1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet. PPv2 support was
29 updated based on HAProxy spec change in May 2014.
31 JH/02 Add ${listextract {number}{list}{success}{fail}}.
33 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
34 Properly escape header and check for NULL return.
36 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
39 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
41 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
42 characters in header names, implemented as a verify condition.
43 Contributed by Michael Fischer v. Mollard.
45 TL/05 Rename SPF condition results err_perm and err_temp to standardized
46 results permerror and temperror. Previous values are deprecated but
47 still accepted. In a future release, err_perm and err_temp will be
48 completely removed, which will be a backward incompatibility if the
49 ACL tests for either of these two old results. Patch contributed by
50 user bes-internal on the mailing list.
52 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
54 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
55 selectors, in both main and reject logs.
57 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
60 JH/07 Add malware type "sock" for talking to simple daemon.
62 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
64 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
65 routers/transports under cutthrough routing.
67 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
68 numbers. Touch up "bool" conditional to keep the same definition.
70 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
72 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
74 JH/12 Expand items in router/transport headers_add or headers_remove lists
75 individually rather than the list as a whole. Bug 1452.
77 Required for reasonable handling of multiple headers_ options when
78 they may be empty; requires that headers_remove items with embedded
79 colons must have them doubled (or the list-separator changed).
81 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
82 view the policy declared in the DMARC record. Currently, $dmarc_status
83 is a combined value of both the record presence and the result of the
86 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
88 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
89 dnslookup router and the smtp transport (applying to the forward
92 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
93 of ldap servers used for a specific lookup. Patch provided by Heiko
96 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
97 New variable $lookup_dnssec_authenticated for observability.
99 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
100 Patch submitted by Lars Timman.
102 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
104 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
105 Requires trusted mode and valid format message id, aborts otherwise.
106 Patch contributed by Heiko Schlichting.
108 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
109 certextract with support for various fields. Bug 1358.
111 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
112 is requested by default, modifiable by smtp transport option
115 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
116 operate on certificate variables to give certificate fingerprints
117 Also new ${sha256:cert_variable}.
119 JH/23 The PRDR feature is moved from being Experimental into the mainline.
121 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
122 Christian Aistleitner.
124 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
126 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
127 file. Patch from Wolfgang Breyha.
129 JH/25 Expand the coverage of the delivery $host and $host_address to
130 client authenticators run in verify callout. Bug 1476.
132 JH/26 Port service names are now accepted for tls_on_connect_ports, to
133 align with daemon_smtp_ports. Bug 72.
135 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
136 support and error reporting did not work properly.
138 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
139 and is readable. Patch from Andrew Colin Kissa.
141 TL/14 Enhance documentation of ${run expansion and how it parses the
142 commandline after expansion, particularly in the case when an
143 unquoted variable expansion results in an empty value.
145 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
151 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
153 PP/02 Make -n do something, by making it not do something.
154 When combined with -bP, the name of an option is not output.
156 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
159 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
160 $sender_host_name and config options to manage this, and basic check
163 PP/05 DSCP support for outbound connections and control modifier for inbound.
165 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
166 (Only plugin which currently uses this is kerberos4, which nobody should
167 be using, but we should make it available and other future plugins might
168 conceivably use it, even though it would break NAT; stuff *should* be
169 using channel bindings instead).
171 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
172 name; added for Sendmail compatibility; requires admin caller.
173 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
174 just ignore it); requires trusted caller.
175 Also parse but ignore: -Ac -Am -X<logfile>
178 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
180 TL/02 Add +smtp_confirmation as a default logging option.
182 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
183 Patch by Magnus Holmgren from 2007-02-20.
185 TL/04 Bugzilla 1281 - Spec typo.
186 Bugzilla 1283 - Spec typo.
187 Bugzilla 1290 - Spec grammar fixes.
189 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
191 TL/06 Add Experimental DMARC support using libopendmarc libraries.
193 TL/07 Fix an out of order global option causing a segfault. Reported to dev
194 mailing list by by Dmitry Isaikin.
196 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
198 JH/02 Support "G" suffix to numbers in ${if comparisons.
200 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
202 NM/01 Bugzilla 1197 - Spec typo
203 Bugzilla 1196 - Spec examples corrections
205 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
207 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
208 gnutls_enable_pkcs11, but renamed to more accurately indicate its
211 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
212 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
214 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
215 "acl {{name}{arg}...}", and optional args on acl condition
218 JH/05 Permit multiple router/transport headers_add/remove lines.
220 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
222 JH/07 Avoid using a waiting database for a single-message-only transport.
223 Performance patch from Paul Fisher. Bugzilla 1262.
225 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
228 JH/09 Add $headers_added variable, with content from use of ACL modifier
229 add_header (but not yet added to the message). Bugzilla 199.
231 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
232 Pulled from Bugzilla 817 by Wolfgang Breyha.
234 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
236 (nb: this is the same fix as in Exim 4.80.1)
238 JH/11 Add A= logging on delivery lines, and a client_set_id option on
241 JH/12 Add optional authenticated_sender logging to A= and a log_selector
244 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
246 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
247 advertise SMTP AUTH mechanism to us, instead of a generic
248 protocol violation error. Also, make Exim more robust to bad
249 data from the Dovecot auth socket.
251 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
253 When a queue runner is handling a message, Exim first routes the
254 recipient addresses, during which it prunes them based on the retry
255 hints database. After that it attempts to deliver the message to
256 any remaining recipients. It then updates the hints database using
259 So if a recipient address works intermittently, it can get repeatedly
260 deferred at routing time. The retry hints record remains fresh so the
261 address never reaches the final cutoff time.
263 This is a fairly common occurrence when a user is bumping up against
264 their storage quota. Exim had some logic in its local delivery code
265 to deal with this. However it did not apply to per-recipient defers
266 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
268 This change adds a proper retry rule check during routing so that the
269 final cutoff time is checked against the message's age. We only do
270 this check if there is an address retry record and there is not a
271 domain retry record; this implies that previous attempts to handle
272 the address had the retry_use_local_parts option turned on. We use
273 this as an approximation for the destination being like a local
274 delivery, as in LMTP.
276 I suspect this new check makes the old local delivery cutoff check
277 redundant, but I have not verified this so I left the code in place.
279 TF/02 Correct gecos expansion when From: is a prefix of the username.
281 Test 0254 submits a message to Exim with the header
285 When I ran the test suite under the user fanf2, Exim expanded
286 the header to contain my full name, whereas it should have added
287 a Resent-Sender: header. It erroneously treats any prefix of the
288 username as equal to the username.
290 This change corrects that bug.
292 GF/01 DCC debug and logging tidyup
293 Error conditions log to paniclog rather than rejectlog.
294 Debug lines prefixed by "DCC: " to remove any ambiguity.
296 TF/03 Avoid unnecessary rebuilds of lookup-related code.
298 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
299 Bug spotted by Jeremy Harris; was flawed since initial commit.
300 Would have resulted in OCSP responses post-SNI triggering an Exim
301 NULL dereference and crash.
303 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
305 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
306 Bug detection, analysis and fix by Samuel Thibault.
307 Bugzilla 1331, Debian bug #698092.
309 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
311 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
312 Server implementation by Todd Lyons, client by JH.
313 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
314 config variable "prdr_enable" controls whether the server
315 advertises the facility. If the client requests PRDR a new
316 acl_data_smtp_prdr ACL is called once for each recipient, after
317 the body content is received and before the acl_smtp_data ACL.
318 The client is controlled by bolth of: a hosts_try_prdr option
319 on the smtp transport, and the server advertisement.
320 Default client logging of deliveries and rejections involving
321 PRDR are flagged with the string "PRDR".
323 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
324 fclose(). Diagnosis by Todd Lyons.
326 PP/17 Update configure.default to handle IPv6 localhost better.
327 Patch by Alain Williams (plus minor tweaks).
330 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
331 This is now consistent with GnuTLS, and is now documented: the
332 previous undocumented portable approach to treating the option as
333 unset was to force an expansion failure. That still works, and
334 an empty string is now equivalent.
336 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
337 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
338 not performing validation itself.
340 PP/20 Added force_command boolean option to pipe transport.
341 Patch from Nick Koston, of cPanel Inc.
343 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
346 TF/04 Added udpsend ACL modifer and hexquote expansion operator
348 PP/21 Fix eximon continuous updating with timestamped log-files.
349 Broken in a format-string cleanup in 4.80, missed when I repaired the
350 other false fix of the same issue.
351 Report and fix from Heiko Schlichting.
354 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
355 Report from Prashanth Katuri.
357 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
358 It's SecureTransport, so affects any MacOS clients which use the
359 system-integrated TLS libraries, including email clients.
361 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
362 using a MIME ACL for non-SMTP local injection.
363 Report and assistance in diagnosis by Warren Baker.
365 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
367 JH/16 Fix comparisons for 64b. Bugzilla 1385.
369 TL/09 Add expansion variable $authenticated_fail_id to keep track of
370 last id that failed so it may be referenced in subsequent ACL's.
372 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
375 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
376 ldap library initialization, allowing self-signed CA's to be
377 used. Also properly sets require_cert option later in code by
378 using NULL (global ldap config) instead of ldap handle (per
379 session). Bug diagnosis and testing by alxgomz.
381 TL/12 Enhanced documentation in the ratelimit.pl script provided in
382 the src/util/ subdirectory.
384 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
385 renamed to Transport Post Delivery Action by Jeremy Harris, as
388 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
389 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
390 redis_servers = needs to be configured which will be used by the redis
391 lookup. Patch from Warren Baker, of The Packet Hub.
393 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
395 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
396 hostname or reverse DNS when processing a host list. Used suggestions
397 from multiple comments on this bug.
399 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
401 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
402 Missed a few lines, added it to make the runtest require no keyboard
405 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
406 contains upper case chars. Make router use caseful_local_part.
408 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
409 support when GnuTLS has been built with p11-kit.
415 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
417 This, or similar/improved, will also be change PP/11 of 4.82.
423 PP/01 Handle short writes when writing local log-files.
424 In practice, only affects FreeBSD (8 onwards).
425 Bugzilla 1053, with thanks to Dmitry Isaikin.
427 NM/01 Bugzilla 949 - Documentation tweak
429 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
432 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
434 PP/02 Implemented gsasl authenticator.
436 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
438 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
439 `pkg-config foo` for cflags/libs.
441 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
442 with rest of GSASL and with heimdal_gssapi.
444 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
445 `pkg-config foo` for cflags/libs for the TLS implementation.
447 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
448 properties get this fed in as external SSF. A number of robustness
449 and debugging improvements to the cyrus_sasl authenticator.
451 PP/08 cyrus_sasl server now expands the server_realm option.
453 PP/09 Bugzilla 1214 - Log authentication information in reject log.
454 Patch by Jeremy Harris.
456 PP/10 Added dbmjz lookup type.
458 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
460 PP/12 MAIL args handles TAB as well as SP, for better interop with
461 non-compliant senders.
462 Analysis and variant patch by Todd Lyons.
464 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
465 Bug report from Lars Müller <lars@samba.org> (via SUSE),
466 Patch from Dirk Mueller <dmueller@suse.com>
468 PP/13 tls_peerdn now print-escaped for spool files.
469 Observed some $tls_peerdn in wild which contained \n, which resulted
470 in spool file corruption.
472 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
473 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
474 or write after TLS renegotiation, which otherwise led to messages
477 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
478 as a tracking header (ie: a signed header comes before the signature).
479 Patch from Wolfgang Breyha.
481 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
482 comma-sep list; embedded commas doubled.
484 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
486 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
488 Report and patch from Dmitry Banschikov.
490 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
491 Removed SSL_clear() after SSL_new() which led to protocol negotiation
492 failures. We appear to now support TLS1.1+ with Exim.
494 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
495 lets Exim select keys and certificates based upon TLS SNI from client.
496 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
497 before an outbound SMTP session. New log_selector, +tls_sni.
499 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
500 NULL dereference. Report and patch from Alun Jones.
502 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
503 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
504 Not seeing resolver debug output on NetBSD, but suspect this is a
505 resolver implementation change.
507 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
508 Left warnings. Added "eximon gdb" invocation mode.
510 PP/21 Defaulting "accept_8bitmime" to true, not false.
512 PP/22 Added -bw for inetd wait mode support.
514 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
515 locate the relevant includes and libraries. Made this the default.
517 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
518 Bugzilla 1246, report and most of solution from Tomasz Kusy.
520 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
521 This may cause build issues on older platforms.
523 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
524 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
525 gnutls_require_mac & gnutls_require_protocols (no longer supported).
526 Added SNI support via GnuTLS too.
527 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
529 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
531 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
532 Applied second patch from Janne, implementing suggestion to default
533 multiple-strings-in-record handling to match SPF spec.
535 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
537 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
538 read-only, out of scope).
539 Patch from Wolfgang Breyha, report from Stuart Northfield.
541 PP/29 Fix three issues highlighted by clang analyser static analysis.
542 Only crash-plausible issue would require the Cambridge-specific
543 iplookup router and a misconfiguration.
544 Report from Marcin Mirosław.
546 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
548 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
549 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
550 As part of this, removing so much warning spew let me fix some minor
551 real issues in debug logging.
553 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
554 assignment on my part. Fixed.
556 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
557 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
558 Janne Snabb (who went above and beyond: thank you).
560 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
561 string otherwise requires a connection and a bunch more work and it's
562 relatively easy to get wrong. Should also expose TLS library linkage
565 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
566 64-bit ${eval} (JH/03).
568 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
569 GNU libc to support some of the 64-bit stuff, should not lead to
570 conflicts. Defined before os.h is pulled in, so if a given platform
571 needs to override this, it can.
573 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
574 protection layer was required, which is not implemented.
575 Bugzilla 1254, patch from Wolfgang Breyha.
577 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
578 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
579 tls_dhparam take prime identifiers. Also unbreak combination of
580 OpenSSL+DH_params+TLSSNI.
582 PP/39 Disable SSLv2 by default in OpenSSL support.
588 PP/01 Solaris build fix for Oracle's LDAP libraries.
589 Bugzilla 1109, patch from Stephen Usher.
591 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
593 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
596 TF/02 Fix a couple more cases where we did not log the error message
597 when unlink() failed. See also change 4.74-TF/03.
599 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
600 lock up or crash if it happened to be inside a call to libc when it
601 got a SIGUSR1 from exiwhat.
603 The SIGUSR1 handler appends the current process status to the process
604 log which is later printed by exiwhat. It used to use the general
605 purpose logging code to do this, but several functions it calls are
606 not safe for signals.
608 The new output code in the SIGUSR1 handler is specific to the process
609 log, and simple enough that it's easy to inspect for signal safety.
610 Removing some special cases also simplifies the general logging code.
611 Removing the spurious timestamps from the process log simplifies
614 TF/04 Improved ratelimit ACL condition.
616 The /noupdate option has been deprecated in favour of /readonly which
617 has clearer semantics. The /leaky, /strict, and /readonly update modes
618 are mutually exclusive. The update mode is no longer included in the
619 database key; it just determines when the database is updated. (This
620 means that when you upgrde Exim will forget old rate measurements.)
622 Exim now checks that the per_* options are used with an update mode that
623 makes sense for the current ACL. For example, when Exim is processing a
624 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
625 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
626 must specify per_mail/readonly. If you omit the update mode it defaults to
627 /leaky where that makes sense (as before) or /readonly where required.
629 The /noupdate option is now undocumented but still supported for
630 backwards compatibility. It is equivalent to /readonly except that in
631 ACLs where /readonly is required you may specify /leaky/noupdate or
632 /strict/noupdate which are treated the same as /readonly.
634 A useful new feature is the /count= option. This is a generalization
635 of the per_byte option, so that you can measure the throughput of other
636 aggregate values. For example, the per_byte option is now equivalent
637 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
639 The per_rcpt option has been generalized using the /count= mechanism
640 (though it's more complicated than the per_byte equivalence). When it is
641 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
642 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
643 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
644 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
645 Note that using per_rcpt with a non-readonly update mode in more than
646 one ACL will cause the recipients to be double-counted. (The per_mail
647 and per_byte options don't have this problem.)
649 The handling of very low rates has changed slightly. If the computed rate
650 is less than the event's count (usually one) then this event is the first
651 after a long gap. In this case the rate is set to the same as this event's
652 count, so that the first message of a spam run is counted properly.
654 The major new feature is a mechanism for counting the rate of unique
655 events. The new per_addr option counts the number of different
656 recipients that someone has sent messages to in the last time period. It
657 behaves like per_rcpt if all the recipient addresses are different, but
658 duplicate recipient addresses do not increase the measured rate. Like
659 the /count= option this is a general mechanism, so the per_addr option
660 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
661 example, measure the rate that a client uses different sender addresses
662 with the options per_mail/unique=$sender_address. There are further
663 details in the main documentation.
665 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
667 TF/06 Removed a few PCRE remnants.
669 TF/07 Automatically extract Exim's version number from tags in the git
670 repository when doing development or release builds.
672 PP/02 Raise smtp_cmd_buffer_size to 16kB.
673 Bugzilla 879. Patch from Paul Fisher.
675 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
676 Heavily based on revision 40f9a89a from Simon Arlott's tree.
679 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
681 PP/05 Variable $av_failed, true if the AV scanner deferred.
682 Bugzilla 1078. Patch from John Horne.
684 PP/06 Stop make process more reliably on build failure.
685 Bugzilla 1087. Patch from Heiko Schlittermann.
687 PP/07 Make maildir_use_size_file an _expandable_ boolean.
688 Bugzilla 1089. Patch from Heiko Schlittermann.
690 PP/08 Handle ${run} returning more data than OS pipe buffer size.
691 Bugzilla 1131. Patch from Holger Weiß.
693 PP/09 Handle IPv6 addresses with SPF.
694 Bugzilla 860. Patch from Wolfgang Breyha.
696 PP/10 GnuTLS: support TLS 1.2 & 1.1.
698 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
701 PP/11 match_* no longer expand right-hand-side by default.
702 New compile-time build option, EXPAND_LISTMATCH_RHS.
703 New expansion conditions, "inlist", "inlisti".
705 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
707 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
709 PP/14 fix log_write() format string regression from TF/03.
710 Bugzilla 1152. Patch from Dmitry Isaikin.
716 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
718 PP/02 Harmonised TLS library version reporting; only show if debugging.
719 Layout now matches that introduced for other libraries in 4.74 PP/03.
721 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
723 PP/04 New "dns_use_edns0" global option.
725 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
728 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
729 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
731 TK/01 Updated PolarSSL code to 0.14.2.
732 Bugzilla 1097. Patch from Andreas Metzler.
734 PP/07 Catch divide-by-zero in ${eval:...}.
737 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
740 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
741 format-string attack -- SECURITY: remote arbitrary code execution.
743 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
744 time unintentionally subject to list matching rules, letting the header
745 cause arbitrary Exim lookups (of items which can occur in lists, *not*
746 arbitrary string expansion). This allowed for information disclosure.
748 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
749 INT_MIN/-1 -- value coerced to INT_MAX.
755 NM/01 Workround for PCRE version dependency in version reporting
758 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
759 This fixes portability to compilers other than gcc, notably
760 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
762 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
763 makefiles for portability to HP-UX and POSIX correctness.
765 PP/01 Permit LOOKUP_foo enabling on the make command-line.
766 Also via indented variable definition in the Makefile.
767 (Debugging by Oliver Heesakkers).
769 PP/02 Restore caching of spamd results with expanded spamd_address.
770 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
772 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
773 Improves build reliability. Fix from: Frank Elsner
775 NM/02 Fix wide character breakage in the rfc2047 coding
776 Fixes bug 1064. Patch from Andrey N. Oktyabrski
778 NM/03 Allow underscore in dnslist lookups
779 Fixes bug 1026. Patch from Graeme Fowler
781 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
782 Code patches from Adam Ciarcinski of NetBSD.
784 NM/04 Fixed exiqgrep to cope with mailq missing size issue
787 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
788 is logged, to avoid truncation. Patch from John Horne.
790 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
791 Patch from Jakob Hirsch.
793 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
794 SQL string expansion failure details.
795 Patch from Andrey Oktyabrski.
797 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
798 Patch from Simon Arlott.
800 PP/09 New lookups functionality failed to compile on old gcc which rejects
801 extern declarations in function scope.
802 Patch from Oliver Fleischmann
804 PP/10 Use sig_atomic_t for flags set from signal handlers.
805 Check getgroups() return and improve debugging.
806 Fixed developed for diagnosis in bug 927 (which turned out to be
809 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
810 Patch from Mark Zealey.
812 PP/12 Bugzilla 1056: Improved spamd server selection.
813 Patch from Mark Zealey.
815 PP/13 Bugzilla 1086: Deal with maildir quota file races.
816 Based on patch from Heiko Schlittermann.
818 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
819 Patch from Uwe Doering, sign-off by Michael Haardt.
821 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
822 variable declaration deep within a block. Bug and patch from
825 PP/15 lookups-Makefile IRIX compatibilty coercion.
827 PP/16 Make DISABLE_DKIM build knob functional.
829 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
830 Patch by Simon Arlott
832 TF/03 Fix valgrind.h portability to C89 compilers that do not support
833 variable argument macros. Our copy now differs from upstream.
839 TF/01 Failure to get a lock on a hints database can have serious
840 consequences so log it to the panic log.
842 TF/02 Log LMTP confirmation messages in the same way as SMTP,
843 controlled using the smtp_confirmation log selector.
845 TF/03 Include the error message when we fail to unlink a spool file.
847 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
848 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
849 for maintaining out-of-tree patches for some time.
851 PP/01 Bugzilla 139: Documentation and portability issues.
852 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
853 Handle per-OS dynamic-module compilation flags.
855 PP/02 Let /dev/null have normal permissions.
856 The 4.73 fixes were a little too stringent and complained about the
857 permissions on /dev/null. Exempt it from some checks.
858 Reported by Andreas M. Kirchwitz.
860 PP/03 Report version information for many libraries, including
861 Exim version information for dynamically loaded libraries. Created
862 version.h, now support a version extension string for distributors
863 who patch heavily. Dynamic module ABI change.
865 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
866 privilege escalation vulnerability whereby the Exim run-time user
867 can cause root to append content of the attacker's choosing to
870 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
873 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
874 If dropping privileges for untrusted macros, we disabled normal logging
875 on the basis that it would fail; for the Exim run-time user, this is not
876 the case, and it resulted in successful deliveries going unlogged.
877 Fixed. Reported by Andreas Metzler.
883 PP/01 Date: & Message-Id: revert to normally being appended to a message,
884 only prepend for the Resent-* case. Fixes regression introduced in
885 Exim 4.70 by NM/22 for Bugzilla 607.
887 PP/02 Include check_rfc2047_length in configure.default because we're seeing
888 increasing numbers of administrators be bitten by this.
890 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
892 PP/03 Bugzilla 994: added openssl_options main configuration option.
894 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
896 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
898 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
900 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
901 without a peer certificate, leading to a segfault because of an
902 assumption that peers always have certificates. Be a little more
903 paranoid. Problem reported by Martin Tscholak.
905 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
906 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
907 NB: ClamAV planning to remove STREAM in "middle of 2010".
908 CL also introduces -bmalware, various -d+acl logging additions and
909 more caution in buffer sizes.
911 PP/09 Implemented reverse_ip expansion operator.
913 PP/10 Bugzilla 937: provide a "debug" ACL control.
915 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
917 PP/12 Bugzilla 973: Implement --version.
919 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
921 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
923 PP/15 Bugzilla 816: support multiple condition rules on Routers.
925 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
926 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
927 ignore trailing whitespace.
929 JJ/02 prevent non-panic DKIM error from being sent to paniclog
931 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
934 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
935 Notification from Dr Andrew Aitchison.
937 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
938 ExtendedDetectionInfo response format.
939 Notification from John Horne.
941 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
944 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
945 XSL and documented dependency on system catalogs, with examples of how
948 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
951 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
952 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
953 configuration file which is writeable by the Exim user or group.
955 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
956 of configuration files to cover files specified with the -C option if
957 they are going to be used with root privileges, not just the default
960 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
961 option (effectively making it always true).
963 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
964 files to be used while preserving root privileges.
966 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
967 that rogue child processes cannot use them.
969 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
970 run-time user, instead of root.
972 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
973 Exim run-time user without dropping privileges.
975 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
976 result string, instead of calling string_vformat() twice with the same
979 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
980 for other users. Others should always drop root privileges if they use
981 -C on the command line, even for a whitelisted configure file.
983 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
985 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
991 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
992 $data_path, and $header_path variables; fixed documentation bugs and
995 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
996 exipick to access non-standard spools, including the "frozen" queue
999 NM/01 Bugzilla 965: Support mysql stored procedures.
1000 Patch from Alain Williams
1002 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1004 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1005 Patch from Andreas Metzler
1007 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1008 Patch from Kirill Miazine
1010 NM/05 Bugzilla 671: Added umask to procmail example.
1012 JJ/03 installed exipick 20100323.0, fixing doc bug
1014 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1015 directory. Notification and patch from Dan Rosenberg.
1017 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1019 TK/02 Improve log output when DKIM signing operation fails.
1021 MH/01 Treat the transport option dkim_domain as a colon separated
1022 list, not as a single string, and sign the message with each element,
1023 omitting multiple occurences of the same signer.
1025 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1026 Bugzilla 985, 986. Patch by Simon Arlott
1028 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1029 Patch by Simon Arlott
1031 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1032 MBX locking. Notification from Dan Rosenberg.
1038 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1040 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1042 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1044 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1046 NM/04 Bugzilla 915: Flag broken perl installation during build.
1052 TK/01 Added patch by Johannes Berg that expands the main option
1053 "spamd_address" if it starts with a dollar sign.
1055 TK/02 Write list of recipients to X-Envelope-Sender header when building
1056 the mbox-format spool file for content scanning (suggested by Jakob
1059 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1060 (http://www.dcc-servers.net/) support via dccifd. Activated by
1061 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1063 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1064 by Mark Daniel Reidel <mr@df.eu>.
1066 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1067 When building exim an external PCRE library is now needed -
1068 PCRE is a system library on the majority of modern systems.
1069 See entry on PCRE_LIBS in EDITME file.
1071 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1072 conversation. Added nologin parameter to request.
1073 Patch contributed by Kirill Miazine.
1075 TF/01 Do not log submission mode rewrites if they do not change the address.
1077 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1079 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1080 log files in place. Contributed by Roberto Lima.
1082 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1084 TF/03 Bugzilla 615: When checking the local_parts router precondition
1085 after a local_part_suffix or local_part_prefix option, Exim now
1086 does not use the address's named list lookup cache, since this
1087 contains cached lookups for the whole local part.
1089 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1090 Robert Millan. Documentation is in experimental-spec.txt.
1092 TF/04 Bugzilla 668: Fix parallel build (make -j).
1094 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1096 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1097 Patch provided by Jan Srzednicki.
1099 TF/05 Leading white space used to be stripped from $spam_report which
1100 wrecked the formatting. Now it is preserved.
1102 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1103 that they are available at delivery time.
1105 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1107 TF/08 TLS error reporting now respects the incoming_interface and
1108 incoming_port log selectors.
1110 TF/09 Produce a more useful error message if an SMTP transport's hosts
1111 setting expands to an empty string.
1113 NM/06 Bugzilla 744: EXPN did not work under TLS.
1114 Patch provided by Phil Pennock.
1116 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1117 Patch provided by Richard Godbee.
1119 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1120 acl_smtp_notquit, added index entry.
1122 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1123 Patch provided by Eugene Bujak.
1125 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1126 accept(). Patch provided by Maxim Dounin.
1128 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1129 Patch provided by Phil Pennock.
1131 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1133 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1134 Patch provided by Brad "anomie" Jorsch.
1136 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1137 Patch provided by Dean Brooks.
1139 TK/05 Add native DKIM support (does not depend on external libraries).
1141 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1142 Patch provided by Graeme Fowler.
1144 NM/16 Bugzilla 851: Documentation example syntax fix.
1146 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1148 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1151 NM/19 Bugzilla 745: TLS version reporting.
1152 Patch provided by Phil Pennock.
1154 NM/20 Bugzilla 167: bool: condition support.
1155 Patch provided by Phil Pennock.
1157 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1158 clients. Patch provided by Phil Pennock.
1160 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1161 Patch provided by Brad "anomie" Jorsch.
1163 NM/23 Bugzilla 687: Fix misparses in eximstats.
1164 Patch provided by Heiko Schlittermann.
1166 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1167 Patch provided by Heiko Schlittermann.
1169 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1170 plus update to original patch.
1172 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1174 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1175 Patch provided by David Brownlee.
1177 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1179 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1181 NM/30 Bugzilla 888: TLS documentation bugfixes.
1183 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1185 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1186 Unlike the original bugzilla I have changed all shell scripts in src tree.
1188 NM/33 Bugzilla 898: Transport filter timeout fix.
1189 Patch by Todd Rinaldo.
1191 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1192 Patch by Serge Demonchaux.
1194 NM/35 Bugzilla 39: Base64 decode bug fixes.
1195 Patch by Jakob Hirsch.
1197 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1199 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1201 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1203 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1209 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1210 ALT-N's libdkim that I have put here:
1211 http://duncanthrax.net/exim-experimental/
1213 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1214 were called 'true' and it seems that C99 defines that as a reserved
1215 keyword to be used with 'bool' variable types. That means you could
1216 not include C99-style headers which use bools without triggering
1217 build errors in sieve.c.
1219 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1220 as mailq or other aliases. Changed the --help handling significantly
1221 to do whats expected. exim_usage() emits usage/help information.
1223 SC/01 Added the -bylocaldomain option to eximstats.
1225 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1227 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1229 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1235 PH/01 Another patch from the Sieve maintainer.
1237 PH/02 When an IPv6 address is converted to a string for single-key lookup
1238 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1239 dots are used instead of colons so that keys in lsearch files need not
1240 contain colons. This was done some time before quoting was made available
1241 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1242 (notated using the quote facility) so as to distinguish them from IPv4
1243 keys. This meant that lookups for IP addresses in host lists did not work
1244 for iplsearch lookups.
1246 This has been fixed by arranging for IPv6 addresses to be expressed with
1247 colons if the lookup type is iplsearch. This is not incompatible, because
1248 previously such lookups could never work.
1250 The situation is now rather anomolous, since one *can* have colons in
1251 ordinary lsearch keys. However, making the change in all cases is
1252 incompatible and would probably break a number of configurations.
1254 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1257 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1258 conversion specification without a maximum field width, thereby enabling
1259 a rogue spamd server to cause a buffer overflow. While nobody in their
1260 right mind would setup Exim to query an untrusted spamd server, an
1261 attacker that gains access to a server running spamd could potentially
1262 exploit this vulnerability to run arbitrary code as the Exim user.
1264 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1265 $primary_hostname instead of what libspf2 thinks the hosts name is.
1267 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1268 a directory entry by the name of the lookup key. Previously, if a
1269 symlink pointed to a non-existing file or a file in a directory that
1270 Exim lacked permissions to read, a lookup for a key matching that
1271 symlink would fail. Now it is enough that a matching directory entry
1272 exists, symlink or not. (Bugzilla 503.)
1274 PH/03 The body_linecount and body_zerocount variables are now exported in the
1277 PH/04 Added the $dnslist_matched variable.
1279 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1280 This means they are set thereafter only if the connection becomes
1283 PH/06 Added the client_condition to authenticators so that some can be skipped
1284 by clients under certain conditions.
1286 PH/07 The error message for a badly-placed control=no_multiline_responses left
1287 "_responses" off the end of the name.
1289 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1291 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1292 (without spaces) instead of just copying the configuration text.
1294 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1296 PH/11 Added $max_received_linelength.
1298 PH/12 Added +ignore_defer and +include_defer to host lists.
1300 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1301 way in which PCRE > 7.0 is built.
1303 PH/14 Implemented queue_only_load_latch.
1305 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1306 MAIL command. The effect was to mangle the value on 64-bit systems.
1308 PH/16 Another patch from the Sieve maintainer.
1310 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1312 PH/18 If a system quota error occurred while trying to create the file for
1313 a maildir delivery, the message "Mailbox is full" was not appended to the
1314 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1315 only to a quota excession during the actual writing of the file.
1317 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1318 characters?) which causes problems in log lines. The DN values are now
1319 passed through string_printing() before being added to log lines.
1321 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1322 and InterBase are left for another time.)
1324 PH/21 Added message_body_newlines option.
1326 PH/22 Guard against possible overflow in moan_check_errorcopy().
1328 PH/23 POSIX allows open() to be a macro; guard against that.
1330 PH/24 If the recipient of an error message contained an @ in the local part
1331 (suitably quoted, of course), incorrect values were put in $domain and
1332 $local_part during the evaluation of errors_copy.
1338 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1339 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1342 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1343 issue a MAIL command.
1345 PH/02 In an ACL statement such as
1347 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1349 if a client was not listed at all, or was listed with a value other than
1350 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1351 the condition was not true (as it should be), so access was not denied.
1352 The bug was that the ! inversion was incorrectly passed on to the second
1353 item. This has been fixed.
1355 PH/03 Added additional dnslists conditions == and =& which are different from
1356 = and & when the dns lookup returns more than one IP address.
1358 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1359 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1361 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1362 FSYNC, which compiles an option called disable_fsync that allows for
1363 bypassing fsync(). The documentation is heavily laced with warnings.
1365 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1367 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1368 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1369 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1370 including adding "make clean"; (3) Added -fPIC when compiling the test
1371 dynamically loaded module, to get rid of a warning.
1373 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1374 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1375 = 0s. The bug is otherwise harmless.
1377 PH/07 There was a bug in the dovecot authenticator such that the value of
1378 $auth1 could be overwritten, and so not correctly preserved, after a
1379 successful authentication. This usually meant that the value preserved by
1380 the server_setid option was incorrect.
1382 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1384 PH/09 Installed PCRE release 7.0.
1386 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1387 run for batched SMTP input. It is now run at the start of every message
1388 in the batch. While fixing this I discovered that the process information
1389 (output by running exiwhat) was not always getting set for -bs and -bS
1390 input. This is fixed, and it now also says "batched" for BSMTP.
1392 PH/11 Added control=no_pipelining.
1394 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1395 patch, slightly modified), and move the expansion of helo_data till after
1396 the connection is made in the smtp transport (so it can use these
1399 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1401 PH/14 Added log_selector = +pid.
1403 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1405 PH/16 Add ${if forany and ${if forall.
1407 PH/17 Added dsn_from option to vary the From: line in DSNs.
1409 PH/18 Flush SMTP output before performing a callout, unless control =
1410 no_callout_flush is set.
1412 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1413 was true (the default) a successful delivery failed to delete the retry
1414 item, thus causing premature timeout of the address. The bug is now
1417 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1419 PH/21 Long custom messages for fakedefer and fakereject are now split up
1420 into multiline reponses in the same way that messages for "deny" and
1421 other ACL rejections are.
1423 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1424 with slight modification.
1426 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1427 draft, changing the syntax and factoring some duplicate code".
1429 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1430 for deliveries of the second and subsequent messages over the same SMTP
1433 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1434 ${reduce, with only minor "tidies".
1436 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1438 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1439 expansion side effects.
1441 PH/27 When a message times out after an over-quota error from an Exim-imposed
1442 quota, the bounce message says "mailbox is full". This message was not
1443 being given when it was a system quota that was exceeded. It now should
1446 MH/03 Made $recipients available in local_scan(). local_scan() already has
1447 better access to the recipient list through recipients_list[], but
1448 $recipients can be useful in postmaster-provided expansion strings.
1450 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1451 in the case of a MAIL command with additional options following the
1452 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1453 were accidentally chopped off.
1455 PH/29 SMTP synchronization checks are implemented when a command is read -
1456 there is a check that no more input is waiting when there shouldn't be
1457 any. However, for some commands, a delay in an ACL can mean that it is
1458 some time before the response is written. In this time, more input might
1459 arrive, invalidly. So now there are extra checks after an ACL has run for
1460 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1461 pipelining has not been advertised.
1463 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1465 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1466 returned in the SMTP response when smtp_return_error_details was false.
1467 This has been fixed.
1469 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1470 instead of the C I/O that was originally supplied, because problems were
1471 reported on Solaris.
1473 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1474 Exim which did not show up earlier: it was assuming that a call to
1475 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1476 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1477 was a macro that became an assignment, so it seemed to work. This has
1478 changed to a proper function call with a void return, hence the compile
1479 error. Exim's code has been fixed.
1481 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1484 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1486 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1488 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1489 because it turns out to be too big or there is a timeout) there is no
1490 "Completed" line in the log. When some messages of this type were
1491 selected by exigrep, they were listed as "not completed". Others were
1492 picked up by some special patterns. I have improved the selection
1493 criteria to be more general.
1495 PH/38 The host_find_failed option in the manualroute router can now be set
1496 to "ignore", to completely ignore a host whose IP address cannot be
1497 found. If all hosts are ignored, the behaviour is controlled by the new
1498 host_all_ignored option.
1500 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1501 homing or because of multiple MX records with /mx) generated more than
1502 one IP address, and the following item turned out to be the local host,
1503 all the secondary addresses of the first item were incorrectly removed
1504 from the list, along with the local host and any following hosts (which
1505 is what is supposed to happen).
1507 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1508 whoever called Exim into the -H file. In the case of the daemon it was
1509 behaving confusingly. When first started, it used values for whoever
1510 started the daemon, but after a SIGHUP it used the Exim user (because it
1511 calls itself on a restart). I have changed the code so that it now always
1514 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1515 message are rejected with the same error (e.g. no authentication or bad
1516 sender address), and a DATA command is nevertheless sent (as can happen
1517 with PIPELINING or a stupid MUA), the error message that was given to the
1518 RCPT commands is included in the rejection of the DATA command. This is
1519 intended to be helpful for MUAs that show only the final error to their
1522 PH/42 Another patch from the Sieve maintainer.
1524 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1525 Eximstats - Fixed some broken HTML links and added missing column headers
1527 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1528 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1530 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1532 PH/43 Yet another patch from the Sieve maintainer.
1534 PH/44 I found a way to check for a TCP/IP connection going away before sending
1535 the response to the final '.' that terminates a message, but only in the
1536 case where the client has not sent further data following the '.'
1537 (unfortunately, this is allowed). However, in many cases there won't be
1538 any further data because there won't be any more messages to send. A call
1539 to select() can be used: if it shows that the input is "ready", there is
1540 either input waiting, or the socket has been closed. An attempt to read
1541 the next input character can distinguish the two cases. Previously, Exim
1542 would have sent an OK response which the client would never have see.
1543 This could lead to message repetition. This fix should cure that, at
1544 least in a lot of common cases.
1546 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1547 advertised in response to EHLO.
1553 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1554 fixed by 4.65/MH/01 (is this a record?) are fixed:
1556 (i) An empty string was always treated as zero by the numeric comparison
1557 operators. This behaviour has been restored.
1559 (ii) It is documented that the numeric comparison operators always treat
1560 their arguments as decimal numbers. This was broken in that numbers
1561 starting with 0 were being interpreted as octal.
1563 While fixing these problems I realized that there was another issue that
1564 hadn't been noticed. Values of message_size_limit (both the global option
1565 and the transport option) were treated as octal if they started with 0.
1566 The documentation was vague. These values are now always treated as
1567 decimal, and I will make that clear in the documentation.
1573 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1574 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1577 MH/01 Don't check that the operands of numeric comparison operators are
1578 integers when their expansion is in "skipping" mode (fixes bug
1579 introduced by 4.64-PH/07).
1581 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1582 child addresses, Exim now panics and dies. Previously, because the count
1583 is held in a short int, deliveries were likely to be lost. As such a
1584 large number of recipients for a single message is ridiculous
1585 (performance will be very, very poor), I have chosen to impose a limit
1586 rather than extend the field.
1592 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1593 leftover -K file (the existence of which was triggered by #402).
1594 While we were at it, introduced process PID as part of the -K
1595 filename. This should rule out race conditions when creating
1598 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1599 processing considerably. Previous code took too long for large mails,
1600 triggering a timeout which in turn triggers #401.
1602 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1603 in the DK code in transports.c. sendfile() is not really portable,
1604 hence the _LINUX specificness.
1606 TF/01 In the add_headers option to the mail command in an Exim filter,
1607 there was a bug that Exim would claim a syntax error in any
1608 header after the first one which had an odd number of characters
1611 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1612 callout verification, Exim cached a "reject" for the entire domain. This
1613 is correct for most verifications, but it is not correct for a recipient
1614 verification with use_sender or use_postmaster set, because in that case
1615 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1616 case of MAIL FROM:<> rejection from other early rejections (e.g.
1617 rejection of HELO). When verifying a recipient using a non-null MAIL
1618 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1619 Whatever the result of the callout, the value of the domain cache is
1620 left unchanged (for any other kind of callout, getting as far as trying
1621 RCPT means that the domain itself is ok).
1623 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1626 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1627 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1628 session, the daemon ignores ECONNECT errors and logs others; it now
1629 ignores EPIPE as well.
1631 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1632 (quoted-printable decoding).
1634 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1635 later the small subsequent patch to fix an introduced bug.
1637 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1639 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1641 PH/08 An error is now given if message_size_limit is specified negative.
1643 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1644 to be given (somewhat) arbitrary names.
1646 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1649 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1650 miscellaneous code fixes
1652 PH/10 Added the log_reject_target ACL modifier to specify where to log
1655 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1656 hostname. This is wrong, because it relates to the incoming message (and
1657 probably the interface on which it is arriving) and not to the outgoing
1658 callout (which could be using a different interface). This has been
1659 changed to use the value of the helo_data option from the smtp transport
1660 instead - this is what is used when a message is actually being sent. If
1661 there is no remote transport (possible with a router that sets up host
1662 addresses), $smtp_active_hostname is used.
1664 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1665 tweaks were necessary in order to get it to work (see also 21 below):
1666 (a) The code assumed that strncpy() returns a negative number on buffer
1667 overflow, which isn't the case. Replaced with Exim's string_format()
1669 (b) There were several signed/unsigned issues. I just did the minimum
1670 hacking in of casts. There is scope for a larger refactoring.
1671 (c) The code used strcasecmp() which is not a standard C function.
1672 Replaced with Exim's strcmpic() function.
1673 (d) The code set only $1; it now sets $auth1 as well.
1674 (e) A simple test gave the error "authentication client didn't specify
1675 service in request". It would seem that Dovecot has changed its
1676 interface. Fortunately there's a specification; I followed it and
1677 changed what the client sends and it appears to be working now.
1679 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1682 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1683 address A is aliased to B and C, where B exists and C does not. Without
1684 -v the output is "A verified" because verification stops after a
1685 successful redirection if more than one address is generated. However,
1686 with -v the child addresses are also verified. Exim was outputting "A
1687 failed to verify" and then showing the successful verification for C,
1688 with its parentage. It now outputs "B failed to verify", showing B's
1689 parentage before showing the successful verification of C.
1691 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1692 look up a TXT record in a specific list after matching in a combined
1695 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1696 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1697 they consult the DNS. I had assumed they would set it the way they
1698 wanted; and indeed my experiments on Linux seem to show that in some
1699 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1700 To be on the safe side, however, I have now made the interface to
1701 host_find_byname() similar to host_find_bydns(), with an argument
1702 containing the DNS resolver options. The host_find_byname() function now
1703 sets these options at its start, just as host_find_bydns() does. The smtp
1704 transport options dns_qualify_single and dns_search_parents are passed to
1705 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1706 of host_find_byname() use the default settings of RES_DEFNAMES
1707 (qualify_single) but not RES_DNSRCH (search_parents).
1709 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1710 spool_read_header() do less string testing, by means of a preliminary
1711 switch on the second character of optional "-foo" lines. (This is
1712 overdue, caused by the large number of possibilities that now exist.
1713 Originally there were few.) While I was there, I also converted the
1714 str(n)cmp tests so they don't re-test the leading "-" and the first
1715 character, in the hope this might squeeze out yet more improvement.
1717 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1718 flag allowing group syntax was set by the header_syntax check but not
1719 turned off, possible causing trouble later; (2) The flag was not being
1720 set at all for the header_verify test, causing "group"-style headers to
1721 be rejected. I have now set it in this case, and also caused header_
1722 verify to ignore an empty address taken from a group. While doing this, I
1723 came across some other cases where the code for allowing group syntax
1724 while scanning a header line wasn't quite right (mostly, not resetting
1725 the flag correctly in the right place). These bugs could have caused
1726 trouble for malformed header lines. I hope it is now all correct.
1728 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1729 with the "reply" argument non-NULL. The code, however (which originally
1730 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1731 but it didn't always do it. This confused somebody who was copying the
1732 code for some other use. I have removed all the tests.
1734 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1735 feature that was used to support insecure browsers during the U.S. crypto
1736 embargo. It requires special client support, and Exim is probably the
1737 only MTA that supported it -- and would never use it because real RSA is
1738 always available. This code has been removed, because it had the bad
1739 effect of slowing Exim down by computing (never used) parameters for the
1740 RSA_EXPORT functionality.
1742 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1743 authenticator to fail if there's a tab character in the incoming data
1744 (there should never be unless someone is messing about, as it's supposed
1745 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1746 if the connection is using TLS or if the remote IP is the same as the
1747 local IP, and the "valid-client-cert option" if a client certificate has
1750 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1751 authenticators. This can be used for authorization after authentication
1752 succeeds. (In the case of plaintext, it servers for both authentication
1755 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1756 if any retry times were supplied.
1758 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1759 connection, where there is no client IP address to check. In this
1760 situation, the verify now always succeeds.
1762 PH/25 Applied John Jetmore's -Mset patch.
1764 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1766 PH/27 In a string expansion for a processed (not raw) header when multiple
1767 headers of the same name were present, leading whitespace was being
1768 removed from all of them, but trailing whitespace was being removed only
1769 from the last one. Now trailing whitespace is removed from each header
1770 before concatenation. Completely empty headers in a concatenation (as
1771 before) are ignored.
1773 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1774 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1776 PH/29 [Removed. This was a change that I later backed out, and forgot to
1777 correct the ChangeLog entry (that I had efficiently created) before
1778 committing the later change.]
1780 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1781 address errors (4xx response to RCPT) over the same connection as other
1782 messages routed to the same hosts. Such deliveries are always "forced",
1783 so retry times are not inspected. This resulted in far too many retries
1784 for the affected addresses. The effect occurred only when there were more
1785 hosts than the hosts_max_try setting in the smtp transport when it had
1786 the 4xx errors. Those hosts that it had tried were not added to the list
1787 of hosts for which the message was waiting, so if all were tried, there
1788 was no problem. Two fixes have been applied:
1790 (i) If there are any address or message errors in an SMTP delivery, none
1791 of the hosts (tried or untried) are now added to the list of hosts
1792 for which the message is waiting, so the message should not be a
1793 candidate for sending over the same connection that was used for a
1794 successful delivery of some other message. This seems entirely
1795 reasonable: after all the message is NOT "waiting for some host".
1796 This is so "obvious" that I'm not sure why it wasn't done
1797 previously. Hope I haven't missed anything, but it can't do any
1798 harm, as the worst effect is to miss an optimization.
1800 (ii) If, despite (i), such a delivery is accidentally attempted, the
1801 routing retry time is respected, so at least it doesn't keep
1802 hammering the server.
1804 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1805 in ${readsocket because some servers need this prod.
1807 PH/32 Added some extra debug output when updating a wait-xxx database.
1809 PH/33 The hint "could be header name not terminated by colon", which has been
1810 given for certain expansion errors for a long time, was not being given
1811 for the ${if def:h_colon_omitted{... case.
1813 PH/34 The spec says: "With one important exception, whenever a domain list is
1814 being scanned, $domain contains the subject domain." There was at least
1815 one case where this was not true.
1817 PH/35 The error "getsockname() failed: connection reset by peer" was being
1818 written to the panic log as well as the main log, but it isn't really
1819 panic-worthy as it just means the connection died rather early on. I have
1820 removed the panic log writing for the ECONNRESET error when getsockname()
1823 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1824 runs only) independently of the message's sender address. This meant
1825 that, if the 4xx error was in fact related to the sender, a different
1826 message to the same recipient with a different sender could confuse
1827 things. In particualar, this can happen when sending to a greylisting
1828 server, but other circumstances could also provoke similar problems.
1829 I have changed the default so that the retry time for these errors is now
1830 based a combination of the sender and recipient addresses. This change
1831 can be overridden by setting address_retry_include_sender=false in the
1834 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1835 remote server are returned as part of bounce messages. This was not
1836 happening for LMTP over a pipe (the lmtp transport), but now it is the
1837 same for both kinds of LMTP.
1839 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1840 in header lines that were in fact CNAMEs. This is no longer the case.
1842 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1843 and queue runs started by the daemon processed all messages. This has
1844 been fixed so that -R and -S can now usefully be given with -q<time>.
1846 PH/40 Import PCRE release 6.7 (fixes some bugs).
1848 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1850 PH/42 Give an error if -q is specified more than once.
1852 PH/43 Renamed the variables $interface_address and $interface_port as
1853 $received_ip_address and $received_port, to make it clear that these
1854 values apply to message reception, and not to the outgoing interface when
1855 a message is delivered. (The old names remain recognized, of course.)
1857 PH/44 There was no timeout on the connect() call when using a Unix domain
1858 socket in the ${readsocket expansion. There now is.
1860 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1861 be meaningful with "accept".
1863 SC/01 Eximstats V1.43
1864 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1866 SC/02 Eximstats V1.44
1867 Use a glob alias rather than an array ref in the generated
1868 parser. This improves both readability and performance.
1870 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1871 Collect SpamAssassin and rejection statistics.
1872 Don't display local sender or destination tables unless
1873 there is data to show.
1874 Added average volumes into the top table text output.
1876 SC/04 Eximstats V1.46
1877 Collect data on the number of addresses (recipients)
1878 as well as the number of messages.
1880 SC/05 Eximstats V1.47
1881 Added 'Message too big' to the list of mail rejection
1882 reasons (thanks to Marco Gaiarin).
1884 SC/06 Eximstats V1.48
1885 Mainlog lines which have GMT offsets and are too short to
1886 have a flag are now skipped.
1888 SC/07 Eximstats V1.49 (Alain Williams)
1889 Added the -emptyok flag.
1891 SC/08 Eximstats V1.50
1892 Fixes for obtaining the IP address from reject messages.
1894 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1895 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1896 whitesspace changes from 4.64-PH/27
1898 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1901 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1902 are found, allow negative numbers in numeric criteria)
1904 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1906 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1907 to match changes made in 4.64-PH/43
1909 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1911 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1912 "domain" is actually an IP address, and force a failure. This locks out
1913 those revolvers/nameservers that support "A-for-A" lookups, in
1914 contravention of the specifications.
1916 PH/48 When a host name was looked up from an IP address, and the subsequent
1917 forward lookup of the name timed out, the host name was left in
1918 $sender_host_name, contrary to the specification.
1920 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1921 restricted to single-key lookups, Exim was not diagnosing an error if
1922 * or *@ was used with a query-style lookup.
1924 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1926 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1927 long ago, but noone interested enough thought of it. Let's just say that
1928 the "1.1" means that there are some new functions that weren't there at
1929 some point in the past.
1931 PH/51 Error processing for expansion failure of helo_data from an smtp
1932 transport during callout processing was broken.
1934 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1935 tested/used via the -bh/-bhc/-bs options.
1937 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1938 bug, fixed in subsequent PCRE releases).
1940 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1941 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1943 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1949 SC/01 Use a glob alias rather than an array ref in eximstats generated
1950 parser. This improves both readability and performance.
1952 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1953 Don't display local sender or destination tables in eximstats unless
1954 there is data to show.
1955 Added average volumes into the eximstats top table text output.
1957 SC/03 Collect data on the number of addresses (recipients) as well
1958 as the number of messages in eximstats.
1960 TF/01 Correct an error in the documentation for the redirect router. Exim
1961 does (usually) call initgroups() when daemonizing.
1963 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1964 with consistent privilege compared to when running as a daemon.
1966 TF/03 Note in the spec that $authenticated_id is not set for local
1967 submissions from trusted users.
1969 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1970 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1972 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1973 by adding some example configuration directives to the default
1974 configuration file. A little bit of work is required to uncomment the
1975 directives and define how usernames and passwords are checked, but
1976 there is now a framework to start from.
1978 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1979 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1980 without this. I don't know how relevant this is to other LDAP libraries.
1982 PH/02 Add the verb name to the "unknown ACL verb" error.
1984 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1986 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1988 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1989 directory not expanded when it should be if an expanded home directory
1990 was set for the address (which is overridden by the transport).
1992 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1995 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1996 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1997 because it is too late at that time, and has no effect.
1999 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2000 security issue with \' (bugzilla #107). I could not use the
2001 PQescapeStringConn() function, because it needs a PGconn value as one of
2004 PH/08 When testing addresses using -bt, indicate those final addresses that
2005 are duplicates that would not cause an additional delivery. At least one
2006 person was confused, thinking that -bt output corresponded to deliveries.
2007 (Suppressing duplicates isn't a good idea as you lose the information
2008 about possibly different redirections that led to the duplicates.)
2010 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2011 systems where poll() doesn't work, in particular OS X.
2013 PH/10 Added more information to debugging output for retry time not reached.
2015 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2016 operations in malware.c.
2018 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2021 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2022 syslog_duplication set false, log lines that would normally be written
2023 both the the main log and to the reject log were not written to syslog at
2026 PH/14 In the default configuration, change the use of "message" in ACL warn
2027 statements to "add_header".
2029 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2030 not followed by a command (e.g. "seen endif").
2032 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2033 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2036 PH/17 Added extra conditions to the default value of delay_warning_condition
2040 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2041 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2042 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2045 The Auto-Submitted: and various List- headers are standardised, whereas I
2046 don't think Precedence: ever was.
2048 PH/18 Refactored debugging code in route_finduser() to show more information,
2049 in particular, the error code if getpwnam() issues one.
2051 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2052 This is apparently needed in addition to the PH/07 change above to avoid
2053 any possible encoding problems.
2055 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2056 but not after initializing Perl.
2058 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2059 output them only if debugging. By default they are written stderr,
2060 apparently, which is not desirable.
2062 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2065 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2068 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2070 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2071 authenticated or an ident call has been made. Suppress the default
2072 values for $authenticated_id and $authenticated_sender (but permit -oMai
2073 and -oMas) when testing with -bh.
2075 PH/24 Re-jigged the order of the tests in the default configuration so that the
2076 tests for valid domains and recipients precede the DNS black list and CSA
2077 tests, on the grounds that those ones are more expensive.
2079 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2080 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2081 This bug exists in every version of Exim that I still have, right back to
2084 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2085 However, an attempt to turn on case-sensitivity in a regex key by
2086 including (?-i) didn't work because the subject string was already
2087 lowercased, and the effects were non-intuitive. It turns out that a
2088 one-line patch can be used to allow (?-i) to work as expected.
2094 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2095 other effects) broke the use of negated acl sub-conditions.
2097 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2100 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2101 "Deny" causes Exim to reject the incoming connection with a 554 error.
2102 Unfortunately, if there is a major crisis, such as a disk failure,
2103 tcp-wrappers gives "deny", whereas what one would like would be some
2104 kind of temporary error. A kludge has been added to help with this.
2105 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2106 554 error is used if errno is still zero or contains ENOENT (which occurs
2107 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2110 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2112 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2113 errors. Otherwise a message that provokes a temporary error (when other
2114 messages do not) can cause a whole host to time out.
2116 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2117 addresses were routed directly to files or pipes from a redirect router.
2118 File deliveries just didn't batch; pipe deliveries might have suffered
2121 PH/06 A failure to get a lock for a hints database would erroneously always say
2122 "Failed to get write lock", even when it was really a read lock.
2124 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2125 of 0600. This has been changed to use the value of the lockfile_mode
2126 option (which defaults to 0600).
2128 PH/08 Applied small patch from the Sieve maintainer.
2130 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2131 folder from quota calculations, a direct delivery into this folder messed
2132 up the contents of the maildirsize file. This was because the regex was
2133 used only to exclude .Trash (or whatever) when the size of the mailbox
2134 was calculated. There was no check that a delivery was happening into an
2135 excluded directory. This bug has been fixed by ignoring all quota
2136 processing for deliveries into excluded directories.
2138 PH/10 Added the maildirfolder_create_regex option to appendfile.
2144 PH/01 The code for finding all the local interface addresses on a FreeBSD
2145 system running IPv6 was broken. This may well have applied to all BSD
2146 systems, as well as to others that have similar system calls. The broken
2147 code found IPv4 interfaces correctly, but gave incorrect values for the
2148 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2149 that it would not match correctly against @[] and not recognize the IPv6
2152 PH/02 The ipliteral router was not recognizing addresses of the form user@
2153 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2155 PH/03 Added disable_ipv6.
2157 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2158 decoded form, because it is most often used to construct To: headers
2159 lines in autoreplies, and the decoded form may well be syntactically
2160 invalid. However, $reply_address has leading white space removed, and all
2161 newlines turned into spaces so that the autoreply transport does not
2164 PH/05 If group was specified without a user on a router, and no group or user
2165 was specified on a transport, the group from the router was ignored.
2167 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2168 for visible compile-time settings that can be used to change these
2169 numbers, for those that want even more. Backwards compatibility with old
2170 spool files has been maintained. However, going back to a previous Exim
2171 release will lost any variables that are in spool files.
2173 PH/07 Two small changes when running in the test harness: increase delay when
2174 passing a TCP/IP connection to a new process, in case the original
2175 process has to generate a bounce, and remove special handling of
2176 127.0.0.2 (sic), which is no longer necessary.
2178 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2179 be the same on different OS.
2181 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2184 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2185 whether --show-vars was specified or not
2187 JJ/02 exipick: Added support for new ACL variable spool format introduced
2190 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2191 syntactically invalid From: or Reply-to: line, and a filter used this to
2192 generate an autoreply, and therefore failed to obtain an address for the
2193 autoreply, Exim could try to deliver to a non-existent relative file
2194 name, causing unrelated and misleading errors. What now happens is that
2195 it logs this as a hard delivery error, but does not attempt to create a
2198 PH/11 The exinext utility has a -C option for testing purposes, but although
2199 the given file was scanned by exinext itself; it wasn't being passed on
2200 when Exim was called.
2202 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2203 an end-of-file indication when reading a command response.
2205 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2206 compiled. In many other places in Exim, IPv6 addresses are always
2207 recognized, so I have changed this. It also means that IPv4 domain
2208 literals of the form [IPV4:n.n.n.n] are now always recognized.
2210 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2211 used if the router is not running as root, for example, when verifying at
2212 ACL time, or when using -bh. The debugging output from this situation was
2213 non-existent - all you got was a failure to exec. I have made two
2216 (a) Failures to set uid/gid, the current directory, or a process leader
2217 in a subprocess such as that created by queryprogram now generate
2218 suitable debugging ouput when -d is set.
2220 (b) The queryprogram router detects when it is not running as root,
2221 outputs suitable debugging information if -d is set, and then runs
2222 the subprocess without attempting to change uid/gid.
2224 PH/15 Minor change to Makefile for building test_host (undocumented testing
2227 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2228 additional section of a DNS packet that returns MX or SRV records.
2229 Instead, it always explicitly searches for A/AAAA records. This avoids
2230 major problems that occur when a DNS server includes only records of one
2231 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2232 fixed another bug: if SRV records were looked up and the corresponding
2233 address records were *not* found in the additional section, the port
2234 values from the SRV records were lost.
2236 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2237 using the correct key (the original address) when searching the retry
2238 rules in order to find which one to use for generating the retry hint.
2240 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2241 adding the default one. Similarly, if it contains a Reply-To: header, the
2242 errors_reply_to option, if set, is not used.
2244 PH/19 When calculating a retry time, Exim used to measure the "time since
2245 failure" by looking at the "first failed" field in the retry record. Now
2246 it does not use this if it is later than than the arrival time of the
2247 message. Instead it uses the arrival time. This makes for better
2248 behaviour in cases where some deliveries succeed, thus re-setting the
2249 "first failed" field. An example is a quota failure for a huge message
2250 when small messages continue to be delivered. Without this change, the
2251 "time since failure" will always be short, possible causing more frequent
2252 delivery attempts for the huge message than are intended.
2253 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2255 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2256 $1, $2, $3) because the numerical variables can be reset during some
2257 expansion items (e.g. "match"), thereby losing the authentication data.
2259 PH/21 Make -bV show the size of off_t variables so that the test suite can
2260 decide whether to run tests for quotas > 2G.
2262 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2263 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2264 filecount value is greater than 2G or if a quota value is greater than 2G
2265 on a system where the size of off_t is not greater than 4, a panic error
2268 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2269 never match. The debug and -bh output now contains an explicit error
2270 message indicating a malformed IPv4 address or mask.
2272 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2273 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2274 PH/23 above applies.
2276 PH/25 Do not write to syslog when running in the test harness. The only
2277 occasion when this arises is a failure to open the main or panic logs
2278 (for which there is an explicit test).
2280 PH/26 Added the /no_tell option to "control=freeze".
2282 PH/27 If a host name lookup failed very early in a connection, for example, if
2283 the IP address matched host_lookup and the reverse lookup yielded a name
2284 that did not have a forward lookup, an error message of the form "no IP
2285 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2286 could be logged. Now it outputs the IP address instead of "NULL".
2288 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2289 allows the sender and the authenticated sender to be set when
2290 submitting a message from within Exim. Since child_open_exim() is
2291 documented for local_scan(), the new function should be too.
2293 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2294 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2295 results in an empty string is now treated as unset.
2297 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2299 PH/31 Added sender_verify_fail logging option.
2301 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2302 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2303 tidied the source and removed it altogether.
2305 PH/33 When a queue run was abandoned because the load average was too high, a
2306 log line was always written; now it is written only if the queue_run log
2307 selector is set. In addition, the log line for abandonment now contains
2308 information about the queue run such as the pid. This is always present
2309 in "start" and "stop" lines but was omitted from the "abandon" line.
2311 PH/34 Omit spaces between a header name and the colon in the error message that
2312 is given when verify = headers_syntax fails (if there are lots of them,
2313 the message gets confusing).
2315 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2316 names, as there are now some PTR records that contain slashes. This check
2317 is only to protect against broken name servers that fall over on strange
2318 characters, so the fact that it applies to all lookups doesn't matter.
2320 PH/36 Now that the new test suite is complete, we can remove some of the
2321 special code in Exim that was needed for the old test suite. For example,
2322 sorting DNS records because real resolvers return them in an arbitrary
2323 order. The new test suite's fake resolver always returns records in the
2326 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2327 bounces) except when queue_only is set, to avoid logging races between
2328 the different processes.
2330 PH/38 Panic-die if .include specifies a non-absolute path.
2332 PH/39 A tweak to the "H" retry rule from its user.
2334 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2335 a label. They prevented compilation on older perls.
2337 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2338 a warning to be raised on newish perls.
2340 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2341 on queue. Changes to match documented behaviour of showing count of
2342 messages matching specified criteria.
2344 PH/40 Changed the default ident timeout from 30s to 5s.
2346 PH/41 Added support for the use of login_cap features, on those BSD systems
2347 that have them, for controlling the resources used by pipe deliveries.
2349 PH/42 The content-scanning code uses fopen() to create files in which to put
2350 message data. Previously it was not paying any attention to the mode of
2351 the files. Exim runs with umask(0) because the rest of the code creates
2352 files with open(), and sets the required mode explicitly. Thus, these
2353 files were ending up world-writeable. This was not a big issue, because,
2354 being within the spool directory, they were not world-accessible. I have
2355 created a function called modefopen, which takes an additional mode
2356 argument. It sets umask(777), creates the file, chmods it to the required
2357 mode, then resets the umask. All the relevant calls to fopen() in the
2358 content scanning code have been changed to use this function.
2360 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2361 to 24 hours. This avoids potential overflow problems when processing G
2362 and H retry rules. I suspect nobody ever tinkers with this value.
2364 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2366 PH/45 When the plaintext authenticator is running as a client, the server's
2367 challenges are checked to ensure they are valid base64 strings. By
2368 default, the authentication attempt is cancelled if an invalid string is
2369 received. Setting client_ignore_invalid_base64 true ignores these errors.
2370 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2371 they are received. Thus, the responses can be made to depend on the
2372 challenges. If an invalid string is ignored, an empty string is placed in
2375 PH/46 Messages that are created by the autoreply transport now contains a
2376 References: header, in accordance with RFCs 2822 and 3834.
2378 PH/47 Added authenticated_sender_force to the smtp transport.
2380 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2382 PH/49 Installed latest patch from the Sieve maintainer.
2384 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2385 was also set, the appendfile transport was unnecessarily scanning a
2386 directory of message files (e.g. for maildir delivery) to find the count
2387 of files (along with the size), even though it did not need this
2388 information. It now does the scan only if it needs to find either the
2389 size of the count of files.
2391 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2393 PH/52 Two bugs concerned with error handling when the smtp transport is
2396 (i) Exim was not creating retry information for temporary errors given
2397 for individual recipients after the DATA command when the smtp transport
2398 was used in LMTP mode. This meant that they could be retried too
2399 frequently, and not timed out correctly.
2401 (ii) Exim was setting the flag that allows error details to be returned
2402 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2403 recipients that were returned after the DATA command.
2405 PH/53 This is related to PH/52, but is more general: for any failing address,
2406 when detailed error information was permitted to be returned to the
2407 sender, but the error was temporary, then after the final timeout, only
2408 "retry timeout exceeded" was returned. Now it returns the full error as
2409 well as "retry timeout exceeded".
2411 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2412 do this, and (what is worse) MTAs that accept it.
2414 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2415 will now be deprecated.
2417 PH/56 New os.c-cygwin from the Cygwin maintainer.
2419 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2420 formats (previously only available in exim formats via -bpr, -bpru,
2421 and -bpra. Now also available in native and exiqgrep formats)
2423 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2424 with very large, slow to parse queues
2426 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2428 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2430 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2431 responses to authentication challenges, though it was showing the
2432 challenges; (ii) I've removed the CR characters from the debug output for
2435 PH/58 Allow for the insertion of a newline as well as a space when a string
2436 is turned into more than one encoded-word during RFC 2047 encoding. The
2437 Sieve code now uses this.
2439 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2440 data_4xx, lost_connection, tls_required.
2442 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2443 message was being sent as an SMTP response.
2445 PH/61 Add -l and -k options to exicyclog.
2447 PH/62 When verifying, if an address was redirected to one new address, so that
2448 verification continued, and the new address failed or deferred after
2449 having set something in $address_data, the value of $address_data was not
2450 passed back to the ACL. This was different to the case when no
2451 redirection occurred. The value is now passed back in both cases.
2453 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2454 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2455 use login_cap.h, so on its own it isn't the distinguishing feature. The
2456 new name refers directly to the setclassresources() function.
2458 PH/65 Added configuration files for NetBSD3.
2460 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2462 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2463 is preferred over IPv4.
2465 PH/68 The bounce_return_message and bounce_return_body options were not being
2466 honoured for bounces generated during the reception of non-SMTP messages.
2467 In particular, this applied to messages rejected by the ACL. This bug has
2468 been fixed. However, if bounce_return_message is true and bounce_return_
2469 body is false, the headers that are returned for a non-SMTP message
2470 include only those that have been read before the error was detected.
2471 (In the case of an ACL rejection, they have all been read.)
2473 PH/69 The HTML version of the specification is now built in a directory called
2474 spec_html instead of spec.html, because the latter looks like a path with
2475 a MIME-type, and this confuses some software.
2477 PH/70 Catch two compiler warnings in sieve.c.
2479 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2480 function verify_get_ident() calls ip_connect() to connect a socket, but
2481 if the "connect()" function timed out, ip_connect() used to close the
2482 socket. However, verify_get_ident() also closes the socket later, and in
2483 between Exim writes to the log, which may get opened at this point. When
2484 the socket was closed in ip_connect(), the log could get the same file
2485 descriptor number as the socket. This naturally causes chaos. The fix is
2486 not to close the socket in ip_connect(); the socket should be closed by
2487 the function that creates it. There was only one place in the code where
2488 this was missing, in the iplookup router, which I don't think anybody now
2489 uses, but I've fixed it anyway.
2491 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2492 well as to direct DNS lookups. Otherwise the handling of names in host
2493 lists is inconsistent and therefore confusing.
2499 PH/01 Two changes to the default runtime configuration:
2501 (1) Move the checks for relay_from_hosts and authenticated clients from
2502 after to before the (commented out) DNS black list checks.
2504 (2) Add control=submission to the relay_from_hosts and authenticated
2505 clients checks, on the grounds that messages accepted by these
2506 statements are most likely to be submissions.
2508 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2510 (1) Generate an error if the third argument for the ${prvs expansion is
2513 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2516 (3) Reset the variables that are obtained from the first argument of
2517 ${prvscheck and used in the second argument before leaving the code,
2518 because their memory is reclaimed, so using them afterwards may do
2521 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2522 one (it's much easier than Tom thought :-).
2524 (5) Because of (4), we can now allow for the use of $prvscheck_result
2525 inside the third argument.
2527 PH/03 For some reason, the default setting of PATH when running a command from
2528 a pipe transport was just "/usr/bin". I have changed it to
2531 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2532 anything to be listed in the output from -bV.
2534 PH/05 When a filter generated an autoreply, the entire To: header line was
2535 quoted in the delivery log line, like this:
2537 => >A.N.Other <ano@some.domain> <original@ddress> ...
2539 This has been changed so that it extracts the operative address. There
2540 may be more than one such address. If so, they are comma-separated, like
2543 => >ano@some.domain,ona@other.domain <original@ddress> ...
2545 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2546 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2547 not being looked up in the rDNS to get a host name, Exim was showing the
2548 IP address twice in Received: lines, even though the IP addresses were
2549 identical. For example:
2551 Received: from [1.2.3.4] (helo=[1.2.3.4])
2553 However, if the real host name was known, it was omitting the HELO data
2554 if it matched the actual IP address. This has been tidied up so that it
2555 doesn't show the same IP address twice.
2557 PH/07 When both +timestamp and +memory debugging was on, the value given by
2558 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2559 called by the debug printing, thereby overwriting the timestamp buffer.
2560 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2562 PH/08 When the original message was included in an autoreply transport, it
2563 always said "this is a copy of the message, including all the headers",
2564 even if body_only or headers_only was set. It now gives an appropriate
2567 PH/09 Applied a patch from the Sieve maintainer which:
2569 o fixes some comments
2570 o adds the (disabled) notify extension core
2571 o adds some debug output for the result of if/elsif tests
2572 o points to the current vacation draft in the documentation
2573 and documents the missing references header update
2577 o fixes a bug in processing the envelope test (when testing
2578 multiple envelope elements, the last element determinted the
2581 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2582 Electronic Mail") by including:
2584 Auto-submitted: auto-generated
2586 in the messages that it generates (bounce messages and others, such as
2587 warnings). In the case of bounce messages for non-SMTP mesages, there was
2588 also a typo: it was using "Auto_submitted" (underscore instead of
2589 hyphen). Since every message generated by Exim is necessarily in response
2590 to another message, thes have all been changed to:
2592 Auto-Submitted: auto-replied
2594 in accordance with these statements in the RFC:
2596 The auto-replied keyword:
2598 - SHOULD be used on messages sent in direct response to another
2599 message by an automatic process,
2601 - MUST NOT be used on manually-generated messages,
2603 - MAY be used on Delivery Status Notifications (DSNs) and Message
2604 Disposition Notifications (MDNs),
2606 - MUST NOT be used on messages generated by automatic or periodic
2607 processes, except for messages which are automatic responses to
2610 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2611 to the default Received: header definition.
2613 PH/12 Added log selector acl_warn_skipped (default on).
2615 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2616 variables because (a) they are in the wrong storage pool and (b) even if
2617 they were copied, it wouldn't work properly because of the caching.
2619 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2620 checking when decoding. Apparently there are clients that generate
2621 overlong encoded strings. Why am I not surprised?
2623 PH/15 If the first argument of "${if match_address" was not empty, but did not
2624 contain an "@" character, Exim crashed. Now it writes a panic log message
2625 and treats the condition as false.
2627 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2629 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2630 extension "envelope-auth". The code is finished and in agreement with
2631 other implementations, but there is no documentation so far and in fact,
2632 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2633 not changing the active code.
2635 Print executed "if" and "elsif" statements when debugging is used. This
2636 helps a great deal to understand what a filter does.
2638 Document more things not specified clearly in RFC3028. I had all this
2639 sorted out, when out of a sudden new issues came to my mind. Oops."
2641 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2644 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2645 canonical form (as documented). However, after a host name lookup from
2646 the IP address, check_host() was doing a simple string comparison with
2647 addresses acquired from the DNS when checking that the found name did
2648 have the original IP as one of its addresses. Since any found IPv6
2649 addresses are likely to be in abbreviated form, the comparison could
2650 fail. Luckily, there already exists a function for doing the comparison
2651 by converting both addresses to binary, so now that is used instead of
2652 the text comparison.
2654 PH/20 There was another similar case to PH/19, when a complete host name was
2655 given in a host list; looking up its IP address could give an abbreviated
2656 form, whereas the current host's name might or might not be abbreviated.
2657 The same fix has been applied.
2663 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2664 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2667 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2668 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2670 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2672 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2673 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2674 Also, for "Auto-Submitted", treat anything other than "no" as
2675 non-personal, in accordance with RFC 3834. (Previously it treated
2676 anything starting "auto-" as non-personal.)
2678 TF/01 The control=submission/name=... option had a problem with syntax
2679 errors if the name included a slash character. The /name= option
2680 now slurps the rest of the string, so it can include any characters
2681 but it must come last in the list of options (after /sender_retain
2684 PH/05 Some modifications to the interface to the fake nameserver for the new
2692 TK/01 Added the "success_on_redirect" address verification option. See
2693 NewStuff for rationale and an example.
2695 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2697 PH/02 Patch to exigrep to allow it to work on syslog lines.
2699 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2700 fread() to skip over the body file's header line, because in Cygwin the
2701 header line is locked and is inaccessible.
2703 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2704 co-exist for some time) to make it clear that it is the Exim ID that is
2705 referenced, not the Message-ID: header line.
2707 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2708 string_format() function, because snprintf() does not exist on all
2711 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2712 ${stat: expansion item.
2714 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2715 protocol synchronization error", to keep the pedants happy.
2717 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2718 well as for IRIX systems, when gcc is being used. See the host.c source
2721 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2723 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2726 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2727 transport and to the smtp transport in LMTP mode.
2729 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2731 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2732 run a filter in a subprocess. This could lead to confusion in subsequent
2733 lookups in the parent process. There should also be a search_tidyup() at
2734 the end of the subprocess.
2736 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2737 only if the host matched helo_try_verify_hosts, which caused the
2738 verification to occur when the EHLO/HELO command was issued. The ACL just
2739 tested the remembered result. Now, if a previous verification attempt has
2740 not happened, "verify = helo" does it there and then.
2742 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2744 TK/03 Fix log output including CR from clamd.
2746 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2747 exist provoked a memory error which could cause a segfault.
2749 PH/15 Installed PCRE 6.2
2751 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2753 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2754 of the problem. Specifically, suggested +O2 rather than +O1 for the
2757 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2759 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2760 record, the port was not correctly logged when the outgoing_port log
2761 selector was set (it logged the transort's default port).
2763 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2764 fallback_hosts, and "hosts" in the smtp transport.
2766 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2767 host errors such as "Connection refused".
2769 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2770 authentication with radiusclient 0.4.9:
2772 - Error returned from rc_read_config was caught wrongly
2773 - Username/password not passed on to radius server due to wrong length.
2775 The presumption is that some radiusclient API changes for 4.51/PH/17
2776 were not taken care of correctly. The code is still untested by me (my
2777 Linux distribution still has 0.3.2 of radiusclient), but it was
2778 contributed by a Radius user.
2780 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2781 expanding the "port" option of the smtp transport.
2783 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2784 while reading a MIME header. Thanks to Tom Hughes for a patch.
2786 PH/24 Include config.h inside local_scan.h so that configuration settings are
2789 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2790 that in an ACL for RCPT (for example), you can examine exactly what was
2793 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2794 commands, but it was not correctly comparing the address with the actual
2795 client host address. Thus, it would show the EHLO address in Received:
2796 header lines when this was not necessary.
2798 PH/27 Added the % operator to ${eval:}.
2800 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2801 it should be ignoring failures (because with -C, for example, it has lost
2802 privilege). It wasn't ignoring creation failures other than "already
2805 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2808 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2809 because an input line was too long, either on its own, or by virtue of
2810 too many continuations, the temporary file was not being removed, and the
2811 return code was incorrect.
2813 PH/31 Missing "BOOL" in function definition in filtertest.c.
2815 PH/32 Applied Sieve patches from the maintainer.
2817 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2819 PH/33 Added "verify = not_blind".
2821 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2822 Local/Makefile (with some defaults set). These are used in built scripts
2823 such as exicyclog, but they have never been used in the exim_install
2824 script (though there are many overriding facilities there). I have
2825 arranged that the exim_install script now takes note of these two
2828 PH/35 Installed configuration files for Dragonfly.
2830 PH/36 When a locally submitted message by a trusted user did not contain a
2831 From: header, and the sender address was obtained from -f or from an SMTP
2832 MAIL command, and the trusted user did not use -F to supply a sender
2833 name, $originator_name was incorrectly used when constructing a From:
2834 header. Furthermore, $originator_name was used for submission mode
2835 messages from external hosts without From: headers in a similar way,
2836 which is clearly wrong.
2838 PH/37 Added control=suppress_local_fixups.
2840 PH/38 When log_selector = +received_sender was set, and the addition of the
2841 sender made the log line's construction buffer exactly full, or one byte
2842 less than full, an overflow happened when the terminating "\n" was
2845 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2846 when the result of a list match is failure because a DNS lookup failed.
2848 PH/40 RM_COMMAND is now used in the building process.
2850 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2851 the "build-* directories that it finds.
2853 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2854 address in a domain literal was a prefix of an interface address.
2856 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2857 when verifying a sender address, unless rewrite_headers is false.
2859 PH/44 Wrote a long comment about why errors_to addresses are verified as
2860 recipients, not senders.
2862 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2863 the ratelimit ACL was added.
2865 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2867 PH/46 Added extra information about PostgreSQL errors to the error string.
2869 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2870 avoiding the need to install special zones in a real server. This is
2871 backwards compatible; if it can't find the fake resolver, it drops back.
2872 Thus, both old and new test suites can be run.
2874 TF/02 Added util/ratelimit.pl
2876 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2877 clock is set back in time.
2879 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2880 Candler <B.Candler@pobox.com>.
2882 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2883 OK for addresses that are the result of redirections.
2885 PH/48 A number of further additions for the benefit of the new test suite,
2886 including a fake gethostbyname() that interfaces to the fake DNS resolver
2889 TF/06 The fix for widen_domains has also been applied to qualify_single and
2890 search_parents which are the other dnslookup options that can cause
2893 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2896 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2898 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2899 service" if it encountered an MX record with an empty target hostname.
2900 The message is now "an MX or SRV record indicated no SMTP service".
2902 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2903 if the DNS of the sending site is misconfigured. This is quite a
2904 common situation. This change restores the behaviour of treating a
2905 helo verification defer as a failure.
2907 PH/51 If self=fail was set on a router, the bounce message did not include the
2908 actual error message.
2914 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2916 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2917 command itself ended in error, the underlying message about the transport
2918 filter timeout was being overwritten with the pipe command error. Now the
2919 underlying error message should be appended to the second error message.
2921 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2923 PH/02 Reduce the amount of output that "make" produces by default. Full output
2924 can still be requested.
2926 PH/03 The warning log line about a condition test deferring for a "warn" verb
2927 was being output only once per connection, rather than after each
2928 occurrence (because it was using the same function as for successful
2929 "warn" verbs). This seems wrong, so I have changed it.
2931 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2932 it should not have, which might have caused a crash in the right
2933 circumstances, but probably never did.
2935 PH/04 Installed a modified version of Tony Finch's patch to make submission
2936 mode fix the return path as well as the Sender: header line, and to
2937 add a /name= option so that you can make the user's friendly name appear
2940 TF/03 Added the control = fakedefer ACL modifier.
2942 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2943 Mark Lowes for thorough testing.
2945 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2947 TK/03 Merged latest SRS patch from Miles Wilton.
2949 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2950 with the definition in sysexits.h (which is #included earlier).
2951 Fortunately, Exim does not actually use EX_OK. The code used to try to
2952 preserve the sysexits.h value, by assumimg that macro definitions were
2953 scanned for macro replacements. I have been disabused of this notion,
2954 so now the code just undefines EX_OK before #including unistd.h.
2956 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2957 in the smtp transport. When a block could not be written in a single
2958 write() function, the timeout was being re-applied to each part-write.
2959 This seems wrong - if the receiver was accepting one byte at a time it
2960 would take for ever. The timeout is now adjusted when this happens. It
2961 doesn't have to be particularly precise.
2963 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2964 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2966 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2967 without a domain if the check to <postmaster@domain> fails.
2969 SC/01 Eximstats: added -xls and the ability to specify output files
2970 (patch written by Frank Heydlauf).
2972 SC/02 Eximstats: use FileHandles for outputing results.
2974 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2976 SC/04 Eximstats: fixed display of large numbers with -nvr option
2978 SC/05 Eximstats: fixed merging of reports with empty tables.
2980 SC/06 Eximstats: added the -include_original_destination flag
2982 SC/07 Eximstats: removed tabs and trailing whitespace.
2984 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2986 TK/06 MBOX spool code: Add real "From " MBOX separator line
2987 so the .eml file is really in mbox format (even though
2988 most programs do not really care). Patch from Alex Miller.
2990 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2991 The latter is generated from $received_to and is only set if the
2992 message has one envelope recipient. SA can use these headers,
2993 obviously out-of-the-box. Patch from Alex Miller.
2995 PH/08 The ${def test on a variable was returning false if the variable's
2996 value was "0", contrary to what the specification has always said!
2997 The result should be true unless the variable is empty.
2999 PH/09 The syntax error of a character other than { following "${if
3000 def:variable_name" (after optional whitespace) was not being diagnosed.
3001 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3002 accidental colon was present, for example, could give incorrect results.
3004 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3005 result is used (not including appendfile, where other changes are about
3008 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3009 This involved changing a lot of size variables from int to off_t. It
3010 should work with maildirs and everything.
3012 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3013 spamd dying while we are connected to it.
3015 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3018 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3019 function for BDB 4.3.
3021 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3023 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3024 the sources for PCRE have been re-arranged and more files are now
3027 PH/15 The code I had for printing potentially long long variables in PH/11
3028 above was not the best (it lost precision). The length of off_t variables
3029 is now inspected at build time, and an appropriate printing format (%ld
3030 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3031 to be "long long int" or "long int". This is needed for the internal
3032 formatting function string_vformat().
3034 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3035 the configuration file to be ":syslog", then the script "guesses" where
3036 the logs files are, rather than using the compiled in default. In our
3037 case the guess is not the same as the compiled default, so the script
3038 suddenly stopped working when I started to use syslog. The patch checks
3039 to see if log_file_path is "". If so, it attempts to read it from exim
3040 with no configuration file to get the compiled in version, before it
3041 falls back to the previous guessing code."
3043 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3044 implementing BATV in an Exim configuration. See NewStuff for the gory
3047 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3048 Makefile that are specific to HP-UX.
3050 PH/18 If the "use_postmaster" option was set for a recipient callout together
3051 with the "random" option, the postmaster address was used as the MAIL
3052 FROM address for the random test, but not for the subsequent recipient
3053 test. It is now used for both.
3055 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3056 patch removes a few documentation additions to RFC 3028, because the
3057 latest draft now contains them. It adds the new en;ascii-case comparator
3058 and a new error check for 8bit text in MIME parts. Comparator and
3059 require names are now matched exactly. I enabled the subaddress
3060 extension, but it is not well tested yet (read: it works for me)."
3062 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3063 rework some of the code of TK/09 above to avoid the hardwired use of
3064 "%lld" and "long long". Replaced the call to snprintf() with a call to
3067 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3068 records point to non-existent hosts", "retry timeout exceeded", and
3069 "retry time not reached for any host after a long failure period".
3071 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3072 experimental DomainKeys support:
3074 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3075 (2) On an error such as an illegally used "control", the wrong name for
3076 the control was given.
3078 These problems did NOT occur unless DomainKeys support was compiled.
3080 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3082 PH/24 Added ${if match_ip condition.
3084 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3085 databases so that it will be absolutely obvious if a crash occurs in the
3086 DB library. This is a regular occurrence (often caused by mis-matched
3089 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3090 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3091 picked up on a user's system that detects such things. There doesn't seem
3092 to be a gcc warning option for this - only an attribute that has to be
3093 put on the function's prototype. It seems that in Fedora Core 4 they have
3094 set this on a number of new functions. No doubt there will be more in due
3097 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3098 specify a transport. However, if an address that was verified by such a
3099 router was the subject of a callout, Exim crashed because it tried to
3100 read the rcpt_include_affixes from the non-existent transport. Now it
3101 just assumes that the setting of that option is false. This bug was
3102 introduced by 4.51/PH/31.
3104 PH/28 Changed -d+all to exclude +memory, because that information is very
3105 rarely of interest, but it makes the output a lot bigger. People tend to
3106 do -d+all out of habit.
3108 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3109 code in os-type was giving problems when libc.so lives in lib64, like on
3112 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3113 aren't the modern standard, and it seems that some systems' include files
3114 don't always have them. Exim was already checking for some of the newer
3115 ones like T_AAAA, and defining it itself. I've added checks for all the
3116 record types that Exim uses.
3118 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3119 not automatically generating a new one, as it is supposed to. This
3120 prevented TLS from working. If the file did exist, but contained invalid
3121 data, a new version was generated, as expected. It was only the case of a
3122 non-existent file that was broken.
3124 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3125 with a change in libdomainkeys > 0.64.
3127 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3128 from DNS. If the selector record carries the flag, it now has
3129 precedence over the domain-wide flag.
3131 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3133 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3134 the use of a port name that isn't defined in /etc/services) occurred, the
3135 message was deferred as in a normal delivery, and thus remained on the
3136 spool, instead of being failed because of the mua_wrapper setting. This
3137 is now fixed, and I tidied up some of the mua_wrapper messages at the
3140 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3141 the messages in a hash of arrays rather than using individual hashes.
3142 This is a bit cleaner and results in dramatic memory savings, albeit
3143 at a slight CPU cost.
3145 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3146 as requested by Marc Sherman.
3148 SC/10 Eximstats: added histograms for user specified patterns as requested
3151 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3153 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3154 fopen() in the content-scanning modules that did not already have it.
3160 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3161 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3163 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3165 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3167 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3168 to test Sieve filters that use "vacation".
3170 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3171 that changes the way the GnuTLS parameters are stored in the cache file.
3172 The new format can be generated externally. For backward compatibility,
3173 if the data in the cache doesn't make sense, Exim assumes it has read an
3174 old-format file, and it generates new data and writes a new file. This
3175 means that you can't go back to an older release without removing the
3178 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3179 work if there are any delivery delays because "one_time" forces the
3180 parent to be marked "delivered", so its unseen clone is never tried
3181 again. For this reason, Exim now forbids the simultaneous setting of
3184 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3185 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3186 entry if you want to know the details. The fix, however, neglected to
3187 consider the case when local delivery batching is involved. The test for
3188 "previously delivered" was not happening when checking to see if an
3189 address could be batched with a previous (undelivered) one; under
3190 certain circumstances this could lead to multiple deliveries to the same
3193 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3194 in its include files, and this causes problems building Exim.
3196 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3197 header_syntax) but Exim was just ignoring anything given after a slash.
3198 In particular, this caused confusion with an attempt to use "verify =
3199 reverse_host_lookup/defer_ok". An error is now given when options are
3200 supplied for verify items that do not have them. (Maybe reverse_host_
3201 lookup should have a defer_ok option, but that's a different point.)
3203 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3204 defined by RFC 821) to 2048, because there were problems with some AUTH
3205 commands, and RFC 1869 says the size should be increased for extended
3206 SMTP commands that take arguments.
3208 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3211 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3212 "unknown" error; now it says that the functionality isn't in the binary.
3214 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3215 an address' error message when a string expansion fails (syntax or
3216 whatever). Otherwise the password may appear in the log. Following change
3217 PH/42 below, there is no longer a chance of it appearing in a bounce
3220 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3222 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3223 of its addresses was ever tried. (Bugzilla bug #2.)
3225 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3226 the result incorrectly in the debug output. (It correctly added a newline
3227 to what was transported.)
3229 TF/01 Added $received_time.
3231 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3232 commented out examples of how to interface to a virus scanner and to
3233 SpamAssassin. Also added commented examples of av_scanner and
3234 spamd_address settings.
3236 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3237 and controls are allowed in which ACLs. There were a couple of minor
3238 errors. Some of the entries in the conditions table (which is a table of
3239 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3240 negation of where the condition IS allowed.
3242 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3244 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3245 header file does not have a version number, so I've had to invent a new
3246 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3247 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3248 radiusclient), but it was contributed by a Radius user.
3250 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3251 files or directories, for OpenSSL.
3253 PH/19 When an Exim process that is running as root has to create an Exim log
3254 file, it does so in a subprocess that runs as exim:exim so as to get the
3255 ownership right at creation (otherwise, other Exim processes might see
3256 the file with the wrong ownership). There was no test for failure of this
3257 fork() call, which would lead to the process getting stuck as it waited
3258 for a non-existent subprocess. Forks do occasionally fail when resources
3259 run out. I reviewed all the other calls to fork(); they all seem to check
3262 PH/20 When checking for unexpected SMTP input at connect time (before writing
3263 the banner), Exim was not dealing correctly with a non-positive return
3264 from the read() function. If the client had disconnected by this time,
3265 the result was a log entry for a synchronization error with an empty
3266 string after "input=" when read() returned zero. If read() returned -1
3267 (an event I could not check), uninitialized data bytes were printed.
3268 There were reports of junk text (parts of files, etc) appearing after
3271 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3273 PH/22 Added support for macro redefinition, and (re)definition in between
3274 driver and ACL definitions.
3276 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3277 forgetting to use the resulting value; it was using the unexpanded value.
3279 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3280 hadn't been configured. The fix is from Juergen Kreileder, who
3281 understands it better than I do:
3283 "Here's what I see happening with three configured cyrus_sasl
3284 authenticators configured (plain, login, cram-md5):
3286 On startup auth_cyrus_sasl_init() gets called for each of these.
3287 This means three calls to sasl_listmech() without a specified mech_list.
3288 => SASL tests which mechs of all available mechs actually work
3289 => three warnings about OTP not working
3290 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3292 With the patch, sasl_listmech() also gets called three times. But now
3293 SASL's mech_list option is set to the server_mech specified in the the
3294 authenticator. Or in other words, the answer from sasl_listmech()
3295 gets limited to just the mech you're testing for (which is different
3297 => the return list contains just 'plain' or 'login', 'cram-md5' or
3298 nothing depending on the value of ob->server_mech.
3300 I've just tested the patch: Authentication still works fine,
3301 unavailable mechs specified in the exim configuration are still
3302 caught, and the auth.log warnings about OTP are gone."
3304 PH/25 When debugging is enabled, the contents of the command line are added
3305 to the debugging output, even when log_selector=+arguments is not
3308 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3309 answer is "GNU", and only if the return is "GNU/something" is the answer
3312 PH/27 $acl_verify_message is now set immediately after the failure of a
3313 verification in an ACL, and so is available in subsequent modifiers. In
3314 particular, the message can be preserved by coding like this:
3316 warn !verify = sender
3317 set acl_m0 = $acl_verify_message
3319 Previously, $acl_verify_message was set only while expanding "message"
3320 and "log_message" when a very denied access.
3322 PH/28 Modified OS/os.c-Linux with
3324 -#ifndef OS_LOAD_AVERAGE
3325 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3327 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3328 nomenclature these days.)
3330 PH/29 Installed patch from the Sieve maintainer that adds the options
3331 sieve_useraddress and sieve_subaddress to the redirect router.
3333 PH/30 In these circumstances:
3334 . Two addresses routed to the same list of hosts;
3335 . First host does not offer TLS;
3336 . First host accepts first address;
3337 . First host gives temporary error to second address;
3338 . Second host offers TLS and a TLS session is established;
3339 . Second host accepts second address.
3340 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3341 and peerdn, if requested) that were in fact used only for the second
3344 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3345 attention to any local part prefix or suffix that was matched by the
3346 router that accepted the address. It now behaves in the same way as it
3347 does for delivery: the affixes are removed from the local part unless
3348 rcpt_include_affixes is set on the transport.
3350 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3351 timeout during the DATA phase of an incoming message.
3353 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3354 applied a patch sanctioned by the Sieve maintainer.
3356 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3357 the uid or gid is negative. A case of a negative gid caused this to be
3358 noticed. The fix allows for either to be negative.
3360 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3361 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3364 PH/36 The change PH/12 above was broken. Fixed it.
3366 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3367 the grounds that routing the same address twice would always produce the
3368 same answer. This might have been true once, but it is certainly no
3369 longer true now. Routing a child address may depend on the previous
3370 routing that produced that child. Some complicated redirection strategies
3371 went wrong when messages had multiple recipients, and made Exim's
3372 behaviour dependent on the order in which the addresses were given.
3374 I have moved the duplicate checking until after the routing is complete.
3375 Exim scans the addresses that are assigned to local and remote
3376 transports, and removes any duplicates. This means that more work will be
3377 done, as duplicates will always all be routed, but duplicates are
3378 presumably rare, so I don't expect this is of any significance.
3380 For deliveries to pipes, files, and autoreplies, the duplicate checking
3381 still happens during the routing process, since they are not going to be
3384 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3385 It corrects a timeout issue with spamd. This is Ian's comment: "The
3386 background is that sometimes spamd either never reads data from a
3387 connection it has accepted, or it never writes response data. The exiscan
3388 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3389 blindly assumes that writes won't block so it may never time out."
3391 PH/39 Allow G after quota size as well as K and M.
3393 PH/40 The value set for $authenticated_id in an authenticator may not contain
3394 binary zeroes or newlines because the value is written to log lines and
3395 to spool files. There was no check on this. Now the value is run through
3396 the string_printing() function so that such characters are converted to
3397 printable escape sequences.
3399 PH/41 $message_linecount is a new variable that contains the total number of
3400 lines in the message. Compare $body_linecount, which is the count for the
3403 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3404 bounce and delay warning messages, except in certain special cases, which
3407 (a) An SMTP error message from a remote host;
3408 (b) A message specified in a :fail: redirection;
3409 (c) A message specified in a "fail" command in a system filter;
3410 (d) A message specified in a FAIL return from the queryprogram router;
3411 (e) A message specified by the cannot_route_message router option.
3413 In these cases only, Exim does include the error details in bounce and
3414 warning messages. There are also a few cases where bland messages such
3415 as "unrouteable address" or "local delivery error" are given.
3417 PH/43 $value is now also set for the "else" part of a ${run expansion.
3419 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3420 being worked on, but at least Exim now implements the latest version to
3423 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3424 process to complete was treated as a delivery failure, a timeout while
3425 writing the message to the pipe was logged, but erroneously treated as a
3426 successful delivery. Such timeouts include transport filter timeouts. For
3427 consistency with the overall process timeout, these timeouts are now
3428 treated as errors, giving rise to delivery failures by default. However,
3429 there is now a new Boolean option for the pipe transport called
3430 timeout_defer, which, if set TRUE, converts the failures into defers for
3431 both kinds of timeout. A transport filter timeout is now identified in
3434 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3435 systems where "make" and "gmake" are different, calling "gmake" at top
3436 level broke things. I've arranged for the value of $(MAKE) to be passed
3437 from the Makefile to this script so that it can call the same version of
3441 A note about Exim versions 4.44 and 4.50
3442 ----------------------------------------
3444 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3445 changes of various kinds. As a consequence, a big documentation update was
3446 needed. This delayed the release for rather longer than seemed good, especially
3447 in the light of a couple of (minor) security issues. Therefore, the changes
3448 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3449 release. So 4.44 and 4.50 are in effect two different branches that both start
3452 I have left the 4.50 change log unchanged; it contains all the changes since
3453 4.43. The change log for 4.44 is below; many of its items are identical to
3454 those for 4.50. This seems to be the most sensible way to preserve the
3455 historical information.
3461 1. Minor wording change to the doc/README.SIEVE file.
3463 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3464 computation of the current number of files was incorrect.
3466 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3467 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3470 4. Give more explanation in the error message when the command for a transport
3471 filter fails to execute.
3473 5. There are several places where Exim runs a non-Exim command in a
3474 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3475 was being done only for the command run by the queryprogram router. It is
3476 now done for all such subprocesses. The other cases are: ${run, transport
3477 filters, and the commands run by the lmtp and pipe transports.
3479 6. Added CONFIGURE_GROUP build-time option.
3481 7. Some older OS have a limit of 256 on the maximum number of file
3482 descriptors. Exim was using setrlimit() to set 1000 as a large value
3483 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3484 systems. I've change it so that if it can't get 1000, it tries for 256.
3486 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3487 was an oversight, and furthermore, ever since the addition of extra
3488 controls (e.g. 4.43/32), the checks on when to allow different forms of
3489 "control" were broken. There should now be diagnostics for all cases when a
3490 control that does not make sense is encountered.
3492 9. Added the /retain_sender option to "control=submission".
3494 10. $recipients is now available in the predata ACL (oversight).
3496 11. Tidy the search cache before the fork to do a delivery from a message
3497 received from the command line. Otherwise the child will trigger a lookup
3498 failure and thereby defer the delivery if it tries to use (for example) a
3499 cached ldap connection that the parent has called unbind on.
3501 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3502 of $address_data from the recipient verification was clobbered by the
3503 sender verification.
3505 13. The value of address_data from a sender verification is now available in
3506 $sender_address_data in subsequent conditions in the ACL statement.
3508 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3510 15. Added a new option "connect=<time>" to callout options, to set a different
3513 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3514 was its contents. (It was OK if the option was not defined at all.)
3516 17. A "Completed" log line is now written for messages that are removed from
3517 the spool by the -Mrm option.
3519 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3520 information about exactly what failed.
3522 19. Added -dd to debug only the daemon process.
3524 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3525 handles timeouts, both on the server side and network timeouts. Renamed the
3526 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3528 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3529 It is now set to "smtps".
3531 22. $host_address is now set to the target address during the checking of
3532 ignore_target_hosts.
3534 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3535 being passed; this would have caused $sender_host_name to have been used if
3536 matching the list had actually called for a host name (not very likely,
3537 since this list is usually IP addresses). A host name is now passed as
3540 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3541 code that specifies a non-restarting handler (typically sigaction() in
3542 modern systems) in an attempt to fix a rare and obscure crash bug.
3544 25. Narrowed the window for a race in the daemon that could cause it to ignore
3545 SIGCHLD signals. This is not a major problem, because they are used only to
3546 wake it up if nothing else does.
3548 26. A malformed maildirsize file could cause Exim to calculate negative values
3549 for the mailbox size or file count. Odd effects could occur as a result.
3550 The maildirsize information is now recalculated if the size or filecount
3553 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3554 support for a long time. Removed HAVE_SYS_VFS_H.
3556 28. Installed the latest version of exipick from John Jetmore.
3558 29. In an address list, if the pattern was not a regular expression, an empty
3559 subject address (from a bounce message) matched only if the pattern was an
3560 empty string. Non-empty patterns were not even tested. This was the wrong
3561 because it is perfectly reasonable to use an empty address as part of a
3562 database query. An empty address is now tested by patterns that are
3563 lookups. However, all the other forms of pattern expect the subject to
3564 contain a local part and a domain, and therefore, for them, an empty
3565 address still always fails if the pattern is not itself empty.
3567 30. Exim went into a mad DNS loop when attempting to do a callout where the
3568 host was specified on an smtp transport, and looking it up yielded more
3569 than one IP address.
3571 31. Re-factored the code for checking spool and log partition space into a
3572 function that finds that data and another that does the check. The former
3573 is then used to implement four new variables: $spool_space, $log_space,
3574 $spool_inodes, and $log_inodes.
3576 32. The RFC2047 encoding function was originally intended for short strings
3577 such as real names; it was not keeping to the 75-character limit for
3578 encoded words that the RFC imposes. It now respects the limit, and
3579 generates multiple encoded words if necessary. To be on the safe side, I
3580 have increased the buffer size for the ${rfc2047: expansion operator from
3583 33. It is now permitted to omit both strings after an "if" condition; if the
3584 condition is true, the result is "true". As before, when the second string
3585 is omitted, a false condition yields an empty string. This makes it less
3586 cumbersome to write custom ACL and router conditions.
3588 34. Failure to deliver a bounce message always caused it to be frozen, even if
3589 there was an errors_to setting on the router. The errors_to setting is now
3592 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3593 canonical form (fully expanded) before being placed in
3594 $sender_host_address.
3596 36. The table in the code that translates DNS record types into text (T_A to
3597 "A" for instance) was missing entries for NS and CNAME. It is just possible
3598 that this could have caused confusion if both these types were looked up
3599 for the same domain, because the text type is used as part of Exim's
3600 per-process caching. But the chance of anyone hitting this buglet seems
3603 37. The dnsdb lookup has been extended in a number of ways.
3605 (1) There is a new type, "zns", which walks up the domain tree until it
3606 finds some nameserver records. It should be used with care.
3608 (2) There is a new type, "mxh", which is like "mx" except that it returns
3609 just the host names, not the priorities.
3611 (3) It is now possible to give a list of domains (or IP addresses) to be
3612 looked up. The behaviour when one of the lookups defers can be
3613 controlled by a keyword.
3615 (4) It is now possible to specify the separator character for use when
3616 multiple records are returned.
3618 38. The dnslists ACL condition has been extended: it is now possible to supply
3619 a list of IP addresses and/or domains to be looked up in a particular DNS
3622 39. Added log_selector=+queue_time_overall.
3624 40. When running the queue in the test harness, wait just a tad after forking a
3625 delivery process, to get repeatability of debugging output.
3627 41. Include certificate and key file names in error message when GnuTLS fails
3628 to set them up, because the GnuTLS error message doesn't include the name
3629 of the failing file when there is a problem reading it.
3631 42. Allow both -bf and -bF in the same test run.
3633 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3635 44. The "Exiscan patch" is now merged into the mainline Exim source.
3637 45. Sometimes the final signoff response after QUIT could fail to get
3638 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3639 < 0 before doing a fflush(). This bug looks as though it goes back to the
3640 introduction of TLS in release 3.20, but "sometimes" must have been rare
3641 because the tests only now provoked it.
3643 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3644 (this can affect the format of dates).
3646 47. exim_tidydb, when checking for the continued existence of a message for
3647 which it has found a message-specific retry record, was not finding
3648 messages that were in split spool directories. Consequently, it was
3649 deleting retry records that should have stayed in existence.
3651 48. Steve fixed some bugs in eximstats.
3653 49. The SPA authentication driver was not abandoning authentication and moving
3654 on to the next authenticator when an expansion was forced to fail,
3655 contradicting the general specification for all authenticators. Instead it
3656 was generating a temporary error. It now behaves as specified.
3658 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3659 (the order specifies the preference for clients). The order is now AES256,
3660 AES128, 3DES, ARCFOUR128.
3662 51. Small patch to Sieve code - explicitly set From: when generating an
3665 52. Exim crashed if a remote delivery caused a very long error message to be
3666 recorded - for instance if somebody sent an entire SpamAssassin report back
3667 as a large number of 550 error lines. This bug was coincidentally fixed by
3668 increasing the size of one of Exim's internal buffers (big_buffer) that
3669 happened as part of the Exiscan merge. However, to be on the safe side, I
3670 have made the code more robust (and fixed the comments that describe what
3673 53. Now that there can be additional text after "Completed" in log lines (if
3674 the queue_time_overall log selector is set), a one-byte patch to exigrep
3675 was needed to allow it to recognize "Completed" as not the last thing in
3678 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3679 patch that reportedly fixes this has been added. I am not expert enough to
3680 create a test for it. This is what the patch creator wrote:
3682 "I found a little strange behaviour of ldap code when working with
3683 Windows 2003 AD Domain, where users was placed in more than one
3684 Organization Units. When I tried to give exim partial DN, the exit code
3685 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3686 But simultaneously result of request was absolutely normal ldap result,
3687 so I produce this patch..."
3689 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3690 REFERENCE, so I have modified the code to exclude the patch when that macro
3693 55. Some experimental protocols are using DNS PTR records for new purposes. The
3694 keys for these records are domain names, not reversed IP addresses. The
3695 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3696 leaves it alone. Component reversal etc. now happens only for IP addresses.
3699 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3701 57. Double the size of the debug message buffer (to 2048) so that more of very
3702 long debug lines gets shown.
3704 58. The exicyclog utility now does better if the number of log files to keep
3705 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3707 59. Two changes related to the smtp_active_hostname option:
3709 (1) $smtp_active_hostname is now available as a variable.
3710 (2) The default for smtp_banner uses $smtp_active_hostname instead
3711 of $primary_hostname.
3713 60. The host_aton() function is supposed to be passed a string that is known
3714 to be a valid IP address. However, in the case of IPv6 addresses, it was
3715 not checking this. This is a hostage to fortune. Exim now panics and dies
3716 if the condition is not met. A case was found where this could be provoked
3717 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3718 components; fortuitously, this particular loophole had already been fixed
3719 by change 4.50/55 above.
3721 If there are any other similar loopholes, the new check in host_aton()
3722 itself should stop them being exploited. The report I received stated that
3723 data on the command line could provoke the exploit when Exim was running as
3724 exim, but did not say which command line option was involved. All I could
3725 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3726 running as the user.
3729 61. There was a buffer overflow vulnerability in the SPA authentication code
3730 (which came originally from the Samba project). I have added a test to the
3731 spa_base64_to_bits() function which I hope fixes it.
3734 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3735 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3737 63. The daemon start-up calls getloadavg() while still root for those OS that
3738 need the first call to be done as root, but it missed one case: when
3739 deliver_queue_load_max is set with deliver_drop_privilege. This is
3740 necessary for the benefit of the queue runner, because there is no re-exec
3741 when deliver_drop_privilege is set.
3743 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3744 This has been fixed.
3746 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3747 was in use, was not putting the data itself into the right store pool;
3748 consequently, it could be overwritten for a subsequent message in the same
3749 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3752 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3754 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3755 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3756 the function were treating the return as a boolean value, which happened to
3757 work because 0=false and not-0=true, but is not correct code.
3759 68. The host_aton() function was not handling scoped IPv6 addresses (those
3760 with, for example, "%eth0" on the end) correctly.
3762 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3763 negated items (that is, ~something) in unsigned ints. Some compilers
3764 apparently mutter when there is no cast.
3766 70. If an address verification called from an ACL failed, and did not produce a
3767 user-specific message (i.e. there was only a "system" message), nothing was
3768 put in $acl_verify_message. In this situation, it now puts the system
3771 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3772 session; change 4.31/43 added the unwanted input to the log line - except
3773 that it did not do this in the start of session case. It now does.
3775 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3776 This is wrong and can cause the other end to generate a synchronization
3777 error if it is another Exim or anything else that does the synchronization
3778 check. A QUIT command is no longer sent after a timeout.
3780 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3781 during host lookups.
3783 74. The defer_ok option of callout verification was not working if it was used
3784 when verifying addresses in header lines, that is, for this case:
3786 verify = header_sender/callout=defer_ok
3788 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3789 those file descriptors could be used for SMTP connections. If anything
3790 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3791 could be sent to the SMTP client, causing chaos. The daemon now opens
3792 stdin, stdout, and stderr to /dev/null when it puts itself into the
3795 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3796 log by default. The user can override this with suitable Perl magic.
3798 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3799 the log message when discard triggers, was not working for the DATA ACL or
3800 for the non-SMTP ACL.
3802 78. Error message wording change in sieve.c.
3804 79. If smtp_accept_max_per_host was set, the number of connections could be
3805 restricted to fewer than expected, because the daemon was trying to set up
3806 a new connection before checking whether the processes handling previous
3807 connections had finished. The check for completed processes is now done
3808 earlier. On busy systems, this bug wouldn't be noticed because something
3809 else would have woken the daemon, and it would have reaped the completed
3812 80. If a message was submitted locally by a user whose login name contained one
3813 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3814 It caused a spool format error. I have fixed the spool reading code. A
3815 related problem was that the "from" clause in the Received: line became
3816 illegal because of the space(s). It is now covered by ${quote_local_part.
3818 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3821 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3823 83. Patch from Sieve maintainer to fix unterminated string problem in
3824 "vacation" handling.
3826 84. Some minor changes to the Linux configuration files to help with other
3827 OS variants using glibc.
3829 85. One more patch for Sieve to update vacation handling to latest spec.
3832 ----------------------------------------------------
3833 See the note above about the 4.44 and 4.50 releases.
3834 ----------------------------------------------------
3840 1. Change 4.43/35 introduced a bug that caused file counts to be
3841 incorrectly computed when quota_filecount was set in an appendfile
3844 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3845 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3848 3. Give more explanation in the error message when the command for a transport
3849 filter fails to execute.
3851 4. There are several places where Exim runs a non-Exim command in a
3852 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3853 was being done only for the command run by the queryprogram router. It is
3854 now done for all such subprocesses. The other cases are: ${run, transport
3855 filters, and the commands run by the lmtp and pipe transports.
3857 5. Some older OS have a limit of 256 on the maximum number of file
3858 descriptors. Exim was using setrlimit() to set 1000 as a large value
3859 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3860 systems. I've change it so that if it can't get 1000, it tries for 256.
3862 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3863 was an oversight, and furthermore, ever since the addition of extra
3864 controls (e.g. 4.43/32), the checks on when to allow different forms of
3865 "control" were broken. There should now be diagnostics for all cases when a
3866 control that does not make sense is encountered.
3868 7. $recipients is now available in the predata ACL (oversight).
3870 8. Tidy the search cache before the fork to do a delivery from a message
3871 received from the command line. Otherwise the child will trigger a lookup
3872 failure and thereby defer the delivery if it tries to use (for example) a
3873 cached ldap connection that the parent has called unbind on.
3875 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3876 of $address_data from the recipient verification was clobbered by the
3877 sender verification.
3879 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3880 was its contents. (It was OK if the option was not defined at all.)
3882 11. A "Completed" log line is now written for messages that are removed from
3883 the spool by the -Mrm option.
3885 12. $host_address is now set to the target address during the checking of
3886 ignore_target_hosts.
3888 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3889 being passed; this would have caused $sender_host_name to have been used if
3890 matching the list had actually called for a host name (not very likely,
3891 since this list is usually IP addresses). A host name is now passed as
3894 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3895 code that specifies a non-restarting handler (typically sigaction() in
3896 modern systems) in an attempt to fix a rare and obscure crash bug.
3898 15. Narrowed the window for a race in the daemon that could cause it to ignore
3899 SIGCHLD signals. This is not a major problem, because they are used only to
3900 wake it up if nothing else does.
3902 16. A malformed maildirsize file could cause Exim to calculate negative values
3903 for the mailbox size or file count. Odd effects could occur as a result.
3904 The maildirsize information is now recalculated if the size or filecount
3907 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3908 support for a long time. Removed HAVE_SYS_VFS_H.
3910 18. Updated exipick to current release from John Jetmore.
3912 19. Allow an empty sender to be matched against a lookup in an address list.
3913 Previously the only cases considered were a regular expression, or an
3916 20. Exim went into a mad DNS lookup loop when doing a callout where the
3917 host was specified on the transport, if the DNS lookup yielded more than
3920 21. The RFC2047 encoding function was originally intended for short strings
3921 such as real names; it was not keeping to the 75-character limit for
3922 encoded words that the RFC imposes. It now respects the limit, and
3923 generates multiple encoded words if necessary. To be on the safe side, I
3924 have increased the buffer size for the ${rfc2047: expansion operator from
3927 22. Failure to deliver a bounce message always caused it to be frozen, even if
3928 there was an errors_to setting on the router. The errors_to setting is now
3931 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3932 canonical form (fully expanded) before being placed in
3933 $sender_host_address.
3935 24. Updated eximstats to version 1.33
3937 25. Include certificate and key file names in error message when GnuTLS fails
3938 to set them up, because the GnuTLS error message doesn't include the name
3939 of the failing file when there is a problem reading it.
3941 26. Expand error message when OpenSSL has problems setting up cert/key files.
3944 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3945 (this can affect the format of dates).
3947 28. exim_tidydb, when checking for the continued existence of a message for
3948 which it has found a message-specific retry record, was not finding
3949 messages that were in split spool directories. Consequently, it was
3950 deleting retry records that should have stayed in existence.
3952 29. eximstats updated to version 1.35
3953 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3954 1.35 - bugfix such that pie charts by volume are generated correctly
3956 30. The SPA authentication driver was not abandoning authentication and moving
3957 on to the next authenticator when an expansion was forced to fail,
3958 contradicting the general specification for all authenticators. Instead it
3959 was generating a temporary error. It now behaves as specified.
3961 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3962 (the order specifies the preference for clients). The order is now AES256,
3963 AES128, 3DES, ARCFOUR128.
3965 31. Small patch to Sieve code - explicitly set From: when generating an
3968 32. Exim crashed if a remote delivery caused a very long error message to be
3969 recorded - for instance if somebody sent an entire SpamAssassin report back
3970 as a large number of 550 error lines. This bug was coincidentally fixed by
3971 increasing the size of one of Exim's internal buffers (big_buffer) that
3972 happened as part of the Exiscan merge. However, to be on the safe side, I
3973 have made the code more robust (and fixed the comments that describe what
3976 33. Some experimental protocols are using DNS PTR records for new purposes. The
3977 keys for these records are domain names, not reversed IP addresses. The
3978 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3979 leaves it alone. Component reversal etc. now happens only for IP addresses.
3982 34. The host_aton() function is supposed to be passed a string that is known
3983 to be a valid IP address. However, in the case of IPv6 addresses, it was
3984 not checking this. This is a hostage to fortune. Exim now panics and dies
3985 if the condition is not met. A case was found where this could be provoked
3986 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3987 components; fortuitously, this particular loophole had already been fixed
3988 by change 4.50/55 or 4.44/33 above.
3990 If there are any other similar loopholes, the new check in host_aton()
3991 itself should stop them being exploited. The report I received stated that
3992 data on the command line could provoke the exploit when Exim was running as
3993 exim, but did not say which command line option was involved. All I could
3994 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3995 running as the user.
3998 35. There was a buffer overflow vulnerability in the SPA authentication code
3999 (which came originally from the Samba project). I have added a test to the
4000 spa_base64_to_bits() function which I hope fixes it.
4003 36. The daemon start-up calls getloadavg() while still root for those OS that
4004 need the first call to be done as root, but it missed one case: when
4005 deliver_queue_load_max is set with deliver_drop_privilege. This is
4006 necessary for the benefit of the queue runner, because there is no re-exec
4007 when deliver_drop_privilege is set.
4009 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4010 was in use, was not putting the data itself into the right store pool;
4011 consequently, it could be overwritten for a subsequent message in the same
4012 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4015 38. Sometimes the final signoff response after QUIT could fail to get
4016 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4017 < 0 before doing a fflush(). This bug looks as though it goes back to the
4018 introduction of TLS in release 3.20, but "sometimes" must have been rare
4019 because the tests only now provoked it.
4025 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4026 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4027 Now it returns an integer. A number of calls were still expecting a T/F
4028 return. Fortuitously, in all cases, the tests worked in OK situations,
4029 which is the norm. However, things would have gone wrong on any write
4030 failures on the smtp file descriptor. This function is used when sending
4031 messages over SMTP and also when doing verify callouts.
4033 2. When Exim is called to do synchronous delivery of a locally submitted
4034 message (the -odf or -odi options), it no longer closes stderr before doing
4037 3. Implemented the mua_wrapper option.
4039 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4041 5. Implemented the functions header_remove(), header_testname(),
4042 header_add_at_position(), and receive_remove_recipient(), and exported them
4045 6. If an ACL "warn" statement specified the addition of headers, Exim already
4046 inserted X-ACL-Warn: at the start if there was no header name. However, it
4047 was not making this test for the second and subsequent header lines if
4048 there were newlines in the string. This meant that an invalid header could
4049 be inserted if Exim was badly configured.
4051 7. Allow an ACL "warn" statement to add header lines at the start or after all
4052 the Received: headers, as well as at the end.
4054 8. Added the rcpt_4xx retry error code.
4056 9. Added postmaster_mailfrom=xxx to callout verification option.
4058 10. Added mailfrom=xxxx to the callout verification option, for verify=
4061 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4062 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4064 12. Inserted some casts to stop certain compilers warning when using pointer
4065 differences as field lengths or precisions in printf-type calls (mostly
4066 affecting debugging statements).
4068 13. Added optional readline() support for -be (dynamically loaded).
4070 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4071 same clock tick as a message's arrival, so that its received time was the
4072 same as the "first fail" time on the retry record, and that message
4073 remained on the queue past the ultimate address timeout, every queue runner
4074 would try a delivery (because it was past the ultimate address timeout) but
4075 after another failure, the ultimate address timeout, which should have then
4076 bounced the address, did not kick in. This was a "< instead of <=" error;
4077 in most cases the first failure would have been in the next clock tick
4078 after the received time, and all would be well.
4080 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4081 being recognized when the domain list was tested by the match_domain
4082 condition in an expansion string.
4084 16. Added the ${str2b64: operator.
4086 17. Exim was always calling setrlimit() to set a large limit for the number of
4087 processes, without checking whether the existing limit was already
4088 adequate. (It did check for the limit on file descriptors.) Furthermore,
4089 errors from getrlimit() and setrlimit() were being ignored. Now they are
4090 logged to the main and panic logs, but Exim does carry on, to try to do its
4091 job under whatever limits there are.
4093 18. Imported PCRE 5.0.
4095 19. Trivial typo in log message " temporarily refused connection" (the leading
4098 20. If the log selector return_path_on_delivery was set and an address was
4099 redirected to /dev/null, the delivery process crashed because it assumed
4100 that a return path would always be set for a "successful" delivery. In this
4101 case, the whole delivery is bypassed as an optimization, and therefore no
4104 21. Internal re-arrangement: the function for sending a challenge and reading
4105 a response while authentication was assuming a zero-terminated challenge
4106 string. It's now changed to take a pointer and a length, to allow for
4107 binary data in such strings.
4109 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4111 23. Exim was not respecting finduser_retries when seeking the login of the
4112 uid under which it was called; it was always trying 10 times. (The default
4113 setting of finduser_retries is zero.) Also, it was sleeping after the final
4114 failure, which is pointless.
4116 24. Implemented tls_on_connect_ports.
4118 25. Implemented acl_smtp_predata.
4120 26. If the domain in control=submission is set empty, Exim assumes that the
4121 authenticated id is a complete email address when it generates From: or
4122 Sender: header lines.
4124 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4125 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4126 chown and chgrp in /bin and hostname in /usr/bin.
4128 28. Exim was keeping the "process log" file open after each use, just as it
4129 does for the main log. This opens the possibility of it remaining open for
4130 long periods when the USR1 signal hits a daemon. Occasional processlog
4131 errors were reported, that could have been caused by this. Anyway, it seems
4132 much more sensible not to leave this file open at all, so that is what now
4135 29. The long-running daemon process does not normally write to the log once it
4136 has entered its main loop, and it closes the log before doing so. This is
4137 so that log files can straightforwardly be renamed and moved. However,
4138 there are a couple of unusual error situations where the daemon does write
4139 log entries, and I had neglected to close the log afterwards.
4141 30. The text of an SMTP error response that was received during a remote
4142 delivery was being truncated at 512 bytes. This is too short for some of
4143 the long messages that one sometimes sees. I've increased the limit to
4146 31. It is now possible to make retry rules that apply only when a message has a
4147 specific sender, in particular, an empty sender.
4149 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4150 it possible to be selective about when SMTP synchronization is enforced.
4152 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4154 32. Implemented hosts_connection_nolog.
4156 33. Added an ACL for QUIT.
4158 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4161 35. Added mailbox_size and mailbox_filecount to appendfile.
4163 36. Added control = no_multiline_responses to ACLs.
4165 37. There was a bug in the logic of the code that waits for the clock to tick
4166 in the case where the clock went backwards by a substantial amount such
4167 that the microsecond fraction of "now" was more than the microsecond
4168 fraction of "then" (but the whole seconds number was less).
4170 38. Added support for the libradius Radius client library this is found on
4171 FreeBSD (previously only the radiusclient library was supported).
4177 1. When certain lookups returned multiple values in the form name=value, the
4178 quoting of the values was not always being done properly. Specifically:
4179 (a) If the value started with a double quote, but contained no whitespace,
4181 (b) If the value contained whitespace other than a space character (i.e.
4182 tabs or newlines or carriage returns) it was not quoted.
4183 This fix has been applied to the mysql and pgsql lookups by writing a
4184 separate quoting function and calling it from the lookup code. The fix
4185 should probably also be applied to nisplus, ibase and oracle lookups, but
4186 since I cannot test any of those, I have not disturbed their existing code.
4188 2. A hit in the callout cache for a specific address caused a log line with no
4189 reason for rejecting RCPT. Now it says "Previous (cached) callout
4190 verification failure".
4192 3. There was an off-by-one bug in the queryprogram router. An over-long
4193 return line was truncated at 256 instead of 255 characters, thereby
4194 overflowing its buffer with the terminating zero. As well as fixing this, I
4195 have increased the buffer size to 1024 (and made a note to document this).
4197 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4198 when Exim is waiting for an SMTP response from a remote server, Exim
4199 restarts its select() call on the socket, thereby resetting its timeout.
4200 This is not a problem when such interrupts are rare. Somebody set up a cron
4201 job to run exiwhat every 2 minutes, which is less than the normal select()
4202 timeout (5 or 10 minutes). This meant that the select() timeout never
4203 kicked in because it was always reset. I have fixed this by comparing the
4204 time when an interrupt arrives with the time at the start of the first call
4205 to select(). If more time than the timeout has elapsed, the interrupt is
4206 treated as a timeout.
4208 5. Some internal re-factoring in preparation for the addition of Sieve
4209 extensions (by MH). In particular, the "personal" test is moved to a
4210 separate function, and given an option for scanning Cc: and Bcc: (which is
4211 not set for Exim filters).
4213 6. When Exim created an email address using the login of the caller as the
4214 local part (e.g. when creating a From: or Sender: header line), it was not
4215 quoting the local part when it contained special characters such as @.
4217 7. Installed new OpenBSD configuration files.
4219 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4220 try to make them clearer.
4222 9. Callout options, other than the timeout value, were being ignored when
4223 verifying sender addresses in header lines. For example, when using
4225 verify = header_sender/callout=no_cache
4227 the cache was (incorrectly) being used.
4229 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4230 only the Cygwin environment.
4232 11. When return_path_on_delivery was set as a log selector, if different remote
4233 addresses in the same message used different return paths and parallel
4234 remote delivery occurred, the wrong values would sometimes be logged.
4235 (Whenever a remote delivery process finished, the return path value from
4236 the most recently started remote delivery process was logged.)
4238 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4239 lines when AUTH and/or TLS are in use. This is the "received protocol"
4240 field. Exim used to use "asmtp" for authenticated SMTP, without any
4241 indication (in the protocol name) for TLS use. Now it follows the RFC and
4242 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4243 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4244 names appear in log lines as well as in Received: header lines.
4246 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4247 extensions, and comparison tests, and to fix some bugs.
4249 14. Changes to the "personal" filter test:
4251 (1) The test was buggy in that it was just doing the equivalent of
4252 "contains" tests on header lines. For example, if a user's address was
4253 anne@some.where, the "personal" test would incorrectly be true for
4255 To: susanne@some.where
4257 This test is now done by extracting each address from the header in turn,
4258 and checking the entire address. Other tests that are part of "personal"
4259 are now done using regular expressions (for example, to check local parts
4260 of addresses in From: header lines).
4262 (2) The list of non-personal local parts in From: addresses has been
4263 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4264 taken from the Sieve specification recommendations.
4266 (3) If the message contains any header line starting with "List-" it is
4267 treated as non-personal.
4269 (4) The test for "circular" in the Subject: header line has been removed
4270 because it now seems ill-conceived.
4272 15. Minor typos in src/EDITME comments corrected.
4274 16. Installed latest exipick from John Jetmore.
4276 17. If headers_add on a router specified a text string that was too long for
4277 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4278 of string_sprintf() is now avoided.
4280 18. $message_body_size was not set (it was always zero) when running the DATA
4281 ACL and the local_scan() function.
4283 19. For the "mail" command in an Exim filter, no default was being set for
4284 the once_repeat time, causing a random time value to be used if "once" was
4285 specified. (If the value happened to be <= 0, no repeat happened.) The
4286 default is now 0s, meaning "never repeat". The "vacation" command was OK
4287 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4288 (I found it when inspecting the code).
4290 20. There is now an overall timeout for performing a callout verification. It
4291 defaults to 4 times the callout timeout, which applies to individual SMTP
4292 commands during the callout. The overall timeout applies when there is more
4293 than one host that can be tried. The timeout is checked before trying the
4294 next host. This prevents very long delays if there are a large number of
4295 hosts and all are timing out (e.g. when the network connections are timing
4296 out). The value of the overall timeout can be changed by specifying an
4297 additional sub-option for "callout", called "maxwait". For example:
4299 verify = sender/callout=5s,maxwait=20s
4301 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4302 to the end before writing, but this should make it even safer).
4304 22. Exim was forgetting that it had advertised PIPELINING for the second and
4305 subsequent messages on an SMTP connection. It was also not resetting its
4306 memory on STARTTLS and an internal HELO.
4308 23. When Exim logs an SMTP synchronization error within a session, it now
4309 records whether PIPELINING has been advertised or not.
4311 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4312 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4313 rather than long int.
4315 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4317 26. Added the never_mail option to autoreply.
4323 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4324 crash if the getsockname() call failed; this can happen if a connection is
4325 closed very soon after it is established. The problem was simply in the
4326 order in which certain operations were done, causing Exim to try to write
4327 to the SMTP stream before it had set up the file descriptor. The bug has
4328 been fixed by making things happen in the correct order.
4334 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4335 before the connection was closed, thus losing the rejection response.
4337 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4338 some early Solaris releases, but causes trouble in current releases where
4339 socklen_t is defined.
4341 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4344 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4347 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4348 that releasing the top of it at the end releases what was used for sub-
4349 expansions (unless the block got too big). However, discard this block if
4350 the first thing is a variable or header, so that we can use its block when
4351 it is dynamic (useful for very large $message_headers, for example).
4353 6. Lookups now cache *every* query, not just the most recent. A new, separate
4354 store pool is used for this. It can be recovered when all lookup caches are
4355 flushed. Lookups now release memory at the end of their result strings.
4356 This has involved some general refactoring of the lookup sources.
4358 7. Some code has been added to the store_xxx() functions to reduce the amount
4359 of flapping under certain conditions.
4361 8. log_incoming_interface used to affect only the <= reception log lines. Now
4362 it causes the local interface and port to be added to several more SMTP log
4363 lines, for example "SMTP connection from", and rejection lines.
4365 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4367 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4369 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4370 could be overwritten at the end of the current message (or the start of a
4371 new message if it was set in a HELO ACL). The value is now preserved for
4372 the duration of the SMTP connection.
4374 12. If a transport had a headers_rewrite setting, and a matching header line
4375 contained an unqualified address, that address was qualified, even if it
4376 did not match any rewriting rules. The underlying bug was that the values
4377 of the flags that permit the existence of unqualified sender and recipient
4378 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4379 non-local messages, and by -bnq for local messages) were not being
4380 preserved with the message after it was received.
4382 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4383 "next input=" as part of the text comprising the host identity instead of
4384 the correct text. The code was using the same buffer for two different
4385 strings. However, depending on which order the printing function evaluated
4386 its arguments, the bug did not always show up. Under Linux, for example, my
4387 test suite worked just fine.
4389 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4390 doesn't work with some older versions of Perl. It has been changed to "my",
4391 which in any case is probably the better facility to use.
4393 15. A really picky compiler found some instances of statements for creating
4394 error messages that either had too many or two few arguments for the format
4397 16. The size of the buffer for calls to the DNS resolver has been increased
4398 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4399 for addresses that have a lot of PTR records. This alleviates a problem; it
4400 does not fully solve it.
4402 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4403 buffer now truncates the list and logs the incident, which is the same
4404 action as happens when Exim is looking up a host name and its aliases.
4405 Previously in this situation something unpredictable would happen;
4406 sometimes it was "internal error: store_reset failed".
4408 18. If a server dropped the connection unexpectedly when an Exim client was
4409 using GnuTLS and trying to read a response, the client delivery process
4410 crashed while trying to generate an error log message.
4412 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4413 string, for example:
4415 warn message = H1: something\nH2: something
4417 the text was added as a single header line from Exim's point of view
4418 though it ended up OK in the delivered message. However, searching for the
4419 second and subsequent header lines using $h_h2: did not work. This has been
4420 fixed. Similarly, if a system filter added multiple headers in this way,
4421 the routers could not see them.
4423 20. Expanded the error message when iplsearch is called with an invalid key to
4424 suggest using net-iplsearch in a host list.
4426 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4427 ACLs are no longer actually imposed (and a message to that effect is
4430 22. If a "gecos" field in a passwd entry contained escaped characters, in
4431 particular, if it contained a \" sequence, Exim got it wrong when building
4432 a From: or a Sender: header from that name. A second bug also caused
4433 incorrect handling when an unquoted " was present following a character
4434 that needed quoting.
4436 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4437 was not being matched caselessly.
4439 24. Arranged for all hyphens in the exim.8 source to be escaped with
4442 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4443 properly. Recipient callout cache records were still being keyed to include
4444 the sender, even when use_sender was set false. This led to far more
4445 callouts that were necessary. The sender is no longer included in the key
4446 when use_sender is false.
4448 26. Added "control = submission" modifier to ACLs.
4450 27. Added the ${base62d: operator to decode base 62 numbers.
4452 28. dnsdb lookups can now access SRV records.
4454 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4455 the configuration file.
4457 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4458 (-v) mode. This makes the output for a verbose queue run more intelligible.
4460 31. Added a use_postmaster feature to recipient callouts.
4462 32. Added the $body_zerocount variable, containing the number of binary zero
4463 bytes in the message body.
4465 33. The time of last modification of the "new" subdirectory is now used as the
4466 "mailbox time last read" when there is a quota error for a maildir
4469 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4471 35. Added +ignore_unknown as a special item in host lists.
4473 36. Code for decoding IPv6 addresses in host lists is now included, even if
4474 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4475 address was recognized as an IP address, but was then not correctly decoded
4476 into binary, causing unexpected and incorrect effects when compared with
4483 1. Very minor rewording of debugging text in manualroute to say "list of
4484 hosts" instead of "hostlist".
4486 2. If verify=header_syntax was set, and a header line with an unqualified
4487 address (no domain) and a large number of spaces between the end of the
4488 name and the colon was received, the reception process suffered a buffer
4489 overflow, and (when I tested it) crashed. This was caused by some obsolete
4490 code that should have been removed. The fix is to remove it!
4492 3. When running in the test harness, delay a bit after writing a bounce
4493 message to get a bit more predictability in the log output.
4495 4. Added a call to search_tidyup() just before forking a reception process. In
4496 theory, someone could use a lookup in the expansion of smtp_accept_max_
4497 per_host which, without the tidyup, could leave open a database connection.
4499 5. Added the variables $recipient_data and $sender_data which get set from a
4500 lookup success in an ACL "recipients" or "senders" condition, or a router
4501 "senders" option, similar to $domain_data and $local_part_data.
4503 6. Moved the writing of debug_print from before to after the "senders" test
4506 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4507 problems for message scanning, either using a data ACL, or using
4508 local_scan() because the Received: header was not generated till after they
4509 were called (in order to set the time as the time of reception completion).
4510 I have revised the way this works. The header is now generated after the
4511 body is received, but before the ACL or local_scan() are called. After they
4512 are run, the timestamp in the header is updated.
4518 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4519 before starting a queue runner without re-exec. This happened only when
4520 deliver_drop_privilege was set or when the Exim user was set to root. The
4521 effect of the bug was that timeouts during subsequent deliveries caused
4522 crashes instead of being properly handled. The handler is now left at its
4523 default (and expected) setting.
4525 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4526 message, again when deliver_drop_privilege is set or Exim is run as root.
4527 The bug described in (1) was not present in this case, but the tidying up
4528 of the other signals was missing. I have made the two cases consistent.
4530 3. The ignore_target_hosts setting on a manualroute router was being ignored
4531 for hosts that were looked up using the /MX notation.
4533 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4536 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4537 operated on the sender address. After changing the $sender_address to <>
4538 for the sender address verify, Exim was re-instated it as the original
4539 (before rewriting) address, but remembering that it had rewritten it, so it
4540 wasn't rewriting it again. This bug also had the effect of breaking the
4541 sender address verification caching when the sender address was rewritten.
4543 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4544 This has been changed so that if the ip literal address matches
4545 ignore_target_hosts, the router declines.
4547 7. Added expansion conditions match_domain, match_address, and match_local_
4548 part (NOT match_host).
4550 8. The placeholder for the Received: header didn't have a length field set.
4552 9. Added code to Exim itself and to exim_lock to test for a specific race
4553 condition that could lead to file corruption when using MBX delivery. The
4554 issue is with the lockfile that is created in /tmp. If this file is removed
4555 after a process has opened it but before that process has acquired a lock,
4556 there is the potential for a second process to recreate the file and also
4557 acquire a lock. This could lead to two Exim processes writing to the file
4558 at the same time. The added code performs the same test as UW imapd; it
4559 checks after acquiring the lock that its file descriptor still refers to
4560 the same named file.
4562 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4563 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4566 11. Added the smtp_active_hostname option. If used, this will typically be made
4567 to depend on the incoming interface address. Because $interface_address is
4568 not set up until the daemon has forked a reception process, error responses
4569 that can happen earlier (such as "too many connections") no longer contain
4572 12. If an expansion in a condition on a "warn" statement fails because a lookup
4573 defers, the "warn" statement is abandoned, and the next ACL statement is
4574 processed. Previously this caused the whole ACL to be aborted.
4576 13. Added the iplsearch lookup type.
4578 14. Added ident_timeout as a log selector.
4580 15. Added tls_certificate_verified as a log selector.
4582 16. Added a global option tls_require_ciphers (compare the smtp transport
4583 option of the same name). This controls incoming TLS connections.
4585 17. I finally figured out how to make tls_require_ciphers do a similar thing
4586 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4587 before starting the TLS session.
4589 18. Tabs are now shown as \t in -bP output.
4591 19. If the log selector return_path_on_delivery was set, Exim crashed when
4592 bouncing a message because it had too many Received: header lines.
4594 20. If two routers both had headers_remove settings, and the first one included
4595 a superfluous trailing colon, the final name in the first list and the
4596 first name in the second list were incorrectly joined into one item (with a
4597 colon in the middle).
4603 1. Added -C and -D options to the exinext utility, mainly to make it easier
4604 to include in the automated testing, but these could be helpful when
4605 multiple configurations are in use.
4607 2. The exinext utility was not formatting the output nicely when there was
4608 an alternate port involved in the retry record key, nor when there was a
4609 message id as well (for retries that were specific to a specific message
4610 and a specific host). It was also confused by IPv6 addresses, because of
4611 the additional colons they contain. I have fixed the IPv4 problem, and
4612 patched it up to do a reasonable job for IPv6.
4614 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4615 delivery, the log line now contains "pipelined" if PIPELINING was used.
4617 4. An SMTP transport process used to panic and die if the bind() call to set
4618 an explicit outgoing interface failed. This has been changed; it is now
4619 treated in the same way as a connect() failure.
4621 5. A reference to $sender_host_name in the part of a conditional expansion
4622 that was being skipped was still causing a DNS lookup. This no longer
4625 6. The def: expansion condition was not recognizing references to header lines
4626 that used bh_ and bheader_.
4628 7. Added the _cache feature to named lists.
4630 8. The code for checking quota_filecount in the appendfile transport was
4631 allowing one more file than it should have been.
4633 9. For compatibility with Sendmail, the command line option
4641 and sets the incoming protocol and host name (for trusted callers). The
4642 host name and its colon can be omitted when only the protocol is to be set.
4643 Note the Exim already has two private options, -pd and -ps, that refer to
4644 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4645 "s", but I don't think that's a major issue.
4647 10. A number of refactoring changes to the code, none of which should affect
4650 (a) The number of logging options was getting close to filling up the
4651 32-bit word that was used as a bit map. I have split them into two classes:
4652 those that are passed in the argument to log_write(), and those that are
4653 only ever tested independently outside of that function. These are now in
4654 separate 32-bit words, so there is plenty of room for expansion again.
4655 There is no change in the user interface or the logging behaviour.
4657 (b) When building, for example, log lines, the code previously used a
4658 macro that called string_cat() twice, in order to add two strings. This is
4659 not really sufficiently general. Furthermore, there was one instance where
4660 it was actually wrong because one of the argument was used twice, and in
4661 one call a function was used. (As it happened, calling the function twice
4662 did not affect the overall behaviour.) The macro has been replaced by a
4663 function that can join an arbitrary number of extra strings onto a growing
4666 (c) The code for expansion conditions now uses a table and a binary chop
4667 instead of a serial search (which was left over from when there were very
4668 few conditions). Also, it now recognizes conditions like "pam" even when
4669 the relevant support is not compiled in: a suitably worded error message is
4670 given if an attempt is made to use such a condition.
4672 11. Added ${time_interval:xxxxx}.
4674 12. A bug was causing one of the ddress fields not to be passed back correctly
4675 from remote delivery subprocesses. The field in question was not being
4676 subsequently used, so this caused to problems in practice.
4678 13. Added new log selectors queue_time and deliver_time.
4680 14. Might have fixed a bug in maildirsizefile handling that threw up
4681 "unexpected character" debug warnings, and recalculated the data
4682 unnecessarily. In any case, I expanded the warning message to give more
4685 15. Added the message "Restricted characters in address" to the statements in
4686 the default ACL that block characters like @ and % in local parts.
4688 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4689 Three changes have been made:
4691 (a) There was a serious bug; a negative response to MAIL caused the whole
4692 recipient domain to be cached as invalid, thereby blocking all messages
4693 to all local parts at the same domain, from all senders. This bug has
4694 been fixed. The domain is no longer cached after a negative response to
4695 MAIL if the sender used is not empty.
4697 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4700 (c) A new callout option, "use_sender" has been added for people who want
4701 the modified behaviour.
4707 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4710 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4711 indeed breaks things for older releases.
4713 3. Added additional logging to the case where there is a problem reading data
4714 from a filter that is running in a subprocess using a pipe, in order to
4715 try to track down a specific problem.
4717 4. Testing facility fudge: when running in the test harness and attempting
4718 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4719 getting "No route to host". Convert this to a timeout.
4721 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4724 6. Some OS don't have socklen_t but use size_t instead. This affects the
4725 fifth argument of getsockopt() amongst other things. This is now
4726 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4727 can be set for individual OS. I have set it for SunOS5, OSF1, and
4728 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4729 some earlier ones do not.
4731 7. Change 4.30/15 was not doing the test caselessly.
4733 8. The standard form for an IPv6 address literal was being rejected by address
4734 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4735 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4736 this, as well as the form without the "IPv6" on the front (but only when
4737 address literals are enabled, of course).
4739 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4741 10. Exim crashed if a message with an empty sender address specified by -f
4742 encountered a router with an errors_to setting. This could be provoked only
4743 by a command such as
4747 where an empty string was supplied; "<>" did not hit this bug.
4749 11. Installed PCRE release 4.5.
4751 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4752 remained set. It is now erased.
4754 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4755 times from message ids (which are base 36 rather than the normal 62).
4757 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4758 were being counted as actual protocol errors, and logged if the log
4759 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4760 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4761 and DATA following a set of rejected RCPTs do not count as protocol errors.
4762 In other words, Exim assumes they were pipelined, though this may not
4763 actually be the case. Of course, in all cases the client gets an
4764 appropriate error code.
4766 15. If a lookup fails in an ACL condition, a message about the failure may
4767 be available; it is used if testing the ACL cannot continue, because most
4768 such messages specify what the cause of the deferral is. However, some
4769 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4770 that caused an old message to be retained and used if a later statement
4771 caused a defer, replacing the real cause of the deferral.
4773 16. If an IP address had so many PTR records that the DNS lookup buffer
4774 was not large enough to hold them, Exim could crash while trying to process
4775 the truncated data. It now detects and logs this case.
4777 17. Further to 4.21/58, another change has been made: if (and only if) the
4778 first line of a message (the first header line) ends with CRLF, a bare LF
4779 in a subsequent header line has a space inserted after it, so as not to
4780 terminate the header.
4782 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4783 unnecessarily, used atoi() instead of strtol(), and didn't check the
4784 termination when getting file sizes from file names by regex.
4786 19. Completely re-implemented the support for maildirsize files, in the light
4787 of a number of problems with the previous contributed implementation
4788 (4.30/29). In particular:
4790 . If the quota is zero, the maildirsize file is maintained, but no quota is
4793 . If the maildir directory does not exist, it is created before any attempt
4794 to write a maildirsize file.
4796 . The quota value in the file is just a cache; if the quota is changed in
4797 the transport, the new value overrides.
4799 . A regular expression is available for excluding directories from the
4802 20. The autoreply transport checks the characters in options that define the
4803 message's headers; it allows continued headers, but it was checking with
4804 isspace() after an embedded newline instead of explicitly looking for a
4807 21. If all the "regular" hosts to which an address was routed had passed their
4808 expiry times, and had not reached their retry times, the address was
4809 bounced, even if fallback hosts were defined. Now Exim should go on to try
4812 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4813 equivalent code in the SMTP transport. Some hosts send humungous responses
4814 to HELO/EHLO, more than 1024 it seems.
4816 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4817 gives compiler warnings in some environments. I've now done it "properly",
4820 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4821 (because of problems with the built-in one) was declared to return uschar *
4822 instead of char *, causing compiler failure.
4824 25. Fixed a file descriptor leak when processing alias/forward files.
4826 26. Fixed a minor format string issue in dbfn.c.
4828 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4830 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4831 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4834 29. When a sender address is verified, it is cached, to save repeating the test
4835 when there is more than one recipient in a message. However, when the
4836 verification involves a callout, it is possible for different callout
4837 options to be set for different recipients. It is too complicated to keep
4838 track of this in the cache, so now Exim always runs a verification when a
4839 callout is required, relying on the callout cache for the optimization.
4840 The overhead is duplication of the address routing, but this should not be
4843 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4844 to be verified with callout=postmaster, and the main callout worked but the
4845 postmaster check failed, the verification correctly failed. However, if a
4846 subsequent RCPT command asked for sender verification *without* the
4847 postmaster check, incorrect caching caused this verification also to fail,
4850 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4851 it was not caching the DNS options (qualify_single, search_parents) that
4852 were used when the lookup failed. A subsequent lookup with different
4853 options therefore always gave the same answer, though there were cases
4854 where it should not have. (Example: a "domains = !$mx_any" option on a
4855 dnslookup router: the "domains" option is always processed without any
4856 widening, but the router might have qualify_single set.) Now Exim uses the
4857 cached value only when the same options are set.
4859 32. Added John Jetmore's "exipick" utility to the distribution.
4861 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4862 than a timeout (e.g. a certificate is required, and is not provided), an
4863 Exim server now closes the connection immediately. Previously it waited for
4864 the client to close - but if the client is SSL, it seems that they each
4865 wait for each other, leading to a delay before one of them times out.
4867 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4868 maintained 0.8.x compatibility because I don't think many are using it, and
4869 it is clearly obsolete.
4871 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4874 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4875 client certificate was expired. A simple patch fixes this, though I don't
4876 understand the full logic of why the verify callback is called multiple
4879 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4880 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4881 which causes problems with some clients (such as the Certicom SSL Plus
4882 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4883 disables the coutermeasure allowing Eudora to connect."
4885 38. Exim was not checking that a write() to a log file succeeded. This could
4886 lead to Bad Things if a log got too big, in particular if it hit a file
4887 size limit. Exim now panics and dies if it cannot write to a log file, just
4888 as it does if it cannot open a log file.
4890 39. Modified OS/Makefile-Linux so that it now contains
4892 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4894 The two -D definitions ensure that Exim is compiled with large file
4895 support, which makes it possible to handle log files that are bigger than
4898 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4899 instance) a domain was checked against a named list that involved a lookup,
4900 causing $domain_data to be set, then another domain was checked against the
4901 same list, then the first domain was re-checked, the value of $domain_data
4902 after the final check could be wrong. In particular, if the second check
4903 failed, it could be set empty. This bug probably also applied to
4906 41. The strip_trailing_dot option was not being applied to the address given
4907 with the -f command-line option.
4909 42. The code for reading a message's header from the spool was incrementing
4910 $received_count, but never initializing it. This meant that the value was
4911 incorrect (doubled) while delivering a message in the same process in which
4912 it was received. In the most common configuration of Exim, this never
4913 happens - a fresh exec is done - but it can happen when
4914 deliver_drop_privilege is set.
4916 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4917 it now includes up to 150 characters of the unexpected data in the log
4920 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4921 and building data strings. The size of both of these buffers was 10 000
4922 bytes - far larger than anybody would *ever* want, thought I. Needless to
4923 say, somebody hit the limit. I have increased the maximum line length to
4924 20 000 and the maximum data length of concatenated lines to 100 000. I have
4925 also fixed two bugs, because there was no checking on these buffers. Tsk,
4926 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4927 buffer is too small.
4929 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4930 lsearch lookups. Now it does.
4932 46. When parsing a route_list item in a manualroute router, a fixed-length
4933 buffer was used for the list of hosts. I made this 1024 bytes long,
4934 thinking that nobody would ever have a list of hosts that long. Wrong.
4935 Somebody had a whole pile of complicated expansion conditions, and the
4936 string was silently truncated, leading to an expansion error. It turns out
4937 that it is easier to change to an unlimited length (owing to other changes
4938 that have happened since this code was originally written) than to build
4939 structure for giving a limitation error. The length of the item that
4940 expands into the list of hosts is now unlimited.
4942 47. The lsearch lookup could not handle data where the length of text line was
4943 more than 4095 characters. Such lines were truncated, leading to shortened
4944 data being returned. It should now handle lines of any length.
4946 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4947 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4950 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4951 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4952 no longer happens in comment lines. A list of replacements is now placed
4953 at the head of all of the source files, except those whose only change is
4954 to replace PERL_COMMAND in the very first #! line.
4956 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4957 messages on the queue, with a bottom-up merge sort, using code contributed
4958 by Michael Haardt. This should make operations like -bp somewhat faster on
4959 large queues. It won't affect queue runners, except when queue_run_in_order
4962 51. Installed eximstats 1.31 in the distribution.
4964 52. Added support for SRV lookups to the dnslookup router.
4966 53. If an ACL referred to $message_body or $message_body_end, the value was not
4967 reset for any messages that followed in the same SMTP session.
4969 54. The store-handling optimization for building very long strings was not
4970 differentiating between the different store pools. I don't think this
4971 actually made any difference in practice, but I've tidied it.
4973 55. While running the routers to verify a sender address, $sender_address
4974 was still set to the sender address. This is wrong, because when routing to
4975 send a bounce to the sender, it would be empty. Therefore, I have changed
4976 it so that, while verifying a sender address, $sender_address is set to <>.
4977 (There is no change to what happens when verifying a recipient address.)
4979 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4980 target A or AAAA records (if not already returned) without resetting the
4981 qualify_single or search_parents options of the DNS resolver. These are
4982 inappropriate in this case because the targets of MX and SRV records must
4983 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4984 target that, when qualified, matched something in the local domain. These
4985 two options are now turned off when doing these lookups.
4987 57. It seems that at least some releases of Reiserfs (which does not have the
4988 concept of a fixed number of inodes) returns zero and not -1 for the
4989 number of available inodes. This interacted badly with check_spool_inodes,
4990 which assumed that -1 was the "no such thing" setting. What I have done is
4991 to check that the total number of inodes is greater than zero before doing
4992 the test of how many are available.
4994 58. When a "warn" ACL statement has a log_message modifier, the message is
4995 remembered, and not repeated. This is to avoid a lot of repetition when a
4996 message has many recipients that cause the same warning to be written.
4997 Howewer, Exim was preserving the list of already written lines for an
4998 entire SMTP session, which doesn't seem right. The memory is now reset if a
4999 new message is started.
5001 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5002 debugging output unless log_rewrite was also set.
5004 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5005 of (int)(handle) when we know that handle contains (void *)(-1).
5007 61. The Exim daemon panic-logs an error return when it closes the incoming
5008 connection. However "connection reset by peer" seems to be common, and
5009 isn't really an error worthy of noting specially, so that particular error
5012 62. When Exim is trying to find all the local interfaces, it used to panic and
5013 die if the ioctl to get the interface flags failed. However, it seems that
5014 on at least one OS (Solaris 9) it is possible to have an interface that is
5015 included in the list of interfaces, but for which you get a failure error
5016 for this call. This happens when the interface is not "plumbed" into a
5017 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5018 failure of the "get flags" call assumes that the interface is down.
5020 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5021 makes life easier for people who are doing arithmetic on fields extracted
5022 from dates, where you often get leading zeros that should not be
5023 interpreted as octal.
5025 64. Added qualify_domain to the redirect router, to override the global
5028 65. If a pathologically long header line contained very many addresses (the
5029 report of this problem mentioned 10 000) and each of them was rewritten,
5030 Exim could use up a very large amount of memory. (It kept on making new
5031 copies of the header line as it rewrote, and never released the old ones.)
5032 At the expense of a bit more processing, the header rewriting function has
5033 been changed so that it no longer eats memory in this way.
5035 66. The generation of the Received: header has been moved from the time that a
5036 message starts to be received, to the time that it finishes. The timestamp
5037 in the Received: header should now be very close to that of the <= log
5038 line. There are two side-effects of this change:
5040 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5041 logged header lines no longer include the local Received: line, because
5042 it has not yet been created. The same applies to a copy of the message
5043 that is returned to a non-SMTP sender when a message is rejected.
5045 (b) When a filter file is tested using -bf, no additional Received: header
5046 is added to the test message. After some thought, I decided that this
5049 This change does not affect the value of $received_for. It is still set
5050 after address rewriting, but before local_scan() is called.
5052 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5054 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5055 gave an unhelpful panic error message, and a defer error. I have managed to
5056 change this behaviour so that it now rejects any supplied certificate,
5057 which seems right, as the list of acceptable certificates is empty.
5059 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5060 gave an unhelpful defer error. I have not managed to make this reject any
5061 supplied certificates, but the error message it gives is "no certificate
5062 supplied", which is not helpful.
5064 70. exigrep's output now also includes lines that are not associated with any
5065 message, but which match the given pattern. Implemented by a patch from
5066 Martin Sluka, which also tidied up the Perl a bit.
5068 71. Recipient callout verification, like sender verification, was using <> in
5069 the MAIL FROM command. This isn't really the right thing, since the actual
5070 sender may affect whether the remote host accepts the recipient or not. I
5071 have changed it to use the actual sender in the callout; this means that
5072 the cache record is now keyed on a recipient/sender pair, not just the
5073 recipient address. There doesn't seem to be a real danger of callout loops,
5074 since a callout by the remote host to check the sender would use <>.
5075 [SEE ABOVE: changed after hitting problems.]
5077 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5078 temporary errors. However, in the case of such a code being given after
5079 the end of a data transmission (i.e. after ".") Exim was failing to write
5080 a retry record for the message. (Yes, there was some broken host that was
5081 actually sending 8xx at this point.)
5083 73. An unknown lookup type in a host list could cause Exim to panic-die when
5084 the list was checked. (An example that provoked this was putting <; in the
5085 middle of a list instead of at the start.) If this happened during a DATA
5086 ACL check, a -D file could be left lying around. This kind of configuration
5087 error no longer causes Exim to die; instead it causes a defer errror. The
5088 incident is still logged to the main and panic logs.
5090 74. Buglet left over from Exim 3 conversion. The message "too many messages
5091 in one connection" was written to the rejectlog but not the mainlog, except
5092 when address rewriting (yes!) was being logged.
5094 75. Added write_rejectlog option.
5096 76. When a system filter was run not as root (that is, when system_filter_user
5097 was set), the values of the $n variables were not being returned to the
5098 main process; thus, they were not subsequently available in the $sn
5101 77. Added +return_path_on_delivery log selector.
5103 78. A connection timeout was being treated differently from recipients deferred
5104 when testing hosts_max_try with a message that was older than the host's
5105 retry timeout. (The host should not be counted, thus allowing all hosts to
5106 be tried at least once before bouncing.) This may have been the cause of an
5107 occasionally reported bug whereby a message would remain on the queue
5108 longer than the retry timeout, but would be bounced if a delivery was
5109 forced. I say "may" because I never totally pinned down the problem;
5110 setting up timeout/retry tests is difficult. See also the next item.
5112 79. The ultimate address timeout was not being applied to errors that involved
5113 a combination of host plus message (for example, a timeout on a MAIL
5114 command). When an address resolved to a number of possible hosts, and they
5115 were not all tried for each delivery (e.g. because of hosts_max_try), a
5116 message could remain on the queue longer than the retry timeout.
5118 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5121 81. Fixed an obscure SMTP outgoing bug which required at least the following
5122 conditions: (a) there was another message waiting for the same server;
5123 (b) the server returned 5xx to all RCPT commands in the first message so
5124 that the message was not completed; (c) the server dropped the connection
5125 or gave a negative response to the RSET that Exim sends to abort the
5126 transaction. The observed case was a dropped connection after DATA that had
5127 been sent in pipelining mode. That is, the server had advertised PIPELINING
5128 but was not implementing it correctly. The effect of the bug was incorrect
5129 behaviour, such as trying another host, and this could lead to a crash.
5135 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5136 and daemon.c were passed as pointers to ints; they should have been
5137 pointers to socklen_t variables (which are typically unsigned ints).
5139 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5142 3. Fixed a really odd bug that affected only the testing scheme; patching a
5143 certain fixed string in the binary changed the value of another string that
5144 happened to be identical to the end of the original first string.
5146 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5147 name", it returns that address as the IP address. On some operating
5148 systems (e.g. Solaris), it also passes back the IP address string as the
5149 "host name". However, on others (e.g. Linux), it passes back an empty
5150 string. Exim wasn't checking for this, and was changing the host name to an
5151 empty string, assuming it had been canonicized.
5153 5. Although rare, it is permitted to have more than one PTR record for a given
5154 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5155 all the names associated with an address, because they do in Solaris.
5156 However, it seems that they do not in Linux for data that comes from the
5157 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5158 I found this out when I moved to a new Linux workstation and tried to run
5159 the Exim test suite.
5161 To get round this problem I have changed the code so that it now does its
5162 own call to the DNS to look up PTR records when searching for a host name.
5163 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5164 addresses that are only in /etc/hosts are still found.
5166 This behaviour is, however, controlled by an option called host_lookup_
5167 order, which defaults to "bydns:byaddr". If people want to use the other
5168 order, or indeed, just use one or the other means of lookup, they can
5169 specify it in this variable.
5171 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5172 some operating systems, this comes back from gethostbyaddr() as an empty
5173 string, and this is what Exim used to test for. However, it seems that in
5174 other systems, "." is yielded. Exim now tests for this case too.
5176 7. The values of check_spool_space and check_log_space are now held internally
5177 as a number of kilobytes instead of an absolute number of bytes. If a
5178 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5179 kilobyte. This means that much larger values can be stored.
5181 8. Exim monitor: an attempt to get the action menu when not actually pointing
5182 at a message produces an empty menu entitled "No message selected". This
5183 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5184 no entries in it ("Shell widget menu has zero width and/or height"). So I
5185 have added a single, blank menu entry in this case.
5187 9. Added ${quote_local_part.
5189 10. MIME decoding is now applied to the contents of Subject: header lines when
5192 11. Now that a reference to $sender_host_address automatically causes a reverse
5193 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5194 host lookup before query-style lookups in lists that might use this
5195 variable. This has therefore been abolished, and the "net-" prefix is no
5196 longer necessary for query-style lookups.
5198 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5199 have been a typo for LFLAGS, so it has been changed.
5201 13. The install script calls Exim with "-C /dev/null" in order to find the
5202 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5203 to be output. Howeve, since Exim outputs its version number before the
5204 error, it didn't break the script. It just looked ugly. I fixed this by
5205 always allowing "-C /dev/null" if the caller is root.
5207 14. Ignore overlarge ACL variable number when reading spool file - insurance
5208 against a later release with more variables having written the file.
5210 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5211 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5212 this, as well as the form without the "IPv6" on the front.
5214 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5215 OS/Makefile-Darwin file.
5217 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5218 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5220 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5221 variable. However, some OS use ints for inodes. I've added cast to long int
5222 to get rid of the compiler warning.
5224 19. I had forgotten to lock out "/../" in configuration file names when
5225 ALT_CONFIG_PREFIX was set.
5227 20. Routers used for verification do not need to specify transports. However,
5228 if such a router generated a host list, and callout was configured, Exim
5229 crashed, because it could not find a port number from the (non-existent)
5230 transport. It now assumes port 25 in this circumstance.
5232 21. Added the -t option to exigrep.
5234 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5235 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5236 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5237 them is set, LOOKUP_LSEARCH is forced.
5239 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5241 24. Added sender and host information to the "rejected by local_scan()" log
5242 line; previously there was no indication of these.
5244 25. Added .include_if_exists.
5246 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5247 Linux. It turns out that not all file systems support this. Apparently some
5248 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5249 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5250 not supported on the file descriptor, is now ignored when Exim is trying to
5251 sync a directory. This applies only to Linux.
5253 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5255 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5256 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5259 29. Added support for maildirsize files from supplied patch (modified a bit).
5261 30. The use of :fail: followed by an empty string could lead Exim to respond to
5262 sender verification failures with (e.g.):
5264 550 Verification failed for <xxx>
5265 550 Sender verify failed
5267 where the first response line was missing the '-' that indicates it is not
5268 the final line of the response.
5270 31. The loop for finding the name of the user that called Exim had a hardwired
5271 limit of 10; it now uses the value of finduser_retries, which is used for
5272 all other user lookups.
5274 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5277 33. Exim was neglecting to zero errno before one call of strtol() when
5278 expanding a string and expecting an integer value. On some systems this
5279 resulted in spurious "integer overflow" errors. Also, it was casting the
5280 result into an int without checking.
5282 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5283 did not work. The code looks as if it has *never* worked, though it appears
5284 to have been documented since at least releast 1.62. I have made it work.
5286 35. The "timeout_DNS" error in retry rules, also documented since at least
5287 1.62, also never worked. As it isn't clear exactly what this means, and
5288 clearly it isn't a major issue, I have abolished the feature by treating it
5289 as "timeout", and writing a warning to the main and panic logs.
5291 36. The display of retry rules for -brt wasn't always showing the error code
5294 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5295 timeout_connect_A, timeout_connect_MX.
5297 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5298 to the empty sender.
5300 39. The daemon was not analysing the content of -oX till after it had closed
5301 stderr and disconnected from the controlling terminal. This meant that any
5302 syntax errors were only noted on the panic log, and the return code from
5303 the command was 0. By re-arranging the code a little, I've made the
5304 decoding happen first, so such errors now appear on stderr, and the return
5305 code is 1. However, the actual setting up of the sockets still happens in
5306 the disconnected process, so errors there are still only recorded on the
5309 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5310 connections (as happens on some IP stacks) was logged at start up time as
5311 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5312 it from "IPv6 and IPv4", which means that two separate sockets are being
5315 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5316 says whether AF_INET or AF_INET6 was passed as an argument.
5318 42. Exiwhat output was messed up when time zones were included in log
5321 43. Exiwhat now gives more information about the daemon's listening ports,
5322 and whether -tls-on-connect was used.
5324 44. The "port" option of the smtp transport is now expanded.
5326 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5327 silently ignored. Now an error message is written to the main and panic
5330 46. There's a new ACL modifier called "logwrite" which writes to a log file
5331 as soon as it is encountered.
5333 47. Added $local_user_uid and $local_user_gid at routing time.
5335 48. Exim crashed when trying to verify a sender address that was being
5338 49. Exim was recognizing only a space character after ".include". It now also
5339 recognizes a tab character.
5341 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5342 extracting the relevant information from the specification. The man page no
5343 longer contains scrambled data for the -d option, and I've added a section
5344 at the front about calling Exim under different names.
5346 51. Added "extra_headers" argument to the "mail" command in filter files.
5348 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5351 53. Installed eximstats 1.29.
5353 54. Added transport_filter_timeout as a generic transport option.
5355 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5356 Cc: header lines. This was required by RFC 822, but it not required by RFC
5359 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5360 incoming messages that did not have them. Now it does so only if the
5361 message originates locally, that is, if there is no associated remote host
5362 address. When Resent- header lines are present, this applies to the Resent-
5363 lines rather than the non-Resent- lines.
5365 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5366 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5368 58. Messages for configuration errors now include the name of the main
5369 configuration file - useful now that there may be more than one file in a
5370 list (.included file names were always shown).
5372 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5373 for those rare installations that do not start the daemon as root or run it
5374 setuid root. I've cut out the call to initgroups() if the daemon is not
5377 60. The Exim user and group can now be bound into the binary as text strings
5378 that are looked up at the start of Exim's processing.
5380 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5382 62. Added $mailstore_basename variable.
5384 63. Installed patch to sieve.c from Michael Haardt.
5386 64. When Exim failed to open the panic log after failing to open the main log,
5387 the original message it was trying to log was written to stderr and debug
5388 output, but if they were not available (the usual case in production), it
5389 was lost. Now it is written to syslog before the two lines that record the
5390 failures to open the logs.
5392 65. Users' Exim filters run in subprocesses under the user's uid. It is
5393 possible for a "deliver" command or an alias in a "personal" command to
5394 provoke an address rewrite. If logging of address rewriting is configured,
5395 this fails because the process is not running as root or exim. There may be
5396 a better way of dealing with this, but for the moment (because 4.30 needs
5397 to be released), I have disabled address rewrite logging when running a
5398 filter in a non-root, non-exim process.
5404 1. The buildconfig auxiliary program wasn't quoting the value set for
5405 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5406 not defined. This bug was masked in 4.22 by the effect that was fixed in
5409 2. Some messages that were rejected after a message id was allocated were
5410 shown as "incomplete" by exigrep. It no longer does this for messages that
5411 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5413 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5414 have allow_domain_literals set, the ID did not get logged in the <= line.
5415 Domain literals are now always recognized in Message-ID: header lines.
5417 4. The first argument for a ${extract expansion item is the key name or field
5418 number. Leading and trailing spaces in this item were not being ignored,
5419 causing some misleading effects.
5421 5. When deliver_drop_privilege was set, single queue runner processes started
5422 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5423 same command in the process it spins off) were not dropping privilege.
5425 6. When the daemon running as "exim" started a queue runner, it always
5426 re-executed Exim in the spun-off process. This is a waste of effort when
5427 deliver_drop_privilege is set. The new process now just calls the
5428 queue-runner function directly.
5434 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5437 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5438 ignored. Though the use of -oP was forcing the writing of a pid file, it
5439 was always written to the default place.
5441 3. If the message "no IP address found for host xxxx" is generated during
5442 incoming verification, it is now followed by identification of the incoming
5443 connection (so you can more easily find what provoked it).
5445 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5447 5. Added some features to "harden" Exim a bit more against certain attacks:
5449 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5450 be put in Local/Makefile. This is like the never_users runtime option,
5451 but it cannot be overridden. The default setting is "root".
5453 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5454 prefix string with which any file named in a -C command line option
5457 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5458 is retained for -C and -D only if the caller of Exim is root. Without
5459 it, the exim user may also use -C and -D and retain privilege.
5461 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5462 command line option is disabled.
5464 6. Macro names set by the -D option must start with an upper case letter, just
5465 like macro names defined in the configuration file.
5467 7. Added "dereference=" facility to LDAP.
5469 8. Two instances of the typo "uknown" in the source files are fixed.
5471 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5472 the Configure-Makefile script screwed up while processing it.
5474 10. Incorporated PCRE 4.4.
5476 11. The SMTP synchronization check was not operating right at the start of an
5477 SMTP session. For example, it could not catch a HELO sent before the client
5478 waited for the greeting. There is now a check for outstanding input at the
5479 point when the greeting is written. Because of the duplex, asynchronous
5480 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5481 way, but not yet received, when the check is performed.
5483 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5484 on TCP/IP sockets, because this apparently causes some broken clients to
5487 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5488 unchanged) from the Cygwin maintainer.
5490 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5491 was supported instead of testing for mbx. Effectively a typo.
5493 15. The spa authenticator server code was not checking that the input it
5494 received was valid base64.
5496 16. The debug output line for the "set" modifier in ACLs was not showing the
5497 name of the variable that was being set.
5499 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5501 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5502 lookup on its own. Something else was needed to trigger the lookup. For
5503 example, a match in host_lookup or the need for a host name in a host list.
5504 Now, if $sender_host_name is referenced and the host name has not yet been
5505 looked up, a lookup is performed. If the lookup fails, the variable remains
5506 empty, and $host_lookup_failed is set to "1".
5508 19. Added "eqi" as a case-independent comparison operator.
5510 20. The saslauthd authentication condition could segfault if neither service
5511 nor realm was specified.
5513 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5514 error message that was logged was misleading, and incoming SMTP
5515 connections were dropped. The message is now more accurate, and temporary
5516 errors are given to SMTP connections.
5518 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5519 (except RSET) with a 421 error, until QUIT is received. However, it was
5520 failing to send a response to QUIT.
5522 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5523 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5524 ACL is now run after the helo_try_verify_hosts code.
5526 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5527 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5528 case-independent, so other case variants are also recognized). Apparently
5529 some systems use these upper case variants.
5531 25. If more than two messages were waiting for the same host, and a transport
5532 filter was specified for the transport, Exim sent two messages over the
5533 same TCP/IP connection, and then failed with "socket operation on non-
5534 socket" when it tried to send the third.
5536 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5538 27. The extern definition of crypt16() in expand.c was not being excluded when
5539 the OS had its own crypt16() function.
5541 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5542 as a preferred synonym for return_size_limit, both as an option and as an
5545 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5547 30. Changed the default configuration ACL to relax the local part checking rule
5548 for addresses that are not in any local domains. For these addresses,
5549 slashes and pipe symbols are allowed within local parts, but the sequence
5550 /../ is explicitly forbidden.
5552 31. SPA server authentication was not clearing the challenge buffer before
5555 32. log_message in a "warn" ACL statement was writing to the reject log as
5556 well as to the main log, which contradicts the documentation and doesn't
5557 seem right (because no rejection is happening). So I have stopped it.
5559 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5560 However, I am unable to do any testing of this.
5562 34. Fixed an infelicity in the appendfile transport. When checking directories
5563 for a mailbox, to see if any needed to be created, it was accidentally
5564 using path names with one or more superfluous leading slashes; tracing
5565 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5567 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5568 contents are added to the log line that is written for every discarded
5569 recipient. (Previously a log_message setting was ignored.)
5571 36. The ${quote: operator now quotes the string if it is empty.
5573 37. The install script runs exim in order to find its version number. If for
5574 some reason other than non-existence or emptiness, which it checks, it
5575 could not run './exim', it was installing it with an empty version number,
5576 i.e. as "exim-". This error state is now caught, and the installation is
5579 38. An argument was missing from the function that creates an error message
5580 when Exim fails to connect to the socket for saslauthd authentication.
5581 This could cause Exim to crash, or give a corrupted message.
5583 39. Added isip, isip4, and isip6 to ${if conditions.
5585 40. The ACL variables $acl_xx are now saved with the message, and can be
5586 accessed later in routers, transports, and filters.
5588 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5589 strings in the file are not string-expanded.
5591 42. If a MAIL command specified a SIZE value that was too large to fit into an
5592 int variable, the check against message_size_limit failed. Such values are
5593 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5594 day this will have to be increased, but I don't think I want to be around
5595 when emails are that large.
5602 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5603 iconv() is not standard in FreeBSD.
5605 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5606 IPv6 enabled. The observed symptom was a segmentation fault on return from
5607 the function os_common_find_running_interfaces() in src/os.c.
5609 3. In the check_special_case() function in daemon.c I had used "errno" as an
5610 argument name, which causes warnings on some systems. This was basically a
5611 typo, since it was named "eno" in the comments!
5613 4. The code that waits for the clock to tick (at a resolution of some fraction
5614 of a second) so as to ensure message-id uniqueness was always waiting for
5615 at least one whole tick, when it could have waited for less. [This is
5616 almost certainly not relevant at current processor speeds, where it is
5617 unlikely to ever wait at all. But we try to future-proof.]
5619 5. The function that sleeps for a time interval that includes fractions of a
5620 second contained a race. It did not block SIGALRM between setting the
5621 timer, and suspending (a couple of lines later). If the interval was short
5622 and the sigsuspend() was delayed until after it had expired, the suspension
5623 never ended. On busy systems this could lead to processes getting stuck for
5626 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5627 process, before it forks any delivery processes. The open lookup caching
5628 mechanism meant that the open file or database connection was passed into
5629 the delivery process. The problem was that delivery processes always tidy
5630 up cached lookup data. This could cause a problem for the next delivery
5631 process started by the queue runner, because the external queue runner
5632 process does not know about the closure. So the next delivery process
5633 still has data in the lookup cache. In the case of a file lookup, there was
5634 no problem because closing a file descriptor in a subprocess doesn't affect
5635 the parent. However, if the lookup was caching a connection to a database,
5636 the connection was closed, and the second delivery process was likely to
5637 see errors such as "PGSQL: query failed: server closed the connection
5638 unexpectedly". The problem has been fixed by closing all cached lookups
5639 in a queue runner before running a delivery process.
5641 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5642 seem to have the "const" qualifier which it has on other OS. I've
5645 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5646 *of the same type* (client or server) with the same public name that an
5647 error should be diagnosed.
5649 9. When Exim looked up a host name for an IP address, but failed to find the
5650 original IP address when looking up the host name (a safety check), it
5651 output the message "<ip address> does not match any IP for NULL", which was
5652 confusing, to say the least. The bug was that the host name should have
5653 appeared instead of "NULL".
5655 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5656 user that is built into the binary, and the -C or -D options is used, root
5657 privilege is dropped before the configuration file is read. In addition,
5658 logging is switched to stderr instead of the normal log files. If the
5659 configuration then re-defines the Exim user, the unprivileged environment
5660 is probably not what is expected, so Exim logs a panic warning message (but
5663 However, if deliver_drop_privilege is set, the unprivileged state may well
5664 be exactly what is intended, so the warning has been cut out in that case,
5665 and Exim is allowed to try to write to its normal log files.
5671 1. smtp_return_error_details was not giving details for temporary sender
5672 or receiver verification errors.
5674 2. Diagnose a configuration error if two authenticators have the same public
5677 3. Exim used not to create the message log file for a message until the first
5678 delivery attempt. This could be confusing when incoming messages were held
5679 for policy or load reasons. The message log file is now created at the time
5680 the message is received, and an initial "Received" line is written to it.
5682 4. The automatically generated man page for command line options had a minor
5683 bug that caused no ill effects; however, a more serious problem was that
5684 the procedure for building the man page automatically didn't always
5685 operate. Consequently, release 4.20 contains an out-of-date version. This
5686 shouldn't happen again.
5688 5. When building Exim with embedded Perl support, the script that builds the
5689 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5690 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5692 6. The freeze_tell option was not being used for messages that were frozen on
5693 arrival, either by an ACL or by local_scan().
5695 7. Added the smtp_incomplete_transaction log selector.
5697 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5698 was accepting AUTH without a new EHLO.
5700 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5701 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5704 10. Logging of TCP/IP connections (when configured) now happens in the main
5705 daemon process instead of the child process, so that the TCP/IP connection
5706 count is more accurate (but it can never be perfect).
5708 11. The use of "drop" in a nested ACL was not being handled correctly in the
5709 outer ACL. Now, if condition failure induced by the nested "drop" causes
5710 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5711 or "require"), the connection is dropped.
5713 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5714 that yield "discard" can now be used with an "accept" or a "discard" verb,
5715 but an error is generated for any others (because I can't see a useful way
5716 to define what should happen).
5718 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5719 are now processed in the same way as lines in the Exim configuration file.
5720 In particular, continuation lines are supported.
5722 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5724 15. Added -ti meaning -t -i.
5726 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5727 domains, and warn by logging if others are found.
5729 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5730 returned from GIFCONF when Exim is trying to find the list of interfaces on
5731 a host. The code in os.c has been modified to copy each ifreq to an aligned
5732 structure in all cases.
5734 Also, in some cases, the returned ifreq's were being copied to a 'struct
5735 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5736 means the last couple of bytes of an IPv6 address could be chopped if the
5737 ifreq contained only a normal sockaddr (14 bytes storage).
5739 18. Named domain lists were not supported in the hosts_treat_as_local option.
5740 An entry such as +xxxx was not recognized, and was treated as a literal
5743 19. Ensure that header lines added by a DATA ACL are included in the reject log
5744 if the ACL subsequently rejects the message.
5746 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5747 MD5 (which is deprecated).
5749 21. When testing a filter file using -bf, Exim was writing a message when it
5750 took the sender from a "From " line in the message, but it was not doing so
5751 when it took $return_path from a Return-Path: header line. It now does.
5753 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5754 with a valid header line field name (a series of printing characters
5755 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5757 23. Changed "disc" in the source to "disk" to conform to the documentation and
5758 the book and for uniformity.
5760 24. Ignore Sendmail's -Ooption=value command line item.
5762 25. When execve() failed while trying to run a command in a pipe transport,
5763 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5764 could be confused with a return value of 69 from the command itself. This
5765 has been changed to 127, the value the shell returns if it is asked to run
5766 a non-existent command. The wording for the related log line suggests a
5767 non-existent command as the problem.
5769 26. If received_header_text expands to an empty string, do not add a Received:
5770 header line to the message. (Well, it adds a token one on the spool, but
5771 marks it "old" so that it doesn't get used or transmitted.)
5773 27. Installed eximstats 1.28 (addition of -nt option).
5775 28. There was no check for failure on the call to getsockname() in the daemon
5776 code. This can fail if there is a shortage of resources on the system, with
5777 ENOMEM, for example. A temporary error is now given on failure.
5779 29. Contrary to the C standard, it seems that in some environments, the
5780 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5781 program. Exim now does this explicitly; it affects the formatting of
5782 timestamps using strftime().
5784 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5785 complaints. I've now initialized all the variables, to avoid this.
5787 32. Header lines added by a system filter were not being "seen" during
5788 transport-time rewrites.
5790 33. The info_callback() function passed to OpenSSL is set up with type void
5791 (*)(SSL *, int, int), as described somewhere. However, when calling the
5792 function (actually a macro) that sets it up, the type void(*)() is
5793 expected. I've put in a cast to prevent warnings from picky compilers.
5795 34. If a DNS black list lookup found a CNAME record, but there were no A
5796 records associated with the domain it pointed at, Exim crashed.
5798 35. If a DNS black list lookup returned more than one A record, Exim ignored
5799 all but the first. It now scans all returned addresses if a particular IP
5800 value is being sought. In this situation, the contents of the
5801 $dnslist_value variable are a list of all the addresses, separated by a
5804 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5805 to accept a host name and all its aliases if the forward lookup for any of
5806 them yielded the IP address of the incoming connection. Now it accepts only
5807 those names whose forward lookup yields the correct IP address. Any other
5808 names are discarded. This closes a loophole whereby a rogue DNS
5809 administrator could create reverse DNS records to break through a
5810 wildcarded host restriction in an ACL.
5812 37. If a user filter or a system filter that ran in a subprocess used any of
5813 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5814 the wrong values were passed to the pipe command ($thisaddress had the
5815 value of $0, $0 had the value of $1, etc). This bug was introduced by
5816 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5818 38. Improved the line breaking for long SMTP error messages from ACLs.
5819 Previously, if there was no break point between 40 and 75 characters, Exim
5820 left the rest of the message alone. Two changes have been made: (a) I've
5821 reduced the minimum length to 35 characters; (b) if it can't find a break
5822 point between 35 and 75 characters, it looks ahead and uses the first one
5823 that it finds. This may give the occasional overlong line, but at least the
5824 remaining text gets split now.
5826 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5827 file descriptors might be low, and that setting 1000 would always raise it.
5828 It turns out that in some environments, the limit is already over 1000 and
5829 that lowering it causes trouble. So now Exim takes care not to decrease it.
5831 40. When delivering a message, the value of $return_path is set to $sender_
5832 address at the start of routing (routers may change the value). By an
5833 oversight, this default was not being set up when an address was tested by
5834 -bt or -bv, which affected the outcome if any router or filter referred to
5837 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5838 log only when "message" or "log_message" are set. However, if one of the
5839 conditions was an address verification, or a call to a nested ACL, the
5840 messages generated by the underlying test were being passed through. This
5841 no longer happens. The underlying message is available in $acl_verify_
5842 message for both "message" and "log_message" expansions, so it can be
5843 passed through if needed.
5845 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5846 new expansion $bh_ to give the encoded byte string without charset
5847 translation. Translation happens only if iconv() is available; HAVE_ICONV
5848 indicates this at build time. HEADERS_CHARSET gives the charset to
5849 translate to; headers_charset can change it in the configuration, and
5850 "headers charset" can change it in an individual filter file.
5852 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5853 that creates RFC 2047 encoded "words" labels them as that charset instead
5854 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5855 expansion operator; (ii) when Exim creates a From: line for a local
5856 message; (iii) when a header line is rewritten to include a "phrase" part.
5858 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5859 buggy, causing it to skip the first lines of messages whose message ID
5860 ended in 'D'. This would not have bitten before Exim release 4.14, because
5861 message IDs were unlikely to end in 'D' before then. The effect was to have
5862 incorrect size information for certain domains.
5864 45. #include "config.h" was missing at the start of the crypt16.c module. This
5865 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5868 46. If there was a timeout during a "random" callout check, Exim treated it as
5869 a failure of the random address, and carried on sending RSET and the real
5870 address. If the delay was just some slowness somewhere, the response to the
5871 original RCPT would be taken as a response to RSET and so on, causing
5872 mayhem of various kinds.
5874 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5875 when I implemented it. It didn't allow for the fact that some option values
5876 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5877 the right test for positive values.
5879 48. Domain names in DNS records are case-independent. Exim always looks them up
5880 in lower case. Some resolvers return domain names in exactly the case they
5881 appear in the zone file, that is, they may contain uppercase letters. Not
5882 all resolvers do this - some return always lower case. Exim was treating a
5883 change of case by a resolver as a change of domain, similar to a widening
5884 of a domain abbreviation. This triggered its re-routing code and so it was
5885 trying to route what was effectively the same domain again. This normally
5886 caused routing to fail (because the router wouldn't handle the domain
5887 twice). Now Exim checks for this case specially, and just changes the
5888 casing of the domain that it ultimately uses when it transmits the message
5891 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5894 50. If a filter generated a file delivery with a non-absolute name (possible if
5895 no home directory exists for the router), the forbid_file option was not
5898 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5899 the existing equality matching.
5901 52. Exim was using ints instead of ino_t variables in some places where it was
5902 dealing with inode numbers.
5904 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5905 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5906 called TMPDIR, and if it finds it is different, it changes its value.
5908 54. The smtp_printf() function is now made available to local_scan() so
5909 additional output lines can be written before returning. There is also an
5910 smtp_fflush() function to enable the detection of a dropped connection.
5911 The variables smtp_input and smtp_batched_input are exported to
5914 55. Changed the default runtime configuration: the message "Unknown user"
5915 has been removed from the ACL, and instead placed on the localuser router,
5916 using the cannot_route_message feature. This means that any verification
5917 failures that generate their own messages won't get overridden. Similarly,
5918 the "Unrouteable address" message that was in the ACL for unverifiable
5919 relay addresses has also been removed.
5921 56. Added hosts_avoid_esmtp to the smtp transport.
5923 57. The exicyclog script was not checking for the esoteric option
5924 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5925 will work only if exicyclog is run under the appropriate euid.
5927 58. Following a discussion on the list, the rules by which Exim recognises line
5928 endings on incoming messages have been changed. The -dropcr and drop_cr
5929 options are now no-ops, retained only for backwards compatibility. The
5930 following line terminators are recognized: LF CRLF CR. However, special
5931 processing applies to CR:
5933 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5934 nor a local message in the state where . is a terminator.
5936 (ii) If a bare CR is encountered in a header line, an extra space is added
5937 after the line terminator so as not to end the header. The reasoning
5938 behind this is that bare CRs in header lines are most likely either
5939 to be mistakes, or people trying to play silly games.
5941 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5942 was being incorrectly given as 18 bytes larger than it should have been.
5943 This is a VOB (very old bug).
5945 60. This may never have affected anything current, but just in case it has:
5946 When the local host is found other than at the start of a list of hosts,
5947 the local host, those with the same MX, and any that follow, are discarded.
5948 When the list in question was part of a longer list of hosts, the following
5949 hosts (not currently being processed) were also being discarded. This no
5950 longer happens. I'm not sure if this situation could ever has previously
5953 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5956 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5957 header. This is something that is recommended in a new Internet Draft, and
5958 is something that is documented as being done by Sendmail. There are two
5959 possible values. For messages generated by the autoreply transport, Exim
5962 Auto-Submitted: auto-replied
5964 whereas for all other generated messages (e.g. bounces) it adds
5966 Auto-Submitted: auto-generated
5968 63. The "personal" condition in filters now includes a test for the
5969 Auto-Submitted: header. If it contains the string "auto-" the message it
5970 not considered personal.
5972 64. Added rcpt_include_affixes as a generic transport option.
5974 65. Added queue_only_override (default true).
5976 66. Added the syslog_duplication option.
5978 67. If what should have been the first header line of a message consisted of
5979 a space followed by a colon, Exim was mis-interpreting it as a header line.
5980 It isn't of course - it is syntactically invalid and should therefore be
5981 treated as the start of the message body. The misbehaviour could have
5982 caused a number of strange effects, including loss of data in subsequent
5983 header lines, and spool format errors.
5985 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5986 client host had authenticated. This control can now be exercised by an ACL
5987 for more flexibility.
5989 69. By default, callouts do not happen when testing with -bh. There is now a
5990 variant, -bhc, which does actually run the callout code, including
5991 consulting and updating the callout cache.
5993 70. Added support for saslauthd authentication, courtesy of Alexander
5996 71. If statvfs() failed on the spool or log directories while checking their
5997 size for availability, Exim confusingly gave the error "space shortage".
5998 Furthermore, in debugging mode it crashed with a floating point exception.
5999 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6000 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6001 serious problem, Exim now writes to the main and panic logs when this
6002 happens, with details of the failure. It then refuses to accept the
6003 incoming message, giving the message "spool directory problem" or "log
6004 directory problem" with a 421 code for SMTP messages.
6006 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6007 0, 1, and 2 exist, because some OS complain for execs without them (see
6008 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6009 descriptors. However, the code omitted to check that the open succeeded,
6010 causing mysterious errors if for some reason the permissions on /dev/null
6011 got screwed. Now Exim writes a message to the main and panic logs, and
6012 bombs out if it can't open /dev/null.
6014 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6015 interact so that it is all more flexible. It is supposed to remain
6016 backwards compatible. Also added extra_local_interfaces.
6018 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6019 to bomb out with an assertion failure - to the client this appears as a
6020 connection drop. This problem occurs in the part of the code that was taken
6021 from the Samba project. Fortunately, the assertion is in a very simple
6022 function, so I have fixed this by reproducing the function inline in the
6023 one place where it is called, and arranging for authentication to fail
6024 instead of killing the process with assert().
6026 75. The SPA client code was not working when the server requested OEM rather
6027 than Unicode encoding.
6029 76. Added code to make require_files with a specific uid setting more usable in
6030 the case where statting the file as root fails - usually a non-root-mounted
6031 NFS file system. When this happens and the failure is EACCES, Exim now
6032 forks a subprocess and does the per-uid checking as the relevant uid.
6034 77. Added process_log_path.
6036 78. If log_file_path was not explicitly set, a setting of check_log_space or
6037 check_log_inodes was ignored.
6039 79. If a space check for the spool or log partitions fails, the incident is now
6040 logged. Of course, in the latter case the data may get lost...
6042 80. Added the %p formatting code to string_format() so that it can be used to
6043 print addresses in debug_print(). Adjusted all the address printing in the
6044 debugging in store.c to use %p rather than %d.
6046 81. There was a concern that a line of code in smtp_in.c could overflow a
6047 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6048 initially expressed, the concern was not well-founded, because trailing
6049 spaces are removed early. However, if the trailing spaces were followed by
6050 a NULL, they did not get removed, so the overflow was possible. Two fixes
6053 (a) I re-wrote the offending code in a cleaner fashion.
6054 (b) If an incoming SMTP command contains a NULL character, it is rejected
6057 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6058 runs initgroups(), so that if the Exim user is in any additional groups,
6059 they will be used during message reception.
6065 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob