentation bugs and
1055 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1056 exipick to access non-standard spools, including the "frozen" queue
1059 NM/01 Bugzilla 965: Support mysql stored procedures.
1060 Patch from Alain Williams
1062 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1064 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1065 Patch from Andreas Metzler
1067 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1068 Patch from Kirill Miazine
1070 NM/05 Bugzilla 671: Added umask to procmail example.
1072 JJ/03 installed exipick 20100323.0, fixing doc bug
1074 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1075 directory. Notification and patch from Dan Rosenberg.
1077 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1079 TK/02 Improve log output when DKIM signing operation fails.
1081 MH/01 Treat the transport option dkim_domain as a colon separated
1082 list, not as a single string, and sign the message with each element,
1083 omitting multiple occurences of the same signer.
1085 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1086 Bugzilla 985, 986. Patch by Simon Arlott
1088 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1089 Patch by Simon Arlott
1091 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1092 MBX locking. Notification from Dan Rosenberg.
1098 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1100 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1102 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1104 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1106 NM/04 Bugzilla 915: Flag broken perl installation during build.
1112 TK/01 Added patch by Johannes Berg that expands the main option
1113 "spamd_address" if it starts with a dollar sign.
1115 TK/02 Write list of recipients to X-Envelope-Sender header when building
1116 the mbox-format spool file for content scanning (suggested by Jakob
1119 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1120 (http://www.dcc-servers.net/) support via dccifd. Activated by
1121 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1123 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1124 by Mark Daniel Reidel <mr@df.eu>.
1126 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1127 When building exim an external PCRE library is now needed -
1128 PCRE is a system library on the majority of modern systems.
1129 See entry on PCRE_LIBS in EDITME file.
1131 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1132 conversation. Added nologin parameter to request.
1133 Patch contributed by Kirill Miazine.
1135 TF/01 Do not log submission mode rewrites if they do not change the address.
1137 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1139 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1140 log files in place. Contributed by Roberto Lima.
1142 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1144 TF/03 Bugzilla 615: When checking the local_parts router precondition
1145 after a local_part_suffix or local_part_prefix option, Exim now
1146 does not use the address's named list lookup cache, since this
1147 contains cached lookups for the whole local part.
1149 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1150 Robert Millan. Documentation is in experimental-spec.txt.
1152 TF/04 Bugzilla 668: Fix parallel build (make -j).
1154 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1156 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1157 Patch provided by Jan Srzednicki.
1159 TF/05 Leading white space used to be stripped from $spam_report which
1160 wrecked the formatting. Now it is preserved.
1162 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1163 that they are available at delivery time.
1165 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1167 TF/08 TLS error reporting now respects the incoming_interface and
1168 incoming_port log selectors.
1170 TF/09 Produce a more useful error message if an SMTP transport's hosts
1171 setting expands to an empty string.
1173 NM/06 Bugzilla 744: EXPN did not work under TLS.
1174 Patch provided by Phil Pennock.
1176 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1177 Patch provided by Richard Godbee.
1179 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1180 acl_smtp_notquit, added index entry.
1182 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1183 Patch provided by Eugene Bujak.
1185 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1186 accept(). Patch provided by Maxim Dounin.
1188 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1189 Patch provided by Phil Pennock.
1191 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1193 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1194 Patch provided by Brad "anomie" Jorsch.
1196 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1197 Patch provided by Dean Brooks.
1199 TK/05 Add native DKIM support (does not depend on external libraries).
1201 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1202 Patch provided by Graeme Fowler.
1204 NM/16 Bugzilla 851: Documentation example syntax fix.
1206 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1208 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1211 NM/19 Bugzilla 745: TLS version reporting.
1212 Patch provided by Phil Pennock.
1214 NM/20 Bugzilla 167: bool: condition support.
1215 Patch provided by Phil Pennock.
1217 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1218 clients. Patch provided by Phil Pennock.
1220 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1221 Patch provided by Brad "anomie" Jorsch.
1223 NM/23 Bugzilla 687: Fix misparses in eximstats.
1224 Patch provided by Heiko Schlittermann.
1226 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1227 Patch provided by Heiko Schlittermann.
1229 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1230 plus update to original patch.
1232 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1234 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1235 Patch provided by David Brownlee.
1237 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1239 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1241 NM/30 Bugzilla 888: TLS documentation bugfixes.
1243 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1245 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1246 Unlike the original bugzilla I have changed all shell scripts in src tree.
1248 NM/33 Bugzilla 898: Transport filter timeout fix.
1249 Patch by Todd Rinaldo.
1251 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1252 Patch by Serge Demonchaux.
1254 NM/35 Bugzilla 39: Base64 decode bug fixes.
1255 Patch by Jakob Hirsch.
1257 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1259 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1261 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1263 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1269 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1270 ALT-N's libdkim that I have put here:
1271 http://duncanthrax.net/exim-experimental/
1273 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1274 were called 'true' and it seems that C99 defines that as a reserved
1275 keyword to be used with 'bool' variable types. That means you could
1276 not include C99-style headers which use bools without triggering
1277 build errors in sieve.c.
1279 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1280 as mailq or other aliases. Changed the --help handling significantly
1281 to do whats expected. exim_usage() emits usage/help information.
1283 SC/01 Added the -bylocaldomain option to eximstats.
1285 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1287 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1289 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1295 PH/01 Another patch from the Sieve maintainer.
1297 PH/02 When an IPv6 address is converted to a string for single-key lookup
1298 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1299 dots are used instead of colons so that keys in lsearch files need not
1300 contain colons. This was done some time before quoting was made available
1301 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1302 (notated using the quote facility) so as to distinguish them from IPv4
1303 keys. This meant that lookups for IP addresses in host lists did not work
1304 for iplsearch lookups.
1306 This has been fixed by arranging for IPv6 addresses to be expressed with
1307 colons if the lookup type is iplsearch. This is not incompatible, because
1308 previously such lookups could never work.
1310 The situation is now rather anomolous, since one *can* have colons in
1311 ordinary lsearch keys. However, making the change in all cases is
1312 incompatible and would probably break a number of configurations.
1314 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1317 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1318 conversion specification without a maximum field width, thereby enabling
1319 a rogue spamd server to cause a buffer overflow. While nobody in their
1320 right mind would setup Exim to query an untrusted spamd server, an
1321 attacker that gains access to a server running spamd could potentially
1322 exploit this vulnerability to run arbitrary code as the Exim user.
1324 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1325 $primary_hostname instead of what libspf2 thinks the hosts name is.
1327 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1328 a directory entry by the name of the lookup key. Previously, if a
1329 symlink pointed to a non-existing file or a file in a directory that
1330 Exim lacked permissions to read, a lookup for a key matching that
1331 symlink would fail. Now it is enough that a matching directory entry
1332 exists, symlink or not. (Bugzilla 503.)
1334 PH/03 The body_linecount and body_zerocount variables are now exported in the
1337 PH/04 Added the $dnslist_matched variable.
1339 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1340 This means they are set thereafter only if the connection becomes
1343 PH/06 Added the client_condition to authenticators so that some can be skipped
1344 by clients under certain conditions.
1346 PH/07 The error message for a badly-placed control=no_multiline_responses left
1347 "_responses" off the end of the name.
1349 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1351 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1352 (without spaces) instead of just copying the configuration text.
1354 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1356 PH/11 Added $max_received_linelength.
1358 PH/12 Added +ignore_defer and +include_defer to host lists.
1360 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1361 way in which PCRE > 7.0 is built.
1363 PH/14 Implemented queue_only_load_latch.
1365 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1366 MAIL command. The effect was to mangle the value on 64-bit systems.
1368 PH/16 Another patch from the Sieve maintainer.
1370 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1372 PH/18 If a system quota error occurred while trying to create the file for
1373 a maildir delivery, the message "Mailbox is full" was not appended to the
1374 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1375 only to a quota excession during the actual writing of the file.
1377 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1378 characters?) which causes problems in log lines. The DN values are now
1379 passed through string_printing() before being added to log lines.
1381 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1382 and InterBase are left for another time.)
1384 PH/21 Added message_body_newlines option.
1386 PH/22 Guard against possible overflow in moan_check_errorcopy().
1388 PH/23 POSIX allows open() to be a macro; guard against that.
1390 PH/24 If the recipient of an error message contained an @ in the local part
1391 (suitably quoted, of course), incorrect values were put in $domain and
1392 $local_part during the evaluation of errors_copy.
1398 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1399 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1402 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1403 issue a MAIL command.
1405 PH/02 In an ACL statement such as
1407 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1409 if a client was not listed at all, or was listed with a value other than
1410 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1411 the condition was not true (as it should be), so access was not denied.
1412 The bug was that the ! inversion was incorrectly passed on to the second
1413 item. This has been fixed.
1415 PH/03 Added additional dnslists conditions == and =& which are different from
1416 = and & when the dns lookup returns more than one IP address.
1418 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1419 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1421 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1422 FSYNC, which compiles an option called disable_fsync that allows for
1423 bypassing fsync(). The documentation is heavily laced with warnings.
1425 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1427 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1428 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1429 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1430 including adding "make clean"; (3) Added -fPIC when compiling the test
1431 dynamically loaded module, to get rid of a warning.
1433 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1434 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1435 = 0s. The bug is otherwise harmless.
1437 PH/07 There was a bug in the dovecot authenticator such that the value of
1438 $auth1 could be overwritten, and so not correctly preserved, after a
1439 successful authentication. This usually meant that the value preserved by
1440 the server_setid option was incorrect.
1442 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1444 PH/09 Installed PCRE release 7.0.
1446 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1447 run for batched SMTP input. It is now run at the start of every message
1448 in the batch. While fixing this I discovered that the process information
1449 (output by running exiwhat) was not always getting set for -bs and -bS
1450 input. This is fixed, and it now also says "batched" for BSMTP.
1452 PH/11 Added control=no_pipelining.
1454 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1455 patch, slightly modified), and move the expansion of helo_data till after
1456 the connection is made in the smtp transport (so it can use these
1459 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1461 PH/14 Added log_selector = +pid.
1463 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1465 PH/16 Add ${if forany and ${if forall.
1467 PH/17 Added dsn_from option to vary the From: line in DSNs.
1469 PH/18 Flush SMTP output before performing a callout, unless control =
1470 no_callout_flush is set.
1472 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1473 was true (the default) a successful delivery failed to delete the retry
1474 item, thus causing premature timeout of the address. The bug is now
1477 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1479 PH/21 Long custom messages for fakedefer and fakereject are now split up
1480 into multiline reponses in the same way that messages for "deny" and
1481 other ACL rejections are.
1483 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1484 with slight modification.
1486 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1487 draft, changing the syntax and factoring some duplicate code".
1489 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1490 for deliveries of the second and subsequent messages over the same SMTP
1493 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1494 ${reduce, with only minor "tidies".
1496 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1498 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1499 expansion side effects.
1501 PH/27 When a message times out after an over-quota error from an Exim-imposed
1502 quota, the bounce message says "mailbox is full". This message was not
1503 being given when it was a system quota that was exceeded. It now should
1506 MH/03 Made $recipients available in local_scan(). local_scan() already has
1507 better access to the recipient list through recipients_list[], but
1508 $recipients can be useful in postmaster-provided expansion strings.
1510 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1511 in the case of a MAIL command with additional options following the
1512 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1513 were accidentally chopped off.
1515 PH/29 SMTP synchronization checks are implemented when a command is read -
1516 there is a check that no more input is waiting when there shouldn't be
1517 any. However, for some commands, a delay in an ACL can mean that it is
1518 some time before the response is written. In this time, more input might
1519 arrive, invalidly. So now there are extra checks after an ACL has run for
1520 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1521 pipelining has not been advertised.
1523 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1525 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1526 returned in the SMTP response when smtp_return_error_details was false.
1527 This has been fixed.
1529 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1530 instead of the C I/O that was originally supplied, because problems were
1531 reported on Solaris.
1533 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1534 Exim which did not show up earlier: it was assuming that a call to
1535 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1536 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1537 was a macro that became an assignment, so it seemed to work. This has
1538 changed to a proper function call with a void return, hence the compile
1539 error. Exim's code has been fixed.
1541 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1544 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1546 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1548 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1549 because it turns out to be too big or there is a timeout) there is no
1550 "Completed" line in the log. When some messages of this type were
1551 selected by exigrep, they were listed as "not completed". Others were
1552 picked up by some special patterns. I have improved the selection
1553 criteria to be more general.
1555 PH/38 The host_find_failed option in the manualroute router can now be set
1556 to "ignore", to completely ignore a host whose IP address cannot be
1557 found. If all hosts are ignored, the behaviour is controlled by the new
1558 host_all_ignored option.
1560 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1561 homing or because of multiple MX records with /mx) generated more than
1562 one IP address, and the following item turned out to be the local host,
1563 all the secondary addresses of the first item were incorrectly removed
1564 from the list, along with the local host and any following hosts (which
1565 is what is supposed to happen).
1567 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1568 whoever called Exim into the -H file. In the case of the daemon it was
1569 behaving confusingly. When first started, it used values for whoever
1570 started the daemon, but after a SIGHUP it used the Exim user (because it
1571 calls itself on a restart). I have changed the code so that it now always
1574 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1575 message are rejected with the same error (e.g. no authentication or bad
1576 sender address), and a DATA command is nevertheless sent (as can happen
1577 with PIPELINING or a stupid MUA), the error message that was given to the
1578 RCPT commands is included in the rejection of the DATA command. This is
1579 intended to be helpful for MUAs that show only the final error to their
1582 PH/42 Another patch from the Sieve maintainer.
1584 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1585 Eximstats - Fixed some broken HTML links and added missing column headers
1587 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1588 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1590 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1592 PH/43 Yet another patch from the Sieve maintainer.
1594 PH/44 I found a way to check for a TCP/IP connection going away before sending
1595 the response to the final '.' that terminates a message, but only in the
1596 case where the client has not sent further data following the '.'
1597 (unfortunately, this is allowed). However, in many cases there won't be
1598 any further data because there won't be any more messages to send. A call
1599 to select() can be used: if it shows that the input is "ready", there is
1600 either input waiting, or the socket has been closed. An attempt to read
1601 the next input character can distinguish the two cases. Previously, Exim
1602 would have sent an OK response which the client would never have see.
1603 This could lead to message repetition. This fix should cure that, at
1604 least in a lot of common cases.
1606 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1607 advertised in response to EHLO.
1613 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1614 fixed by 4.65/MH/01 (is this a record?) are fixed:
1616 (i) An empty string was always treated as zero by the numeric comparison
1617 operators. This behaviour has been restored.
1619 (ii) It is documented that the numeric comparison operators always treat
1620 their arguments as decimal numbers. This was broken in that numbers
1621 starting with 0 were being interpreted as octal.
1623 While fixing these problems I realized that there was another issue that
1624 hadn't been noticed. Values of message_size_limit (both the global option
1625 and the transport option) were treated as octal if they started with 0.
1626 The documentation was vague. These values are now always treated as
1627 decimal, and I will make that clear in the documentation.
1633 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1634 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1637 MH/01 Don't check that the operands of numeric comparison operators are
1638 integers when their expansion is in "skipping" mode (fixes bug
1639 introduced by 4.64-PH/07).
1641 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1642 child addresses, Exim now panics and dies. Previously, because the count
1643 is held in a short int, deliveries were likely to be lost. As such a
1644 large number of recipients for a single message is ridiculous
1645 (performance will be very, very poor), I have chosen to impose a limit
1646 rather than extend the field.
1652 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1653 leftover -K file (the existence of which was triggered by #402).
1654 While we were at it, introduced process PID as part of the -K
1655 filename. This should rule out race conditions when creating
1658 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1659 processing considerably. Previous code took too long for large mails,
1660 triggering a timeout which in turn triggers #401.
1662 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1663 in the DK code in transports.c. sendfile() is not really portable,
1664 hence the _LINUX specificness.
1666 TF/01 In the add_headers option to the mail command in an Exim filter,
1667 there was a bug that Exim would claim a syntax error in any
1668 header after the first one which had an odd number of characters
1671 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1672 callout verification, Exim cached a "reject" for the entire domain. This
1673 is correct for most verifications, but it is not correct for a recipient
1674 verification with use_sender or use_postmaster set, because in that case
1675 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1676 case of MAIL FROM:<> rejection from other early rejections (e.g.
1677 rejection of HELO). When verifying a recipient using a non-null MAIL
1678 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1679 Whatever the result of the callout, the value of the domain cache is
1680 left unchanged (for any other kind of callout, getting as far as trying
1681 RCPT means that the domain itself is ok).
1683 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1686 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1687 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1688 session, the daemon ignores ECONNECT errors and logs others; it now
1689 ignores EPIPE as well.
1691 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1692 (quoted-printable decoding).
1694 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1695 later the small subsequent patch to fix an introduced bug.
1697 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1699 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1701 PH/08 An error is now given if message_size_limit is specified negative.
1703 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1704 to be given (somewhat) arbitrary names.
1706 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1709 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1710 miscellaneous code fixes
1712 PH/10 Added the log_reject_target ACL modifier to specify where to log
1715 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1716 hostname. This is wrong, because it relates to the incoming message (and
1717 probably the interface on which it is arriving) and not to the outgoing
1718 callout (which could be using a different interface). This has been
1719 changed to use the value of the helo_data option from the smtp transport
1720 instead - this is what is used when a message is actually being sent. If
1721 there is no remote transport (possible with a router that sets up host
1722 addresses), $smtp_active_hostname is used.
1724 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1725 tweaks were necessary in order to get it to work (see also 21 below):
1726 (a) The code assumed that strncpy() returns a negative number on buffer
1727 overflow, which isn't the case. Replaced with Exim's string_format()
1729 (b) There were several signed/unsigned issues. I just did the minimum
1730 hacking in of casts. There is scope for a larger refactoring.
1731 (c) The code used strcasecmp() which is not a standard C function.
1732 Replaced with Exim's strcmpic() function.
1733 (d) The code set only $1; it now sets $auth1 as well.
1734 (e) A simple test gave the error "authentication client didn't specify
1735 service in request". It would seem that Dovecot has changed its
1736 interface. Fortunately there's a specification; I followed it and
1737 changed what the client sends and it appears to be working now.
1739 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1742 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1743 address A is aliased to B and C, where B exists and C does not. Without
1744 -v the output is "A verified" because verification stops after a
1745 successful redirection if more than one address is generated. However,
1746 with -v the child addresses are also verified. Exim was outputting "A
1747 failed to verify" and then showing the successful verification for C,
1748 with its parentage. It now outputs "B failed to verify", showing B's
1749 parentage before showing the successful verification of C.
1751 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1752 look up a TXT record in a specific list after matching in a combined
1755 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1756 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1757 they consult the DNS. I had assumed they would set it the way they
1758 wanted; and indeed my experiments on Linux seem to show that in some
1759 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1760 To be on the safe side, however, I have now made the interface to
1761 host_find_byname() similar to host_find_bydns(), with an argument
1762 containing the DNS resolver options. The host_find_byname() function now
1763 sets these options at its start, just as host_find_bydns() does. The smtp
1764 transport options dns_qualify_single and dns_search_parents are passed to
1765 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1766 of host_find_byname() use the default settings of RES_DEFNAMES
1767 (qualify_single) but not RES_DNSRCH (search_parents).
1769 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1770 spool_read_header() do less string testing, by means of a preliminary
1771 switch on the second character of optional "-foo" lines. (This is
1772 overdue, caused by the large number of possibilities that now exist.
1773 Originally there were few.) While I was there, I also converted the
1774 str(n)cmp tests so they don't re-test the leading "-" and the first
1775 character, in the hope this might squeeze out yet more improvement.
1777 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1778 flag allowing group syntax was set by the header_syntax check but not
1779 turned off, possible causing trouble later; (2) The flag was not being
1780 set at all for the header_verify test, causing "group"-style headers to
1781 be rejected. I have now set it in this case, and also caused header_
1782 verify to ignore an empty address taken from a group. While doing this, I
1783 came across some other cases where the code for allowing group syntax
1784 while scanning a header line wasn't quite right (mostly, not resetting
1785 the flag correctly in the right place). These bugs could have caused
1786 trouble for malformed header lines. I hope it is now all correct.
1788 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1789 with the "reply" argument non-NULL. The code, however (which originally
1790 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1791 but it didn't always do it. This confused somebody who was copying the
1792 code for some other use. I have removed all the tests.
1794 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1795 feature that was used to support insecure browsers during the U.S. crypto
1796 embargo. It requires special client support, and Exim is probably the
1797 only MTA that supported it -- and would never use it because real RSA is
1798 always available. This code has been removed, because it had the bad
1799 effect of slowing Exim down by computing (never used) parameters for the
1800 RSA_EXPORT functionality.
1802 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1803 authenticator to fail if there's a tab character in the incoming data
1804 (there should never be unless someone is messing about, as it's supposed
1805 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1806 if the connection is using TLS or if the remote IP is the same as the
1807 local IP, and the "valid-client-cert option" if a client certificate has
1810 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1811 authenticators. This can be used for authorization after authentication
1812 succeeds. (In the case of plaintext, it servers for both authentication
1815 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1816 if any retry times were supplied.
1818 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1819 connection, where there is no client IP address to check. In this
1820 situation, the verify now always succeeds.
1822 PH/25 Applied John Jetmore's -Mset patch.
1824 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1826 PH/27 In a string expansion for a processed (not raw) header when multiple
1827 headers of the same name were present, leading whitespace was being
1828 removed from all of them, but trailing whitespace was being removed only
1829 from the last one. Now trailing whitespace is removed from each header
1830 before concatenation. Completely empty headers in a concatenation (as
1831 before) are ignored.
1833 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1834 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1836 PH/29 [Removed. This was a change that I later backed out, and forgot to
1837 correct the ChangeLog entry (that I had efficiently created) before
1838 committing the later change.]
1840 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1841 address errors (4xx response to RCPT) over the same connection as other
1842 messages routed to the same hosts. Such deliveries are always "forced",
1843 so retry times are not inspected. This resulted in far too many retries
1844 for the affected addresses. The effect occurred only when there were more
1845 hosts than the hosts_max_try setting in the smtp transport when it had
1846 the 4xx errors. Those hosts that it had tried were not added to the list
1847 of hosts for which the message was waiting, so if all were tried, there
1848 was no problem. Two fixes have been applied:
1850 (i) If there are any address or message errors in an SMTP delivery, none
1851 of the hosts (tried or untried) are now added to the list of hosts
1852 for which the message is waiting, so the message should not be a
1853 candidate for sending over the same connection that was used for a
1854 successful delivery of some other message. This seems entirely
1855 reasonable: after all the message is NOT "waiting for some host".
1856 This is so "obvious" that I'm not sure why it wasn't done
1857 previously. Hope I haven't missed anything, but it can't do any
1858 harm, as the worst effect is to miss an optimization.
1860 (ii) If, despite (i), such a delivery is accidentally attempted, the
1861 routing retry time is respected, so at least it doesn't keep
1862 hammering the server.
1864 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1865 in ${readsocket because some servers need this prod.
1867 PH/32 Added some extra debug output when updating a wait-xxx database.
1869 PH/33 The hint "could be header name not terminated by colon", which has been
1870 given for certain expansion errors for a long time, was not being given
1871 for the ${if def:h_colon_omitted{... case.
1873 PH/34 The spec says: "With one important exception, whenever a domain list is
1874 being scanned, $domain contains the subject domain." There was at least
1875 one case where this was not true.
1877 PH/35 The error "getsockname() failed: connection reset by peer" was being
1878 written to the panic log as well as the main log, but it isn't really
1879 panic-worthy as it just means the connection died rather early on. I have
1880 removed the panic log writing for the ECONNRESET error when getsockname()
1883 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1884 runs only) independently of the message's sender address. This meant
1885 that, if the 4xx error was in fact related to the sender, a different
1886 message to the same recipient with a different sender could confuse
1887 things. In particualar, this can happen when sending to a greylisting
1888 server, but other circumstances could also provoke similar problems.
1889 I have changed the default so that the retry time for these errors is now
1890 based a combination of the sender and recipient addresses. This change
1891 can be overridden by setting address_retry_include_sender=false in the
1894 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1895 remote server are returned as part of bounce messages. This was not
1896 happening for LMTP over a pipe (the lmtp transport), but now it is the
1897 same for both kinds of LMTP.
1899 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1900 in header lines that were in fact CNAMEs. This is no longer the case.
1902 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1903 and queue runs started by the daemon processed all messages. This has
1904 been fixed so that -R and -S can now usefully be given with -q<time>.
1906 PH/40 Import PCRE release 6.7 (fixes some bugs).
1908 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1910 PH/42 Give an error if -q is specified more than once.
1912 PH/43 Renamed the variables $interface_address and $interface_port as
1913 $received_ip_address and $received_port, to make it clear that these
1914 values apply to message reception, and not to the outgoing interface when
1915 a message is delivered. (The old names remain recognized, of course.)
1917 PH/44 There was no timeout on the connect() call when using a Unix domain
1918 socket in the ${readsocket expansion. There now is.
1920 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1921 be meaningful with "accept".
1923 SC/01 Eximstats V1.43
1924 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1926 SC/02 Eximstats V1.44
1927 Use a glob alias rather than an array ref in the generated
1928 parser. This improves both readability and performance.
1930 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1931 Collect SpamAssassin and rejection statistics.
1932 Don't display local sender or destination tables unless
1933 there is data to show.
1934 Added average volumes into the top table text output.
1936 SC/04 Eximstats V1.46
1937 Collect data on the number of addresses (recipients)
1938 as well as the number of messages.
1940 SC/05 Eximstats V1.47
1941 Added 'Message too big' to the list of mail rejection
1942 reasons (thanks to Marco Gaiarin).
1944 SC/06 Eximstats V1.48
1945 Mainlog lines which have GMT offsets and are too short to
1946 have a flag are now skipped.
1948 SC/07 Eximstats V1.49 (Alain Williams)
1949 Added the -emptyok flag.
1951 SC/08 Eximstats V1.50
1952 Fixes for obtaining the IP address from reject messages.
1954 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1955 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1956 whitesspace changes from 4.64-PH/27
1958 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1961 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1962 are found, allow negative numbers in numeric criteria)
1964 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1966 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1967 to match changes made in 4.64-PH/43
1969 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1971 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1972 "domain" is actually an IP address, and force a failure. This locks out
1973 those revolvers/nameservers that support "A-for-A" lookups, in
1974 contravention of the specifications.
1976 PH/48 When a host name was looked up from an IP address, and the subsequent
1977 forward lookup of the name timed out, the host name was left in
1978 $sender_host_name, contrary to the specification.
1980 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1981 restricted to single-key lookups, Exim was not diagnosing an error if
1982 * or *@ was used with a query-style lookup.
1984 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1986 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1987 long ago, but noone interested enough thought of it. Let's just say that
1988 the "1.1" means that there are some new functions that weren't there at
1989 some point in the past.
1991 PH/51 Error processing for expansion failure of helo_data from an smtp
1992 transport during callout processing was broken.
1994 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1995 tested/used via the -bh/-bhc/-bs options.
1997 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1998 bug, fixed in subsequent PCRE releases).
2000 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
2001 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
2003 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
2009 SC/01 Use a glob alias rather than an array ref in eximstats generated
2010 parser. This improves both readability and performance.
2012 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
2013 Don't display local sender or destination tables in eximstats unless
2014 there is data to show.
2015 Added average volumes into the eximstats top table text output.
2017 SC/03 Collect data on the number of addresses (recipients) as well
2018 as the number of messages in eximstats.
2020 TF/01 Correct an error in the documentation for the redirect router. Exim
2021 does (usually) call initgroups() when daemonizing.
2023 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2024 with consistent privilege compared to when running as a daemon.
2026 TF/03 Note in the spec that $authenticated_id is not set for local
2027 submissions from trusted users.
2029 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2030 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2032 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2033 by adding some example configuration directives to the default
2034 configuration file. A little bit of work is required to uncomment the
2035 directives and define how usernames and passwords are checked, but
2036 there is now a framework to start from.
2038 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2039 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2040 without this. I don't know how relevant this is to other LDAP libraries.
2042 PH/02 Add the verb name to the "unknown ACL verb" error.
2044 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2046 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2048 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2049 directory not expanded when it should be if an expanded home directory
2050 was set for the address (which is overridden by the transport).
2052 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2055 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2056 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2057 because it is too late at that time, and has no effect.
2059 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2060 security issue with \' (bugzilla #107). I could not use the
2061 PQescapeStringConn() function, because it needs a PGconn value as one of
2064 PH/08 When testing addresses using -bt, indicate those final addresses that
2065 are duplicates that would not cause an additional delivery. At least one
2066 person was confused, thinking that -bt output corresponded to deliveries.
2067 (Suppressing duplicates isn't a good idea as you lose the information
2068 about possibly different redirections that led to the duplicates.)
2070 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2071 systems where poll() doesn't work, in particular OS X.
2073 PH/10 Added more information to debugging output for retry time not reached.
2075 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2076 operations in malware.c.
2078 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2081 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2082 syslog_duplication set false, log lines that would normally be written
2083 both the the main log and to the reject log were not written to syslog at
2086 PH/14 In the default configuration, change the use of "message" in ACL warn
2087 statements to "add_header".
2089 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2090 not followed by a command (e.g. "seen endif").
2092 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2093 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2096 PH/17 Added extra conditions to the default value of delay_warning_condition
2100 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2101 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2102 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2105 The Auto-Submitted: and various List- headers are standardised, whereas I
2106 don't think Precedence: ever was.
2108 PH/18 Refactored debugging code in route_finduser() to show more information,
2109 in particular, the error code if getpwnam() issues one.
2111 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2112 This is apparently needed in addition to the PH/07 change above to avoid
2113 any possible encoding problems.
2115 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2116 but not after initializing Perl.
2118 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2119 output them only if debugging. By default they are written stderr,
2120 apparently, which is not desirable.
2122 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2125 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2128 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2130 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2131 authenticated or an ident call has been made. Suppress the default
2132 values for $authenticated_id and $authenticated_sender (but permit -oMai
2133 and -oMas) when testing with -bh.
2135 PH/24 Re-jigged the order of the tests in the default configuration so that the
2136 tests for valid domains and recipients precede the DNS black list and CSA
2137 tests, on the grounds that those ones are more expensive.
2139 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2140 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2141 This bug exists in every version of Exim that I still have, right back to
2144 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2145 However, an attempt to turn on case-sensitivity in a regex key by
2146 including (?-i) didn't work because the subject string was already
2147 lowercased, and the effects were non-intuitive. It turns out that a
2148 one-line patch can be used to allow (?-i) to work as expected.
2154 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2155 other effects) broke the use of negated acl sub-conditions.
2157 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2160 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2161 "Deny" causes Exim to reject the incoming connection with a 554 error.
2162 Unfortunately, if there is a major crisis, such as a disk failure,
2163 tcp-wrappers gives "deny", whereas what one would like would be some
2164 kind of temporary error. A kludge has been added to help with this.
2165 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2166 554 error is used if errno is still zero or contains ENOENT (which occurs
2167 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2170 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2172 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2173 errors. Otherwise a message that provokes a temporary error (when other
2174 messages do not) can cause a whole host to time out.
2176 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2177 addresses were routed directly to files or pipes from a redirect router.
2178 File deliveries just didn't batch; pipe deliveries might have suffered
2181 PH/06 A failure to get a lock for a hints database would erroneously always say
2182 "Failed to get write lock", even when it was really a read lock.
2184 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2185 of 0600. This has been changed to use the value of the lockfile_mode
2186 option (which defaults to 0600).
2188 PH/08 Applied small patch from the Sieve maintainer.
2190 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2191 folder from quota calculations, a direct delivery into this folder messed
2192 up the contents of the maildirsize file. This was because the regex was
2193 used only to exclude .Trash (or whatever) when the size of the mailbox
2194 was calculated. There was no check that a delivery was happening into an
2195 excluded directory. This bug has been fixed by ignoring all quota
2196 processing for deliveries into excluded directories.
2198 PH/10 Added the maildirfolder_create_regex option to appendfile.
2204 PH/01 The code for finding all the local interface addresses on a FreeBSD
2205 system running IPv6 was broken. This may well have applied to all BSD
2206 systems, as well as to others that have similar system calls. The broken
2207 code found IPv4 interfaces correctly, but gave incorrect values for the
2208 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2209 that it would not match correctly against @[] and not recognize the IPv6
2212 PH/02 The ipliteral router was not recognizing addresses of the form user@
2213 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2215 PH/03 Added disable_ipv6.
2217 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2218 decoded form, because it is most often used to construct To: headers
2219 lines in autoreplies, and the decoded form may well be syntactically
2220 invalid. However, $reply_address has leading white space removed, and all
2221 newlines turned into spaces so that the autoreply transport does not
2224 PH/05 If group was specified without a user on a router, and no group or user
2225 was specified on a transport, the group from the router was ignored.
2227 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2228 for visible compile-time settings that can be used to change these
2229 numbers, for those that want even more. Backwards compatibility with old
2230 spool files has been maintained. However, going back to a previous Exim
2231 release will lost any variables that are in spool files.
2233 PH/07 Two small changes when running in the test harness: increase delay when
2234 passing a TCP/IP connection to a new process, in case the original
2235 process has to generate a bounce, and remove special handling of
2236 127.0.0.2 (sic), which is no longer necessary.
2238 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2239 be the same on different OS.
2241 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2244 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2245 whether --show-vars was specified or not
2247 JJ/02 exipick: Added support for new ACL variable spool format introduced
2250 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2251 syntactically invalid From: or Reply-to: line, and a filter used this to
2252 generate an autoreply, and therefore failed to obtain an address for the
2253 autoreply, Exim could try to deliver to a non-existent relative file
2254 name, causing unrelated and misleading errors. What now happens is that
2255 it logs this as a hard delivery error, but does not attempt to create a
2258 PH/11 The exinext utility has a -C option for testing purposes, but although
2259 the given file was scanned by exinext itself; it wasn't being passed on
2260 when Exim was called.
2262 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2263 an end-of-file indication when reading a command response.
2265 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2266 compiled. In many other places in Exim, IPv6 addresses are always
2267 recognized, so I have changed this. It also means that IPv4 domain
2268 literals of the form [IPV4:n.n.n.n] are now always recognized.
2270 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2271 used if the router is not running as root, for example, when verifying at
2272 ACL time, or when using -bh. The debugging output from this situation was
2273 non-existent - all you got was a failure to exec. I have made two
2276 (a) Failures to set uid/gid, the current directory, or a process leader
2277 in a subprocess such as that created by queryprogram now generate
2278 suitable debugging ouput when -d is set.
2280 (b) The queryprogram router detects when it is not running as root,
2281 outputs suitable debugging information if -d is set, and then runs
2282 the subprocess without attempting to change uid/gid.
2284 PH/15 Minor change to Makefile for building test_host (undocumented testing
2287 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2288 additional section of a DNS packet that returns MX or SRV records.
2289 Instead, it always explicitly searches for A/AAAA records. This avoids
2290 major problems that occur when a DNS server includes only records of one
2291 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2292 fixed another bug: if SRV records were looked up and the corresponding
2293 address records were *not* found in the additional section, the port
2294 values from the SRV records were lost.
2296 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2297 using the correct key (the original address) when searching the retry
2298 rules in order to find which one to use for generating the retry hint.
2300 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2301 adding the default one. Similarly, if it contains a Reply-To: header, the
2302 errors_reply_to option, if set, is not used.
2304 PH/19 When calculating a retry time, Exim used to measure the "time since
2305 failure" by looking at the "first failed" field in the retry record. Now
2306 it does not use this if it is later than than the arrival time of the
2307 message. Instead it uses the arrival time. This makes for better
2308 behaviour in cases where some deliveries succeed, thus re-setting the
2309 "first failed" field. An example is a quota failure for a huge message
2310 when small messages continue to be delivered. Without this change, the
2311 "time since failure" will always be short, possible causing more frequent
2312 delivery attempts for the huge message than are intended.
2313 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2315 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2316 $1, $2, $3) because the numerical variables can be reset during some
2317 expansion items (e.g. "match"), thereby losing the authentication data.
2319 PH/21 Make -bV show the size of off_t variables so that the test suite can
2320 decide whether to run tests for quotas > 2G.
2322 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2323 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2324 filecount value is greater than 2G or if a quota value is greater than 2G
2325 on a system where the size of off_t is not greater than 4, a panic error
2328 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2329 never match. The debug and -bh output now contains an explicit error
2330 message indicating a malformed IPv4 address or mask.
2332 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2333 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2334 PH/23 above applies.
2336 PH/25 Do not write to syslog when running in the test harness. The only
2337 occasion when this arises is a failure to open the main or panic logs
2338 (for which there is an explicit test).
2340 PH/26 Added the /no_tell option to "control=freeze".
2342 PH/27 If a host name lookup failed very early in a connection, for example, if
2343 the IP address matched host_lookup and the reverse lookup yielded a name
2344 that did not have a forward lookup, an error message of the form "no IP
2345 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2346 could be logged. Now it outputs the IP address instead of "NULL".
2348 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2349 allows the sender and the authenticated sender to be set when
2350 submitting a message from within Exim. Since child_open_exim() is
2351 documented for local_scan(), the new function should be too.
2353 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2354 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2355 results in an empty string is now treated as unset.
2357 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2359 PH/31 Added sender_verify_fail logging option.
2361 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2362 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2363 tidied the source and removed it altogether.
2365 PH/33 When a queue run was abandoned because the load average was too high, a
2366 log line was always written; now it is written only if the queue_run log
2367 selector is set. In addition, the log line for abandonment now contains
2368 information about the queue run such as the pid. This is always present
2369 in "start" and "stop" lines but was omitted from the "abandon" line.
2371 PH/34 Omit spaces between a header name and the colon in the error message that
2372 is given when verify = headers_syntax fails (if there are lots of them,
2373 the message gets confusing).
2375 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2376 names, as there are now some PTR records that contain slashes. This check
2377 is only to protect against broken name servers that fall over on strange
2378 characters, so the fact that it applies to all lookups doesn't matter.
2380 PH/36 Now that the new test suite is complete, we can remove some of the
2381 special code in Exim that was needed for the old test suite. For example,
2382 sorting DNS records because real resolvers return them in an arbitrary
2383 order. The new test suite's fake resolver always returns records in the
2386 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2387 bounces) except when queue_only is set, to avoid logging races between
2388 the different processes.
2390 PH/38 Panic-die if .include specifies a non-absolute path.
2392 PH/39 A tweak to the "H" retry rule from its user.
2394 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2395 a label. They prevented compilation on older perls.
2397 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2398 a warning to be raised on newish perls.
2400 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2401 on queue. Changes to match documented behaviour of showing count of
2402 messages matching specified criteria.
2404 PH/40 Changed the default ident timeout from 30s to 5s.
2406 PH/41 Added support for the use of login_cap features, on those BSD systems
2407 that have them, for controlling the resources used by pipe deliveries.
2409 PH/42 The content-scanning code uses fopen() to create files in which to put
2410 message data. Previously it was not paying any attention to the mode of
2411 the files. Exim runs with umask(0) because the rest of the code creates
2412 files with open(), and sets the required mode explicitly. Thus, these
2413 files were ending up world-writeable. This was not a big issue, because,
2414 being within the spool directory, they were not world-accessible. I have
2415 created a function called modefopen, which takes an additional mode
2416 argument. It sets umask(777), creates the file, chmods it to the required
2417 mode, then resets the umask. All the relevant calls to fopen() in the
2418 content scanning code have been changed to use this function.
2420 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2421 to 24 hours. This avoids potential overflow problems when processing G
2422 and H retry rules. I suspect nobody ever tinkers with this value.
2424 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2426 PH/45 When the plaintext authenticator is running as a client, the server's
2427 challenges are checked to ensure they are valid base64 strings. By
2428 default, the authentication attempt is cancelled if an invalid string is
2429 received. Setting client_ignore_invalid_base64 true ignores these errors.
2430 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2431 they are received. Thus, the responses can be made to depend on the
2432 challenges. If an invalid string is ignored, an empty string is placed in
2435 PH/46 Messages that are created by the autoreply transport now contains a
2436 References: header, in accordance with RFCs 2822 and 3834.
2438 PH/47 Added authenticated_sender_force to the smtp transport.
2440 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2442 PH/49 Installed latest patch from the Sieve maintainer.
2444 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2445 was also set, the appendfile transport was unnecessarily scanning a
2446 directory of message files (e.g. for maildir delivery) to find the count
2447 of files (along with the size), even though it did not need this
2448 information. It now does the scan only if it needs to find either the
2449 size of the count of files.
2451 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2453 PH/52 Two bugs concerned with error handling when the smtp transport is
2456 (i) Exim was not creating retry information for temporary errors given
2457 for individual recipients after the DATA command when the smtp transport
2458 was used in LMTP mode. This meant that they could be retried too
2459 frequently, and not timed out correctly.
2461 (ii) Exim was setting the flag that allows error details to be returned
2462 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2463 recipients that were returned after the DATA command.
2465 PH/53 This is related to PH/52, but is more general: for any failing address,
2466 when detailed error information was permitted to be returned to the
2467 sender, but the error was temporary, then after the final timeout, only
2468 "retry timeout exceeded" was returned. Now it returns the full error as
2469 well as "retry timeout exceeded".
2471 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2472 do this, and (what is worse) MTAs that accept it.
2474 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2475 will now be deprecated.
2477 PH/56 New os.c-cygwin from the Cygwin maintainer.
2479 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2480 formats (previously only available in exim formats via -bpr, -bpru,
2481 and -bpra. Now also available in native and exiqgrep formats)
2483 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2484 with very large, slow to parse queues
2486 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2488 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2490 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2491 responses to authentication challenges, though it was showing the
2492 challenges; (ii) I've removed the CR characters from the debug output for
2495 PH/58 Allow for the insertion of a newline as well as a space when a string
2496 is turned into more than one encoded-word during RFC 2047 encoding. The
2497 Sieve code now uses this.
2499 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2500 data_4xx, lost_connection, tls_required.
2502 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2503 message was being sent as an SMTP response.
2505 PH/61 Add -l and -k options to exicyclog.
2507 PH/62 When verifying, if an address was redirected to one new address, so that
2508 verification continued, and the new address failed or deferred after
2509 having set something in $address_data, the value of $address_data was not
2510 passed back to the ACL. This was different to the case when no
2511 redirection occurred. The value is now passed back in both cases.
2513 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2514 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2515 use login_cap.h, so on its own it isn't the distinguishing feature. The
2516 new name refers directly to the setclassresources() function.
2518 PH/65 Added configuration files for NetBSD3.
2520 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2522 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2523 is preferred over IPv4.
2525 PH/68 The bounce_return_message and bounce_return_body options were not being
2526 honoured for bounces generated during the reception of non-SMTP messages.
2527 In particular, this applied to messages rejected by the ACL. This bug has
2528 been fixed. However, if bounce_return_message is true and bounce_return_
2529 body is false, the headers that are returned for a non-SMTP message
2530 include only those that have been read before the error was detected.
2531 (In the case of an ACL rejection, they have all been read.)
2533 PH/69 The HTML version of the specification is now built in a directory called
2534 spec_html instead of spec.html, because the latter looks like a path with
2535 a MIME-type, and this confuses some software.
2537 PH/70 Catch two compiler warnings in sieve.c.
2539 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2540 function verify_get_ident() calls ip_connect() to connect a socket, but
2541 if the "connect()" function timed out, ip_connect() used to close the
2542 socket. However, verify_get_ident() also closes the socket later, and in
2543 between Exim writes to the log, which may get opened at this point. When
2544 the socket was closed in ip_connect(), the log could get the same file
2545 descriptor number as the socket. This naturally causes chaos. The fix is
2546 not to close the socket in ip_connect(); the socket should be closed by
2547 the function that creates it. There was only one place in the code where
2548 this was missing, in the iplookup router, which I don't think anybody now
2549 uses, but I've fixed it anyway.
2551 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2552 well as to direct DNS lookups. Otherwise the handling of names in host
2553 lists is inconsistent and therefore confusing.
2559 PH/01 Two changes to the default runtime configuration:
2561 (1) Move the checks for relay_from_hosts and authenticated clients from
2562 after to before the (commented out) DNS black list checks.
2564 (2) Add control=submission to the relay_from_hosts and authenticated
2565 clients checks, on the grounds that messages accepted by these
2566 statements are most likely to be submissions.
2568 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2570 (1) Generate an error if the third argument for the ${prvs expansion is
2573 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2576 (3) Reset the variables that are obtained from the first argument of
2577 ${prvscheck and used in the second argument before leaving the code,
2578 because their memory is reclaimed, so using them afterwards may do
2581 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2582 one (it's much easier than Tom thought :-).
2584 (5) Because of (4), we can now allow for the use of $prvscheck_result
2585 inside the third argument.
2587 PH/03 For some reason, the default setting of PATH when running a command from
2588 a pipe transport was just "/usr/bin". I have changed it to
2591 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2592 anything to be listed in the output from -bV.
2594 PH/05 When a filter generated an autoreply, the entire To: header line was
2595 quoted in the delivery log line, like this:
2597 => >A.N.Other <ano@some.domain> <original@ddress> ...
2599 This has been changed so that it extracts the operative address. There
2600 may be more than one such address. If so, they are comma-separated, like
2603 => >ano@some.domain,ona@other.domain <original@ddress> ...
2605 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2606 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2607 not being looked up in the rDNS to get a host name, Exim was showing the
2608 IP address twice in Received: lines, even though the IP addresses were
2609 identical. For example:
2611 Received: from [1.2.3.4] (helo=[1.2.3.4])
2613 However, if the real host name was known, it was omitting the HELO data
2614 if it matched the actual IP address. This has been tidied up so that it
2615 doesn't show the same IP address twice.
2617 PH/07 When both +timestamp and +memory debugging was on, the value given by
2618 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2619 called by the debug printing, thereby overwriting the timestamp buffer.
2620 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2622 PH/08 When the original message was included in an autoreply transport, it
2623 always said "this is a copy of the message, including all the headers",
2624 even if body_only or headers_only was set. It now gives an appropriate
2627 PH/09 Applied a patch from the Sieve maintainer which:
2629 o fixes some comments
2630 o adds the (disabled) notify extension core
2631 o adds some debug output for the result of if/elsif tests
2632 o points to the current vacation draft in the documentation
2633 and documents the missing references header update
2637 o fixes a bug in processing the envelope test (when testing
2638 multiple envelope elements, the last element determinted the
2641 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2642 Electronic Mail") by including:
2644 Auto-submitted: auto-generated
2646 in the messages that it generates (bounce messages and others, such as
2647 warnings). In the case of bounce messages for non-SMTP mesages, there was
2648 also a typo: it was using "Auto_submitted" (underscore instead of
2649 hyphen). Since every message generated by Exim is necessarily in response
2650 to another message, thes have all been changed to:
2652 Auto-Submitted: auto-replied
2654 in accordance with these statements in the RFC:
2656 The auto-replied keyword:
2658 - SHOULD be used on messages sent in direct response to another
2659 message by an automatic process,
2661 - MUST NOT be used on manually-generated messages,
2663 - MAY be used on Delivery Status Notifications (DSNs) and Message
2664 Disposition Notifications (MDNs),
2666 - MUST NOT be used on messages generated by automatic or periodic
2667 processes, except for messages which are automatic responses to
2670 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2671 to the default Received: header definition.
2673 PH/12 Added log selector acl_warn_skipped (default on).
2675 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2676 variables because (a) they are in the wrong storage pool and (b) even if
2677 they were copied, it wouldn't work properly because of the caching.
2679 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2680 checking when decoding. Apparently there are clients that generate
2681 overlong encoded strings. Why am I not surprised?
2683 PH/15 If the first argument of "${if match_address" was not empty, but did not
2684 contain an "@" character, Exim crashed. Now it writes a panic log message
2685 and treats the condition as false.
2687 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2689 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2690 extension "envelope-auth". The code is finished and in agreement with
2691 other implementations, but there is no documentation so far and in fact,
2692 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2693 not changing the active code.
2695 Print executed "if" and "elsif" statements when debugging is used. This
2696 helps a great deal to understand what a filter does.
2698 Document more things not specified clearly in RFC3028. I had all this
2699 sorted out, when out of a sudden new issues came to my mind. Oops."
2701 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2704 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2705 canonical form (as documented). However, after a host name lookup from
2706 the IP address, check_host() was doing a simple string comparison with
2707 addresses acquired from the DNS when checking that the found name did
2708 have the original IP as one of its addresses. Since any found IPv6
2709 addresses are likely to be in abbreviated form, the comparison could
2710 fail. Luckily, there already exists a function for doing the comparison
2711 by converting both addresses to binary, so now that is used instead of
2712 the text comparison.
2714 PH/20 There was another similar case to PH/19, when a complete host name was
2715 given in a host list; looking up its IP address could give an abbreviated
2716 form, whereas the current host's name might or might not be abbreviated.
2717 The same fix has been applied.
2723 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2724 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2727 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2728 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2730 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2732 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2733 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2734 Also, for "Auto-Submitted", treat anything other than "no" as
2735 non-personal, in accordance with RFC 3834. (Previously it treated
2736 anything starting "auto-" as non-personal.)
2738 TF/01 The control=submission/name=... option had a problem with syntax
2739 errors if the name included a slash character. The /name= option
2740 now slurps the rest of the string, so it can include any characters
2741 but it must come last in the list of options (after /sender_retain
2744 PH/05 Some modifications to the interface to the fake nameserver for the new
2752 TK/01 Added the "success_on_redirect" address verification option. See
2753 NewStuff for rationale and an example.
2755 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2757 PH/02 Patch to exigrep to allow it to work on syslog lines.
2759 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2760 fread() to skip over the body file's header line, because in Cygwin the
2761 header line is locked and is inaccessible.
2763 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2764 co-exist for some time) to make it clear that it is the Exim ID that is
2765 referenced, not the Message-ID: header line.
2767 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2768 string_format() function, because snprintf() does not exist on all
2771 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2772 ${stat: expansion item.
2774 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2775 protocol synchronization error", to keep the pedants happy.
2777 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2778 well as for IRIX systems, when gcc is being used. See the host.c source
2781 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2783 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2786 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2787 transport and to the smtp transport in LMTP mode.
2789 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2791 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2792 run a filter in a subprocess. This could lead to confusion in subsequent
2793 lookups in the parent process. There should also be a search_tidyup() at
2794 the end of the subprocess.
2796 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2797 only if the host matched helo_try_verify_hosts, which caused the
2798 verification to occur when the EHLO/HELO command was issued. The ACL just
2799 tested the remembered result. Now, if a previous verification attempt has
2800 not happened, "verify = helo" does it there and then.
2802 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2804 TK/03 Fix log output including CR from clamd.
2806 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2807 exist provoked a memory error which could cause a segfault.
2809 PH/15 Installed PCRE 6.2
2811 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2813 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2814 of the problem. Specifically, suggested +O2 rather than +O1 for the
2817 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2819 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2820 record, the port was not correctly logged when the outgoing_port log
2821 selector was set (it logged the transort's default port).
2823 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2824 fallback_hosts, and "hosts" in the smtp transport.
2826 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2827 host errors such as "Connection refused".
2829 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2830 authentication with radiusclient 0.4.9:
2832 - Error returned from rc_read_config was caught wrongly
2833 - Username/password not passed on to radius server due to wrong length.
2835 The presumption is that some radiusclient API changes for 4.51/PH/17
2836 were not taken care of correctly. The code is still untested by me (my
2837 Linux distribution still has 0.3.2 of radiusclient), but it was
2838 contributed by a Radius user.
2840 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2841 expanding the "port" option of the smtp transport.
2843 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2844 while reading a MIME header. Thanks to Tom Hughes for a patch.
2846 PH/24 Include config.h inside local_scan.h so that configuration settings are
2849 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2850 that in an ACL for RCPT (for example), you can examine exactly what was
2853 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2854 commands, but it was not correctly comparing the address with the actual
2855 client host address. Thus, it would show the EHLO address in Received:
2856 header lines when this was not necessary.
2858 PH/27 Added the % operator to ${eval:}.
2860 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2861 it should be ignoring failures (because with -C, for example, it has lost
2862 privilege). It wasn't ignoring creation failures other than "already
2865 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2868 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2869 because an input line was too long, either on its own, or by virtue of
2870 too many continuations, the temporary file was not being removed, and the
2871 return code was incorrect.
2873 PH/31 Missing "BOOL" in function definition in filtertest.c.
2875 PH/32 Applied Sieve patches from the maintainer.
2877 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2879 PH/33 Added "verify = not_blind".
2881 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2882 Local/Makefile (with some defaults set). These are used in built scripts
2883 such as exicyclog, but they have never been used in the exim_install
2884 script (though there are many overriding facilities there). I have
2885 arranged that the exim_install script now takes note of these two
2888 PH/35 Installed configuration files for Dragonfly.
2890 PH/36 When a locally submitted message by a trusted user did not contain a
2891 From: header, and the sender address was obtained from -f or from an SMTP
2892 MAIL command, and the trusted user did not use -F to supply a sender
2893 name, $originator_name was incorrectly used when constructing a From:
2894 header. Furthermore, $originator_name was used for submission mode
2895 messages from external hosts without From: headers in a similar way,
2896 which is clearly wrong.
2898 PH/37 Added control=suppress_local_fixups.
2900 PH/38 When log_selector = +received_sender was set, and the addition of the
2901 sender made the log line's construction buffer exactly full, or one byte
2902 less than full, an overflow happened when the terminating "\n" was
2905 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2906 when the result of a list match is failure because a DNS lookup failed.
2908 PH/40 RM_COMMAND is now used in the building process.
2910 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2911 the "build-* directories that it finds.
2913 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2914 address in a domain literal was a prefix of an interface address.
2916 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2917 when verifying a sender address, unless rewrite_headers is false.
2919 PH/44 Wrote a long comment about why errors_to addresses are verified as
2920 recipients, not senders.
2922 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2923 the ratelimit ACL was added.
2925 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2927 PH/46 Added extra information about PostgreSQL errors to the error string.
2929 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2930 avoiding the need to install special zones in a real server. This is
2931 backwards compatible; if it can't find the fake resolver, it drops back.
2932 Thus, both old and new test suites can be run.
2934 TF/02 Added util/ratelimit.pl
2936 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2937 clock is set back in time.
2939 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2940 Candler <B.Candler@pobox.com>.
2942 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2943 OK for addresses that are the result of redirections.
2945 PH/48 A number of further additions for the benefit of the new test suite,
2946 including a fake gethostbyname() that interfaces to the fake DNS resolver
2949 TF/06 The fix for widen_domains has also been applied to qualify_single and
2950 search_parents which are the other dnslookup options that can cause
2953 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2956 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2958 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2959 service" if it encountered an MX record with an empty target hostname.
2960 The message is now "an MX or SRV record indicated no SMTP service".
2962 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2963 if the DNS of the sending site is misconfigured. This is quite a
2964 common situation. This change restores the behaviour of treating a
2965 helo verification defer as a failure.
2967 PH/51 If self=fail was set on a router, the bounce message did not include the
2968 actual error message.
2974 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2976 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2977 command itself ended in error, the underlying message about the transport
2978 filter timeout was being overwritten with the pipe command error. Now the
2979 underlying error message should be appended to the second error message.
2981 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2983 PH/02 Reduce the amount of output that "make" produces by default. Full output
2984 can still be requested.
2986 PH/03 The warning log line about a condition test deferring for a "warn" verb
2987 was being output only once per connection, rather than after each
2988 occurrence (because it was using the same function as for successful
2989 "warn" verbs). This seems wrong, so I have changed it.
2991 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2992 it should not have, which might have caused a crash in the right
2993 circumstances, but probably never did.
2995 PH/04 Installed a modified version of Tony Finch's patch to make submission
2996 mode fix the return path as well as the Sender: header line, and to
2997 add a /name= option so that you can make the user's friendly name appear
3000 TF/03 Added the control = fakedefer ACL modifier.
3002 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
3003 Mark Lowes for thorough testing.
3005 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
3007 TK/03 Merged latest SRS patch from Miles Wilton.
3009 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
3010 with the definition in sysexits.h (which is #included earlier).
3011 Fortunately, Exim does not actually use EX_OK. The code used to try to
3012 preserve the sysexits.h value, by assumimg that macro definitions were
3013 scanned for macro replacements. I have been disabused of this notion,
3014 so now the code just undefines EX_OK before #including unistd.h.
3016 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3017 in the smtp transport. When a block could not be written in a single
3018 write() function, the timeout was being re-applied to each part-write.
3019 This seems wrong - if the receiver was accepting one byte at a time it
3020 would take for ever. The timeout is now adjusted when this happens. It
3021 doesn't have to be particularly precise.
3023 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3024 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3026 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3027 without a domain if the check to <postmaster@domain> fails.
3029 SC/01 Eximstats: added -xls and the ability to specify output files
3030 (patch written by Frank Heydlauf).
3032 SC/02 Eximstats: use FileHandles for outputing results.
3034 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3036 SC/04 Eximstats: fixed display of large numbers with -nvr option
3038 SC/05 Eximstats: fixed merging of reports with empty tables.
3040 SC/06 Eximstats: added the -include_original_destination flag
3042 SC/07 Eximstats: removed tabs and trailing whitespace.
3044 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3046 TK/06 MBOX spool code: Add real "From " MBOX separator line
3047 so the .eml file is really in mbox format (even though
3048 most programs do not really care). Patch from Alex Miller.
3050 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3051 The latter is generated from $received_to and is only set if the
3052 message has one envelope recipient. SA can use these headers,
3053 obviously out-of-the-box. Patch from Alex Miller.
3055 PH/08 The ${def test on a variable was returning false if the variable's
3056 value was "0", contrary to what the specification has always said!
3057 The result should be true unless the variable is empty.
3059 PH/09 The syntax error of a character other than { following "${if
3060 def:variable_name" (after optional whitespace) was not being diagnosed.
3061 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3062 accidental colon was present, for example, could give incorrect results.
3064 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3065 result is used (not including appendfile, where other changes are about
3068 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3069 This involved changing a lot of size variables from int to off_t. It
3070 should work with maildirs and everything.
3072 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3073 spamd dying while we are connected to it.
3075 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3078 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3079 function for BDB 4.3.
3081 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3083 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3084 the sources for PCRE have been re-arranged and more files are now
3087 PH/15 The code I had for printing potentially long long variables in PH/11
3088 above was not the best (it lost precision). The length of off_t variables
3089 is now inspected at build time, and an appropriate printing format (%ld
3090 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3091 to be "long long int" or "long int". This is needed for the internal
3092 formatting function string_vformat().
3094 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3095 the configuration file to be ":syslog", then the script "guesses" where
3096 the logs files are, rather than using the compiled in default. In our
3097 case the guess is not the same as the compiled default, so the script
3098 suddenly stopped working when I started to use syslog. The patch checks
3099 to see if log_file_path is "". If so, it attempts to read it from exim
3100 with no configuration file to get the compiled in version, before it
3101 falls back to the previous guessing code."
3103 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3104 implementing BATV in an Exim configuration. See NewStuff for the gory
3107 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3108 Makefile that are specific to HP-UX.
3110 PH/18 If the "use_postmaster" option was set for a recipient callout together
3111 with the "random" option, the postmaster address was used as the MAIL
3112 FROM address for the random test, but not for the subsequent recipient
3113 test. It is now used for both.
3115 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3116 patch removes a few documentation additions to RFC 3028, because the
3117 latest draft now contains them. It adds the new en;ascii-case comparator
3118 and a new error check for 8bit text in MIME parts. Comparator and
3119 require names are now matched exactly. I enabled the subaddress
3120 extension, but it is not well tested yet (read: it works for me)."
3122 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3123 rework some of the code of TK/09 above to avoid the hardwired use of
3124 "%lld" and "long long". Replaced the call to snprintf() with a call to
3127 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3128 records point to non-existent hosts", "retry timeout exceeded", and
3129 "retry time not reached for any host after a long failure period".
3131 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3132 experimental DomainKeys support:
3134 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3135 (2) On an error such as an illegally used "control", the wrong name for
3136 the control was given.
3138 These problems did NOT occur unless DomainKeys support was compiled.
3140 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3142 PH/24 Added ${if match_ip condition.
3144 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3145 databases so that it will be absolutely obvious if a crash occurs in the
3146 DB library. This is a regular occurrence (often caused by mis-matched
3149 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3150 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3151 picked up on a user's system that detects such things. There doesn't seem
3152 to be a gcc warning option for this - only an attribute that has to be
3153 put on the function's prototype. It seems that in Fedora Core 4 they have
3154 set this on a number of new functions. No doubt there will be more in due
3157 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3158 specify a transport. However, if an address that was verified by such a
3159 router was the subject of a callout, Exim crashed because it tried to
3160 read the rcpt_include_affixes from the non-existent transport. Now it
3161 just assumes that the setting of that option is false. This bug was
3162 introduced by 4.51/PH/31.
3164 PH/28 Changed -d+all to exclude +memory, because that information is very
3165 rarely of interest, but it makes the output a lot bigger. People tend to
3166 do -d+all out of habit.
3168 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3169 code in os-type was giving problems when libc.so lives in lib64, like on
3172 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3173 aren't the modern standard, and it seems that some systems' include files
3174 don't always have them. Exim was already checking for some of the newer
3175 ones like T_AAAA, and defining it itself. I've added checks for all the
3176 record types that Exim uses.
3178 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3179 not automatically generating a new one, as it is supposed to. This
3180 prevented TLS from working. If the file did exist, but contained invalid
3181 data, a new version was generated, as expected. It was only the case of a
3182 non-existent file that was broken.
3184 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3185 with a change in libdomainkeys > 0.64.
3187 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3188 from DNS. If the selector record carries the flag, it now has
3189 precedence over the domain-wide flag.
3191 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3193 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3194 the use of a port name that isn't defined in /etc/services) occurred, the
3195 message was deferred as in a normal delivery, and thus remained on the
3196 spool, instead of being failed because of the mua_wrapper setting. This
3197 is now fixed, and I tidied up some of the mua_wrapper messages at the
3200 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3201 the messages in a hash of arrays rather than using individual hashes.
3202 This is a bit cleaner and results in dramatic memory savings, albeit
3203 at a slight CPU cost.
3205 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3206 as requested by Marc Sherman.
3208 SC/10 Eximstats: added histograms for user specified patterns as requested
3211 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3213 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3214 fopen() in the content-scanning modules that did not already have it.
3220 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3221 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3223 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3225 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3227 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3228 to test Sieve filters that use "vacation".
3230 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3231 that changes the way the GnuTLS parameters are stored in the cache file.
3232 The new format can be generated externally. For backward compatibility,
3233 if the data in the cache doesn't make sense, Exim assumes it has read an
3234 old-format file, and it generates new data and writes a new file. This
3235 means that you can't go back to an older release without removing the
3238 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3239 work if there are any delivery delays because "one_time" forces the
3240 parent to be marked "delivered", so its unseen clone is never tried
3241 again. For this reason, Exim now forbids the simultaneous setting of
3244 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3245 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3246 entry if you want to know the details. The fix, however, neglected to
3247 consider the case when local delivery batching is involved. The test for
3248 "previously delivered" was not happening when checking to see if an
3249 address could be batched with a previous (undelivered) one; under
3250 certain circumstances this could lead to multiple deliveries to the same
3253 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3254 in its include files, and this causes problems building Exim.
3256 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3257 header_syntax) but Exim was just ignoring anything given after a slash.
3258 In particular, this caused confusion with an attempt to use "verify =
3259 reverse_host_lookup/defer_ok". An error is now given when options are
3260 supplied for verify items that do not have them. (Maybe reverse_host_
3261 lookup should have a defer_ok option, but that's a different point.)
3263 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3264 defined by RFC 821) to 2048, because there were problems with some AUTH
3265 commands, and RFC 1869 says the size should be increased for extended
3266 SMTP commands that take arguments.
3268 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3271 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3272 "unknown" error; now it says that the functionality isn't in the binary.
3274 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3275 an address' error message when a string expansion fails (syntax or
3276 whatever). Otherwise the password may appear in the log. Following change
3277 PH/42 below, there is no longer a chance of it appearing in a bounce
3280 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3282 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3283 of its addresses was ever tried. (Bugzilla bug #2.)
3285 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3286 the result incorrectly in the debug output. (It correctly added a newline
3287 to what was transported.)
3289 TF/01 Added $received_time.
3291 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3292 commented out examples of how to interface to a virus scanner and to
3293 SpamAssassin. Also added commented examples of av_scanner and
3294 spamd_address settings.
3296 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3297 and controls are allowed in which ACLs. There were a couple of minor
3298 errors. Some of the entries in the conditions table (which is a table of
3299 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3300 negation of where the condition IS allowed.
3302 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3304 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3305 header file does not have a version number, so I've had to invent a new
3306 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3307 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3308 radiusclient), but it was contributed by a Radius user.
3310 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3311 files or directories, for OpenSSL.
3313 PH/19 When an Exim process that is running as root has to create an Exim log
3314 file, it does so in a subprocess that runs as exim:exim so as to get the
3315 ownership right at creation (otherwise, other Exim processes might see
3316 the file with the wrong ownership). There was no test for failure of this
3317 fork() call, which would lead to the process getting stuck as it waited
3318 for a non-existent subprocess. Forks do occasionally fail when resources
3319 run out. I reviewed all the other calls to fork(); they all seem to check
3322 PH/20 When checking for unexpected SMTP input at connect time (before writing
3323 the banner), Exim was not dealing correctly with a non-positive return
3324 from the read() function. If the client had disconnected by this time,
3325 the result was a log entry for a synchronization error with an empty
3326 string after "input=" when read() returned zero. If read() returned -1
3327 (an event I could not check), uninitialized data bytes were printed.
3328 There were reports of junk text (parts of files, etc) appearing after
3331 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3333 PH/22 Added support for macro redefinition, and (re)definition in between
3334 driver and ACL definitions.
3336 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3337 forgetting to use the resulting value; it was using the unexpanded value.
3339 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3340 hadn't been configured. The fix is from Juergen Kreileder, who
3341 understands it better than I do:
3343 "Here's what I see happening with three configured cyrus_sasl
3344 authenticators configured (plain, login, cram-md5):
3346 On startup auth_cyrus_sasl_init() gets called for each of these.
3347 This means three calls to sasl_listmech() without a specified mech_list.
3348 => SASL tests which mechs of all available mechs actually work
3349 => three warnings about OTP not working
3350 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3352 With the patch, sasl_listmech() also gets called three times. But now
3353 SASL's mech_list option is set to the server_mech specified in the the
3354 authenticator. Or in other words, the answer from sasl_listmech()
3355 gets limited to just the mech you're testing for (which is different
3357 => the return list contains just 'plain' or 'login', 'cram-md5' or
3358 nothing depending on the value of ob->server_mech.
3360 I've just tested the patch: Authentication still works fine,
3361 unavailable mechs specified in the exim configuration are still
3362 caught, and the auth.log warnings about OTP are gone."
3364 PH/25 When debugging is enabled, the contents of the command line are added
3365 to the debugging output, even when log_selector=+arguments is not
3368 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3369 answer is "GNU", and only if the return is "GNU/something" is the answer
3372 PH/27 $acl_verify_message is now set immediately after the failure of a
3373 verification in an ACL, and so is available in subsequent modifiers. In
3374 particular, the message can be preserved by coding like this:
3376 warn !verify = sender
3377 set acl_m0 = $acl_verify_message
3379 Previously, $acl_verify_message was set only while expanding "message"
3380 and "log_message" when a very denied access.
3382 PH/28 Modified OS/os.c-Linux with
3384 -#ifndef OS_LOAD_AVERAGE
3385 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3387 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3388 nomenclature these days.)
3390 PH/29 Installed patch from the Sieve maintainer that adds the options
3391 sieve_useraddress and sieve_subaddress to the redirect router.
3393 PH/30 In these circumstances:
3394 . Two addresses routed to the same list of hosts;
3395 . First host does not offer TLS;
3396 . First host accepts first address;
3397 . First host gives temporary error to second address;
3398 . Second host offers TLS and a TLS session is established;
3399 . Second host accepts second address.
3400 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3401 and peerdn, if requested) that were in fact used only for the second
3404 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3405 attention to any local part prefix or suffix that was matched by the
3406 router that accepted the address. It now behaves in the same way as it
3407 does for delivery: the affixes are removed from the local part unless
3408 rcpt_include_affixes is set on the transport.
3410 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3411 timeout during the DATA phase of an incoming message.
3413 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3414 applied a patch sanctioned by the Sieve maintainer.
3416 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3417 the uid or gid is negative. A case of a negative gid caused this to be
3418 noticed. The fix allows for either to be negative.
3420 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3421 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3424 PH/36 The change PH/12 above was broken. Fixed it.
3426 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3427 the grounds that routing the same address twice would always produce the
3428 same answer. This might have been true once, but it is certainly no
3429 longer true now. Routing a child address may depend on the previous
3430 routing that produced that child. Some complicated redirection strategies
3431 went wrong when messages had multiple recipients, and made Exim's
3432 behaviour dependent on the order in which the addresses were given.
3434 I have moved the duplicate checking until after the routing is complete.
3435 Exim scans the addresses that are assigned to local and remote
3436 transports, and removes any duplicates. This means that more work will be
3437 done, as duplicates will always all be routed, but duplicates are
3438 presumably rare, so I don't expect this is of any significance.
3440 For deliveries to pipes, files, and autoreplies, the duplicate checking
3441 still happens during the routing process, since they are not going to be
3444 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3445 It corrects a timeout issue with spamd. This is Ian's comment: "The
3446 background is that sometimes spamd either never reads data from a
3447 connection it has accepted, or it never writes response data. The exiscan
3448 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3449 blindly assumes that writes won't block so it may never time out."
3451 PH/39 Allow G after quota size as well as K and M.
3453 PH/40 The value set for $authenticated_id in an authenticator may not contain
3454 binary zeroes or newlines because the value is written to log lines and
3455 to spool files. There was no check on this. Now the value is run through
3456 the string_printing() function so that such characters are converted to
3457 printable escape sequences.
3459 PH/41 $message_linecount is a new variable that contains the total number of
3460 lines in the message. Compare $body_linecount, which is the count for the
3463 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3464 bounce and delay warning messages, except in certain special cases, which
3467 (a) An SMTP error message from a remote host;
3468 (b) A message specified in a :fail: redirection;
3469 (c) A message specified in a "fail" command in a system filter;
3470 (d) A message specified in a FAIL return from the queryprogram router;
3471 (e) A message specified by the cannot_route_message router option.
3473 In these cases only, Exim does include the error details in bounce and
3474 warning messages. There are also a few cases where bland messages such
3475 as "unrouteable address" or "local delivery error" are given.
3477 PH/43 $value is now also set for the "else" part of a ${run expansion.
3479 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3480 being worked on, but at least Exim now implements the latest version to
3483 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3484 process to complete was treated as a delivery failure, a timeout while
3485 writing the message to the pipe was logged, but erroneously treated as a
3486 successful delivery. Such timeouts include transport filter timeouts. For
3487 consistency with the overall process timeout, these timeouts are now
3488 treated as errors, giving rise to delivery failures by default. However,
3489 there is now a new Boolean option for the pipe transport called
3490 timeout_defer, which, if set TRUE, converts the failures into defers for
3491 both kinds of timeout. A transport filter timeout is now identified in
3494 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3495 systems where "make" and "gmake" are different, calling "gmake" at top
3496 level broke things. I've arranged for the value of $(MAKE) to be passed
3497 from the Makefile to this script so that it can call the same version of
3501 A note about Exim versions 4.44 and 4.50
3502 ----------------------------------------
3504 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3505 changes of various kinds. As a consequence, a big documentation update was
3506 needed. This delayed the release for rather longer than seemed good, especially
3507 in the light of a couple of (minor) security issues. Therefore, the changes
3508 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3509 release. So 4.44 and 4.50 are in effect two different branches that both start
3512 I have left the 4.50 change log unchanged; it contains all the changes since
3513 4.43. The change log for 4.44 is below; many of its items are identical to
3514 those for 4.50. This seems to be the most sensible way to preserve the
3515 historical information.
3521 1. Minor wording change to the doc/README.SIEVE file.
3523 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3524 computation of the current number of files was incorrect.
3526 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3527 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3530 4. Give more explanation in the error message when the command for a transport
3531 filter fails to execute.
3533 5. There are several places where Exim runs a non-Exim command in a
3534 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3535 was being done only for the command run by the queryprogram router. It is
3536 now done for all such subprocesses. The other cases are: ${run, transport
3537 filters, and the commands run by the lmtp and pipe transports.
3539 6. Added CONFIGURE_GROUP build-time option.
3541 7. Some older OS have a limit of 256 on the maximum number of file
3542 descriptors. Exim was using setrlimit() to set 1000 as a large value
3543 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3544 systems. I've change it so that if it can't get 1000, it tries for 256.
3546 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3547 was an oversight, and furthermore, ever since the addition of extra
3548 controls (e.g. 4.43/32), the checks on when to allow different forms of
3549 "control" were broken. There should now be diagnostics for all cases when a
3550 control that does not make sense is encountered.
3552 9. Added the /retain_sender option to "control=submission".
3554 10. $recipients is now available in the predata ACL (oversight).
3556 11. Tidy the search cache before the fork to do a delivery from a message
3557 received from the command line. Otherwise the child will trigger a lookup
3558 failure and thereby defer the delivery if it tries to use (for example) a
3559 cached ldap connection that the parent has called unbind on.
3561 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3562 of $address_data from the recipient verification was clobbered by the
3563 sender verification.
3565 13. The value of address_data from a sender verification is now available in
3566 $sender_address_data in subsequent conditions in the ACL statement.
3568 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3570 15. Added a new option "connect=<time>" to callout options, to set a different
3573 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3574 was its contents. (It was OK if the option was not defined at all.)
3576 17. A "Completed" log line is now written for messages that are removed from
3577 the spool by the -Mrm option.
3579 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3580 information about exactly what failed.
3582 19. Added -dd to debug only the daemon process.
3584 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3585 handles timeouts, both on the server side and network timeouts. Renamed the
3586 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3588 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3589 It is now set to "smtps".
3591 22. $host_address is now set to the target address during the checking of
3592 ignore_target_hosts.
3594 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3595 being passed; this would have caused $sender_host_name to have been used if
3596 matching the list had actually called for a host name (not very likely,
3597 since this list is usually IP addresses). A host name is now passed as
3600 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3601 code that specifies a non-restarting handler (typically sigaction() in
3602 modern systems) in an attempt to fix a rare and obscure crash bug.
3604 25. Narrowed the window for a race in the daemon that could cause it to ignore
3605 SIGCHLD signals. This is not a major problem, because they are used only to
3606 wake it up if nothing else does.
3608 26. A malformed maildirsize file could cause Exim to calculate negative values
3609 for the mailbox size or file count. Odd effects could occur as a result.
3610 The maildirsize information is now recalculated if the size or filecount
3613 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3614 support for a long time. Removed HAVE_SYS_VFS_H.
3616 28. Installed the latest version of exipick from John Jetmore.
3618 29. In an address list, if the pattern was not a regular expression, an empty
3619 subject address (from a bounce message) matched only if the pattern was an
3620 empty string. Non-empty patterns were not even tested. This was the wrong
3621 because it is perfectly reasonable to use an empty address as part of a
3622 database query. An empty address is now tested by patterns that are
3623 lookups. However, all the other forms of pattern expect the subject to
3624 contain a local part and a domain, and therefore, for them, an empty
3625 address still always fails if the pattern is not itself empty.
3627 30. Exim went into a mad DNS loop when attempting to do a callout where the
3628 host was specified on an smtp transport, and looking it up yielded more
3629 than one IP address.
3631 31. Re-factored the code for checking spool and log partition space into a
3632 function that finds that data and another that does the check. The former
3633 is then used to implement four new variables: $spool_space, $log_space,
3634 $spool_inodes, and $log_inodes.
3636 32. The RFC2047 encoding function was originally intended for short strings
3637 such as real names; it was not keeping to the 75-character limit for
3638 encoded words that the RFC imposes. It now respects the limit, and
3639 generates multiple encoded words if necessary. To be on the safe side, I
3640 have increased the buffer size for the ${rfc2047: expansion operator from
3643 33. It is now permitted to omit both strings after an "if" condition; if the
3644 condition is true, the result is "true". As before, when the second string
3645 is omitted, a false condition yields an empty string. This makes it less
3646 cumbersome to write custom ACL and router conditions.
3648 34. Failure to deliver a bounce message always caused it to be frozen, even if
3649 there was an errors_to setting on the router. The errors_to setting is now
3652 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3653 canonical form (fully expanded) before being placed in
3654 $sender_host_address.
3656 36. The table in the code that translates DNS record types into text (T_A to
3657 "A" for instance) was missing entries for NS and CNAME. It is just possible
3658 that this could have caused confusion if both these types were looked up
3659 for the same domain, because the text type is used as part of Exim's
3660 per-process caching. But the chance of anyone hitting this buglet seems
3663 37. The dnsdb lookup has been extended in a number of ways.
3665 (1) There is a new type, "zns", which walks up the domain tree until it
3666 finds some nameserver records. It should be used with care.
3668 (2) There is a new type, "mxh", which is like "mx" except that it returns
3669 just the host names, not the priorities.
3671 (3) It is now possible to give a list of domains (or IP addresses) to be
3672 looked up. The behaviour when one of the lookups defers can be
3673 controlled by a keyword.
3675 (4) It is now possible to specify the separator character for use when
3676 multiple records are returned.
3678 38. The dnslists ACL condition has been extended: it is now possible to supply
3679 a list of IP addresses and/or domains to be looked up in a particular DNS
3682 39. Added log_selector=+queue_time_overall.
3684 40. When running the queue in the test harness, wait just a tad after forking a
3685 delivery process, to get repeatability of debugging output.
3687 41. Include certificate and key file names in error message when GnuTLS fails
3688 to set them up, because the GnuTLS error message doesn't include the name
3689 of the failing file when there is a problem reading it.
3691 42. Allow both -bf and -bF in the same test run.
3693 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3695 44. The "Exiscan patch" is now merged into the mainline Exim source.
3697 45. Sometimes the final signoff response after QUIT could fail to get
3698 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3699 < 0 before doing a fflush(). This bug looks as though it goes back to the
3700 introduction of TLS in release 3.20, but "sometimes" must have been rare
3701 because the tests only now provoked it.
3703 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3704 (this can affect the format of dates).
3706 47. exim_tidydb, when checking for the continued existence of a message for
3707 which it has found a message-specific retry record, was not finding
3708 messages that were in split spool directories. Consequently, it was
3709 deleting retry records that should have stayed in existence.
3711 48. Steve fixed some bugs in eximstats.
3713 49. The SPA authentication driver was not abandoning authentication and moving
3714 on to the next authenticator when an expansion was forced to fail,
3715 contradicting the general specification for all authenticators. Instead it
3716 was generating a temporary error. It now behaves as specified.
3718 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3719 (the order specifies the preference for clients). The order is now AES256,
3720 AES128, 3DES, ARCFOUR128.
3722 51. Small patch to Sieve code - explicitly set From: when generating an
3725 52. Exim crashed if a remote delivery caused a very long error message to be
3726 recorded - for instance if somebody sent an entire SpamAssassin report back
3727 as a large number of 550 error lines. This bug was coincidentally fixed by
3728 increasing the size of one of Exim's internal buffers (big_buffer) that
3729 happened as part of the Exiscan merge. However, to be on the safe side, I
3730 have made the code more robust (and fixed the comments that describe what
3733 53. Now that there can be additional text after "Completed" in log lines (if
3734 the queue_time_overall log selector is set), a one-byte patch to exigrep
3735 was needed to allow it to recognize "Completed" as not the last thing in
3738 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3739 patch that reportedly fixes this has been added. I am not expert enough to
3740 create a test for it. This is what the patch creator wrote:
3742 "I found a little strange behaviour of ldap code when working with
3743 Windows 2003 AD Domain, where users was placed in more than one
3744 Organization Units. When I tried to give exim partial DN, the exit code
3745 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3746 But simultaneously result of request was absolutely normal ldap result,
3747 so I produce this patch..."
3749 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3750 REFERENCE, so I have modified the code to exclude the patch when that macro
3753 55. Some experimental protocols are using DNS PTR records for new purposes. The
3754 keys for these records are domain names, not reversed IP addresses. The
3755 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3756 leaves it alone. Component reversal etc. now happens only for IP addresses.
3759 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3761 57. Double the size of the debug message buffer (to 2048) so that more of very
3762 long debug lines gets shown.
3764 58. The exicyclog utility now does better if the number of log files to keep
3765 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3767 59. Two changes related to the smtp_active_hostname option:
3769 (1) $smtp_active_hostname is now available as a variable.
3770 (2) The default for smtp_banner uses $smtp_active_hostname instead
3771 of $primary_hostname.
3773 60. The host_aton() function is supposed to be passed a string that is known
3774 to be a valid IP address. However, in the case of IPv6 addresses, it was
3775 not checking this. This is a hostage to fortune. Exim now panics and dies
3776 if the condition is not met. A case was found where this could be provoked
3777 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3778 components; fortuitously, this particular loophole had already been fixed
3779 by change 4.50/55 above.
3781 If there are any other similar loopholes, the new check in host_aton()
3782 itself should stop them being exploited. The report I received stated that
3783 data on the command line could provoke the exploit when Exim was running as
3784 exim, but did not say which command line option was involved. All I could
3785 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3786 running as the user.
3789 61. There was a buffer overflow vulnerability in the SPA authentication code
3790 (which came originally from the Samba project). I have added a test to the
3791 spa_base64_to_bits() function which I hope fixes it.
3794 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3795 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3797 63. The daemon start-up calls getloadavg() while still root for those OS that
3798 need the first call to be done as root, but it missed one case: when
3799 deliver_queue_load_max is set with deliver_drop_privilege. This is
3800 necessary for the benefit of the queue runner, because there is no re-exec
3801 when deliver_drop_privilege is set.
3803 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3804 This has been fixed.
3806 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3807 was in use, was not putting the data itself into the right store pool;
3808 consequently, it could be overwritten for a subsequent message in the same
3809 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3812 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3814 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3815 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3816 the function were treating the return as a boolean value, which happened to
3817 work because 0=false and not-0=true, but is not correct code.
3819 68. The host_aton() function was not handling scoped IPv6 addresses (those
3820 with, for example, "%eth0" on the end) correctly.
3822 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3823 negated items (that is, ~something) in unsigned ints. Some compilers
3824 apparently mutter when there is no cast.
3826 70. If an address verification called from an ACL failed, and did not produce a
3827 user-specific message (i.e. there was only a "system" message), nothing was
3828 put in $acl_verify_message. In this situation, it now puts the system
3831 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3832 session; change 4.31/43 added the unwanted input to the log line - except
3833 that it did not do this in the start of session case. It now does.
3835 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3836 This is wrong and can cause the other end to generate a synchronization
3837 error if it is another Exim or anything else that does the synchronization
3838 check. A QUIT command is no longer sent after a timeout.
3840 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3841 during host lookups.
3843 74. The defer_ok option of callout verification was not working if it was used
3844 when verifying addresses in header lines, that is, for this case:
3846 verify = header_sender/callout=defer_ok
3848 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3849 those file descriptors could be used for SMTP connections. If anything
3850 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3851 could be sent to the SMTP client, causing chaos. The daemon now opens
3852 stdin, stdout, and stderr to /dev/null when it puts itself into the
3855 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3856 log by default. The user can override this with suitable Perl magic.
3858 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3859 the log message when discard triggers, was not working for the DATA ACL or
3860 for the non-SMTP ACL.
3862 78. Error message wording change in sieve.c.
3864 79. If smtp_accept_max_per_host was set, the number of connections could be
3865 restricted to fewer than expected, because the daemon was trying to set up
3866 a new connection before checking whether the processes handling previous
3867 connections had finished. The check for completed processes is now done
3868 earlier. On busy systems, this bug wouldn't be noticed because something
3869 else would have woken the daemon, and it would have reaped the completed
3872 80. If a message was submitted locally by a user whose login name contained one
3873 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3874 It caused a spool format error. I have fixed the spool reading code. A
3875 related problem was that the "from" clause in the Received: line became
3876 illegal because of the space(s). It is now covered by ${quote_local_part.
3878 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3881 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3883 83. Patch from Sieve maintainer to fix unterminated string problem in
3884 "vacation" handling.
3886 84. Some minor changes to the Linux configuration files to help with other
3887 OS variants using glibc.
3889 85. One more patch for Sieve to update vacation handling to latest spec.
3892 ----------------------------------------------------
3893 See the note above about the 4.44 and 4.50 releases.
3894 ----------------------------------------------------
3900 1. Change 4.43/35 introduced a bug that caused file counts to be
3901 incorrectly computed when quota_filecount was set in an appendfile
3904 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3905 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3908 3. Give more explanation in the error message when the command for a transport
3909 filter fails to execute.
3911 4. There are several places where Exim runs a non-Exim command in a
3912 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3913 was being done only for the command run by the queryprogram router. It is
3914 now done for all such subprocesses. The other cases are: ${run, transport
3915 filters, and the commands run by the lmtp and pipe transports.
3917 5. Some older OS have a limit of 256 on the maximum number of file
3918 descriptors. Exim was using setrlimit() to set 1000 as a large value
3919 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3920 systems. I've change it so that if it can't get 1000, it tries for 256.
3922 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3923 was an oversight, and furthermore, ever since the addition of extra
3924 controls (e.g. 4.43/32), the checks on when to allow different forms of
3925 "control" were broken. There should now be diagnostics for all cases when a
3926 control that does not make sense is encountered.
3928 7. $recipients is now available in the predata ACL (oversight).
3930 8. Tidy the search cache before the fork to do a delivery from a message
3931 received from the command line. Otherwise the child will trigger a lookup
3932 failure and thereby defer the delivery if it tries to use (for example) a
3933 cached ldap connection that the parent has called unbind on.
3935 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3936 of $address_data from the recipient verification was clobbered by the
3937 sender verification.
3939 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3940 was its contents. (It was OK if the option was not defined at all.)
3942 11. A "Completed" log line is now written for messages that are removed from
3943 the spool by the -Mrm option.
3945 12. $host_address is now set to the target address during the checking of
3946 ignore_target_hosts.
3948 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3949 being passed; this would have caused $sender_host_name to have been used if
3950 matching the list had actually called for a host name (not very likely,
3951 since this list is usually IP addresses). A host name is now passed as
3954 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3955 code that specifies a non-restarting handler (typically sigaction() in
3956 modern systems) in an attempt to fix a rare and obscure crash bug.
3958 15. Narrowed the window for a race in the daemon that could cause it to ignore
3959 SIGCHLD signals. This is not a major problem, because they are used only to
3960 wake it up if nothing else does.
3962 16. A malformed maildirsize file could cause Exim to calculate negative values
3963 for the mailbox size or file count. Odd effects could occur as a result.
3964 The maildirsize information is now recalculated if the size or filecount
3967 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3968 support for a long time. Removed HAVE_SYS_VFS_H.
3970 18. Updated exipick to current release from John Jetmore.
3972 19. Allow an empty sender to be matched against a lookup in an address list.
3973 Previously the only cases considered were a regular expression, or an
3976 20. Exim went into a mad DNS lookup loop when doing a callout where the
3977 host was specified on the transport, if the DNS lookup yielded more than
3980 21. The RFC2047 encoding function was originally intended for short strings
3981 such as real names; it was not keeping to the 75-character limit for
3982 encoded words that the RFC imposes. It now respects the limit, and
3983 generates multiple encoded words if necessary. To be on the safe side, I
3984 have increased the buffer size for the ${rfc2047: expansion operator from
3987 22. Failure to deliver a bounce message always caused it to be frozen, even if
3988 there was an errors_to setting on the router. The errors_to setting is now
3991 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3992 canonical form (fully expanded) before being placed in
3993 $sender_host_address.
3995 24. Updated eximstats to version 1.33
3997 25. Include certificate and key file names in error message when GnuTLS fails
3998 to set them up, because the GnuTLS error message doesn't include the name
3999 of the failing file when there is a problem reading it.
4001 26. Expand error message when OpenSSL has problems setting up cert/key files.
4004 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
4005 (this can affect the format of dates).
4007 28. exim_tidydb, when checking for the continued existence of a message for
4008 which it has found a message-specific retry record, was not finding
4009 messages that were in split spool directories. Consequently, it was
4010 deleting retry records that should have stayed in existence.
4012 29. eximstats updated to version 1.35
4013 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
4014 1.35 - bugfix such that pie charts by volume are generated correctly
4016 30. The SPA authentication driver was not abandoning authentication and moving
4017 on to the next authenticator when an expansion was forced to fail,
4018 contradicting the general specification for all authenticators. Instead it
4019 was generating a temporary error. It now behaves as specified.
4021 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4022 (the order specifies the preference for clients). The order is now AES256,
4023 AES128, 3DES, ARCFOUR128.
4025 31. Small patch to Sieve code - explicitly set From: when generating an
4028 32. Exim crashed if a remote delivery caused a very long error message to be
4029 recorded - for instance if somebody sent an entire SpamAssassin report back
4030 as a large number of 550 error lines. This bug was coincidentally fixed by
4031 increasing the size of one of Exim's internal buffers (big_buffer) that
4032 happened as part of the Exiscan merge. However, to be on the safe side, I
4033 have made the code more robust (and fixed the comments that describe what
4036 33. Some experimental protocols are using DNS PTR records for new purposes. The
4037 keys for these records are domain names, not reversed IP addresses. The
4038 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4039 leaves it alone. Component reversal etc. now happens only for IP addresses.
4042 34. The host_aton() function is supposed to be passed a string that is known
4043 to be a valid IP address. However, in the case of IPv6 addresses, it was
4044 not checking this. This is a hostage to fortune. Exim now panics and dies
4045 if the condition is not met. A case was found where this could be provoked
4046 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4047 components; fortuitously, this particular loophole had already been fixed
4048 by change 4.50/55 or 4.44/33 above.
4050 If there are any other similar loopholes, the new check in host_aton()
4051 itself should stop them being exploited. The report I received stated that
4052 data on the command line could provoke the exploit when Exim was running as
4053 exim, but did not say which command line option was involved. All I could
4054 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4055 running as the user.
4058 35. There was a buffer overflow vulnerability in the SPA authentication code
4059 (which came originally from the Samba project). I have added a test to the
4060 spa_base64_to_bits() function which I hope fixes it.
4063 36. The daemon start-up calls getloadavg() while still root for those OS that
4064 need the first call to be done as root, but it missed one case: when
4065 deliver_queue_load_max is set with deliver_drop_privilege. This is
4066 necessary for the benefit of the queue runner, because there is no re-exec
4067 when deliver_drop_privilege is set.
4069 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4070 was in use, was not putting the data itself into the right store pool;
4071 consequently, it could be overwritten for a subsequent message in the same
4072 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4075 38. Sometimes the final signoff response after QUIT could fail to get
4076 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4077 < 0 before doing a fflush(). This bug looks as though it goes back to the
4078 introduction of TLS in release 3.20, but "sometimes" must have been rare
4079 because the tests only now provoked it.
4085 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4086 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4087 Now it returns an integer. A number of calls were still expecting a T/F
4088 return. Fortuitously, in all cases, the tests worked in OK situations,
4089 which is the norm. However, things would have gone wrong on any write
4090 failures on the smtp file descriptor. This function is used when sending
4091 messages over SMTP and also when doing verify callouts.
4093 2. When Exim is called to do synchronous delivery of a locally submitted
4094 message (the -odf or -odi options), it no longer closes stderr before doing
4097 3. Implemented the mua_wrapper option.
4099 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4101 5. Implemented the functions header_remove(), header_testname(),
4102 header_add_at_position(), and receive_remove_recipient(), and exported them
4105 6. If an ACL "warn" statement specified the addition of headers, Exim already
4106 inserted X-ACL-Warn: at the start if there was no header name. However, it
4107 was not making this test for the second and subsequent header lines if
4108 there were newlines in the string. This meant that an invalid header could
4109 be inserted if Exim was badly configured.
4111 7. Allow an ACL "warn" statement to add header lines at the start or after all
4112 the Received: headers, as well as at the end.
4114 8. Added the rcpt_4xx retry error code.
4116 9. Added postmaster_mailfrom=xxx to callout verification option.
4118 10. Added mailfrom=xxxx to the callout verification option, for verify=
4121 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4122 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4124 12. Inserted some casts to stop certain compilers warning when using pointer
4125 differences as field lengths or precisions in printf-type calls (mostly
4126 affecting debugging statements).
4128 13. Added optional readline() support for -be (dynamically loaded).
4130 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4131 same clock tick as a message's arrival, so that its received time was the
4132 same as the "first fail" time on the retry record, and that message
4133 remained on the queue past the ultimate address timeout, every queue runner
4134 would try a delivery (because it was past the ultimate address timeout) but
4135 after another failure, the ultimate address timeout, which should have then
4136 bounced the address, did not kick in. This was a "< instead of <=" error;
4137 in most cases the first failure would have been in the next clock tick
4138 after the received time, and all would be well.
4140 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4141 being recognized when the domain list was tested by the match_domain
4142 condition in an expansion string.
4144 16. Added the ${str2b64: operator.
4146 17. Exim was always calling setrlimit() to set a large limit for the number of
4147 processes, without checking whether the existing limit was already
4148 adequate. (It did check for the limit on file descriptors.) Furthermore,
4149 errors from getrlimit() and setrlimit() were being ignored. Now they are
4150 logged to the main and panic logs, but Exim does carry on, to try to do its
4151 job under whatever limits there are.
4153 18. Imported PCRE 5.0.
4155 19. Trivial typo in log message " temporarily refused connection" (the leading
4158 20. If the log selector return_path_on_delivery was set and an address was
4159 redirected to /dev/null, the delivery process crashed because it assumed
4160 that a return path would always be set for a "successful" delivery. In this
4161 case, the whole delivery is bypassed as an optimization, and therefore no
4164 21. Internal re-arrangement: the function for sending a challenge and reading
4165 a response while authentication was assuming a zero-terminated challenge
4166 string. It's now changed to take a pointer and a length, to allow for
4167 binary data in such strings.
4169 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4171 23. Exim was not respecting finduser_retries when seeking the login of the
4172 uid under which it was called; it was always trying 10 times. (The default
4173 setting of finduser_retries is zero.) Also, it was sleeping after the final
4174 failure, which is pointless.
4176 24. Implemented tls_on_connect_ports.
4178 25. Implemented acl_smtp_predata.
4180 26. If the domain in control=submission is set empty, Exim assumes that the
4181 authenticated id is a complete email address when it generates From: or
4182 Sender: header lines.
4184 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4185 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4186 chown and chgrp in /bin and hostname in /usr/bin.
4188 28. Exim was keeping the "process log" file open after each use, just as it
4189 does for the main log. This opens the possibility of it remaining open for
4190 long periods when the USR1 signal hits a daemon. Occasional processlog
4191 errors were reported, that could have been caused by this. Anyway, it seems
4192 much more sensible not to leave this file open at all, so that is what now
4195 29. The long-running daemon process does not normally write to the log once it
4196 has entered its main loop, and it closes the log before doing so. This is
4197 so that log files can straightforwardly be renamed and moved. However,
4198 there are a couple of unusual error situations where the daemon does write
4199 log entries, and I had neglected to close the log afterwards.
4201 30. The text of an SMTP error response that was received during a remote
4202 delivery was being truncated at 512 bytes. This is too short for some of
4203 the long messages that one sometimes sees. I've increased the limit to
4206 31. It is now possible to make retry rules that apply only when a message has a
4207 specific sender, in particular, an empty sender.
4209 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4210 it possible to be selective about when SMTP synchronization is enforced.
4212 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4214 32. Implemented hosts_connection_nolog.
4216 33. Added an ACL for QUIT.
4218 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4221 35. Added mailbox_size and mailbox_filecount to appendfile.
4223 36. Added control = no_multiline_responses to ACLs.
4225 37. There was a bug in the logic of the code that waits for the clock to tick
4226 in the case where the clock went backwards by a substantial amount such
4227 that the microsecond fraction of "now" was more than the microsecond
4228 fraction of "then" (but the whole seconds number was less).
4230 38. Added support for the libradius Radius client library this is found on
4231 FreeBSD (previously only the radiusclient library was supported).
4237 1. When certain lookups returned multiple values in the form name=value, the
4238 quoting of the values was not always being done properly. Specifically:
4239 (a) If the value started with a double quote, but contained no whitespace,
4241 (b) If the value contained whitespace other than a space character (i.e.
4242 tabs or newlines or carriage returns) it was not quoted.
4243 This fix has been applied to the mysql and pgsql lookups by writing a
4244 separate quoting function and calling it from the lookup code. The fix
4245 should probably also be applied to nisplus, ibase and oracle lookups, but
4246 since I cannot test any of those, I have not disturbed their existing code.
4248 2. A hit in the callout cache for a specific address caused a log line with no
4249 reason for rejecting RCPT. Now it says "Previous (cached) callout
4250 verification failure".
4252 3. There was an off-by-one bug in the queryprogram router. An over-long
4253 return line was truncated at 256 instead of 255 characters, thereby
4254 overflowing its buffer with the terminating zero. As well as fixing this, I
4255 have increased the buffer size to 1024 (and made a note to document this).
4257 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4258 when Exim is waiting for an SMTP response from a remote server, Exim
4259 restarts its select() call on the socket, thereby resetting its timeout.
4260 This is not a problem when such interrupts are rare. Somebody set up a cron
4261 job to run exiwhat every 2 minutes, which is less than the normal select()
4262 timeout (5 or 10 minutes). This meant that the select() timeout never
4263 kicked in because it was always reset. I have fixed this by comparing the
4264 time when an interrupt arrives with the time at the start of the first call
4265 to select(). If more time than the timeout has elapsed, the interrupt is
4266 treated as a timeout.
4268 5. Some internal re-factoring in preparation for the addition of Sieve
4269 extensions (by MH). In particular, the "personal" test is moved to a
4270 separate function, and given an option for scanning Cc: and Bcc: (which is
4271 not set for Exim filters).
4273 6. When Exim created an email address using the login of the caller as the
4274 local part (e.g. when creating a From: or Sender: header line), it was not
4275 quoting the local part when it contained special characters such as @.
4277 7. Installed new OpenBSD configuration files.
4279 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4280 try to make them clearer.
4282 9. Callout options, other than the timeout value, were being ignored when
4283 verifying sender addresses in header lines. For example, when using
4285 verify = header_sender/callout=no_cache
4287 the cache was (incorrectly) being used.
4289 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4290 only the Cygwin environment.
4292 11. When return_path_on_delivery was set as a log selector, if different remote
4293 addresses in the same message used different return paths and parallel
4294 remote delivery occurred, the wrong values would sometimes be logged.
4295 (Whenever a remote delivery process finished, the return path value from
4296 the most recently started remote delivery process was logged.)
4298 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4299 lines when AUTH and/or TLS are in use. This is the "received protocol"
4300 field. Exim used to use "asmtp" for authenticated SMTP, without any
4301 indication (in the protocol name) for TLS use. Now it follows the RFC and
4302 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4303 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4304 names appear in log lines as well as in Received: header lines.
4306 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4307 extensions, and comparison tests, and to fix some bugs.
4309 14. Changes to the "personal" filter test:
4311 (1) The test was buggy in that it was just doing the equivalent of
4312 "contains" tests on header lines. For example, if a user's address was
4313 anne@some.where, the "personal" test would incorrectly be true for
4315 To: susanne@some.where
4317 This test is now done by extracting each address from the header in turn,
4318 and checking the entire address. Other tests that are part of "personal"
4319 are now done using regular expressions (for example, to check local parts
4320 of addresses in From: header lines).
4322 (2) The list of non-personal local parts in From: addresses has been
4323 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4324 taken from the Sieve specification recommendations.
4326 (3) If the message contains any header line starting with "List-" it is
4327 treated as non-personal.
4329 (4) The test for "circular" in the Subject: header line has been removed
4330 because it now seems ill-conceived.
4332 15. Minor typos in src/EDITME comments corrected.
4334 16. Installed latest exipick from John Jetmore.
4336 17. If headers_add on a router specified a text string that was too long for
4337 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4338 of string_sprintf() is now avoided.
4340 18. $message_body_size was not set (it was always zero) when running the DATA
4341 ACL and the local_scan() function.
4343 19. For the "mail" command in an Exim filter, no default was being set for
4344 the once_repeat time, causing a random time value to be used if "once" was
4345 specified. (If the value happened to be <= 0, no repeat happened.) The
4346 default is now 0s, meaning "never repeat". The "vacation" command was OK
4347 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4348 (I found it when inspecting the code).
4350 20. There is now an overall timeout for performing a callout verification. It
4351 defaults to 4 times the callout timeout, which applies to individual SMTP
4352 commands during the callout. The overall timeout applies when there is more
4353 than one host that can be tried. The timeout is checked before trying the
4354 next host. This prevents very long delays if there are a large number of
4355 hosts and all are timing out (e.g. when the network connections are timing
4356 out). The value of the overall timeout can be changed by specifying an
4357 additional sub-option for "callout", called "maxwait". For example:
4359 verify = sender/callout=5s,maxwait=20s
4361 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4362 to the end before writing, but this should make it even safer).
4364 22. Exim was forgetting that it had advertised PIPELINING for the second and
4365 subsequent messages on an SMTP connection. It was also not resetting its
4366 memory on STARTTLS and an internal HELO.
4368 23. When Exim logs an SMTP synchronization error within a session, it now
4369 records whether PIPELINING has been advertised or not.
4371 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4372 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4373 rather than long int.
4375 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4377 26. Added the never_mail option to autoreply.
4383 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4384 crash if the getsockname() call failed; this can happen if a connection is
4385 closed very soon after it is established. The problem was simply in the
4386 order in which certain operations were done, causing Exim to try to write
4387 to the SMTP stream before it had set up the file descriptor. The bug has
4388 been fixed by making things happen in the correct order.
4394 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4395 before the connection was closed, thus losing the rejection response.
4397 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4398 some early Solaris releases, but causes trouble in current releases where
4399 socklen_t is defined.
4401 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4404 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4407 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4408 that releasing the top of it at the end releases what was used for sub-
4409 expansions (unless the block got too big). However, discard this block if
4410 the first thing is a variable or header, so that we can use its block when
4411 it is dynamic (useful for very large $message_headers, for example).
4413 6. Lookups now cache *every* query, not just the most recent. A new, separate
4414 store pool is used for this. It can be recovered when all lookup caches are
4415 flushed. Lookups now release memory at the end of their result strings.
4416 This has involved some general refactoring of the lookup sources.
4418 7. Some code has been added to the store_xxx() functions to reduce the amount
4419 of flapping under certain conditions.
4421 8. log_incoming_interface used to affect only the <= reception log lines. Now
4422 it causes the local interface and port to be added to several more SMTP log
4423 lines, for example "SMTP connection from", and rejection lines.
4425 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4427 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4429 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4430 could be overwritten at the end of the current message (or the start of a
4431 new message if it was set in a HELO ACL). The value is now preserved for
4432 the duration of the SMTP connection.
4434 12. If a transport had a headers_rewrite setting, and a matching header line
4435 contained an unqualified address, that address was qualified, even if it
4436 did not match any rewriting rules. The underlying bug was that the values
4437 of the flags that permit the existence of unqualified sender and recipient
4438 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4439 non-local messages, and by -bnq for local messages) were not being
4440 preserved with the message after it was received.
4442 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4443 "next input=" as part of the text comprising the host identity instead of
4444 the correct text. The code was using the same buffer for two different
4445 strings. However, depending on which order the printing function evaluated
4446 its arguments, the bug did not always show up. Under Linux, for example, my
4447 test suite worked just fine.
4449 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4450 doesn't work with some older versions of Perl. It has been changed to "my",
4451 which in any case is probably the better facility to use.
4453 15. A really picky compiler found some instances of statements for creating
4454 error messages that either had too many or two few arguments for the format
4457 16. The size of the buffer for calls to the DNS resolver has been increased
4458 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4459 for addresses that have a lot of PTR records. This alleviates a problem; it
4460 does not fully solve it.
4462 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4463 buffer now truncates the list and logs the incident, which is the same
4464 action as happens when Exim is looking up a host name and its aliases.
4465 Previously in this situation something unpredictable would happen;
4466 sometimes it was "internal error: store_reset failed".
4468 18. If a server dropped the connection unexpectedly when an Exim client was
4469 using GnuTLS and trying to read a response, the client delivery process
4470 crashed while trying to generate an error log message.
4472 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4473 string, for example:
4475 warn message = H1: something\nH2: something
4477 the text was added as a single header line from Exim's point of view
4478 though it ended up OK in the delivered message. However, searching for the
4479 second and subsequent header lines using $h_h2: did not work. This has been
4480 fixed. Similarly, if a system filter added multiple headers in this way,
4481 the routers could not see them.
4483 20. Expanded the error message when iplsearch is called with an invalid key to
4484 suggest using net-iplsearch in a host list.
4486 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4487 ACLs are no longer actually imposed (and a message to that effect is
4490 22. If a "gecos" field in a passwd entry contained escaped characters, in
4491 particular, if it contained a \" sequence, Exim got it wrong when building
4492 a From: or a Sender: header from that name. A second bug also caused
4493 incorrect handling when an unquoted " was present following a character
4494 that needed quoting.
4496 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4497 was not being matched caselessly.
4499 24. Arranged for all hyphens in the exim.8 source to be escaped with
4502 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4503 properly. Recipient callout cache records were still being keyed to include
4504 the sender, even when use_sender was set false. This led to far more
4505 callouts that were necessary. The sender is no longer included in the key
4506 when use_sender is false.
4508 26. Added "control = submission" modifier to ACLs.
4510 27. Added the ${base62d: operator to decode base 62 numbers.
4512 28. dnsdb lookups can now access SRV records.
4514 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4515 the configuration file.
4517 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4518 (-v) mode. This makes the output for a verbose queue run more intelligible.
4520 31. Added a use_postmaster feature to recipient callouts.
4522 32. Added the $body_zerocount variable, containing the number of binary zero
4523 bytes in the message body.
4525 33. The time of last modification of the "new" subdirectory is now used as the
4526 "mailbox time last read" when there is a quota error for a maildir
4529 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4531 35. Added +ignore_unknown as a special item in host lists.
4533 36. Code for decoding IPv6 addresses in host lists is now included, even if
4534 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4535 address was recognized as an IP address, but was then not correctly decoded
4536 into binary, causing unexpected and incorrect effects when compared with
4543 1. Very minor rewording of debugging text in manualroute to say "list of
4544 hosts" instead of "hostlist".
4546 2. If verify=header_syntax was set, and a header line with an unqualified
4547 address (no domain) and a large number of spaces between the end of the
4548 name and the colon was received, the reception process suffered a buffer
4549 overflow, and (when I tested it) crashed. This was caused by some obsolete
4550 code that should have been removed. The fix is to remove it!
4552 3. When running in the test harness, delay a bit after writing a bounce
4553 message to get a bit more predictability in the log output.
4555 4. Added a call to search_tidyup() just before forking a reception process. In
4556 theory, someone could use a lookup in the expansion of smtp_accept_max_
4557 per_host which, without the tidyup, could leave open a database connection.
4559 5. Added the variables $recipient_data and $sender_data which get set from a
4560 lookup success in an ACL "recipients" or "senders" condition, or a router
4561 "senders" option, similar to $domain_data and $local_part_data.
4563 6. Moved the writing of debug_print from before to after the "senders" test
4566 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4567 problems for message scanning, either using a data ACL, or using
4568 local_scan() because the Received: header was not generated till after they
4569 were called (in order to set the time as the time of reception completion).
4570 I have revised the way this works. The header is now generated after the
4571 body is received, but before the ACL or local_scan() are called. After they
4572 are run, the timestamp in the header is updated.
4578 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4579 before starting a queue runner without re-exec. This happened only when
4580 deliver_drop_privilege was set or when the Exim user was set to root. The
4581 effect of the bug was that timeouts during subsequent deliveries caused
4582 crashes instead of being properly handled. The handler is now left at its
4583 default (and expected) setting.
4585 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4586 message, again when deliver_drop_privilege is set or Exim is run as root.
4587 The bug described in (1) was not present in this case, but the tidying up
4588 of the other signals was missing. I have made the two cases consistent.
4590 3. The ignore_target_hosts setting on a manualroute router was being ignored
4591 for hosts that were looked up using the /MX notation.
4593 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4596 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4597 operated on the sender address. After changing the $sender_address to <>
4598 for the sender address verify, Exim was re-instated it as the original
4599 (before rewriting) address, but remembering that it had rewritten it, so it
4600 wasn't rewriting it again. This bug also had the effect of breaking the
4601 sender address verification caching when the sender address was rewritten.
4603 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4604 This has been changed so that if the ip literal address matches
4605 ignore_target_hosts, the router declines.
4607 7. Added expansion conditions match_domain, match_address, and match_local_
4608 part (NOT match_host).
4610 8. The placeholder for the Received: header didn't have a length field set.
4612 9. Added code to Exim itself and to exim_lock to test for a specific race
4613 condition that could lead to file corruption when using MBX delivery. The
4614 issue is with the lockfile that is created in /tmp. If this file is removed
4615 after a process has opened it but before that process has acquired a lock,
4616 there is the potential for a second process to recreate the file and also
4617 acquire a lock. This could lead to two Exim processes writing to the file
4618 at the same time. The added code performs the same test as UW imapd; it
4619 checks after acquiring the lock that its file descriptor still refers to
4620 the same named file.
4622 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4623 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4626 11. Added the smtp_active_hostname option. If used, this will typically be made
4627 to depend on the incoming interface address. Because $interface_address is
4628 not set up until the daemon has forked a reception process, error responses
4629 that can happen earlier (such as "too many connections") no longer contain
4632 12. If an expansion in a condition on a "warn" statement fails because a lookup
4633 defers, the "warn" statement is abandoned, and the next ACL statement is
4634 processed. Previously this caused the whole ACL to be aborted.
4636 13. Added the iplsearch lookup type.
4638 14. Added ident_timeout as a log selector.
4640 15. Added tls_certificate_verified as a log selector.
4642 16. Added a global option tls_require_ciphers (compare the smtp transport
4643 option of the same name). This controls incoming TLS connections.
4645 17. I finally figured out how to make tls_require_ciphers do a similar thing
4646 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4647 before starting the TLS session.
4649 18. Tabs are now shown as \t in -bP output.
4651 19. If the log selector return_path_on_delivery was set, Exim crashed when
4652 bouncing a message because it had too many Received: header lines.
4654 20. If two routers both had headers_remove settings, and the first one included
4655 a superfluous trailing colon, the final name in the first list and the
4656 first name in the second list were incorrectly joined into one item (with a
4657 colon in the middle).
4663 1. Added -C and -D options to the exinext utility, mainly to make it easier
4664 to include in the automated testing, but these could be helpful when
4665 multiple configurations are in use.
4667 2. The exinext utility was not formatting the output nicely when there was
4668 an alternate port involved in the retry record key, nor when there was a
4669 message id as well (for retries that were specific to a specific message
4670 and a specific host). It was also confused by IPv6 addresses, because of
4671 the additional colons they contain. I have fixed the IPv4 problem, and
4672 patched it up to do a reasonable job for IPv6.
4674 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4675 delivery, the log line now contains "pipelined" if PIPELINING was used.
4677 4. An SMTP transport process used to panic and die if the bind() call to set
4678 an explicit outgoing interface failed. This has been changed; it is now
4679 treated in the same way as a connect() failure.
4681 5. A reference to $sender_host_name in the part of a conditional expansion
4682 that was being skipped was still causing a DNS lookup. This no longer
4685 6. The def: expansion condition was not recognizing references to header lines
4686 that used bh_ and bheader_.
4688 7. Added the _cache feature to named lists.
4690 8. The code for checking quota_filecount in the appendfile transport was
4691 allowing one more file than it should have been.
4693 9. For compatibility with Sendmail, the command line option
4701 and sets the incoming protocol and host name (for trusted callers). The
4702 host name and its colon can be omitted when only the protocol is to be set.
4703 Note the Exim already has two private options, -pd and -ps, that refer to
4704 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4705 "s", but I don't think that's a major issue.
4707 10. A number of refactoring changes to the code, none of which should affect
4710 (a) The number of logging options was getting close to filling up the
4711 32-bit word that was used as a bit map. I have split them into two classes:
4712 those that are passed in the argument to log_write(), and those that are
4713 only ever tested independently outside of that function. These are now in
4714 separate 32-bit words, so there is plenty of room for expansion again.
4715 There is no change in the user interface or the logging behaviour.
4717 (b) When building, for example, log lines, the code previously used a
4718 macro that called string_cat() twice, in order to add two strings. This is
4719 not really sufficiently general. Furthermore, there was one instance where
4720 it was actually wrong because one of the argument was used twice, and in
4721 one call a function was used. (As it happened, calling the function twice
4722 did not affect the overall behaviour.) The macro has been replaced by a
4723 function that can join an arbitrary number of extra strings onto a growing
4726 (c) The code for expansion conditions now uses a table and a binary chop
4727 instead of a serial search (which was left over from when there were very
4728 few conditions). Also, it now recognizes conditions like "pam" even when
4729 the relevant support is not compiled in: a suitably worded error message is
4730 given if an attempt is made to use such a condition.
4732 11. Added ${time_interval:xxxxx}.
4734 12. A bug was causing one of the ddress fields not to be passed back correctly
4735 from remote delivery subprocesses. The field in question was not being
4736 subsequently used, so this caused to problems in practice.
4738 13. Added new log selectors queue_time and deliver_time.
4740 14. Might have fixed a bug in maildirsizefile handling that threw up
4741 "unexpected character" debug warnings, and recalculated the data
4742 unnecessarily. In any case, I expanded the warning message to give more
4745 15. Added the message "Restricted characters in address" to the statements in
4746 the default ACL that block characters like @ and % in local parts.
4748 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4749 Three changes have been made:
4751 (a) There was a serious bug; a negative response to MAIL caused the whole
4752 recipient domain to be cached as invalid, thereby blocking all messages
4753 to all local parts at the same domain, from all senders. This bug has
4754 been fixed. The domain is no longer cached after a negative response to
4755 MAIL if the sender used is not empty.
4757 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4760 (c) A new callout option, "use_sender" has been added for people who want
4761 the modified behaviour.
4767 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4770 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4771 indeed breaks things for older releases.
4773 3. Added additional logging to the case where there is a problem reading data
4774 from a filter that is running in a subprocess using a pipe, in order to
4775 try to track down a specific problem.
4777 4. Testing facility fudge: when running in the test harness and attempting
4778 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4779 getting "No route to host". Convert this to a timeout.
4781 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4784 6. Some OS don't have socklen_t but use size_t instead. This affects the
4785 fifth argument of getsockopt() amongst other things. This is now
4786 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4787 can be set for individual OS. I have set it for SunOS5, OSF1, and
4788 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4789 some earlier ones do not.
4791 7. Change 4.30/15 was not doing the test caselessly.
4793 8. The standard form for an IPv6 address literal was being rejected by address
4794 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4795 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4796 this, as well as the form without the "IPv6" on the front (but only when
4797 address literals are enabled, of course).
4799 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4801 10. Exim crashed if a message with an empty sender address specified by -f
4802 encountered a router with an errors_to setting. This could be provoked only
4803 by a command such as
4807 where an empty string was supplied; "<>" did not hit this bug.
4809 11. Installed PCRE release 4.5.
4811 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4812 remained set. It is now erased.
4814 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4815 times from message ids (which are base 36 rather than the normal 62).
4817 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4818 were being counted as actual protocol errors, and logged if the log
4819 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4820 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4821 and DATA following a set of rejected RCPTs do not count as protocol errors.
4822 In other words, Exim assumes they were pipelined, though this may not
4823 actually be the case. Of course, in all cases the client gets an
4824 appropriate error code.
4826 15. If a lookup fails in an ACL condition, a message about the failure may
4827 be available; it is used if testing the ACL cannot continue, because most
4828 such messages specify what the cause of the deferral is. However, some
4829 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4830 that caused an old message to be retained and used if a later statement
4831 caused a defer, replacing the real cause of the deferral.
4833 16. If an IP address had so many PTR records that the DNS lookup buffer
4834 was not large enough to hold them, Exim could crash while trying to process
4835 the truncated data. It now detects and logs this case.
4837 17. Further to 4.21/58, another change has been made: if (and only if) the
4838 first line of a message (the first header line) ends with CRLF, a bare LF
4839 in a subsequent header line has a space inserted after it, so as not to
4840 terminate the header.
4842 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4843 unnecessarily, used atoi() instead of strtol(), and didn't check the
4844 termination when getting file sizes from file names by regex.
4846 19. Completely re-implemented the support for maildirsize files, in the light
4847 of a number of problems with the previous contributed implementation
4848 (4.30/29). In particular:
4850 . If the quota is zero, the maildirsize file is maintained, but no quota is
4853 . If the maildir directory does not exist, it is created before any attempt
4854 to write a maildirsize file.
4856 . The quota value in the file is just a cache; if the quota is changed in
4857 the transport, the new value overrides.
4859 . A regular expression is available for excluding directories from the
4862 20. The autoreply transport checks the characters in options that define the
4863 message's headers; it allows continued headers, but it was checking with
4864 isspace() after an embedded newline instead of explicitly looking for a
4867 21. If all the "regular" hosts to which an address was routed had passed their
4868 expiry times, and had not reached their retry times, the address was
4869 bounced, even if fallback hosts were defined. Now Exim should go on to try
4872 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4873 equivalent code in the SMTP transport. Some hosts send humungous responses
4874 to HELO/EHLO, more than 1024 it seems.
4876 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4877 gives compiler warnings in some environments. I've now done it "properly",
4880 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4881 (because of problems with the built-in one) was declared to return uschar *
4882 instead of char *, causing compiler failure.
4884 25. Fixed a file descriptor leak when processing alias/forward files.
4886 26. Fixed a minor format string issue in dbfn.c.
4888 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4890 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4891 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4894 29. When a sender address is verified, it is cached, to save repeating the test
4895 when there is more than one recipient in a message. However, when the
4896 verification involves a callout, it is possible for different callout
4897 options to be set for different recipients. It is too complicated to keep
4898 track of this in the cache, so now Exim always runs a verification when a
4899 callout is required, relying on the callout cache for the optimization.
4900 The overhead is duplication of the address routing, but this should not be
4903 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4904 to be verified with callout=postmaster, and the main callout worked but the
4905 postmaster check failed, the verification correctly failed. However, if a
4906 subsequent RCPT command asked for sender verification *without* the
4907 postmaster check, incorrect caching caused this verification also to fail,
4910 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4911 it was not caching the DNS options (qualify_single, search_parents) that
4912 were used when the lookup failed. A subsequent lookup with different
4913 options therefore always gave the same answer, though there were cases
4914 where it should not have. (Example: a "domains = !$mx_any" option on a
4915 dnslookup router: the "domains" option is always processed without any
4916 widening, but the router might have qualify_single set.) Now Exim uses the
4917 cached value only when the same options are set.
4919 32. Added John Jetmore's "exipick" utility to the distribution.
4921 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4922 than a timeout (e.g. a certificate is required, and is not provided), an
4923 Exim server now closes the connection immediately. Previously it waited for
4924 the client to close - but if the client is SSL, it seems that they each
4925 wait for each other, leading to a delay before one of them times out.
4927 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4928 maintained 0.8.x compatibility because I don't think many are using it, and
4929 it is clearly obsolete.
4931 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4934 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4935 client certificate was expired. A simple patch fixes this, though I don't
4936 understand the full logic of why the verify callback is called multiple
4939 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4940 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4941 which causes problems with some clients (such as the Certicom SSL Plus
4942 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4943 disables the coutermeasure allowing Eudora to connect."
4945 38. Exim was not checking that a write() to a log file succeeded. This could
4946 lead to Bad Things if a log got too big, in particular if it hit a file
4947 size limit. Exim now panics and dies if it cannot write to a log file, just
4948 as it does if it cannot open a log file.
4950 39. Modified OS/Makefile-Linux so that it now contains
4952 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4954 The two -D definitions ensure that Exim is compiled with large file
4955 support, which makes it possible to handle log files that are bigger than
4958 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4959 instance) a domain was checked against a named list that involved a lookup,
4960 causing $domain_data to be set, then another domain was checked against the
4961 same list, then the first domain was re-checked, the value of $domain_data
4962 after the final check could be wrong. In particular, if the second check
4963 failed, it could be set empty. This bug probably also applied to
4966 41. The strip_trailing_dot option was not being applied to the address given
4967 with the -f command-line option.
4969 42. The code for reading a message's header from the spool was incrementing
4970 $received_count, but never initializing it. This meant that the value was
4971 incorrect (doubled) while delivering a message in the same process in which
4972 it was received. In the most common configuration of Exim, this never
4973 happens - a fresh exec is done - but it can happen when
4974 deliver_drop_privilege is set.
4976 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4977 it now includes up to 150 characters of the unexpected data in the log
4980 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4981 and building data strings. The size of both of these buffers was 10 000
4982 bytes - far larger than anybody would *ever* want, thought I. Needless to
4983 say, somebody hit the limit. I have increased the maximum line length to
4984 20 000 and the maximum data length of concatenated lines to 100 000. I have
4985 also fixed two bugs, because there was no checking on these buffers. Tsk,
4986 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4987 buffer is too small.
4989 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4990 lsearch lookups. Now it does.
4992 46. When parsing a route_list item in a manualroute router, a fixed-length
4993 buffer was used for the list of hosts. I made this 1024 bytes long,
4994 thinking that nobody would ever have a list of hosts that long. Wrong.
4995 Somebody had a whole pile of complicated expansion conditions, and the
4996 string was silently truncated, leading to an expansion error. It turns out
4997 that it is easier to change to an unlimited length (owing to other changes
4998 that have happened since this code was originally written) than to build
4999 structure for giving a limitation error. The length of the item that
5000 expands into the list of hosts is now unlimited.
5002 47. The lsearch lookup could not handle data where the length of text line was
5003 more than 4095 characters. Such lines were truncated, leading to shortened
5004 data being returned. It should now handle lines of any length.
5006 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
5007 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
5010 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
5011 replacing strings such as BIN_DIRECTORY in a source file: the replacement
5012 no longer happens in comment lines. A list of replacements is now placed
5013 at the head of all of the source files, except those whose only change is
5014 to replace PERL_COMMAND in the very first #! line.
5016 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5017 messages on the queue, with a bottom-up merge sort, using code contributed
5018 by Michael Haardt. This should make operations like -bp somewhat faster on
5019 large queues. It won't affect queue runners, except when queue_run_in_order
5022 51. Installed eximstats 1.31 in the distribution.
5024 52. Added support for SRV lookups to the dnslookup router.
5026 53. If an ACL referred to $message_body or $message_body_end, the value was not
5027 reset for any messages that followed in the same SMTP session.
5029 54. The store-handling optimization for building very long strings was not
5030 differentiating between the different store pools. I don't think this
5031 actually made any difference in practice, but I've tidied it.
5033 55. While running the routers to verify a sender address, $sender_address
5034 was still set to the sender address. This is wrong, because when routing to
5035 send a bounce to the sender, it would be empty. Therefore, I have changed
5036 it so that, while verifying a sender address, $sender_address is set to <>.
5037 (There is no change to what happens when verifying a recipient address.)
5039 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5040 target A or AAAA records (if not already returned) without resetting the
5041 qualify_single or search_parents options of the DNS resolver. These are
5042 inappropriate in this case because the targets of MX and SRV records must
5043 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5044 target that, when qualified, matched something in the local domain. These
5045 two options are now turned off when doing these lookups.
5047 57. It seems that at least some releases of Reiserfs (which does not have the
5048 concept of a fixed number of inodes) returns zero and not -1 for the
5049 number of available inodes. This interacted badly with check_spool_inodes,
5050 which assumed that -1 was the "no such thing" setting. What I have done is
5051 to check that the total number of inodes is greater than zero before doing
5052 the test of how many are available.
5054 58. When a "warn" ACL statement has a log_message modifier, the message is
5055 remembered, and not repeated. This is to avoid a lot of repetition when a
5056 message has many recipients that cause the same warning to be written.
5057 Howewer, Exim was preserving the list of already written lines for an
5058 entire SMTP session, which doesn't seem right. The memory is now reset if a
5059 new message is started.
5061 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5062 debugging output unless log_rewrite was also set.
5064 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5065 of (int)(handle) when we know that handle contains (void *)(-1).
5067 61. The Exim daemon panic-logs an error return when it closes the incoming
5068 connection. However "connection reset by peer" seems to be common, and
5069 isn't really an error worthy of noting specially, so that particular error
5072 62. When Exim is trying to find all the local interfaces, it used to panic and
5073 die if the ioctl to get the interface flags failed. However, it seems that
5074 on at least one OS (Solaris 9) it is possible to have an interface that is
5075 included in the list of interfaces, but for which you get a failure error
5076 for this call. This happens when the interface is not "plumbed" into a
5077 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5078 failure of the "get flags" call assumes that the interface is down.
5080 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5081 makes life easier for people who are doing arithmetic on fields extracted
5082 from dates, where you often get leading zeros that should not be
5083 interpreted as octal.
5085 64. Added qualify_domain to the redirect router, to override the global
5088 65. If a pathologically long header line contained very many addresses (the
5089 report of this problem mentioned 10 000) and each of them was rewritten,
5090 Exim could use up a very large amount of memory. (It kept on making new
5091 copies of the header line as it rewrote, and never released the old ones.)
5092 At the expense of a bit more processing, the header rewriting function has
5093 been changed so that it no longer eats memory in this way.
5095 66. The generation of the Received: header has been moved from the time that a
5096 message starts to be received, to the time that it finishes. The timestamp
5097 in the Received: header should now be very close to that of the <= log
5098 line. There are two side-effects of this change:
5100 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5101 logged header lines no longer include the local Received: line, because
5102 it has not yet been created. The same applies to a copy of the message
5103 that is returned to a non-SMTP sender when a message is rejected.
5105 (b) When a filter file is tested using -bf, no additional Received: header
5106 is added to the test message. After some thought, I decided that this
5109 This change does not affect the value of $received_for. It is still set
5110 after address rewriting, but before local_scan() is called.
5112 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5114 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5115 gave an unhelpful panic error message, and a defer error. I have managed to
5116 change this behaviour so that it now rejects any supplied certificate,
5117 which seems right, as the list of acceptable certificates is empty.
5119 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5120 gave an unhelpful defer error. I have not managed to make this reject any
5121 supplied certificates, but the error message it gives is "no certificate
5122 supplied", which is not helpful.
5124 70. exigrep's output now also includes lines that are not associated with any
5125 message, but which match the given pattern. Implemented by a patch from
5126 Martin Sluka, which also tidied up the Perl a bit.
5128 71. Recipient callout verification, like sender verification, was using <> in
5129 the MAIL FROM command. This isn't really the right thing, since the actual
5130 sender may affect whether the remote host accepts the recipient or not. I
5131 have changed it to use the actual sender in the callout; this means that
5132 the cache record is now keyed on a recipient/sender pair, not just the
5133 recipient address. There doesn't seem to be a real danger of callout loops,
5134 since a callout by the remote host to check the sender would use <>.
5135 [SEE ABOVE: changed after hitting problems.]
5137 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5138 temporary errors. However, in the case of such a code being given after
5139 the end of a data transmission (i.e. after ".") Exim was failing to write
5140 a retry record for the message. (Yes, there was some broken host that was
5141 actually sending 8xx at this point.)
5143 73. An unknown lookup type in a host list could cause Exim to panic-die when
5144 the list was checked. (An example that provoked this was putting <; in the
5145 middle of a list instead of at the start.) If this happened during a DATA
5146 ACL check, a -D file could be left lying around. This kind of configuration
5147 error no longer causes Exim to die; instead it causes a defer errror. The
5148 incident is still logged to the main and panic logs.
5150 74. Buglet left over from Exim 3 conversion. The message "too many messages
5151 in one connection" was written to the rejectlog but not the mainlog, except
5152 when address rewriting (yes!) was being logged.
5154 75. Added write_rejectlog option.
5156 76. When a system filter was run not as root (that is, when system_filter_user
5157 was set), the values of the $n variables were not being returned to the
5158 main process; thus, they were not subsequently available in the $sn
5161 77. Added +return_path_on_delivery log selector.
5163 78. A connection timeout was being treated differently from recipients deferred
5164 when testing hosts_max_try with a message that was older than the host's
5165 retry timeout. (The host should not be counted, thus allowing all hosts to
5166 be tried at least once before bouncing.) This may have been the cause of an
5167 occasionally reported bug whereby a message would remain on the queue
5168 longer than the retry timeout, but would be bounced if a delivery was
5169 forced. I say "may" because I never totally pinned down the problem;
5170 setting up timeout/retry tests is difficult. See also the next item.
5172 79. The ultimate address timeout was not being applied to errors that involved
5173 a combination of host plus message (for example, a timeout on a MAIL
5174 command). When an address resolved to a number of possible hosts, and they
5175 were not all tried for each delivery (e.g. because of hosts_max_try), a
5176 message could remain on the queue longer than the retry timeout.
5178 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5181 81. Fixed an obscure SMTP outgoing bug which required at least the following
5182 conditions: (a) there was another message waiting for the same server;
5183 (b) the server returned 5xx to all RCPT commands in the first message so
5184 that the message was not completed; (c) the server dropped the connection
5185 or gave a negative response to the RSET that Exim sends to abort the
5186 transaction. The observed case was a dropped connection after DATA that had
5187 been sent in pipelining mode. That is, the server had advertised PIPELINING
5188 but was not implementing it correctly. The effect of the bug was incorrect
5189 behaviour, such as trying another host, and this could lead to a crash.
5195 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5196 and daemon.c were passed as pointers to ints; they should have been
5197 pointers to socklen_t variables (which are typically unsigned ints).
5199 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5202 3. Fixed a really odd bug that affected only the testing scheme; patching a
5203 certain fixed string in the binary changed the value of another string that
5204 happened to be identical to the end of the original first string.
5206 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5207 name", it returns that address as the IP address. On some operating
5208 systems (e.g. Solaris), it also passes back the IP address string as the
5209 "host name". However, on others (e.g. Linux), it passes back an empty
5210 string. Exim wasn't checking for this, and was changing the host name to an
5211 empty string, assuming it had been canonicized.
5213 5. Although rare, it is permitted to have more than one PTR record for a given
5214 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5215 all the names associated with an address, because they do in Solaris.
5216 However, it seems that they do not in Linux for data that comes from the
5217 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5218 I found this out when I moved to a new Linux workstation and tried to run
5219 the Exim test suite.
5221 To get round this problem I have changed the code so that it now does its
5222 own call to the DNS to look up PTR records when searching for a host name.
5223 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5224 addresses that are only in /etc/hosts are still found.
5226 This behaviour is, however, controlled by an option called host_lookup_
5227 order, which defaults to "bydns:byaddr". If people want to use the other
5228 order, or indeed, just use one or the other means of lookup, they can
5229 specify it in this variable.
5231 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5232 some operating systems, this comes back from gethostbyaddr() as an empty
5233 string, and this is what Exim used to test for. However, it seems that in
5234 other systems, "." is yielded. Exim now tests for this case too.
5236 7. The values of check_spool_space and check_log_space are now held internally
5237 as a number of kilobytes instead of an absolute number of bytes. If a
5238 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5239 kilobyte. This means that much larger values can be stored.
5241 8. Exim monitor: an attempt to get the action menu when not actually pointing
5242 at a message produces an empty menu entitled "No message selected". This
5243 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5244 no entries in it ("Shell widget menu has zero width and/or height"). So I
5245 have added a single, blank menu entry in this case.
5247 9. Added ${quote_local_part.
5249 10. MIME decoding is now applied to the contents of Subject: header lines when
5252 11. Now that a reference to $sender_host_address automatically causes a reverse
5253 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5254 host lookup before query-style lookups in lists that might use this
5255 variable. This has therefore been abolished, and the "net-" prefix is no
5256 longer necessary for query-style lookups.
5258 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5259 have been a typo for LFLAGS, so it has been changed.
5261 13. The install script calls Exim with "-C /dev/null" in order to find the
5262 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5263 to be output. Howeve, since Exim outputs its version number before the
5264 error, it didn't break the script. It just looked ugly. I fixed this by
5265 always allowing "-C /dev/null" if the caller is root.
5267 14. Ignore overlarge ACL variable number when reading spool file - insurance
5268 against a later release with more variables having written the file.
5270 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5271 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5272 this, as well as the form without the "IPv6" on the front.
5274 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5275 OS/Makefile-Darwin file.
5277 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5278 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5280 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5281 variable. However, some OS use ints for inodes. I've added cast to long int
5282 to get rid of the compiler warning.
5284 19. I had forgotten to lock out "/../" in configuration file names when
5285 ALT_CONFIG_PREFIX was set.
5287 20. Routers used for verification do not need to specify transports. However,
5288 if such a router generated a host list, and callout was configured, Exim
5289 crashed, because it could not find a port number from the (non-existent)
5290 transport. It now assumes port 25 in this circumstance.
5292 21. Added the -t option to exigrep.
5294 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5295 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5296 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5297 them is set, LOOKUP_LSEARCH is forced.
5299 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5301 24. Added sender and host information to the "rejected by local_scan()" log
5302 line; previously there was no indication of these.
5304 25. Added .include_if_exists.
5306 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5307 Linux. It turns out that not all file systems support this. Apparently some
5308 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5309 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5310 not supported on the file descriptor, is now ignored when Exim is trying to
5311 sync a directory. This applies only to Linux.
5313 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5315 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5316 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5319 29. Added support for maildirsize files from supplied patch (modified a bit).
5321 30. The use of :fail: followed by an empty string could lead Exim to respond to
5322 sender verification failures with (e.g.):
5324 550 Verification failed for <xxx>
5325 550 Sender verify failed
5327 where the first response line was missing the '-' that indicates it is not
5328 the final line of the response.
5330 31. The loop for finding the name of the user that called Exim had a hardwired
5331 limit of 10; it now uses the value of finduser_retries, which is used for
5332 all other user lookups.
5334 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5337 33. Exim was neglecting to zero errno before one call of strtol() when
5338 expanding a string and expecting an integer value. On some systems this
5339 resulted in spurious "integer overflow" errors. Also, it was casting the
5340 result into an int without checking.
5342 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5343 did not work. The code looks as if it has *never* worked, though it appears
5344 to have been documented since at least releast 1.62. I have made it work.
5346 35. The "timeout_DNS" error in retry rules, also documented since at least
5347 1.62, also never worked. As it isn't clear exactly what this means, and
5348 clearly it isn't a major issue, I have abolished the feature by treating it
5349 as "timeout", and writing a warning to the main and panic logs.
5351 36. The display of retry rules for -brt wasn't always showing the error code
5354 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5355 timeout_connect_A, timeout_connect_MX.
5357 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5358 to the empty sender.
5360 39. The daemon was not analysing the content of -oX till after it had closed
5361 stderr and disconnected from the controlling terminal. This meant that any
5362 syntax errors were only noted on the panic log, and the return code from
5363 the command was 0. By re-arranging the code a little, I've made the
5364 decoding happen first, so such errors now appear on stderr, and the return
5365 code is 1. However, the actual setting up of the sockets still happens in
5366 the disconnected process, so errors there are still only recorded on the
5369 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5370 connections (as happens on some IP stacks) was logged at start up time as
5371 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5372 it from "IPv6 and IPv4", which means that two separate sockets are being
5375 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5376 says whether AF_INET or AF_INET6 was passed as an argument.
5378 42. Exiwhat output was messed up when time zones were included in log
5381 43. Exiwhat now gives more information about the daemon's listening ports,
5382 and whether -tls-on-connect was used.
5384 44. The "port" option of the smtp transport is now expanded.
5386 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5387 silently ignored. Now an error message is written to the main and panic
5390 46. There's a new ACL modifier called "logwrite" which writes to a log file
5391 as soon as it is encountered.
5393 47. Added $local_user_uid and $local_user_gid at routing time.
5395 48. Exim crashed when trying to verify a sender address that was being
5398 49. Exim was recognizing only a space character after ".include". It now also
5399 recognizes a tab character.
5401 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5402 extracting the relevant information from the specification. The man page no
5403 longer contains scrambled data for the -d option, and I've added a section
5404 at the front about calling Exim under different names.
5406 51. Added "extra_headers" argument to the "mail" command in filter files.
5408 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5411 53. Installed eximstats 1.29.
5413 54. Added transport_filter_timeout as a generic transport option.
5415 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5416 Cc: header lines. This was required by RFC 822, but it not required by RFC
5419 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5420 incoming messages that did not have them. Now it does so only if the
5421 message originates locally, that is, if there is no associated remote host
5422 address. When Resent- header lines are present, this applies to the Resent-
5423 lines rather than the non-Resent- lines.
5425 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5426 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5428 58. Messages for configuration errors now include the name of the main
5429 configuration file - useful now that there may be more than one file in a
5430 list (.included file names were always shown).
5432 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5433 for those rare installations that do not start the daemon as root or run it
5434 setuid root. I've cut out the call to initgroups() if the daemon is not
5437 60. The Exim user and group can now be bound into the binary as text strings
5438 that are looked up at the start of Exim's processing.
5440 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5442 62. Added $mailstore_basename variable.
5444 63. Installed patch to sieve.c from Michael Haardt.
5446 64. When Exim failed to open the panic log after failing to open the main log,
5447 the original message it was trying to log was written to stderr and debug
5448 output, but if they were not available (the usual case in production), it
5449 was lost. Now it is written to syslog before the two lines that record the
5450 failures to open the logs.
5452 65. Users' Exim filters run in subprocesses under the user's uid. It is
5453 possible for a "deliver" command or an alias in a "personal" command to
5454 provoke an address rewrite. If logging of address rewriting is configured,
5455 this fails because the process is not running as root or exim. There may be
5456 a better way of dealing with this, but for the moment (because 4.30 needs
5457 to be released), I have disabled address rewrite logging when running a
5458 filter in a non-root, non-exim process.
5464 1. The buildconfig auxiliary program wasn't quoting the value set for
5465 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5466 not defined. This bug was masked in 4.22 by the effect that was fixed in
5469 2. Some messages that were rejected after a message id was allocated were
5470 shown as "incomplete" by exigrep. It no longer does this for messages that
5471 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5473 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5474 have allow_domain_literals set, the ID did not get logged in the <= line.
5475 Domain literals are now always recognized in Message-ID: header lines.
5477 4. The first argument for a ${extract expansion item is the key name or field
5478 number. Leading and trailing spaces in this item were not being ignored,
5479 causing some misleading effects.
5481 5. When deliver_drop_privilege was set, single queue runner processes started
5482 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5483 same command in the process it spins off) were not dropping privilege.
5485 6. When the daemon running as "exim" started a queue runner, it always
5486 re-executed Exim in the spun-off process. This is a waste of effort when
5487 deliver_drop_privilege is set. The new process now just calls the
5488 queue-runner function directly.
5494 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5497 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5498 ignored. Though the use of -oP was forcing the writing of a pid file, it
5499 was always written to the default place.
5501 3. If the message "no IP address found for host xxxx" is generated during
5502 incoming verification, it is now followed by identification of the incoming
5503 connection (so you can more easily find what provoked it).
5505 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5507 5. Added some features to "harden" Exim a bit more against certain attacks:
5509 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5510 be put in Local/Makefile. This is like the never_users runtime option,
5511 but it cannot be overridden. The default setting is "root".
5513 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5514 prefix string with which any file named in a -C command line option
5517 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5518 is retained for -C and -D only if the caller of Exim is root. Without
5519 it, the exim user may also use -C and -D and retain privilege.
5521 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5522 command line option is disabled.
5524 6. Macro names set by the -D option must start with an upper case letter, just
5525 like macro names defined in the configuration file.
5527 7. Added "dereference=" facility to LDAP.
5529 8. Two instances of the typo "uknown" in the source files are fixed.
5531 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5532 the Configure-Makefile script screwed up while processing it.
5534 10. Incorporated PCRE 4.4.
5536 11. The SMTP synchronization check was not operating right at the start of an
5537 SMTP session. For example, it could not catch a HELO sent before the client
5538 waited for the greeting. There is now a check for outstanding input at the
5539 point when the greeting is written. Because of the duplex, asynchronous
5540 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5541 way, but not yet received, when the check is performed.
5543 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5544 on TCP/IP sockets, because this apparently causes some broken clients to
5547 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5548 unchanged) from the Cygwin maintainer.
5550 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5551 was supported instead of testing for mbx. Effectively a typo.
5553 15. The spa authenticator server code was not checking that the input it
5554 received was valid base64.
5556 16. The debug output line for the "set" modifier in ACLs was not showing the
5557 name of the variable that was being set.
5559 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5561 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5562 lookup on its own. Something else was needed to trigger the lookup. For
5563 example, a match in host_lookup or the need for a host name in a host list.
5564 Now, if $sender_host_name is referenced and the host name has not yet been
5565 looked up, a lookup is performed. If the lookup fails, the variable remains
5566 empty, and $host_lookup_failed is set to "1".
5568 19. Added "eqi" as a case-independent comparison operator.
5570 20. The saslauthd authentication condition could segfault if neither service
5571 nor realm was specified.
5573 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5574 error message that was logged was misleading, and incoming SMTP
5575 connections were dropped. The message is now more accurate, and temporary
5576 errors are given to SMTP connections.
5578 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5579 (except RSET) with a 421 error, until QUIT is received. However, it was
5580 failing to send a response to QUIT.
5582 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5583 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5584 ACL is now run after the helo_try_verify_hosts code.
5586 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5587 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5588 case-independent, so other case variants are also recognized). Apparently
5589 some systems use these upper case variants.
5591 25. If more than two messages were waiting for the same host, and a transport
5592 filter was specified for the transport, Exim sent two messages over the
5593 same TCP/IP connection, and then failed with "socket operation on non-
5594 socket" when it tried to send the third.
5596 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5598 27. The extern definition of crypt16() in expand.c was not being excluded when
5599 the OS had its own crypt16() function.
5601 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5602 as a preferred synonym for return_size_limit, both as an option and as an
5605 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5607 30. Changed the default configuration ACL to relax the local part checking rule
5608 for addresses that are not in any local domains. For these addresses,
5609 slashes and pipe symbols are allowed within local parts, but the sequence
5610 /../ is explicitly forbidden.
5612 31. SPA server authentication was not clearing the challenge buffer before
5615 32. log_message in a "warn" ACL statement was writing to the reject log as
5616 well as to the main log, which contradicts the documentation and doesn't
5617 seem right (because no rejection is happening). So I have stopped it.
5619 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5620 However, I am unable to do any testing of this.
5622 34. Fixed an infelicity in the appendfile transport. When checking directories
5623 for a mailbox, to see if any needed to be created, it was accidentally
5624 using path names with one or more superfluous leading slashes; tracing
5625 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5627 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5628 contents are added to the log line that is written for every discarded
5629 recipient. (Previously a log_message setting was ignored.)
5631 36. The ${quote: operator now quotes the string if it is empty.
5633 37. The install script runs exim in order to find its version number. If for
5634 some reason other than non-existence or emptiness, which it checks, it
5635 could not run './exim', it was installing it with an empty version number,
5636 i.e. as "exim-". This error state is now caught, and the installation is
5639 38. An argument was missing from the function that creates an error message
5640 when Exim fails to connect to the socket for saslauthd authentication.
5641 This could cause Exim to crash, or give a corrupted message.
5643 39. Added isip, isip4, and isip6 to ${if conditions.
5645 40. The ACL variables $acl_xx are now saved with the message, and can be
5646 accessed later in routers, transports, and filters.
5648 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5649 strings in the file are not string-expanded.
5651 42. If a MAIL command specified a SIZE value that was too large to fit into an
5652 int variable, the check against message_size_limit failed. Such values are
5653 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5654 day this will have to be increased, but I don't think I want to be around
5655 when emails are that large.
5662 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5663 iconv() is not standard in FreeBSD.
5665 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5666 IPv6 enabled. The observed symptom was a segmentation fault on return from
5667 the function os_common_find_running_interfaces() in src/os.c.
5669 3. In the check_special_case() function in daemon.c I had used "errno" as an
5670 argument name, which causes warnings on some systems. This was basically a
5671 typo, since it was named "eno" in the comments!
5673 4. The code that waits for the clock to tick (at a resolution of some fraction
5674 of a second) so as to ensure message-id uniqueness was always waiting for
5675 at least one whole tick, when it could have waited for less. [This is
5676 almost certainly not relevant at current processor speeds, where it is
5677 unlikely to ever wait at all. But we try to future-proof.]
5679 5. The function that sleeps for a time interval that includes fractions of a
5680 second contained a race. It did not block SIGALRM between setting the
5681 timer, and suspending (a couple of lines later). If the interval was short
5682 and the sigsuspend() was delayed until after it had expired, the suspension
5683 never ended. On busy systems this could lead to processes getting stuck for
5686 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5687 process, before it forks any delivery processes. The open lookup caching
5688 mechanism meant that the open file or database connection was passed into
5689 the delivery process. The problem was that delivery processes always tidy
5690 up cached lookup data. This could cause a problem for the next delivery
5691 process started by the queue runner, because the external queue runner
5692 process does not know about the closure. So the next delivery process
5693 still has data in the lookup cache. In the case of a file lookup, there was
5694 no problem because closing a file descriptor in a subprocess doesn't affect
5695 the parent. However, if the lookup was caching a connection to a database,
5696 the connection was closed, and the second delivery process was likely to
5697 see errors such as "PGSQL: query failed: server closed the connection
5698 unexpectedly". The problem has been fixed by closing all cached lookups
5699 in a queue runner before running a delivery process.
5701 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5702 seem to have the "const" qualifier which it has on other OS. I've
5705 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5706 *of the same type* (client or server) with the same public name that an
5707 error should be diagnosed.
5709 9. When Exim looked up a host name for an IP address, but failed to find the
5710 original IP address when looking up the host name (a safety check), it
5711 output the message "<ip address> does not match any IP for NULL", which was
5712 confusing, to say the least. The bug was that the host name should have
5713 appeared instead of "NULL".
5715 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5716 user that is built into the binary, and the -C or -D options is used, root
5717 privilege is dropped before the configuration file is read. In addition,
5718 logging is switched to stderr instead of the normal log files. If the
5719 configuration then re-defines the Exim user, the unprivileged environment
5720 is probably not what is expected, so Exim logs a panic warning message (but
5723 However, if deliver_drop_privilege is set, the unprivileged state may well
5724 be exactly what is intended, so the warning has been cut out in that case,
5725 and Exim is allowed to try to write to its normal log files.
5731 1. smtp_return_error_details was not giving details for temporary sender
5732 or receiver verification errors.
5734 2. Diagnose a configuration error if two authenticators have the same public
5737 3. Exim used not to create the message log file for a message until the first
5738 delivery attempt. This could be confusing when incoming messages were held
5739 for policy or load reasons. The message log file is now created at the time
5740 the message is received, and an initial "Received" line is written to it.
5742 4. The automatically generated man page for command line options had a minor
5743 bug that caused no ill effects; however, a more serious problem was that
5744 the procedure for building the man page automatically didn't always
5745 operate. Consequently, release 4.20 contains an out-of-date version. This
5746 shouldn't happen again.
5748 5. When building Exim with embedded Perl support, the script that builds the
5749 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5750 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5752 6. The freeze_tell option was not being used for messages that were frozen on
5753 arrival, either by an ACL or by local_scan().
5755 7. Added the smtp_incomplete_transaction log selector.
5757 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5758 was accepting AUTH without a new EHLO.
5760 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5761 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5764 10. Logging of TCP/IP connections (when configured) now happens in the main
5765 daemon process instead of the child process, so that the TCP/IP connection
5766 count is more accurate (but it can never be perfect).
5768 11. The use of "drop" in a nested ACL was not being handled correctly in the
5769 outer ACL. Now, if condition failure induced by the nested "drop" causes
5770 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5771 or "require"), the connection is dropped.
5773 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5774 that yield "discard" can now be used with an "accept" or a "discard" verb,
5775 but an error is generated for any others (because I can't see a useful way
5776 to define what should happen).
5778 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5779 are now processed in the same way as lines in the Exim configuration file.
5780 In particular, continuation lines are supported.
5782 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5784 15. Added -ti meaning -t -i.
5786 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5787 domains, and warn by logging if others are found.
5789 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5790 returned from GIFCONF when Exim is trying to find the list of interfaces on
5791 a host. The code in os.c has been modified to copy each ifreq to an aligned
5792 structure in all cases.
5794 Also, in some cases, the returned ifreq's were being copied to a 'struct
5795 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5796 means the last couple of bytes of an IPv6 address could be chopped if the
5797 ifreq contained only a normal sockaddr (14 bytes storage).
5799 18. Named domain lists were not supported in the hosts_treat_as_local option.
5800 An entry such as +xxxx was not recognized, and was treated as a literal
5803 19. Ensure that header lines added by a DATA ACL are included in the reject log
5804 if the ACL subsequently rejects the message.
5806 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5807 MD5 (which is deprecated).
5809 21. When testing a filter file using -bf, Exim was writing a message when it
5810 took the sender from a "From " line in the message, but it was not doing so
5811 when it took $return_path from a Return-Path: header line. It now does.
5813 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5814 with a valid header line field name (a series of printing characters
5815 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5817 23. Changed "disc" in the source to "disk" to conform to the documentation and
5818 the book and for uniformity.
5820 24. Ignore Sendmail's -Ooption=value command line item.
5822 25. When execve() failed while trying to run a command in a pipe transport,
5823 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5824 could be confused with a return value of 69 from the command itself. This
5825 has been changed to 127, the value the shell returns if it is asked to run
5826 a non-existent command. The wording for the related log line suggests a
5827 non-existent command as the problem.
5829 26. If received_header_text expands to an empty string, do not add a Received:
5830 header line to the message. (Well, it adds a token one on the spool, but
5831 marks it "old" so that it doesn't get used or transmitted.)
5833 27. Installed eximstats 1.28 (addition of -nt option).
5835 28. There was no check for failure on the call to getsockname() in the daemon
5836 code. This can fail if there is a shortage of resources on the system, with
5837 ENOMEM, for example. A temporary error is now given on failure.
5839 29. Contrary to the C standard, it seems that in some environments, the
5840 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5841 program. Exim now does this explicitly; it affects the formatting of
5842 timestamps using strftime().
5844 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5845 complaints. I've now initialized all the variables, to avoid this.
5847 32. Header lines added by a system filter were not being "seen" during
5848 transport-time rewrites.
5850 33. The info_callback() function passed to OpenSSL is set up with type void
5851 (*)(SSL *, int, int), as described somewhere. However, when calling the
5852 function (actually a macro) that sets it up, the type void(*)() is
5853 expected. I've put in a cast to prevent warnings from picky compilers.
5855 34. If a DNS black list lookup found a CNAME record, but there were no A
5856 records associated with the domain it pointed at, Exim crashed.
5858 35. If a DNS black list lookup returned more than one A record, Exim ignored
5859 all but the first. It now scans all returned addresses if a particular IP
5860 value is being sought. In this situation, the contents of the
5861 $dnslist_value variable are a list of all the addresses, separated by a
5864 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5865 to accept a host name and all its aliases if the forward lookup for any of
5866 them yielded the IP address of the incoming connection. Now it accepts only
5867 those names whose forward lookup yields the correct IP address. Any other
5868 names are discarded. This closes a loophole whereby a rogue DNS
5869 administrator could create reverse DNS records to break through a
5870 wildcarded host restriction in an ACL.
5872 37. If a user filter or a system filter that ran in a subprocess used any of
5873 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5874 the wrong values were passed to the pipe command ($thisaddress had the
5875 value of $0, $0 had the value of $1, etc). This bug was introduced by
5876 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5878 38. Improved the line breaking for long SMTP error messages from ACLs.
5879 Previously, if there was no break point between 40 and 75 characters, Exim
5880 left the rest of the message alone. Two changes have been made: (a) I've
5881 reduced the minimum length to 35 characters; (b) if it can't find a break
5882 point between 35 and 75 characters, it looks ahead and uses the first one
5883 that it finds. This may give the occasional overlong line, but at least the
5884 remaining text gets split now.
5886 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5887 file descriptors might be low, and that setting 1000 would always raise it.
5888 It turns out that in some environments, the limit is already over 1000 and
5889 that lowering it causes trouble. So now Exim takes care not to decrease it.
5891 40. When delivering a message, the value of $return_path is set to $sender_
5892 address at the start of routing (routers may change the value). By an
5893 oversight, this default was not being set up when an address was tested by
5894 -bt or -bv, which affected the outcome if any router or filter referred to
5897 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5898 log only when "message" or "log_message" are set. However, if one of the
5899 conditions was an address verification, or a call to a nested ACL, the
5900 messages generated by the underlying test were being passed through. This
5901 no longer happens. The underlying message is available in $acl_verify_
5902 message for both "message" and "log_message" expansions, so it can be
5903 passed through if needed.
5905 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5906 new expansion $bh_ to give the encoded byte string without charset
5907 translation. Translation happens only if iconv() is available; HAVE_ICONV
5908 indicates this at build time. HEADERS_CHARSET gives the charset to
5909 translate to; headers_charset can change it in the configuration, and
5910 "headers charset" can change it in an individual filter file.
5912 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5913 that creates RFC 2047 encoded "words" labels them as that charset instead
5914 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5915 expansion operator; (ii) when Exim creates a From: line for a local
5916 message; (iii) when a header line is rewritten to include a "phrase" part.
5918 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5919 buggy, causing it to skip the first lines of messages whose message ID
5920 ended in 'D'. This would not have bitten before Exim release 4.14, because
5921 message IDs were unlikely to end in 'D' before then. The effect was to have
5922 incorrect size information for certain domains.
5924 45. #include "config.h" was missing at the start of the crypt16.c module. This
5925 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5928 46. If there was a timeout during a "random" callout check, Exim treated it as
5929 a failure of the random address, and carried on sending RSET and the real
5930 address. If the delay was just some slowness somewhere, the response to the
5931 original RCPT would be taken as a response to RSET and so on, causing
5932 mayhem of various kinds.
5934 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5935 when I implemented it. It didn't allow for the fact that some option values
5936 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5937 the right test for positive values.
5939 48. Domain names in DNS records are case-independent. Exim always looks them up
5940 in lower case. Some resolvers return domain names in exactly the case they
5941 appear in the zone file, that is, they may contain uppercase letters. Not
5942 all resolvers do this - some return always lower case. Exim was treating a
5943 change of case by a resolver as a change of domain, similar to a widening
5944 of a domain abbreviation. This triggered its re-routing code and so it was
5945 trying to route what was effectively the same domain again. This normally
5946 caused routing to fail (because the router wouldn't handle the domain
5947 twice). Now Exim checks for this case specially, and just changes the
5948 casing of the domain that it ultimately uses when it transmits the message
5951 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5954 50. If a filter generated a file delivery with a non-absolute name (possible if
5955 no home directory exists for the router), the forbid_file option was not
5958 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5959 the existing equality matching.
5961 52. Exim was using ints instead of ino_t variables in some places where it was
5962 dealing with inode numbers.
5964 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5965 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5966 called TMPDIR, and if it finds it is different, it changes its value.
5968 54. The smtp_printf() function is now made available to local_scan() so
5969 additional output lines can be written before returning. There is also an
5970 smtp_fflush() function to enable the detection of a dropped connection.
5971 The variables smtp_input and smtp_batched_input are exported to
5974 55. Changed the default runtime configuration: the message "Unknown user"
5975 has been removed from the ACL, and instead placed on the localuser router,
5976 using the cannot_route_message feature. This means that any verification
5977 failures that generate their own messages won't get overridden. Similarly,
5978 the "Unrouteable address" message that was in the ACL for unverifiable
5979 relay addresses has also been removed.
5981 56. Added hosts_avoid_esmtp to the smtp transport.
5983 57. The exicyclog script was not checking for the esoteric option
5984 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5985 will work only if exicyclog is run under the appropriate euid.
5987 58. Following a discussion on the list, the rules by which Exim recognises line
5988 endings on incoming messages have been changed. The -dropcr and drop_cr
5989 options are now no-ops, retained only for backwards compatibility. The
5990 following line terminators are recognized: LF CRLF CR. However, special
5991 processing applies to CR:
5993 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5994 nor a local message in the state where . is a terminator.
5996 (ii) If a bare CR is encountered in a header line, an extra space is added
5997 after the line terminator so as not to end the header. The reasoning
5998 behind this is that bare CRs in header lines are most likely either
5999 to be mistakes, or people trying to play silly games.
6001 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
6002 was being incorrectly given as 18 bytes larger than it should have been.
6003 This is a VOB (very old bug).
6005 60. This may never have affected anything current, but just in case it has:
6006 When the local host is found other than at the start of a list of hosts,
6007 the local host, those with the same MX, and any that follow, are discarded.
6008 When the list in question was part of a longer list of hosts, the following
6009 hosts (not currently being processed) were also being discarded. This no
6010 longer happens. I'm not sure if this situation could ever has previously
6013 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6016 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6017 header. This is something that is recommended in a new Internet Draft, and
6018 is something that is documented as being done by Sendmail. There are two
6019 possible values. For messages generated by the autoreply transport, Exim
6022 Auto-Submitted: auto-replied
6024 whereas for all other generated messages (e.g. bounces) it adds
6026 Auto-Submitted: auto-generated
6028 63. The "personal" condition in filters now includes a test for the
6029 Auto-Submitted: header. If it contains the string "auto-" the message it
6030 not considered personal.
6032 64. Added rcpt_include_affixes as a generic transport option.
6034 65. Added queue_only_override (default true).
6036 66. Added the syslog_duplication option.
6038 67. If what should have been the first header line of a message consisted of
6039 a space followed by a colon, Exim was mis-interpreting it as a header line.
6040 It isn't of course - it is syntactically invalid and should therefore be
6041 treated as the start of the message body. The misbehaviour could have
6042 caused a number of strange effects, including loss of data in subsequent
6043 header lines, and spool format errors.
6045 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6046 client host had authenticated. This control can now be exercised by an ACL
6047 for more flexibility.
6049 69. By default, callouts do not happen when testing with -bh. There is now a
6050 variant, -bhc, which does actually run the callout code, including
6051 consulting and updating the callout cache.
6053 70. Added support for saslauthd authentication, courtesy of Alexander
6056 71. If statvfs() failed on the spool or log directories while checking their
6057 size for availability, Exim confusingly gave the error "space shortage".
6058 Furthermore, in debugging mode it crashed with a floating point exception.
6059 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6060 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6061 serious problem, Exim now writes to the main and panic logs when this
6062 happens, with details of the failure. It then refuses to accept the
6063 incoming message, giving the message "spool directory problem" or "log
6064 directory problem" with a 421 code for SMTP messages.
6066 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6067 0, 1, and 2 exist, because some OS complain for execs without them (see
6068 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6069 descriptors. However, the code omitted to check that the open succeeded,
6070 causing mysterious errors if for some reason the permissions on /dev/null
6071 got screwed. Now Exim writes a message to the main and panic logs, and
6072 bombs out if it can't open /dev/null.
6074 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6075 interact so that it is all more flexible. It is supposed to remain
6076 backwards compatible. Also added extra_local_interfaces.
6078 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6079 to bomb out with an assertion failure - to the client this appears as a
6080 connection drop. This problem occurs in the part of the code that was taken
6081 from the Samba project. Fortunately, the assertion is in a very simple
6082 function, so I have fixed this by reproducing the function inline in the
6083 one place where it is called, and arranging for authentication to fail
6084 instead of killing the process with assert().
6086 75. The SPA client code was not working when the server requested OEM rather
6087 than Unicode encoding.
6089 76. Added code to make require_files with a specific uid setting more usable in
6090 the case where statting the file as root fails - usually a non-root-mounted
6091 NFS file system. When this happens and the failure is EACCES, Exim now
6092 forks a subprocess and does the per-uid checking as the relevant uid.
6094 77. Added process_log_path.
6096 78. If log_file_path was not explicitly set, a setting of check_log_space or
6097 check_log_inodes was ignored.
6099 79. If a space check for the spool or log partitions fails, the incident is now
6100 logged. Of course, in the latter case the data may get lost...
6102 80. Added the %p formatting code to string_format() so that it can be used to
6103 print addresses in debug_print(). Adjusted all the address printing in the
6104 debugging in store.c to use %p rather than %d.
6106 81. There was a concern that a line of code in smtp_in.c could overflow a
6107 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6108 initially expressed, the concern was not well-founded, because trailing
6109 spaces are removed early. However, if the trailing spaces were followed by
6110 a NULL, they did not get removed, so the overflow was possible. Two fixes
6113 (a) I re-wrote the offending code in a cleaner fashion.
6114 (b) If an incoming SMTP command contains a NULL character, it is rejected
6117 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6118 runs initgroups(), so that if the Exim user is in any additional groups,
6119 they will be used during message reception.
6125 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob