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.
149 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
151 PP/02 Make -n do something, by making it not do something.
152 When combined with -bP, the name of an option is not output.
154 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
157 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
158 $sender_host_name and config options to manage this, and basic check
161 PP/05 DSCP support for outbound connections and control modifier for inbound.
163 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
164 (Only plugin which currently uses this is kerberos4, which nobody should
165 be using, but we should make it available and other future plugins might
166 conceivably use it, even though it would break NAT; stuff *should* be
167 using channel bindings instead).
169 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
170 name; added for Sendmail compatibility; requires admin caller.
171 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
172 just ignore it); requires trusted caller.
173 Also parse but ignore: -Ac -Am -X<logfile>
176 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
178 TL/02 Add +smtp_confirmation as a default logging option.
180 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
181 Patch by Magnus Holmgren from 2007-02-20.
183 TL/04 Bugzilla 1281 - Spec typo.
184 Bugzilla 1283 - Spec typo.
185 Bugzilla 1290 - Spec grammar fixes.
187 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
189 TL/06 Add Experimental DMARC support using libopendmarc libraries.
191 TL/07 Fix an out of order global option causing a segfault. Reported to dev
192 mailing list by by Dmitry Isaikin.
194 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
196 JH/02 Support "G" suffix to numbers in ${if comparisons.
198 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
200 NM/01 Bugzilla 1197 - Spec typo
201 Bugzilla 1196 - Spec examples corrections
203 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
205 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
206 gnutls_enable_pkcs11, but renamed to more accurately indicate its
209 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
210 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
212 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
213 "acl {{name}{arg}...}", and optional args on acl condition
216 JH/05 Permit multiple router/transport headers_add/remove lines.
218 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
220 JH/07 Avoid using a waiting database for a single-message-only transport.
221 Performance patch from Paul Fisher. Bugzilla 1262.
223 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
226 JH/09 Add $headers_added variable, with content from use of ACL modifier
227 add_header (but not yet added to the message). Bugzilla 199.
229 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
230 Pulled from Bugzilla 817 by Wolfgang Breyha.
232 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
234 (nb: this is the same fix as in Exim 4.80.1)
236 JH/11 Add A= logging on delivery lines, and a client_set_id option on
239 JH/12 Add optional authenticated_sender logging to A= and a log_selector
242 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
244 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
245 advertise SMTP AUTH mechanism to us, instead of a generic
246 protocol violation error. Also, make Exim more robust to bad
247 data from the Dovecot auth socket.
249 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
251 When a queue runner is handling a message, Exim first routes the
252 recipient addresses, during which it prunes them based on the retry
253 hints database. After that it attempts to deliver the message to
254 any remaining recipients. It then updates the hints database using
257 So if a recipient address works intermittently, it can get repeatedly
258 deferred at routing time. The retry hints record remains fresh so the
259 address never reaches the final cutoff time.
261 This is a fairly common occurrence when a user is bumping up against
262 their storage quota. Exim had some logic in its local delivery code
263 to deal with this. However it did not apply to per-recipient defers
264 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
266 This change adds a proper retry rule check during routing so that the
267 final cutoff time is checked against the message's age. We only do
268 this check if there is an address retry record and there is not a
269 domain retry record; this implies that previous attempts to handle
270 the address had the retry_use_local_parts option turned on. We use
271 this as an approximation for the destination being like a local
272 delivery, as in LMTP.
274 I suspect this new check makes the old local delivery cutoff check
275 redundant, but I have not verified this so I left the code in place.
277 TF/02 Correct gecos expansion when From: is a prefix of the username.
279 Test 0254 submits a message to Exim with the header
283 When I ran the test suite under the user fanf2, Exim expanded
284 the header to contain my full name, whereas it should have added
285 a Resent-Sender: header. It erroneously treats any prefix of the
286 username as equal to the username.
288 This change corrects that bug.
290 GF/01 DCC debug and logging tidyup
291 Error conditions log to paniclog rather than rejectlog.
292 Debug lines prefixed by "DCC: " to remove any ambiguity.
294 TF/03 Avoid unnecessary rebuilds of lookup-related code.
296 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
297 Bug spotted by Jeremy Harris; was flawed since initial commit.
298 Would have resulted in OCSP responses post-SNI triggering an Exim
299 NULL dereference and crash.
301 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
303 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
304 Bug detection, analysis and fix by Samuel Thibault.
305 Bugzilla 1331, Debian bug #698092.
307 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
309 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
310 Server implementation by Todd Lyons, client by JH.
311 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
312 config variable "prdr_enable" controls whether the server
313 advertises the facility. If the client requests PRDR a new
314 acl_data_smtp_prdr ACL is called once for each recipient, after
315 the body content is received and before the acl_smtp_data ACL.
316 The client is controlled by bolth of: a hosts_try_prdr option
317 on the smtp transport, and the server advertisement.
318 Default client logging of deliveries and rejections involving
319 PRDR are flagged with the string "PRDR".
321 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
322 fclose(). Diagnosis by Todd Lyons.
324 PP/17 Update configure.default to handle IPv6 localhost better.
325 Patch by Alain Williams (plus minor tweaks).
328 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
329 This is now consistent with GnuTLS, and is now documented: the
330 previous undocumented portable approach to treating the option as
331 unset was to force an expansion failure. That still works, and
332 an empty string is now equivalent.
334 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
335 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
336 not performing validation itself.
338 PP/20 Added force_command boolean option to pipe transport.
339 Patch from Nick Koston, of cPanel Inc.
341 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
344 TF/04 Added udpsend ACL modifer and hexquote expansion operator
346 PP/21 Fix eximon continuous updating with timestamped log-files.
347 Broken in a format-string cleanup in 4.80, missed when I repaired the
348 other false fix of the same issue.
349 Report and fix from Heiko Schlichting.
352 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
353 Report from Prashanth Katuri.
355 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
356 It's SecureTransport, so affects any MacOS clients which use the
357 system-integrated TLS libraries, including email clients.
359 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
360 using a MIME ACL for non-SMTP local injection.
361 Report and assistance in diagnosis by Warren Baker.
363 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
365 JH/16 Fix comparisons for 64b. Bugzilla 1385.
367 TL/09 Add expansion variable $authenticated_fail_id to keep track of
368 last id that failed so it may be referenced in subsequent ACL's.
370 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
373 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
374 ldap library initialization, allowing self-signed CA's to be
375 used. Also properly sets require_cert option later in code by
376 using NULL (global ldap config) instead of ldap handle (per
377 session). Bug diagnosis and testing by alxgomz.
379 TL/12 Enhanced documentation in the ratelimit.pl script provided in
380 the src/util/ subdirectory.
382 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
383 renamed to Transport Post Delivery Action by Jeremy Harris, as
386 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
387 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
388 redis_servers = needs to be configured which will be used by the redis
389 lookup. Patch from Warren Baker, of The Packet Hub.
391 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
393 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
394 hostname or reverse DNS when processing a host list. Used suggestions
395 from multiple comments on this bug.
397 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
399 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
400 Missed a few lines, added it to make the runtest require no keyboard
403 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
404 contains upper case chars. Make router use caseful_local_part.
406 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
407 support when GnuTLS has been built with p11-kit.
413 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
415 This, or similar/improved, will also be change PP/11 of 4.82.
421 PP/01 Handle short writes when writing local log-files.
422 In practice, only affects FreeBSD (8 onwards).
423 Bugzilla 1053, with thanks to Dmitry Isaikin.
425 NM/01 Bugzilla 949 - Documentation tweak
427 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
430 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
432 PP/02 Implemented gsasl authenticator.
434 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
436 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
437 `pkg-config foo` for cflags/libs.
439 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
440 with rest of GSASL and with heimdal_gssapi.
442 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
443 `pkg-config foo` for cflags/libs for the TLS implementation.
445 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
446 properties get this fed in as external SSF. A number of robustness
447 and debugging improvements to the cyrus_sasl authenticator.
449 PP/08 cyrus_sasl server now expands the server_realm option.
451 PP/09 Bugzilla 1214 - Log authentication information in reject log.
452 Patch by Jeremy Harris.
454 PP/10 Added dbmjz lookup type.
456 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
458 PP/12 MAIL args handles TAB as well as SP, for better interop with
459 non-compliant senders.
460 Analysis and variant patch by Todd Lyons.
462 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
463 Bug report from Lars Müller <lars@samba.org> (via SUSE),
464 Patch from Dirk Mueller <dmueller@suse.com>
466 PP/13 tls_peerdn now print-escaped for spool files.
467 Observed some $tls_peerdn in wild which contained \n, which resulted
468 in spool file corruption.
470 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
471 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
472 or write after TLS renegotiation, which otherwise led to messages
475 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
476 as a tracking header (ie: a signed header comes before the signature).
477 Patch from Wolfgang Breyha.
479 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
480 comma-sep list; embedded commas doubled.
482 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
484 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
486 Report and patch from Dmitry Banschikov.
488 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
489 Removed SSL_clear() after SSL_new() which led to protocol negotiation
490 failures. We appear to now support TLS1.1+ with Exim.
492 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
493 lets Exim select keys and certificates based upon TLS SNI from client.
494 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
495 before an outbound SMTP session. New log_selector, +tls_sni.
497 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
498 NULL dereference. Report and patch from Alun Jones.
500 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
501 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
502 Not seeing resolver debug output on NetBSD, but suspect this is a
503 resolver implementation change.
505 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
506 Left warnings. Added "eximon gdb" invocation mode.
508 PP/21 Defaulting "accept_8bitmime" to true, not false.
510 PP/22 Added -bw for inetd wait mode support.
512 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
513 locate the relevant includes and libraries. Made this the default.
515 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
516 Bugzilla 1246, report and most of solution from Tomasz Kusy.
518 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
519 This may cause build issues on older platforms.
521 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
522 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
523 gnutls_require_mac & gnutls_require_protocols (no longer supported).
524 Added SNI support via GnuTLS too.
525 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
527 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
529 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
530 Applied second patch from Janne, implementing suggestion to default
531 multiple-strings-in-record handling to match SPF spec.
533 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
535 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
536 read-only, out of scope).
537 Patch from Wolfgang Breyha, report from Stuart Northfield.
539 PP/29 Fix three issues highlighted by clang analyser static analysis.
540 Only crash-plausible issue would require the Cambridge-specific
541 iplookup router and a misconfiguration.
542 Report from Marcin Mirosław.
544 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
546 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
547 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
548 As part of this, removing so much warning spew let me fix some minor
549 real issues in debug logging.
551 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
552 assignment on my part. Fixed.
554 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
555 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
556 Janne Snabb (who went above and beyond: thank you).
558 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
559 string otherwise requires a connection and a bunch more work and it's
560 relatively easy to get wrong. Should also expose TLS library linkage
563 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
564 64-bit ${eval} (JH/03).
566 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
567 GNU libc to support some of the 64-bit stuff, should not lead to
568 conflicts. Defined before os.h is pulled in, so if a given platform
569 needs to override this, it can.
571 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
572 protection layer was required, which is not implemented.
573 Bugzilla 1254, patch from Wolfgang Breyha.
575 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
576 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
577 tls_dhparam take prime identifiers. Also unbreak combination of
578 OpenSSL+DH_params+TLSSNI.
580 PP/39 Disable SSLv2 by default in OpenSSL support.
586 PP/01 Solaris build fix for Oracle's LDAP libraries.
587 Bugzilla 1109, patch from Stephen Usher.
589 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
591 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
594 TF/02 Fix a couple more cases where we did not log the error message
595 when unlink() failed. See also change 4.74-TF/03.
597 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
598 lock up or crash if it happened to be inside a call to libc when it
599 got a SIGUSR1 from exiwhat.
601 The SIGUSR1 handler appends the current process status to the process
602 log which is later printed by exiwhat. It used to use the general
603 purpose logging code to do this, but several functions it calls are
604 not safe for signals.
606 The new output code in the SIGUSR1 handler is specific to the process
607 log, and simple enough that it's easy to inspect for signal safety.
608 Removing some special cases also simplifies the general logging code.
609 Removing the spurious timestamps from the process log simplifies
612 TF/04 Improved ratelimit ACL condition.
614 The /noupdate option has been deprecated in favour of /readonly which
615 has clearer semantics. The /leaky, /strict, and /readonly update modes
616 are mutually exclusive. The update mode is no longer included in the
617 database key; it just determines when the database is updated. (This
618 means that when you upgrde Exim will forget old rate measurements.)
620 Exim now checks that the per_* options are used with an update mode that
621 makes sense for the current ACL. For example, when Exim is processing a
622 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
623 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
624 must specify per_mail/readonly. If you omit the update mode it defaults to
625 /leaky where that makes sense (as before) or /readonly where required.
627 The /noupdate option is now undocumented but still supported for
628 backwards compatibility. It is equivalent to /readonly except that in
629 ACLs where /readonly is required you may specify /leaky/noupdate or
630 /strict/noupdate which are treated the same as /readonly.
632 A useful new feature is the /count= option. This is a generalization
633 of the per_byte option, so that you can measure the throughput of other
634 aggregate values. For example, the per_byte option is now equivalent
635 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
637 The per_rcpt option has been generalized using the /count= mechanism
638 (though it's more complicated than the per_byte equivalence). When it is
639 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
640 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
641 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
642 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
643 Note that using per_rcpt with a non-readonly update mode in more than
644 one ACL will cause the recipients to be double-counted. (The per_mail
645 and per_byte options don't have this problem.)
647 The handling of very low rates has changed slightly. If the computed rate
648 is less than the event's count (usually one) then this event is the first
649 after a long gap. In this case the rate is set to the same as this event's
650 count, so that the first message of a spam run is counted properly.
652 The major new feature is a mechanism for counting the rate of unique
653 events. The new per_addr option counts the number of different
654 recipients that someone has sent messages to in the last time period. It
655 behaves like per_rcpt if all the recipient addresses are different, but
656 duplicate recipient addresses do not increase the measured rate. Like
657 the /count= option this is a general mechanism, so the per_addr option
658 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
659 example, measure the rate that a client uses different sender addresses
660 with the options per_mail/unique=$sender_address. There are further
661 details in the main documentation.
663 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
665 TF/06 Removed a few PCRE remnants.
667 TF/07 Automatically extract Exim's version number from tags in the git
668 repository when doing development or release builds.
670 PP/02 Raise smtp_cmd_buffer_size to 16kB.
671 Bugzilla 879. Patch from Paul Fisher.
673 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
674 Heavily based on revision 40f9a89a from Simon Arlott's tree.
677 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
679 PP/05 Variable $av_failed, true if the AV scanner deferred.
680 Bugzilla 1078. Patch from John Horne.
682 PP/06 Stop make process more reliably on build failure.
683 Bugzilla 1087. Patch from Heiko Schlittermann.
685 PP/07 Make maildir_use_size_file an _expandable_ boolean.
686 Bugzilla 1089. Patch from Heiko Schlittermann.
688 PP/08 Handle ${run} returning more data than OS pipe buffer size.
689 Bugzilla 1131. Patch from Holger Weiß.
691 PP/09 Handle IPv6 addresses with SPF.
692 Bugzilla 860. Patch from Wolfgang Breyha.
694 PP/10 GnuTLS: support TLS 1.2 & 1.1.
696 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
699 PP/11 match_* no longer expand right-hand-side by default.
700 New compile-time build option, EXPAND_LISTMATCH_RHS.
701 New expansion conditions, "inlist", "inlisti".
703 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
705 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
707 PP/14 fix log_write() format string regression from TF/03.
708 Bugzilla 1152. Patch from Dmitry Isaikin.
714 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
716 PP/02 Harmonised TLS library version reporting; only show if debugging.
717 Layout now matches that introduced for other libraries in 4.74 PP/03.
719 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
721 PP/04 New "dns_use_edns0" global option.
723 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
726 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
727 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
729 TK/01 Updated PolarSSL code to 0.14.2.
730 Bugzilla 1097. Patch from Andreas Metzler.
732 PP/07 Catch divide-by-zero in ${eval:...}.
735 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
738 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
739 format-string attack -- SECURITY: remote arbitrary code execution.
741 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
742 time unintentionally subject to list matching rules, letting the header
743 cause arbitrary Exim lookups (of items which can occur in lists, *not*
744 arbitrary string expansion). This allowed for information disclosure.
746 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
747 INT_MIN/-1 -- value coerced to INT_MAX.
753 NM/01 Workround for PCRE version dependency in version reporting
756 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
757 This fixes portability to compilers other than gcc, notably
758 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
760 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
761 makefiles for portability to HP-UX and POSIX correctness.
763 PP/01 Permit LOOKUP_foo enabling on the make command-line.
764 Also via indented variable definition in the Makefile.
765 (Debugging by Oliver Heesakkers).
767 PP/02 Restore caching of spamd results with expanded spamd_address.
768 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
770 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
771 Improves build reliability. Fix from: Frank Elsner
773 NM/02 Fix wide character breakage in the rfc2047 coding
774 Fixes bug 1064. Patch from Andrey N. Oktyabrski
776 NM/03 Allow underscore in dnslist lookups
777 Fixes bug 1026. Patch from Graeme Fowler
779 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
780 Code patches from Adam Ciarcinski of NetBSD.
782 NM/04 Fixed exiqgrep to cope with mailq missing size issue
785 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
786 is logged, to avoid truncation. Patch from John Horne.
788 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
789 Patch from Jakob Hirsch.
791 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
792 SQL string expansion failure details.
793 Patch from Andrey Oktyabrski.
795 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
796 Patch from Simon Arlott.
798 PP/09 New lookups functionality failed to compile on old gcc which rejects
799 extern declarations in function scope.
800 Patch from Oliver Fleischmann
802 PP/10 Use sig_atomic_t for flags set from signal handlers.
803 Check getgroups() return and improve debugging.
804 Fixed developed for diagnosis in bug 927 (which turned out to be
807 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
808 Patch from Mark Zealey.
810 PP/12 Bugzilla 1056: Improved spamd server selection.
811 Patch from Mark Zealey.
813 PP/13 Bugzilla 1086: Deal with maildir quota file races.
814 Based on patch from Heiko Schlittermann.
816 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
817 Patch from Uwe Doering, sign-off by Michael Haardt.
819 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
820 variable declaration deep within a block. Bug and patch from
823 PP/15 lookups-Makefile IRIX compatibilty coercion.
825 PP/16 Make DISABLE_DKIM build knob functional.
827 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
828 Patch by Simon Arlott
830 TF/03 Fix valgrind.h portability to C89 compilers that do not support
831 variable argument macros. Our copy now differs from upstream.
837 TF/01 Failure to get a lock on a hints database can have serious
838 consequences so log it to the panic log.
840 TF/02 Log LMTP confirmation messages in the same way as SMTP,
841 controlled using the smtp_confirmation log selector.
843 TF/03 Include the error message when we fail to unlink a spool file.
845 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
846 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
847 for maintaining out-of-tree patches for some time.
849 PP/01 Bugzilla 139: Documentation and portability issues.
850 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
851 Handle per-OS dynamic-module compilation flags.
853 PP/02 Let /dev/null have normal permissions.
854 The 4.73 fixes were a little too stringent and complained about the
855 permissions on /dev/null. Exempt it from some checks.
856 Reported by Andreas M. Kirchwitz.
858 PP/03 Report version information for many libraries, including
859 Exim version information for dynamically loaded libraries. Created
860 version.h, now support a version extension string for distributors
861 who patch heavily. Dynamic module ABI change.
863 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
864 privilege escalation vulnerability whereby the Exim run-time user
865 can cause root to append content of the attacker's choosing to
868 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
871 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
872 If dropping privileges for untrusted macros, we disabled normal logging
873 on the basis that it would fail; for the Exim run-time user, this is not
874 the case, and it resulted in successful deliveries going unlogged.
875 Fixed. Reported by Andreas Metzler.
881 PP/01 Date: & Message-Id: revert to normally being appended to a message,
882 only prepend for the Resent-* case. Fixes regression introduced in
883 Exim 4.70 by NM/22 for Bugzilla 607.
885 PP/02 Include check_rfc2047_length in configure.default because we're seeing
886 increasing numbers of administrators be bitten by this.
888 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
890 PP/03 Bugzilla 994: added openssl_options main configuration option.
892 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
894 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
896 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
898 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
899 without a peer certificate, leading to a segfault because of an
900 assumption that peers always have certificates. Be a little more
901 paranoid. Problem reported by Martin Tscholak.
903 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
904 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
905 NB: ClamAV planning to remove STREAM in "middle of 2010".
906 CL also introduces -bmalware, various -d+acl logging additions and
907 more caution in buffer sizes.
909 PP/09 Implemented reverse_ip expansion operator.
911 PP/10 Bugzilla 937: provide a "debug" ACL control.
913 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
915 PP/12 Bugzilla 973: Implement --version.
917 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
919 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
921 PP/15 Bugzilla 816: support multiple condition rules on Routers.
923 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
924 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
925 ignore trailing whitespace.
927 JJ/02 prevent non-panic DKIM error from being sent to paniclog
929 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
932 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
933 Notification from Dr Andrew Aitchison.
935 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
936 ExtendedDetectionInfo response format.
937 Notification from John Horne.
939 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
942 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
943 XSL and documented dependency on system catalogs, with examples of how
946 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
949 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
950 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
951 configuration file which is writeable by the Exim user or group.
953 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
954 of configuration files to cover files specified with the -C option if
955 they are going to be used with root privileges, not just the default
958 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
959 option (effectively making it always true).
961 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
962 files to be used while preserving root privileges.
964 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
965 that rogue child processes cannot use them.
967 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
968 run-time user, instead of root.
970 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
971 Exim run-time user without dropping privileges.
973 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
974 result string, instead of calling string_vformat() twice with the same
977 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
978 for other users. Others should always drop root privileges if they use
979 -C on the command line, even for a whitelisted configure file.
981 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
983 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
989 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
990 $data_path, and $header_path variables; fixed documentation bugs and
993 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
994 exipick to access non-standard spools, including the "frozen" queue
997 NM/01 Bugzilla 965: Support mysql stored procedures.
998 Patch from Alain Williams
1000 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1002 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1003 Patch from Andreas Metzler
1005 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1006 Patch from Kirill Miazine
1008 NM/05 Bugzilla 671: Added umask to procmail example.
1010 JJ/03 installed exipick 20100323.0, fixing doc bug
1012 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1013 directory. Notification and patch from Dan Rosenberg.
1015 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1017 TK/02 Improve log output when DKIM signing operation fails.
1019 MH/01 Treat the transport option dkim_domain as a colon separated
1020 list, not as a single string, and sign the message with each element,
1021 omitting multiple occurences of the same signer.
1023 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1024 Bugzilla 985, 986. Patch by Simon Arlott
1026 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1027 Patch by Simon Arlott
1029 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1030 MBX locking. Notification from Dan Rosenberg.
1036 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1038 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1040 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1042 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1044 NM/04 Bugzilla 915: Flag broken perl installation during build.
1050 TK/01 Added patch by Johannes Berg that expands the main option
1051 "spamd_address" if it starts with a dollar sign.
1053 TK/02 Write list of recipients to X-Envelope-Sender header when building
1054 the mbox-format spool file for content scanning (suggested by Jakob
1057 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1058 (http://www.dcc-servers.net/) support via dccifd. Activated by
1059 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1061 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1062 by Mark Daniel Reidel <mr@df.eu>.
1064 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1065 When building exim an external PCRE library is now needed -
1066 PCRE is a system library on the majority of modern systems.
1067 See entry on PCRE_LIBS in EDITME file.
1069 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1070 conversation. Added nologin parameter to request.
1071 Patch contributed by Kirill Miazine.
1073 TF/01 Do not log submission mode rewrites if they do not change the address.
1075 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1077 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1078 log files in place. Contributed by Roberto Lima.
1080 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1082 TF/03 Bugzilla 615: When checking the local_parts router precondition
1083 after a local_part_suffix or local_part_prefix option, Exim now
1084 does not use the address's named list lookup cache, since this
1085 contains cached lookups for the whole local part.
1087 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1088 Robert Millan. Documentation is in experimental-spec.txt.
1090 TF/04 Bugzilla 668: Fix parallel build (make -j).
1092 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1094 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1095 Patch provided by Jan Srzednicki.
1097 TF/05 Leading white space used to be stripped from $spam_report which
1098 wrecked the formatting. Now it is preserved.
1100 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1101 that they are available at delivery time.
1103 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1105 TF/08 TLS error reporting now respects the incoming_interface and
1106 incoming_port log selectors.
1108 TF/09 Produce a more useful error message if an SMTP transport's hosts
1109 setting expands to an empty string.
1111 NM/06 Bugzilla 744: EXPN did not work under TLS.
1112 Patch provided by Phil Pennock.
1114 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1115 Patch provided by Richard Godbee.
1117 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1118 acl_smtp_notquit, added index entry.
1120 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1121 Patch provided by Eugene Bujak.
1123 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1124 accept(). Patch provided by Maxim Dounin.
1126 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1127 Patch provided by Phil Pennock.
1129 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1131 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1132 Patch provided by Brad "anomie" Jorsch.
1134 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1135 Patch provided by Dean Brooks.
1137 TK/05 Add native DKIM support (does not depend on external libraries).
1139 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1140 Patch provided by Graeme Fowler.
1142 NM/16 Bugzilla 851: Documentation example syntax fix.
1144 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1146 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1149 NM/19 Bugzilla 745: TLS version reporting.
1150 Patch provided by Phil Pennock.
1152 NM/20 Bugzilla 167: bool: condition support.
1153 Patch provided by Phil Pennock.
1155 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1156 clients. Patch provided by Phil Pennock.
1158 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1159 Patch provided by Brad "anomie" Jorsch.
1161 NM/23 Bugzilla 687: Fix misparses in eximstats.
1162 Patch provided by Heiko Schlittermann.
1164 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1165 Patch provided by Heiko Schlittermann.
1167 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1168 plus update to original patch.
1170 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1172 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1173 Patch provided by David Brownlee.
1175 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1177 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1179 NM/30 Bugzilla 888: TLS documentation bugfixes.
1181 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1183 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1184 Unlike the original bugzilla I have changed all shell scripts in src tree.
1186 NM/33 Bugzilla 898: Transport filter timeout fix.
1187 Patch by Todd Rinaldo.
1189 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1190 Patch by Serge Demonchaux.
1192 NM/35 Bugzilla 39: Base64 decode bug fixes.
1193 Patch by Jakob Hirsch.
1195 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1197 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1199 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1201 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1207 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1208 ALT-N's libdkim that I have put here:
1209 http://duncanthrax.net/exim-experimental/
1211 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1212 were called 'true' and it seems that C99 defines that as a reserved
1213 keyword to be used with 'bool' variable types. That means you could
1214 not include C99-style headers which use bools without triggering
1215 build errors in sieve.c.
1217 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1218 as mailq or other aliases. Changed the --help handling significantly
1219 to do whats expected. exim_usage() emits usage/help information.
1221 SC/01 Added the -bylocaldomain option to eximstats.
1223 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1225 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1227 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1233 PH/01 Another patch from the Sieve maintainer.
1235 PH/02 When an IPv6 address is converted to a string for single-key lookup
1236 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1237 dots are used instead of colons so that keys in lsearch files need not
1238 contain colons. This was done some time before quoting was made available
1239 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1240 (notated using the quote facility) so as to distinguish them from IPv4
1241 keys. This meant that lookups for IP addresses in host lists did not work
1242 for iplsearch lookups.
1244 This has been fixed by arranging for IPv6 addresses to be expressed with
1245 colons if the lookup type is iplsearch. This is not incompatible, because
1246 previously such lookups could never work.
1248 The situation is now rather anomolous, since one *can* have colons in
1249 ordinary lsearch keys. However, making the change in all cases is
1250 incompatible and would probably break a number of configurations.
1252 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1255 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1256 conversion specification without a maximum field width, thereby enabling
1257 a rogue spamd server to cause a buffer overflow. While nobody in their
1258 right mind would setup Exim to query an untrusted spamd server, an
1259 attacker that gains access to a server running spamd could potentially
1260 exploit this vulnerability to run arbitrary code as the Exim user.
1262 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1263 $primary_hostname instead of what libspf2 thinks the hosts name is.
1265 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1266 a directory entry by the name of the lookup key. Previously, if a
1267 symlink pointed to a non-existing file or a file in a directory that
1268 Exim lacked permissions to read, a lookup for a key matching that
1269 symlink would fail. Now it is enough that a matching directory entry
1270 exists, symlink or not. (Bugzilla 503.)
1272 PH/03 The body_linecount and body_zerocount variables are now exported in the
1275 PH/04 Added the $dnslist_matched variable.
1277 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1278 This means they are set thereafter only if the connection becomes
1281 PH/06 Added the client_condition to authenticators so that some can be skipped
1282 by clients under certain conditions.
1284 PH/07 The error message for a badly-placed control=no_multiline_responses left
1285 "_responses" off the end of the name.
1287 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1289 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1290 (without spaces) instead of just copying the configuration text.
1292 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1294 PH/11 Added $max_received_linelength.
1296 PH/12 Added +ignore_defer and +include_defer to host lists.
1298 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1299 way in which PCRE > 7.0 is built.
1301 PH/14 Implemented queue_only_load_latch.
1303 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1304 MAIL command. The effect was to mangle the value on 64-bit systems.
1306 PH/16 Another patch from the Sieve maintainer.
1308 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1310 PH/18 If a system quota error occurred while trying to create the file for
1311 a maildir delivery, the message "Mailbox is full" was not appended to the
1312 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1313 only to a quota excession during the actual writing of the file.
1315 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1316 characters?) which causes problems in log lines. The DN values are now
1317 passed through string_printing() before being added to log lines.
1319 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1320 and InterBase are left for another time.)
1322 PH/21 Added message_body_newlines option.
1324 PH/22 Guard against possible overflow in moan_check_errorcopy().
1326 PH/23 POSIX allows open() to be a macro; guard against that.
1328 PH/24 If the recipient of an error message contained an @ in the local part
1329 (suitably quoted, of course), incorrect values were put in $domain and
1330 $local_part during the evaluation of errors_copy.
1336 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1337 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1340 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1341 issue a MAIL command.
1343 PH/02 In an ACL statement such as
1345 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1347 if a client was not listed at all, or was listed with a value other than
1348 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1349 the condition was not true (as it should be), so access was not denied.
1350 The bug was that the ! inversion was incorrectly passed on to the second
1351 item. This has been fixed.
1353 PH/03 Added additional dnslists conditions == and =& which are different from
1354 = and & when the dns lookup returns more than one IP address.
1356 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1357 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1359 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1360 FSYNC, which compiles an option called disable_fsync that allows for
1361 bypassing fsync(). The documentation is heavily laced with warnings.
1363 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1365 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1366 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1367 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1368 including adding "make clean"; (3) Added -fPIC when compiling the test
1369 dynamically loaded module, to get rid of a warning.
1371 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1372 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1373 = 0s. The bug is otherwise harmless.
1375 PH/07 There was a bug in the dovecot authenticator such that the value of
1376 $auth1 could be overwritten, and so not correctly preserved, after a
1377 successful authentication. This usually meant that the value preserved by
1378 the server_setid option was incorrect.
1380 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1382 PH/09 Installed PCRE release 7.0.
1384 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1385 run for batched SMTP input. It is now run at the start of every message
1386 in the batch. While fixing this I discovered that the process information
1387 (output by running exiwhat) was not always getting set for -bs and -bS
1388 input. This is fixed, and it now also says "batched" for BSMTP.
1390 PH/11 Added control=no_pipelining.
1392 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1393 patch, slightly modified), and move the expansion of helo_data till after
1394 the connection is made in the smtp transport (so it can use these
1397 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1399 PH/14 Added log_selector = +pid.
1401 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1403 PH/16 Add ${if forany and ${if forall.
1405 PH/17 Added dsn_from option to vary the From: line in DSNs.
1407 PH/18 Flush SMTP output before performing a callout, unless control =
1408 no_callout_flush is set.
1410 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1411 was true (the default) a successful delivery failed to delete the retry
1412 item, thus causing premature timeout of the address. The bug is now
1415 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1417 PH/21 Long custom messages for fakedefer and fakereject are now split up
1418 into multiline reponses in the same way that messages for "deny" and
1419 other ACL rejections are.
1421 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1422 with slight modification.
1424 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1425 draft, changing the syntax and factoring some duplicate code".
1427 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1428 for deliveries of the second and subsequent messages over the same SMTP
1431 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1432 ${reduce, with only minor "tidies".
1434 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1436 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1437 expansion side effects.
1439 PH/27 When a message times out after an over-quota error from an Exim-imposed
1440 quota, the bounce message says "mailbox is full". This message was not
1441 being given when it was a system quota that was exceeded. It now should
1444 MH/03 Made $recipients available in local_scan(). local_scan() already has
1445 better access to the recipient list through recipients_list[], but
1446 $recipients can be useful in postmaster-provided expansion strings.
1448 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1449 in the case of a MAIL command with additional options following the
1450 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1451 were accidentally chopped off.
1453 PH/29 SMTP synchronization checks are implemented when a command is read -
1454 there is a check that no more input is waiting when there shouldn't be
1455 any. However, for some commands, a delay in an ACL can mean that it is
1456 some time before the response is written. In this time, more input might
1457 arrive, invalidly. So now there are extra checks after an ACL has run for
1458 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1459 pipelining has not been advertised.
1461 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1463 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1464 returned in the SMTP response when smtp_return_error_details was false.
1465 This has been fixed.
1467 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1468 instead of the C I/O that was originally supplied, because problems were
1469 reported on Solaris.
1471 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1472 Exim which did not show up earlier: it was assuming that a call to
1473 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1474 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1475 was a macro that became an assignment, so it seemed to work. This has
1476 changed to a proper function call with a void return, hence the compile
1477 error. Exim's code has been fixed.
1479 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1482 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1484 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1486 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1487 because it turns out to be too big or there is a timeout) there is no
1488 "Completed" line in the log. When some messages of this type were
1489 selected by exigrep, they were listed as "not completed". Others were
1490 picked up by some special patterns. I have improved the selection
1491 criteria to be more general.
1493 PH/38 The host_find_failed option in the manualroute router can now be set
1494 to "ignore", to completely ignore a host whose IP address cannot be
1495 found. If all hosts are ignored, the behaviour is controlled by the new
1496 host_all_ignored option.
1498 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1499 homing or because of multiple MX records with /mx) generated more than
1500 one IP address, and the following item turned out to be the local host,
1501 all the secondary addresses of the first item were incorrectly removed
1502 from the list, along with the local host and any following hosts (which
1503 is what is supposed to happen).
1505 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1506 whoever called Exim into the -H file. In the case of the daemon it was
1507 behaving confusingly. When first started, it used values for whoever
1508 started the daemon, but after a SIGHUP it used the Exim user (because it
1509 calls itself on a restart). I have changed the code so that it now always
1512 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1513 message are rejected with the same error (e.g. no authentication or bad
1514 sender address), and a DATA command is nevertheless sent (as can happen
1515 with PIPELINING or a stupid MUA), the error message that was given to the
1516 RCPT commands is included in the rejection of the DATA command. This is
1517 intended to be helpful for MUAs that show only the final error to their
1520 PH/42 Another patch from the Sieve maintainer.
1522 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1523 Eximstats - Fixed some broken HTML links and added missing column headers
1525 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1526 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1528 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1530 PH/43 Yet another patch from the Sieve maintainer.
1532 PH/44 I found a way to check for a TCP/IP connection going away before sending
1533 the response to the final '.' that terminates a message, but only in the
1534 case where the client has not sent further data following the '.'
1535 (unfortunately, this is allowed). However, in many cases there won't be
1536 any further data because there won't be any more messages to send. A call
1537 to select() can be used: if it shows that the input is "ready", there is
1538 either input waiting, or the socket has been closed. An attempt to read
1539 the next input character can distinguish the two cases. Previously, Exim
1540 would have sent an OK response which the client would never have see.
1541 This could lead to message repetition. This fix should cure that, at
1542 least in a lot of common cases.
1544 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1545 advertised in response to EHLO.
1551 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1552 fixed by 4.65/MH/01 (is this a record?) are fixed:
1554 (i) An empty string was always treated as zero by the numeric comparison
1555 operators. This behaviour has been restored.
1557 (ii) It is documented that the numeric comparison operators always treat
1558 their arguments as decimal numbers. This was broken in that numbers
1559 starting with 0 were being interpreted as octal.
1561 While fixing these problems I realized that there was another issue that
1562 hadn't been noticed. Values of message_size_limit (both the global option
1563 and the transport option) were treated as octal if they started with 0.
1564 The documentation was vague. These values are now always treated as
1565 decimal, and I will make that clear in the documentation.
1571 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1572 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1575 MH/01 Don't check that the operands of numeric comparison operators are
1576 integers when their expansion is in "skipping" mode (fixes bug
1577 introduced by 4.64-PH/07).
1579 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1580 child addresses, Exim now panics and dies. Previously, because the count
1581 is held in a short int, deliveries were likely to be lost. As such a
1582 large number of recipients for a single message is ridiculous
1583 (performance will be very, very poor), I have chosen to impose a limit
1584 rather than extend the field.
1590 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1591 leftover -K file (the existence of which was triggered by #402).
1592 While we were at it, introduced process PID as part of the -K
1593 filename. This should rule out race conditions when creating
1596 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1597 processing considerably. Previous code took too long for large mails,
1598 triggering a timeout which in turn triggers #401.
1600 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1601 in the DK code in transports.c. sendfile() is not really portable,
1602 hence the _LINUX specificness.
1604 TF/01 In the add_headers option to the mail command in an Exim filter,
1605 there was a bug that Exim would claim a syntax error in any
1606 header after the first one which had an odd number of characters
1609 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1610 callout verification, Exim cached a "reject" for the entire domain. This
1611 is correct for most verifications, but it is not correct for a recipient
1612 verification with use_sender or use_postmaster set, because in that case
1613 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1614 case of MAIL FROM:<> rejection from other early rejections (e.g.
1615 rejection of HELO). When verifying a recipient using a non-null MAIL
1616 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1617 Whatever the result of the callout, the value of the domain cache is
1618 left unchanged (for any other kind of callout, getting as far as trying
1619 RCPT means that the domain itself is ok).
1621 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1624 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1625 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1626 session, the daemon ignores ECONNECT errors and logs others; it now
1627 ignores EPIPE as well.
1629 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1630 (quoted-printable decoding).
1632 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1633 later the small subsequent patch to fix an introduced bug.
1635 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1637 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1639 PH/08 An error is now given if message_size_limit is specified negative.
1641 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1642 to be given (somewhat) arbitrary names.
1644 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1647 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1648 miscellaneous code fixes
1650 PH/10 Added the log_reject_target ACL modifier to specify where to log
1653 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1654 hostname. This is wrong, because it relates to the incoming message (and
1655 probably the interface on which it is arriving) and not to the outgoing
1656 callout (which could be using a different interface). This has been
1657 changed to use the value of the helo_data option from the smtp transport
1658 instead - this is what is used when a message is actually being sent. If
1659 there is no remote transport (possible with a router that sets up host
1660 addresses), $smtp_active_hostname is used.
1662 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1663 tweaks were necessary in order to get it to work (see also 21 below):
1664 (a) The code assumed that strncpy() returns a negative number on buffer
1665 overflow, which isn't the case. Replaced with Exim's string_format()
1667 (b) There were several signed/unsigned issues. I just did the minimum
1668 hacking in of casts. There is scope for a larger refactoring.
1669 (c) The code used strcasecmp() which is not a standard C function.
1670 Replaced with Exim's strcmpic() function.
1671 (d) The code set only $1; it now sets $auth1 as well.
1672 (e) A simple test gave the error "authentication client didn't specify
1673 service in request". It would seem that Dovecot has changed its
1674 interface. Fortunately there's a specification; I followed it and
1675 changed what the client sends and it appears to be working now.
1677 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1680 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1681 address A is aliased to B and C, where B exists and C does not. Without
1682 -v the output is "A verified" because verification stops after a
1683 successful redirection if more than one address is generated. However,
1684 with -v the child addresses are also verified. Exim was outputting "A
1685 failed to verify" and then showing the successful verification for C,
1686 with its parentage. It now outputs "B failed to verify", showing B's
1687 parentage before showing the successful verification of C.
1689 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1690 look up a TXT record in a specific list after matching in a combined
1693 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1694 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1695 they consult the DNS. I had assumed they would set it the way they
1696 wanted; and indeed my experiments on Linux seem to show that in some
1697 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1698 To be on the safe side, however, I have now made the interface to
1699 host_find_byname() similar to host_find_bydns(), with an argument
1700 containing the DNS resolver options. The host_find_byname() function now
1701 sets these options at its start, just as host_find_bydns() does. The smtp
1702 transport options dns_qualify_single and dns_search_parents are passed to
1703 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1704 of host_find_byname() use the default settings of RES_DEFNAMES
1705 (qualify_single) but not RES_DNSRCH (search_parents).
1707 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1708 spool_read_header() do less string testing, by means of a preliminary
1709 switch on the second character of optional "-foo" lines. (This is
1710 overdue, caused by the large number of possibilities that now exist.
1711 Originally there were few.) While I was there, I also converted the
1712 str(n)cmp tests so they don't re-test the leading "-" and the first
1713 character, in the hope this might squeeze out yet more improvement.
1715 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1716 flag allowing group syntax was set by the header_syntax check but not
1717 turned off, possible causing trouble later; (2) The flag was not being
1718 set at all for the header_verify test, causing "group"-style headers to
1719 be rejected. I have now set it in this case, and also caused header_
1720 verify to ignore an empty address taken from a group. While doing this, I
1721 came across some other cases where the code for allowing group syntax
1722 while scanning a header line wasn't quite right (mostly, not resetting
1723 the flag correctly in the right place). These bugs could have caused
1724 trouble for malformed header lines. I hope it is now all correct.
1726 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1727 with the "reply" argument non-NULL. The code, however (which originally
1728 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1729 but it didn't always do it. This confused somebody who was copying the
1730 code for some other use. I have removed all the tests.
1732 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1733 feature that was used to support insecure browsers during the U.S. crypto
1734 embargo. It requires special client support, and Exim is probably the
1735 only MTA that supported it -- and would never use it because real RSA is
1736 always available. This code has been removed, because it had the bad
1737 effect of slowing Exim down by computing (never used) parameters for the
1738 RSA_EXPORT functionality.
1740 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1741 authenticator to fail if there's a tab character in the incoming data
1742 (there should never be unless someone is messing about, as it's supposed
1743 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1744 if the connection is using TLS or if the remote IP is the same as the
1745 local IP, and the "valid-client-cert option" if a client certificate has
1748 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1749 authenticators. This can be used for authorization after authentication
1750 succeeds. (In the case of plaintext, it servers for both authentication
1753 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1754 if any retry times were supplied.
1756 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1757 connection, where there is no client IP address to check. In this
1758 situation, the verify now always succeeds.
1760 PH/25 Applied John Jetmore's -Mset patch.
1762 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1764 PH/27 In a string expansion for a processed (not raw) header when multiple
1765 headers of the same name were present, leading whitespace was being
1766 removed from all of them, but trailing whitespace was being removed only
1767 from the last one. Now trailing whitespace is removed from each header
1768 before concatenation. Completely empty headers in a concatenation (as
1769 before) are ignored.
1771 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1772 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1774 PH/29 [Removed. This was a change that I later backed out, and forgot to
1775 correct the ChangeLog entry (that I had efficiently created) before
1776 committing the later change.]
1778 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1779 address errors (4xx response to RCPT) over the same connection as other
1780 messages routed to the same hosts. Such deliveries are always "forced",
1781 so retry times are not inspected. This resulted in far too many retries
1782 for the affected addresses. The effect occurred only when there were more
1783 hosts than the hosts_max_try setting in the smtp transport when it had
1784 the 4xx errors. Those hosts that it had tried were not added to the list
1785 of hosts for which the message was waiting, so if all were tried, there
1786 was no problem. Two fixes have been applied:
1788 (i) If there are any address or message errors in an SMTP delivery, none
1789 of the hosts (tried or untried) are now added to the list of hosts
1790 for which the message is waiting, so the message should not be a
1791 candidate for sending over the same connection that was used for a
1792 successful delivery of some other message. This seems entirely
1793 reasonable: after all the message is NOT "waiting for some host".
1794 This is so "obvious" that I'm not sure why it wasn't done
1795 previously. Hope I haven't missed anything, but it can't do any
1796 harm, as the worst effect is to miss an optimization.
1798 (ii) If, despite (i), such a delivery is accidentally attempted, the
1799 routing retry time is respected, so at least it doesn't keep
1800 hammering the server.
1802 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1803 in ${readsocket because some servers need this prod.
1805 PH/32 Added some extra debug output when updating a wait-xxx database.
1807 PH/33 The hint "could be header name not terminated by colon", which has been
1808 given for certain expansion errors for a long time, was not being given
1809 for the ${if def:h_colon_omitted{... case.
1811 PH/34 The spec says: "With one important exception, whenever a domain list is
1812 being scanned, $domain contains the subject domain." There was at least
1813 one case where this was not true.
1815 PH/35 The error "getsockname() failed: connection reset by peer" was being
1816 written to the panic log as well as the main log, but it isn't really
1817 panic-worthy as it just means the connection died rather early on. I have
1818 removed the panic log writing for the ECONNRESET error when getsockname()
1821 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1822 runs only) independently of the message's sender address. This meant
1823 that, if the 4xx error was in fact related to the sender, a different
1824 message to the same recipient with a different sender could confuse
1825 things. In particualar, this can happen when sending to a greylisting
1826 server, but other circumstances could also provoke similar problems.
1827 I have changed the default so that the retry time for these errors is now
1828 based a combination of the sender and recipient addresses. This change
1829 can be overridden by setting address_retry_include_sender=false in the
1832 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1833 remote server are returned as part of bounce messages. This was not
1834 happening for LMTP over a pipe (the lmtp transport), but now it is the
1835 same for both kinds of LMTP.
1837 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1838 in header lines that were in fact CNAMEs. This is no longer the case.
1840 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1841 and queue runs started by the daemon processed all messages. This has
1842 been fixed so that -R and -S can now usefully be given with -q<time>.
1844 PH/40 Import PCRE release 6.7 (fixes some bugs).
1846 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1848 PH/42 Give an error if -q is specified more than once.
1850 PH/43 Renamed the variables $interface_address and $interface_port as
1851 $received_ip_address and $received_port, to make it clear that these
1852 values apply to message reception, and not to the outgoing interface when
1853 a message is delivered. (The old names remain recognized, of course.)
1855 PH/44 There was no timeout on the connect() call when using a Unix domain
1856 socket in the ${readsocket expansion. There now is.
1858 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1859 be meaningful with "accept".
1861 SC/01 Eximstats V1.43
1862 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1864 SC/02 Eximstats V1.44
1865 Use a glob alias rather than an array ref in the generated
1866 parser. This improves both readability and performance.
1868 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1869 Collect SpamAssassin and rejection statistics.
1870 Don't display local sender or destination tables unless
1871 there is data to show.
1872 Added average volumes into the top table text output.
1874 SC/04 Eximstats V1.46
1875 Collect data on the number of addresses (recipients)
1876 as well as the number of messages.
1878 SC/05 Eximstats V1.47
1879 Added 'Message too big' to the list of mail rejection
1880 reasons (thanks to Marco Gaiarin).
1882 SC/06 Eximstats V1.48
1883 Mainlog lines which have GMT offsets and are too short to
1884 have a flag are now skipped.
1886 SC/07 Eximstats V1.49 (Alain Williams)
1887 Added the -emptyok flag.
1889 SC/08 Eximstats V1.50
1890 Fixes for obtaining the IP address from reject messages.
1892 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1893 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1894 whitesspace changes from 4.64-PH/27
1896 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1899 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1900 are found, allow negative numbers in numeric criteria)
1902 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1904 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1905 to match changes made in 4.64-PH/43
1907 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1909 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1910 "domain" is actually an IP address, and force a failure. This locks out
1911 those revolvers/nameservers that support "A-for-A" lookups, in
1912 contravention of the specifications.
1914 PH/48 When a host name was looked up from an IP address, and the subsequent
1915 forward lookup of the name timed out, the host name was left in
1916 $sender_host_name, contrary to the specification.
1918 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1919 restricted to single-key lookups, Exim was not diagnosing an error if
1920 * or *@ was used with a query-style lookup.
1922 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1924 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1925 long ago, but noone interested enough thought of it. Let's just say that
1926 the "1.1" means that there are some new functions that weren't there at
1927 some point in the past.
1929 PH/51 Error processing for expansion failure of helo_data from an smtp
1930 transport during callout processing was broken.
1932 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1933 tested/used via the -bh/-bhc/-bs options.
1935 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1936 bug, fixed in subsequent PCRE releases).
1938 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1939 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1941 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1947 SC/01 Use a glob alias rather than an array ref in eximstats generated
1948 parser. This improves both readability and performance.
1950 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1951 Don't display local sender or destination tables in eximstats unless
1952 there is data to show.
1953 Added average volumes into the eximstats top table text output.
1955 SC/03 Collect data on the number of addresses (recipients) as well
1956 as the number of messages in eximstats.
1958 TF/01 Correct an error in the documentation for the redirect router. Exim
1959 does (usually) call initgroups() when daemonizing.
1961 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1962 with consistent privilege compared to when running as a daemon.
1964 TF/03 Note in the spec that $authenticated_id is not set for local
1965 submissions from trusted users.
1967 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1968 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1970 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1971 by adding some example configuration directives to the default
1972 configuration file. A little bit of work is required to uncomment the
1973 directives and define how usernames and passwords are checked, but
1974 there is now a framework to start from.
1976 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1977 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1978 without this. I don't know how relevant this is to other LDAP libraries.
1980 PH/02 Add the verb name to the "unknown ACL verb" error.
1982 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1984 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1986 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1987 directory not expanded when it should be if an expanded home directory
1988 was set for the address (which is overridden by the transport).
1990 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1993 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1994 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1995 because it is too late at that time, and has no effect.
1997 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1998 security issue with \' (bugzilla #107). I could not use the
1999 PQescapeStringConn() function, because it needs a PGconn value as one of
2002 PH/08 When testing addresses using -bt, indicate those final addresses that
2003 are duplicates that would not cause an additional delivery. At least one
2004 person was confused, thinking that -bt output corresponded to deliveries.
2005 (Suppressing duplicates isn't a good idea as you lose the information
2006 about possibly different redirections that led to the duplicates.)
2008 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2009 systems where poll() doesn't work, in particular OS X.
2011 PH/10 Added more information to debugging output for retry time not reached.
2013 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2014 operations in malware.c.
2016 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2019 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2020 syslog_duplication set false, log lines that would normally be written
2021 both the the main log and to the reject log were not written to syslog at
2024 PH/14 In the default configuration, change the use of "message" in ACL warn
2025 statements to "add_header".
2027 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2028 not followed by a command (e.g. "seen endif").
2030 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2031 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2034 PH/17 Added extra conditions to the default value of delay_warning_condition
2038 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2039 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2040 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2043 The Auto-Submitted: and various List- headers are standardised, whereas I
2044 don't think Precedence: ever was.
2046 PH/18 Refactored debugging code in route_finduser() to show more information,
2047 in particular, the error code if getpwnam() issues one.
2049 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2050 This is apparently needed in addition to the PH/07 change above to avoid
2051 any possible encoding problems.
2053 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2054 but not after initializing Perl.
2056 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2057 output them only if debugging. By default they are written stderr,
2058 apparently, which is not desirable.
2060 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2063 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2066 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2068 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2069 authenticated or an ident call has been made. Suppress the default
2070 values for $authenticated_id and $authenticated_sender (but permit -oMai
2071 and -oMas) when testing with -bh.
2073 PH/24 Re-jigged the order of the tests in the default configuration so that the
2074 tests for valid domains and recipients precede the DNS black list and CSA
2075 tests, on the grounds that those ones are more expensive.
2077 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2078 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2079 This bug exists in every version of Exim that I still have, right back to
2082 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2083 However, an attempt to turn on case-sensitivity in a regex key by
2084 including (?-i) didn't work because the subject string was already
2085 lowercased, and the effects were non-intuitive. It turns out that a
2086 one-line patch can be used to allow (?-i) to work as expected.
2092 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2093 other effects) broke the use of negated acl sub-conditions.
2095 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2098 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2099 "Deny" causes Exim to reject the incoming connection with a 554 error.
2100 Unfortunately, if there is a major crisis, such as a disk failure,
2101 tcp-wrappers gives "deny", whereas what one would like would be some
2102 kind of temporary error. A kludge has been added to help with this.
2103 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2104 554 error is used if errno is still zero or contains ENOENT (which occurs
2105 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2108 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2110 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2111 errors. Otherwise a message that provokes a temporary error (when other
2112 messages do not) can cause a whole host to time out.
2114 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2115 addresses were routed directly to files or pipes from a redirect router.
2116 File deliveries just didn't batch; pipe deliveries might have suffered
2119 PH/06 A failure to get a lock for a hints database would erroneously always say
2120 "Failed to get write lock", even when it was really a read lock.
2122 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2123 of 0600. This has been changed to use the value of the lockfile_mode
2124 option (which defaults to 0600).
2126 PH/08 Applied small patch from the Sieve maintainer.
2128 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2129 folder from quota calculations, a direct delivery into this folder messed
2130 up the contents of the maildirsize file. This was because the regex was
2131 used only to exclude .Trash (or whatever) when the size of the mailbox
2132 was calculated. There was no check that a delivery was happening into an
2133 excluded directory. This bug has been fixed by ignoring all quota
2134 processing for deliveries into excluded directories.
2136 PH/10 Added the maildirfolder_create_regex option to appendfile.
2142 PH/01 The code for finding all the local interface addresses on a FreeBSD
2143 system running IPv6 was broken. This may well have applied to all BSD
2144 systems, as well as to others that have similar system calls. The broken
2145 code found IPv4 interfaces correctly, but gave incorrect values for the
2146 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2147 that it would not match correctly against @[] and not recognize the IPv6
2150 PH/02 The ipliteral router was not recognizing addresses of the form user@
2151 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2153 PH/03 Added disable_ipv6.
2155 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2156 decoded form, because it is most often used to construct To: headers
2157 lines in autoreplies, and the decoded form may well be syntactically
2158 invalid. However, $reply_address has leading white space removed, and all
2159 newlines turned into spaces so that the autoreply transport does not
2162 PH/05 If group was specified without a user on a router, and no group or user
2163 was specified on a transport, the group from the router was ignored.
2165 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2166 for visible compile-time settings that can be used to change these
2167 numbers, for those that want even more. Backwards compatibility with old
2168 spool files has been maintained. However, going back to a previous Exim
2169 release will lost any variables that are in spool files.
2171 PH/07 Two small changes when running in the test harness: increase delay when
2172 passing a TCP/IP connection to a new process, in case the original
2173 process has to generate a bounce, and remove special handling of
2174 127.0.0.2 (sic), which is no longer necessary.
2176 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2177 be the same on different OS.
2179 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2182 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2183 whether --show-vars was specified or not
2185 JJ/02 exipick: Added support for new ACL variable spool format introduced
2188 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2189 syntactically invalid From: or Reply-to: line, and a filter used this to
2190 generate an autoreply, and therefore failed to obtain an address for the
2191 autoreply, Exim could try to deliver to a non-existent relative file
2192 name, causing unrelated and misleading errors. What now happens is that
2193 it logs this as a hard delivery error, but does not attempt to create a
2196 PH/11 The exinext utility has a -C option for testing purposes, but although
2197 the given file was scanned by exinext itself; it wasn't being passed on
2198 when Exim was called.
2200 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2201 an end-of-file indication when reading a command response.
2203 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2204 compiled. In many other places in Exim, IPv6 addresses are always
2205 recognized, so I have changed this. It also means that IPv4 domain
2206 literals of the form [IPV4:n.n.n.n] are now always recognized.
2208 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2209 used if the router is not running as root, for example, when verifying at
2210 ACL time, or when using -bh. The debugging output from this situation was
2211 non-existent - all you got was a failure to exec. I have made two
2214 (a) Failures to set uid/gid, the current directory, or a process leader
2215 in a subprocess such as that created by queryprogram now generate
2216 suitable debugging ouput when -d is set.
2218 (b) The queryprogram router detects when it is not running as root,
2219 outputs suitable debugging information if -d is set, and then runs
2220 the subprocess without attempting to change uid/gid.
2222 PH/15 Minor change to Makefile for building test_host (undocumented testing
2225 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2226 additional section of a DNS packet that returns MX or SRV records.
2227 Instead, it always explicitly searches for A/AAAA records. This avoids
2228 major problems that occur when a DNS server includes only records of one
2229 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2230 fixed another bug: if SRV records were looked up and the corresponding
2231 address records were *not* found in the additional section, the port
2232 values from the SRV records were lost.
2234 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2235 using the correct key (the original address) when searching the retry
2236 rules in order to find which one to use for generating the retry hint.
2238 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2239 adding the default one. Similarly, if it contains a Reply-To: header, the
2240 errors_reply_to option, if set, is not used.
2242 PH/19 When calculating a retry time, Exim used to measure the "time since
2243 failure" by looking at the "first failed" field in the retry record. Now
2244 it does not use this if it is later than than the arrival time of the
2245 message. Instead it uses the arrival time. This makes for better
2246 behaviour in cases where some deliveries succeed, thus re-setting the
2247 "first failed" field. An example is a quota failure for a huge message
2248 when small messages continue to be delivered. Without this change, the
2249 "time since failure" will always be short, possible causing more frequent
2250 delivery attempts for the huge message than are intended.
2251 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2253 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2254 $1, $2, $3) because the numerical variables can be reset during some
2255 expansion items (e.g. "match"), thereby losing the authentication data.
2257 PH/21 Make -bV show the size of off_t variables so that the test suite can
2258 decide whether to run tests for quotas > 2G.
2260 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2261 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2262 filecount value is greater than 2G or if a quota value is greater than 2G
2263 on a system where the size of off_t is not greater than 4, a panic error
2266 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2267 never match. The debug and -bh output now contains an explicit error
2268 message indicating a malformed IPv4 address or mask.
2270 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2271 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2272 PH/23 above applies.
2274 PH/25 Do not write to syslog when running in the test harness. The only
2275 occasion when this arises is a failure to open the main or panic logs
2276 (for which there is an explicit test).
2278 PH/26 Added the /no_tell option to "control=freeze".
2280 PH/27 If a host name lookup failed very early in a connection, for example, if
2281 the IP address matched host_lookup and the reverse lookup yielded a name
2282 that did not have a forward lookup, an error message of the form "no IP
2283 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2284 could be logged. Now it outputs the IP address instead of "NULL".
2286 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2287 allows the sender and the authenticated sender to be set when
2288 submitting a message from within Exim. Since child_open_exim() is
2289 documented for local_scan(), the new function should be too.
2291 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2292 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2293 results in an empty string is now treated as unset.
2295 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2297 PH/31 Added sender_verify_fail logging option.
2299 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2300 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2301 tidied the source and removed it altogether.
2303 PH/33 When a queue run was abandoned because the load average was too high, a
2304 log line was always written; now it is written only if the queue_run log
2305 selector is set. In addition, the log line for abandonment now contains
2306 information about the queue run such as the pid. This is always present
2307 in "start" and "stop" lines but was omitted from the "abandon" line.
2309 PH/34 Omit spaces between a header name and the colon in the error message that
2310 is given when verify = headers_syntax fails (if there are lots of them,
2311 the message gets confusing).
2313 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2314 names, as there are now some PTR records that contain slashes. This check
2315 is only to protect against broken name servers that fall over on strange
2316 characters, so the fact that it applies to all lookups doesn't matter.
2318 PH/36 Now that the new test suite is complete, we can remove some of the
2319 special code in Exim that was needed for the old test suite. For example,
2320 sorting DNS records because real resolvers return them in an arbitrary
2321 order. The new test suite's fake resolver always returns records in the
2324 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2325 bounces) except when queue_only is set, to avoid logging races between
2326 the different processes.
2328 PH/38 Panic-die if .include specifies a non-absolute path.
2330 PH/39 A tweak to the "H" retry rule from its user.
2332 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2333 a label. They prevented compilation on older perls.
2335 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2336 a warning to be raised on newish perls.
2338 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2339 on queue. Changes to match documented behaviour of showing count of
2340 messages matching specified criteria.
2342 PH/40 Changed the default ident timeout from 30s to 5s.
2344 PH/41 Added support for the use of login_cap features, on those BSD systems
2345 that have them, for controlling the resources used by pipe deliveries.
2347 PH/42 The content-scanning code uses fopen() to create files in which to put
2348 message data. Previously it was not paying any attention to the mode of
2349 the files. Exim runs with umask(0) because the rest of the code creates
2350 files with open(), and sets the required mode explicitly. Thus, these
2351 files were ending up world-writeable. This was not a big issue, because,
2352 being within the spool directory, they were not world-accessible. I have
2353 created a function called modefopen, which takes an additional mode
2354 argument. It sets umask(777), creates the file, chmods it to the required
2355 mode, then resets the umask. All the relevant calls to fopen() in the
2356 content scanning code have been changed to use this function.
2358 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2359 to 24 hours. This avoids potential overflow problems when processing G
2360 and H retry rules. I suspect nobody ever tinkers with this value.
2362 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2364 PH/45 When the plaintext authenticator is running as a client, the server's
2365 challenges are checked to ensure they are valid base64 strings. By
2366 default, the authentication attempt is cancelled if an invalid string is
2367 received. Setting client_ignore_invalid_base64 true ignores these errors.
2368 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2369 they are received. Thus, the responses can be made to depend on the
2370 challenges. If an invalid string is ignored, an empty string is placed in
2373 PH/46 Messages that are created by the autoreply transport now contains a
2374 References: header, in accordance with RFCs 2822 and 3834.
2376 PH/47 Added authenticated_sender_force to the smtp transport.
2378 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2380 PH/49 Installed latest patch from the Sieve maintainer.
2382 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2383 was also set, the appendfile transport was unnecessarily scanning a
2384 directory of message files (e.g. for maildir delivery) to find the count
2385 of files (along with the size), even though it did not need this
2386 information. It now does the scan only if it needs to find either the
2387 size of the count of files.
2389 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2391 PH/52 Two bugs concerned with error handling when the smtp transport is
2394 (i) Exim was not creating retry information for temporary errors given
2395 for individual recipients after the DATA command when the smtp transport
2396 was used in LMTP mode. This meant that they could be retried too
2397 frequently, and not timed out correctly.
2399 (ii) Exim was setting the flag that allows error details to be returned
2400 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2401 recipients that were returned after the DATA command.
2403 PH/53 This is related to PH/52, but is more general: for any failing address,
2404 when detailed error information was permitted to be returned to the
2405 sender, but the error was temporary, then after the final timeout, only
2406 "retry timeout exceeded" was returned. Now it returns the full error as
2407 well as "retry timeout exceeded".
2409 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2410 do this, and (what is worse) MTAs that accept it.
2412 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2413 will now be deprecated.
2415 PH/56 New os.c-cygwin from the Cygwin maintainer.
2417 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2418 formats (previously only available in exim formats via -bpr, -bpru,
2419 and -bpra. Now also available in native and exiqgrep formats)
2421 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2422 with very large, slow to parse queues
2424 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2426 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2428 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2429 responses to authentication challenges, though it was showing the
2430 challenges; (ii) I've removed the CR characters from the debug output for
2433 PH/58 Allow for the insertion of a newline as well as a space when a string
2434 is turned into more than one encoded-word during RFC 2047 encoding. The
2435 Sieve code now uses this.
2437 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2438 data_4xx, lost_connection, tls_required.
2440 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2441 message was being sent as an SMTP response.
2443 PH/61 Add -l and -k options to exicyclog.
2445 PH/62 When verifying, if an address was redirected to one new address, so that
2446 verification continued, and the new address failed or deferred after
2447 having set something in $address_data, the value of $address_data was not
2448 passed back to the ACL. This was different to the case when no
2449 redirection occurred. The value is now passed back in both cases.
2451 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2452 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2453 use login_cap.h, so on its own it isn't the distinguishing feature. The
2454 new name refers directly to the setclassresources() function.
2456 PH/65 Added configuration files for NetBSD3.
2458 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2460 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2461 is preferred over IPv4.
2463 PH/68 The bounce_return_message and bounce_return_body options were not being
2464 honoured for bounces generated during the reception of non-SMTP messages.
2465 In particular, this applied to messages rejected by the ACL. This bug has
2466 been fixed. However, if bounce_return_message is true and bounce_return_
2467 body is false, the headers that are returned for a non-SMTP message
2468 include only those that have been read before the error was detected.
2469 (In the case of an ACL rejection, they have all been read.)
2471 PH/69 The HTML version of the specification is now built in a directory called
2472 spec_html instead of spec.html, because the latter looks like a path with
2473 a MIME-type, and this confuses some software.
2475 PH/70 Catch two compiler warnings in sieve.c.
2477 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2478 function verify_get_ident() calls ip_connect() to connect a socket, but
2479 if the "connect()" function timed out, ip_connect() used to close the
2480 socket. However, verify_get_ident() also closes the socket later, and in
2481 between Exim writes to the log, which may get opened at this point. When
2482 the socket was closed in ip_connect(), the log could get the same file
2483 descriptor number as the socket. This naturally causes chaos. The fix is
2484 not to close the socket in ip_connect(); the socket should be closed by
2485 the function that creates it. There was only one place in the code where
2486 this was missing, in the iplookup router, which I don't think anybody now
2487 uses, but I've fixed it anyway.
2489 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2490 well as to direct DNS lookups. Otherwise the handling of names in host
2491 lists is inconsistent and therefore confusing.
2497 PH/01 Two changes to the default runtime configuration:
2499 (1) Move the checks for relay_from_hosts and authenticated clients from
2500 after to before the (commented out) DNS black list checks.
2502 (2) Add control=submission to the relay_from_hosts and authenticated
2503 clients checks, on the grounds that messages accepted by these
2504 statements are most likely to be submissions.
2506 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2508 (1) Generate an error if the third argument for the ${prvs expansion is
2511 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2514 (3) Reset the variables that are obtained from the first argument of
2515 ${prvscheck and used in the second argument before leaving the code,
2516 because their memory is reclaimed, so using them afterwards may do
2519 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2520 one (it's much easier than Tom thought :-).
2522 (5) Because of (4), we can now allow for the use of $prvscheck_result
2523 inside the third argument.
2525 PH/03 For some reason, the default setting of PATH when running a command from
2526 a pipe transport was just "/usr/bin". I have changed it to
2529 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2530 anything to be listed in the output from -bV.
2532 PH/05 When a filter generated an autoreply, the entire To: header line was
2533 quoted in the delivery log line, like this:
2535 => >A.N.Other <ano@some.domain> <original@ddress> ...
2537 This has been changed so that it extracts the operative address. There
2538 may be more than one such address. If so, they are comma-separated, like
2541 => >ano@some.domain,ona@other.domain <original@ddress> ...
2543 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2544 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2545 not being looked up in the rDNS to get a host name, Exim was showing the
2546 IP address twice in Received: lines, even though the IP addresses were
2547 identical. For example:
2549 Received: from [1.2.3.4] (helo=[1.2.3.4])
2551 However, if the real host name was known, it was omitting the HELO data
2552 if it matched the actual IP address. This has been tidied up so that it
2553 doesn't show the same IP address twice.
2555 PH/07 When both +timestamp and +memory debugging was on, the value given by
2556 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2557 called by the debug printing, thereby overwriting the timestamp buffer.
2558 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2560 PH/08 When the original message was included in an autoreply transport, it
2561 always said "this is a copy of the message, including all the headers",
2562 even if body_only or headers_only was set. It now gives an appropriate
2565 PH/09 Applied a patch from the Sieve maintainer which:
2567 o fixes some comments
2568 o adds the (disabled) notify extension core
2569 o adds some debug output for the result of if/elsif tests
2570 o points to the current vacation draft in the documentation
2571 and documents the missing references header update
2575 o fixes a bug in processing the envelope test (when testing
2576 multiple envelope elements, the last element determinted the
2579 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2580 Electronic Mail") by including:
2582 Auto-submitted: auto-generated
2584 in the messages that it generates (bounce messages and others, such as
2585 warnings). In the case of bounce messages for non-SMTP mesages, there was
2586 also a typo: it was using "Auto_submitted" (underscore instead of
2587 hyphen). Since every message generated by Exim is necessarily in response
2588 to another message, thes have all been changed to:
2590 Auto-Submitted: auto-replied
2592 in accordance with these statements in the RFC:
2594 The auto-replied keyword:
2596 - SHOULD be used on messages sent in direct response to another
2597 message by an automatic process,
2599 - MUST NOT be used on manually-generated messages,
2601 - MAY be used on Delivery Status Notifications (DSNs) and Message
2602 Disposition Notifications (MDNs),
2604 - MUST NOT be used on messages generated by automatic or periodic
2605 processes, except for messages which are automatic responses to
2608 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2609 to the default Received: header definition.
2611 PH/12 Added log selector acl_warn_skipped (default on).
2613 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2614 variables because (a) they are in the wrong storage pool and (b) even if
2615 they were copied, it wouldn't work properly because of the caching.
2617 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2618 checking when decoding. Apparently there are clients that generate
2619 overlong encoded strings. Why am I not surprised?
2621 PH/15 If the first argument of "${if match_address" was not empty, but did not
2622 contain an "@" character, Exim crashed. Now it writes a panic log message
2623 and treats the condition as false.
2625 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2627 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2628 extension "envelope-auth". The code is finished and in agreement with
2629 other implementations, but there is no documentation so far and in fact,
2630 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2631 not changing the active code.
2633 Print executed "if" and "elsif" statements when debugging is used. This
2634 helps a great deal to understand what a filter does.
2636 Document more things not specified clearly in RFC3028. I had all this
2637 sorted out, when out of a sudden new issues came to my mind. Oops."
2639 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2642 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2643 canonical form (as documented). However, after a host name lookup from
2644 the IP address, check_host() was doing a simple string comparison with
2645 addresses acquired from the DNS when checking that the found name did
2646 have the original IP as one of its addresses. Since any found IPv6
2647 addresses are likely to be in abbreviated form, the comparison could
2648 fail. Luckily, there already exists a function for doing the comparison
2649 by converting both addresses to binary, so now that is used instead of
2650 the text comparison.
2652 PH/20 There was another similar case to PH/19, when a complete host name was
2653 given in a host list; looking up its IP address could give an abbreviated
2654 form, whereas the current host's name might or might not be abbreviated.
2655 The same fix has been applied.
2661 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2662 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2665 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2666 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2668 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2670 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2671 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2672 Also, for "Auto-Submitted", treat anything other than "no" as
2673 non-personal, in accordance with RFC 3834. (Previously it treated
2674 anything starting "auto-" as non-personal.)
2676 TF/01 The control=submission/name=... option had a problem with syntax
2677 errors if the name included a slash character. The /name= option
2678 now slurps the rest of the string, so it can include any characters
2679 but it must come last in the list of options (after /sender_retain
2682 PH/05 Some modifications to the interface to the fake nameserver for the new
2690 TK/01 Added the "success_on_redirect" address verification option. See
2691 NewStuff for rationale and an example.
2693 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2695 PH/02 Patch to exigrep to allow it to work on syslog lines.
2697 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2698 fread() to skip over the body file's header line, because in Cygwin the
2699 header line is locked and is inaccessible.
2701 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2702 co-exist for some time) to make it clear that it is the Exim ID that is
2703 referenced, not the Message-ID: header line.
2705 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2706 string_format() function, because snprintf() does not exist on all
2709 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2710 ${stat: expansion item.
2712 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2713 protocol synchronization error", to keep the pedants happy.
2715 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2716 well as for IRIX systems, when gcc is being used. See the host.c source
2719 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2721 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2724 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2725 transport and to the smtp transport in LMTP mode.
2727 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2729 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2730 run a filter in a subprocess. This could lead to confusion in subsequent
2731 lookups in the parent process. There should also be a search_tidyup() at
2732 the end of the subprocess.
2734 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2735 only if the host matched helo_try_verify_hosts, which caused the
2736 verification to occur when the EHLO/HELO command was issued. The ACL just
2737 tested the remembered result. Now, if a previous verification attempt has
2738 not happened, "verify = helo" does it there and then.
2740 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2742 TK/03 Fix log output including CR from clamd.
2744 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2745 exist provoked a memory error which could cause a segfault.
2747 PH/15 Installed PCRE 6.2
2749 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2751 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2752 of the problem. Specifically, suggested +O2 rather than +O1 for the
2755 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2757 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2758 record, the port was not correctly logged when the outgoing_port log
2759 selector was set (it logged the transort's default port).
2761 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2762 fallback_hosts, and "hosts" in the smtp transport.
2764 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2765 host errors such as "Connection refused".
2767 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2768 authentication with radiusclient 0.4.9:
2770 - Error returned from rc_read_config was caught wrongly
2771 - Username/password not passed on to radius server due to wrong length.
2773 The presumption is that some radiusclient API changes for 4.51/PH/17
2774 were not taken care of correctly. The code is still untested by me (my
2775 Linux distribution still has 0.3.2 of radiusclient), but it was
2776 contributed by a Radius user.
2778 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2779 expanding the "port" option of the smtp transport.
2781 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2782 while reading a MIME header. Thanks to Tom Hughes for a patch.
2784 PH/24 Include config.h inside local_scan.h so that configuration settings are
2787 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2788 that in an ACL for RCPT (for example), you can examine exactly what was
2791 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2792 commands, but it was not correctly comparing the address with the actual
2793 client host address. Thus, it would show the EHLO address in Received:
2794 header lines when this was not necessary.
2796 PH/27 Added the % operator to ${eval:}.
2798 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2799 it should be ignoring failures (because with -C, for example, it has lost
2800 privilege). It wasn't ignoring creation failures other than "already
2803 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2806 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2807 because an input line was too long, either on its own, or by virtue of
2808 too many continuations, the temporary file was not being removed, and the
2809 return code was incorrect.
2811 PH/31 Missing "BOOL" in function definition in filtertest.c.
2813 PH/32 Applied Sieve patches from the maintainer.
2815 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2817 PH/33 Added "verify = not_blind".
2819 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2820 Local/Makefile (with some defaults set). These are used in built scripts
2821 such as exicyclog, but they have never been used in the exim_install
2822 script (though there are many overriding facilities there). I have
2823 arranged that the exim_install script now takes note of these two
2826 PH/35 Installed configuration files for Dragonfly.
2828 PH/36 When a locally submitted message by a trusted user did not contain a
2829 From: header, and the sender address was obtained from -f or from an SMTP
2830 MAIL command, and the trusted user did not use -F to supply a sender
2831 name, $originator_name was incorrectly used when constructing a From:
2832 header. Furthermore, $originator_name was used for submission mode
2833 messages from external hosts without From: headers in a similar way,
2834 which is clearly wrong.
2836 PH/37 Added control=suppress_local_fixups.
2838 PH/38 When log_selector = +received_sender was set, and the addition of the
2839 sender made the log line's construction buffer exactly full, or one byte
2840 less than full, an overflow happened when the terminating "\n" was
2843 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2844 when the result of a list match is failure because a DNS lookup failed.
2846 PH/40 RM_COMMAND is now used in the building process.
2848 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2849 the "build-* directories that it finds.
2851 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2852 address in a domain literal was a prefix of an interface address.
2854 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2855 when verifying a sender address, unless rewrite_headers is false.
2857 PH/44 Wrote a long comment about why errors_to addresses are verified as
2858 recipients, not senders.
2860 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2861 the ratelimit ACL was added.
2863 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2865 PH/46 Added extra information about PostgreSQL errors to the error string.
2867 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2868 avoiding the need to install special zones in a real server. This is
2869 backwards compatible; if it can't find the fake resolver, it drops back.
2870 Thus, both old and new test suites can be run.
2872 TF/02 Added util/ratelimit.pl
2874 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2875 clock is set back in time.
2877 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2878 Candler <B.Candler@pobox.com>.
2880 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2881 OK for addresses that are the result of redirections.
2883 PH/48 A number of further additions for the benefit of the new test suite,
2884 including a fake gethostbyname() that interfaces to the fake DNS resolver
2887 TF/06 The fix for widen_domains has also been applied to qualify_single and
2888 search_parents which are the other dnslookup options that can cause
2891 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2894 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2896 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2897 service" if it encountered an MX record with an empty target hostname.
2898 The message is now "an MX or SRV record indicated no SMTP service".
2900 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2901 if the DNS of the sending site is misconfigured. This is quite a
2902 common situation. This change restores the behaviour of treating a
2903 helo verification defer as a failure.
2905 PH/51 If self=fail was set on a router, the bounce message did not include the
2906 actual error message.
2912 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2914 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2915 command itself ended in error, the underlying message about the transport
2916 filter timeout was being overwritten with the pipe command error. Now the
2917 underlying error message should be appended to the second error message.
2919 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2921 PH/02 Reduce the amount of output that "make" produces by default. Full output
2922 can still be requested.
2924 PH/03 The warning log line about a condition test deferring for a "warn" verb
2925 was being output only once per connection, rather than after each
2926 occurrence (because it was using the same function as for successful
2927 "warn" verbs). This seems wrong, so I have changed it.
2929 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2930 it should not have, which might have caused a crash in the right
2931 circumstances, but probably never did.
2933 PH/04 Installed a modified version of Tony Finch's patch to make submission
2934 mode fix the return path as well as the Sender: header line, and to
2935 add a /name= option so that you can make the user's friendly name appear
2938 TF/03 Added the control = fakedefer ACL modifier.
2940 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2941 Mark Lowes for thorough testing.
2943 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2945 TK/03 Merged latest SRS patch from Miles Wilton.
2947 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2948 with the definition in sysexits.h (which is #included earlier).
2949 Fortunately, Exim does not actually use EX_OK. The code used to try to
2950 preserve the sysexits.h value, by assumimg that macro definitions were
2951 scanned for macro replacements. I have been disabused of this notion,
2952 so now the code just undefines EX_OK before #including unistd.h.
2954 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2955 in the smtp transport. When a block could not be written in a single
2956 write() function, the timeout was being re-applied to each part-write.
2957 This seems wrong - if the receiver was accepting one byte at a time it
2958 would take for ever. The timeout is now adjusted when this happens. It
2959 doesn't have to be particularly precise.
2961 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2962 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2964 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2965 without a domain if the check to <postmaster@domain> fails.
2967 SC/01 Eximstats: added -xls and the ability to specify output files
2968 (patch written by Frank Heydlauf).
2970 SC/02 Eximstats: use FileHandles for outputing results.
2972 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2974 SC/04 Eximstats: fixed display of large numbers with -nvr option
2976 SC/05 Eximstats: fixed merging of reports with empty tables.
2978 SC/06 Eximstats: added the -include_original_destination flag
2980 SC/07 Eximstats: removed tabs and trailing whitespace.
2982 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2984 TK/06 MBOX spool code: Add real "From " MBOX separator line
2985 so the .eml file is really in mbox format (even though
2986 most programs do not really care). Patch from Alex Miller.
2988 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2989 The latter is generated from $received_to and is only set if the
2990 message has one envelope recipient. SA can use these headers,
2991 obviously out-of-the-box. Patch from Alex Miller.
2993 PH/08 The ${def test on a variable was returning false if the variable's
2994 value was "0", contrary to what the specification has always said!
2995 The result should be true unless the variable is empty.
2997 PH/09 The syntax error of a character other than { following "${if
2998 def:variable_name" (after optional whitespace) was not being diagnosed.
2999 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3000 accidental colon was present, for example, could give incorrect results.
3002 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3003 result is used (not including appendfile, where other changes are about
3006 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3007 This involved changing a lot of size variables from int to off_t. It
3008 should work with maildirs and everything.
3010 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3011 spamd dying while we are connected to it.
3013 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3016 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3017 function for BDB 4.3.
3019 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3021 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3022 the sources for PCRE have been re-arranged and more files are now
3025 PH/15 The code I had for printing potentially long long variables in PH/11
3026 above was not the best (it lost precision). The length of off_t variables
3027 is now inspected at build time, and an appropriate printing format (%ld
3028 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3029 to be "long long int" or "long int". This is needed for the internal
3030 formatting function string_vformat().
3032 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3033 the configuration file to be ":syslog", then the script "guesses" where
3034 the logs files are, rather than using the compiled in default. In our
3035 case the guess is not the same as the compiled default, so the script
3036 suddenly stopped working when I started to use syslog. The patch checks
3037 to see if log_file_path is "". If so, it attempts to read it from exim
3038 with no configuration file to get the compiled in version, before it
3039 falls back to the previous guessing code."
3041 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3042 implementing BATV in an Exim configuration. See NewStuff for the gory
3045 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3046 Makefile that are specific to HP-UX.
3048 PH/18 If the "use_postmaster" option was set for a recipient callout together
3049 with the "random" option, the postmaster address was used as the MAIL
3050 FROM address for the random test, but not for the subsequent recipient
3051 test. It is now used for both.
3053 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3054 patch removes a few documentation additions to RFC 3028, because the
3055 latest draft now contains them. It adds the new en;ascii-case comparator
3056 and a new error check for 8bit text in MIME parts. Comparator and
3057 require names are now matched exactly. I enabled the subaddress
3058 extension, but it is not well tested yet (read: it works for me)."
3060 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3061 rework some of the code of TK/09 above to avoid the hardwired use of
3062 "%lld" and "long long". Replaced the call to snprintf() with a call to
3065 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3066 records point to non-existent hosts", "retry timeout exceeded", and
3067 "retry time not reached for any host after a long failure period".
3069 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3070 experimental DomainKeys support:
3072 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3073 (2) On an error such as an illegally used "control", the wrong name for
3074 the control was given.
3076 These problems did NOT occur unless DomainKeys support was compiled.
3078 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3080 PH/24 Added ${if match_ip condition.
3082 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3083 databases so that it will be absolutely obvious if a crash occurs in the
3084 DB library. This is a regular occurrence (often caused by mis-matched
3087 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3088 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3089 picked up on a user's system that detects such things. There doesn't seem
3090 to be a gcc warning option for this - only an attribute that has to be
3091 put on the function's prototype. It seems that in Fedora Core 4 they have
3092 set this on a number of new functions. No doubt there will be more in due
3095 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3096 specify a transport. However, if an address that was verified by such a
3097 router was the subject of a callout, Exim crashed because it tried to
3098 read the rcpt_include_affixes from the non-existent transport. Now it
3099 just assumes that the setting of that option is false. This bug was
3100 introduced by 4.51/PH/31.
3102 PH/28 Changed -d+all to exclude +memory, because that information is very
3103 rarely of interest, but it makes the output a lot bigger. People tend to
3104 do -d+all out of habit.
3106 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3107 code in os-type was giving problems when libc.so lives in lib64, like on
3110 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3111 aren't the modern standard, and it seems that some systems' include files
3112 don't always have them. Exim was already checking for some of the newer
3113 ones like T_AAAA, and defining it itself. I've added checks for all the
3114 record types that Exim uses.
3116 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3117 not automatically generating a new one, as it is supposed to. This
3118 prevented TLS from working. If the file did exist, but contained invalid
3119 data, a new version was generated, as expected. It was only the case of a
3120 non-existent file that was broken.
3122 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3123 with a change in libdomainkeys > 0.64.
3125 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3126 from DNS. If the selector record carries the flag, it now has
3127 precedence over the domain-wide flag.
3129 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3131 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3132 the use of a port name that isn't defined in /etc/services) occurred, the
3133 message was deferred as in a normal delivery, and thus remained on the
3134 spool, instead of being failed because of the mua_wrapper setting. This
3135 is now fixed, and I tidied up some of the mua_wrapper messages at the
3138 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3139 the messages in a hash of arrays rather than using individual hashes.
3140 This is a bit cleaner and results in dramatic memory savings, albeit
3141 at a slight CPU cost.
3143 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3144 as requested by Marc Sherman.
3146 SC/10 Eximstats: added histograms for user specified patterns as requested
3149 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3151 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3152 fopen() in the content-scanning modules that did not already have it.
3158 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3159 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3161 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3163 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3165 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3166 to test Sieve filters that use "vacation".
3168 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3169 that changes the way the GnuTLS parameters are stored in the cache file.
3170 The new format can be generated externally. For backward compatibility,
3171 if the data in the cache doesn't make sense, Exim assumes it has read an
3172 old-format file, and it generates new data and writes a new file. This
3173 means that you can't go back to an older release without removing the
3176 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3177 work if there are any delivery delays because "one_time" forces the
3178 parent to be marked "delivered", so its unseen clone is never tried
3179 again. For this reason, Exim now forbids the simultaneous setting of
3182 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3183 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3184 entry if you want to know the details. The fix, however, neglected to
3185 consider the case when local delivery batching is involved. The test for
3186 "previously delivered" was not happening when checking to see if an
3187 address could be batched with a previous (undelivered) one; under
3188 certain circumstances this could lead to multiple deliveries to the same
3191 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3192 in its include files, and this causes problems building Exim.
3194 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3195 header_syntax) but Exim was just ignoring anything given after a slash.
3196 In particular, this caused confusion with an attempt to use "verify =
3197 reverse_host_lookup/defer_ok". An error is now given when options are
3198 supplied for verify items that do not have them. (Maybe reverse_host_
3199 lookup should have a defer_ok option, but that's a different point.)
3201 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3202 defined by RFC 821) to 2048, because there were problems with some AUTH
3203 commands, and RFC 1869 says the size should be increased for extended
3204 SMTP commands that take arguments.
3206 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3209 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3210 "unknown" error; now it says that the functionality isn't in the binary.
3212 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3213 an address' error message when a string expansion fails (syntax or
3214 whatever). Otherwise the password may appear in the log. Following change
3215 PH/42 below, there is no longer a chance of it appearing in a bounce
3218 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3220 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3221 of its addresses was ever tried. (Bugzilla bug #2.)
3223 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3224 the result incorrectly in the debug output. (It correctly added a newline
3225 to what was transported.)
3227 TF/01 Added $received_time.
3229 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3230 commented out examples of how to interface to a virus scanner and to
3231 SpamAssassin. Also added commented examples of av_scanner and
3232 spamd_address settings.
3234 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3235 and controls are allowed in which ACLs. There were a couple of minor
3236 errors. Some of the entries in the conditions table (which is a table of
3237 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3238 negation of where the condition IS allowed.
3240 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3242 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3243 header file does not have a version number, so I've had to invent a new
3244 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3245 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3246 radiusclient), but it was contributed by a Radius user.
3248 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3249 files or directories, for OpenSSL.
3251 PH/19 When an Exim process that is running as root has to create an Exim log
3252 file, it does so in a subprocess that runs as exim:exim so as to get the
3253 ownership right at creation (otherwise, other Exim processes might see
3254 the file with the wrong ownership). There was no test for failure of this
3255 fork() call, which would lead to the process getting stuck as it waited
3256 for a non-existent subprocess. Forks do occasionally fail when resources
3257 run out. I reviewed all the other calls to fork(); they all seem to check
3260 PH/20 When checking for unexpected SMTP input at connect time (before writing
3261 the banner), Exim was not dealing correctly with a non-positive return
3262 from the read() function. If the client had disconnected by this time,
3263 the result was a log entry for a synchronization error with an empty
3264 string after "input=" when read() returned zero. If read() returned -1
3265 (an event I could not check), uninitialized data bytes were printed.
3266 There were reports of junk text (parts of files, etc) appearing after
3269 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3271 PH/22 Added support for macro redefinition, and (re)definition in between
3272 driver and ACL definitions.
3274 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3275 forgetting to use the resulting value; it was using the unexpanded value.
3277 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3278 hadn't been configured. The fix is from Juergen Kreileder, who
3279 understands it better than I do:
3281 "Here's what I see happening with three configured cyrus_sasl
3282 authenticators configured (plain, login, cram-md5):
3284 On startup auth_cyrus_sasl_init() gets called for each of these.
3285 This means three calls to sasl_listmech() without a specified mech_list.
3286 => SASL tests which mechs of all available mechs actually work
3287 => three warnings about OTP not working
3288 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3290 With the patch, sasl_listmech() also gets called three times. But now
3291 SASL's mech_list option is set to the server_mech specified in the the
3292 authenticator. Or in other words, the answer from sasl_listmech()
3293 gets limited to just the mech you're testing for (which is different
3295 => the return list contains just 'plain' or 'login', 'cram-md5' or
3296 nothing depending on the value of ob->server_mech.
3298 I've just tested the patch: Authentication still works fine,
3299 unavailable mechs specified in the exim configuration are still
3300 caught, and the auth.log warnings about OTP are gone."
3302 PH/25 When debugging is enabled, the contents of the command line are added
3303 to the debugging output, even when log_selector=+arguments is not
3306 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3307 answer is "GNU", and only if the return is "GNU/something" is the answer
3310 PH/27 $acl_verify_message is now set immediately after the failure of a
3311 verification in an ACL, and so is available in subsequent modifiers. In
3312 particular, the message can be preserved by coding like this:
3314 warn !verify = sender
3315 set acl_m0 = $acl_verify_message
3317 Previously, $acl_verify_message was set only while expanding "message"
3318 and "log_message" when a very denied access.
3320 PH/28 Modified OS/os.c-Linux with
3322 -#ifndef OS_LOAD_AVERAGE
3323 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3325 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3326 nomenclature these days.)
3328 PH/29 Installed patch from the Sieve maintainer that adds the options
3329 sieve_useraddress and sieve_subaddress to the redirect router.
3331 PH/30 In these circumstances:
3332 . Two addresses routed to the same list of hosts;
3333 . First host does not offer TLS;
3334 . First host accepts first address;
3335 . First host gives temporary error to second address;
3336 . Second host offers TLS and a TLS session is established;
3337 . Second host accepts second address.
3338 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3339 and peerdn, if requested) that were in fact used only for the second
3342 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3343 attention to any local part prefix or suffix that was matched by the
3344 router that accepted the address. It now behaves in the same way as it
3345 does for delivery: the affixes are removed from the local part unless
3346 rcpt_include_affixes is set on the transport.
3348 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3349 timeout during the DATA phase of an incoming message.
3351 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3352 applied a patch sanctioned by the Sieve maintainer.
3354 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3355 the uid or gid is negative. A case of a negative gid caused this to be
3356 noticed. The fix allows for either to be negative.
3358 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3359 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3362 PH/36 The change PH/12 above was broken. Fixed it.
3364 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3365 the grounds that routing the same address twice would always produce the
3366 same answer. This might have been true once, but it is certainly no
3367 longer true now. Routing a child address may depend on the previous
3368 routing that produced that child. Some complicated redirection strategies
3369 went wrong when messages had multiple recipients, and made Exim's
3370 behaviour dependent on the order in which the addresses were given.
3372 I have moved the duplicate checking until after the routing is complete.
3373 Exim scans the addresses that are assigned to local and remote
3374 transports, and removes any duplicates. This means that more work will be
3375 done, as duplicates will always all be routed, but duplicates are
3376 presumably rare, so I don't expect this is of any significance.
3378 For deliveries to pipes, files, and autoreplies, the duplicate checking
3379 still happens during the routing process, since they are not going to be
3382 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3383 It corrects a timeout issue with spamd. This is Ian's comment: "The
3384 background is that sometimes spamd either never reads data from a
3385 connection it has accepted, or it never writes response data. The exiscan
3386 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3387 blindly assumes that writes won't block so it may never time out."
3389 PH/39 Allow G after quota size as well as K and M.
3391 PH/40 The value set for $authenticated_id in an authenticator may not contain
3392 binary zeroes or newlines because the value is written to log lines and
3393 to spool files. There was no check on this. Now the value is run through
3394 the string_printing() function so that such characters are converted to
3395 printable escape sequences.
3397 PH/41 $message_linecount is a new variable that contains the total number of
3398 lines in the message. Compare $body_linecount, which is the count for the
3401 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3402 bounce and delay warning messages, except in certain special cases, which
3405 (a) An SMTP error message from a remote host;
3406 (b) A message specified in a :fail: redirection;
3407 (c) A message specified in a "fail" command in a system filter;
3408 (d) A message specified in a FAIL return from the queryprogram router;
3409 (e) A message specified by the cannot_route_message router option.
3411 In these cases only, Exim does include the error details in bounce and
3412 warning messages. There are also a few cases where bland messages such
3413 as "unrouteable address" or "local delivery error" are given.
3415 PH/43 $value is now also set for the "else" part of a ${run expansion.
3417 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3418 being worked on, but at least Exim now implements the latest version to
3421 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3422 process to complete was treated as a delivery failure, a timeout while
3423 writing the message to the pipe was logged, but erroneously treated as a
3424 successful delivery. Such timeouts include transport filter timeouts. For
3425 consistency with the overall process timeout, these timeouts are now
3426 treated as errors, giving rise to delivery failures by default. However,
3427 there is now a new Boolean option for the pipe transport called
3428 timeout_defer, which, if set TRUE, converts the failures into defers for
3429 both kinds of timeout. A transport filter timeout is now identified in
3432 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3433 systems where "make" and "gmake" are different, calling "gmake" at top
3434 level broke things. I've arranged for the value of $(MAKE) to be passed
3435 from the Makefile to this script so that it can call the same version of
3439 A note about Exim versions 4.44 and 4.50
3440 ----------------------------------------
3442 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3443 changes of various kinds. As a consequence, a big documentation update was
3444 needed. This delayed the release for rather longer than seemed good, especially
3445 in the light of a couple of (minor) security issues. Therefore, the changes
3446 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3447 release. So 4.44 and 4.50 are in effect two different branches that both start
3450 I have left the 4.50 change log unchanged; it contains all the changes since
3451 4.43. The change log for 4.44 is below; many of its items are identical to
3452 those for 4.50. This seems to be the most sensible way to preserve the
3453 historical information.
3459 1. Minor wording change to the doc/README.SIEVE file.
3461 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3462 computation of the current number of files was incorrect.
3464 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3465 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3468 4. Give more explanation in the error message when the command for a transport
3469 filter fails to execute.
3471 5. There are several places where Exim runs a non-Exim command in a
3472 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3473 was being done only for the command run by the queryprogram router. It is
3474 now done for all such subprocesses. The other cases are: ${run, transport
3475 filters, and the commands run by the lmtp and pipe transports.
3477 6. Added CONFIGURE_GROUP build-time option.
3479 7. Some older OS have a limit of 256 on the maximum number of file
3480 descriptors. Exim was using setrlimit() to set 1000 as a large value
3481 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3482 systems. I've change it so that if it can't get 1000, it tries for 256.
3484 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3485 was an oversight, and furthermore, ever since the addition of extra
3486 controls (e.g. 4.43/32), the checks on when to allow different forms of
3487 "control" were broken. There should now be diagnostics for all cases when a
3488 control that does not make sense is encountered.
3490 9. Added the /retain_sender option to "control=submission".
3492 10. $recipients is now available in the predata ACL (oversight).
3494 11. Tidy the search cache before the fork to do a delivery from a message
3495 received from the command line. Otherwise the child will trigger a lookup
3496 failure and thereby defer the delivery if it tries to use (for example) a
3497 cached ldap connection that the parent has called unbind on.
3499 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3500 of $address_data from the recipient verification was clobbered by the
3501 sender verification.
3503 13. The value of address_data from a sender verification is now available in
3504 $sender_address_data in subsequent conditions in the ACL statement.
3506 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3508 15. Added a new option "connect=<time>" to callout options, to set a different
3511 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3512 was its contents. (It was OK if the option was not defined at all.)
3514 17. A "Completed" log line is now written for messages that are removed from
3515 the spool by the -Mrm option.
3517 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3518 information about exactly what failed.
3520 19. Added -dd to debug only the daemon process.
3522 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3523 handles timeouts, both on the server side and network timeouts. Renamed the
3524 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3526 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3527 It is now set to "smtps".
3529 22. $host_address is now set to the target address during the checking of
3530 ignore_target_hosts.
3532 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3533 being passed; this would have caused $sender_host_name to have been used if
3534 matching the list had actually called for a host name (not very likely,
3535 since this list is usually IP addresses). A host name is now passed as
3538 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3539 code that specifies a non-restarting handler (typically sigaction() in
3540 modern systems) in an attempt to fix a rare and obscure crash bug.
3542 25. Narrowed the window for a race in the daemon that could cause it to ignore
3543 SIGCHLD signals. This is not a major problem, because they are used only to
3544 wake it up if nothing else does.
3546 26. A malformed maildirsize file could cause Exim to calculate negative values
3547 for the mailbox size or file count. Odd effects could occur as a result.
3548 The maildirsize information is now recalculated if the size or filecount
3551 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3552 support for a long time. Removed HAVE_SYS_VFS_H.
3554 28. Installed the latest version of exipick from John Jetmore.
3556 29. In an address list, if the pattern was not a regular expression, an empty
3557 subject address (from a bounce message) matched only if the pattern was an
3558 empty string. Non-empty patterns were not even tested. This was the wrong
3559 because it is perfectly reasonable to use an empty address as part of a
3560 database query. An empty address is now tested by patterns that are
3561 lookups. However, all the other forms of pattern expect the subject to
3562 contain a local part and a domain, and therefore, for them, an empty
3563 address still always fails if the pattern is not itself empty.
3565 30. Exim went into a mad DNS loop when attempting to do a callout where the
3566 host was specified on an smtp transport, and looking it up yielded more
3567 than one IP address.
3569 31. Re-factored the code for checking spool and log partition space into a
3570 function that finds that data and another that does the check. The former
3571 is then used to implement four new variables: $spool_space, $log_space,
3572 $spool_inodes, and $log_inodes.
3574 32. The RFC2047 encoding function was originally intended for short strings
3575 such as real names; it was not keeping to the 75-character limit for
3576 encoded words that the RFC imposes. It now respects the limit, and
3577 generates multiple encoded words if necessary. To be on the safe side, I
3578 have increased the buffer size for the ${rfc2047: expansion operator from
3581 33. It is now permitted to omit both strings after an "if" condition; if the
3582 condition is true, the result is "true". As before, when the second string
3583 is omitted, a false condition yields an empty string. This makes it less
3584 cumbersome to write custom ACL and router conditions.
3586 34. Failure to deliver a bounce message always caused it to be frozen, even if
3587 there was an errors_to setting on the router. The errors_to setting is now
3590 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3591 canonical form (fully expanded) before being placed in
3592 $sender_host_address.
3594 36. The table in the code that translates DNS record types into text (T_A to
3595 "A" for instance) was missing entries for NS and CNAME. It is just possible
3596 that this could have caused confusion if both these types were looked up
3597 for the same domain, because the text type is used as part of Exim's
3598 per-process caching. But the chance of anyone hitting this buglet seems
3601 37. The dnsdb lookup has been extended in a number of ways.
3603 (1) There is a new type, "zns", which walks up the domain tree until it
3604 finds some nameserver records. It should be used with care.
3606 (2) There is a new type, "mxh", which is like "mx" except that it returns
3607 just the host names, not the priorities.
3609 (3) It is now possible to give a list of domains (or IP addresses) to be
3610 looked up. The behaviour when one of the lookups defers can be
3611 controlled by a keyword.
3613 (4) It is now possible to specify the separator character for use when
3614 multiple records are returned.
3616 38. The dnslists ACL condition has been extended: it is now possible to supply
3617 a list of IP addresses and/or domains to be looked up in a particular DNS
3620 39. Added log_selector=+queue_time_overall.
3622 40. When running the queue in the test harness, wait just a tad after forking a
3623 delivery process, to get repeatability of debugging output.
3625 41. Include certificate and key file names in error message when GnuTLS fails
3626 to set them up, because the GnuTLS error message doesn't include the name
3627 of the failing file when there is a problem reading it.
3629 42. Allow both -bf and -bF in the same test run.
3631 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3633 44. The "Exiscan patch" is now merged into the mainline Exim source.
3635 45. Sometimes the final signoff response after QUIT could fail to get
3636 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3637 < 0 before doing a fflush(). This bug looks as though it goes back to the
3638 introduction of TLS in release 3.20, but "sometimes" must have been rare
3639 because the tests only now provoked it.
3641 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3642 (this can affect the format of dates).
3644 47. exim_tidydb, when checking for the continued existence of a message for
3645 which it has found a message-specific retry record, was not finding
3646 messages that were in split spool directories. Consequently, it was
3647 deleting retry records that should have stayed in existence.
3649 48. Steve fixed some bugs in eximstats.
3651 49. The SPA authentication driver was not abandoning authentication and moving
3652 on to the next authenticator when an expansion was forced to fail,
3653 contradicting the general specification for all authenticators. Instead it
3654 was generating a temporary error. It now behaves as specified.
3656 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3657 (the order specifies the preference for clients). The order is now AES256,
3658 AES128, 3DES, ARCFOUR128.
3660 51. Small patch to Sieve code - explicitly set From: when generating an
3663 52. Exim crashed if a remote delivery caused a very long error message to be
3664 recorded - for instance if somebody sent an entire SpamAssassin report back
3665 as a large number of 550 error lines. This bug was coincidentally fixed by
3666 increasing the size of one of Exim's internal buffers (big_buffer) that
3667 happened as part of the Exiscan merge. However, to be on the safe side, I
3668 have made the code more robust (and fixed the comments that describe what
3671 53. Now that there can be additional text after "Completed" in log lines (if
3672 the queue_time_overall log selector is set), a one-byte patch to exigrep
3673 was needed to allow it to recognize "Completed" as not the last thing in
3676 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3677 patch that reportedly fixes this has been added. I am not expert enough to
3678 create a test for it. This is what the patch creator wrote:
3680 "I found a little strange behaviour of ldap code when working with
3681 Windows 2003 AD Domain, where users was placed in more than one
3682 Organization Units. When I tried to give exim partial DN, the exit code
3683 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3684 But simultaneously result of request was absolutely normal ldap result,
3685 so I produce this patch..."
3687 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3688 REFERENCE, so I have modified the code to exclude the patch when that macro
3691 55. Some experimental protocols are using DNS PTR records for new purposes. The
3692 keys for these records are domain names, not reversed IP addresses. The
3693 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3694 leaves it alone. Component reversal etc. now happens only for IP addresses.
3697 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3699 57. Double the size of the debug message buffer (to 2048) so that more of very
3700 long debug lines gets shown.
3702 58. The exicyclog utility now does better if the number of log files to keep
3703 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3705 59. Two changes related to the smtp_active_hostname option:
3707 (1) $smtp_active_hostname is now available as a variable.
3708 (2) The default for smtp_banner uses $smtp_active_hostname instead
3709 of $primary_hostname.
3711 60. The host_aton() function is supposed to be passed a string that is known
3712 to be a valid IP address. However, in the case of IPv6 addresses, it was
3713 not checking this. This is a hostage to fortune. Exim now panics and dies
3714 if the condition is not met. A case was found where this could be provoked
3715 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3716 components; fortuitously, this particular loophole had already been fixed
3717 by change 4.50/55 above.
3719 If there are any other similar loopholes, the new check in host_aton()
3720 itself should stop them being exploited. The report I received stated that
3721 data on the command line could provoke the exploit when Exim was running as
3722 exim, but did not say which command line option was involved. All I could
3723 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3724 running as the user.
3727 61. There was a buffer overflow vulnerability in the SPA authentication code
3728 (which came originally from the Samba project). I have added a test to the
3729 spa_base64_to_bits() function which I hope fixes it.
3732 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3733 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3735 63. The daemon start-up calls getloadavg() while still root for those OS that
3736 need the first call to be done as root, but it missed one case: when
3737 deliver_queue_load_max is set with deliver_drop_privilege. This is
3738 necessary for the benefit of the queue runner, because there is no re-exec
3739 when deliver_drop_privilege is set.
3741 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3742 This has been fixed.
3744 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3745 was in use, was not putting the data itself into the right store pool;
3746 consequently, it could be overwritten for a subsequent message in the same
3747 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3750 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3752 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3753 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3754 the function were treating the return as a boolean value, which happened to
3755 work because 0=false and not-0=true, but is not correct code.
3757 68. The host_aton() function was not handling scoped IPv6 addresses (those
3758 with, for example, "%eth0" on the end) correctly.
3760 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3761 negated items (that is, ~something) in unsigned ints. Some compilers
3762 apparently mutter when there is no cast.
3764 70. If an address verification called from an ACL failed, and did not produce a
3765 user-specific message (i.e. there was only a "system" message), nothing was
3766 put in $acl_verify_message. In this situation, it now puts the system
3769 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3770 session; change 4.31/43 added the unwanted input to the log line - except
3771 that it did not do this in the start of session case. It now does.
3773 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3774 This is wrong and can cause the other end to generate a synchronization
3775 error if it is another Exim or anything else that does the synchronization
3776 check. A QUIT command is no longer sent after a timeout.
3778 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3779 during host lookups.
3781 74. The defer_ok option of callout verification was not working if it was used
3782 when verifying addresses in header lines, that is, for this case:
3784 verify = header_sender/callout=defer_ok
3786 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3787 those file descriptors could be used for SMTP connections. If anything
3788 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3789 could be sent to the SMTP client, causing chaos. The daemon now opens
3790 stdin, stdout, and stderr to /dev/null when it puts itself into the
3793 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3794 log by default. The user can override this with suitable Perl magic.
3796 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3797 the log message when discard triggers, was not working for the DATA ACL or
3798 for the non-SMTP ACL.
3800 78. Error message wording change in sieve.c.
3802 79. If smtp_accept_max_per_host was set, the number of connections could be
3803 restricted to fewer than expected, because the daemon was trying to set up
3804 a new connection before checking whether the processes handling previous
3805 connections had finished. The check for completed processes is now done
3806 earlier. On busy systems, this bug wouldn't be noticed because something
3807 else would have woken the daemon, and it would have reaped the completed
3810 80. If a message was submitted locally by a user whose login name contained one
3811 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3812 It caused a spool format error. I have fixed the spool reading code. A
3813 related problem was that the "from" clause in the Received: line became
3814 illegal because of the space(s). It is now covered by ${quote_local_part.
3816 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3819 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3821 83. Patch from Sieve maintainer to fix unterminated string problem in
3822 "vacation" handling.
3824 84. Some minor changes to the Linux configuration files to help with other
3825 OS variants using glibc.
3827 85. One more patch for Sieve to update vacation handling to latest spec.
3830 ----------------------------------------------------
3831 See the note above about the 4.44 and 4.50 releases.
3832 ----------------------------------------------------
3838 1. Change 4.43/35 introduced a bug that caused file counts to be
3839 incorrectly computed when quota_filecount was set in an appendfile
3842 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3843 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3846 3. Give more explanation in the error message when the command for a transport
3847 filter fails to execute.
3849 4. There are several places where Exim runs a non-Exim command in a
3850 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3851 was being done only for the command run by the queryprogram router. It is
3852 now done for all such subprocesses. The other cases are: ${run, transport
3853 filters, and the commands run by the lmtp and pipe transports.
3855 5. Some older OS have a limit of 256 on the maximum number of file
3856 descriptors. Exim was using setrlimit() to set 1000 as a large value
3857 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3858 systems. I've change it so that if it can't get 1000, it tries for 256.
3860 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3861 was an oversight, and furthermore, ever since the addition of extra
3862 controls (e.g. 4.43/32), the checks on when to allow different forms of
3863 "control" were broken. There should now be diagnostics for all cases when a
3864 control that does not make sense is encountered.
3866 7. $recipients is now available in the predata ACL (oversight).
3868 8. Tidy the search cache before the fork to do a delivery from a message
3869 received from the command line. Otherwise the child will trigger a lookup
3870 failure and thereby defer the delivery if it tries to use (for example) a
3871 cached ldap connection that the parent has called unbind on.
3873 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3874 of $address_data from the recipient verification was clobbered by the
3875 sender verification.
3877 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3878 was its contents. (It was OK if the option was not defined at all.)
3880 11. A "Completed" log line is now written for messages that are removed from
3881 the spool by the -Mrm option.
3883 12. $host_address is now set to the target address during the checking of
3884 ignore_target_hosts.
3886 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3887 being passed; this would have caused $sender_host_name to have been used if
3888 matching the list had actually called for a host name (not very likely,
3889 since this list is usually IP addresses). A host name is now passed as
3892 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3893 code that specifies a non-restarting handler (typically sigaction() in
3894 modern systems) in an attempt to fix a rare and obscure crash bug.
3896 15. Narrowed the window for a race in the daemon that could cause it to ignore
3897 SIGCHLD signals. This is not a major problem, because they are used only to
3898 wake it up if nothing else does.
3900 16. A malformed maildirsize file could cause Exim to calculate negative values
3901 for the mailbox size or file count. Odd effects could occur as a result.
3902 The maildirsize information is now recalculated if the size or filecount
3905 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3906 support for a long time. Removed HAVE_SYS_VFS_H.
3908 18. Updated exipick to current release from John Jetmore.
3910 19. Allow an empty sender to be matched against a lookup in an address list.
3911 Previously the only cases considered were a regular expression, or an
3914 20. Exim went into a mad DNS lookup loop when doing a callout where the
3915 host was specified on the transport, if the DNS lookup yielded more than
3918 21. The RFC2047 encoding function was originally intended for short strings
3919 such as real names; it was not keeping to the 75-character limit for
3920 encoded words that the RFC imposes. It now respects the limit, and
3921 generates multiple encoded words if necessary. To be on the safe side, I
3922 have increased the buffer size for the ${rfc2047: expansion operator from
3925 22. Failure to deliver a bounce message always caused it to be frozen, even if
3926 there was an errors_to setting on the router. The errors_to setting is now
3929 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3930 canonical form (fully expanded) before being placed in
3931 $sender_host_address.
3933 24. Updated eximstats to version 1.33
3935 25. Include certificate and key file names in error message when GnuTLS fails
3936 to set them up, because the GnuTLS error message doesn't include the name
3937 of the failing file when there is a problem reading it.
3939 26. Expand error message when OpenSSL has problems setting up cert/key files.
3942 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3943 (this can affect the format of dates).
3945 28. exim_tidydb, when checking for the continued existence of a message for
3946 which it has found a message-specific retry record, was not finding
3947 messages that were in split spool directories. Consequently, it was
3948 deleting retry records that should have stayed in existence.
3950 29. eximstats updated to version 1.35
3951 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3952 1.35 - bugfix such that pie charts by volume are generated correctly
3954 30. The SPA authentication driver was not abandoning authentication and moving
3955 on to the next authenticator when an expansion was forced to fail,
3956 contradicting the general specification for all authenticators. Instead it
3957 was generating a temporary error. It now behaves as specified.
3959 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3960 (the order specifies the preference for clients). The order is now AES256,
3961 AES128, 3DES, ARCFOUR128.
3963 31. Small patch to Sieve code - explicitly set From: when generating an
3966 32. Exim crashed if a remote delivery caused a very long error message to be
3967 recorded - for instance if somebody sent an entire SpamAssassin report back
3968 as a large number of 550 error lines. This bug was coincidentally fixed by
3969 increasing the size of one of Exim's internal buffers (big_buffer) that
3970 happened as part of the Exiscan merge. However, to be on the safe side, I
3971 have made the code more robust (and fixed the comments that describe what
3974 33. Some experimental protocols are using DNS PTR records for new purposes. The
3975 keys for these records are domain names, not reversed IP addresses. The
3976 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3977 leaves it alone. Component reversal etc. now happens only for IP addresses.
3980 34. The host_aton() function is supposed to be passed a string that is known
3981 to be a valid IP address. However, in the case of IPv6 addresses, it was
3982 not checking this. This is a hostage to fortune. Exim now panics and dies
3983 if the condition is not met. A case was found where this could be provoked
3984 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3985 components; fortuitously, this particular loophole had already been fixed
3986 by change 4.50/55 or 4.44/33 above.
3988 If there are any other similar loopholes, the new check in host_aton()
3989 itself should stop them being exploited. The report I received stated that
3990 data on the command line could provoke the exploit when Exim was running as
3991 exim, but did not say which command line option was involved. All I could
3992 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3993 running as the user.
3996 35. There was a buffer overflow vulnerability in the SPA authentication code
3997 (which came originally from the Samba project). I have added a test to the
3998 spa_base64_to_bits() function which I hope fixes it.
4001 36. The daemon start-up calls getloadavg() while still root for those OS that
4002 need the first call to be done as root, but it missed one case: when
4003 deliver_queue_load_max is set with deliver_drop_privilege. This is
4004 necessary for the benefit of the queue runner, because there is no re-exec
4005 when deliver_drop_privilege is set.
4007 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4008 was in use, was not putting the data itself into the right store pool;
4009 consequently, it could be overwritten for a subsequent message in the same
4010 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4013 38. Sometimes the final signoff response after QUIT could fail to get
4014 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4015 < 0 before doing a fflush(). This bug looks as though it goes back to the
4016 introduction of TLS in release 3.20, but "sometimes" must have been rare
4017 because the tests only now provoked it.
4023 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4024 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4025 Now it returns an integer. A number of calls were still expecting a T/F
4026 return. Fortuitously, in all cases, the tests worked in OK situations,
4027 which is the norm. However, things would have gone wrong on any write
4028 failures on the smtp file descriptor. This function is used when sending
4029 messages over SMTP and also when doing verify callouts.
4031 2. When Exim is called to do synchronous delivery of a locally submitted
4032 message (the -odf or -odi options), it no longer closes stderr before doing
4035 3. Implemented the mua_wrapper option.
4037 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4039 5. Implemented the functions header_remove(), header_testname(),
4040 header_add_at_position(), and receive_remove_recipient(), and exported them
4043 6. If an ACL "warn" statement specified the addition of headers, Exim already
4044 inserted X-ACL-Warn: at the start if there was no header name. However, it
4045 was not making this test for the second and subsequent header lines if
4046 there were newlines in the string. This meant that an invalid header could
4047 be inserted if Exim was badly configured.
4049 7. Allow an ACL "warn" statement to add header lines at the start or after all
4050 the Received: headers, as well as at the end.
4052 8. Added the rcpt_4xx retry error code.
4054 9. Added postmaster_mailfrom=xxx to callout verification option.
4056 10. Added mailfrom=xxxx to the callout verification option, for verify=
4059 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4060 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4062 12. Inserted some casts to stop certain compilers warning when using pointer
4063 differences as field lengths or precisions in printf-type calls (mostly
4064 affecting debugging statements).
4066 13. Added optional readline() support for -be (dynamically loaded).
4068 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4069 same clock tick as a message's arrival, so that its received time was the
4070 same as the "first fail" time on the retry record, and that message
4071 remained on the queue past the ultimate address timeout, every queue runner
4072 would try a delivery (because it was past the ultimate address timeout) but
4073 after another failure, the ultimate address timeout, which should have then
4074 bounced the address, did not kick in. This was a "< instead of <=" error;
4075 in most cases the first failure would have been in the next clock tick
4076 after the received time, and all would be well.
4078 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4079 being recognized when the domain list was tested by the match_domain
4080 condition in an expansion string.
4082 16. Added the ${str2b64: operator.
4084 17. Exim was always calling setrlimit() to set a large limit for the number of
4085 processes, without checking whether the existing limit was already
4086 adequate. (It did check for the limit on file descriptors.) Furthermore,
4087 errors from getrlimit() and setrlimit() were being ignored. Now they are
4088 logged to the main and panic logs, but Exim does carry on, to try to do its
4089 job under whatever limits there are.
4091 18. Imported PCRE 5.0.
4093 19. Trivial typo in log message " temporarily refused connection" (the leading
4096 20. If the log selector return_path_on_delivery was set and an address was
4097 redirected to /dev/null, the delivery process crashed because it assumed
4098 that a return path would always be set for a "successful" delivery. In this
4099 case, the whole delivery is bypassed as an optimization, and therefore no
4102 21. Internal re-arrangement: the function for sending a challenge and reading
4103 a response while authentication was assuming a zero-terminated challenge
4104 string. It's now changed to take a pointer and a length, to allow for
4105 binary data in such strings.
4107 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4109 23. Exim was not respecting finduser_retries when seeking the login of the
4110 uid under which it was called; it was always trying 10 times. (The default
4111 setting of finduser_retries is zero.) Also, it was sleeping after the final
4112 failure, which is pointless.
4114 24. Implemented tls_on_connect_ports.
4116 25. Implemented acl_smtp_predata.
4118 26. If the domain in control=submission is set empty, Exim assumes that the
4119 authenticated id is a complete email address when it generates From: or
4120 Sender: header lines.
4122 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4123 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4124 chown and chgrp in /bin and hostname in /usr/bin.
4126 28. Exim was keeping the "process log" file open after each use, just as it
4127 does for the main log. This opens the possibility of it remaining open for
4128 long periods when the USR1 signal hits a daemon. Occasional processlog
4129 errors were reported, that could have been caused by this. Anyway, it seems
4130 much more sensible not to leave this file open at all, so that is what now
4133 29. The long-running daemon process does not normally write to the log once it
4134 has entered its main loop, and it closes the log before doing so. This is
4135 so that log files can straightforwardly be renamed and moved. However,
4136 there are a couple of unusual error situations where the daemon does write
4137 log entries, and I had neglected to close the log afterwards.
4139 30. The text of an SMTP error response that was received during a remote
4140 delivery was being truncated at 512 bytes. This is too short for some of
4141 the long messages that one sometimes sees. I've increased the limit to
4144 31. It is now possible to make retry rules that apply only when a message has a
4145 specific sender, in particular, an empty sender.
4147 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4148 it possible to be selective about when SMTP synchronization is enforced.
4150 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4152 32. Implemented hosts_connection_nolog.
4154 33. Added an ACL for QUIT.
4156 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4159 35. Added mailbox_size and mailbox_filecount to appendfile.
4161 36. Added control = no_multiline_responses to ACLs.
4163 37. There was a bug in the logic of the code that waits for the clock to tick
4164 in the case where the clock went backwards by a substantial amount such
4165 that the microsecond fraction of "now" was more than the microsecond
4166 fraction of "then" (but the whole seconds number was less).
4168 38. Added support for the libradius Radius client library this is found on
4169 FreeBSD (previously only the radiusclient library was supported).
4175 1. When certain lookups returned multiple values in the form name=value, the
4176 quoting of the values was not always being done properly. Specifically:
4177 (a) If the value started with a double quote, but contained no whitespace,
4179 (b) If the value contained whitespace other than a space character (i.e.
4180 tabs or newlines or carriage returns) it was not quoted.
4181 This fix has been applied to the mysql and pgsql lookups by writing a
4182 separate quoting function and calling it from the lookup code. The fix
4183 should probably also be applied to nisplus, ibase and oracle lookups, but
4184 since I cannot test any of those, I have not disturbed their existing code.
4186 2. A hit in the callout cache for a specific address caused a log line with no
4187 reason for rejecting RCPT. Now it says "Previous (cached) callout
4188 verification failure".
4190 3. There was an off-by-one bug in the queryprogram router. An over-long
4191 return line was truncated at 256 instead of 255 characters, thereby
4192 overflowing its buffer with the terminating zero. As well as fixing this, I
4193 have increased the buffer size to 1024 (and made a note to document this).
4195 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4196 when Exim is waiting for an SMTP response from a remote server, Exim
4197 restarts its select() call on the socket, thereby resetting its timeout.
4198 This is not a problem when such interrupts are rare. Somebody set up a cron
4199 job to run exiwhat every 2 minutes, which is less than the normal select()
4200 timeout (5 or 10 minutes). This meant that the select() timeout never
4201 kicked in because it was always reset. I have fixed this by comparing the
4202 time when an interrupt arrives with the time at the start of the first call
4203 to select(). If more time than the timeout has elapsed, the interrupt is
4204 treated as a timeout.
4206 5. Some internal re-factoring in preparation for the addition of Sieve
4207 extensions (by MH). In particular, the "personal" test is moved to a
4208 separate function, and given an option for scanning Cc: and Bcc: (which is
4209 not set for Exim filters).
4211 6. When Exim created an email address using the login of the caller as the
4212 local part (e.g. when creating a From: or Sender: header line), it was not
4213 quoting the local part when it contained special characters such as @.
4215 7. Installed new OpenBSD configuration files.
4217 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4218 try to make them clearer.
4220 9. Callout options, other than the timeout value, were being ignored when
4221 verifying sender addresses in header lines. For example, when using
4223 verify = header_sender/callout=no_cache
4225 the cache was (incorrectly) being used.
4227 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4228 only the Cygwin environment.
4230 11. When return_path_on_delivery was set as a log selector, if different remote
4231 addresses in the same message used different return paths and parallel
4232 remote delivery occurred, the wrong values would sometimes be logged.
4233 (Whenever a remote delivery process finished, the return path value from
4234 the most recently started remote delivery process was logged.)
4236 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4237 lines when AUTH and/or TLS are in use. This is the "received protocol"
4238 field. Exim used to use "asmtp" for authenticated SMTP, without any
4239 indication (in the protocol name) for TLS use. Now it follows the RFC and
4240 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4241 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4242 names appear in log lines as well as in Received: header lines.
4244 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4245 extensions, and comparison tests, and to fix some bugs.
4247 14. Changes to the "personal" filter test:
4249 (1) The test was buggy in that it was just doing the equivalent of
4250 "contains" tests on header lines. For example, if a user's address was
4251 anne@some.where, the "personal" test would incorrectly be true for
4253 To: susanne@some.where
4255 This test is now done by extracting each address from the header in turn,
4256 and checking the entire address. Other tests that are part of "personal"
4257 are now done using regular expressions (for example, to check local parts
4258 of addresses in From: header lines).
4260 (2) The list of non-personal local parts in From: addresses has been
4261 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4262 taken from the Sieve specification recommendations.
4264 (3) If the message contains any header line starting with "List-" it is
4265 treated as non-personal.
4267 (4) The test for "circular" in the Subject: header line has been removed
4268 because it now seems ill-conceived.
4270 15. Minor typos in src/EDITME comments corrected.
4272 16. Installed latest exipick from John Jetmore.
4274 17. If headers_add on a router specified a text string that was too long for
4275 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4276 of string_sprintf() is now avoided.
4278 18. $message_body_size was not set (it was always zero) when running the DATA
4279 ACL and the local_scan() function.
4281 19. For the "mail" command in an Exim filter, no default was being set for
4282 the once_repeat time, causing a random time value to be used if "once" was
4283 specified. (If the value happened to be <= 0, no repeat happened.) The
4284 default is now 0s, meaning "never repeat". The "vacation" command was OK
4285 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4286 (I found it when inspecting the code).
4288 20. There is now an overall timeout for performing a callout verification. It
4289 defaults to 4 times the callout timeout, which applies to individual SMTP
4290 commands during the callout. The overall timeout applies when there is more
4291 than one host that can be tried. The timeout is checked before trying the
4292 next host. This prevents very long delays if there are a large number of
4293 hosts and all are timing out (e.g. when the network connections are timing
4294 out). The value of the overall timeout can be changed by specifying an
4295 additional sub-option for "callout", called "maxwait". For example:
4297 verify = sender/callout=5s,maxwait=20s
4299 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4300 to the end before writing, but this should make it even safer).
4302 22. Exim was forgetting that it had advertised PIPELINING for the second and
4303 subsequent messages on an SMTP connection. It was also not resetting its
4304 memory on STARTTLS and an internal HELO.
4306 23. When Exim logs an SMTP synchronization error within a session, it now
4307 records whether PIPELINING has been advertised or not.
4309 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4310 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4311 rather than long int.
4313 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4315 26. Added the never_mail option to autoreply.
4321 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4322 crash if the getsockname() call failed; this can happen if a connection is
4323 closed very soon after it is established. The problem was simply in the
4324 order in which certain operations were done, causing Exim to try to write
4325 to the SMTP stream before it had set up the file descriptor. The bug has
4326 been fixed by making things happen in the correct order.
4332 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4333 before the connection was closed, thus losing the rejection response.
4335 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4336 some early Solaris releases, but causes trouble in current releases where
4337 socklen_t is defined.
4339 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4342 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4345 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4346 that releasing the top of it at the end releases what was used for sub-
4347 expansions (unless the block got too big). However, discard this block if
4348 the first thing is a variable or header, so that we can use its block when
4349 it is dynamic (useful for very large $message_headers, for example).
4351 6. Lookups now cache *every* query, not just the most recent. A new, separate
4352 store pool is used for this. It can be recovered when all lookup caches are
4353 flushed. Lookups now release memory at the end of their result strings.
4354 This has involved some general refactoring of the lookup sources.
4356 7. Some code has been added to the store_xxx() functions to reduce the amount
4357 of flapping under certain conditions.
4359 8. log_incoming_interface used to affect only the <= reception log lines. Now
4360 it causes the local interface and port to be added to several more SMTP log
4361 lines, for example "SMTP connection from", and rejection lines.
4363 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4365 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4367 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4368 could be overwritten at the end of the current message (or the start of a
4369 new message if it was set in a HELO ACL). The value is now preserved for
4370 the duration of the SMTP connection.
4372 12. If a transport had a headers_rewrite setting, and a matching header line
4373 contained an unqualified address, that address was qualified, even if it
4374 did not match any rewriting rules. The underlying bug was that the values
4375 of the flags that permit the existence of unqualified sender and recipient
4376 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4377 non-local messages, and by -bnq for local messages) were not being
4378 preserved with the message after it was received.
4380 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4381 "next input=" as part of the text comprising the host identity instead of
4382 the correct text. The code was using the same buffer for two different
4383 strings. However, depending on which order the printing function evaluated
4384 its arguments, the bug did not always show up. Under Linux, for example, my
4385 test suite worked just fine.
4387 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4388 doesn't work with some older versions of Perl. It has been changed to "my",
4389 which in any case is probably the better facility to use.
4391 15. A really picky compiler found some instances of statements for creating
4392 error messages that either had too many or two few arguments for the format
4395 16. The size of the buffer for calls to the DNS resolver has been increased
4396 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4397 for addresses that have a lot of PTR records. This alleviates a problem; it
4398 does not fully solve it.
4400 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4401 buffer now truncates the list and logs the incident, which is the same
4402 action as happens when Exim is looking up a host name and its aliases.
4403 Previously in this situation something unpredictable would happen;
4404 sometimes it was "internal error: store_reset failed".
4406 18. If a server dropped the connection unexpectedly when an Exim client was
4407 using GnuTLS and trying to read a response, the client delivery process
4408 crashed while trying to generate an error log message.
4410 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4411 string, for example:
4413 warn message = H1: something\nH2: something
4415 the text was added as a single header line from Exim's point of view
4416 though it ended up OK in the delivered message. However, searching for the
4417 second and subsequent header lines using $h_h2: did not work. This has been
4418 fixed. Similarly, if a system filter added multiple headers in this way,
4419 the routers could not see them.
4421 20. Expanded the error message when iplsearch is called with an invalid key to
4422 suggest using net-iplsearch in a host list.
4424 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4425 ACLs are no longer actually imposed (and a message to that effect is
4428 22. If a "gecos" field in a passwd entry contained escaped characters, in
4429 particular, if it contained a \" sequence, Exim got it wrong when building
4430 a From: or a Sender: header from that name. A second bug also caused
4431 incorrect handling when an unquoted " was present following a character
4432 that needed quoting.
4434 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4435 was not being matched caselessly.
4437 24. Arranged for all hyphens in the exim.8 source to be escaped with
4440 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4441 properly. Recipient callout cache records were still being keyed to include
4442 the sender, even when use_sender was set false. This led to far more
4443 callouts that were necessary. The sender is no longer included in the key
4444 when use_sender is false.
4446 26. Added "control = submission" modifier to ACLs.
4448 27. Added the ${base62d: operator to decode base 62 numbers.
4450 28. dnsdb lookups can now access SRV records.
4452 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4453 the configuration file.
4455 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4456 (-v) mode. This makes the output for a verbose queue run more intelligible.
4458 31. Added a use_postmaster feature to recipient callouts.
4460 32. Added the $body_zerocount variable, containing the number of binary zero
4461 bytes in the message body.
4463 33. The time of last modification of the "new" subdirectory is now used as the
4464 "mailbox time last read" when there is a quota error for a maildir
4467 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4469 35. Added +ignore_unknown as a special item in host lists.
4471 36. Code for decoding IPv6 addresses in host lists is now included, even if
4472 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4473 address was recognized as an IP address, but was then not correctly decoded
4474 into binary, causing unexpected and incorrect effects when compared with
4481 1. Very minor rewording of debugging text in manualroute to say "list of
4482 hosts" instead of "hostlist".
4484 2. If verify=header_syntax was set, and a header line with an unqualified
4485 address (no domain) and a large number of spaces between the end of the
4486 name and the colon was received, the reception process suffered a buffer
4487 overflow, and (when I tested it) crashed. This was caused by some obsolete
4488 code that should have been removed. The fix is to remove it!
4490 3. When running in the test harness, delay a bit after writing a bounce
4491 message to get a bit more predictability in the log output.
4493 4. Added a call to search_tidyup() just before forking a reception process. In
4494 theory, someone could use a lookup in the expansion of smtp_accept_max_
4495 per_host which, without the tidyup, could leave open a database connection.
4497 5. Added the variables $recipient_data and $sender_data which get set from a
4498 lookup success in an ACL "recipients" or "senders" condition, or a router
4499 "senders" option, similar to $domain_data and $local_part_data.
4501 6. Moved the writing of debug_print from before to after the "senders" test
4504 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4505 problems for message scanning, either using a data ACL, or using
4506 local_scan() because the Received: header was not generated till after they
4507 were called (in order to set the time as the time of reception completion).
4508 I have revised the way this works. The header is now generated after the
4509 body is received, but before the ACL or local_scan() are called. After they
4510 are run, the timestamp in the header is updated.
4516 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4517 before starting a queue runner without re-exec. This happened only when
4518 deliver_drop_privilege was set or when the Exim user was set to root. The
4519 effect of the bug was that timeouts during subsequent deliveries caused
4520 crashes instead of being properly handled. The handler is now left at its
4521 default (and expected) setting.
4523 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4524 message, again when deliver_drop_privilege is set or Exim is run as root.
4525 The bug described in (1) was not present in this case, but the tidying up
4526 of the other signals was missing. I have made the two cases consistent.
4528 3. The ignore_target_hosts setting on a manualroute router was being ignored
4529 for hosts that were looked up using the /MX notation.
4531 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4534 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4535 operated on the sender address. After changing the $sender_address to <>
4536 for the sender address verify, Exim was re-instated it as the original
4537 (before rewriting) address, but remembering that it had rewritten it, so it
4538 wasn't rewriting it again. This bug also had the effect of breaking the
4539 sender address verification caching when the sender address was rewritten.
4541 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4542 This has been changed so that if the ip literal address matches
4543 ignore_target_hosts, the router declines.
4545 7. Added expansion conditions match_domain, match_address, and match_local_
4546 part (NOT match_host).
4548 8. The placeholder for the Received: header didn't have a length field set.
4550 9. Added code to Exim itself and to exim_lock to test for a specific race
4551 condition that could lead to file corruption when using MBX delivery. The
4552 issue is with the lockfile that is created in /tmp. If this file is removed
4553 after a process has opened it but before that process has acquired a lock,
4554 there is the potential for a second process to recreate the file and also
4555 acquire a lock. This could lead to two Exim processes writing to the file
4556 at the same time. The added code performs the same test as UW imapd; it
4557 checks after acquiring the lock that its file descriptor still refers to
4558 the same named file.
4560 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4561 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4564 11. Added the smtp_active_hostname option. If used, this will typically be made
4565 to depend on the incoming interface address. Because $interface_address is
4566 not set up until the daemon has forked a reception process, error responses
4567 that can happen earlier (such as "too many connections") no longer contain
4570 12. If an expansion in a condition on a "warn" statement fails because a lookup
4571 defers, the "warn" statement is abandoned, and the next ACL statement is
4572 processed. Previously this caused the whole ACL to be aborted.
4574 13. Added the iplsearch lookup type.
4576 14. Added ident_timeout as a log selector.
4578 15. Added tls_certificate_verified as a log selector.
4580 16. Added a global option tls_require_ciphers (compare the smtp transport
4581 option of the same name). This controls incoming TLS connections.
4583 17. I finally figured out how to make tls_require_ciphers do a similar thing
4584 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4585 before starting the TLS session.
4587 18. Tabs are now shown as \t in -bP output.
4589 19. If the log selector return_path_on_delivery was set, Exim crashed when
4590 bouncing a message because it had too many Received: header lines.
4592 20. If two routers both had headers_remove settings, and the first one included
4593 a superfluous trailing colon, the final name in the first list and the
4594 first name in the second list were incorrectly joined into one item (with a
4595 colon in the middle).
4601 1. Added -C and -D options to the exinext utility, mainly to make it easier
4602 to include in the automated testing, but these could be helpful when
4603 multiple configurations are in use.
4605 2. The exinext utility was not formatting the output nicely when there was
4606 an alternate port involved in the retry record key, nor when there was a
4607 message id as well (for retries that were specific to a specific message
4608 and a specific host). It was also confused by IPv6 addresses, because of
4609 the additional colons they contain. I have fixed the IPv4 problem, and
4610 patched it up to do a reasonable job for IPv6.
4612 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4613 delivery, the log line now contains "pipelined" if PIPELINING was used.
4615 4. An SMTP transport process used to panic and die if the bind() call to set
4616 an explicit outgoing interface failed. This has been changed; it is now
4617 treated in the same way as a connect() failure.
4619 5. A reference to $sender_host_name in the part of a conditional expansion
4620 that was being skipped was still causing a DNS lookup. This no longer
4623 6. The def: expansion condition was not recognizing references to header lines
4624 that used bh_ and bheader_.
4626 7. Added the _cache feature to named lists.
4628 8. The code for checking quota_filecount in the appendfile transport was
4629 allowing one more file than it should have been.
4631 9. For compatibility with Sendmail, the command line option
4639 and sets the incoming protocol and host name (for trusted callers). The
4640 host name and its colon can be omitted when only the protocol is to be set.
4641 Note the Exim already has two private options, -pd and -ps, that refer to
4642 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4643 "s", but I don't think that's a major issue.
4645 10. A number of refactoring changes to the code, none of which should affect
4648 (a) The number of logging options was getting close to filling up the
4649 32-bit word that was used as a bit map. I have split them into two classes:
4650 those that are passed in the argument to log_write(), and those that are
4651 only ever tested independently outside of that function. These are now in
4652 separate 32-bit words, so there is plenty of room for expansion again.
4653 There is no change in the user interface or the logging behaviour.
4655 (b) When building, for example, log lines, the code previously used a
4656 macro that called string_cat() twice, in order to add two strings. This is
4657 not really sufficiently general. Furthermore, there was one instance where
4658 it was actually wrong because one of the argument was used twice, and in
4659 one call a function was used. (As it happened, calling the function twice
4660 did not affect the overall behaviour.) The macro has been replaced by a
4661 function that can join an arbitrary number of extra strings onto a growing
4664 (c) The code for expansion conditions now uses a table and a binary chop
4665 instead of a serial search (which was left over from when there were very
4666 few conditions). Also, it now recognizes conditions like "pam" even when
4667 the relevant support is not compiled in: a suitably worded error message is
4668 given if an attempt is made to use such a condition.
4670 11. Added ${time_interval:xxxxx}.
4672 12. A bug was causing one of the ddress fields not to be passed back correctly
4673 from remote delivery subprocesses. The field in question was not being
4674 subsequently used, so this caused to problems in practice.
4676 13. Added new log selectors queue_time and deliver_time.
4678 14. Might have fixed a bug in maildirsizefile handling that threw up
4679 "unexpected character" debug warnings, and recalculated the data
4680 unnecessarily. In any case, I expanded the warning message to give more
4683 15. Added the message "Restricted characters in address" to the statements in
4684 the default ACL that block characters like @ and % in local parts.
4686 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4687 Three changes have been made:
4689 (a) There was a serious bug; a negative response to MAIL caused the whole
4690 recipient domain to be cached as invalid, thereby blocking all messages
4691 to all local parts at the same domain, from all senders. This bug has
4692 been fixed. The domain is no longer cached after a negative response to
4693 MAIL if the sender used is not empty.
4695 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4698 (c) A new callout option, "use_sender" has been added for people who want
4699 the modified behaviour.
4705 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4708 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4709 indeed breaks things for older releases.
4711 3. Added additional logging to the case where there is a problem reading data
4712 from a filter that is running in a subprocess using a pipe, in order to
4713 try to track down a specific problem.
4715 4. Testing facility fudge: when running in the test harness and attempting
4716 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4717 getting "No route to host". Convert this to a timeout.
4719 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4722 6. Some OS don't have socklen_t but use size_t instead. This affects the
4723 fifth argument of getsockopt() amongst other things. This is now
4724 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4725 can be set for individual OS. I have set it for SunOS5, OSF1, and
4726 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4727 some earlier ones do not.
4729 7. Change 4.30/15 was not doing the test caselessly.
4731 8. The standard form for an IPv6 address literal was being rejected by address
4732 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4733 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4734 this, as well as the form without the "IPv6" on the front (but only when
4735 address literals are enabled, of course).
4737 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4739 10. Exim crashed if a message with an empty sender address specified by -f
4740 encountered a router with an errors_to setting. This could be provoked only
4741 by a command such as
4745 where an empty string was supplied; "<>" did not hit this bug.
4747 11. Installed PCRE release 4.5.
4749 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4750 remained set. It is now erased.
4752 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4753 times from message ids (which are base 36 rather than the normal 62).
4755 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4756 were being counted as actual protocol errors, and logged if the log
4757 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4758 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4759 and DATA following a set of rejected RCPTs do not count as protocol errors.
4760 In other words, Exim assumes they were pipelined, though this may not
4761 actually be the case. Of course, in all cases the client gets an
4762 appropriate error code.
4764 15. If a lookup fails in an ACL condition, a message about the failure may
4765 be available; it is used if testing the ACL cannot continue, because most
4766 such messages specify what the cause of the deferral is. However, some
4767 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4768 that caused an old message to be retained and used if a later statement
4769 caused a defer, replacing the real cause of the deferral.
4771 16. If an IP address had so many PTR records that the DNS lookup buffer
4772 was not large enough to hold them, Exim could crash while trying to process
4773 the truncated data. It now detects and logs this case.
4775 17. Further to 4.21/58, another change has been made: if (and only if) the
4776 first line of a message (the first header line) ends with CRLF, a bare LF
4777 in a subsequent header line has a space inserted after it, so as not to
4778 terminate the header.
4780 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4781 unnecessarily, used atoi() instead of strtol(), and didn't check the
4782 termination when getting file sizes from file names by regex.
4784 19. Completely re-implemented the support for maildirsize files, in the light
4785 of a number of problems with the previous contributed implementation
4786 (4.30/29). In particular:
4788 . If the quota is zero, the maildirsize file is maintained, but no quota is
4791 . If the maildir directory does not exist, it is created before any attempt
4792 to write a maildirsize file.
4794 . The quota value in the file is just a cache; if the quota is changed in
4795 the transport, the new value overrides.
4797 . A regular expression is available for excluding directories from the
4800 20. The autoreply transport checks the characters in options that define the
4801 message's headers; it allows continued headers, but it was checking with
4802 isspace() after an embedded newline instead of explicitly looking for a
4805 21. If all the "regular" hosts to which an address was routed had passed their
4806 expiry times, and had not reached their retry times, the address was
4807 bounced, even if fallback hosts were defined. Now Exim should go on to try
4810 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4811 equivalent code in the SMTP transport. Some hosts send humungous responses
4812 to HELO/EHLO, more than 1024 it seems.
4814 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4815 gives compiler warnings in some environments. I've now done it "properly",
4818 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4819 (because of problems with the built-in one) was declared to return uschar *
4820 instead of char *, causing compiler failure.
4822 25. Fixed a file descriptor leak when processing alias/forward files.
4824 26. Fixed a minor format string issue in dbfn.c.
4826 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4828 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4829 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4832 29. When a sender address is verified, it is cached, to save repeating the test
4833 when there is more than one recipient in a message. However, when the
4834 verification involves a callout, it is possible for different callout
4835 options to be set for different recipients. It is too complicated to keep
4836 track of this in the cache, so now Exim always runs a verification when a
4837 callout is required, relying on the callout cache for the optimization.
4838 The overhead is duplication of the address routing, but this should not be
4841 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4842 to be verified with callout=postmaster, and the main callout worked but the
4843 postmaster check failed, the verification correctly failed. However, if a
4844 subsequent RCPT command asked for sender verification *without* the
4845 postmaster check, incorrect caching caused this verification also to fail,
4848 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4849 it was not caching the DNS options (qualify_single, search_parents) that
4850 were used when the lookup failed. A subsequent lookup with different
4851 options therefore always gave the same answer, though there were cases
4852 where it should not have. (Example: a "domains = !$mx_any" option on a
4853 dnslookup router: the "domains" option is always processed without any
4854 widening, but the router might have qualify_single set.) Now Exim uses the
4855 cached value only when the same options are set.
4857 32. Added John Jetmore's "exipick" utility to the distribution.
4859 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4860 than a timeout (e.g. a certificate is required, and is not provided), an
4861 Exim server now closes the connection immediately. Previously it waited for
4862 the client to close - but if the client is SSL, it seems that they each
4863 wait for each other, leading to a delay before one of them times out.
4865 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4866 maintained 0.8.x compatibility because I don't think many are using it, and
4867 it is clearly obsolete.
4869 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4872 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4873 client certificate was expired. A simple patch fixes this, though I don't
4874 understand the full logic of why the verify callback is called multiple
4877 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4878 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4879 which causes problems with some clients (such as the Certicom SSL Plus
4880 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4881 disables the coutermeasure allowing Eudora to connect."
4883 38. Exim was not checking that a write() to a log file succeeded. This could
4884 lead to Bad Things if a log got too big, in particular if it hit a file
4885 size limit. Exim now panics and dies if it cannot write to a log file, just
4886 as it does if it cannot open a log file.
4888 39. Modified OS/Makefile-Linux so that it now contains
4890 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4892 The two -D definitions ensure that Exim is compiled with large file
4893 support, which makes it possible to handle log files that are bigger than
4896 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4897 instance) a domain was checked against a named list that involved a lookup,
4898 causing $domain_data to be set, then another domain was checked against the
4899 same list, then the first domain was re-checked, the value of $domain_data
4900 after the final check could be wrong. In particular, if the second check
4901 failed, it could be set empty. This bug probably also applied to
4904 41. The strip_trailing_dot option was not being applied to the address given
4905 with the -f command-line option.
4907 42. The code for reading a message's header from the spool was incrementing
4908 $received_count, but never initializing it. This meant that the value was
4909 incorrect (doubled) while delivering a message in the same process in which
4910 it was received. In the most common configuration of Exim, this never
4911 happens - a fresh exec is done - but it can happen when
4912 deliver_drop_privilege is set.
4914 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4915 it now includes up to 150 characters of the unexpected data in the log
4918 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4919 and building data strings. The size of both of these buffers was 10 000
4920 bytes - far larger than anybody would *ever* want, thought I. Needless to
4921 say, somebody hit the limit. I have increased the maximum line length to
4922 20 000 and the maximum data length of concatenated lines to 100 000. I have
4923 also fixed two bugs, because there was no checking on these buffers. Tsk,
4924 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4925 buffer is too small.
4927 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4928 lsearch lookups. Now it does.
4930 46. When parsing a route_list item in a manualroute router, a fixed-length
4931 buffer was used for the list of hosts. I made this 1024 bytes long,
4932 thinking that nobody would ever have a list of hosts that long. Wrong.
4933 Somebody had a whole pile of complicated expansion conditions, and the
4934 string was silently truncated, leading to an expansion error. It turns out
4935 that it is easier to change to an unlimited length (owing to other changes
4936 that have happened since this code was originally written) than to build
4937 structure for giving a limitation error. The length of the item that
4938 expands into the list of hosts is now unlimited.
4940 47. The lsearch lookup could not handle data where the length of text line was
4941 more than 4095 characters. Such lines were truncated, leading to shortened
4942 data being returned. It should now handle lines of any length.
4944 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4945 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4948 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4949 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4950 no longer happens in comment lines. A list of replacements is now placed
4951 at the head of all of the source files, except those whose only change is
4952 to replace PERL_COMMAND in the very first #! line.
4954 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4955 messages on the queue, with a bottom-up merge sort, using code contributed
4956 by Michael Haardt. This should make operations like -bp somewhat faster on
4957 large queues. It won't affect queue runners, except when queue_run_in_order
4960 51. Installed eximstats 1.31 in the distribution.
4962 52. Added support for SRV lookups to the dnslookup router.
4964 53. If an ACL referred to $message_body or $message_body_end, the value was not
4965 reset for any messages that followed in the same SMTP session.
4967 54. The store-handling optimization for building very long strings was not
4968 differentiating between the different store pools. I don't think this
4969 actually made any difference in practice, but I've tidied it.
4971 55. While running the routers to verify a sender address, $sender_address
4972 was still set to the sender address. This is wrong, because when routing to
4973 send a bounce to the sender, it would be empty. Therefore, I have changed
4974 it so that, while verifying a sender address, $sender_address is set to <>.
4975 (There is no change to what happens when verifying a recipient address.)
4977 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4978 target A or AAAA records (if not already returned) without resetting the
4979 qualify_single or search_parents options of the DNS resolver. These are
4980 inappropriate in this case because the targets of MX and SRV records must
4981 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4982 target that, when qualified, matched something in the local domain. These
4983 two options are now turned off when doing these lookups.
4985 57. It seems that at least some releases of Reiserfs (which does not have the
4986 concept of a fixed number of inodes) returns zero and not -1 for the
4987 number of available inodes. This interacted badly with check_spool_inodes,
4988 which assumed that -1 was the "no such thing" setting. What I have done is
4989 to check that the total number of inodes is greater than zero before doing
4990 the test of how many are available.
4992 58. When a "warn" ACL statement has a log_message modifier, the message is
4993 remembered, and not repeated. This is to avoid a lot of repetition when a
4994 message has many recipients that cause the same warning to be written.
4995 Howewer, Exim was preserving the list of already written lines for an
4996 entire SMTP session, which doesn't seem right. The memory is now reset if a
4997 new message is started.
4999 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5000 debugging output unless log_rewrite was also set.
5002 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5003 of (int)(handle) when we know that handle contains (void *)(-1).
5005 61. The Exim daemon panic-logs an error return when it closes the incoming
5006 connection. However "connection reset by peer" seems to be common, and
5007 isn't really an error worthy of noting specially, so that particular error
5010 62. When Exim is trying to find all the local interfaces, it used to panic and
5011 die if the ioctl to get the interface flags failed. However, it seems that
5012 on at least one OS (Solaris 9) it is possible to have an interface that is
5013 included in the list of interfaces, but for which you get a failure error
5014 for this call. This happens when the interface is not "plumbed" into a
5015 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5016 failure of the "get flags" call assumes that the interface is down.
5018 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5019 makes life easier for people who are doing arithmetic on fields extracted
5020 from dates, where you often get leading zeros that should not be
5021 interpreted as octal.
5023 64. Added qualify_domain to the redirect router, to override the global
5026 65. If a pathologically long header line contained very many addresses (the
5027 report of this problem mentioned 10 000) and each of them was rewritten,
5028 Exim could use up a very large amount of memory. (It kept on making new
5029 copies of the header line as it rewrote, and never released the old ones.)
5030 At the expense of a bit more processing, the header rewriting function has
5031 been changed so that it no longer eats memory in this way.
5033 66. The generation of the Received: header has been moved from the time that a
5034 message starts to be received, to the time that it finishes. The timestamp
5035 in the Received: header should now be very close to that of the <= log
5036 line. There are two side-effects of this change:
5038 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5039 logged header lines no longer include the local Received: line, because
5040 it has not yet been created. The same applies to a copy of the message
5041 that is returned to a non-SMTP sender when a message is rejected.
5043 (b) When a filter file is tested using -bf, no additional Received: header
5044 is added to the test message. After some thought, I decided that this
5047 This change does not affect the value of $received_for. It is still set
5048 after address rewriting, but before local_scan() is called.
5050 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5052 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5053 gave an unhelpful panic error message, and a defer error. I have managed to
5054 change this behaviour so that it now rejects any supplied certificate,
5055 which seems right, as the list of acceptable certificates is empty.
5057 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5058 gave an unhelpful defer error. I have not managed to make this reject any
5059 supplied certificates, but the error message it gives is "no certificate
5060 supplied", which is not helpful.
5062 70. exigrep's output now also includes lines that are not associated with any
5063 message, but which match the given pattern. Implemented by a patch from
5064 Martin Sluka, which also tidied up the Perl a bit.
5066 71. Recipient callout verification, like sender verification, was using <> in
5067 the MAIL FROM command. This isn't really the right thing, since the actual
5068 sender may affect whether the remote host accepts the recipient or not. I
5069 have changed it to use the actual sender in the callout; this means that
5070 the cache record is now keyed on a recipient/sender pair, not just the
5071 recipient address. There doesn't seem to be a real danger of callout loops,
5072 since a callout by the remote host to check the sender would use <>.
5073 [SEE ABOVE: changed after hitting problems.]
5075 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5076 temporary errors. However, in the case of such a code being given after
5077 the end of a data transmission (i.e. after ".") Exim was failing to write
5078 a retry record for the message. (Yes, there was some broken host that was
5079 actually sending 8xx at this point.)
5081 73. An unknown lookup type in a host list could cause Exim to panic-die when
5082 the list was checked. (An example that provoked this was putting <; in the
5083 middle of a list instead of at the start.) If this happened during a DATA
5084 ACL check, a -D file could be left lying around. This kind of configuration
5085 error no longer causes Exim to die; instead it causes a defer errror. The
5086 incident is still logged to the main and panic logs.
5088 74. Buglet left over from Exim 3 conversion. The message "too many messages
5089 in one connection" was written to the rejectlog but not the mainlog, except
5090 when address rewriting (yes!) was being logged.
5092 75. Added write_rejectlog option.
5094 76. When a system filter was run not as root (that is, when system_filter_user
5095 was set), the values of the $n variables were not being returned to the
5096 main process; thus, they were not subsequently available in the $sn
5099 77. Added +return_path_on_delivery log selector.
5101 78. A connection timeout was being treated differently from recipients deferred
5102 when testing hosts_max_try with a message that was older than the host's
5103 retry timeout. (The host should not be counted, thus allowing all hosts to
5104 be tried at least once before bouncing.) This may have been the cause of an
5105 occasionally reported bug whereby a message would remain on the queue
5106 longer than the retry timeout, but would be bounced if a delivery was
5107 forced. I say "may" because I never totally pinned down the problem;
5108 setting up timeout/retry tests is difficult. See also the next item.
5110 79. The ultimate address timeout was not being applied to errors that involved
5111 a combination of host plus message (for example, a timeout on a MAIL
5112 command). When an address resolved to a number of possible hosts, and they
5113 were not all tried for each delivery (e.g. because of hosts_max_try), a
5114 message could remain on the queue longer than the retry timeout.
5116 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5119 81. Fixed an obscure SMTP outgoing bug which required at least the following
5120 conditions: (a) there was another message waiting for the same server;
5121 (b) the server returned 5xx to all RCPT commands in the first message so
5122 that the message was not completed; (c) the server dropped the connection
5123 or gave a negative response to the RSET that Exim sends to abort the
5124 transaction. The observed case was a dropped connection after DATA that had
5125 been sent in pipelining mode. That is, the server had advertised PIPELINING
5126 but was not implementing it correctly. The effect of the bug was incorrect
5127 behaviour, such as trying another host, and this could lead to a crash.
5133 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5134 and daemon.c were passed as pointers to ints; they should have been
5135 pointers to socklen_t variables (which are typically unsigned ints).
5137 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5140 3. Fixed a really odd bug that affected only the testing scheme; patching a
5141 certain fixed string in the binary changed the value of another string that
5142 happened to be identical to the end of the original first string.
5144 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5145 name", it returns that address as the IP address. On some operating
5146 systems (e.g. Solaris), it also passes back the IP address string as the
5147 "host name". However, on others (e.g. Linux), it passes back an empty
5148 string. Exim wasn't checking for this, and was changing the host name to an
5149 empty string, assuming it had been canonicized.
5151 5. Although rare, it is permitted to have more than one PTR record for a given
5152 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5153 all the names associated with an address, because they do in Solaris.
5154 However, it seems that they do not in Linux for data that comes from the
5155 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5156 I found this out when I moved to a new Linux workstation and tried to run
5157 the Exim test suite.
5159 To get round this problem I have changed the code so that it now does its
5160 own call to the DNS to look up PTR records when searching for a host name.
5161 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5162 addresses that are only in /etc/hosts are still found.
5164 This behaviour is, however, controlled by an option called host_lookup_
5165 order, which defaults to "bydns:byaddr". If people want to use the other
5166 order, or indeed, just use one or the other means of lookup, they can
5167 specify it in this variable.
5169 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5170 some operating systems, this comes back from gethostbyaddr() as an empty
5171 string, and this is what Exim used to test for. However, it seems that in
5172 other systems, "." is yielded. Exim now tests for this case too.
5174 7. The values of check_spool_space and check_log_space are now held internally
5175 as a number of kilobytes instead of an absolute number of bytes. If a
5176 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5177 kilobyte. This means that much larger values can be stored.
5179 8. Exim monitor: an attempt to get the action menu when not actually pointing
5180 at a message produces an empty menu entitled "No message selected". This
5181 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5182 no entries in it ("Shell widget menu has zero width and/or height"). So I
5183 have added a single, blank menu entry in this case.
5185 9. Added ${quote_local_part.
5187 10. MIME decoding is now applied to the contents of Subject: header lines when
5190 11. Now that a reference to $sender_host_address automatically causes a reverse
5191 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5192 host lookup before query-style lookups in lists that might use this
5193 variable. This has therefore been abolished, and the "net-" prefix is no
5194 longer necessary for query-style lookups.
5196 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5197 have been a typo for LFLAGS, so it has been changed.
5199 13. The install script calls Exim with "-C /dev/null" in order to find the
5200 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5201 to be output. Howeve, since Exim outputs its version number before the
5202 error, it didn't break the script. It just looked ugly. I fixed this by
5203 always allowing "-C /dev/null" if the caller is root.
5205 14. Ignore overlarge ACL variable number when reading spool file - insurance
5206 against a later release with more variables having written the file.
5208 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5209 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5210 this, as well as the form without the "IPv6" on the front.
5212 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5213 OS/Makefile-Darwin file.
5215 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5216 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5218 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5219 variable. However, some OS use ints for inodes. I've added cast to long int
5220 to get rid of the compiler warning.
5222 19. I had forgotten to lock out "/../" in configuration file names when
5223 ALT_CONFIG_PREFIX was set.
5225 20. Routers used for verification do not need to specify transports. However,
5226 if such a router generated a host list, and callout was configured, Exim
5227 crashed, because it could not find a port number from the (non-existent)
5228 transport. It now assumes port 25 in this circumstance.
5230 21. Added the -t option to exigrep.
5232 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5233 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5234 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5235 them is set, LOOKUP_LSEARCH is forced.
5237 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5239 24. Added sender and host information to the "rejected by local_scan()" log
5240 line; previously there was no indication of these.
5242 25. Added .include_if_exists.
5244 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5245 Linux. It turns out that not all file systems support this. Apparently some
5246 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5247 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5248 not supported on the file descriptor, is now ignored when Exim is trying to
5249 sync a directory. This applies only to Linux.
5251 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5253 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5254 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5257 29. Added support for maildirsize files from supplied patch (modified a bit).
5259 30. The use of :fail: followed by an empty string could lead Exim to respond to
5260 sender verification failures with (e.g.):
5262 550 Verification failed for <xxx>
5263 550 Sender verify failed
5265 where the first response line was missing the '-' that indicates it is not
5266 the final line of the response.
5268 31. The loop for finding the name of the user that called Exim had a hardwired
5269 limit of 10; it now uses the value of finduser_retries, which is used for
5270 all other user lookups.
5272 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5275 33. Exim was neglecting to zero errno before one call of strtol() when
5276 expanding a string and expecting an integer value. On some systems this
5277 resulted in spurious "integer overflow" errors. Also, it was casting the
5278 result into an int without checking.
5280 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5281 did not work. The code looks as if it has *never* worked, though it appears
5282 to have been documented since at least releast 1.62. I have made it work.
5284 35. The "timeout_DNS" error in retry rules, also documented since at least
5285 1.62, also never worked. As it isn't clear exactly what this means, and
5286 clearly it isn't a major issue, I have abolished the feature by treating it
5287 as "timeout", and writing a warning to the main and panic logs.
5289 36. The display of retry rules for -brt wasn't always showing the error code
5292 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5293 timeout_connect_A, timeout_connect_MX.
5295 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5296 to the empty sender.
5298 39. The daemon was not analysing the content of -oX till after it had closed
5299 stderr and disconnected from the controlling terminal. This meant that any
5300 syntax errors were only noted on the panic log, and the return code from
5301 the command was 0. By re-arranging the code a little, I've made the
5302 decoding happen first, so such errors now appear on stderr, and the return
5303 code is 1. However, the actual setting up of the sockets still happens in
5304 the disconnected process, so errors there are still only recorded on the
5307 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5308 connections (as happens on some IP stacks) was logged at start up time as
5309 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5310 it from "IPv6 and IPv4", which means that two separate sockets are being
5313 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5314 says whether AF_INET or AF_INET6 was passed as an argument.
5316 42. Exiwhat output was messed up when time zones were included in log
5319 43. Exiwhat now gives more information about the daemon's listening ports,
5320 and whether -tls-on-connect was used.
5322 44. The "port" option of the smtp transport is now expanded.
5324 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5325 silently ignored. Now an error message is written to the main and panic
5328 46. There's a new ACL modifier called "logwrite" which writes to a log file
5329 as soon as it is encountered.
5331 47. Added $local_user_uid and $local_user_gid at routing time.
5333 48. Exim crashed when trying to verify a sender address that was being
5336 49. Exim was recognizing only a space character after ".include". It now also
5337 recognizes a tab character.
5339 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5340 extracting the relevant information from the specification. The man page no
5341 longer contains scrambled data for the -d option, and I've added a section
5342 at the front about calling Exim under different names.
5344 51. Added "extra_headers" argument to the "mail" command in filter files.
5346 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5349 53. Installed eximstats 1.29.
5351 54. Added transport_filter_timeout as a generic transport option.
5353 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5354 Cc: header lines. This was required by RFC 822, but it not required by RFC
5357 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5358 incoming messages that did not have them. Now it does so only if the
5359 message originates locally, that is, if there is no associated remote host
5360 address. When Resent- header lines are present, this applies to the Resent-
5361 lines rather than the non-Resent- lines.
5363 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5364 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5366 58. Messages for configuration errors now include the name of the main
5367 configuration file - useful now that there may be more than one file in a
5368 list (.included file names were always shown).
5370 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5371 for those rare installations that do not start the daemon as root or run it
5372 setuid root. I've cut out the call to initgroups() if the daemon is not
5375 60. The Exim user and group can now be bound into the binary as text strings
5376 that are looked up at the start of Exim's processing.
5378 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5380 62. Added $mailstore_basename variable.
5382 63. Installed patch to sieve.c from Michael Haardt.
5384 64. When Exim failed to open the panic log after failing to open the main log,
5385 the original message it was trying to log was written to stderr and debug
5386 output, but if they were not available (the usual case in production), it
5387 was lost. Now it is written to syslog before the two lines that record the
5388 failures to open the logs.
5390 65. Users' Exim filters run in subprocesses under the user's uid. It is
5391 possible for a "deliver" command or an alias in a "personal" command to
5392 provoke an address rewrite. If logging of address rewriting is configured,
5393 this fails because the process is not running as root or exim. There may be
5394 a better way of dealing with this, but for the moment (because 4.30 needs
5395 to be released), I have disabled address rewrite logging when running a
5396 filter in a non-root, non-exim process.
5402 1. The buildconfig auxiliary program wasn't quoting the value set for
5403 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5404 not defined. This bug was masked in 4.22 by the effect that was fixed in
5407 2. Some messages that were rejected after a message id was allocated were
5408 shown as "incomplete" by exigrep. It no longer does this for messages that
5409 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5411 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5412 have allow_domain_literals set, the ID did not get logged in the <= line.
5413 Domain literals are now always recognized in Message-ID: header lines.
5415 4. The first argument for a ${extract expansion item is the key name or field
5416 number. Leading and trailing spaces in this item were not being ignored,
5417 causing some misleading effects.
5419 5. When deliver_drop_privilege was set, single queue runner processes started
5420 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5421 same command in the process it spins off) were not dropping privilege.
5423 6. When the daemon running as "exim" started a queue runner, it always
5424 re-executed Exim in the spun-off process. This is a waste of effort when
5425 deliver_drop_privilege is set. The new process now just calls the
5426 queue-runner function directly.
5432 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5435 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5436 ignored. Though the use of -oP was forcing the writing of a pid file, it
5437 was always written to the default place.
5439 3. If the message "no IP address found for host xxxx" is generated during
5440 incoming verification, it is now followed by identification of the incoming
5441 connection (so you can more easily find what provoked it).
5443 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5445 5. Added some features to "harden" Exim a bit more against certain attacks:
5447 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5448 be put in Local/Makefile. This is like the never_users runtime option,
5449 but it cannot be overridden. The default setting is "root".
5451 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5452 prefix string with which any file named in a -C command line option
5455 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5456 is retained for -C and -D only if the caller of Exim is root. Without
5457 it, the exim user may also use -C and -D and retain privilege.
5459 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5460 command line option is disabled.
5462 6. Macro names set by the -D option must start with an upper case letter, just
5463 like macro names defined in the configuration file.
5465 7. Added "dereference=" facility to LDAP.
5467 8. Two instances of the typo "uknown" in the source files are fixed.
5469 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5470 the Configure-Makefile script screwed up while processing it.
5472 10. Incorporated PCRE 4.4.
5474 11. The SMTP synchronization check was not operating right at the start of an
5475 SMTP session. For example, it could not catch a HELO sent before the client
5476 waited for the greeting. There is now a check for outstanding input at the
5477 point when the greeting is written. Because of the duplex, asynchronous
5478 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5479 way, but not yet received, when the check is performed.
5481 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5482 on TCP/IP sockets, because this apparently causes some broken clients to
5485 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5486 unchanged) from the Cygwin maintainer.
5488 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5489 was supported instead of testing for mbx. Effectively a typo.
5491 15. The spa authenticator server code was not checking that the input it
5492 received was valid base64.
5494 16. The debug output line for the "set" modifier in ACLs was not showing the
5495 name of the variable that was being set.
5497 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5499 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5500 lookup on its own. Something else was needed to trigger the lookup. For
5501 example, a match in host_lookup or the need for a host name in a host list.
5502 Now, if $sender_host_name is referenced and the host name has not yet been
5503 looked up, a lookup is performed. If the lookup fails, the variable remains
5504 empty, and $host_lookup_failed is set to "1".
5506 19. Added "eqi" as a case-independent comparison operator.
5508 20. The saslauthd authentication condition could segfault if neither service
5509 nor realm was specified.
5511 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5512 error message that was logged was misleading, and incoming SMTP
5513 connections were dropped. The message is now more accurate, and temporary
5514 errors are given to SMTP connections.
5516 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5517 (except RSET) with a 421 error, until QUIT is received. However, it was
5518 failing to send a response to QUIT.
5520 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5521 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5522 ACL is now run after the helo_try_verify_hosts code.
5524 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5525 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5526 case-independent, so other case variants are also recognized). Apparently
5527 some systems use these upper case variants.
5529 25. If more than two messages were waiting for the same host, and a transport
5530 filter was specified for the transport, Exim sent two messages over the
5531 same TCP/IP connection, and then failed with "socket operation on non-
5532 socket" when it tried to send the third.
5534 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5536 27. The extern definition of crypt16() in expand.c was not being excluded when
5537 the OS had its own crypt16() function.
5539 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5540 as a preferred synonym for return_size_limit, both as an option and as an
5543 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5545 30. Changed the default configuration ACL to relax the local part checking rule
5546 for addresses that are not in any local domains. For these addresses,
5547 slashes and pipe symbols are allowed within local parts, but the sequence
5548 /../ is explicitly forbidden.
5550 31. SPA server authentication was not clearing the challenge buffer before
5553 32. log_message in a "warn" ACL statement was writing to the reject log as
5554 well as to the main log, which contradicts the documentation and doesn't
5555 seem right (because no rejection is happening). So I have stopped it.
5557 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5558 However, I am unable to do any testing of this.
5560 34. Fixed an infelicity in the appendfile transport. When checking directories
5561 for a mailbox, to see if any needed to be created, it was accidentally
5562 using path names with one or more superfluous leading slashes; tracing
5563 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5565 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5566 contents are added to the log line that is written for every discarded
5567 recipient. (Previously a log_message setting was ignored.)
5569 36. The ${quote: operator now quotes the string if it is empty.
5571 37. The install script runs exim in order to find its version number. If for
5572 some reason other than non-existence or emptiness, which it checks, it
5573 could not run './exim', it was installing it with an empty version number,
5574 i.e. as "exim-". This error state is now caught, and the installation is
5577 38. An argument was missing from the function that creates an error message
5578 when Exim fails to connect to the socket for saslauthd authentication.
5579 This could cause Exim to crash, or give a corrupted message.
5581 39. Added isip, isip4, and isip6 to ${if conditions.
5583 40. The ACL variables $acl_xx are now saved with the message, and can be
5584 accessed later in routers, transports, and filters.
5586 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5587 strings in the file are not string-expanded.
5589 42. If a MAIL command specified a SIZE value that was too large to fit into an
5590 int variable, the check against message_size_limit failed. Such values are
5591 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5592 day this will have to be increased, but I don't think I want to be around
5593 when emails are that large.
5600 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5601 iconv() is not standard in FreeBSD.
5603 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5604 IPv6 enabled. The observed symptom was a segmentation fault on return from
5605 the function os_common_find_running_interfaces() in src/os.c.
5607 3. In the check_special_case() function in daemon.c I had used "errno" as an
5608 argument name, which causes warnings on some systems. This was basically a
5609 typo, since it was named "eno" in the comments!
5611 4. The code that waits for the clock to tick (at a resolution of some fraction
5612 of a second) so as to ensure message-id uniqueness was always waiting for
5613 at least one whole tick, when it could have waited for less. [This is
5614 almost certainly not relevant at current processor speeds, where it is
5615 unlikely to ever wait at all. But we try to future-proof.]
5617 5. The function that sleeps for a time interval that includes fractions of a
5618 second contained a race. It did not block SIGALRM between setting the
5619 timer, and suspending (a couple of lines later). If the interval was short
5620 and the sigsuspend() was delayed until after it had expired, the suspension
5621 never ended. On busy systems this could lead to processes getting stuck for
5624 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5625 process, before it forks any delivery processes. The open lookup caching
5626 mechanism meant that the open file or database connection was passed into
5627 the delivery process. The problem was that delivery processes always tidy
5628 up cached lookup data. This could cause a problem for the next delivery
5629 process started by the queue runner, because the external queue runner
5630 process does not know about the closure. So the next delivery process
5631 still has data in the lookup cache. In the case of a file lookup, there was
5632 no problem because closing a file descriptor in a subprocess doesn't affect
5633 the parent. However, if the lookup was caching a connection to a database,
5634 the connection was closed, and the second delivery process was likely to
5635 see errors such as "PGSQL: query failed: server closed the connection
5636 unexpectedly". The problem has been fixed by closing all cached lookups
5637 in a queue runner before running a delivery process.
5639 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5640 seem to have the "const" qualifier which it has on other OS. I've
5643 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5644 *of the same type* (client or server) with the same public name that an
5645 error should be diagnosed.
5647 9. When Exim looked up a host name for an IP address, but failed to find the
5648 original IP address when looking up the host name (a safety check), it
5649 output the message "<ip address> does not match any IP for NULL", which was
5650 confusing, to say the least. The bug was that the host name should have
5651 appeared instead of "NULL".
5653 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5654 user that is built into the binary, and the -C or -D options is used, root
5655 privilege is dropped before the configuration file is read. In addition,
5656 logging is switched to stderr instead of the normal log files. If the
5657 configuration then re-defines the Exim user, the unprivileged environment
5658 is probably not what is expected, so Exim logs a panic warning message (but
5661 However, if deliver_drop_privilege is set, the unprivileged state may well
5662 be exactly what is intended, so the warning has been cut out in that case,
5663 and Exim is allowed to try to write to its normal log files.
5669 1. smtp_return_error_details was not giving details for temporary sender
5670 or receiver verification errors.
5672 2. Diagnose a configuration error if two authenticators have the same public
5675 3. Exim used not to create the message log file for a message until the first
5676 delivery attempt. This could be confusing when incoming messages were held
5677 for policy or load reasons. The message log file is now created at the time
5678 the message is received, and an initial "Received" line is written to it.
5680 4. The automatically generated man page for command line options had a minor
5681 bug that caused no ill effects; however, a more serious problem was that
5682 the procedure for building the man page automatically didn't always
5683 operate. Consequently, release 4.20 contains an out-of-date version. This
5684 shouldn't happen again.
5686 5. When building Exim with embedded Perl support, the script that builds the
5687 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5688 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5690 6. The freeze_tell option was not being used for messages that were frozen on
5691 arrival, either by an ACL or by local_scan().
5693 7. Added the smtp_incomplete_transaction log selector.
5695 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5696 was accepting AUTH without a new EHLO.
5698 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5699 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5702 10. Logging of TCP/IP connections (when configured) now happens in the main
5703 daemon process instead of the child process, so that the TCP/IP connection
5704 count is more accurate (but it can never be perfect).
5706 11. The use of "drop" in a nested ACL was not being handled correctly in the
5707 outer ACL. Now, if condition failure induced by the nested "drop" causes
5708 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5709 or "require"), the connection is dropped.
5711 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5712 that yield "discard" can now be used with an "accept" or a "discard" verb,
5713 but an error is generated for any others (because I can't see a useful way
5714 to define what should happen).
5716 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5717 are now processed in the same way as lines in the Exim configuration file.
5718 In particular, continuation lines are supported.
5720 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5722 15. Added -ti meaning -t -i.
5724 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5725 domains, and warn by logging if others are found.
5727 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5728 returned from GIFCONF when Exim is trying to find the list of interfaces on
5729 a host. The code in os.c has been modified to copy each ifreq to an aligned
5730 structure in all cases.
5732 Also, in some cases, the returned ifreq's were being copied to a 'struct
5733 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5734 means the last couple of bytes of an IPv6 address could be chopped if the
5735 ifreq contained only a normal sockaddr (14 bytes storage).
5737 18. Named domain lists were not supported in the hosts_treat_as_local option.
5738 An entry such as +xxxx was not recognized, and was treated as a literal
5741 19. Ensure that header lines added by a DATA ACL are included in the reject log
5742 if the ACL subsequently rejects the message.
5744 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5745 MD5 (which is deprecated).
5747 21. When testing a filter file using -bf, Exim was writing a message when it
5748 took the sender from a "From " line in the message, but it was not doing so
5749 when it took $return_path from a Return-Path: header line. It now does.
5751 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5752 with a valid header line field name (a series of printing characters
5753 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5755 23. Changed "disc" in the source to "disk" to conform to the documentation and
5756 the book and for uniformity.
5758 24. Ignore Sendmail's -Ooption=value command line item.
5760 25. When execve() failed while trying to run a command in a pipe transport,
5761 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5762 could be confused with a return value of 69 from the command itself. This
5763 has been changed to 127, the value the shell returns if it is asked to run
5764 a non-existent command. The wording for the related log line suggests a
5765 non-existent command as the problem.
5767 26. If received_header_text expands to an empty string, do not add a Received:
5768 header line to the message. (Well, it adds a token one on the spool, but
5769 marks it "old" so that it doesn't get used or transmitted.)
5771 27. Installed eximstats 1.28 (addition of -nt option).
5773 28. There was no check for failure on the call to getsockname() in the daemon
5774 code. This can fail if there is a shortage of resources on the system, with
5775 ENOMEM, for example. A temporary error is now given on failure.
5777 29. Contrary to the C standard, it seems that in some environments, the
5778 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5779 program. Exim now does this explicitly; it affects the formatting of
5780 timestamps using strftime().
5782 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5783 complaints. I've now initialized all the variables, to avoid this.
5785 32. Header lines added by a system filter were not being "seen" during
5786 transport-time rewrites.
5788 33. The info_callback() function passed to OpenSSL is set up with type void
5789 (*)(SSL *, int, int), as described somewhere. However, when calling the
5790 function (actually a macro) that sets it up, the type void(*)() is
5791 expected. I've put in a cast to prevent warnings from picky compilers.
5793 34. If a DNS black list lookup found a CNAME record, but there were no A
5794 records associated with the domain it pointed at, Exim crashed.
5796 35. If a DNS black list lookup returned more than one A record, Exim ignored
5797 all but the first. It now scans all returned addresses if a particular IP
5798 value is being sought. In this situation, the contents of the
5799 $dnslist_value variable are a list of all the addresses, separated by a
5802 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5803 to accept a host name and all its aliases if the forward lookup for any of
5804 them yielded the IP address of the incoming connection. Now it accepts only
5805 those names whose forward lookup yields the correct IP address. Any other
5806 names are discarded. This closes a loophole whereby a rogue DNS
5807 administrator could create reverse DNS records to break through a
5808 wildcarded host restriction in an ACL.
5810 37. If a user filter or a system filter that ran in a subprocess used any of
5811 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5812 the wrong values were passed to the pipe command ($thisaddress had the
5813 value of $0, $0 had the value of $1, etc). This bug was introduced by
5814 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5816 38. Improved the line breaking for long SMTP error messages from ACLs.
5817 Previously, if there was no break point between 40 and 75 characters, Exim
5818 left the rest of the message alone. Two changes have been made: (a) I've
5819 reduced the minimum length to 35 characters; (b) if it can't find a break
5820 point between 35 and 75 characters, it looks ahead and uses the first one
5821 that it finds. This may give the occasional overlong line, but at least the
5822 remaining text gets split now.
5824 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5825 file descriptors might be low, and that setting 1000 would always raise it.
5826 It turns out that in some environments, the limit is already over 1000 and
5827 that lowering it causes trouble. So now Exim takes care not to decrease it.
5829 40. When delivering a message, the value of $return_path is set to $sender_
5830 address at the start of routing (routers may change the value). By an
5831 oversight, this default was not being set up when an address was tested by
5832 -bt or -bv, which affected the outcome if any router or filter referred to
5835 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5836 log only when "message" or "log_message" are set. However, if one of the
5837 conditions was an address verification, or a call to a nested ACL, the
5838 messages generated by the underlying test were being passed through. This
5839 no longer happens. The underlying message is available in $acl_verify_
5840 message for both "message" and "log_message" expansions, so it can be
5841 passed through if needed.
5843 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5844 new expansion $bh_ to give the encoded byte string without charset
5845 translation. Translation happens only if iconv() is available; HAVE_ICONV
5846 indicates this at build time. HEADERS_CHARSET gives the charset to
5847 translate to; headers_charset can change it in the configuration, and
5848 "headers charset" can change it in an individual filter file.
5850 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5851 that creates RFC 2047 encoded "words" labels them as that charset instead
5852 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5853 expansion operator; (ii) when Exim creates a From: line for a local
5854 message; (iii) when a header line is rewritten to include a "phrase" part.
5856 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5857 buggy, causing it to skip the first lines of messages whose message ID
5858 ended in 'D'. This would not have bitten before Exim release 4.14, because
5859 message IDs were unlikely to end in 'D' before then. The effect was to have
5860 incorrect size information for certain domains.
5862 45. #include "config.h" was missing at the start of the crypt16.c module. This
5863 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5866 46. If there was a timeout during a "random" callout check, Exim treated it as
5867 a failure of the random address, and carried on sending RSET and the real
5868 address. If the delay was just some slowness somewhere, the response to the
5869 original RCPT would be taken as a response to RSET and so on, causing
5870 mayhem of various kinds.
5872 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5873 when I implemented it. It didn't allow for the fact that some option values
5874 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5875 the right test for positive values.
5877 48. Domain names in DNS records are case-independent. Exim always looks them up
5878 in lower case. Some resolvers return domain names in exactly the case they
5879 appear in the zone file, that is, they may contain uppercase letters. Not
5880 all resolvers do this - some return always lower case. Exim was treating a
5881 change of case by a resolver as a change of domain, similar to a widening
5882 of a domain abbreviation. This triggered its re-routing code and so it was
5883 trying to route what was effectively the same domain again. This normally
5884 caused routing to fail (because the router wouldn't handle the domain
5885 twice). Now Exim checks for this case specially, and just changes the
5886 casing of the domain that it ultimately uses when it transmits the message
5889 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5892 50. If a filter generated a file delivery with a non-absolute name (possible if
5893 no home directory exists for the router), the forbid_file option was not
5896 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5897 the existing equality matching.
5899 52. Exim was using ints instead of ino_t variables in some places where it was
5900 dealing with inode numbers.
5902 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5903 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5904 called TMPDIR, and if it finds it is different, it changes its value.
5906 54. The smtp_printf() function is now made available to local_scan() so
5907 additional output lines can be written before returning. There is also an
5908 smtp_fflush() function to enable the detection of a dropped connection.
5909 The variables smtp_input and smtp_batched_input are exported to
5912 55. Changed the default runtime configuration: the message "Unknown user"
5913 has been removed from the ACL, and instead placed on the localuser router,
5914 using the cannot_route_message feature. This means that any verification
5915 failures that generate their own messages won't get overridden. Similarly,
5916 the "Unrouteable address" message that was in the ACL for unverifiable
5917 relay addresses has also been removed.
5919 56. Added hosts_avoid_esmtp to the smtp transport.
5921 57. The exicyclog script was not checking for the esoteric option
5922 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5923 will work only if exicyclog is run under the appropriate euid.
5925 58. Following a discussion on the list, the rules by which Exim recognises line
5926 endings on incoming messages have been changed. The -dropcr and drop_cr
5927 options are now no-ops, retained only for backwards compatibility. The
5928 following line terminators are recognized: LF CRLF CR. However, special
5929 processing applies to CR:
5931 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5932 nor a local message in the state where . is a terminator.
5934 (ii) If a bare CR is encountered in a header line, an extra space is added
5935 after the line terminator so as not to end the header. The reasoning
5936 behind this is that bare CRs in header lines are most likely either
5937 to be mistakes, or people trying to play silly games.
5939 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5940 was being incorrectly given as 18 bytes larger than it should have been.
5941 This is a VOB (very old bug).
5943 60. This may never have affected anything current, but just in case it has:
5944 When the local host is found other than at the start of a list of hosts,
5945 the local host, those with the same MX, and any that follow, are discarded.
5946 When the list in question was part of a longer list of hosts, the following
5947 hosts (not currently being processed) were also being discarded. This no
5948 longer happens. I'm not sure if this situation could ever has previously
5951 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5954 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5955 header. This is something that is recommended in a new Internet Draft, and
5956 is something that is documented as being done by Sendmail. There are two
5957 possible values. For messages generated by the autoreply transport, Exim
5960 Auto-Submitted: auto-replied
5962 whereas for all other generated messages (e.g. bounces) it adds
5964 Auto-Submitted: auto-generated
5966 63. The "personal" condition in filters now includes a test for the
5967 Auto-Submitted: header. If it contains the string "auto-" the message it
5968 not considered personal.
5970 64. Added rcpt_include_affixes as a generic transport option.
5972 65. Added queue_only_override (default true).
5974 66. Added the syslog_duplication option.
5976 67. If what should have been the first header line of a message consisted of
5977 a space followed by a colon, Exim was mis-interpreting it as a header line.
5978 It isn't of course - it is syntactically invalid and should therefore be
5979 treated as the start of the message body. The misbehaviour could have
5980 caused a number of strange effects, including loss of data in subsequent
5981 header lines, and spool format errors.
5983 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5984 client host had authenticated. This control can now be exercised by an ACL
5985 for more flexibility.
5987 69. By default, callouts do not happen when testing with -bh. There is now a
5988 variant, -bhc, which does actually run the callout code, including
5989 consulting and updating the callout cache.
5991 70. Added support for saslauthd authentication, courtesy of Alexander
5994 71. If statvfs() failed on the spool or log directories while checking their
5995 size for availability, Exim confusingly gave the error "space shortage".
5996 Furthermore, in debugging mode it crashed with a floating point exception.
5997 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5998 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5999 serious problem, Exim now writes to the main and panic logs when this
6000 happens, with details of the failure. It then refuses to accept the
6001 incoming message, giving the message "spool directory problem" or "log
6002 directory problem" with a 421 code for SMTP messages.
6004 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6005 0, 1, and 2 exist, because some OS complain for execs without them (see
6006 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6007 descriptors. However, the code omitted to check that the open succeeded,
6008 causing mysterious errors if for some reason the permissions on /dev/null
6009 got screwed. Now Exim writes a message to the main and panic logs, and
6010 bombs out if it can't open /dev/null.
6012 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6013 interact so that it is all more flexible. It is supposed to remain
6014 backwards compatible. Also added extra_local_interfaces.
6016 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6017 to bomb out with an assertion failure - to the client this appears as a
6018 connection drop. This problem occurs in the part of the code that was taken
6019 from the Samba project. Fortunately, the assertion is in a very simple
6020 function, so I have fixed this by reproducing the function inline in the
6021 one place where it is called, and arranging for authentication to fail
6022 instead of killing the process with assert().
6024 75. The SPA client code was not working when the server requested OEM rather
6025 than Unicode encoding.
6027 76. Added code to make require_files with a specific uid setting more usable in
6028 the case where statting the file as root fails - usually a non-root-mounted
6029 NFS file system. When this happens and the failure is EACCES, Exim now
6030 forks a subprocess and does the per-uid checking as the relevant uid.
6032 77. Added process_log_path.
6034 78. If log_file_path was not explicitly set, a setting of check_log_space or
6035 check_log_inodes was ignored.
6037 79. If a space check for the spool or log partitions fails, the incident is now
6038 logged. Of course, in the latter case the data may get lost...
6040 80. Added the %p formatting code to string_format() so that it can be used to
6041 print addresses in debug_print(). Adjusted all the address printing in the
6042 debugging in store.c to use %p rather than %d.
6044 81. There was a concern that a line of code in smtp_in.c could overflow a
6045 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6046 initially expressed, the concern was not well-founded, because trailing
6047 spaces are removed early. However, if the trailing spaces were followed by
6048 a NULL, they did not get removed, so the overflow was possible. Two fixes
6051 (a) I re-wrote the offending code in a cleaner fashion.
6052 (b) If an incoming SMTP command contains a NULL character, it is rejected
6055 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6056 runs initgroups(), so that if the Exim user is in any additional groups,
6057 they will be used during message reception.
6063 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob