1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 TL/01 When running the test suite, the README says that variables such as
8 no_msglog_check are global and can be placed anywhere in a specific
9 test's script, however it was observed that placement needed to be near
10 the beginning for it to behave that way. Changed the runtest perl
11 script to read through the entire script once to detect and set these
12 variables, reset to the beginning of the script, and then run through
13 the script parsing/test process like normal.
15 TL/02 The BSD's have an arc4random API. One of the functions to induce
16 adding randomness was arc4random_stir(), but it has been removed in
17 OpenBSD 5.5. Detect this OpenBSD version and skip calling this
18 function when detected.
20 JH/01 Expand the EXPERIMENTAL_TPDA feature. Several different events now
21 cause callback expansion.
26 TL/01 Bugzilla 1506: Re-add a 'return NULL' to silence complaints from static
27 checkers that were complaining about end of non-void function with no
30 JH/01 Bug 1513: Fix parsing of quoted parameter values in MIME headers.
31 This was a regression intruduced in 4.83 by another bugfix.
33 JH/02 Fix broken compilation when EXPERIMENTAL_DSN is enabled.
35 TL/02 Bug 1509: Fix exipick for enhanced spoolfile specification used when
36 EXPERIMENTAL_DSN is enabled. Fix from Wolfgang Breyha.
42 TF/01 Correctly close the server side of TLS when forking for delivery.
44 When a message was received over SMTP with TLS, Exim failed to clear up
45 the incoming connection properly after forking off the child process to
46 deliver the message. In some situations the subsequent outgoing
47 delivery connection happened to have the same fd number as the incoming
48 connection previously had. Exim would try to use TLS and fail, logging
49 a "Bad file descriptor" error.
51 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
52 utilities have not been installed.
54 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
55 temporary space as the ACL may create new global variables.
57 TL/01 LDAP support uses per connection or global context settings, depending
58 upon the detected version of the libraries at build time.
60 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
61 to extract and use the src ip:port in logging and expansions as if it
62 were a direct connection from the outside internet. PPv2 support was
63 updated based on HAProxy spec change in May 2014.
65 JH/02 Add ${listextract {number}{list}{success}{fail}}.
67 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
68 Properly escape header and check for NULL return.
70 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
73 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
75 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
76 characters in header names, implemented as a verify condition.
77 Contributed by Michael Fischer v. Mollard.
79 TL/05 Rename SPF condition results err_perm and err_temp to standardized
80 results permerror and temperror. Previous values are deprecated but
81 still accepted. In a future release, err_perm and err_temp will be
82 completely removed, which will be a backward incompatibility if the
83 ACL tests for either of these two old results. Patch contributed by
84 user bes-internal on the mailing list.
86 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
88 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
89 selectors, in both main and reject logs.
91 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
94 JH/07 Add malware type "sock" for talking to simple daemon.
96 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
98 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
99 routers/transports under cutthrough routing.
101 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
102 numbers. Touch up "bool" conditional to keep the same definition.
104 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
106 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
108 JH/12 Expand items in router/transport headers_add or headers_remove lists
109 individually rather than the list as a whole. Bug 1452.
111 Required for reasonable handling of multiple headers_ options when
112 they may be empty; requires that headers_remove items with embedded
113 colons must have them doubled (or the list-separator changed).
115 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
116 view the policy declared in the DMARC record. Currently, $dmarc_status
117 is a combined value of both the record presence and the result of the
120 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
122 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
123 dnslookup router and the smtp transport (applying to the forward
126 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
127 of ldap servers used for a specific lookup. Patch provided by Heiko
130 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
131 New variable $lookup_dnssec_authenticated for observability.
133 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
134 Patch submitted by Lars Timman.
136 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
138 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
139 Requires trusted mode and valid format message id, aborts otherwise.
140 Patch contributed by Heiko Schlichting.
142 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
143 certextract with support for various fields. Bug 1358.
145 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
146 is requested by default, modifiable by smtp transport option
149 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
150 operate on certificate variables to give certificate fingerprints
151 Also new ${sha256:cert_variable}.
153 JH/23 The PRDR feature is moved from being Experimental into the mainline.
155 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
156 Christian Aistleitner.
158 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
160 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
161 file. Patch from Wolfgang Breyha.
163 JH/25 Expand the coverage of the delivery $host and $host_address to
164 client authenticators run in verify callout. Bug 1476.
166 JH/26 Port service names are now accepted for tls_on_connect_ports, to
167 align with daemon_smtp_ports. Bug 72.
169 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
170 support and error reporting did not work properly.
172 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
173 and is readable. Patch from Andrew Colin Kissa.
175 TL/14 Enhance documentation of ${run expansion and how it parses the
176 commandline after expansion, particularly in the case when an
177 unquoted variable expansion results in an empty value.
179 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
181 PP/02 Fix internal collision of T_APL on systems which support RFC3123
182 by renaming away from it. Addresses GH issue 15, reported by
185 JH/28 Fix parsing of MIME headers for parameters with quoted semicolons.
187 TL/15 SECURITY: prevent double expansion in math comparison functions
188 (can expand unsanitized data). Not remotely exploitable.
195 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
197 PP/02 Make -n do something, by making it not do something.
198 When combined with -bP, the name of an option is not output.
200 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
203 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
204 $sender_host_name and config options to manage this, and basic check
207 PP/05 DSCP support for outbound connections and control modifier for inbound.
209 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
210 (Only plugin which currently uses this is kerberos4, which nobody should
211 be using, but we should make it available and other future plugins might
212 conceivably use it, even though it would break NAT; stuff *should* be
213 using channel bindings instead).
215 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
216 name; added for Sendmail compatibility; requires admin caller.
217 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
218 just ignore it); requires trusted caller.
219 Also parse but ignore: -Ac -Am -X<logfile>
222 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
224 TL/02 Add +smtp_confirmation as a default logging option.
226 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
227 Patch by Magnus Holmgren from 2007-02-20.
229 TL/04 Bugzilla 1281 - Spec typo.
230 Bugzilla 1283 - Spec typo.
231 Bugzilla 1290 - Spec grammar fixes.
233 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
235 TL/06 Add Experimental DMARC support using libopendmarc libraries.
237 TL/07 Fix an out of order global option causing a segfault. Reported to dev
238 mailing list by by Dmitry Isaikin.
240 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
242 JH/02 Support "G" suffix to numbers in ${if comparisons.
244 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
246 NM/01 Bugzilla 1197 - Spec typo
247 Bugzilla 1196 - Spec examples corrections
249 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
251 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
252 gnutls_enable_pkcs11, but renamed to more accurately indicate its
255 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
256 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
258 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
259 "acl {{name}{arg}...}", and optional args on acl condition
262 JH/05 Permit multiple router/transport headers_add/remove lines.
264 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
266 JH/07 Avoid using a waiting database for a single-message-only transport.
267 Performance patch from Paul Fisher. Bugzilla 1262.
269 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
272 JH/09 Add $headers_added variable, with content from use of ACL modifier
273 add_header (but not yet added to the message). Bugzilla 199.
275 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
276 Pulled from Bugzilla 817 by Wolfgang Breyha.
278 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
280 (nb: this is the same fix as in Exim 4.80.1)
282 JH/11 Add A= logging on delivery lines, and a client_set_id option on
285 JH/12 Add optional authenticated_sender logging to A= and a log_selector
288 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
290 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
291 advertise SMTP AUTH mechanism to us, instead of a generic
292 protocol violation error. Also, make Exim more robust to bad
293 data from the Dovecot auth socket.
295 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
297 When a queue runner is handling a message, Exim first routes the
298 recipient addresses, during which it prunes them based on the retry
299 hints database. After that it attempts to deliver the message to
300 any remaining recipients. It then updates the hints database using
303 So if a recipient address works intermittently, it can get repeatedly
304 deferred at routing time. The retry hints record remains fresh so the
305 address never reaches the final cutoff time.
307 This is a fairly common occurrence when a user is bumping up against
308 their storage quota. Exim had some logic in its local delivery code
309 to deal with this. However it did not apply to per-recipient defers
310 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
312 This change adds a proper retry rule check during routing so that the
313 final cutoff time is checked against the message's age. We only do
314 this check if there is an address retry record and there is not a
315 domain retry record; this implies that previous attempts to handle
316 the address had the retry_use_local_parts option turned on. We use
317 this as an approximation for the destination being like a local
318 delivery, as in LMTP.
320 I suspect this new check makes the old local delivery cutoff check
321 redundant, but I have not verified this so I left the code in place.
323 TF/02 Correct gecos expansion when From: is a prefix of the username.
325 Test 0254 submits a message to Exim with the header
329 When I ran the test suite under the user fanf2, Exim expanded
330 the header to contain my full name, whereas it should have added
331 a Resent-Sender: header. It erroneously treats any prefix of the
332 username as equal to the username.
334 This change corrects that bug.
336 GF/01 DCC debug and logging tidyup
337 Error conditions log to paniclog rather than rejectlog.
338 Debug lines prefixed by "DCC: " to remove any ambiguity.
340 TF/03 Avoid unnecessary rebuilds of lookup-related code.
342 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
343 Bug spotted by Jeremy Harris; was flawed since initial commit.
344 Would have resulted in OCSP responses post-SNI triggering an Exim
345 NULL dereference and crash.
347 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
349 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
350 Bug detection, analysis and fix by Samuel Thibault.
351 Bugzilla 1331, Debian bug #698092.
353 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
355 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
356 Server implementation by Todd Lyons, client by JH.
357 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
358 config variable "prdr_enable" controls whether the server
359 advertises the facility. If the client requests PRDR a new
360 acl_data_smtp_prdr ACL is called once for each recipient, after
361 the body content is received and before the acl_smtp_data ACL.
362 The client is controlled by bolth of: a hosts_try_prdr option
363 on the smtp transport, and the server advertisement.
364 Default client logging of deliveries and rejections involving
365 PRDR are flagged with the string "PRDR".
367 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
368 fclose(). Diagnosis by Todd Lyons.
370 PP/17 Update configure.default to handle IPv6 localhost better.
371 Patch by Alain Williams (plus minor tweaks).
374 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
375 This is now consistent with GnuTLS, and is now documented: the
376 previous undocumented portable approach to treating the option as
377 unset was to force an expansion failure. That still works, and
378 an empty string is now equivalent.
380 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
381 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
382 not performing validation itself.
384 PP/20 Added force_command boolean option to pipe transport.
385 Patch from Nick Koston, of cPanel Inc.
387 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
390 TF/04 Added udpsend ACL modifer and hexquote expansion operator
392 PP/21 Fix eximon continuous updating with timestamped log-files.
393 Broken in a format-string cleanup in 4.80, missed when I repaired the
394 other false fix of the same issue.
395 Report and fix from Heiko Schlichting.
398 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
399 Report from Prashanth Katuri.
401 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
402 It's SecureTransport, so affects any MacOS clients which use the
403 system-integrated TLS libraries, including email clients.
405 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
406 using a MIME ACL for non-SMTP local injection.
407 Report and assistance in diagnosis by Warren Baker.
409 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
411 JH/16 Fix comparisons for 64b. Bugzilla 1385.
413 TL/09 Add expansion variable $authenticated_fail_id to keep track of
414 last id that failed so it may be referenced in subsequent ACL's.
416 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
419 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
420 ldap library initialization, allowing self-signed CA's to be
421 used. Also properly sets require_cert option later in code by
422 using NULL (global ldap config) instead of ldap handle (per
423 session). Bug diagnosis and testing by alxgomz.
425 TL/12 Enhanced documentation in the ratelimit.pl script provided in
426 the src/util/ subdirectory.
428 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
429 renamed to Transport Post Delivery Action by Jeremy Harris, as
432 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
433 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
434 redis_servers = needs to be configured which will be used by the redis
435 lookup. Patch from Warren Baker, of The Packet Hub.
437 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
439 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
440 hostname or reverse DNS when processing a host list. Used suggestions
441 from multiple comments on this bug.
443 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
445 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
446 Missed a few lines, added it to make the runtest require no keyboard
449 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
450 contains upper case chars. Make router use caseful_local_part.
452 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
453 support when GnuTLS has been built with p11-kit.
459 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
461 This, or similar/improved, will also be change PP/11 of 4.82.
467 PP/01 Handle short writes when writing local log-files.
468 In practice, only affects FreeBSD (8 onwards).
469 Bugzilla 1053, with thanks to Dmitry Isaikin.
471 NM/01 Bugzilla 949 - Documentation tweak
473 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
476 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
478 PP/02 Implemented gsasl authenticator.
480 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
482 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
483 `pkg-config foo` for cflags/libs.
485 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
486 with rest of GSASL and with heimdal_gssapi.
488 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
489 `pkg-config foo` for cflags/libs for the TLS implementation.
491 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
492 properties get this fed in as external SSF. A number of robustness
493 and debugging improvements to the cyrus_sasl authenticator.
495 PP/08 cyrus_sasl server now expands the server_realm option.
497 PP/09 Bugzilla 1214 - Log authentication information in reject log.
498 Patch by Jeremy Harris.
500 PP/10 Added dbmjz lookup type.
502 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
504 PP/12 MAIL args handles TAB as well as SP, for better interop with
505 non-compliant senders.
506 Analysis and variant patch by Todd Lyons.
508 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
509 Bug report from Lars Müller <lars@samba.org> (via SUSE),
510 Patch from Dirk Mueller <dmueller@suse.com>
512 PP/13 tls_peerdn now print-escaped for spool files.
513 Observed some $tls_peerdn in wild which contained \n, which resulted
514 in spool file corruption.
516 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
517 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
518 or write after TLS renegotiation, which otherwise led to messages
521 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
522 as a tracking header (ie: a signed header comes before the signature).
523 Patch from Wolfgang Breyha.
525 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
526 comma-sep list; embedded commas doubled.
528 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
530 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
532 Report and patch from Dmitry Banschikov.
534 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
535 Removed SSL_clear() after SSL_new() which led to protocol negotiation
536 failures. We appear to now support TLS1.1+ with Exim.
538 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
539 lets Exim select keys and certificates based upon TLS SNI from client.
540 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
541 before an outbound SMTP session. New log_selector, +tls_sni.
543 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
544 NULL dereference. Report and patch from Alun Jones.
546 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
547 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
548 Not seeing resolver debug output on NetBSD, but suspect this is a
549 resolver implementation change.
551 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
552 Left warnings. Added "eximon gdb" invocation mode.
554 PP/21 Defaulting "accept_8bitmime" to true, not false.
556 PP/22 Added -bw for inetd wait mode support.
558 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
559 locate the relevant includes and libraries. Made this the default.
561 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
562 Bugzilla 1246, report and most of solution from Tomasz Kusy.
564 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
565 This may cause build issues on older platforms.
567 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
568 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
569 gnutls_require_mac & gnutls_require_protocols (no longer supported).
570 Added SNI support via GnuTLS too.
571 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
573 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
575 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
576 Applied second patch from Janne, implementing suggestion to default
577 multiple-strings-in-record handling to match SPF spec.
579 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
581 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
582 read-only, out of scope).
583 Patch from Wolfgang Breyha, report from Stuart Northfield.
585 PP/29 Fix three issues highlighted by clang analyser static analysis.
586 Only crash-plausible issue would require the Cambridge-specific
587 iplookup router and a misconfiguration.
588 Report from Marcin Mirosław.
590 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
592 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
593 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
594 As part of this, removing so much warning spew let me fix some minor
595 real issues in debug logging.
597 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
598 assignment on my part. Fixed.
600 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
601 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
602 Janne Snabb (who went above and beyond: thank you).
604 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
605 string otherwise requires a connection and a bunch more work and it's
606 relatively easy to get wrong. Should also expose TLS library linkage
609 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
610 64-bit ${eval} (JH/03).
612 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
613 GNU libc to support some of the 64-bit stuff, should not lead to
614 conflicts. Defined before os.h is pulled in, so if a given platform
615 needs to override this, it can.
617 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
618 protection layer was required, which is not implemented.
619 Bugzilla 1254, patch from Wolfgang Breyha.
621 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
622 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
623 tls_dhparam take prime identifiers. Also unbreak combination of
624 OpenSSL+DH_params+TLSSNI.
626 PP/39 Disable SSLv2 by default in OpenSSL support.
632 PP/01 Solaris build fix for Oracle's LDAP libraries.
633 Bugzilla 1109, patch from Stephen Usher.
635 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
637 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
640 TF/02 Fix a couple more cases where we did not log the error message
641 when unlink() failed. See also change 4.74-TF/03.
643 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
644 lock up or crash if it happened to be inside a call to libc when it
645 got a SIGUSR1 from exiwhat.
647 The SIGUSR1 handler appends the current process status to the process
648 log which is later printed by exiwhat. It used to use the general
649 purpose logging code to do this, but several functions it calls are
650 not safe for signals.
652 The new output code in the SIGUSR1 handler is specific to the process
653 log, and simple enough that it's easy to inspect for signal safety.
654 Removing some special cases also simplifies the general logging code.
655 Removing the spurious timestamps from the process log simplifies
658 TF/04 Improved ratelimit ACL condition.
660 The /noupdate option has been deprecated in favour of /readonly which
661 has clearer semantics. The /leaky, /strict, and /readonly update modes
662 are mutually exclusive. The update mode is no longer included in the
663 database key; it just determines when the database is updated. (This
664 means that when you upgrde Exim will forget old rate measurements.)
666 Exim now checks that the per_* options are used with an update mode that
667 makes sense for the current ACL. For example, when Exim is processing a
668 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
669 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
670 must specify per_mail/readonly. If you omit the update mode it defaults to
671 /leaky where that makes sense (as before) or /readonly where required.
673 The /noupdate option is now undocumented but still supported for
674 backwards compatibility. It is equivalent to /readonly except that in
675 ACLs where /readonly is required you may specify /leaky/noupdate or
676 /strict/noupdate which are treated the same as /readonly.
678 A useful new feature is the /count= option. This is a generalization
679 of the per_byte option, so that you can measure the throughput of other
680 aggregate values. For example, the per_byte option is now equivalent
681 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
683 The per_rcpt option has been generalized using the /count= mechanism
684 (though it's more complicated than the per_byte equivalence). When it is
685 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
686 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
687 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
688 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
689 Note that using per_rcpt with a non-readonly update mode in more than
690 one ACL will cause the recipients to be double-counted. (The per_mail
691 and per_byte options don't have this problem.)
693 The handling of very low rates has changed slightly. If the computed rate
694 is less than the event's count (usually one) then this event is the first
695 after a long gap. In this case the rate is set to the same as this event's
696 count, so that the first message of a spam run is counted properly.
698 The major new feature is a mechanism for counting the rate of unique
699 events. The new per_addr option counts the number of different
700 recipients that someone has sent messages to in the last time period. It
701 behaves like per_rcpt if all the recipient addresses are different, but
702 duplicate recipient addresses do not increase the measured rate. Like
703 the /count= option this is a general mechanism, so the per_addr option
704 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
705 example, measure the rate that a client uses different sender addresses
706 with the options per_mail/unique=$sender_address. There are further
707 details in the main documentation.
709 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
711 TF/06 Removed a few PCRE remnants.
713 TF/07 Automatically extract Exim's version number from tags in the git
714 repository when doing development or release builds.
716 PP/02 Raise smtp_cmd_buffer_size to 16kB.
717 Bugzilla 879. Patch from Paul Fisher.
719 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
720 Heavily based on revision 40f9a89a from Simon Arlott's tree.
723 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
725 PP/05 Variable $av_failed, true if the AV scanner deferred.
726 Bugzilla 1078. Patch from John Horne.
728 PP/06 Stop make process more reliably on build failure.
729 Bugzilla 1087. Patch from Heiko Schlittermann.
731 PP/07 Make maildir_use_size_file an _expandable_ boolean.
732 Bugzilla 1089. Patch from Heiko Schlittermann.
734 PP/08 Handle ${run} returning more data than OS pipe buffer size.
735 Bugzilla 1131. Patch from Holger Weiß.
737 PP/09 Handle IPv6 addresses with SPF.
738 Bugzilla 860. Patch from Wolfgang Breyha.
740 PP/10 GnuTLS: support TLS 1.2 & 1.1.
742 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
745 PP/11 match_* no longer expand right-hand-side by default.
746 New compile-time build option, EXPAND_LISTMATCH_RHS.
747 New expansion conditions, "inlist", "inlisti".
749 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
751 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
753 PP/14 fix log_write() format string regression from TF/03.
754 Bugzilla 1152. Patch from Dmitry Isaikin.
760 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
762 PP/02 Harmonised TLS library version reporting; only show if debugging.
763 Layout now matches that introduced for other libraries in 4.74 PP/03.
765 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
767 PP/04 New "dns_use_edns0" global option.
769 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
772 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
773 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
775 TK/01 Updated PolarSSL code to 0.14.2.
776 Bugzilla 1097. Patch from Andreas Metzler.
778 PP/07 Catch divide-by-zero in ${eval:...}.
781 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
784 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
785 format-string attack -- SECURITY: remote arbitrary code execution.
787 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
788 time unintentionally subject to list matching rules, letting the header
789 cause arbitrary Exim lookups (of items which can occur in lists, *not*
790 arbitrary string expansion). This allowed for information disclosure.
792 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
793 INT_MIN/-1 -- value coerced to INT_MAX.
799 NM/01 Workround for PCRE version dependency in version reporting
802 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
803 This fixes portability to compilers other than gcc, notably
804 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
806 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
807 makefiles for portability to HP-UX and POSIX correctness.
809 PP/01 Permit LOOKUP_foo enabling on the make command-line.
810 Also via indented variable definition in the Makefile.
811 (Debugging by Oliver Heesakkers).
813 PP/02 Restore caching of spamd results with expanded spamd_address.
814 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
816 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
817 Improves build reliability. Fix from: Frank Elsner
819 NM/02 Fix wide character breakage in the rfc2047 coding
820 Fixes bug 1064. Patch from Andrey N. Oktyabrski
822 NM/03 Allow underscore in dnslist lookups
823 Fixes bug 1026. Patch from Graeme Fowler
825 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
826 Code patches from Adam Ciarcinski of NetBSD.
828 NM/04 Fixed exiqgrep to cope with mailq missing size issue
831 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
832 is logged, to avoid truncation. Patch from John Horne.
834 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
835 Patch from Jakob Hirsch.
837 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
838 SQL string expansion failure details.
839 Patch from Andrey Oktyabrski.
841 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
842 Patch from Simon Arlott.
844 PP/09 New lookups functionality failed to compile on old gcc which rejects
845 extern declarations in function scope.
846 Patch from Oliver Fleischmann
848 PP/10 Use sig_atomic_t for flags set from signal handlers.
849 Check getgroups() return and improve debugging.
850 Fixed developed for diagnosis in bug 927 (which turned out to be
853 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
854 Patch from Mark Zealey.
856 PP/12 Bugzilla 1056: Improved spamd server selection.
857 Patch from Mark Zealey.
859 PP/13 Bugzilla 1086: Deal with maildir quota file races.
860 Based on patch from Heiko Schlittermann.
862 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
863 Patch from Uwe Doering, sign-off by Michael Haardt.
865 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
866 variable declaration deep within a block. Bug and patch from
869 PP/15 lookups-Makefile IRIX compatibilty coercion.
871 PP/16 Make DISABLE_DKIM build knob functional.
873 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
874 Patch by Simon Arlott
876 TF/03 Fix valgrind.h portability to C89 compilers that do not support
877 variable argument macros. Our copy now differs from upstream.
883 TF/01 Failure to get a lock on a hints database can have serious
884 consequences so log it to the panic log.
886 TF/02 Log LMTP confirmation messages in the same way as SMTP,
887 controlled using the smtp_confirmation log selector.
889 TF/03 Include the error message when we fail to unlink a spool file.
891 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
892 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
893 for maintaining out-of-tree patches for some time.
895 PP/01 Bugzilla 139: Documentation and portability issues.
896 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
897 Handle per-OS dynamic-module compilation flags.
899 PP/02 Let /dev/null have normal permissions.
900 The 4.73 fixes were a little too stringent and complained about the
901 permissions on /dev/null. Exempt it from some checks.
902 Reported by Andreas M. Kirchwitz.
904 PP/03 Report version information for many libraries, including
905 Exim version information for dynamically loaded libraries. Created
906 version.h, now support a version extension string for distributors
907 who patch heavily. Dynamic module ABI change.
909 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
910 privilege escalation vulnerability whereby the Exim run-time user
911 can cause root to append content of the attacker's choosing to
914 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
917 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
918 If dropping privileges for untrusted macros, we disabled normal logging
919 on the basis that it would fail; for the Exim run-time user, this is not
920 the case, and it resulted in successful deliveries going unlogged.
921 Fixed. Reported by Andreas Metzler.
927 PP/01 Date: & Message-Id: revert to normally being appended to a message,
928 only prepend for the Resent-* case. Fixes regression introduced in
929 Exim 4.70 by NM/22 for Bugzilla 607.
931 PP/02 Include check_rfc2047_length in configure.default because we're seeing
932 increasing numbers of administrators be bitten by this.
934 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
936 PP/03 Bugzilla 994: added openssl_options main configuration option.
938 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
940 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
942 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
944 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
945 without a peer certificate, leading to a segfault because of an
946 assumption that peers always have certificates. Be a little more
947 paranoid. Problem reported by Martin Tscholak.
949 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
950 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
951 NB: ClamAV planning to remove STREAM in "middle of 2010".
952 CL also introduces -bmalware, various -d+acl logging additions and
953 more caution in buffer sizes.
955 PP/09 Implemented reverse_ip expansion operator.
957 PP/10 Bugzilla 937: provide a "debug" ACL control.
959 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
961 PP/12 Bugzilla 973: Implement --version.
963 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
965 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
967 PP/15 Bugzilla 816: support multiple condition rules on Routers.
969 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
970 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
971 ignore trailing whitespace.
973 JJ/02 prevent non-panic DKIM error from being sent to paniclog
975 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
978 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
979 Notification from Dr Andrew Aitchison.
981 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
982 ExtendedDetectionInfo response format.
983 Notification from John Horne.
985 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
988 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
989 XSL and documented dependency on system catalogs, with examples of how
992 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
995 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
996 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
997 configuration file which is writeable by the Exim user or group.
999 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
1000 of configuration files to cover files specified with the -C option if
1001 they are going to be used with root privileges, not just the default
1004 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
1005 option (effectively making it always true).
1007 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
1008 files to be used while preserving root privileges.
1010 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
1011 that rogue child processes cannot use them.
1013 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
1014 run-time user, instead of root.
1016 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
1017 Exim run-time user without dropping privileges.
1019 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
1020 result string, instead of calling string_vformat() twice with the same
1023 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
1024 for other users. Others should always drop root privileges if they use
1025 -C on the command line, even for a whitelisted configure file.
1027 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
1029 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
1035 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
1036 $data_path, and $header_path variables; fixed documentation bugs and
1039 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1040 exipick to access non-standard spools, including the "frozen" queue
1043 NM/01 Bugzilla 965: Support mysql stored procedures.
1044 Patch from Alain Williams
1046 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1048 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1049 Patch from Andreas Metzler
1051 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1052 Patch from Kirill Miazine
1054 NM/05 Bugzilla 671: Added umask to procmail example.
1056 JJ/03 installed exipick 20100323.0, fixing doc bug
1058 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1059 directory. Notification and patch from Dan Rosenberg.
1061 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1063 TK/02 Improve log output when DKIM signing operation fails.
1065 MH/01 Treat the transport option dkim_domain as a colon separated
1066 list, not as a single string, and sign the message with each element,
1067 omitting multiple occurences of the same signer.
1069 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1070 Bugzilla 985, 986. Patch by Simon Arlott
1072 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1073 Patch by Simon Arlott
1075 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1076 MBX locking. Notification from Dan Rosenberg.
1082 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1084 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1086 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1088 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1090 NM/04 Bugzilla 915: Flag broken perl installation during build.
1096 TK/01 Added patch by Johannes Berg that expands the main option
1097 "spamd_address" if it starts with a dollar sign.
1099 TK/02 Write list of recipients to X-Envelope-Sender header when building
1100 the mbox-format spool file for content scanning (suggested by Jakob
1103 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1104 (http://www.dcc-servers.net/) support via dccifd. Activated by
1105 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1107 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1108 by Mark Daniel Reidel <mr@df.eu>.
1110 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1111 When building exim an external PCRE library is now needed -
1112 PCRE is a system library on the majority of modern systems.
1113 See entry on PCRE_LIBS in EDITME file.
1115 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1116 conversation. Added nologin parameter to request.
1117 Patch contributed by Kirill Miazine.
1119 TF/01 Do not log submission mode rewrites if they do not change the address.
1121 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1123 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1124 log files in place. Contributed by Roberto Lima.
1126 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1128 TF/03 Bugzilla 615: When checking the local_parts router precondition
1129 after a local_part_suffix or local_part_prefix option, Exim now
1130 does not use the address's named list lookup cache, since this
1131 contains cached lookups for the whole local part.
1133 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1134 Robert Millan. Documentation is in experimental-spec.txt.
1136 TF/04 Bugzilla 668: Fix parallel build (make -j).
1138 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1140 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1141 Patch provided by Jan Srzednicki.
1143 TF/05 Leading white space used to be stripped from $spam_report which
1144 wrecked the formatting. Now it is preserved.
1146 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1147 that they are available at delivery time.
1149 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1151 TF/08 TLS error reporting now respects the incoming_interface and
1152 incoming_port log selectors.
1154 TF/09 Produce a more useful error message if an SMTP transport's hosts
1155 setting expands to an empty string.
1157 NM/06 Bugzilla 744: EXPN did not work under TLS.
1158 Patch provided by Phil Pennock.
1160 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1161 Patch provided by Richard Godbee.
1163 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1164 acl_smtp_notquit, added index entry.
1166 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1167 Patch provided by Eugene Bujak.
1169 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1170 accept(). Patch provided by Maxim Dounin.
1172 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1173 Patch provided by Phil Pennock.
1175 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1177 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1178 Patch provided by Brad "anomie" Jorsch.
1180 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1181 Patch provided by Dean Brooks.
1183 TK/05 Add native DKIM support (does not depend on external libraries).
1185 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1186 Patch provided by Graeme Fowler.
1188 NM/16 Bugzilla 851: Documentation example syntax fix.
1190 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1192 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1195 NM/19 Bugzilla 745: TLS version reporting.
1196 Patch provided by Phil Pennock.
1198 NM/20 Bugzilla 167: bool: condition support.
1199 Patch provided by Phil Pennock.
1201 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1202 clients. Patch provided by Phil Pennock.
1204 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1205 Patch provided by Brad "anomie" Jorsch.
1207 NM/23 Bugzilla 687: Fix misparses in eximstats.
1208 Patch provided by Heiko Schlittermann.
1210 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1211 Patch provided by Heiko Schlittermann.
1213 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1214 plus update to original patch.
1216 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1218 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1219 Patch provided by David Brownlee.
1221 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1223 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1225 NM/30 Bugzilla 888: TLS documentation bugfixes.
1227 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1229 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1230 Unlike the original bugzilla I have changed all shell scripts in src tree.
1232 NM/33 Bugzilla 898: Transport filter timeout fix.
1233 Patch by Todd Rinaldo.
1235 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1236 Patch by Serge Demonchaux.
1238 NM/35 Bugzilla 39: Base64 decode bug fixes.
1239 Patch by Jakob Hirsch.
1241 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1243 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1245 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1247 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1253 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1254 ALT-N's libdkim that I have put here:
1255 http://duncanthrax.net/exim-experimental/
1257 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1258 were called 'true' and it seems that C99 defines that as a reserved
1259 keyword to be used with 'bool' variable types. That means you could
1260 not include C99-style headers which use bools without triggering
1261 build errors in sieve.c.
1263 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1264 as mailq or other aliases. Changed the --help handling significantly
1265 to do whats expected. exim_usage() emits usage/help information.
1267 SC/01 Added the -bylocaldomain option to eximstats.
1269 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1271 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1273 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1279 PH/01 Another patch from the Sieve maintainer.
1281 PH/02 When an IPv6 address is converted to a string for single-key lookup
1282 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1283 dots are used instead of colons so that keys in lsearch files need not
1284 contain colons. This was done some time before quoting was made available
1285 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1286 (notated using the quote facility) so as to distinguish them from IPv4
1287 keys. This meant that lookups for IP addresses in host lists did not work
1288 for iplsearch lookups.
1290 This has been fixed by arranging for IPv6 addresses to be expressed with
1291 colons if the lookup type is iplsearch. This is not incompatible, because
1292 previously such lookups could never work.
1294 The situation is now rather anomolous, since one *can* have colons in
1295 ordinary lsearch keys. However, making the change in all cases is
1296 incompatible and would probably break a number of configurations.
1298 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1301 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1302 conversion specification without a maximum field width, thereby enabling
1303 a rogue spamd server to cause a buffer overflow. While nobody in their
1304 right mind would setup Exim to query an untrusted spamd server, an
1305 attacker that gains access to a server running spamd could potentially
1306 exploit this vulnerability to run arbitrary code as the Exim user.
1308 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1309 $primary_hostname instead of what libspf2 thinks the hosts name is.
1311 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1312 a directory entry by the name of the lookup key. Previously, if a
1313 symlink pointed to a non-existing file or a file in a directory that
1314 Exim lacked permissions to read, a lookup for a key matching that
1315 symlink would fail. Now it is enough that a matching directory entry
1316 exists, symlink or not. (Bugzilla 503.)
1318 PH/03 The body_linecount and body_zerocount variables are now exported in the
1321 PH/04 Added the $dnslist_matched variable.
1323 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1324 This means they are set thereafter only if the connection becomes
1327 PH/06 Added the client_condition to authenticators so that some can be skipped
1328 by clients under certain conditions.
1330 PH/07 The error message for a badly-placed control=no_multiline_responses left
1331 "_responses" off the end of the name.
1333 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1335 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1336 (without spaces) instead of just copying the configuration text.
1338 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1340 PH/11 Added $max_received_linelength.
1342 PH/12 Added +ignore_defer and +include_defer to host lists.
1344 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1345 way in which PCRE > 7.0 is built.
1347 PH/14 Implemented queue_only_load_latch.
1349 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1350 MAIL command. The effect was to mangle the value on 64-bit systems.
1352 PH/16 Another patch from the Sieve maintainer.
1354 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1356 PH/18 If a system quota error occurred while trying to create the file for
1357 a maildir delivery, the message "Mailbox is full" was not appended to the
1358 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1359 only to a quota excession during the actual writing of the file.
1361 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1362 characters?) which causes problems in log lines. The DN values are now
1363 passed through string_printing() before being added to log lines.
1365 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1366 and InterBase are left for another time.)
1368 PH/21 Added message_body_newlines option.
1370 PH/22 Guard against possible overflow in moan_check_errorcopy().
1372 PH/23 POSIX allows open() to be a macro; guard against that.
1374 PH/24 If the recipient of an error message contained an @ in the local part
1375 (suitably quoted, of course), incorrect values were put in $domain and
1376 $local_part during the evaluation of errors_copy.
1382 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1383 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1386 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1387 issue a MAIL command.
1389 PH/02 In an ACL statement such as
1391 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1393 if a client was not listed at all, or was listed with a value other than
1394 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1395 the condition was not true (as it should be), so access was not denied.
1396 The bug was that the ! inversion was incorrectly passed on to the second
1397 item. This has been fixed.
1399 PH/03 Added additional dnslists conditions == and =& which are different from
1400 = and & when the dns lookup returns more than one IP address.
1402 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1403 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1405 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1406 FSYNC, which compiles an option called disable_fsync that allows for
1407 bypassing fsync(). The documentation is heavily laced with warnings.
1409 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1411 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1412 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1413 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1414 including adding "make clean"; (3) Added -fPIC when compiling the test
1415 dynamically loaded module, to get rid of a warning.
1417 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1418 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1419 = 0s. The bug is otherwise harmless.
1421 PH/07 There was a bug in the dovecot authenticator such that the value of
1422 $auth1 could be overwritten, and so not correctly preserved, after a
1423 successful authentication. This usually meant that the value preserved by
1424 the server_setid option was incorrect.
1426 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1428 PH/09 Installed PCRE release 7.0.
1430 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1431 run for batched SMTP input. It is now run at the start of every message
1432 in the batch. While fixing this I discovered that the process information
1433 (output by running exiwhat) was not always getting set for -bs and -bS
1434 input. This is fixed, and it now also says "batched" for BSMTP.
1436 PH/11 Added control=no_pipelining.
1438 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1439 patch, slightly modified), and move the expansion of helo_data till after
1440 the connection is made in the smtp transport (so it can use these
1443 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1445 PH/14 Added log_selector = +pid.
1447 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1449 PH/16 Add ${if forany and ${if forall.
1451 PH/17 Added dsn_from option to vary the From: line in DSNs.
1453 PH/18 Flush SMTP output before performing a callout, unless control =
1454 no_callout_flush is set.
1456 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1457 was true (the default) a successful delivery failed to delete the retry
1458 item, thus causing premature timeout of the address. The bug is now
1461 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1463 PH/21 Long custom messages for fakedefer and fakereject are now split up
1464 into multiline reponses in the same way that messages for "deny" and
1465 other ACL rejections are.
1467 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1468 with slight modification.
1470 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1471 draft, changing the syntax and factoring some duplicate code".
1473 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1474 for deliveries of the second and subsequent messages over the same SMTP
1477 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1478 ${reduce, with only minor "tidies".
1480 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1482 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1483 expansion side effects.
1485 PH/27 When a message times out after an over-quota error from an Exim-imposed
1486 quota, the bounce message says "mailbox is full". This message was not
1487 being given when it was a system quota that was exceeded. It now should
1490 MH/03 Made $recipients available in local_scan(). local_scan() already has
1491 better access to the recipient list through recipients_list[], but
1492 $recipients can be useful in postmaster-provided expansion strings.
1494 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1495 in the case of a MAIL command with additional options following the
1496 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1497 were accidentally chopped off.
1499 PH/29 SMTP synchronization checks are implemented when a command is read -
1500 there is a check that no more input is waiting when there shouldn't be
1501 any. However, for some commands, a delay in an ACL can mean that it is
1502 some time before the response is written. In this time, more input might
1503 arrive, invalidly. So now there are extra checks after an ACL has run for
1504 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1505 pipelining has not been advertised.
1507 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1509 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1510 returned in the SMTP response when smtp_return_error_details was false.
1511 This has been fixed.
1513 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1514 instead of the C I/O that was originally supplied, because problems were
1515 reported on Solaris.
1517 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1518 Exim which did not show up earlier: it was assuming that a call to
1519 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1520 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1521 was a macro that became an assignment, so it seemed to work. This has
1522 changed to a proper function call with a void return, hence the compile
1523 error. Exim's code has been fixed.
1525 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1528 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1530 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1532 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1533 because it turns out to be too big or there is a timeout) there is no
1534 "Completed" line in the log. When some messages of this type were
1535 selected by exigrep, they were listed as "not completed". Others were
1536 picked up by some special patterns. I have improved the selection
1537 criteria to be more general.
1539 PH/38 The host_find_failed option in the manualroute router can now be set
1540 to "ignore", to completely ignore a host whose IP address cannot be
1541 found. If all hosts are ignored, the behaviour is controlled by the new
1542 host_all_ignored option.
1544 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1545 homing or because of multiple MX records with /mx) generated more than
1546 one IP address, and the following item turned out to be the local host,
1547 all the secondary addresses of the first item were incorrectly removed
1548 from the list, along with the local host and any following hosts (which
1549 is what is supposed to happen).
1551 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1552 whoever called Exim into the -H file. In the case of the daemon it was
1553 behaving confusingly. When first started, it used values for whoever
1554 started the daemon, but after a SIGHUP it used the Exim user (because it
1555 calls itself on a restart). I have changed the code so that it now always
1558 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1559 message are rejected with the same error (e.g. no authentication or bad
1560 sender address), and a DATA command is nevertheless sent (as can happen
1561 with PIPELINING or a stupid MUA), the error message that was given to the
1562 RCPT commands is included in the rejection of the DATA command. This is
1563 intended to be helpful for MUAs that show only the final error to their
1566 PH/42 Another patch from the Sieve maintainer.
1568 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1569 Eximstats - Fixed some broken HTML links and added missing column headers
1571 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1572 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1574 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1576 PH/43 Yet another patch from the Sieve maintainer.
1578 PH/44 I found a way to check for a TCP/IP connection going away before sending
1579 the response to the final '.' that terminates a message, but only in the
1580 case where the client has not sent further data following the '.'
1581 (unfortunately, this is allowed). However, in many cases there won't be
1582 any further data because there won't be any more messages to send. A call
1583 to select() can be used: if it shows that the input is "ready", there is
1584 either input waiting, or the socket has been closed. An attempt to read
1585 the next input character can distinguish the two cases. Previously, Exim
1586 would have sent an OK response which the client would never have see.
1587 This could lead to message repetition. This fix should cure that, at
1588 least in a lot of common cases.
1590 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1591 advertised in response to EHLO.
1597 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1598 fixed by 4.65/MH/01 (is this a record?) are fixed:
1600 (i) An empty string was always treated as zero by the numeric comparison
1601 operators. This behaviour has been restored.
1603 (ii) It is documented that the numeric comparison operators always treat
1604 their arguments as decimal numbers. This was broken in that numbers
1605 starting with 0 were being interpreted as octal.
1607 While fixing these problems I realized that there was another issue that
1608 hadn't been noticed. Values of message_size_limit (both the global option
1609 and the transport option) were treated as octal if they started with 0.
1610 The documentation was vague. These values are now always treated as
1611 decimal, and I will make that clear in the documentation.
1617 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1618 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1621 MH/01 Don't check that the operands of numeric comparison operators are
1622 integers when their expansion is in "skipping" mode (fixes bug
1623 introduced by 4.64-PH/07).
1625 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1626 child addresses, Exim now panics and dies. Previously, because the count
1627 is held in a short int, deliveries were likely to be lost. As such a
1628 large number of recipients for a single message is ridiculous
1629 (performance will be very, very poor), I have chosen to impose a limit
1630 rather than extend the field.
1636 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1637 leftover -K file (the existence of which was triggered by #402).
1638 While we were at it, introduced process PID as part of the -K
1639 filename. This should rule out race conditions when creating
1642 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1643 processing considerably. Previous code took too long for large mails,
1644 triggering a timeout which in turn triggers #401.
1646 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1647 in the DK code in transports.c. sendfile() is not really portable,
1648 hence the _LINUX specificness.
1650 TF/01 In the add_headers option to the mail command in an Exim filter,
1651 there was a bug that Exim would claim a syntax error in any
1652 header after the first one which had an odd number of characters
1655 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1656 callout verification, Exim cached a "reject" for the entire domain. This
1657 is correct for most verifications, but it is not correct for a recipient
1658 verification with use_sender or use_postmaster set, because in that case
1659 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1660 case of MAIL FROM:<> rejection from other early rejections (e.g.
1661 rejection of HELO). When verifying a recipient using a non-null MAIL
1662 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1663 Whatever the result of the callout, the value of the domain cache is
1664 left unchanged (for any other kind of callout, getting as far as trying
1665 RCPT means that the domain itself is ok).
1667 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1670 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1671 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1672 session, the daemon ignores ECONNECT errors and logs others; it now
1673 ignores EPIPE as well.
1675 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1676 (quoted-printable decoding).
1678 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1679 later the small subsequent patch to fix an introduced bug.
1681 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1683 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1685 PH/08 An error is now given if message_size_limit is specified negative.
1687 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1688 to be given (somewhat) arbitrary names.
1690 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1693 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1694 miscellaneous code fixes
1696 PH/10 Added the log_reject_target ACL modifier to specify where to log
1699 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1700 hostname. This is wrong, because it relates to the incoming message (and
1701 probably the interface on which it is arriving) and not to the outgoing
1702 callout (which could be using a different interface). This has been
1703 changed to use the value of the helo_data option from the smtp transport
1704 instead - this is what is used when a message is actually being sent. If
1705 there is no remote transport (possible with a router that sets up host
1706 addresses), $smtp_active_hostname is used.
1708 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1709 tweaks were necessary in order to get it to work (see also 21 below):
1710 (a) The code assumed that strncpy() returns a negative number on buffer
1711 overflow, which isn't the case. Replaced with Exim's string_format()
1713 (b) There were several signed/unsigned issues. I just did the minimum
1714 hacking in of casts. There is scope for a larger refactoring.
1715 (c) The code used strcasecmp() which is not a standard C function.
1716 Replaced with Exim's strcmpic() function.
1717 (d) The code set only $1; it now sets $auth1 as well.
1718 (e) A simple test gave the error "authentication client didn't specify
1719 service in request". It would seem that Dovecot has changed its
1720 interface. Fortunately there's a specification; I followed it and
1721 changed what the client sends and it appears to be working now.
1723 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1726 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1727 address A is aliased to B and C, where B exists and C does not. Without
1728 -v the output is "A verified" because verification stops after a
1729 successful redirection if more than one address is generated. However,
1730 with -v the child addresses are also verified. Exim was outputting "A
1731 failed to verify" and then showing the successful verification for C,
1732 with its parentage. It now outputs "B failed to verify", showing B's
1733 parentage before showing the successful verification of C.
1735 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1736 look up a TXT record in a specific list after matching in a combined
1739 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1740 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1741 they consult the DNS. I had assumed they would set it the way they
1742 wanted; and indeed my experiments on Linux seem to show that in some
1743 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1744 To be on the safe side, however, I have now made the interface to
1745 host_find_byname() similar to host_find_bydns(), with an argument
1746 containing the DNS resolver options. The host_find_byname() function now
1747 sets these options at its start, just as host_find_bydns() does. The smtp
1748 transport options dns_qualify_single and dns_search_parents are passed to
1749 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1750 of host_find_byname() use the default settings of RES_DEFNAMES
1751 (qualify_single) but not RES_DNSRCH (search_parents).
1753 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1754 spool_read_header() do less string testing, by means of a preliminary
1755 switch on the second character of optional "-foo" lines. (This is
1756 overdue, caused by the large number of possibilities that now exist.
1757 Originally there were few.) While I was there, I also converted the
1758 str(n)cmp tests so they don't re-test the leading "-" and the first
1759 character, in the hope this might squeeze out yet more improvement.
1761 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1762 flag allowing group syntax was set by the header_syntax check but not
1763 turned off, possible causing trouble later; (2) The flag was not being
1764 set at all for the header_verify test, causing "group"-style headers to
1765 be rejected. I have now set it in this case, and also caused header_
1766 verify to ignore an empty address taken from a group. While doing this, I
1767 came across some other cases where the code for allowing group syntax
1768 while scanning a header line wasn't quite right (mostly, not resetting
1769 the flag correctly in the right place). These bugs could have caused
1770 trouble for malformed header lines. I hope it is now all correct.
1772 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1773 with the "reply" argument non-NULL. The code, however (which originally
1774 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1775 but it didn't always do it. This confused somebody who was copying the
1776 code for some other use. I have removed all the tests.
1778 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1779 feature that was used to support insecure browsers during the U.S. crypto
1780 embargo. It requires special client support, and Exim is probably the
1781 only MTA that supported it -- and would never use it because real RSA is
1782 always available. This code has been removed, because it had the bad
1783 effect of slowing Exim down by computing (never used) parameters for the
1784 RSA_EXPORT functionality.
1786 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1787 authenticator to fail if there's a tab character in the incoming data
1788 (there should never be unless someone is messing about, as it's supposed
1789 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1790 if the connection is using TLS or if the remote IP is the same as the
1791 local IP, and the "valid-client-cert option" if a client certificate has
1794 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1795 authenticators. This can be used for authorization after authentication
1796 succeeds. (In the case of plaintext, it servers for both authentication
1799 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1800 if any retry times were supplied.
1802 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1803 connection, where there is no client IP address to check. In this
1804 situation, the verify now always succeeds.
1806 PH/25 Applied John Jetmore's -Mset patch.
1808 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1810 PH/27 In a string expansion for a processed (not raw) header when multiple
1811 headers of the same name were present, leading whitespace was being
1812 removed from all of them, but trailing whitespace was being removed only
1813 from the last one. Now trailing whitespace is removed from each header
1814 before concatenation. Completely empty headers in a concatenation (as
1815 before) are ignored.
1817 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1818 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1820 PH/29 [Removed. This was a change that I later backed out, and forgot to
1821 correct the ChangeLog entry (that I had efficiently created) before
1822 committing the later change.]
1824 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1825 address errors (4xx response to RCPT) over the same connection as other
1826 messages routed to the same hosts. Such deliveries are always "forced",
1827 so retry times are not inspected. This resulted in far too many retries
1828 for the affected addresses. The effect occurred only when there were more
1829 hosts than the hosts_max_try setting in the smtp transport when it had
1830 the 4xx errors. Those hosts that it had tried were not added to the list
1831 of hosts for which the message was waiting, so if all were tried, there
1832 was no problem. Two fixes have been applied:
1834 (i) If there are any address or message errors in an SMTP delivery, none
1835 of the hosts (tried or untried) are now added to the list of hosts
1836 for which the message is waiting, so the message should not be a
1837 candidate for sending over the same connection that was used for a
1838 successful delivery of some other message. This seems entirely
1839 reasonable: after all the message is NOT "waiting for some host".
1840 This is so "obvious" that I'm not sure why it wasn't done
1841 previously. Hope I haven't missed anything, but it can't do any
1842 harm, as the worst effect is to miss an optimization.
1844 (ii) If, despite (i), such a delivery is accidentally attempted, the
1845 routing retry time is respected, so at least it doesn't keep
1846 hammering the server.
1848 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1849 in ${readsocket because some servers need this prod.
1851 PH/32 Added some extra debug output when updating a wait-xxx database.
1853 PH/33 The hint "could be header name not terminated by colon", which has been
1854 given for certain expansion errors for a long time, was not being given
1855 for the ${if def:h_colon_omitted{... case.
1857 PH/34 The spec says: "With one important exception, whenever a domain list is
1858 being scanned, $domain contains the subject domain." There was at least
1859 one case where this was not true.
1861 PH/35 The error "getsockname() failed: connection reset by peer" was being
1862 written to the panic log as well as the main log, but it isn't really
1863 panic-worthy as it just means the connection died rather early on. I have
1864 removed the panic log writing for the ECONNRESET error when getsockname()
1867 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1868 runs only) independently of the message's sender address. This meant
1869 that, if the 4xx error was in fact related to the sender, a different
1870 message to the same recipient with a different sender could confuse
1871 things. In particualar, this can happen when sending to a greylisting
1872 server, but other circumstances could also provoke similar problems.
1873 I have changed the default so that the retry time for these errors is now
1874 based a combination of the sender and recipient addresses. This change
1875 can be overridden by setting address_retry_include_sender=false in the
1878 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1879 remote server are returned as part of bounce messages. This was not
1880 happening for LMTP over a pipe (the lmtp transport), but now it is the
1881 same for both kinds of LMTP.
1883 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1884 in header lines that were in fact CNAMEs. This is no longer the case.
1886 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1887 and queue runs started by the daemon processed all messages. This has
1888 been fixed so that -R and -S can now usefully be given with -q<time>.
1890 PH/40 Import PCRE release 6.7 (fixes some bugs).
1892 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1894 PH/42 Give an error if -q is specified more than once.
1896 PH/43 Renamed the variables $interface_address and $interface_port as
1897 $received_ip_address and $received_port, to make it clear that these
1898 values apply to message reception, and not to the outgoing interface when
1899 a message is delivered. (The old names remain recognized, of course.)
1901 PH/44 There was no timeout on the connect() call when using a Unix domain
1902 socket in the ${readsocket expansion. There now is.
1904 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1905 be meaningful with "accept".
1907 SC/01 Eximstats V1.43
1908 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1910 SC/02 Eximstats V1.44
1911 Use a glob alias rather than an array ref in the generated
1912 parser. This improves both readability and performance.
1914 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1915 Collect SpamAssassin and rejection statistics.
1916 Don't display local sender or destination tables unless
1917 there is data to show.
1918 Added average volumes into the top table text output.
1920 SC/04 Eximstats V1.46
1921 Collect data on the number of addresses (recipients)
1922 as well as the number of messages.
1924 SC/05 Eximstats V1.47
1925 Added 'Message too big' to the list of mail rejection
1926 reasons (thanks to Marco Gaiarin).
1928 SC/06 Eximstats V1.48
1929 Mainlog lines which have GMT offsets and are too short to
1930 have a flag are now skipped.
1932 SC/07 Eximstats V1.49 (Alain Williams)
1933 Added the -emptyok flag.
1935 SC/08 Eximstats V1.50
1936 Fixes for obtaining the IP address from reject messages.
1938 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1939 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1940 whitesspace changes from 4.64-PH/27
1942 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1945 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1946 are found, allow negative numbers in numeric criteria)
1948 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1950 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1951 to match changes made in 4.64-PH/43
1953 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1955 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1956 "domain" is actually an IP address, and force a failure. This locks out
1957 those revolvers/nameservers that support "A-for-A" lookups, in
1958 contravention of the specifications.
1960 PH/48 When a host name was looked up from an IP address, and the subsequent
1961 forward lookup of the name timed out, the host name was left in
1962 $sender_host_name, contrary to the specification.
1964 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1965 restricted to single-key lookups, Exim was not diagnosing an error if
1966 * or *@ was used with a query-style lookup.
1968 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1970 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1971 long ago, but noone interested enough thought of it. Let's just say that
1972 the "1.1" means that there are some new functions that weren't there at
1973 some point in the past.
1975 PH/51 Error processing for expansion failure of helo_data from an smtp
1976 transport during callout processing was broken.
1978 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1979 tested/used via the -bh/-bhc/-bs options.
1981 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1982 bug, fixed in subsequent PCRE releases).
1984 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1985 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1987 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1993 SC/01 Use a glob alias rather than an array ref in eximstats generated
1994 parser. This improves both readability and performance.
1996 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1997 Don't display local sender or destination tables in eximstats unless
1998 there is data to show.
1999 Added average volumes into the eximstats top table text output.
2001 SC/03 Collect data on the number of addresses (recipients) as well
2002 as the number of messages in eximstats.
2004 TF/01 Correct an error in the documentation for the redirect router. Exim
2005 does (usually) call initgroups() when daemonizing.
2007 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
2008 with consistent privilege compared to when running as a daemon.
2010 TF/03 Note in the spec that $authenticated_id is not set for local
2011 submissions from trusted users.
2013 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
2014 Thanks to Dean Brooks <dean@iglou.com> for the patch.
2016 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
2017 by adding some example configuration directives to the default
2018 configuration file. A little bit of work is required to uncomment the
2019 directives and define how usernames and passwords are checked, but
2020 there is now a framework to start from.
2022 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
2023 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
2024 without this. I don't know how relevant this is to other LDAP libraries.
2026 PH/02 Add the verb name to the "unknown ACL verb" error.
2028 PH/03 Magnus Holmgren's patch for filter_prepend_home.
2030 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
2032 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
2033 directory not expanded when it should be if an expanded home directory
2034 was set for the address (which is overridden by the transport).
2036 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
2039 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2040 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2041 because it is too late at that time, and has no effect.
2043 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2044 security issue with \' (bugzilla #107). I could not use the
2045 PQescapeStringConn() function, because it needs a PGconn value as one of
2048 PH/08 When testing addresses using -bt, indicate those final addresses that
2049 are duplicates that would not cause an additional delivery. At least one
2050 person was confused, thinking that -bt output corresponded to deliveries.
2051 (Suppressing duplicates isn't a good idea as you lose the information
2052 about possibly different redirections that led to the duplicates.)
2054 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2055 systems where poll() doesn't work, in particular OS X.
2057 PH/10 Added more information to debugging output for retry time not reached.
2059 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2060 operations in malware.c.
2062 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2065 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2066 syslog_duplication set false, log lines that would normally be written
2067 both the the main log and to the reject log were not written to syslog at
2070 PH/14 In the default configuration, change the use of "message" in ACL warn
2071 statements to "add_header".
2073 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2074 not followed by a command (e.g. "seen endif").
2076 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2077 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2080 PH/17 Added extra conditions to the default value of delay_warning_condition
2084 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2085 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2086 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2089 The Auto-Submitted: and various List- headers are standardised, whereas I
2090 don't think Precedence: ever was.
2092 PH/18 Refactored debugging code in route_finduser() to show more information,
2093 in particular, the error code if getpwnam() issues one.
2095 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2096 This is apparently needed in addition to the PH/07 change above to avoid
2097 any possible encoding problems.
2099 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2100 but not after initializing Perl.
2102 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2103 output them only if debugging. By default they are written stderr,
2104 apparently, which is not desirable.
2106 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2109 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2112 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2114 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2115 authenticated or an ident call has been made. Suppress the default
2116 values for $authenticated_id and $authenticated_sender (but permit -oMai
2117 and -oMas) when testing with -bh.
2119 PH/24 Re-jigged the order of the tests in the default configuration so that the
2120 tests for valid domains and recipients precede the DNS black list and CSA
2121 tests, on the grounds that those ones are more expensive.
2123 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2124 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2125 This bug exists in every version of Exim that I still have, right back to
2128 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2129 However, an attempt to turn on case-sensitivity in a regex key by
2130 including (?-i) didn't work because the subject string was already
2131 lowercased, and the effects were non-intuitive. It turns out that a
2132 one-line patch can be used to allow (?-i) to work as expected.
2138 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2139 other effects) broke the use of negated acl sub-conditions.
2141 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2144 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2145 "Deny" causes Exim to reject the incoming connection with a 554 error.
2146 Unfortunately, if there is a major crisis, such as a disk failure,
2147 tcp-wrappers gives "deny", whereas what one would like would be some
2148 kind of temporary error. A kludge has been added to help with this.
2149 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2150 554 error is used if errno is still zero or contains ENOENT (which occurs
2151 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2154 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2156 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2157 errors. Otherwise a message that provokes a temporary error (when other
2158 messages do not) can cause a whole host to time out.
2160 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2161 addresses were routed directly to files or pipes from a redirect router.
2162 File deliveries just didn't batch; pipe deliveries might have suffered
2165 PH/06 A failure to get a lock for a hints database would erroneously always say
2166 "Failed to get write lock", even when it was really a read lock.
2168 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2169 of 0600. This has been changed to use the value of the lockfile_mode
2170 option (which defaults to 0600).
2172 PH/08 Applied small patch from the Sieve maintainer.
2174 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2175 folder from quota calculations, a direct delivery into this folder messed
2176 up the contents of the maildirsize file. This was because the regex was
2177 used only to exclude .Trash (or whatever) when the size of the mailbox
2178 was calculated. There was no check that a delivery was happening into an
2179 excluded directory. This bug has been fixed by ignoring all quota
2180 processing for deliveries into excluded directories.
2182 PH/10 Added the maildirfolder_create_regex option to appendfile.
2188 PH/01 The code for finding all the local interface addresses on a FreeBSD
2189 system running IPv6 was broken. This may well have applied to all BSD
2190 systems, as well as to others that have similar system calls. The broken
2191 code found IPv4 interfaces correctly, but gave incorrect values for the
2192 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2193 that it would not match correctly against @[] and not recognize the IPv6
2196 PH/02 The ipliteral router was not recognizing addresses of the form user@
2197 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2199 PH/03 Added disable_ipv6.
2201 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2202 decoded form, because it is most often used to construct To: headers
2203 lines in autoreplies, and the decoded form may well be syntactically
2204 invalid. However, $reply_address has leading white space removed, and all
2205 newlines turned into spaces so that the autoreply transport does not
2208 PH/05 If group was specified without a user on a router, and no group or user
2209 was specified on a transport, the group from the router was ignored.
2211 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2212 for visible compile-time settings that can be used to change these
2213 numbers, for those that want even more. Backwards compatibility with old
2214 spool files has been maintained. However, going back to a previous Exim
2215 release will lost any variables that are in spool files.
2217 PH/07 Two small changes when running in the test harness: increase delay when
2218 passing a TCP/IP connection to a new process, in case the original
2219 process has to generate a bounce, and remove special handling of
2220 127.0.0.2 (sic), which is no longer necessary.
2222 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2223 be the same on different OS.
2225 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2228 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2229 whether --show-vars was specified or not
2231 JJ/02 exipick: Added support for new ACL variable spool format introduced
2234 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2235 syntactically invalid From: or Reply-to: line, and a filter used this to
2236 generate an autoreply, and therefore failed to obtain an address for the
2237 autoreply, Exim could try to deliver to a non-existent relative file
2238 name, causing unrelated and misleading errors. What now happens is that
2239 it logs this as a hard delivery error, but does not attempt to create a
2242 PH/11 The exinext utility has a -C option for testing purposes, but although
2243 the given file was scanned by exinext itself; it wasn't being passed on
2244 when Exim was called.
2246 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2247 an end-of-file indication when reading a command response.
2249 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2250 compiled. In many other places in Exim, IPv6 addresses are always
2251 recognized, so I have changed this. It also means that IPv4 domain
2252 literals of the form [IPV4:n.n.n.n] are now always recognized.
2254 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2255 used if the router is not running as root, for example, when verifying at
2256 ACL time, or when using -bh. The debugging output from this situation was
2257 non-existent - all you got was a failure to exec. I have made two
2260 (a) Failures to set uid/gid, the current directory, or a process leader
2261 in a subprocess such as that created by queryprogram now generate
2262 suitable debugging ouput when -d is set.
2264 (b) The queryprogram router detects when it is not running as root,
2265 outputs suitable debugging information if -d is set, and then runs
2266 the subprocess without attempting to change uid/gid.
2268 PH/15 Minor change to Makefile for building test_host (undocumented testing
2271 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2272 additional section of a DNS packet that returns MX or SRV records.
2273 Instead, it always explicitly searches for A/AAAA records. This avoids
2274 major problems that occur when a DNS server includes only records of one
2275 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2276 fixed another bug: if SRV records were looked up and the corresponding
2277 address records were *not* found in the additional section, the port
2278 values from the SRV records were lost.
2280 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2281 using the correct key (the original address) when searching the retry
2282 rules in order to find which one to use for generating the retry hint.
2284 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2285 adding the default one. Similarly, if it contains a Reply-To: header, the
2286 errors_reply_to option, if set, is not used.
2288 PH/19 When calculating a retry time, Exim used to measure the "time since
2289 failure" by looking at the "first failed" field in the retry record. Now
2290 it does not use this if it is later than than the arrival time of the
2291 message. Instead it uses the arrival time. This makes for better
2292 behaviour in cases where some deliveries succeed, thus re-setting the
2293 "first failed" field. An example is a quota failure for a huge message
2294 when small messages continue to be delivered. Without this change, the
2295 "time since failure" will always be short, possible causing more frequent
2296 delivery attempts for the huge message than are intended.
2297 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2299 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2300 $1, $2, $3) because the numerical variables can be reset during some
2301 expansion items (e.g. "match"), thereby losing the authentication data.
2303 PH/21 Make -bV show the size of off_t variables so that the test suite can
2304 decide whether to run tests for quotas > 2G.
2306 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2307 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2308 filecount value is greater than 2G or if a quota value is greater than 2G
2309 on a system where the size of off_t is not greater than 4, a panic error
2312 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2313 never match. The debug and -bh output now contains an explicit error
2314 message indicating a malformed IPv4 address or mask.
2316 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2317 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2318 PH/23 above applies.
2320 PH/25 Do not write to syslog when running in the test harness. The only
2321 occasion when this arises is a failure to open the main or panic logs
2322 (for which there is an explicit test).
2324 PH/26 Added the /no_tell option to "control=freeze".
2326 PH/27 If a host name lookup failed very early in a connection, for example, if
2327 the IP address matched host_lookup and the reverse lookup yielded a name
2328 that did not have a forward lookup, an error message of the form "no IP
2329 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2330 could be logged. Now it outputs the IP address instead of "NULL".
2332 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2333 allows the sender and the authenticated sender to be set when
2334 submitting a message from within Exim. Since child_open_exim() is
2335 documented for local_scan(), the new function should be too.
2337 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2338 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2339 results in an empty string is now treated as unset.
2341 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2343 PH/31 Added sender_verify_fail logging option.
2345 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2346 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2347 tidied the source and removed it altogether.
2349 PH/33 When a queue run was abandoned because the load average was too high, a
2350 log line was always written; now it is written only if the queue_run log
2351 selector is set. In addition, the log line for abandonment now contains
2352 information about the queue run such as the pid. This is always present
2353 in "start" and "stop" lines but was omitted from the "abandon" line.
2355 PH/34 Omit spaces between a header name and the colon in the error message that
2356 is given when verify = headers_syntax fails (if there are lots of them,
2357 the message gets confusing).
2359 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2360 names, as there are now some PTR records that contain slashes. This check
2361 is only to protect against broken name servers that fall over on strange
2362 characters, so the fact that it applies to all lookups doesn't matter.
2364 PH/36 Now that the new test suite is complete, we can remove some of the
2365 special code in Exim that was needed for the old test suite. For example,
2366 sorting DNS records because real resolvers return them in an arbitrary
2367 order. The new test suite's fake resolver always returns records in the
2370 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2371 bounces) except when queue_only is set, to avoid logging races between
2372 the different processes.
2374 PH/38 Panic-die if .include specifies a non-absolute path.
2376 PH/39 A tweak to the "H" retry rule from its user.
2378 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2379 a label. They prevented compilation on older perls.
2381 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2382 a warning to be raised on newish perls.
2384 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2385 on queue. Changes to match documented behaviour of showing count of
2386 messages matching specified criteria.
2388 PH/40 Changed the default ident timeout from 30s to 5s.
2390 PH/41 Added support for the use of login_cap features, on those BSD systems
2391 that have them, for controlling the resources used by pipe deliveries.
2393 PH/42 The content-scanning code uses fopen() to create files in which to put
2394 message data. Previously it was not paying any attention to the mode of
2395 the files. Exim runs with umask(0) because the rest of the code creates
2396 files with open(), and sets the required mode explicitly. Thus, these
2397 files were ending up world-writeable. This was not a big issue, because,
2398 being within the spool directory, they were not world-accessible. I have
2399 created a function called modefopen, which takes an additional mode
2400 argument. It sets umask(777), creates the file, chmods it to the required
2401 mode, then resets the umask. All the relevant calls to fopen() in the
2402 content scanning code have been changed to use this function.
2404 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2405 to 24 hours. This avoids potential overflow problems when processing G
2406 and H retry rules. I suspect nobody ever tinkers with this value.
2408 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2410 PH/45 When the plaintext authenticator is running as a client, the server's
2411 challenges are checked to ensure they are valid base64 strings. By
2412 default, the authentication attempt is cancelled if an invalid string is
2413 received. Setting client_ignore_invalid_base64 true ignores these errors.
2414 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2415 they are received. Thus, the responses can be made to depend on the
2416 challenges. If an invalid string is ignored, an empty string is placed in
2419 PH/46 Messages that are created by the autoreply transport now contains a
2420 References: header, in accordance with RFCs 2822 and 3834.
2422 PH/47 Added authenticated_sender_force to the smtp transport.
2424 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2426 PH/49 Installed latest patch from the Sieve maintainer.
2428 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2429 was also set, the appendfile transport was unnecessarily scanning a
2430 directory of message files (e.g. for maildir delivery) to find the count
2431 of files (along with the size), even though it did not need this
2432 information. It now does the scan only if it needs to find either the
2433 size of the count of files.
2435 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2437 PH/52 Two bugs concerned with error handling when the smtp transport is
2440 (i) Exim was not creating retry information for temporary errors given
2441 for individual recipients after the DATA command when the smtp transport
2442 was used in LMTP mode. This meant that they could be retried too
2443 frequently, and not timed out correctly.
2445 (ii) Exim was setting the flag that allows error details to be returned
2446 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2447 recipients that were returned after the DATA command.
2449 PH/53 This is related to PH/52, but is more general: for any failing address,
2450 when detailed error information was permitted to be returned to the
2451 sender, but the error was temporary, then after the final timeout, only
2452 "retry timeout exceeded" was returned. Now it returns the full error as
2453 well as "retry timeout exceeded".
2455 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2456 do this, and (what is worse) MTAs that accept it.
2458 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2459 will now be deprecated.
2461 PH/56 New os.c-cygwin from the Cygwin maintainer.
2463 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2464 formats (previously only available in exim formats via -bpr, -bpru,
2465 and -bpra. Now also available in native and exiqgrep formats)
2467 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2468 with very large, slow to parse queues
2470 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2472 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2474 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2475 responses to authentication challenges, though it was showing the
2476 challenges; (ii) I've removed the CR characters from the debug output for
2479 PH/58 Allow for the insertion of a newline as well as a space when a string
2480 is turned into more than one encoded-word during RFC 2047 encoding. The
2481 Sieve code now uses this.
2483 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2484 data_4xx, lost_connection, tls_required.
2486 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2487 message was being sent as an SMTP response.
2489 PH/61 Add -l and -k options to exicyclog.
2491 PH/62 When verifying, if an address was redirected to one new address, so that
2492 verification continued, and the new address failed or deferred after
2493 having set something in $address_data, the value of $address_data was not
2494 passed back to the ACL. This was different to the case when no
2495 redirection occurred. The value is now passed back in both cases.
2497 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2498 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2499 use login_cap.h, so on its own it isn't the distinguishing feature. The
2500 new name refers directly to the setclassresources() function.
2502 PH/65 Added configuration files for NetBSD3.
2504 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2506 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2507 is preferred over IPv4.
2509 PH/68 The bounce_return_message and bounce_return_body options were not being
2510 honoured for bounces generated during the reception of non-SMTP messages.
2511 In particular, this applied to messages rejected by the ACL. This bug has
2512 been fixed. However, if bounce_return_message is true and bounce_return_
2513 body is false, the headers that are returned for a non-SMTP message
2514 include only those that have been read before the error was detected.
2515 (In the case of an ACL rejection, they have all been read.)
2517 PH/69 The HTML version of the specification is now built in a directory called
2518 spec_html instead of spec.html, because the latter looks like a path with
2519 a MIME-type, and this confuses some software.
2521 PH/70 Catch two compiler warnings in sieve.c.
2523 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2524 function verify_get_ident() calls ip_connect() to connect a socket, but
2525 if the "connect()" function timed out, ip_connect() used to close the
2526 socket. However, verify_get_ident() also closes the socket later, and in
2527 between Exim writes to the log, which may get opened at this point. When
2528 the socket was closed in ip_connect(), the log could get the same file
2529 descriptor number as the socket. This naturally causes chaos. The fix is
2530 not to close the socket in ip_connect(); the socket should be closed by
2531 the function that creates it. There was only one place in the code where
2532 this was missing, in the iplookup router, which I don't think anybody now
2533 uses, but I've fixed it anyway.
2535 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2536 well as to direct DNS lookups. Otherwise the handling of names in host
2537 lists is inconsistent and therefore confusing.
2543 PH/01 Two changes to the default runtime configuration:
2545 (1) Move the checks for relay_from_hosts and authenticated clients from
2546 after to before the (commented out) DNS black list checks.
2548 (2) Add control=submission to the relay_from_hosts and authenticated
2549 clients checks, on the grounds that messages accepted by these
2550 statements are most likely to be submissions.
2552 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2554 (1) Generate an error if the third argument for the ${prvs expansion is
2557 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2560 (3) Reset the variables that are obtained from the first argument of
2561 ${prvscheck and used in the second argument before leaving the code,
2562 because their memory is reclaimed, so using them afterwards may do
2565 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2566 one (it's much easier than Tom thought :-).
2568 (5) Because of (4), we can now allow for the use of $prvscheck_result
2569 inside the third argument.
2571 PH/03 For some reason, the default setting of PATH when running a command from
2572 a pipe transport was just "/usr/bin". I have changed it to
2575 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2576 anything to be listed in the output from -bV.
2578 PH/05 When a filter generated an autoreply, the entire To: header line was
2579 quoted in the delivery log line, like this:
2581 => >A.N.Other <ano@some.domain> <original@ddress> ...
2583 This has been changed so that it extracts the operative address. There
2584 may be more than one such address. If so, they are comma-separated, like
2587 => >ano@some.domain,ona@other.domain <original@ddress> ...
2589 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2590 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2591 not being looked up in the rDNS to get a host name, Exim was showing the
2592 IP address twice in Received: lines, even though the IP addresses were
2593 identical. For example:
2595 Received: from [1.2.3.4] (helo=[1.2.3.4])
2597 However, if the real host name was known, it was omitting the HELO data
2598 if it matched the actual IP address. This has been tidied up so that it
2599 doesn't show the same IP address twice.
2601 PH/07 When both +timestamp and +memory debugging was on, the value given by
2602 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2603 called by the debug printing, thereby overwriting the timestamp buffer.
2604 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2606 PH/08 When the original message was included in an autoreply transport, it
2607 always said "this is a copy of the message, including all the headers",
2608 even if body_only or headers_only was set. It now gives an appropriate
2611 PH/09 Applied a patch from the Sieve maintainer which:
2613 o fixes some comments
2614 o adds the (disabled) notify extension core
2615 o adds some debug output for the result of if/elsif tests
2616 o points to the current vacation draft in the documentation
2617 and documents the missing references header update
2621 o fixes a bug in processing the envelope test (when testing
2622 multiple envelope elements, the last element determinted the
2625 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2626 Electronic Mail") by including:
2628 Auto-submitted: auto-generated
2630 in the messages that it generates (bounce messages and others, such as
2631 warnings). In the case of bounce messages for non-SMTP mesages, there was
2632 also a typo: it was using "Auto_submitted" (underscore instead of
2633 hyphen). Since every message generated by Exim is necessarily in response
2634 to another message, thes have all been changed to:
2636 Auto-Submitted: auto-replied
2638 in accordance with these statements in the RFC:
2640 The auto-replied keyword:
2642 - SHOULD be used on messages sent in direct response to another
2643 message by an automatic process,
2645 - MUST NOT be used on manually-generated messages,
2647 - MAY be used on Delivery Status Notifications (DSNs) and Message
2648 Disposition Notifications (MDNs),
2650 - MUST NOT be used on messages generated by automatic or periodic
2651 processes, except for messages which are automatic responses to
2654 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2655 to the default Received: header definition.
2657 PH/12 Added log selector acl_warn_skipped (default on).
2659 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2660 variables because (a) they are in the wrong storage pool and (b) even if
2661 they were copied, it wouldn't work properly because of the caching.
2663 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2664 checking when decoding. Apparently there are clients that generate
2665 overlong encoded strings. Why am I not surprised?
2667 PH/15 If the first argument of "${if match_address" was not empty, but did not
2668 contain an "@" character, Exim crashed. Now it writes a panic log message
2669 and treats the condition as false.
2671 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2673 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2674 extension "envelope-auth". The code is finished and in agreement with
2675 other implementations, but there is no documentation so far and in fact,
2676 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2677 not changing the active code.
2679 Print executed "if" and "elsif" statements when debugging is used. This
2680 helps a great deal to understand what a filter does.
2682 Document more things not specified clearly in RFC3028. I had all this
2683 sorted out, when out of a sudden new issues came to my mind. Oops."
2685 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2688 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2689 canonical form (as documented). However, after a host name lookup from
2690 the IP address, check_host() was doing a simple string comparison with
2691 addresses acquired from the DNS when checking that the found name did
2692 have the original IP as one of its addresses. Since any found IPv6
2693 addresses are likely to be in abbreviated form, the comparison could
2694 fail. Luckily, there already exists a function for doing the comparison
2695 by converting both addresses to binary, so now that is used instead of
2696 the text comparison.
2698 PH/20 There was another similar case to PH/19, when a complete host name was
2699 given in a host list; looking up its IP address could give an abbreviated
2700 form, whereas the current host's name might or might not be abbreviated.
2701 The same fix has been applied.
2707 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2708 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2711 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2712 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2714 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2716 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2717 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2718 Also, for "Auto-Submitted", treat anything other than "no" as
2719 non-personal, in accordance with RFC 3834. (Previously it treated
2720 anything starting "auto-" as non-personal.)
2722 TF/01 The control=submission/name=... option had a problem with syntax
2723 errors if the name included a slash character. The /name= option
2724 now slurps the rest of the string, so it can include any characters
2725 but it must come last in the list of options (after /sender_retain
2728 PH/05 Some modifications to the interface to the fake nameserver for the new
2736 TK/01 Added the "success_on_redirect" address verification option. See
2737 NewStuff for rationale and an example.
2739 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2741 PH/02 Patch to exigrep to allow it to work on syslog lines.
2743 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2744 fread() to skip over the body file's header line, because in Cygwin the
2745 header line is locked and is inaccessible.
2747 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2748 co-exist for some time) to make it clear that it is the Exim ID that is
2749 referenced, not the Message-ID: header line.
2751 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2752 string_format() function, because snprintf() does not exist on all
2755 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2756 ${stat: expansion item.
2758 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2759 protocol synchronization error", to keep the pedants happy.
2761 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2762 well as for IRIX systems, when gcc is being used. See the host.c source
2765 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2767 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2770 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2771 transport and to the smtp transport in LMTP mode.
2773 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2775 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2776 run a filter in a subprocess. This could lead to confusion in subsequent
2777 lookups in the parent process. There should also be a search_tidyup() at
2778 the end of the subprocess.
2780 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2781 only if the host matched helo_try_verify_hosts, which caused the
2782 verification to occur when the EHLO/HELO command was issued. The ACL just
2783 tested the remembered result. Now, if a previous verification attempt has
2784 not happened, "verify = helo" does it there and then.
2786 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2788 TK/03 Fix log output including CR from clamd.
2790 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2791 exist provoked a memory error which could cause a segfault.
2793 PH/15 Installed PCRE 6.2
2795 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2797 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2798 of the problem. Specifically, suggested +O2 rather than +O1 for the
2801 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2803 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2804 record, the port was not correctly logged when the outgoing_port log
2805 selector was set (it logged the transort's default port).
2807 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2808 fallback_hosts, and "hosts" in the smtp transport.
2810 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2811 host errors such as "Connection refused".
2813 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2814 authentication with radiusclient 0.4.9:
2816 - Error returned from rc_read_config was caught wrongly
2817 - Username/password not passed on to radius server due to wrong length.
2819 The presumption is that some radiusclient API changes for 4.51/PH/17
2820 were not taken care of correctly. The code is still untested by me (my
2821 Linux distribution still has 0.3.2 of radiusclient), but it was
2822 contributed by a Radius user.
2824 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2825 expanding the "port" option of the smtp transport.
2827 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2828 while reading a MIME header. Thanks to Tom Hughes for a patch.
2830 PH/24 Include config.h inside local_scan.h so that configuration settings are
2833 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2834 that in an ACL for RCPT (for example), you can examine exactly what was
2837 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2838 commands, but it was not correctly comparing the address with the actual
2839 client host address. Thus, it would show the EHLO address in Received:
2840 header lines when this was not necessary.
2842 PH/27 Added the % operator to ${eval:}.
2844 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2845 it should be ignoring failures (because with -C, for example, it has lost
2846 privilege). It wasn't ignoring creation failures other than "already
2849 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2852 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2853 because an input line was too long, either on its own, or by virtue of
2854 too many continuations, the temporary file was not being removed, and the
2855 return code was incorrect.
2857 PH/31 Missing "BOOL" in function definition in filtertest.c.
2859 PH/32 Applied Sieve patches from the maintainer.
2861 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2863 PH/33 Added "verify = not_blind".
2865 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2866 Local/Makefile (with some defaults set). These are used in built scripts
2867 such as exicyclog, but they have never been used in the exim_install
2868 script (though there are many overriding facilities there). I have
2869 arranged that the exim_install script now takes note of these two
2872 PH/35 Installed configuration files for Dragonfly.
2874 PH/36 When a locally submitted message by a trusted user did not contain a
2875 From: header, and the sender address was obtained from -f or from an SMTP
2876 MAIL command, and the trusted user did not use -F to supply a sender
2877 name, $originator_name was incorrectly used when constructing a From:
2878 header. Furthermore, $originator_name was used for submission mode
2879 messages from external hosts without From: headers in a similar way,
2880 which is clearly wrong.
2882 PH/37 Added control=suppress_local_fixups.
2884 PH/38 When log_selector = +received_sender was set, and the addition of the
2885 sender made the log line's construction buffer exactly full, or one byte
2886 less than full, an overflow happened when the terminating "\n" was
2889 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2890 when the result of a list match is failure because a DNS lookup failed.
2892 PH/40 RM_COMMAND is now used in the building process.
2894 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2895 the "build-* directories that it finds.
2897 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2898 address in a domain literal was a prefix of an interface address.
2900 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2901 when verifying a sender address, unless rewrite_headers is false.
2903 PH/44 Wrote a long comment about why errors_to addresses are verified as
2904 recipients, not senders.
2906 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2907 the ratelimit ACL was added.
2909 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2911 PH/46 Added extra information about PostgreSQL errors to the error string.
2913 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2914 avoiding the need to install special zones in a real server. This is
2915 backwards compatible; if it can't find the fake resolver, it drops back.
2916 Thus, both old and new test suites can be run.
2918 TF/02 Added util/ratelimit.pl
2920 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2921 clock is set back in time.
2923 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2924 Candler <B.Candler@pobox.com>.
2926 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2927 OK for addresses that are the result of redirections.
2929 PH/48 A number of further additions for the benefit of the new test suite,
2930 including a fake gethostbyname() that interfaces to the fake DNS resolver
2933 TF/06 The fix for widen_domains has also been applied to qualify_single and
2934 search_parents which are the other dnslookup options that can cause
2937 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2940 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2942 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2943 service" if it encountered an MX record with an empty target hostname.
2944 The message is now "an MX or SRV record indicated no SMTP service".
2946 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2947 if the DNS of the sending site is misconfigured. This is quite a
2948 common situation. This change restores the behaviour of treating a
2949 helo verification defer as a failure.
2951 PH/51 If self=fail was set on a router, the bounce message did not include the
2952 actual error message.
2958 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2960 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2961 command itself ended in error, the underlying message about the transport
2962 filter timeout was being overwritten with the pipe command error. Now the
2963 underlying error message should be appended to the second error message.
2965 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2967 PH/02 Reduce the amount of output that "make" produces by default. Full output
2968 can still be requested.
2970 PH/03 The warning log line about a condition test deferring for a "warn" verb
2971 was being output only once per connection, rather than after each
2972 occurrence (because it was using the same function as for successful
2973 "warn" verbs). This seems wrong, so I have changed it.
2975 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2976 it should not have, which might have caused a crash in the right
2977 circumstances, but probably never did.
2979 PH/04 Installed a modified version of Tony Finch's patch to make submission
2980 mode fix the return path as well as the Sender: header line, and to
2981 add a /name= option so that you can make the user's friendly name appear
2984 TF/03 Added the control = fakedefer ACL modifier.
2986 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2987 Mark Lowes for thorough testing.
2989 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2991 TK/03 Merged latest SRS patch from Miles Wilton.
2993 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2994 with the definition in sysexits.h (which is #included earlier).
2995 Fortunately, Exim does not actually use EX_OK. The code used to try to
2996 preserve the sysexits.h value, by assumimg that macro definitions were
2997 scanned for macro replacements. I have been disabused of this notion,
2998 so now the code just undefines EX_OK before #including unistd.h.
3000 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
3001 in the smtp transport. When a block could not be written in a single
3002 write() function, the timeout was being re-applied to each part-write.
3003 This seems wrong - if the receiver was accepting one byte at a time it
3004 would take for ever. The timeout is now adjusted when this happens. It
3005 doesn't have to be particularly precise.
3007 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
3008 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
3010 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
3011 without a domain if the check to <postmaster@domain> fails.
3013 SC/01 Eximstats: added -xls and the ability to specify output files
3014 (patch written by Frank Heydlauf).
3016 SC/02 Eximstats: use FileHandles for outputing results.
3018 SC/03 Eximstats: allow any combination of xls, txt, and html output.
3020 SC/04 Eximstats: fixed display of large numbers with -nvr option
3022 SC/05 Eximstats: fixed merging of reports with empty tables.
3024 SC/06 Eximstats: added the -include_original_destination flag
3026 SC/07 Eximstats: removed tabs and trailing whitespace.
3028 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
3030 TK/06 MBOX spool code: Add real "From " MBOX separator line
3031 so the .eml file is really in mbox format (even though
3032 most programs do not really care). Patch from Alex Miller.
3034 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
3035 The latter is generated from $received_to and is only set if the
3036 message has one envelope recipient. SA can use these headers,
3037 obviously out-of-the-box. Patch from Alex Miller.
3039 PH/08 The ${def test on a variable was returning false if the variable's
3040 value was "0", contrary to what the specification has always said!
3041 The result should be true unless the variable is empty.
3043 PH/09 The syntax error of a character other than { following "${if
3044 def:variable_name" (after optional whitespace) was not being diagnosed.
3045 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3046 accidental colon was present, for example, could give incorrect results.
3048 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3049 result is used (not including appendfile, where other changes are about
3052 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3053 This involved changing a lot of size variables from int to off_t. It
3054 should work with maildirs and everything.
3056 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3057 spamd dying while we are connected to it.
3059 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3062 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3063 function for BDB 4.3.
3065 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3067 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3068 the sources for PCRE have been re-arranged and more files are now
3071 PH/15 The code I had for printing potentially long long variables in PH/11
3072 above was not the best (it lost precision). The length of off_t variables
3073 is now inspected at build time, and an appropriate printing format (%ld
3074 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3075 to be "long long int" or "long int". This is needed for the internal
3076 formatting function string_vformat().
3078 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3079 the configuration file to be ":syslog", then the script "guesses" where
3080 the logs files are, rather than using the compiled in default. In our
3081 case the guess is not the same as the compiled default, so the script
3082 suddenly stopped working when I started to use syslog. The patch checks
3083 to see if log_file_path is "". If so, it attempts to read it from exim
3084 with no configuration file to get the compiled in version, before it
3085 falls back to the previous guessing code."
3087 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3088 implementing BATV in an Exim configuration. See NewStuff for the gory
3091 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3092 Makefile that are specific to HP-UX.
3094 PH/18 If the "use_postmaster" option was set for a recipient callout together
3095 with the "random" option, the postmaster address was used as the MAIL
3096 FROM address for the random test, but not for the subsequent recipient
3097 test. It is now used for both.
3099 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3100 patch removes a few documentation additions to RFC 3028, because the
3101 latest draft now contains them. It adds the new en;ascii-case comparator
3102 and a new error check for 8bit text in MIME parts. Comparator and
3103 require names are now matched exactly. I enabled the subaddress
3104 extension, but it is not well tested yet (read: it works for me)."
3106 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3107 rework some of the code of TK/09 above to avoid the hardwired use of
3108 "%lld" and "long long". Replaced the call to snprintf() with a call to
3111 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3112 records point to non-existent hosts", "retry timeout exceeded", and
3113 "retry time not reached for any host after a long failure period".
3115 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3116 experimental DomainKeys support:
3118 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3119 (2) On an error such as an illegally used "control", the wrong name for
3120 the control was given.
3122 These problems did NOT occur unless DomainKeys support was compiled.
3124 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3126 PH/24 Added ${if match_ip condition.
3128 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3129 databases so that it will be absolutely obvious if a crash occurs in the
3130 DB library. This is a regular occurrence (often caused by mis-matched
3133 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3134 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3135 picked up on a user's system that detects such things. There doesn't seem
3136 to be a gcc warning option for this - only an attribute that has to be
3137 put on the function's prototype. It seems that in Fedora Core 4 they have
3138 set this on a number of new functions. No doubt there will be more in due
3141 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3142 specify a transport. However, if an address that was verified by such a
3143 router was the subject of a callout, Exim crashed because it tried to
3144 read the rcpt_include_affixes from the non-existent transport. Now it
3145 just assumes that the setting of that option is false. This bug was
3146 introduced by 4.51/PH/31.
3148 PH/28 Changed -d+all to exclude +memory, because that information is very
3149 rarely of interest, but it makes the output a lot bigger. People tend to
3150 do -d+all out of habit.
3152 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3153 code in os-type was giving problems when libc.so lives in lib64, like on
3156 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3157 aren't the modern standard, and it seems that some systems' include files
3158 don't always have them. Exim was already checking for some of the newer
3159 ones like T_AAAA, and defining it itself. I've added checks for all the
3160 record types that Exim uses.
3162 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3163 not automatically generating a new one, as it is supposed to. This
3164 prevented TLS from working. If the file did exist, but contained invalid
3165 data, a new version was generated, as expected. It was only the case of a
3166 non-existent file that was broken.
3168 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3169 with a change in libdomainkeys > 0.64.
3171 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3172 from DNS. If the selector record carries the flag, it now has
3173 precedence over the domain-wide flag.
3175 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3177 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3178 the use of a port name that isn't defined in /etc/services) occurred, the
3179 message was deferred as in a normal delivery, and thus remained on the
3180 spool, instead of being failed because of the mua_wrapper setting. This
3181 is now fixed, and I tidied up some of the mua_wrapper messages at the
3184 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3185 the messages in a hash of arrays rather than using individual hashes.
3186 This is a bit cleaner and results in dramatic memory savings, albeit
3187 at a slight CPU cost.
3189 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3190 as requested by Marc Sherman.
3192 SC/10 Eximstats: added histograms for user specified patterns as requested
3195 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3197 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3198 fopen() in the content-scanning modules that did not already have it.
3204 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3205 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3207 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3209 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3211 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3212 to test Sieve filters that use "vacation".
3214 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3215 that changes the way the GnuTLS parameters are stored in the cache file.
3216 The new format can be generated externally. For backward compatibility,
3217 if the data in the cache doesn't make sense, Exim assumes it has read an
3218 old-format file, and it generates new data and writes a new file. This
3219 means that you can't go back to an older release without removing the
3222 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3223 work if there are any delivery delays because "one_time" forces the
3224 parent to be marked "delivered", so its unseen clone is never tried
3225 again. For this reason, Exim now forbids the simultaneous setting of
3228 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3229 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3230 entry if you want to know the details. The fix, however, neglected to
3231 consider the case when local delivery batching is involved. The test for
3232 "previously delivered" was not happening when checking to see if an
3233 address could be batched with a previous (undelivered) one; under
3234 certain circumstances this could lead to multiple deliveries to the same
3237 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3238 in its include files, and this causes problems building Exim.
3240 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3241 header_syntax) but Exim was just ignoring anything given after a slash.
3242 In particular, this caused confusion with an attempt to use "verify =
3243 reverse_host_lookup/defer_ok". An error is now given when options are
3244 supplied for verify items that do not have them. (Maybe reverse_host_
3245 lookup should have a defer_ok option, but that's a different point.)
3247 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3248 defined by RFC 821) to 2048, because there were problems with some AUTH
3249 commands, and RFC 1869 says the size should be increased for extended
3250 SMTP commands that take arguments.
3252 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3255 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3256 "unknown" error; now it says that the functionality isn't in the binary.
3258 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3259 an address' error message when a string expansion fails (syntax or
3260 whatever). Otherwise the password may appear in the log. Following change
3261 PH/42 below, there is no longer a chance of it appearing in a bounce
3264 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3266 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3267 of its addresses was ever tried. (Bugzilla bug #2.)
3269 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3270 the result incorrectly in the debug output. (It correctly added a newline
3271 to what was transported.)
3273 TF/01 Added $received_time.
3275 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3276 commented out examples of how to interface to a virus scanner and to
3277 SpamAssassin. Also added commented examples of av_scanner and
3278 spamd_address settings.
3280 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3281 and controls are allowed in which ACLs. There were a couple of minor
3282 errors. Some of the entries in the conditions table (which is a table of
3283 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3284 negation of where the condition IS allowed.
3286 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3288 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3289 header file does not have a version number, so I've had to invent a new
3290 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3291 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3292 radiusclient), but it was contributed by a Radius user.
3294 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3295 files or directories, for OpenSSL.
3297 PH/19 When an Exim process that is running as root has to create an Exim log
3298 file, it does so in a subprocess that runs as exim:exim so as to get the
3299 ownership right at creation (otherwise, other Exim processes might see
3300 the file with the wrong ownership). There was no test for failure of this
3301 fork() call, which would lead to the process getting stuck as it waited
3302 for a non-existent subprocess. Forks do occasionally fail when resources
3303 run out. I reviewed all the other calls to fork(); they all seem to check
3306 PH/20 When checking for unexpected SMTP input at connect time (before writing
3307 the banner), Exim was not dealing correctly with a non-positive return
3308 from the read() function. If the client had disconnected by this time,
3309 the result was a log entry for a synchronization error with an empty
3310 string after "input=" when read() returned zero. If read() returned -1
3311 (an event I could not check), uninitialized data bytes were printed.
3312 There were reports of junk text (parts of files, etc) appearing after
3315 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3317 PH/22 Added support for macro redefinition, and (re)definition in between
3318 driver and ACL definitions.
3320 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3321 forgetting to use the resulting value; it was using the unexpanded value.
3323 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3324 hadn't been configured. The fix is from Juergen Kreileder, who
3325 understands it better than I do:
3327 "Here's what I see happening with three configured cyrus_sasl
3328 authenticators configured (plain, login, cram-md5):
3330 On startup auth_cyrus_sasl_init() gets called for each of these.
3331 This means three calls to sasl_listmech() without a specified mech_list.
3332 => SASL tests which mechs of all available mechs actually work
3333 => three warnings about OTP not working
3334 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3336 With the patch, sasl_listmech() also gets called three times. But now
3337 SASL's mech_list option is set to the server_mech specified in the the
3338 authenticator. Or in other words, the answer from sasl_listmech()
3339 gets limited to just the mech you're testing for (which is different
3341 => the return list contains just 'plain' or 'login', 'cram-md5' or
3342 nothing depending on the value of ob->server_mech.
3344 I've just tested the patch: Authentication still works fine,
3345 unavailable mechs specified in the exim configuration are still
3346 caught, and the auth.log warnings about OTP are gone."
3348 PH/25 When debugging is enabled, the contents of the command line are added
3349 to the debugging output, even when log_selector=+arguments is not
3352 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3353 answer is "GNU", and only if the return is "GNU/something" is the answer
3356 PH/27 $acl_verify_message is now set immediately after the failure of a
3357 verification in an ACL, and so is available in subsequent modifiers. In
3358 particular, the message can be preserved by coding like this:
3360 warn !verify = sender
3361 set acl_m0 = $acl_verify_message
3363 Previously, $acl_verify_message was set only while expanding "message"
3364 and "log_message" when a very denied access.
3366 PH/28 Modified OS/os.c-Linux with
3368 -#ifndef OS_LOAD_AVERAGE
3369 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3371 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3372 nomenclature these days.)
3374 PH/29 Installed patch from the Sieve maintainer that adds the options
3375 sieve_useraddress and sieve_subaddress to the redirect router.
3377 PH/30 In these circumstances:
3378 . Two addresses routed to the same list of hosts;
3379 . First host does not offer TLS;
3380 . First host accepts first address;
3381 . First host gives temporary error to second address;
3382 . Second host offers TLS and a TLS session is established;
3383 . Second host accepts second address.
3384 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3385 and peerdn, if requested) that were in fact used only for the second
3388 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3389 attention to any local part prefix or suffix that was matched by the
3390 router that accepted the address. It now behaves in the same way as it
3391 does for delivery: the affixes are removed from the local part unless
3392 rcpt_include_affixes is set on the transport.
3394 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3395 timeout during the DATA phase of an incoming message.
3397 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3398 applied a patch sanctioned by the Sieve maintainer.
3400 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3401 the uid or gid is negative. A case of a negative gid caused this to be
3402 noticed. The fix allows for either to be negative.
3404 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3405 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3408 PH/36 The change PH/12 above was broken. Fixed it.
3410 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3411 the grounds that routing the same address twice would always produce the
3412 same answer. This might have been true once, but it is certainly no
3413 longer true now. Routing a child address may depend on the previous
3414 routing that produced that child. Some complicated redirection strategies
3415 went wrong when messages had multiple recipients, and made Exim's
3416 behaviour dependent on the order in which the addresses were given.
3418 I have moved the duplicate checking until after the routing is complete.
3419 Exim scans the addresses that are assigned to local and remote
3420 transports, and removes any duplicates. This means that more work will be
3421 done, as duplicates will always all be routed, but duplicates are
3422 presumably rare, so I don't expect this is of any significance.
3424 For deliveries to pipes, files, and autoreplies, the duplicate checking
3425 still happens during the routing process, since they are not going to be
3428 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3429 It corrects a timeout issue with spamd. This is Ian's comment: "The
3430 background is that sometimes spamd either never reads data from a
3431 connection it has accepted, or it never writes response data. The exiscan
3432 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3433 blindly assumes that writes won't block so it may never time out."
3435 PH/39 Allow G after quota size as well as K and M.
3437 PH/40 The value set for $authenticated_id in an authenticator may not contain
3438 binary zeroes or newlines because the value is written to log lines and
3439 to spool files. There was no check on this. Now the value is run through
3440 the string_printing() function so that such characters are converted to
3441 printable escape sequences.
3443 PH/41 $message_linecount is a new variable that contains the total number of
3444 lines in the message. Compare $body_linecount, which is the count for the
3447 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3448 bounce and delay warning messages, except in certain special cases, which
3451 (a) An SMTP error message from a remote host;
3452 (b) A message specified in a :fail: redirection;
3453 (c) A message specified in a "fail" command in a system filter;
3454 (d) A message specified in a FAIL return from the queryprogram router;
3455 (e) A message specified by the cannot_route_message router option.
3457 In these cases only, Exim does include the error details in bounce and
3458 warning messages. There are also a few cases where bland messages such
3459 as "unrouteable address" or "local delivery error" are given.
3461 PH/43 $value is now also set for the "else" part of a ${run expansion.
3463 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3464 being worked on, but at least Exim now implements the latest version to
3467 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3468 process to complete was treated as a delivery failure, a timeout while
3469 writing the message to the pipe was logged, but erroneously treated as a
3470 successful delivery. Such timeouts include transport filter timeouts. For
3471 consistency with the overall process timeout, these timeouts are now
3472 treated as errors, giving rise to delivery failures by default. However,
3473 there is now a new Boolean option for the pipe transport called
3474 timeout_defer, which, if set TRUE, converts the failures into defers for
3475 both kinds of timeout. A transport filter timeout is now identified in
3478 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3479 systems where "make" and "gmake" are different, calling "gmake" at top
3480 level broke things. I've arranged for the value of $(MAKE) to be passed
3481 from the Makefile to this script so that it can call the same version of
3485 A note about Exim versions 4.44 and 4.50
3486 ----------------------------------------
3488 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3489 changes of various kinds. As a consequence, a big documentation update was
3490 needed. This delayed the release for rather longer than seemed good, especially
3491 in the light of a couple of (minor) security issues. Therefore, the changes
3492 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3493 release. So 4.44 and 4.50 are in effect two different branches that both start
3496 I have left the 4.50 change log unchanged; it contains all the changes since
3497 4.43. The change log for 4.44 is below; many of its items are identical to
3498 those for 4.50. This seems to be the most sensible way to preserve the
3499 historical information.
3505 1. Minor wording change to the doc/README.SIEVE file.
3507 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3508 computation of the current number of files was incorrect.
3510 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3511 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3514 4. Give more explanation in the error message when the command for a transport
3515 filter fails to execute.
3517 5. There are several places where Exim runs a non-Exim command in a
3518 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3519 was being done only for the command run by the queryprogram router. It is
3520 now done for all such subprocesses. The other cases are: ${run, transport
3521 filters, and the commands run by the lmtp and pipe transports.
3523 6. Added CONFIGURE_GROUP build-time option.
3525 7. Some older OS have a limit of 256 on the maximum number of file
3526 descriptors. Exim was using setrlimit() to set 1000 as a large value
3527 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3528 systems. I've change it so that if it can't get 1000, it tries for 256.
3530 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3531 was an oversight, and furthermore, ever since the addition of extra
3532 controls (e.g. 4.43/32), the checks on when to allow different forms of
3533 "control" were broken. There should now be diagnostics for all cases when a
3534 control that does not make sense is encountered.
3536 9. Added the /retain_sender option to "control=submission".
3538 10. $recipients is now available in the predata ACL (oversight).
3540 11. Tidy the search cache before the fork to do a delivery from a message
3541 received from the command line. Otherwise the child will trigger a lookup
3542 failure and thereby defer the delivery if it tries to use (for example) a
3543 cached ldap connection that the parent has called unbind on.
3545 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3546 of $address_data from the recipient verification was clobbered by the
3547 sender verification.
3549 13. The value of address_data from a sender verification is now available in
3550 $sender_address_data in subsequent conditions in the ACL statement.
3552 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3554 15. Added a new option "connect=<time>" to callout options, to set a different
3557 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3558 was its contents. (It was OK if the option was not defined at all.)
3560 17. A "Completed" log line is now written for messages that are removed from
3561 the spool by the -Mrm option.
3563 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3564 information about exactly what failed.
3566 19. Added -dd to debug only the daemon process.
3568 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3569 handles timeouts, both on the server side and network timeouts. Renamed the
3570 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3572 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3573 It is now set to "smtps".
3575 22. $host_address is now set to the target address during the checking of
3576 ignore_target_hosts.
3578 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3579 being passed; this would have caused $sender_host_name to have been used if
3580 matching the list had actually called for a host name (not very likely,
3581 since this list is usually IP addresses). A host name is now passed as
3584 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3585 code that specifies a non-restarting handler (typically sigaction() in
3586 modern systems) in an attempt to fix a rare and obscure crash bug.
3588 25. Narrowed the window for a race in the daemon that could cause it to ignore
3589 SIGCHLD signals. This is not a major problem, because they are used only to
3590 wake it up if nothing else does.
3592 26. A malformed maildirsize file could cause Exim to calculate negative values
3593 for the mailbox size or file count. Odd effects could occur as a result.
3594 The maildirsize information is now recalculated if the size or filecount
3597 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3598 support for a long time. Removed HAVE_SYS_VFS_H.
3600 28. Installed the latest version of exipick from John Jetmore.
3602 29. In an address list, if the pattern was not a regular expression, an empty
3603 subject address (from a bounce message) matched only if the pattern was an
3604 empty string. Non-empty patterns were not even tested. This was the wrong
3605 because it is perfectly reasonable to use an empty address as part of a
3606 database query. An empty address is now tested by patterns that are
3607 lookups. However, all the other forms of pattern expect the subject to
3608 contain a local part and a domain, and therefore, for them, an empty
3609 address still always fails if the pattern is not itself empty.
3611 30. Exim went into a mad DNS loop when attempting to do a callout where the
3612 host was specified on an smtp transport, and looking it up yielded more
3613 than one IP address.
3615 31. Re-factored the code for checking spool and log partition space into a
3616 function that finds that data and another that does the check. The former
3617 is then used to implement four new variables: $spool_space, $log_space,
3618 $spool_inodes, and $log_inodes.
3620 32. The RFC2047 encoding function was originally intended for short strings
3621 such as real names; it was not keeping to the 75-character limit for
3622 encoded words that the RFC imposes. It now respects the limit, and
3623 generates multiple encoded words if necessary. To be on the safe side, I
3624 have increased the buffer size for the ${rfc2047: expansion operator from
3627 33. It is now permitted to omit both strings after an "if" condition; if the
3628 condition is true, the result is "true". As before, when the second string
3629 is omitted, a false condition yields an empty string. This makes it less
3630 cumbersome to write custom ACL and router conditions.
3632 34. Failure to deliver a bounce message always caused it to be frozen, even if
3633 there was an errors_to setting on the router. The errors_to setting is now
3636 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3637 canonical form (fully expanded) before being placed in
3638 $sender_host_address.
3640 36. The table in the code that translates DNS record types into text (T_A to
3641 "A" for instance) was missing entries for NS and CNAME. It is just possible
3642 that this could have caused confusion if both these types were looked up
3643 for the same domain, because the text type is used as part of Exim's
3644 per-process caching. But the chance of anyone hitting this buglet seems
3647 37. The dnsdb lookup has been extended in a number of ways.
3649 (1) There is a new type, "zns", which walks up the domain tree until it
3650 finds some nameserver records. It should be used with care.
3652 (2) There is a new type, "mxh", which is like "mx" except that it returns
3653 just the host names, not the priorities.
3655 (3) It is now possible to give a list of domains (or IP addresses) to be
3656 looked up. The behaviour when one of the lookups defers can be
3657 controlled by a keyword.
3659 (4) It is now possible to specify the separator character for use when
3660 multiple records are returned.
3662 38. The dnslists ACL condition has been extended: it is now possible to supply
3663 a list of IP addresses and/or domains to be looked up in a particular DNS
3666 39. Added log_selector=+queue_time_overall.
3668 40. When running the queue in the test harness, wait just a tad after forking a
3669 delivery process, to get repeatability of debugging output.
3671 41. Include certificate and key file names in error message when GnuTLS fails
3672 to set them up, because the GnuTLS error message doesn't include the name
3673 of the failing file when there is a problem reading it.
3675 42. Allow both -bf and -bF in the same test run.
3677 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3679 44. The "Exiscan patch" is now merged into the mainline Exim source.
3681 45. Sometimes the final signoff response after QUIT could fail to get
3682 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3683 < 0 before doing a fflush(). This bug looks as though it goes back to the
3684 introduction of TLS in release 3.20, but "sometimes" must have been rare
3685 because the tests only now provoked it.
3687 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3688 (this can affect the format of dates).
3690 47. exim_tidydb, when checking for the continued existence of a message for
3691 which it has found a message-specific retry record, was not finding
3692 messages that were in split spool directories. Consequently, it was
3693 deleting retry records that should have stayed in existence.
3695 48. Steve fixed some bugs in eximstats.
3697 49. The SPA authentication driver was not abandoning authentication and moving
3698 on to the next authenticator when an expansion was forced to fail,
3699 contradicting the general specification for all authenticators. Instead it
3700 was generating a temporary error. It now behaves as specified.
3702 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3703 (the order specifies the preference for clients). The order is now AES256,
3704 AES128, 3DES, ARCFOUR128.
3706 51. Small patch to Sieve code - explicitly set From: when generating an
3709 52. Exim crashed if a remote delivery caused a very long error message to be
3710 recorded - for instance if somebody sent an entire SpamAssassin report back
3711 as a large number of 550 error lines. This bug was coincidentally fixed by
3712 increasing the size of one of Exim's internal buffers (big_buffer) that
3713 happened as part of the Exiscan merge. However, to be on the safe side, I
3714 have made the code more robust (and fixed the comments that describe what
3717 53. Now that there can be additional text after "Completed" in log lines (if
3718 the queue_time_overall log selector is set), a one-byte patch to exigrep
3719 was needed to allow it to recognize "Completed" as not the last thing in
3722 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3723 patch that reportedly fixes this has been added. I am not expert enough to
3724 create a test for it. This is what the patch creator wrote:
3726 "I found a little strange behaviour of ldap code when working with
3727 Windows 2003 AD Domain, where users was placed in more than one
3728 Organization Units. When I tried to give exim partial DN, the exit code
3729 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3730 But simultaneously result of request was absolutely normal ldap result,
3731 so I produce this patch..."
3733 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3734 REFERENCE, so I have modified the code to exclude the patch when that macro
3737 55. Some experimental protocols are using DNS PTR records for new purposes. The
3738 keys for these records are domain names, not reversed IP addresses. The
3739 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3740 leaves it alone. Component reversal etc. now happens only for IP addresses.
3743 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3745 57. Double the size of the debug message buffer (to 2048) so that more of very
3746 long debug lines gets shown.
3748 58. The exicyclog utility now does better if the number of log files to keep
3749 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3751 59. Two changes related to the smtp_active_hostname option:
3753 (1) $smtp_active_hostname is now available as a variable.
3754 (2) The default for smtp_banner uses $smtp_active_hostname instead
3755 of $primary_hostname.
3757 60. The host_aton() function is supposed to be passed a string that is known
3758 to be a valid IP address. However, in the case of IPv6 addresses, it was
3759 not checking this. This is a hostage to fortune. Exim now panics and dies
3760 if the condition is not met. A case was found where this could be provoked
3761 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3762 components; fortuitously, this particular loophole had already been fixed
3763 by change 4.50/55 above.
3765 If there are any other similar loopholes, the new check in host_aton()
3766 itself should stop them being exploited. The report I received stated that
3767 data on the command line could provoke the exploit when Exim was running as
3768 exim, but did not say which command line option was involved. All I could
3769 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3770 running as the user.
3773 61. There was a buffer overflow vulnerability in the SPA authentication code
3774 (which came originally from the Samba project). I have added a test to the
3775 spa_base64_to_bits() function which I hope fixes it.
3778 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3779 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3781 63. The daemon start-up calls getloadavg() while still root for those OS that
3782 need the first call to be done as root, but it missed one case: when
3783 deliver_queue_load_max is set with deliver_drop_privilege. This is
3784 necessary for the benefit of the queue runner, because there is no re-exec
3785 when deliver_drop_privilege is set.
3787 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3788 This has been fixed.
3790 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3791 was in use, was not putting the data itself into the right store pool;
3792 consequently, it could be overwritten for a subsequent message in the same
3793 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3796 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3798 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3799 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3800 the function were treating the return as a boolean value, which happened to
3801 work because 0=false and not-0=true, but is not correct code.
3803 68. The host_aton() function was not handling scoped IPv6 addresses (those
3804 with, for example, "%eth0" on the end) correctly.
3806 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3807 negated items (that is, ~something) in unsigned ints. Some compilers
3808 apparently mutter when there is no cast.
3810 70. If an address verification called from an ACL failed, and did not produce a
3811 user-specific message (i.e. there was only a "system" message), nothing was
3812 put in $acl_verify_message. In this situation, it now puts the system
3815 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3816 session; change 4.31/43 added the unwanted input to the log line - except
3817 that it did not do this in the start of session case. It now does.
3819 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3820 This is wrong and can cause the other end to generate a synchronization
3821 error if it is another Exim or anything else that does the synchronization
3822 check. A QUIT command is no longer sent after a timeout.
3824 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3825 during host lookups.
3827 74. The defer_ok option of callout verification was not working if it was used
3828 when verifying addresses in header lines, that is, for this case:
3830 verify = header_sender/callout=defer_ok
3832 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3833 those file descriptors could be used for SMTP connections. If anything
3834 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3835 could be sent to the SMTP client, causing chaos. The daemon now opens
3836 stdin, stdout, and stderr to /dev/null when it puts itself into the
3839 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3840 log by default. The user can override this with suitable Perl magic.
3842 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3843 the log message when discard triggers, was not working for the DATA ACL or
3844 for the non-SMTP ACL.
3846 78. Error message wording change in sieve.c.
3848 79. If smtp_accept_max_per_host was set, the number of connections could be
3849 restricted to fewer than expected, because the daemon was trying to set up
3850 a new connection before checking whether the processes handling previous
3851 connections had finished. The check for completed processes is now done
3852 earlier. On busy systems, this bug wouldn't be noticed because something
3853 else would have woken the daemon, and it would have reaped the completed
3856 80. If a message was submitted locally by a user whose login name contained one
3857 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3858 It caused a spool format error. I have fixed the spool reading code. A
3859 related problem was that the "from" clause in the Received: line became
3860 illegal because of the space(s). It is now covered by ${quote_local_part.
3862 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3865 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3867 83. Patch from Sieve maintainer to fix unterminated string problem in
3868 "vacation" handling.
3870 84. Some minor changes to the Linux configuration files to help with other
3871 OS variants using glibc.
3873 85. One more patch for Sieve to update vacation handling to latest spec.
3876 ----------------------------------------------------
3877 See the note above about the 4.44 and 4.50 releases.
3878 ----------------------------------------------------
3884 1. Change 4.43/35 introduced a bug that caused file counts to be
3885 incorrectly computed when quota_filecount was set in an appendfile
3888 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3889 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3892 3. Give more explanation in the error message when the command for a transport
3893 filter fails to execute.
3895 4. There are several places where Exim runs a non-Exim command in a
3896 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3897 was being done only for the command run by the queryprogram router. It is
3898 now done for all such subprocesses. The other cases are: ${run, transport
3899 filters, and the commands run by the lmtp and pipe transports.
3901 5. Some older OS have a limit of 256 on the maximum number of file
3902 descriptors. Exim was using setrlimit() to set 1000 as a large value
3903 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3904 systems. I've change it so that if it can't get 1000, it tries for 256.
3906 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3907 was an oversight, and furthermore, ever since the addition of extra
3908 controls (e.g. 4.43/32), the checks on when to allow different forms of
3909 "control" were broken. There should now be diagnostics for all cases when a
3910 control that does not make sense is encountered.
3912 7. $recipients is now available in the predata ACL (oversight).
3914 8. Tidy the search cache before the fork to do a delivery from a message
3915 received from the command line. Otherwise the child will trigger a lookup
3916 failure and thereby defer the delivery if it tries to use (for example) a
3917 cached ldap connection that the parent has called unbind on.
3919 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3920 of $address_data from the recipient verification was clobbered by the
3921 sender verification.
3923 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3924 was its contents. (It was OK if the option was not defined at all.)
3926 11. A "Completed" log line is now written for messages that are removed from
3927 the spool by the -Mrm option.
3929 12. $host_address is now set to the target address during the checking of
3930 ignore_target_hosts.
3932 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3933 being passed; this would have caused $sender_host_name to have been used if
3934 matching the list had actually called for a host name (not very likely,
3935 since this list is usually IP addresses). A host name is now passed as
3938 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3939 code that specifies a non-restarting handler (typically sigaction() in
3940 modern systems) in an attempt to fix a rare and obscure crash bug.
3942 15. Narrowed the window for a race in the daemon that could cause it to ignore
3943 SIGCHLD signals. This is not a major problem, because they are used only to
3944 wake it up if nothing else does.
3946 16. A malformed maildirsize file could cause Exim to calculate negative values
3947 for the mailbox size or file count. Odd effects could occur as a result.
3948 The maildirsize information is now recalculated if the size or filecount
3951 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3952 support for a long time. Removed HAVE_SYS_VFS_H.
3954 18. Updated exipick to current release from John Jetmore.
3956 19. Allow an empty sender to be matched against a lookup in an address list.
3957 Previously the only cases considered were a regular expression, or an
3960 20. Exim went into a mad DNS lookup loop when doing a callout where the
3961 host was specified on the transport, if the DNS lookup yielded more than
3964 21. The RFC2047 encoding function was originally intended for short strings
3965 such as real names; it was not keeping to the 75-character limit for
3966 encoded words that the RFC imposes. It now respects the limit, and
3967 generates multiple encoded words if necessary. To be on the safe side, I
3968 have increased the buffer size for the ${rfc2047: expansion operator from
3971 22. Failure to deliver a bounce message always caused it to be frozen, even if
3972 there was an errors_to setting on the router. The errors_to setting is now
3975 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3976 canonical form (fully expanded) before being placed in
3977 $sender_host_address.
3979 24. Updated eximstats to version 1.33
3981 25. Include certificate and key file names in error message when GnuTLS fails
3982 to set them up, because the GnuTLS error message doesn't include the name
3983 of the failing file when there is a problem reading it.
3985 26. Expand error message when OpenSSL has problems setting up cert/key files.
3988 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3989 (this can affect the format of dates).
3991 28. exim_tidydb, when checking for the continued existence of a message for
3992 which it has found a message-specific retry record, was not finding
3993 messages that were in split spool directories. Consequently, it was
3994 deleting retry records that should have stayed in existence.
3996 29. eximstats updated to version 1.35
3997 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3998 1.35 - bugfix such that pie charts by volume are generated correctly
4000 30. The SPA authentication driver was not abandoning authentication and moving
4001 on to the next authenticator when an expansion was forced to fail,
4002 contradicting the general specification for all authenticators. Instead it
4003 was generating a temporary error. It now behaves as specified.
4005 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
4006 (the order specifies the preference for clients). The order is now AES256,
4007 AES128, 3DES, ARCFOUR128.
4009 31. Small patch to Sieve code - explicitly set From: when generating an
4012 32. Exim crashed if a remote delivery caused a very long error message to be
4013 recorded - for instance if somebody sent an entire SpamAssassin report back
4014 as a large number of 550 error lines. This bug was coincidentally fixed by
4015 increasing the size of one of Exim's internal buffers (big_buffer) that
4016 happened as part of the Exiscan merge. However, to be on the safe side, I
4017 have made the code more robust (and fixed the comments that describe what
4020 33. Some experimental protocols are using DNS PTR records for new purposes. The
4021 keys for these records are domain names, not reversed IP addresses. The
4022 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
4023 leaves it alone. Component reversal etc. now happens only for IP addresses.
4026 34. The host_aton() function is supposed to be passed a string that is known
4027 to be a valid IP address. However, in the case of IPv6 addresses, it was
4028 not checking this. This is a hostage to fortune. Exim now panics and dies
4029 if the condition is not met. A case was found where this could be provoked
4030 from a dnsdb PTR lookup with an IPv6 address that had more than 8
4031 components; fortuitously, this particular loophole had already been fixed
4032 by change 4.50/55 or 4.44/33 above.
4034 If there are any other similar loopholes, the new check in host_aton()
4035 itself should stop them being exploited. The report I received stated that
4036 data on the command line could provoke the exploit when Exim was running as
4037 exim, but did not say which command line option was involved. All I could
4038 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
4039 running as the user.
4042 35. There was a buffer overflow vulnerability in the SPA authentication code
4043 (which came originally from the Samba project). I have added a test to the
4044 spa_base64_to_bits() function which I hope fixes it.
4047 36. The daemon start-up calls getloadavg() while still root for those OS that
4048 need the first call to be done as root, but it missed one case: when
4049 deliver_queue_load_max is set with deliver_drop_privilege. This is
4050 necessary for the benefit of the queue runner, because there is no re-exec
4051 when deliver_drop_privilege is set.
4053 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4054 was in use, was not putting the data itself into the right store pool;
4055 consequently, it could be overwritten for a subsequent message in the same
4056 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4059 38. Sometimes the final signoff response after QUIT could fail to get
4060 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4061 < 0 before doing a fflush(). This bug looks as though it goes back to the
4062 introduction of TLS in release 3.20, but "sometimes" must have been rare
4063 because the tests only now provoked it.
4069 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4070 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4071 Now it returns an integer. A number of calls were still expecting a T/F
4072 return. Fortuitously, in all cases, the tests worked in OK situations,
4073 which is the norm. However, things would have gone wrong on any write
4074 failures on the smtp file descriptor. This function is used when sending
4075 messages over SMTP and also when doing verify callouts.
4077 2. When Exim is called to do synchronous delivery of a locally submitted
4078 message (the -odf or -odi options), it no longer closes stderr before doing
4081 3. Implemented the mua_wrapper option.
4083 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4085 5. Implemented the functions header_remove(), header_testname(),
4086 header_add_at_position(), and receive_remove_recipient(), and exported them
4089 6. If an ACL "warn" statement specified the addition of headers, Exim already
4090 inserted X-ACL-Warn: at the start if there was no header name. However, it
4091 was not making this test for the second and subsequent header lines if
4092 there were newlines in the string. This meant that an invalid header could
4093 be inserted if Exim was badly configured.
4095 7. Allow an ACL "warn" statement to add header lines at the start or after all
4096 the Received: headers, as well as at the end.
4098 8. Added the rcpt_4xx retry error code.
4100 9. Added postmaster_mailfrom=xxx to callout verification option.
4102 10. Added mailfrom=xxxx to the callout verification option, for verify=
4105 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4106 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4108 12. Inserted some casts to stop certain compilers warning when using pointer
4109 differences as field lengths or precisions in printf-type calls (mostly
4110 affecting debugging statements).
4112 13. Added optional readline() support for -be (dynamically loaded).
4114 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4115 same clock tick as a message's arrival, so that its received time was the
4116 same as the "first fail" time on the retry record, and that message
4117 remained on the queue past the ultimate address timeout, every queue runner
4118 would try a delivery (because it was past the ultimate address timeout) but
4119 after another failure, the ultimate address timeout, which should have then
4120 bounced the address, did not kick in. This was a "< instead of <=" error;
4121 in most cases the first failure would have been in the next clock tick
4122 after the received time, and all would be well.
4124 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4125 being recognized when the domain list was tested by the match_domain
4126 condition in an expansion string.
4128 16. Added the ${str2b64: operator.
4130 17. Exim was always calling setrlimit() to set a large limit for the number of
4131 processes, without checking whether the existing limit was already
4132 adequate. (It did check for the limit on file descriptors.) Furthermore,
4133 errors from getrlimit() and setrlimit() were being ignored. Now they are
4134 logged to the main and panic logs, but Exim does carry on, to try to do its
4135 job under whatever limits there are.
4137 18. Imported PCRE 5.0.
4139 19. Trivial typo in log message " temporarily refused connection" (the leading
4142 20. If the log selector return_path_on_delivery was set and an address was
4143 redirected to /dev/null, the delivery process crashed because it assumed
4144 that a return path would always be set for a "successful" delivery. In this
4145 case, the whole delivery is bypassed as an optimization, and therefore no
4148 21. Internal re-arrangement: the function for sending a challenge and reading
4149 a response while authentication was assuming a zero-terminated challenge
4150 string. It's now changed to take a pointer and a length, to allow for
4151 binary data in such strings.
4153 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4155 23. Exim was not respecting finduser_retries when seeking the login of the
4156 uid under which it was called; it was always trying 10 times. (The default
4157 setting of finduser_retries is zero.) Also, it was sleeping after the final
4158 failure, which is pointless.
4160 24. Implemented tls_on_connect_ports.
4162 25. Implemented acl_smtp_predata.
4164 26. If the domain in control=submission is set empty, Exim assumes that the
4165 authenticated id is a complete email address when it generates From: or
4166 Sender: header lines.
4168 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4169 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4170 chown and chgrp in /bin and hostname in /usr/bin.
4172 28. Exim was keeping the "process log" file open after each use, just as it
4173 does for the main log. This opens the possibility of it remaining open for
4174 long periods when the USR1 signal hits a daemon. Occasional processlog
4175 errors were reported, that could have been caused by this. Anyway, it seems
4176 much more sensible not to leave this file open at all, so that is what now
4179 29. The long-running daemon process does not normally write to the log once it
4180 has entered its main loop, and it closes the log before doing so. This is
4181 so that log files can straightforwardly be renamed and moved. However,
4182 there are a couple of unusual error situations where the daemon does write
4183 log entries, and I had neglected to close the log afterwards.
4185 30. The text of an SMTP error response that was received during a remote
4186 delivery was being truncated at 512 bytes. This is too short for some of
4187 the long messages that one sometimes sees. I've increased the limit to
4190 31. It is now possible to make retry rules that apply only when a message has a
4191 specific sender, in particular, an empty sender.
4193 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4194 it possible to be selective about when SMTP synchronization is enforced.
4196 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4198 32. Implemented hosts_connection_nolog.
4200 33. Added an ACL for QUIT.
4202 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4205 35. Added mailbox_size and mailbox_filecount to appendfile.
4207 36. Added control = no_multiline_responses to ACLs.
4209 37. There was a bug in the logic of the code that waits for the clock to tick
4210 in the case where the clock went backwards by a substantial amount such
4211 that the microsecond fraction of "now" was more than the microsecond
4212 fraction of "then" (but the whole seconds number was less).
4214 38. Added support for the libradius Radius client library this is found on
4215 FreeBSD (previously only the radiusclient library was supported).
4221 1. When certain lookups returned multiple values in the form name=value, the
4222 quoting of the values was not always being done properly. Specifically:
4223 (a) If the value started with a double quote, but contained no whitespace,
4225 (b) If the value contained whitespace other than a space character (i.e.
4226 tabs or newlines or carriage returns) it was not quoted.
4227 This fix has been applied to the mysql and pgsql lookups by writing a
4228 separate quoting function and calling it from the lookup code. The fix
4229 should probably also be applied to nisplus, ibase and oracle lookups, but
4230 since I cannot test any of those, I have not disturbed their existing code.
4232 2. A hit in the callout cache for a specific address caused a log line with no
4233 reason for rejecting RCPT. Now it says "Previous (cached) callout
4234 verification failure".
4236 3. There was an off-by-one bug in the queryprogram router. An over-long
4237 return line was truncated at 256 instead of 255 characters, thereby
4238 overflowing its buffer with the terminating zero. As well as fixing this, I
4239 have increased the buffer size to 1024 (and made a note to document this).
4241 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4242 when Exim is waiting for an SMTP response from a remote server, Exim
4243 restarts its select() call on the socket, thereby resetting its timeout.
4244 This is not a problem when such interrupts are rare. Somebody set up a cron
4245 job to run exiwhat every 2 minutes, which is less than the normal select()
4246 timeout (5 or 10 minutes). This meant that the select() timeout never
4247 kicked in because it was always reset. I have fixed this by comparing the
4248 time when an interrupt arrives with the time at the start of the first call
4249 to select(). If more time than the timeout has elapsed, the interrupt is
4250 treated as a timeout.
4252 5. Some internal re-factoring in preparation for the addition of Sieve
4253 extensions (by MH). In particular, the "personal" test is moved to a
4254 separate function, and given an option for scanning Cc: and Bcc: (which is
4255 not set for Exim filters).
4257 6. When Exim created an email address using the login of the caller as the
4258 local part (e.g. when creating a From: or Sender: header line), it was not
4259 quoting the local part when it contained special characters such as @.
4261 7. Installed new OpenBSD configuration files.
4263 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4264 try to make them clearer.
4266 9. Callout options, other than the timeout value, were being ignored when
4267 verifying sender addresses in header lines. For example, when using
4269 verify = header_sender/callout=no_cache
4271 the cache was (incorrectly) being used.
4273 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4274 only the Cygwin environment.
4276 11. When return_path_on_delivery was set as a log selector, if different remote
4277 addresses in the same message used different return paths and parallel
4278 remote delivery occurred, the wrong values would sometimes be logged.
4279 (Whenever a remote delivery process finished, the return path value from
4280 the most recently started remote delivery process was logged.)
4282 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4283 lines when AUTH and/or TLS are in use. This is the "received protocol"
4284 field. Exim used to use "asmtp" for authenticated SMTP, without any
4285 indication (in the protocol name) for TLS use. Now it follows the RFC and
4286 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4287 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4288 names appear in log lines as well as in Received: header lines.
4290 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4291 extensions, and comparison tests, and to fix some bugs.
4293 14. Changes to the "personal" filter test:
4295 (1) The test was buggy in that it was just doing the equivalent of
4296 "contains" tests on header lines. For example, if a user's address was
4297 anne@some.where, the "personal" test would incorrectly be true for
4299 To: susanne@some.where
4301 This test is now done by extracting each address from the header in turn,
4302 and checking the entire address. Other tests that are part of "personal"
4303 are now done using regular expressions (for example, to check local parts
4304 of addresses in From: header lines).
4306 (2) The list of non-personal local parts in From: addresses has been
4307 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4308 taken from the Sieve specification recommendations.
4310 (3) If the message contains any header line starting with "List-" it is
4311 treated as non-personal.
4313 (4) The test for "circular" in the Subject: header line has been removed
4314 because it now seems ill-conceived.
4316 15. Minor typos in src/EDITME comments corrected.
4318 16. Installed latest exipick from John Jetmore.
4320 17. If headers_add on a router specified a text string that was too long for
4321 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4322 of string_sprintf() is now avoided.
4324 18. $message_body_size was not set (it was always zero) when running the DATA
4325 ACL and the local_scan() function.
4327 19. For the "mail" command in an Exim filter, no default was being set for
4328 the once_repeat time, causing a random time value to be used if "once" was
4329 specified. (If the value happened to be <= 0, no repeat happened.) The
4330 default is now 0s, meaning "never repeat". The "vacation" command was OK
4331 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4332 (I found it when inspecting the code).
4334 20. There is now an overall timeout for performing a callout verification. It
4335 defaults to 4 times the callout timeout, which applies to individual SMTP
4336 commands during the callout. The overall timeout applies when there is more
4337 than one host that can be tried. The timeout is checked before trying the
4338 next host. This prevents very long delays if there are a large number of
4339 hosts and all are timing out (e.g. when the network connections are timing
4340 out). The value of the overall timeout can be changed by specifying an
4341 additional sub-option for "callout", called "maxwait". For example:
4343 verify = sender/callout=5s,maxwait=20s
4345 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4346 to the end before writing, but this should make it even safer).
4348 22. Exim was forgetting that it had advertised PIPELINING for the second and
4349 subsequent messages on an SMTP connection. It was also not resetting its
4350 memory on STARTTLS and an internal HELO.
4352 23. When Exim logs an SMTP synchronization error within a session, it now
4353 records whether PIPELINING has been advertised or not.
4355 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4356 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4357 rather than long int.
4359 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4361 26. Added the never_mail option to autoreply.
4367 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4368 crash if the getsockname() call failed; this can happen if a connection is
4369 closed very soon after it is established. The problem was simply in the
4370 order in which certain operations were done, causing Exim to try to write
4371 to the SMTP stream before it had set up the file descriptor. The bug has
4372 been fixed by making things happen in the correct order.
4378 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4379 before the connection was closed, thus losing the rejection response.
4381 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4382 some early Solaris releases, but causes trouble in current releases where
4383 socklen_t is defined.
4385 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4388 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4391 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4392 that releasing the top of it at the end releases what was used for sub-
4393 expansions (unless the block got too big). However, discard this block if
4394 the first thing is a variable or header, so that we can use its block when
4395 it is dynamic (useful for very large $message_headers, for example).
4397 6. Lookups now cache *every* query, not just the most recent. A new, separate
4398 store pool is used for this. It can be recovered when all lookup caches are
4399 flushed. Lookups now release memory at the end of their result strings.
4400 This has involved some general refactoring of the lookup sources.
4402 7. Some code has been added to the store_xxx() functions to reduce the amount
4403 of flapping under certain conditions.
4405 8. log_incoming_interface used to affect only the <= reception log lines. Now
4406 it causes the local interface and port to be added to several more SMTP log
4407 lines, for example "SMTP connection from", and rejection lines.
4409 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4411 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4413 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4414 could be overwritten at the end of the current message (or the start of a
4415 new message if it was set in a HELO ACL). The value is now preserved for
4416 the duration of the SMTP connection.
4418 12. If a transport had a headers_rewrite setting, and a matching header line
4419 contained an unqualified address, that address was qualified, even if it
4420 did not match any rewriting rules. The underlying bug was that the values
4421 of the flags that permit the existence of unqualified sender and recipient
4422 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4423 non-local messages, and by -bnq for local messages) were not being
4424 preserved with the message after it was received.
4426 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4427 "next input=" as part of the text comprising the host identity instead of
4428 the correct text. The code was using the same buffer for two different
4429 strings. However, depending on which order the printing function evaluated
4430 its arguments, the bug did not always show up. Under Linux, for example, my
4431 test suite worked just fine.
4433 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4434 doesn't work with some older versions of Perl. It has been changed to "my",
4435 which in any case is probably the better facility to use.
4437 15. A really picky compiler found some instances of statements for creating
4438 error messages that either had too many or two few arguments for the format
4441 16. The size of the buffer for calls to the DNS resolver has been increased
4442 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4443 for addresses that have a lot of PTR records. This alleviates a problem; it
4444 does not fully solve it.
4446 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4447 buffer now truncates the list and logs the incident, which is the same
4448 action as happens when Exim is looking up a host name and its aliases.
4449 Previously in this situation something unpredictable would happen;
4450 sometimes it was "internal error: store_reset failed".
4452 18. If a server dropped the connection unexpectedly when an Exim client was
4453 using GnuTLS and trying to read a response, the client delivery process
4454 crashed while trying to generate an error log message.
4456 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4457 string, for example:
4459 warn message = H1: something\nH2: something
4461 the text was added as a single header line from Exim's point of view
4462 though it ended up OK in the delivered message. However, searching for the
4463 second and subsequent header lines using $h_h2: did not work. This has been
4464 fixed. Similarly, if a system filter added multiple headers in this way,
4465 the routers could not see them.
4467 20. Expanded the error message when iplsearch is called with an invalid key to
4468 suggest using net-iplsearch in a host list.
4470 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4471 ACLs are no longer actually imposed (and a message to that effect is
4474 22. If a "gecos" field in a passwd entry contained escaped characters, in
4475 particular, if it contained a \" sequence, Exim got it wrong when building
4476 a From: or a Sender: header from that name. A second bug also caused
4477 incorrect handling when an unquoted " was present following a character
4478 that needed quoting.
4480 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4481 was not being matched caselessly.
4483 24. Arranged for all hyphens in the exim.8 source to be escaped with
4486 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4487 properly. Recipient callout cache records were still being keyed to include
4488 the sender, even when use_sender was set false. This led to far more
4489 callouts that were necessary. The sender is no longer included in the key
4490 when use_sender is false.
4492 26. Added "control = submission" modifier to ACLs.
4494 27. Added the ${base62d: operator to decode base 62 numbers.
4496 28. dnsdb lookups can now access SRV records.
4498 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4499 the configuration file.
4501 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4502 (-v) mode. This makes the output for a verbose queue run more intelligible.
4504 31. Added a use_postmaster feature to recipient callouts.
4506 32. Added the $body_zerocount variable, containing the number of binary zero
4507 bytes in the message body.
4509 33. The time of last modification of the "new" subdirectory is now used as the
4510 "mailbox time last read" when there is a quota error for a maildir
4513 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4515 35. Added +ignore_unknown as a special item in host lists.
4517 36. Code for decoding IPv6 addresses in host lists is now included, even if
4518 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4519 address was recognized as an IP address, but was then not correctly decoded
4520 into binary, causing unexpected and incorrect effects when compared with
4527 1. Very minor rewording of debugging text in manualroute to say "list of
4528 hosts" instead of "hostlist".
4530 2. If verify=header_syntax was set, and a header line with an unqualified
4531 address (no domain) and a large number of spaces between the end of the
4532 name and the colon was received, the reception process suffered a buffer
4533 overflow, and (when I tested it) crashed. This was caused by some obsolete
4534 code that should have been removed. The fix is to remove it!
4536 3. When running in the test harness, delay a bit after writing a bounce
4537 message to get a bit more predictability in the log output.
4539 4. Added a call to search_tidyup() just before forking a reception process. In
4540 theory, someone could use a lookup in the expansion of smtp_accept_max_
4541 per_host which, without the tidyup, could leave open a database connection.
4543 5. Added the variables $recipient_data and $sender_data which get set from a
4544 lookup success in an ACL "recipients" or "senders" condition, or a router
4545 "senders" option, similar to $domain_data and $local_part_data.
4547 6. Moved the writing of debug_print from before to after the "senders" test
4550 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4551 problems for message scanning, either using a data ACL, or using
4552 local_scan() because the Received: header was not generated till after they
4553 were called (in order to set the time as the time of reception completion).
4554 I have revised the way this works. The header is now generated after the
4555 body is received, but before the ACL or local_scan() are called. After they
4556 are run, the timestamp in the header is updated.
4562 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4563 before starting a queue runner without re-exec. This happened only when
4564 deliver_drop_privilege was set or when the Exim user was set to root. The
4565 effect of the bug was that timeouts during subsequent deliveries caused
4566 crashes instead of being properly handled. The handler is now left at its
4567 default (and expected) setting.
4569 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4570 message, again when deliver_drop_privilege is set or Exim is run as root.
4571 The bug described in (1) was not present in this case, but the tidying up
4572 of the other signals was missing. I have made the two cases consistent.
4574 3. The ignore_target_hosts setting on a manualroute router was being ignored
4575 for hosts that were looked up using the /MX notation.
4577 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4580 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4581 operated on the sender address. After changing the $sender_address to <>
4582 for the sender address verify, Exim was re-instated it as the original
4583 (before rewriting) address, but remembering that it had rewritten it, so it
4584 wasn't rewriting it again. This bug also had the effect of breaking the
4585 sender address verification caching when the sender address was rewritten.
4587 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4588 This has been changed so that if the ip literal address matches
4589 ignore_target_hosts, the router declines.
4591 7. Added expansion conditions match_domain, match_address, and match_local_
4592 part (NOT match_host).
4594 8. The placeholder for the Received: header didn't have a length field set.
4596 9. Added code to Exim itself and to exim_lock to test for a specific race
4597 condition that could lead to file corruption when using MBX delivery. The
4598 issue is with the lockfile that is created in /tmp. If this file is removed
4599 after a process has opened it but before that process has acquired a lock,
4600 there is the potential for a second process to recreate the file and also
4601 acquire a lock. This could lead to two Exim processes writing to the file
4602 at the same time. The added code performs the same test as UW imapd; it
4603 checks after acquiring the lock that its file descriptor still refers to
4604 the same named file.
4606 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4607 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4610 11. Added the smtp_active_hostname option. If used, this will typically be made
4611 to depend on the incoming interface address. Because $interface_address is
4612 not set up until the daemon has forked a reception process, error responses
4613 that can happen earlier (such as "too many connections") no longer contain
4616 12. If an expansion in a condition on a "warn" statement fails because a lookup
4617 defers, the "warn" statement is abandoned, and the next ACL statement is
4618 processed. Previously this caused the whole ACL to be aborted.
4620 13. Added the iplsearch lookup type.
4622 14. Added ident_timeout as a log selector.
4624 15. Added tls_certificate_verified as a log selector.
4626 16. Added a global option tls_require_ciphers (compare the smtp transport
4627 option of the same name). This controls incoming TLS connections.
4629 17. I finally figured out how to make tls_require_ciphers do a similar thing
4630 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4631 before starting the TLS session.
4633 18. Tabs are now shown as \t in -bP output.
4635 19. If the log selector return_path_on_delivery was set, Exim crashed when
4636 bouncing a message because it had too many Received: header lines.
4638 20. If two routers both had headers_remove settings, and the first one included
4639 a superfluous trailing colon, the final name in the first list and the
4640 first name in the second list were incorrectly joined into one item (with a
4641 colon in the middle).
4647 1. Added -C and -D options to the exinext utility, mainly to make it easier
4648 to include in the automated testing, but these could be helpful when
4649 multiple configurations are in use.
4651 2. The exinext utility was not formatting the output nicely when there was
4652 an alternate port involved in the retry record key, nor when there was a
4653 message id as well (for retries that were specific to a specific message
4654 and a specific host). It was also confused by IPv6 addresses, because of
4655 the additional colons they contain. I have fixed the IPv4 problem, and
4656 patched it up to do a reasonable job for IPv6.
4658 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4659 delivery, the log line now contains "pipelined" if PIPELINING was used.
4661 4. An SMTP transport process used to panic and die if the bind() call to set
4662 an explicit outgoing interface failed. This has been changed; it is now
4663 treated in the same way as a connect() failure.
4665 5. A reference to $sender_host_name in the part of a conditional expansion
4666 that was being skipped was still causing a DNS lookup. This no longer
4669 6. The def: expansion condition was not recognizing references to header lines
4670 that used bh_ and bheader_.
4672 7. Added the _cache feature to named lists.
4674 8. The code for checking quota_filecount in the appendfile transport was
4675 allowing one more file than it should have been.
4677 9. For compatibility with Sendmail, the command line option
4685 and sets the incoming protocol and host name (for trusted callers). The
4686 host name and its colon can be omitted when only the protocol is to be set.
4687 Note the Exim already has two private options, -pd and -ps, that refer to
4688 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4689 "s", but I don't think that's a major issue.
4691 10. A number of refactoring changes to the code, none of which should affect
4694 (a) The number of logging options was getting close to filling up the
4695 32-bit word that was used as a bit map. I have split them into two classes:
4696 those that are passed in the argument to log_write(), and those that are
4697 only ever tested independently outside of that function. These are now in
4698 separate 32-bit words, so there is plenty of room for expansion again.
4699 There is no change in the user interface or the logging behaviour.
4701 (b) When building, for example, log lines, the code previously used a
4702 macro that called string_cat() twice, in order to add two strings. This is
4703 not really sufficiently general. Furthermore, there was one instance where
4704 it was actually wrong because one of the argument was used twice, and in
4705 one call a function was used. (As it happened, calling the function twice
4706 did not affect the overall behaviour.) The macro has been replaced by a
4707 function that can join an arbitrary number of extra strings onto a growing
4710 (c) The code for expansion conditions now uses a table and a binary chop
4711 instead of a serial search (which was left over from when there were very
4712 few conditions). Also, it now recognizes conditions like "pam" even when
4713 the relevant support is not compiled in: a suitably worded error message is
4714 given if an attempt is made to use such a condition.
4716 11. Added ${time_interval:xxxxx}.
4718 12. A bug was causing one of the ddress fields not to be passed back correctly
4719 from remote delivery subprocesses. The field in question was not being
4720 subsequently used, so this caused to problems in practice.
4722 13. Added new log selectors queue_time and deliver_time.
4724 14. Might have fixed a bug in maildirsizefile handling that threw up
4725 "unexpected character" debug warnings, and recalculated the data
4726 unnecessarily. In any case, I expanded the warning message to give more
4729 15. Added the message "Restricted characters in address" to the statements in
4730 the default ACL that block characters like @ and % in local parts.
4732 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4733 Three changes have been made:
4735 (a) There was a serious bug; a negative response to MAIL caused the whole
4736 recipient domain to be cached as invalid, thereby blocking all messages
4737 to all local parts at the same domain, from all senders. This bug has
4738 been fixed. The domain is no longer cached after a negative response to
4739 MAIL if the sender used is not empty.
4741 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4744 (c) A new callout option, "use_sender" has been added for people who want
4745 the modified behaviour.
4751 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4754 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4755 indeed breaks things for older releases.
4757 3. Added additional logging to the case where there is a problem reading data
4758 from a filter that is running in a subprocess using a pipe, in order to
4759 try to track down a specific problem.
4761 4. Testing facility fudge: when running in the test harness and attempting
4762 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4763 getting "No route to host". Convert this to a timeout.
4765 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4768 6. Some OS don't have socklen_t but use size_t instead. This affects the
4769 fifth argument of getsockopt() amongst other things. This is now
4770 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4771 can be set for individual OS. I have set it for SunOS5, OSF1, and
4772 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4773 some earlier ones do not.
4775 7. Change 4.30/15 was not doing the test caselessly.
4777 8. The standard form for an IPv6 address literal was being rejected by address
4778 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4779 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4780 this, as well as the form without the "IPv6" on the front (but only when
4781 address literals are enabled, of course).
4783 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4785 10. Exim crashed if a message with an empty sender address specified by -f
4786 encountered a router with an errors_to setting. This could be provoked only
4787 by a command such as
4791 where an empty string was supplied; "<>" did not hit this bug.
4793 11. Installed PCRE release 4.5.
4795 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4796 remained set. It is now erased.
4798 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4799 times from message ids (which are base 36 rather than the normal 62).
4801 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4802 were being counted as actual protocol errors, and logged if the log
4803 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4804 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4805 and DATA following a set of rejected RCPTs do not count as protocol errors.
4806 In other words, Exim assumes they were pipelined, though this may not
4807 actually be the case. Of course, in all cases the client gets an
4808 appropriate error code.
4810 15. If a lookup fails in an ACL condition, a message about the failure may
4811 be available; it is used if testing the ACL cannot continue, because most
4812 such messages specify what the cause of the deferral is. However, some
4813 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4814 that caused an old message to be retained and used if a later statement
4815 caused a defer, replacing the real cause of the deferral.
4817 16. If an IP address had so many PTR records that the DNS lookup buffer
4818 was not large enough to hold them, Exim could crash while trying to process
4819 the truncated data. It now detects and logs this case.
4821 17. Further to 4.21/58, another change has been made: if (and only if) the
4822 first line of a message (the first header line) ends with CRLF, a bare LF
4823 in a subsequent header line has a space inserted after it, so as not to
4824 terminate the header.
4826 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4827 unnecessarily, used atoi() instead of strtol(), and didn't check the
4828 termination when getting file sizes from file names by regex.
4830 19. Completely re-implemented the support for maildirsize files, in the light
4831 of a number of problems with the previous contributed implementation
4832 (4.30/29). In particular:
4834 . If the quota is zero, the maildirsize file is maintained, but no quota is
4837 . If the maildir directory does not exist, it is created before any attempt
4838 to write a maildirsize file.
4840 . The quota value in the file is just a cache; if the quota is changed in
4841 the transport, the new value overrides.
4843 . A regular expression is available for excluding directories from the
4846 20. The autoreply transport checks the characters in options that define the
4847 message's headers; it allows continued headers, but it was checking with
4848 isspace() after an embedded newline instead of explicitly looking for a
4851 21. If all the "regular" hosts to which an address was routed had passed their
4852 expiry times, and had not reached their retry times, the address was
4853 bounced, even if fallback hosts were defined. Now Exim should go on to try
4856 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4857 equivalent code in the SMTP transport. Some hosts send humungous responses
4858 to HELO/EHLO, more than 1024 it seems.
4860 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4861 gives compiler warnings in some environments. I've now done it "properly",
4864 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4865 (because of problems with the built-in one) was declared to return uschar *
4866 instead of char *, causing compiler failure.
4868 25. Fixed a file descriptor leak when processing alias/forward files.
4870 26. Fixed a minor format string issue in dbfn.c.
4872 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4874 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4875 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4878 29. When a sender address is verified, it is cached, to save repeating the test
4879 when there is more than one recipient in a message. However, when the
4880 verification involves a callout, it is possible for different callout
4881 options to be set for different recipients. It is too complicated to keep
4882 track of this in the cache, so now Exim always runs a verification when a
4883 callout is required, relying on the callout cache for the optimization.
4884 The overhead is duplication of the address routing, but this should not be
4887 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4888 to be verified with callout=postmaster, and the main callout worked but the
4889 postmaster check failed, the verification correctly failed. However, if a
4890 subsequent RCPT command asked for sender verification *without* the
4891 postmaster check, incorrect caching caused this verification also to fail,
4894 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4895 it was not caching the DNS options (qualify_single, search_parents) that
4896 were used when the lookup failed. A subsequent lookup with different
4897 options therefore always gave the same answer, though there were cases
4898 where it should not have. (Example: a "domains = !$mx_any" option on a
4899 dnslookup router: the "domains" option is always processed without any
4900 widening, but the router might have qualify_single set.) Now Exim uses the
4901 cached value only when the same options are set.
4903 32. Added John Jetmore's "exipick" utility to the distribution.
4905 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4906 than a timeout (e.g. a certificate is required, and is not provided), an
4907 Exim server now closes the connection immediately. Previously it waited for
4908 the client to close - but if the client is SSL, it seems that they each
4909 wait for each other, leading to a delay before one of them times out.
4911 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4912 maintained 0.8.x compatibility because I don't think many are using it, and
4913 it is clearly obsolete.
4915 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4918 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4919 client certificate was expired. A simple patch fixes this, though I don't
4920 understand the full logic of why the verify callback is called multiple
4923 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4924 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4925 which causes problems with some clients (such as the Certicom SSL Plus
4926 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4927 disables the coutermeasure allowing Eudora to connect."
4929 38. Exim was not checking that a write() to a log file succeeded. This could
4930 lead to Bad Things if a log got too big, in particular if it hit a file
4931 size limit. Exim now panics and dies if it cannot write to a log file, just
4932 as it does if it cannot open a log file.
4934 39. Modified OS/Makefile-Linux so that it now contains
4936 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4938 The two -D definitions ensure that Exim is compiled with large file
4939 support, which makes it possible to handle log files that are bigger than
4942 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4943 instance) a domain was checked against a named list that involved a lookup,
4944 causing $domain_data to be set, then another domain was checked against the
4945 same list, then the first domain was re-checked, the value of $domain_data
4946 after the final check could be wrong. In particular, if the second check
4947 failed, it could be set empty. This bug probably also applied to
4950 41. The strip_trailing_dot option was not being applied to the address given
4951 with the -f command-line option.
4953 42. The code for reading a message's header from the spool was incrementing
4954 $received_count, but never initializing it. This meant that the value was
4955 incorrect (doubled) while delivering a message in the same process in which
4956 it was received. In the most common configuration of Exim, this never
4957 happens - a fresh exec is done - but it can happen when
4958 deliver_drop_privilege is set.
4960 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4961 it now includes up to 150 characters of the unexpected data in the log
4964 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4965 and building data strings. The size of both of these buffers was 10 000
4966 bytes - far larger than anybody would *ever* want, thought I. Needless to
4967 say, somebody hit the limit. I have increased the maximum line length to
4968 20 000 and the maximum data length of concatenated lines to 100 000. I have
4969 also fixed two bugs, because there was no checking on these buffers. Tsk,
4970 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4971 buffer is too small.
4973 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4974 lsearch lookups. Now it does.
4976 46. When parsing a route_list item in a manualroute router, a fixed-length
4977 buffer was used for the list of hosts. I made this 1024 bytes long,
4978 thinking that nobody would ever have a list of hosts that long. Wrong.
4979 Somebody had a whole pile of complicated expansion conditions, and the
4980 string was silently truncated, leading to an expansion error. It turns out
4981 that it is easier to change to an unlimited length (owing to other changes
4982 that have happened since this code was originally written) than to build
4983 structure for giving a limitation error. The length of the item that
4984 expands into the list of hosts is now unlimited.
4986 47. The lsearch lookup could not handle data where the length of text line was
4987 more than 4095 characters. Such lines were truncated, leading to shortened
4988 data being returned. It should now handle lines of any length.
4990 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4991 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4994 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4995 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4996 no longer happens in comment lines. A list of replacements is now placed
4997 at the head of all of the source files, except those whose only change is
4998 to replace PERL_COMMAND in the very first #! line.
5000 50. Replaced the slow insertion sort in queue.c, for sorting the list of
5001 messages on the queue, with a bottom-up merge sort, using code contributed
5002 by Michael Haardt. This should make operations like -bp somewhat faster on
5003 large queues. It won't affect queue runners, except when queue_run_in_order
5006 51. Installed eximstats 1.31 in the distribution.
5008 52. Added support for SRV lookups to the dnslookup router.
5010 53. If an ACL referred to $message_body or $message_body_end, the value was not
5011 reset for any messages that followed in the same SMTP session.
5013 54. The store-handling optimization for building very long strings was not
5014 differentiating between the different store pools. I don't think this
5015 actually made any difference in practice, but I've tidied it.
5017 55. While running the routers to verify a sender address, $sender_address
5018 was still set to the sender address. This is wrong, because when routing to
5019 send a bounce to the sender, it would be empty. Therefore, I have changed
5020 it so that, while verifying a sender address, $sender_address is set to <>.
5021 (There is no change to what happens when verifying a recipient address.)
5023 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
5024 target A or AAAA records (if not already returned) without resetting the
5025 qualify_single or search_parents options of the DNS resolver. These are
5026 inappropriate in this case because the targets of MX and SRV records must
5027 be FQDNs. A broken DNS record could cause trouble if it happened to have a
5028 target that, when qualified, matched something in the local domain. These
5029 two options are now turned off when doing these lookups.
5031 57. It seems that at least some releases of Reiserfs (which does not have the
5032 concept of a fixed number of inodes) returns zero and not -1 for the
5033 number of available inodes. This interacted badly with check_spool_inodes,
5034 which assumed that -1 was the "no such thing" setting. What I have done is
5035 to check that the total number of inodes is greater than zero before doing
5036 the test of how many are available.
5038 58. When a "warn" ACL statement has a log_message modifier, the message is
5039 remembered, and not repeated. This is to avoid a lot of repetition when a
5040 message has many recipients that cause the same warning to be written.
5041 Howewer, Exim was preserving the list of already written lines for an
5042 entire SMTP session, which doesn't seem right. The memory is now reset if a
5043 new message is started.
5045 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5046 debugging output unless log_rewrite was also set.
5048 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5049 of (int)(handle) when we know that handle contains (void *)(-1).
5051 61. The Exim daemon panic-logs an error return when it closes the incoming
5052 connection. However "connection reset by peer" seems to be common, and
5053 isn't really an error worthy of noting specially, so that particular error
5056 62. When Exim is trying to find all the local interfaces, it used to panic and
5057 die if the ioctl to get the interface flags failed. However, it seems that
5058 on at least one OS (Solaris 9) it is possible to have an interface that is
5059 included in the list of interfaces, but for which you get a failure error
5060 for this call. This happens when the interface is not "plumbed" into a
5061 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5062 failure of the "get flags" call assumes that the interface is down.
5064 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5065 makes life easier for people who are doing arithmetic on fields extracted
5066 from dates, where you often get leading zeros that should not be
5067 interpreted as octal.
5069 64. Added qualify_domain to the redirect router, to override the global
5072 65. If a pathologically long header line contained very many addresses (the
5073 report of this problem mentioned 10 000) and each of them was rewritten,
5074 Exim could use up a very large amount of memory. (It kept on making new
5075 copies of the header line as it rewrote, and never released the old ones.)
5076 At the expense of a bit more processing, the header rewriting function has
5077 been changed so that it no longer eats memory in this way.
5079 66. The generation of the Received: header has been moved from the time that a
5080 message starts to be received, to the time that it finishes. The timestamp
5081 in the Received: header should now be very close to that of the <= log
5082 line. There are two side-effects of this change:
5084 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5085 logged header lines no longer include the local Received: line, because
5086 it has not yet been created. The same applies to a copy of the message
5087 that is returned to a non-SMTP sender when a message is rejected.
5089 (b) When a filter file is tested using -bf, no additional Received: header
5090 is added to the test message. After some thought, I decided that this
5093 This change does not affect the value of $received_for. It is still set
5094 after address rewriting, but before local_scan() is called.
5096 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5098 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5099 gave an unhelpful panic error message, and a defer error. I have managed to
5100 change this behaviour so that it now rejects any supplied certificate,
5101 which seems right, as the list of acceptable certificates is empty.
5103 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5104 gave an unhelpful defer error. I have not managed to make this reject any
5105 supplied certificates, but the error message it gives is "no certificate
5106 supplied", which is not helpful.
5108 70. exigrep's output now also includes lines that are not associated with any
5109 message, but which match the given pattern. Implemented by a patch from
5110 Martin Sluka, which also tidied up the Perl a bit.
5112 71. Recipient callout verification, like sender verification, was using <> in
5113 the MAIL FROM command. This isn't really the right thing, since the actual
5114 sender may affect whether the remote host accepts the recipient or not. I
5115 have changed it to use the actual sender in the callout; this means that
5116 the cache record is now keyed on a recipient/sender pair, not just the
5117 recipient address. There doesn't seem to be a real danger of callout loops,
5118 since a callout by the remote host to check the sender would use <>.
5119 [SEE ABOVE: changed after hitting problems.]
5121 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5122 temporary errors. However, in the case of such a code being given after
5123 the end of a data transmission (i.e. after ".") Exim was failing to write
5124 a retry record for the message. (Yes, there was some broken host that was
5125 actually sending 8xx at this point.)
5127 73. An unknown lookup type in a host list could cause Exim to panic-die when
5128 the list was checked. (An example that provoked this was putting <; in the
5129 middle of a list instead of at the start.) If this happened during a DATA
5130 ACL check, a -D file could be left lying around. This kind of configuration
5131 error no longer causes Exim to die; instead it causes a defer errror. The
5132 incident is still logged to the main and panic logs.
5134 74. Buglet left over from Exim 3 conversion. The message "too many messages
5135 in one connection" was written to the rejectlog but not the mainlog, except
5136 when address rewriting (yes!) was being logged.
5138 75. Added write_rejectlog option.
5140 76. When a system filter was run not as root (that is, when system_filter_user
5141 was set), the values of the $n variables were not being returned to the
5142 main process; thus, they were not subsequently available in the $sn
5145 77. Added +return_path_on_delivery log selector.
5147 78. A connection timeout was being treated differently from recipients deferred
5148 when testing hosts_max_try with a message that was older than the host's
5149 retry timeout. (The host should not be counted, thus allowing all hosts to
5150 be tried at least once before bouncing.) This may have been the cause of an
5151 occasionally reported bug whereby a message would remain on the queue
5152 longer than the retry timeout, but would be bounced if a delivery was
5153 forced. I say "may" because I never totally pinned down the problem;
5154 setting up timeout/retry tests is difficult. See also the next item.
5156 79. The ultimate address timeout was not being applied to errors that involved
5157 a combination of host plus message (for example, a timeout on a MAIL
5158 command). When an address resolved to a number of possible hosts, and they
5159 were not all tried for each delivery (e.g. because of hosts_max_try), a
5160 message could remain on the queue longer than the retry timeout.
5162 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5165 81. Fixed an obscure SMTP outgoing bug which required at least the following
5166 conditions: (a) there was another message waiting for the same server;
5167 (b) the server returned 5xx to all RCPT commands in the first message so
5168 that the message was not completed; (c) the server dropped the connection
5169 or gave a negative response to the RSET that Exim sends to abort the
5170 transaction. The observed case was a dropped connection after DATA that had
5171 been sent in pipelining mode. That is, the server had advertised PIPELINING
5172 but was not implementing it correctly. The effect of the bug was incorrect
5173 behaviour, such as trying another host, and this could lead to a crash.
5179 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5180 and daemon.c were passed as pointers to ints; they should have been
5181 pointers to socklen_t variables (which are typically unsigned ints).
5183 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5186 3. Fixed a really odd bug that affected only the testing scheme; patching a
5187 certain fixed string in the binary changed the value of another string that
5188 happened to be identical to the end of the original first string.
5190 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5191 name", it returns that address as the IP address. On some operating
5192 systems (e.g. Solaris), it also passes back the IP address string as the
5193 "host name". However, on others (e.g. Linux), it passes back an empty
5194 string. Exim wasn't checking for this, and was changing the host name to an
5195 empty string, assuming it had been canonicized.
5197 5. Although rare, it is permitted to have more than one PTR record for a given
5198 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5199 all the names associated with an address, because they do in Solaris.
5200 However, it seems that they do not in Linux for data that comes from the
5201 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5202 I found this out when I moved to a new Linux workstation and tried to run
5203 the Exim test suite.
5205 To get round this problem I have changed the code so that it now does its
5206 own call to the DNS to look up PTR records when searching for a host name.
5207 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5208 addresses that are only in /etc/hosts are still found.
5210 This behaviour is, however, controlled by an option called host_lookup_
5211 order, which defaults to "bydns:byaddr". If people want to use the other
5212 order, or indeed, just use one or the other means of lookup, they can
5213 specify it in this variable.
5215 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5216 some operating systems, this comes back from gethostbyaddr() as an empty
5217 string, and this is what Exim used to test for. However, it seems that in
5218 other systems, "." is yielded. Exim now tests for this case too.
5220 7. The values of check_spool_space and check_log_space are now held internally
5221 as a number of kilobytes instead of an absolute number of bytes. If a
5222 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5223 kilobyte. This means that much larger values can be stored.
5225 8. Exim monitor: an attempt to get the action menu when not actually pointing
5226 at a message produces an empty menu entitled "No message selected". This
5227 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5228 no entries in it ("Shell widget menu has zero width and/or height"). So I
5229 have added a single, blank menu entry in this case.
5231 9. Added ${quote_local_part.
5233 10. MIME decoding is now applied to the contents of Subject: header lines when
5236 11. Now that a reference to $sender_host_address automatically causes a reverse
5237 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5238 host lookup before query-style lookups in lists that might use this
5239 variable. This has therefore been abolished, and the "net-" prefix is no
5240 longer necessary for query-style lookups.
5242 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5243 have been a typo for LFLAGS, so it has been changed.
5245 13. The install script calls Exim with "-C /dev/null" in order to find the
5246 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5247 to be output. Howeve, since Exim outputs its version number before the
5248 error, it didn't break the script. It just looked ugly. I fixed this by
5249 always allowing "-C /dev/null" if the caller is root.
5251 14. Ignore overlarge ACL variable number when reading spool file - insurance
5252 against a later release with more variables having written the file.
5254 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5255 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5256 this, as well as the form without the "IPv6" on the front.
5258 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5259 OS/Makefile-Darwin file.
5261 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5262 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5264 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5265 variable. However, some OS use ints for inodes. I've added cast to long int
5266 to get rid of the compiler warning.
5268 19. I had forgotten to lock out "/../" in configuration file names when
5269 ALT_CONFIG_PREFIX was set.
5271 20. Routers used for verification do not need to specify transports. However,
5272 if such a router generated a host list, and callout was configured, Exim
5273 crashed, because it could not find a port number from the (non-existent)
5274 transport. It now assumes port 25 in this circumstance.
5276 21. Added the -t option to exigrep.
5278 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5279 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5280 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5281 them is set, LOOKUP_LSEARCH is forced.
5283 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5285 24. Added sender and host information to the "rejected by local_scan()" log
5286 line; previously there was no indication of these.
5288 25. Added .include_if_exists.
5290 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5291 Linux. It turns out that not all file systems support this. Apparently some
5292 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5293 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5294 not supported on the file descriptor, is now ignored when Exim is trying to
5295 sync a directory. This applies only to Linux.
5297 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5299 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5300 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5303 29. Added support for maildirsize files from supplied patch (modified a bit).
5305 30. The use of :fail: followed by an empty string could lead Exim to respond to
5306 sender verification failures with (e.g.):
5308 550 Verification failed for <xxx>
5309 550 Sender verify failed
5311 where the first response line was missing the '-' that indicates it is not
5312 the final line of the response.
5314 31. The loop for finding the name of the user that called Exim had a hardwired
5315 limit of 10; it now uses the value of finduser_retries, which is used for
5316 all other user lookups.
5318 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5321 33. Exim was neglecting to zero errno before one call of strtol() when
5322 expanding a string and expecting an integer value. On some systems this
5323 resulted in spurious "integer overflow" errors. Also, it was casting the
5324 result into an int without checking.
5326 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5327 did not work. The code looks as if it has *never* worked, though it appears
5328 to have been documented since at least releast 1.62. I have made it work.
5330 35. The "timeout_DNS" error in retry rules, also documented since at least
5331 1.62, also never worked. As it isn't clear exactly what this means, and
5332 clearly it isn't a major issue, I have abolished the feature by treating it
5333 as "timeout", and writing a warning to the main and panic logs.
5335 36. The display of retry rules for -brt wasn't always showing the error code
5338 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5339 timeout_connect_A, timeout_connect_MX.
5341 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5342 to the empty sender.
5344 39. The daemon was not analysing the content of -oX till after it had closed
5345 stderr and disconnected from the controlling terminal. This meant that any
5346 syntax errors were only noted on the panic log, and the return code from
5347 the command was 0. By re-arranging the code a little, I've made the
5348 decoding happen first, so such errors now appear on stderr, and the return
5349 code is 1. However, the actual setting up of the sockets still happens in
5350 the disconnected process, so errors there are still only recorded on the
5353 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5354 connections (as happens on some IP stacks) was logged at start up time as
5355 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5356 it from "IPv6 and IPv4", which means that two separate sockets are being
5359 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5360 says whether AF_INET or AF_INET6 was passed as an argument.
5362 42. Exiwhat output was messed up when time zones were included in log
5365 43. Exiwhat now gives more information about the daemon's listening ports,
5366 and whether -tls-on-connect was used.
5368 44. The "port" option of the smtp transport is now expanded.
5370 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5371 silently ignored. Now an error message is written to the main and panic
5374 46. There's a new ACL modifier called "logwrite" which writes to a log file
5375 as soon as it is encountered.
5377 47. Added $local_user_uid and $local_user_gid at routing time.
5379 48. Exim crashed when trying to verify a sender address that was being
5382 49. Exim was recognizing only a space character after ".include". It now also
5383 recognizes a tab character.
5385 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5386 extracting the relevant information from the specification. The man page no
5387 longer contains scrambled data for the -d option, and I've added a section
5388 at the front about calling Exim under different names.
5390 51. Added "extra_headers" argument to the "mail" command in filter files.
5392 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5395 53. Installed eximstats 1.29.
5397 54. Added transport_filter_timeout as a generic transport option.
5399 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5400 Cc: header lines. This was required by RFC 822, but it not required by RFC
5403 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5404 incoming messages that did not have them. Now it does so only if the
5405 message originates locally, that is, if there is no associated remote host
5406 address. When Resent- header lines are present, this applies to the Resent-
5407 lines rather than the non-Resent- lines.
5409 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5410 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5412 58. Messages for configuration errors now include the name of the main
5413 configuration file - useful now that there may be more than one file in a
5414 list (.included file names were always shown).
5416 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5417 for those rare installations that do not start the daemon as root or run it
5418 setuid root. I've cut out the call to initgroups() if the daemon is not
5421 60. The Exim user and group can now be bound into the binary as text strings
5422 that are looked up at the start of Exim's processing.
5424 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5426 62. Added $mailstore_basename variable.
5428 63. Installed patch to sieve.c from Michael Haardt.
5430 64. When Exim failed to open the panic log after failing to open the main log,
5431 the original message it was trying to log was written to stderr and debug
5432 output, but if they were not available (the usual case in production), it
5433 was lost. Now it is written to syslog before the two lines that record the
5434 failures to open the logs.
5436 65. Users' Exim filters run in subprocesses under the user's uid. It is
5437 possible for a "deliver" command or an alias in a "personal" command to
5438 provoke an address rewrite. If logging of address rewriting is configured,
5439 this fails because the process is not running as root or exim. There may be
5440 a better way of dealing with this, but for the moment (because 4.30 needs
5441 to be released), I have disabled address rewrite logging when running a
5442 filter in a non-root, non-exim process.
5448 1. The buildconfig auxiliary program wasn't quoting the value set for
5449 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5450 not defined. This bug was masked in 4.22 by the effect that was fixed in
5453 2. Some messages that were rejected after a message id was allocated were
5454 shown as "incomplete" by exigrep. It no longer does this for messages that
5455 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5457 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5458 have allow_domain_literals set, the ID did not get logged in the <= line.
5459 Domain literals are now always recognized in Message-ID: header lines.
5461 4. The first argument for a ${extract expansion item is the key name or field
5462 number. Leading and trailing spaces in this item were not being ignored,
5463 causing some misleading effects.
5465 5. When deliver_drop_privilege was set, single queue runner processes started
5466 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5467 same command in the process it spins off) were not dropping privilege.
5469 6. When the daemon running as "exim" started a queue runner, it always
5470 re-executed Exim in the spun-off process. This is a waste of effort when
5471 deliver_drop_privilege is set. The new process now just calls the
5472 queue-runner function directly.
5478 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5481 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5482 ignored. Though the use of -oP was forcing the writing of a pid file, it
5483 was always written to the default place.
5485 3. If the message "no IP address found for host xxxx" is generated during
5486 incoming verification, it is now followed by identification of the incoming
5487 connection (so you can more easily find what provoked it).
5489 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5491 5. Added some features to "harden" Exim a bit more against certain attacks:
5493 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5494 be put in Local/Makefile. This is like the never_users runtime option,
5495 but it cannot be overridden. The default setting is "root".
5497 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5498 prefix string with which any file named in a -C command line option
5501 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5502 is retained for -C and -D only if the caller of Exim is root. Without
5503 it, the exim user may also use -C and -D and retain privilege.
5505 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5506 command line option is disabled.
5508 6. Macro names set by the -D option must start with an upper case letter, just
5509 like macro names defined in the configuration file.
5511 7. Added "dereference=" facility to LDAP.
5513 8. Two instances of the typo "uknown" in the source files are fixed.
5515 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5516 the Configure-Makefile script screwed up while processing it.
5518 10. Incorporated PCRE 4.4.
5520 11. The SMTP synchronization check was not operating right at the start of an
5521 SMTP session. For example, it could not catch a HELO sent before the client
5522 waited for the greeting. There is now a check for outstanding input at the
5523 point when the greeting is written. Because of the duplex, asynchronous
5524 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5525 way, but not yet received, when the check is performed.
5527 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5528 on TCP/IP sockets, because this apparently causes some broken clients to
5531 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5532 unchanged) from the Cygwin maintainer.
5534 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5535 was supported instead of testing for mbx. Effectively a typo.
5537 15. The spa authenticator server code was not checking that the input it
5538 received was valid base64.
5540 16. The debug output line for the "set" modifier in ACLs was not showing the
5541 name of the variable that was being set.
5543 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5545 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5546 lookup on its own. Something else was needed to trigger the lookup. For
5547 example, a match in host_lookup or the need for a host name in a host list.
5548 Now, if $sender_host_name is referenced and the host name has not yet been
5549 looked up, a lookup is performed. If the lookup fails, the variable remains
5550 empty, and $host_lookup_failed is set to "1".
5552 19. Added "eqi" as a case-independent comparison operator.
5554 20. The saslauthd authentication condition could segfault if neither service
5555 nor realm was specified.
5557 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5558 error message that was logged was misleading, and incoming SMTP
5559 connections were dropped. The message is now more accurate, and temporary
5560 errors are given to SMTP connections.
5562 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5563 (except RSET) with a 421 error, until QUIT is received. However, it was
5564 failing to send a response to QUIT.
5566 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5567 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5568 ACL is now run after the helo_try_verify_hosts code.
5570 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5571 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5572 case-independent, so other case variants are also recognized). Apparently
5573 some systems use these upper case variants.
5575 25. If more than two messages were waiting for the same host, and a transport
5576 filter was specified for the transport, Exim sent two messages over the
5577 same TCP/IP connection, and then failed with "socket operation on non-
5578 socket" when it tried to send the third.
5580 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5582 27. The extern definition of crypt16() in expand.c was not being excluded when
5583 the OS had its own crypt16() function.
5585 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5586 as a preferred synonym for return_size_limit, both as an option and as an
5589 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5591 30. Changed the default configuration ACL to relax the local part checking rule
5592 for addresses that are not in any local domains. For these addresses,
5593 slashes and pipe symbols are allowed within local parts, but the sequence
5594 /../ is explicitly forbidden.
5596 31. SPA server authentication was not clearing the challenge buffer before
5599 32. log_message in a "warn" ACL statement was writing to the reject log as
5600 well as to the main log, which contradicts the documentation and doesn't
5601 seem right (because no rejection is happening). So I have stopped it.
5603 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5604 However, I am unable to do any testing of this.
5606 34. Fixed an infelicity in the appendfile transport. When checking directories
5607 for a mailbox, to see if any needed to be created, it was accidentally
5608 using path names with one or more superfluous leading slashes; tracing
5609 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5611 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5612 contents are added to the log line that is written for every discarded
5613 recipient. (Previously a log_message setting was ignored.)
5615 36. The ${quote: operator now quotes the string if it is empty.
5617 37. The install script runs exim in order to find its version number. If for
5618 some reason other than non-existence or emptiness, which it checks, it
5619 could not run './exim', it was installing it with an empty version number,
5620 i.e. as "exim-". This error state is now caught, and the installation is
5623 38. An argument was missing from the function that creates an error message
5624 when Exim fails to connect to the socket for saslauthd authentication.
5625 This could cause Exim to crash, or give a corrupted message.
5627 39. Added isip, isip4, and isip6 to ${if conditions.
5629 40. The ACL variables $acl_xx are now saved with the message, and can be
5630 accessed later in routers, transports, and filters.
5632 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5633 strings in the file are not string-expanded.
5635 42. If a MAIL command specified a SIZE value that was too large to fit into an
5636 int variable, the check against message_size_limit failed. Such values are
5637 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5638 day this will have to be increased, but I don't think I want to be around
5639 when emails are that large.
5646 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5647 iconv() is not standard in FreeBSD.
5649 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5650 IPv6 enabled. The observed symptom was a segmentation fault on return from
5651 the function os_common_find_running_interfaces() in src/os.c.
5653 3. In the check_special_case() function in daemon.c I had used "errno" as an
5654 argument name, which causes warnings on some systems. This was basically a
5655 typo, since it was named "eno" in the comments!
5657 4. The code that waits for the clock to tick (at a resolution of some fraction
5658 of a second) so as to ensure message-id uniqueness was always waiting for
5659 at least one whole tick, when it could have waited for less. [This is
5660 almost certainly not relevant at current processor speeds, where it is
5661 unlikely to ever wait at all. But we try to future-proof.]
5663 5. The function that sleeps for a time interval that includes fractions of a
5664 second contained a race. It did not block SIGALRM between setting the
5665 timer, and suspending (a couple of lines later). If the interval was short
5666 and the sigsuspend() was delayed until after it had expired, the suspension
5667 never ended. On busy systems this could lead to processes getting stuck for
5670 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5671 process, before it forks any delivery processes. The open lookup caching
5672 mechanism meant that the open file or database connection was passed into
5673 the delivery process. The problem was that delivery processes always tidy
5674 up cached lookup data. This could cause a problem for the next delivery
5675 process started by the queue runner, because the external queue runner
5676 process does not know about the closure. So the next delivery process
5677 still has data in the lookup cache. In the case of a file lookup, there was
5678 no problem because closing a file descriptor in a subprocess doesn't affect
5679 the parent. However, if the lookup was caching a connection to a database,
5680 the connection was closed, and the second delivery process was likely to
5681 see errors such as "PGSQL: query failed: server closed the connection
5682 unexpectedly". The problem has been fixed by closing all cached lookups
5683 in a queue runner before running a delivery process.
5685 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5686 seem to have the "const" qualifier which it has on other OS. I've
5689 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5690 *of the same type* (client or server) with the same public name that an
5691 error should be diagnosed.
5693 9. When Exim looked up a host name for an IP address, but failed to find the
5694 original IP address when looking up the host name (a safety check), it
5695 output the message "<ip address> does not match any IP for NULL", which was
5696 confusing, to say the least. The bug was that the host name should have
5697 appeared instead of "NULL".
5699 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5700 user that is built into the binary, and the -C or -D options is used, root
5701 privilege is dropped before the configuration file is read. In addition,
5702 logging is switched to stderr instead of the normal log files. If the
5703 configuration then re-defines the Exim user, the unprivileged environment
5704 is probably not what is expected, so Exim logs a panic warning message (but
5707 However, if deliver_drop_privilege is set, the unprivileged state may well
5708 be exactly what is intended, so the warning has been cut out in that case,
5709 and Exim is allowed to try to write to its normal log files.
5715 1. smtp_return_error_details was not giving details for temporary sender
5716 or receiver verification errors.
5718 2. Diagnose a configuration error if two authenticators have the same public
5721 3. Exim used not to create the message log file for a message until the first
5722 delivery attempt. This could be confusing when incoming messages were held
5723 for policy or load reasons. The message log file is now created at the time
5724 the message is received, and an initial "Received" line is written to it.
5726 4. The automatically generated man page for command line options had a minor
5727 bug that caused no ill effects; however, a more serious problem was that
5728 the procedure for building the man page automatically didn't always
5729 operate. Consequently, release 4.20 contains an out-of-date version. This
5730 shouldn't happen again.
5732 5. When building Exim with embedded Perl support, the script that builds the
5733 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5734 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5736 6. The freeze_tell option was not being used for messages that were frozen on
5737 arrival, either by an ACL or by local_scan().
5739 7. Added the smtp_incomplete_transaction log selector.
5741 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5742 was accepting AUTH without a new EHLO.
5744 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5745 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5748 10. Logging of TCP/IP connections (when configured) now happens in the main
5749 daemon process instead of the child process, so that the TCP/IP connection
5750 count is more accurate (but it can never be perfect).
5752 11. The use of "drop" in a nested ACL was not being handled correctly in the
5753 outer ACL. Now, if condition failure induced by the nested "drop" causes
5754 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5755 or "require"), the connection is dropped.
5757 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5758 that yield "discard" can now be used with an "accept" or a "discard" verb,
5759 but an error is generated for any others (because I can't see a useful way
5760 to define what should happen).
5762 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5763 are now processed in the same way as lines in the Exim configuration file.
5764 In particular, continuation lines are supported.
5766 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5768 15. Added -ti meaning -t -i.
5770 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5771 domains, and warn by logging if others are found.
5773 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5774 returned from GIFCONF when Exim is trying to find the list of interfaces on
5775 a host. The code in os.c has been modified to copy each ifreq to an aligned
5776 structure in all cases.
5778 Also, in some cases, the returned ifreq's were being copied to a 'struct
5779 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5780 means the last couple of bytes of an IPv6 address could be chopped if the
5781 ifreq contained only a normal sockaddr (14 bytes storage).
5783 18. Named domain lists were not supported in the hosts_treat_as_local option.
5784 An entry such as +xxxx was not recognized, and was treated as a literal
5787 19. Ensure that header lines added by a DATA ACL are included in the reject log
5788 if the ACL subsequently rejects the message.
5790 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5791 MD5 (which is deprecated).
5793 21. When testing a filter file using -bf, Exim was writing a message when it
5794 took the sender from a "From " line in the message, but it was not doing so
5795 when it took $return_path from a Return-Path: header line. It now does.
5797 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5798 with a valid header line field name (a series of printing characters
5799 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5801 23. Changed "disc" in the source to "disk" to conform to the documentation and
5802 the book and for uniformity.
5804 24. Ignore Sendmail's -Ooption=value command line item.
5806 25. When execve() failed while trying to run a command in a pipe transport,
5807 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5808 could be confused with a return value of 69 from the command itself. This
5809 has been changed to 127, the value the shell returns if it is asked to run
5810 a non-existent command. The wording for the related log line suggests a
5811 non-existent command as the problem.
5813 26. If received_header_text expands to an empty string, do not add a Received:
5814 header line to the message. (Well, it adds a token one on the spool, but
5815 marks it "old" so that it doesn't get used or transmitted.)
5817 27. Installed eximstats 1.28 (addition of -nt option).
5819 28. There was no check for failure on the call to getsockname() in the daemon
5820 code. This can fail if there is a shortage of resources on the system, with
5821 ENOMEM, for example. A temporary error is now given on failure.
5823 29. Contrary to the C standard, it seems that in some environments, the
5824 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5825 program. Exim now does this explicitly; it affects the formatting of
5826 timestamps using strftime().
5828 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5829 complaints. I've now initialized all the variables, to avoid this.
5831 32. Header lines added by a system filter were not being "seen" during
5832 transport-time rewrites.
5834 33. The info_callback() function passed to OpenSSL is set up with type void
5835 (*)(SSL *, int, int), as described somewhere. However, when calling the
5836 function (actually a macro) that sets it up, the type void(*)() is
5837 expected. I've put in a cast to prevent warnings from picky compilers.
5839 34. If a DNS black list lookup found a CNAME record, but there were no A
5840 records associated with the domain it pointed at, Exim crashed.
5842 35. If a DNS black list lookup returned more than one A record, Exim ignored
5843 all but the first. It now scans all returned addresses if a particular IP
5844 value is being sought. In this situation, the contents of the
5845 $dnslist_value variable are a list of all the addresses, separated by a
5848 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5849 to accept a host name and all its aliases if the forward lookup for any of
5850 them yielded the IP address of the incoming connection. Now it accepts only
5851 those names whose forward lookup yields the correct IP address. Any other
5852 names are discarded. This closes a loophole whereby a rogue DNS
5853 administrator could create reverse DNS records to break through a
5854 wildcarded host restriction in an ACL.
5856 37. If a user filter or a system filter that ran in a subprocess used any of
5857 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5858 the wrong values were passed to the pipe command ($thisaddress had the
5859 value of $0, $0 had the value of $1, etc). This bug was introduced by
5860 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5862 38. Improved the line breaking for long SMTP error messages from ACLs.
5863 Previously, if there was no break point between 40 and 75 characters, Exim
5864 left the rest of the message alone. Two changes have been made: (a) I've
5865 reduced the minimum length to 35 characters; (b) if it can't find a break
5866 point between 35 and 75 characters, it looks ahead and uses the first one
5867 that it finds. This may give the occasional overlong line, but at least the
5868 remaining text gets split now.
5870 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5871 file descriptors might be low, and that setting 1000 would always raise it.
5872 It turns out that in some environments, the limit is already over 1000 and
5873 that lowering it causes trouble. So now Exim takes care not to decrease it.
5875 40. When delivering a message, the value of $return_path is set to $sender_
5876 address at the start of routing (routers may change the value). By an
5877 oversight, this default was not being set up when an address was tested by
5878 -bt or -bv, which affected the outcome if any router or filter referred to
5881 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5882 log only when "message" or "log_message" are set. However, if one of the
5883 conditions was an address verification, or a call to a nested ACL, the
5884 messages generated by the underlying test were being passed through. This
5885 no longer happens. The underlying message is available in $acl_verify_
5886 message for both "message" and "log_message" expansions, so it can be
5887 passed through if needed.
5889 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5890 new expansion $bh_ to give the encoded byte string without charset
5891 translation. Translation happens only if iconv() is available; HAVE_ICONV
5892 indicates this at build time. HEADERS_CHARSET gives the charset to
5893 translate to; headers_charset can change it in the configuration, and
5894 "headers charset" can change it in an individual filter file.
5896 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5897 that creates RFC 2047 encoded "words" labels them as that charset instead
5898 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5899 expansion operator; (ii) when Exim creates a From: line for a local
5900 message; (iii) when a header line is rewritten to include a "phrase" part.
5902 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5903 buggy, causing it to skip the first lines of messages whose message ID
5904 ended in 'D'. This would not have bitten before Exim release 4.14, because
5905 message IDs were unlikely to end in 'D' before then. The effect was to have
5906 incorrect size information for certain domains.
5908 45. #include "config.h" was missing at the start of the crypt16.c module. This
5909 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5912 46. If there was a timeout during a "random" callout check, Exim treated it as
5913 a failure of the random address, and carried on sending RSET and the real
5914 address. If the delay was just some slowness somewhere, the response to the
5915 original RCPT would be taken as a response to RSET and so on, causing
5916 mayhem of various kinds.
5918 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5919 when I implemented it. It didn't allow for the fact that some option values
5920 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5921 the right test for positive values.
5923 48. Domain names in DNS records are case-independent. Exim always looks them up
5924 in lower case. Some resolvers return domain names in exactly the case they
5925 appear in the zone file, that is, they may contain uppercase letters. Not
5926 all resolvers do this - some return always lower case. Exim was treating a
5927 change of case by a resolver as a change of domain, similar to a widening
5928 of a domain abbreviation. This triggered its re-routing code and so it was
5929 trying to route what was effectively the same domain again. This normally
5930 caused routing to fail (because the router wouldn't handle the domain
5931 twice). Now Exim checks for this case specially, and just changes the
5932 casing of the domain that it ultimately uses when it transmits the message
5935 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5938 50. If a filter generated a file delivery with a non-absolute name (possible if
5939 no home directory exists for the router), the forbid_file option was not
5942 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5943 the existing equality matching.
5945 52. Exim was using ints instead of ino_t variables in some places where it was
5946 dealing with inode numbers.
5948 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5949 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5950 called TMPDIR, and if it finds it is different, it changes its value.
5952 54. The smtp_printf() function is now made available to local_scan() so
5953 additional output lines can be written before returning. There is also an
5954 smtp_fflush() function to enable the detection of a dropped connection.
5955 The variables smtp_input and smtp_batched_input are exported to
5958 55. Changed the default runtime configuration: the message "Unknown user"
5959 has been removed from the ACL, and instead placed on the localuser router,
5960 using the cannot_route_message feature. This means that any verification
5961 failures that generate their own messages won't get overridden. Similarly,
5962 the "Unrouteable address" message that was in the ACL for unverifiable
5963 relay addresses has also been removed.
5965 56. Added hosts_avoid_esmtp to the smtp transport.
5967 57. The exicyclog script was not checking for the esoteric option
5968 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5969 will work only if exicyclog is run under the appropriate euid.
5971 58. Following a discussion on the list, the rules by which Exim recognises line
5972 endings on incoming messages have been changed. The -dropcr and drop_cr
5973 options are now no-ops, retained only for backwards compatibility. The
5974 following line terminators are recognized: LF CRLF CR. However, special
5975 processing applies to CR:
5977 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5978 nor a local message in the state where . is a terminator.
5980 (ii) If a bare CR is encountered in a header line, an extra space is added
5981 after the line terminator so as not to end the header. The reasoning
5982 behind this is that bare CRs in header lines are most likely either
5983 to be mistakes, or people trying to play silly games.
5985 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5986 was being incorrectly given as 18 bytes larger than it should have been.
5987 This is a VOB (very old bug).
5989 60. This may never have affected anything current, but just in case it has:
5990 When the local host is found other than at the start of a list of hosts,
5991 the local host, those with the same MX, and any that follow, are discarded.
5992 When the list in question was part of a longer list of hosts, the following
5993 hosts (not currently being processed) were also being discarded. This no
5994 longer happens. I'm not sure if this situation could ever has previously
5997 61. Added the "/MX" feature to lists of hosts in the manualroute and query
6000 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
6001 header. This is something that is recommended in a new Internet Draft, and
6002 is something that is documented as being done by Sendmail. There are two
6003 possible values. For messages generated by the autoreply transport, Exim
6006 Auto-Submitted: auto-replied
6008 whereas for all other generated messages (e.g. bounces) it adds
6010 Auto-Submitted: auto-generated
6012 63. The "personal" condition in filters now includes a test for the
6013 Auto-Submitted: header. If it contains the string "auto-" the message it
6014 not considered personal.
6016 64. Added rcpt_include_affixes as a generic transport option.
6018 65. Added queue_only_override (default true).
6020 66. Added the syslog_duplication option.
6022 67. If what should have been the first header line of a message consisted of
6023 a space followed by a colon, Exim was mis-interpreting it as a header line.
6024 It isn't of course - it is syntactically invalid and should therefore be
6025 treated as the start of the message body. The misbehaviour could have
6026 caused a number of strange effects, including loss of data in subsequent
6027 header lines, and spool format errors.
6029 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
6030 client host had authenticated. This control can now be exercised by an ACL
6031 for more flexibility.
6033 69. By default, callouts do not happen when testing with -bh. There is now a
6034 variant, -bhc, which does actually run the callout code, including
6035 consulting and updating the callout cache.
6037 70. Added support for saslauthd authentication, courtesy of Alexander
6040 71. If statvfs() failed on the spool or log directories while checking their
6041 size for availability, Exim confusingly gave the error "space shortage".
6042 Furthermore, in debugging mode it crashed with a floating point exception.
6043 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6044 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6045 serious problem, Exim now writes to the main and panic logs when this
6046 happens, with details of the failure. It then refuses to accept the
6047 incoming message, giving the message "spool directory problem" or "log
6048 directory problem" with a 421 code for SMTP messages.
6050 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6051 0, 1, and 2 exist, because some OS complain for execs without them (see
6052 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6053 descriptors. However, the code omitted to check that the open succeeded,
6054 causing mysterious errors if for some reason the permissions on /dev/null
6055 got screwed. Now Exim writes a message to the main and panic logs, and
6056 bombs out if it can't open /dev/null.
6058 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6059 interact so that it is all more flexible. It is supposed to remain
6060 backwards compatible. Also added extra_local_interfaces.
6062 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6063 to bomb out with an assertion failure - to the client this appears as a
6064 connection drop. This problem occurs in the part of the code that was taken
6065 from the Samba project. Fortunately, the assertion is in a very simple
6066 function, so I have fixed this by reproducing the function inline in the
6067 one place where it is called, and arranging for authentication to fail
6068 instead of killing the process with assert().
6070 75. The SPA client code was not working when the server requested OEM rather
6071 than Unicode encoding.
6073 76. Added code to make require_files with a specific uid setting more usable in
6074 the case where statting the file as root fails - usually a non-root-mounted
6075 NFS file system. When this happens and the failure is EACCES, Exim now
6076 forks a subprocess and does the per-uid checking as the relevant uid.
6078 77. Added process_log_path.
6080 78. If log_file_path was not explicitly set, a setting of check_log_space or
6081 check_log_inodes was ignored.
6083 79. If a space check for the spool or log partitions fails, the incident is now
6084 logged. Of course, in the latter case the data may get lost...
6086 80. Added the %p formatting code to string_format() so that it can be used to
6087 print addresses in debug_print(). Adjusted all the address printing in the
6088 debugging in store.c to use %p rather than %d.
6090 81. There was a concern that a line of code in smtp_in.c could overflow a
6091 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6092 initially expressed, the concern was not well-founded, because trailing
6093 spaces are removed early. However, if the trailing spaces were followed by
6094 a NULL, they did not get removed, so the overflow was possible. Two fixes
6097 (a) I re-wrote the offending code in a cleaner fashion.
6098 (b) If an incoming SMTP command contains a NULL character, it is rejected
6101 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6102 runs initgroups(), so that if the Exim user is in any additional groups,
6103 they will be used during message reception.
6109 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob