1 Change log file for Exim from version 4.21
2 -------------------------------------------
8 TF/01 Correctly close the server side of TLS when forking for delivery.
10 When a message was received over SMTP with TLS, Exim failed to clear up
11 the incoming connection properly after forking off the child process to
12 deliver the message. In some situations the subsequent outgoing
13 delivery connection happened to have the same fd number as the incoming
14 connection previously had. Exim would try to use TLS and fail, logging
15 a "Bad file descriptor" error.
17 TF/02 Portability fix for building lookup modules on Solaris when the xpg4
18 utilities have not been installed.
20 JH/01 Fix memory-handling in use of acl as a conditional; avoid free of
21 temporary space as the ACL may create new global variables.
23 TL/01 LDAP support uses per connection or global context settings, depending
24 upon the detected version of the libraries at build time.
26 TL/02 Experimental Proxy Protocol support: allows a proxied SMTP connection
27 to extract and use the src ip:port in logging and expansions as if it
28 were a direct connection from the outside internet. PPv2 support was
29 updated based on HAProxy spec change in May 2014.
31 JH/02 Add ${listextract {number}{list}{success}{fail}}.
33 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
34 Properly escape header and check for NULL return.
36 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
39 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
41 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
42 characters in header names, implemented as a verify condition.
43 Contributed by Michael Fischer v. Mollard.
45 TL/05 Rename SPF condition results err_perm and err_temp to standardized
46 results permerror and temperror. Previous values are deprecated but
47 still accepted. In a future release, err_perm and err_temp will be
48 completely removed, which will be a backward incompatibility if the
49 ACL tests for either of these two old results. Patch contributed by
50 user bes-internal on the mailing list.
52 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
54 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
55 selectors, in both main and reject logs.
57 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
60 JH/07 Add malware type "sock" for talking to simple daemon.
62 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
64 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
65 routers/transports under cutthrough routing.
67 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
68 numbers. Touch up "bool" conditional to keep the same definition.
70 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
72 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
74 JH/12 Expand items in router/transport headers_add or headers_remove lists
75 individually rather than the list as a whole. Bug 1452.
77 Required for reasonable handling of multiple headers_ options when
78 they may be empty; requires that headers_remove items with embedded
79 colons must have them doubled (or the list-separator changed).
81 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
82 view the policy declared in the DMARC record. Currently, $dmarc_status
83 is a combined value of both the record presence and the result of the
86 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
88 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
89 dnslookup router and the smtp transport (applying to the forward
92 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
93 of ldap servers used for a specific lookup. Patch provided by Heiko
96 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
97 New variable $lookup_dnssec_authenticated for observability.
99 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
100 Patch submitted by Lars Timman.
102 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
104 TL/10 Bugzilla 1454: New -oMm option to pass message reference to Exim.
105 Requires trusted mode and valid format message id, aborts otherwise.
106 Patch contributed by Heiko Schlichting.
108 JH/20 New expansion variables tls_(in,out)_(our,peer)cert, and expansion item
109 certextract with support for various fields. Bug 1358.
111 JH/21 Observability of OCSP via variables tls_(in,out)_ocsp. Stapling
112 is requested by default, modifiable by smtp transport option
115 JH/22 Expansion operators ${md5:string} and ${sha1:string} can now
116 operate on certificate variables to give certificate fingerprints
117 Also new ${sha256:cert_variable}.
119 JH/23 The PRDR feature is moved from being Experimental into the mainline.
121 TL/11 Bug 1119: fix memory allocation in string_printing2(). Patch from
122 Christian Aistleitner.
124 JH/24 The OCSP stapling feature is moved from Experimental into the mainline.
126 TL/12 Bug 1444: Fix improper \r\n sequence handling when writing spool
127 file. Patch from Wolfgang Breyha.
129 JH/25 Expand the coverage of the delivery $host and $host_address to
130 client authenticators run in verify callout. Bug 1476.
132 JH/26 Port service names are now accepted for tls_on_connect_ports, to
133 align with daemon_smtp_ports. Bug 72.
135 TF/03 Fix udpsend. The ip_connectedsocket() function's socket type
136 support and error reporting did not work properly.
138 TL/13 Bug 1495: Exiqgrep check if -C config file specified on cli exists
139 and is readable. Patch from Andrew Colin Kissa.
141 TL/14 Enhance documentation of ${run expansion and how it parses the
142 commandline after expansion, particularly in the case when an
143 unquoted variable expansion results in an empty value.
145 JH/27 The TLS SNI feature was broken in 4.82. Fix it.
147 PP/02 Fix internal collision of T_APL on systems which support RFC3123
148 by renaming away from it. Addresses GH issue 15, reported by
155 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
157 PP/02 Make -n do something, by making it not do something.
158 When combined with -bP, the name of an option is not output.
160 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
163 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
164 $sender_host_name and config options to manage this, and basic check
167 PP/05 DSCP support for outbound connections and control modifier for inbound.
169 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
170 (Only plugin which currently uses this is kerberos4, which nobody should
171 be using, but we should make it available and other future plugins might
172 conceivably use it, even though it would break NAT; stuff *should* be
173 using channel bindings instead).
175 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
176 name; added for Sendmail compatibility; requires admin caller.
177 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
178 just ignore it); requires trusted caller.
179 Also parse but ignore: -Ac -Am -X<logfile>
182 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
184 TL/02 Add +smtp_confirmation as a default logging option.
186 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
187 Patch by Magnus Holmgren from 2007-02-20.
189 TL/04 Bugzilla 1281 - Spec typo.
190 Bugzilla 1283 - Spec typo.
191 Bugzilla 1290 - Spec grammar fixes.
193 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
195 TL/06 Add Experimental DMARC support using libopendmarc libraries.
197 TL/07 Fix an out of order global option causing a segfault. Reported to dev
198 mailing list by by Dmitry Isaikin.
200 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
202 JH/02 Support "G" suffix to numbers in ${if comparisons.
204 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
206 NM/01 Bugzilla 1197 - Spec typo
207 Bugzilla 1196 - Spec examples corrections
209 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
211 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
212 gnutls_enable_pkcs11, but renamed to more accurately indicate its
215 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
216 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
218 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
219 "acl {{name}{arg}...}", and optional args on acl condition
222 JH/05 Permit multiple router/transport headers_add/remove lines.
224 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
226 JH/07 Avoid using a waiting database for a single-message-only transport.
227 Performance patch from Paul Fisher. Bugzilla 1262.
229 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
232 JH/09 Add $headers_added variable, with content from use of ACL modifier
233 add_header (but not yet added to the message). Bugzilla 199.
235 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
236 Pulled from Bugzilla 817 by Wolfgang Breyha.
238 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
240 (nb: this is the same fix as in Exim 4.80.1)
242 JH/11 Add A= logging on delivery lines, and a client_set_id option on
245 JH/12 Add optional authenticated_sender logging to A= and a log_selector
248 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
250 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
251 advertise SMTP AUTH mechanism to us, instead of a generic
252 protocol violation error. Also, make Exim more robust to bad
253 data from the Dovecot auth socket.
255 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
257 When a queue runner is handling a message, Exim first routes the
258 recipient addresses, during which it prunes them based on the retry
259 hints database. After that it attempts to deliver the message to
260 any remaining recipients. It then updates the hints database using
263 So if a recipient address works intermittently, it can get repeatedly
264 deferred at routing time. The retry hints record remains fresh so the
265 address never reaches the final cutoff time.
267 This is a fairly common occurrence when a user is bumping up against
268 their storage quota. Exim had some logic in its local delivery code
269 to deal with this. However it did not apply to per-recipient defers
270 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
272 This change adds a proper retry rule check during routing so that the
273 final cutoff time is checked against the message's age. We only do
274 this check if there is an address retry record and there is not a
275 domain retry record; this implies that previous attempts to handle
276 the address had the retry_use_local_parts option turned on. We use
277 this as an approximation for the destination being like a local
278 delivery, as in LMTP.
280 I suspect this new check makes the old local delivery cutoff check
281 redundant, but I have not verified this so I left the code in place.
283 TF/02 Correct gecos expansion when From: is a prefix of the username.
285 Test 0254 submits a message to Exim with the header
289 When I ran the test suite under the user fanf2, Exim expanded
290 the header to contain my full name, whereas it should have added
291 a Resent-Sender: header. It erroneously treats any prefix of the
292 username as equal to the username.
294 This change corrects that bug.
296 GF/01 DCC debug and logging tidyup
297 Error conditions log to paniclog rather than rejectlog.
298 Debug lines prefixed by "DCC: " to remove any ambiguity.
300 TF/03 Avoid unnecessary rebuilds of lookup-related code.
302 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
303 Bug spotted by Jeremy Harris; was flawed since initial commit.
304 Would have resulted in OCSP responses post-SNI triggering an Exim
305 NULL dereference and crash.
307 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
309 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
310 Bug detection, analysis and fix by Samuel Thibault.
311 Bugzilla 1331, Debian bug #698092.
313 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
315 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
316 Server implementation by Todd Lyons, client by JH.
317 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
318 config variable "prdr_enable" controls whether the server
319 advertises the facility. If the client requests PRDR a new
320 acl_data_smtp_prdr ACL is called once for each recipient, after
321 the body content is received and before the acl_smtp_data ACL.
322 The client is controlled by bolth of: a hosts_try_prdr option
323 on the smtp transport, and the server advertisement.
324 Default client logging of deliveries and rejections involving
325 PRDR are flagged with the string "PRDR".
327 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
328 fclose(). Diagnosis by Todd Lyons.
330 PP/17 Update configure.default to handle IPv6 localhost better.
331 Patch by Alain Williams (plus minor tweaks).
334 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
335 This is now consistent with GnuTLS, and is now documented: the
336 previous undocumented portable approach to treating the option as
337 unset was to force an expansion failure. That still works, and
338 an empty string is now equivalent.
340 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
341 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
342 not performing validation itself.
344 PP/20 Added force_command boolean option to pipe transport.
345 Patch from Nick Koston, of cPanel Inc.
347 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
350 TF/04 Added udpsend ACL modifer and hexquote expansion operator
352 PP/21 Fix eximon continuous updating with timestamped log-files.
353 Broken in a format-string cleanup in 4.80, missed when I repaired the
354 other false fix of the same issue.
355 Report and fix from Heiko Schlichting.
358 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
359 Report from Prashanth Katuri.
361 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
362 It's SecureTransport, so affects any MacOS clients which use the
363 system-integrated TLS libraries, including email clients.
365 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
366 using a MIME ACL for non-SMTP local injection.
367 Report and assistance in diagnosis by Warren Baker.
369 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
371 JH/16 Fix comparisons for 64b. Bugzilla 1385.
373 TL/09 Add expansion variable $authenticated_fail_id to keep track of
374 last id that failed so it may be referenced in subsequent ACL's.
376 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
379 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
380 ldap library initialization, allowing self-signed CA's to be
381 used. Also properly sets require_cert option later in code by
382 using NULL (global ldap config) instead of ldap handle (per
383 session). Bug diagnosis and testing by alxgomz.
385 TL/12 Enhanced documentation in the ratelimit.pl script provided in
386 the src/util/ subdirectory.
388 TL/13 Bug 1031 - Imported transport SQL logging patch from Axel Rau
389 renamed to Transport Post Delivery Action by Jeremy Harris, as
392 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
393 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
394 redis_servers = needs to be configured which will be used by the redis
395 lookup. Patch from Warren Baker, of The Packet Hub.
397 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
399 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
400 hostname or reverse DNS when processing a host list. Used suggestions
401 from multiple comments on this bug.
403 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
405 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
406 Missed a few lines, added it to make the runtest require no keyboard
409 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
410 contains upper case chars. Make router use caseful_local_part.
412 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
413 support when GnuTLS has been built with p11-kit.
419 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
421 This, or similar/improved, will also be change PP/11 of 4.82.
427 PP/01 Handle short writes when writing local log-files.
428 In practice, only affects FreeBSD (8 onwards).
429 Bugzilla 1053, with thanks to Dmitry Isaikin.
431 NM/01 Bugzilla 949 - Documentation tweak
433 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
436 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
438 PP/02 Implemented gsasl authenticator.
440 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
442 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
443 `pkg-config foo` for cflags/libs.
445 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
446 with rest of GSASL and with heimdal_gssapi.
448 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
449 `pkg-config foo` for cflags/libs for the TLS implementation.
451 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
452 properties get this fed in as external SSF. A number of robustness
453 and debugging improvements to the cyrus_sasl authenticator.
455 PP/08 cyrus_sasl server now expands the server_realm option.
457 PP/09 Bugzilla 1214 - Log authentication information in reject log.
458 Patch by Jeremy Harris.
460 PP/10 Added dbmjz lookup type.
462 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
464 PP/12 MAIL args handles TAB as well as SP, for better interop with
465 non-compliant senders.
466 Analysis and variant patch by Todd Lyons.
468 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
469 Bug report from Lars Müller <lars@samba.org> (via SUSE),
470 Patch from Dirk Mueller <dmueller@suse.com>
472 PP/13 tls_peerdn now print-escaped for spool files.
473 Observed some $tls_peerdn in wild which contained \n, which resulted
474 in spool file corruption.
476 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
477 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
478 or write after TLS renegotiation, which otherwise led to messages
481 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
482 as a tracking header (ie: a signed header comes before the signature).
483 Patch from Wolfgang Breyha.
485 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
486 comma-sep list; embedded commas doubled.
488 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
490 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
492 Report and patch from Dmitry Banschikov.
494 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
495 Removed SSL_clear() after SSL_new() which led to protocol negotiation
496 failures. We appear to now support TLS1.1+ with Exim.
498 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
499 lets Exim select keys and certificates based upon TLS SNI from client.
500 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
501 before an outbound SMTP session. New log_selector, +tls_sni.
503 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
504 NULL dereference. Report and patch from Alun Jones.
506 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
507 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
508 Not seeing resolver debug output on NetBSD, but suspect this is a
509 resolver implementation change.
511 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
512 Left warnings. Added "eximon gdb" invocation mode.
514 PP/21 Defaulting "accept_8bitmime" to true, not false.
516 PP/22 Added -bw for inetd wait mode support.
518 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
519 locate the relevant includes and libraries. Made this the default.
521 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
522 Bugzilla 1246, report and most of solution from Tomasz Kusy.
524 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
525 This may cause build issues on older platforms.
527 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
528 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
529 gnutls_require_mac & gnutls_require_protocols (no longer supported).
530 Added SNI support via GnuTLS too.
531 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
533 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
535 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
536 Applied second patch from Janne, implementing suggestion to default
537 multiple-strings-in-record handling to match SPF spec.
539 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
541 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
542 read-only, out of scope).
543 Patch from Wolfgang Breyha, report from Stuart Northfield.
545 PP/29 Fix three issues highlighted by clang analyser static analysis.
546 Only crash-plausible issue would require the Cambridge-specific
547 iplookup router and a misconfiguration.
548 Report from Marcin Mirosław.
550 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
552 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
553 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
554 As part of this, removing so much warning spew let me fix some minor
555 real issues in debug logging.
557 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
558 assignment on my part. Fixed.
560 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
561 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
562 Janne Snabb (who went above and beyond: thank you).
564 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
565 string otherwise requires a connection and a bunch more work and it's
566 relatively easy to get wrong. Should also expose TLS library linkage
569 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
570 64-bit ${eval} (JH/03).
572 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
573 GNU libc to support some of the 64-bit stuff, should not lead to
574 conflicts. Defined before os.h is pulled in, so if a given platform
575 needs to override this, it can.
577 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
578 protection layer was required, which is not implemented.
579 Bugzilla 1254, patch from Wolfgang Breyha.
581 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
582 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
583 tls_dhparam take prime identifiers. Also unbreak combination of
584 OpenSSL+DH_params+TLSSNI.
586 PP/39 Disable SSLv2 by default in OpenSSL support.
592 PP/01 Solaris build fix for Oracle's LDAP libraries.
593 Bugzilla 1109, patch from Stephen Usher.
595 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
597 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
600 TF/02 Fix a couple more cases where we did not log the error message
601 when unlink() failed. See also change 4.74-TF/03.
603 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
604 lock up or crash if it happened to be inside a call to libc when it
605 got a SIGUSR1 from exiwhat.
607 The SIGUSR1 handler appends the current process status to the process
608 log which is later printed by exiwhat. It used to use the general
609 purpose logging code to do this, but several functions it calls are
610 not safe for signals.
612 The new output code in the SIGUSR1 handler is specific to the process
613 log, and simple enough that it's easy to inspect for signal safety.
614 Removing some special cases also simplifies the general logging code.
615 Removing the spurious timestamps from the process log simplifies
618 TF/04 Improved ratelimit ACL condition.
620 The /noupdate option has been deprecated in favour of /readonly which
621 has clearer semantics. The /leaky, /strict, and /readonly update modes
622 are mutually exclusive. The update mode is no longer included in the
623 database key; it just determines when the database is updated. (This
624 means that when you upgrde Exim will forget old rate measurements.)
626 Exim now checks that the per_* options are used with an update mode that
627 makes sense for the current ACL. For example, when Exim is processing a
628 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
629 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
630 must specify per_mail/readonly. If you omit the update mode it defaults to
631 /leaky where that makes sense (as before) or /readonly where required.
633 The /noupdate option is now undocumented but still supported for
634 backwards compatibility. It is equivalent to /readonly except that in
635 ACLs where /readonly is required you may specify /leaky/noupdate or
636 /strict/noupdate which are treated the same as /readonly.
638 A useful new feature is the /count= option. This is a generalization
639 of the per_byte option, so that you can measure the throughput of other
640 aggregate values. For example, the per_byte option is now equivalent
641 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
643 The per_rcpt option has been generalized using the /count= mechanism
644 (though it's more complicated than the per_byte equivalence). When it is
645 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
646 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
647 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
648 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
649 Note that using per_rcpt with a non-readonly update mode in more than
650 one ACL will cause the recipients to be double-counted. (The per_mail
651 and per_byte options don't have this problem.)
653 The handling of very low rates has changed slightly. If the computed rate
654 is less than the event's count (usually one) then this event is the first
655 after a long gap. In this case the rate is set to the same as this event's
656 count, so that the first message of a spam run is counted properly.
658 The major new feature is a mechanism for counting the rate of unique
659 events. The new per_addr option counts the number of different
660 recipients that someone has sent messages to in the last time period. It
661 behaves like per_rcpt if all the recipient addresses are different, but
662 duplicate recipient addresses do not increase the measured rate. Like
663 the /count= option this is a general mechanism, so the per_addr option
664 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
665 example, measure the rate that a client uses different sender addresses
666 with the options per_mail/unique=$sender_address. There are further
667 details in the main documentation.
669 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
671 TF/06 Removed a few PCRE remnants.
673 TF/07 Automatically extract Exim's version number from tags in the git
674 repository when doing development or release builds.
676 PP/02 Raise smtp_cmd_buffer_size to 16kB.
677 Bugzilla 879. Patch from Paul Fisher.
679 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
680 Heavily based on revision 40f9a89a from Simon Arlott's tree.
683 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
685 PP/05 Variable $av_failed, true if the AV scanner deferred.
686 Bugzilla 1078. Patch from John Horne.
688 PP/06 Stop make process more reliably on build failure.
689 Bugzilla 1087. Patch from Heiko Schlittermann.
691 PP/07 Make maildir_use_size_file an _expandable_ boolean.
692 Bugzilla 1089. Patch from Heiko Schlittermann.
694 PP/08 Handle ${run} returning more data than OS pipe buffer size.
695 Bugzilla 1131. Patch from Holger Weiß.
697 PP/09 Handle IPv6 addresses with SPF.
698 Bugzilla 860. Patch from Wolfgang Breyha.
700 PP/10 GnuTLS: support TLS 1.2 & 1.1.
702 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
705 PP/11 match_* no longer expand right-hand-side by default.
706 New compile-time build option, EXPAND_LISTMATCH_RHS.
707 New expansion conditions, "inlist", "inlisti".
709 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
711 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
713 PP/14 fix log_write() format string regression from TF/03.
714 Bugzilla 1152. Patch from Dmitry Isaikin.
720 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
722 PP/02 Harmonised TLS library version reporting; only show if debugging.
723 Layout now matches that introduced for other libraries in 4.74 PP/03.
725 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
727 PP/04 New "dns_use_edns0" global option.
729 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
732 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
733 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
735 TK/01 Updated PolarSSL code to 0.14.2.
736 Bugzilla 1097. Patch from Andreas Metzler.
738 PP/07 Catch divide-by-zero in ${eval:...}.
741 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
744 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
745 format-string attack -- SECURITY: remote arbitrary code execution.
747 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
748 time unintentionally subject to list matching rules, letting the header
749 cause arbitrary Exim lookups (of items which can occur in lists, *not*
750 arbitrary string expansion). This allowed for information disclosure.
752 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
753 INT_MIN/-1 -- value coerced to INT_MAX.
759 NM/01 Workround for PCRE version dependency in version reporting
762 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
763 This fixes portability to compilers other than gcc, notably
764 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
766 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
767 makefiles for portability to HP-UX and POSIX correctness.
769 PP/01 Permit LOOKUP_foo enabling on the make command-line.
770 Also via indented variable definition in the Makefile.
771 (Debugging by Oliver Heesakkers).
773 PP/02 Restore caching of spamd results with expanded spamd_address.
774 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
776 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
777 Improves build reliability. Fix from: Frank Elsner
779 NM/02 Fix wide character breakage in the rfc2047 coding
780 Fixes bug 1064. Patch from Andrey N. Oktyabrski
782 NM/03 Allow underscore in dnslist lookups
783 Fixes bug 1026. Patch from Graeme Fowler
785 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
786 Code patches from Adam Ciarcinski of NetBSD.
788 NM/04 Fixed exiqgrep to cope with mailq missing size issue
791 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
792 is logged, to avoid truncation. Patch from John Horne.
794 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
795 Patch from Jakob Hirsch.
797 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
798 SQL string expansion failure details.
799 Patch from Andrey Oktyabrski.
801 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
802 Patch from Simon Arlott.
804 PP/09 New lookups functionality failed to compile on old gcc which rejects
805 extern declarations in function scope.
806 Patch from Oliver Fleischmann
808 PP/10 Use sig_atomic_t for flags set from signal handlers.
809 Check getgroups() return and improve debugging.
810 Fixed developed for diagnosis in bug 927 (which turned out to be
813 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
814 Patch from Mark Zealey.
816 PP/12 Bugzilla 1056: Improved spamd server selection.
817 Patch from Mark Zealey.
819 PP/13 Bugzilla 1086: Deal with maildir quota file races.
820 Based on patch from Heiko Schlittermann.
822 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
823 Patch from Uwe Doering, sign-off by Michael Haardt.
825 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
826 variable declaration deep within a block. Bug and patch from
829 PP/15 lookups-Makefile IRIX compatibilty coercion.
831 PP/16 Make DISABLE_DKIM build knob functional.
833 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
834 Patch by Simon Arlott
836 TF/03 Fix valgrind.h portability to C89 compilers that do not support
837 variable argument macros. Our copy now differs from upstream.
843 TF/01 Failure to get a lock on a hints database can have serious
844 consequences so log it to the panic log.
846 TF/02 Log LMTP confirmation messages in the same way as SMTP,
847 controlled using the smtp_confirmation log selector.
849 TF/03 Include the error message when we fail to unlink a spool file.
851 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
852 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
853 for maintaining out-of-tree patches for some time.
855 PP/01 Bugzilla 139: Documentation and portability issues.
856 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
857 Handle per-OS dynamic-module compilation flags.
859 PP/02 Let /dev/null have normal permissions.
860 The 4.73 fixes were a little too stringent and complained about the
861 permissions on /dev/null. Exempt it from some checks.
862 Reported by Andreas M. Kirchwitz.
864 PP/03 Report version information for many libraries, including
865 Exim version information for dynamically loaded libraries. Created
866 version.h, now support a version extension string for distributors
867 who patch heavily. Dynamic module ABI change.
869 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
870 privilege escalation vulnerability whereby the Exim run-time user
871 can cause root to append content of the attacker's choosing to
874 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
877 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
878 If dropping privileges for untrusted macros, we disabled normal logging
879 on the basis that it would fail; for the Exim run-time user, this is not
880 the case, and it resulted in successful deliveries going unlogged.
881 Fixed. Reported by Andreas Metzler.
887 PP/01 Date: & Message-Id: revert to normally being appended to a message,
888 only prepend for the Resent-* case. Fixes regression introduced in
889 Exim 4.70 by NM/22 for Bugzilla 607.
891 PP/02 Include check_rfc2047_length in configure.default because we're seeing
892 increasing numbers of administrators be bitten by this.
894 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
896 PP/03 Bugzilla 994: added openssl_options main configuration option.
898 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
900 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
902 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
904 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
905 without a peer certificate, leading to a segfault because of an
906 assumption that peers always have certificates. Be a little more
907 paranoid. Problem reported by Martin Tscholak.
909 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
910 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
911 NB: ClamAV planning to remove STREAM in "middle of 2010".
912 CL also introduces -bmalware, various -d+acl logging additions and
913 more caution in buffer sizes.
915 PP/09 Implemented reverse_ip expansion operator.
917 PP/10 Bugzilla 937: provide a "debug" ACL control.
919 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
921 PP/12 Bugzilla 973: Implement --version.
923 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
925 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
927 PP/15 Bugzilla 816: support multiple condition rules on Routers.
929 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
930 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
931 ignore trailing whitespace.
933 JJ/02 prevent non-panic DKIM error from being sent to paniclog
935 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
938 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
939 Notification from Dr Andrew Aitchison.
941 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
942 ExtendedDetectionInfo response format.
943 Notification from John Horne.
945 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
948 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
949 XSL and documented dependency on system catalogs, with examples of how
952 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
955 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
956 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
957 configuration file which is writeable by the Exim user or group.
959 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
960 of configuration files to cover files specified with the -C option if
961 they are going to be used with root privileges, not just the default
964 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
965 option (effectively making it always true).
967 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
968 files to be used while preserving root privileges.
970 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
971 that rogue child processes cannot use them.
973 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
974 run-time user, instead of root.
976 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
977 Exim run-time user without dropping privileges.
979 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
980 result string, instead of calling string_vformat() twice with the same
983 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
984 for other users. Others should always drop root privileges if they use
985 -C on the command line, even for a whitelisted configure file.
987 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
989 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
995 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
996 $data_path, and $header_path variables; fixed documentation bugs and
999 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
1000 exipick to access non-standard spools, including the "frozen" queue
1003 NM/01 Bugzilla 965: Support mysql stored procedures.
1004 Patch from Alain Williams
1006 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
1008 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
1009 Patch from Andreas Metzler
1011 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
1012 Patch from Kirill Miazine
1014 NM/05 Bugzilla 671: Added umask to procmail example.
1016 JJ/03 installed exipick 20100323.0, fixing doc bug
1018 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
1019 directory. Notification and patch from Dan Rosenberg.
1021 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
1023 TK/02 Improve log output when DKIM signing operation fails.
1025 MH/01 Treat the transport option dkim_domain as a colon separated
1026 list, not as a single string, and sign the message with each element,
1027 omitting multiple occurences of the same signer.
1029 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
1030 Bugzilla 985, 986. Patch by Simon Arlott
1032 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
1033 Patch by Simon Arlott
1035 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
1036 MBX locking. Notification from Dan Rosenberg.
1042 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
1044 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
1046 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1048 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1050 NM/04 Bugzilla 915: Flag broken perl installation during build.
1056 TK/01 Added patch by Johannes Berg that expands the main option
1057 "spamd_address" if it starts with a dollar sign.
1059 TK/02 Write list of recipients to X-Envelope-Sender header when building
1060 the mbox-format spool file for content scanning (suggested by Jakob
1063 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1064 (http://www.dcc-servers.net/) support via dccifd. Activated by
1065 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1067 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1068 by Mark Daniel Reidel <mr@df.eu>.
1070 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1071 When building exim an external PCRE library is now needed -
1072 PCRE is a system library on the majority of modern systems.
1073 See entry on PCRE_LIBS in EDITME file.
1075 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1076 conversation. Added nologin parameter to request.
1077 Patch contributed by Kirill Miazine.
1079 TF/01 Do not log submission mode rewrites if they do not change the address.
1081 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1083 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1084 log files in place. Contributed by Roberto Lima.
1086 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1088 TF/03 Bugzilla 615: When checking the local_parts router precondition
1089 after a local_part_suffix or local_part_prefix option, Exim now
1090 does not use the address's named list lookup cache, since this
1091 contains cached lookups for the whole local part.
1093 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1094 Robert Millan. Documentation is in experimental-spec.txt.
1096 TF/04 Bugzilla 668: Fix parallel build (make -j).
1098 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1100 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1101 Patch provided by Jan Srzednicki.
1103 TF/05 Leading white space used to be stripped from $spam_report which
1104 wrecked the formatting. Now it is preserved.
1106 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1107 that they are available at delivery time.
1109 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1111 TF/08 TLS error reporting now respects the incoming_interface and
1112 incoming_port log selectors.
1114 TF/09 Produce a more useful error message if an SMTP transport's hosts
1115 setting expands to an empty string.
1117 NM/06 Bugzilla 744: EXPN did not work under TLS.
1118 Patch provided by Phil Pennock.
1120 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1121 Patch provided by Richard Godbee.
1123 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1124 acl_smtp_notquit, added index entry.
1126 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1127 Patch provided by Eugene Bujak.
1129 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1130 accept(). Patch provided by Maxim Dounin.
1132 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1133 Patch provided by Phil Pennock.
1135 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1137 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1138 Patch provided by Brad "anomie" Jorsch.
1140 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1141 Patch provided by Dean Brooks.
1143 TK/05 Add native DKIM support (does not depend on external libraries).
1145 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1146 Patch provided by Graeme Fowler.
1148 NM/16 Bugzilla 851: Documentation example syntax fix.
1150 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1152 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1155 NM/19 Bugzilla 745: TLS version reporting.
1156 Patch provided by Phil Pennock.
1158 NM/20 Bugzilla 167: bool: condition support.
1159 Patch provided by Phil Pennock.
1161 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1162 clients. Patch provided by Phil Pennock.
1164 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1165 Patch provided by Brad "anomie" Jorsch.
1167 NM/23 Bugzilla 687: Fix misparses in eximstats.
1168 Patch provided by Heiko Schlittermann.
1170 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1171 Patch provided by Heiko Schlittermann.
1173 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1174 plus update to original patch.
1176 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1178 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1179 Patch provided by David Brownlee.
1181 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1183 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1185 NM/30 Bugzilla 888: TLS documentation bugfixes.
1187 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1189 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1190 Unlike the original bugzilla I have changed all shell scripts in src tree.
1192 NM/33 Bugzilla 898: Transport filter timeout fix.
1193 Patch by Todd Rinaldo.
1195 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1196 Patch by Serge Demonchaux.
1198 NM/35 Bugzilla 39: Base64 decode bug fixes.
1199 Patch by Jakob Hirsch.
1201 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1203 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1205 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1207 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1213 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1214 ALT-N's libdkim that I have put here:
1215 http://duncanthrax.net/exim-experimental/
1217 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1218 were called 'true' and it seems that C99 defines that as a reserved
1219 keyword to be used with 'bool' variable types. That means you could
1220 not include C99-style headers which use bools without triggering
1221 build errors in sieve.c.
1223 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1224 as mailq or other aliases. Changed the --help handling significantly
1225 to do whats expected. exim_usage() emits usage/help information.
1227 SC/01 Added the -bylocaldomain option to eximstats.
1229 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1231 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1233 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1239 PH/01 Another patch from the Sieve maintainer.
1241 PH/02 When an IPv6 address is converted to a string for single-key lookup
1242 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1243 dots are used instead of colons so that keys in lsearch files need not
1244 contain colons. This was done some time before quoting was made available
1245 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1246 (notated using the quote facility) so as to distinguish them from IPv4
1247 keys. This meant that lookups for IP addresses in host lists did not work
1248 for iplsearch lookups.
1250 This has been fixed by arranging for IPv6 addresses to be expressed with
1251 colons if the lookup type is iplsearch. This is not incompatible, because
1252 previously such lookups could never work.
1254 The situation is now rather anomolous, since one *can* have colons in
1255 ordinary lsearch keys. However, making the change in all cases is
1256 incompatible and would probably break a number of configurations.
1258 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1261 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1262 conversion specification without a maximum field width, thereby enabling
1263 a rogue spamd server to cause a buffer overflow. While nobody in their
1264 right mind would setup Exim to query an untrusted spamd server, an
1265 attacker that gains access to a server running spamd could potentially
1266 exploit this vulnerability to run arbitrary code as the Exim user.
1268 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1269 $primary_hostname instead of what libspf2 thinks the hosts name is.
1271 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1272 a directory entry by the name of the lookup key. Previously, if a
1273 symlink pointed to a non-existing file or a file in a directory that
1274 Exim lacked permissions to read, a lookup for a key matching that
1275 symlink would fail. Now it is enough that a matching directory entry
1276 exists, symlink or not. (Bugzilla 503.)
1278 PH/03 The body_linecount and body_zerocount variables are now exported in the
1281 PH/04 Added the $dnslist_matched variable.
1283 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1284 This means they are set thereafter only if the connection becomes
1287 PH/06 Added the client_condition to authenticators so that some can be skipped
1288 by clients under certain conditions.
1290 PH/07 The error message for a badly-placed control=no_multiline_responses left
1291 "_responses" off the end of the name.
1293 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1295 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1296 (without spaces) instead of just copying the configuration text.
1298 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1300 PH/11 Added $max_received_linelength.
1302 PH/12 Added +ignore_defer and +include_defer to host lists.
1304 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1305 way in which PCRE > 7.0 is built.
1307 PH/14 Implemented queue_only_load_latch.
1309 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1310 MAIL command. The effect was to mangle the value on 64-bit systems.
1312 PH/16 Another patch from the Sieve maintainer.
1314 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1316 PH/18 If a system quota error occurred while trying to create the file for
1317 a maildir delivery, the message "Mailbox is full" was not appended to the
1318 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1319 only to a quota excession during the actual writing of the file.
1321 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1322 characters?) which causes problems in log lines. The DN values are now
1323 passed through string_printing() before being added to log lines.
1325 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1326 and InterBase are left for another time.)
1328 PH/21 Added message_body_newlines option.
1330 PH/22 Guard against possible overflow in moan_check_errorcopy().
1332 PH/23 POSIX allows open() to be a macro; guard against that.
1334 PH/24 If the recipient of an error message contained an @ in the local part
1335 (suitably quoted, of course), incorrect values were put in $domain and
1336 $local_part during the evaluation of errors_copy.
1342 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1343 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1346 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1347 issue a MAIL command.
1349 PH/02 In an ACL statement such as
1351 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1353 if a client was not listed at all, or was listed with a value other than
1354 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1355 the condition was not true (as it should be), so access was not denied.
1356 The bug was that the ! inversion was incorrectly passed on to the second
1357 item. This has been fixed.
1359 PH/03 Added additional dnslists conditions == and =& which are different from
1360 = and & when the dns lookup returns more than one IP address.
1362 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1363 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1365 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1366 FSYNC, which compiles an option called disable_fsync that allows for
1367 bypassing fsync(). The documentation is heavily laced with warnings.
1369 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1371 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1372 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1373 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1374 including adding "make clean"; (3) Added -fPIC when compiling the test
1375 dynamically loaded module, to get rid of a warning.
1377 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1378 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1379 = 0s. The bug is otherwise harmless.
1381 PH/07 There was a bug in the dovecot authenticator such that the value of
1382 $auth1 could be overwritten, and so not correctly preserved, after a
1383 successful authentication. This usually meant that the value preserved by
1384 the server_setid option was incorrect.
1386 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1388 PH/09 Installed PCRE release 7.0.
1390 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1391 run for batched SMTP input. It is now run at the start of every message
1392 in the batch. While fixing this I discovered that the process information
1393 (output by running exiwhat) was not always getting set for -bs and -bS
1394 input. This is fixed, and it now also says "batched" for BSMTP.
1396 PH/11 Added control=no_pipelining.
1398 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1399 patch, slightly modified), and move the expansion of helo_data till after
1400 the connection is made in the smtp transport (so it can use these
1403 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1405 PH/14 Added log_selector = +pid.
1407 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1409 PH/16 Add ${if forany and ${if forall.
1411 PH/17 Added dsn_from option to vary the From: line in DSNs.
1413 PH/18 Flush SMTP output before performing a callout, unless control =
1414 no_callout_flush is set.
1416 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1417 was true (the default) a successful delivery failed to delete the retry
1418 item, thus causing premature timeout of the address. The bug is now
1421 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1423 PH/21 Long custom messages for fakedefer and fakereject are now split up
1424 into multiline reponses in the same way that messages for "deny" and
1425 other ACL rejections are.
1427 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1428 with slight modification.
1430 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1431 draft, changing the syntax and factoring some duplicate code".
1433 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1434 for deliveries of the second and subsequent messages over the same SMTP
1437 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1438 ${reduce, with only minor "tidies".
1440 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1442 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1443 expansion side effects.
1445 PH/27 When a message times out after an over-quota error from an Exim-imposed
1446 quota, the bounce message says "mailbox is full". This message was not
1447 being given when it was a system quota that was exceeded. It now should
1450 MH/03 Made $recipients available in local_scan(). local_scan() already has
1451 better access to the recipient list through recipients_list[], but
1452 $recipients can be useful in postmaster-provided expansion strings.
1454 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1455 in the case of a MAIL command with additional options following the
1456 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1457 were accidentally chopped off.
1459 PH/29 SMTP synchronization checks are implemented when a command is read -
1460 there is a check that no more input is waiting when there shouldn't be
1461 any. However, for some commands, a delay in an ACL can mean that it is
1462 some time before the response is written. In this time, more input might
1463 arrive, invalidly. So now there are extra checks after an ACL has run for
1464 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1465 pipelining has not been advertised.
1467 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1469 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1470 returned in the SMTP response when smtp_return_error_details was false.
1471 This has been fixed.
1473 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1474 instead of the C I/O that was originally supplied, because problems were
1475 reported on Solaris.
1477 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1478 Exim which did not show up earlier: it was assuming that a call to
1479 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1480 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1481 was a macro that became an assignment, so it seemed to work. This has
1482 changed to a proper function call with a void return, hence the compile
1483 error. Exim's code has been fixed.
1485 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1488 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1490 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1492 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1493 because it turns out to be too big or there is a timeout) there is no
1494 "Completed" line in the log. When some messages of this type were
1495 selected by exigrep, they were listed as "not completed". Others were
1496 picked up by some special patterns. I have improved the selection
1497 criteria to be more general.
1499 PH/38 The host_find_failed option in the manualroute router can now be set
1500 to "ignore", to completely ignore a host whose IP address cannot be
1501 found. If all hosts are ignored, the behaviour is controlled by the new
1502 host_all_ignored option.
1504 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1505 homing or because of multiple MX records with /mx) generated more than
1506 one IP address, and the following item turned out to be the local host,
1507 all the secondary addresses of the first item were incorrectly removed
1508 from the list, along with the local host and any following hosts (which
1509 is what is supposed to happen).
1511 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1512 whoever called Exim into the -H file. In the case of the daemon it was
1513 behaving confusingly. When first started, it used values for whoever
1514 started the daemon, but after a SIGHUP it used the Exim user (because it
1515 calls itself on a restart). I have changed the code so that it now always
1518 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1519 message are rejected with the same error (e.g. no authentication or bad
1520 sender address), and a DATA command is nevertheless sent (as can happen
1521 with PIPELINING or a stupid MUA), the error message that was given to the
1522 RCPT commands is included in the rejection of the DATA command. This is
1523 intended to be helpful for MUAs that show only the final error to their
1526 PH/42 Another patch from the Sieve maintainer.
1528 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1529 Eximstats - Fixed some broken HTML links and added missing column headers
1531 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1532 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1534 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1536 PH/43 Yet another patch from the Sieve maintainer.
1538 PH/44 I found a way to check for a TCP/IP connection going away before sending
1539 the response to the final '.' that terminates a message, but only in the
1540 case where the client has not sent further data following the '.'
1541 (unfortunately, this is allowed). However, in many cases there won't be
1542 any further data because there won't be any more messages to send. A call
1543 to select() can be used: if it shows that the input is "ready", there is
1544 either input waiting, or the socket has been closed. An attempt to read
1545 the next input character can distinguish the two cases. Previously, Exim
1546 would have sent an OK response which the client would never have see.
1547 This could lead to message repetition. This fix should cure that, at
1548 least in a lot of common cases.
1550 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1551 advertised in response to EHLO.
1557 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1558 fixed by 4.65/MH/01 (is this a record?) are fixed:
1560 (i) An empty string was always treated as zero by the numeric comparison
1561 operators. This behaviour has been restored.
1563 (ii) It is documented that the numeric comparison operators always treat
1564 their arguments as decimal numbers. This was broken in that numbers
1565 starting with 0 were being interpreted as octal.
1567 While fixing these problems I realized that there was another issue that
1568 hadn't been noticed. Values of message_size_limit (both the global option
1569 and the transport option) were treated as octal if they started with 0.
1570 The documentation was vague. These values are now always treated as
1571 decimal, and I will make that clear in the documentation.
1577 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1578 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1581 MH/01 Don't check that the operands of numeric comparison operators are
1582 integers when their expansion is in "skipping" mode (fixes bug
1583 introduced by 4.64-PH/07).
1585 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1586 child addresses, Exim now panics and dies. Previously, because the count
1587 is held in a short int, deliveries were likely to be lost. As such a
1588 large number of recipients for a single message is ridiculous
1589 (performance will be very, very poor), I have chosen to impose a limit
1590 rather than extend the field.
1596 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1597 leftover -K file (the existence of which was triggered by #402).
1598 While we were at it, introduced process PID as part of the -K
1599 filename. This should rule out race conditions when creating
1602 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1603 processing considerably. Previous code took too long for large mails,
1604 triggering a timeout which in turn triggers #401.
1606 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1607 in the DK code in transports.c. sendfile() is not really portable,
1608 hence the _LINUX specificness.
1610 TF/01 In the add_headers option to the mail command in an Exim filter,
1611 there was a bug that Exim would claim a syntax error in any
1612 header after the first one which had an odd number of characters
1615 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1616 callout verification, Exim cached a "reject" for the entire domain. This
1617 is correct for most verifications, but it is not correct for a recipient
1618 verification with use_sender or use_postmaster set, because in that case
1619 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1620 case of MAIL FROM:<> rejection from other early rejections (e.g.
1621 rejection of HELO). When verifying a recipient using a non-null MAIL
1622 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1623 Whatever the result of the callout, the value of the domain cache is
1624 left unchanged (for any other kind of callout, getting as far as trying
1625 RCPT means that the domain itself is ok).
1627 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1630 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1631 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1632 session, the daemon ignores ECONNECT errors and logs others; it now
1633 ignores EPIPE as well.
1635 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1636 (quoted-printable decoding).
1638 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1639 later the small subsequent patch to fix an introduced bug.
1641 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1643 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1645 PH/08 An error is now given if message_size_limit is specified negative.
1647 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1648 to be given (somewhat) arbitrary names.
1650 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1653 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1654 miscellaneous code fixes
1656 PH/10 Added the log_reject_target ACL modifier to specify where to log
1659 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1660 hostname. This is wrong, because it relates to the incoming message (and
1661 probably the interface on which it is arriving) and not to the outgoing
1662 callout (which could be using a different interface). This has been
1663 changed to use the value of the helo_data option from the smtp transport
1664 instead - this is what is used when a message is actually being sent. If
1665 there is no remote transport (possible with a router that sets up host
1666 addresses), $smtp_active_hostname is used.
1668 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1669 tweaks were necessary in order to get it to work (see also 21 below):
1670 (a) The code assumed that strncpy() returns a negative number on buffer
1671 overflow, which isn't the case. Replaced with Exim's string_format()
1673 (b) There were several signed/unsigned issues. I just did the minimum
1674 hacking in of casts. There is scope for a larger refactoring.
1675 (c) The code used strcasecmp() which is not a standard C function.
1676 Replaced with Exim's strcmpic() function.
1677 (d) The code set only $1; it now sets $auth1 as well.
1678 (e) A simple test gave the error "authentication client didn't specify
1679 service in request". It would seem that Dovecot has changed its
1680 interface. Fortunately there's a specification; I followed it and
1681 changed what the client sends and it appears to be working now.
1683 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1686 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1687 address A is aliased to B and C, where B exists and C does not. Without
1688 -v the output is "A verified" because verification stops after a
1689 successful redirection if more than one address is generated. However,
1690 with -v the child addresses are also verified. Exim was outputting "A
1691 failed to verify" and then showing the successful verification for C,
1692 with its parentage. It now outputs "B failed to verify", showing B's
1693 parentage before showing the successful verification of C.
1695 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1696 look up a TXT record in a specific list after matching in a combined
1699 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1700 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1701 they consult the DNS. I had assumed they would set it the way they
1702 wanted; and indeed my experiments on Linux seem to show that in some
1703 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1704 To be on the safe side, however, I have now made the interface to
1705 host_find_byname() similar to host_find_bydns(), with an argument
1706 containing the DNS resolver options. The host_find_byname() function now
1707 sets these options at its start, just as host_find_bydns() does. The smtp
1708 transport options dns_qualify_single and dns_search_parents are passed to
1709 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1710 of host_find_byname() use the default settings of RES_DEFNAMES
1711 (qualify_single) but not RES_DNSRCH (search_parents).
1713 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1714 spool_read_header() do less string testing, by means of a preliminary
1715 switch on the second character of optional "-foo" lines. (This is
1716 overdue, caused by the large number of possibilities that now exist.
1717 Originally there were few.) While I was there, I also converted the
1718 str(n)cmp tests so they don't re-test the leading "-" and the first
1719 character, in the hope this might squeeze out yet more improvement.
1721 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1722 flag allowing group syntax was set by the header_syntax check but not
1723 turned off, possible causing trouble later; (2) The flag was not being
1724 set at all for the header_verify test, causing "group"-style headers to
1725 be rejected. I have now set it in this case, and also caused header_
1726 verify to ignore an empty address taken from a group. While doing this, I
1727 came across some other cases where the code for allowing group syntax
1728 while scanning a header line wasn't quite right (mostly, not resetting
1729 the flag correctly in the right place). These bugs could have caused
1730 trouble for malformed header lines. I hope it is now all correct.
1732 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1733 with the "reply" argument non-NULL. The code, however (which originally
1734 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1735 but it didn't always do it. This confused somebody who was copying the
1736 code for some other use. I have removed all the tests.
1738 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1739 feature that was used to support insecure browsers during the U.S. crypto
1740 embargo. It requires special client support, and Exim is probably the
1741 only MTA that supported it -- and would never use it because real RSA is
1742 always available. This code has been removed, because it had the bad
1743 effect of slowing Exim down by computing (never used) parameters for the
1744 RSA_EXPORT functionality.
1746 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1747 authenticator to fail if there's a tab character in the incoming data
1748 (there should never be unless someone is messing about, as it's supposed
1749 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1750 if the connection is using TLS or if the remote IP is the same as the
1751 local IP, and the "valid-client-cert option" if a client certificate has
1754 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1755 authenticators. This can be used for authorization after authentication
1756 succeeds. (In the case of plaintext, it servers for both authentication
1759 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1760 if any retry times were supplied.
1762 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1763 connection, where there is no client IP address to check. In this
1764 situation, the verify now always succeeds.
1766 PH/25 Applied John Jetmore's -Mset patch.
1768 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1770 PH/27 In a string expansion for a processed (not raw) header when multiple
1771 headers of the same name were present, leading whitespace was being
1772 removed from all of them, but trailing whitespace was being removed only
1773 from the last one. Now trailing whitespace is removed from each header
1774 before concatenation. Completely empty headers in a concatenation (as
1775 before) are ignored.
1777 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1778 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1780 PH/29 [Removed. This was a change that I later backed out, and forgot to
1781 correct the ChangeLog entry (that I had efficiently created) before
1782 committing the later change.]
1784 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1785 address errors (4xx response to RCPT) over the same connection as other
1786 messages routed to the same hosts. Such deliveries are always "forced",
1787 so retry times are not inspected. This resulted in far too many retries
1788 for the affected addresses. The effect occurred only when there were more
1789 hosts than the hosts_max_try setting in the smtp transport when it had
1790 the 4xx errors. Those hosts that it had tried were not added to the list
1791 of hosts for which the message was waiting, so if all were tried, there
1792 was no problem. Two fixes have been applied:
1794 (i) If there are any address or message errors in an SMTP delivery, none
1795 of the hosts (tried or untried) are now added to the list of hosts
1796 for which the message is waiting, so the message should not be a
1797 candidate for sending over the same connection that was used for a
1798 successful delivery of some other message. This seems entirely
1799 reasonable: after all the message is NOT "waiting for some host".
1800 This is so "obvious" that I'm not sure why it wasn't done
1801 previously. Hope I haven't missed anything, but it can't do any
1802 harm, as the worst effect is to miss an optimization.
1804 (ii) If, despite (i), such a delivery is accidentally attempted, the
1805 routing retry time is respected, so at least it doesn't keep
1806 hammering the server.
1808 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1809 in ${readsocket because some servers need this prod.
1811 PH/32 Added some extra debug output when updating a wait-xxx database.
1813 PH/33 The hint "could be header name not terminated by colon", which has been
1814 given for certain expansion errors for a long time, was not being given
1815 for the ${if def:h_colon_omitted{... case.
1817 PH/34 The spec says: "With one important exception, whenever a domain list is
1818 being scanned, $domain contains the subject domain." There was at least
1819 one case where this was not true.
1821 PH/35 The error "getsockname() failed: connection reset by peer" was being
1822 written to the panic log as well as the main log, but it isn't really
1823 panic-worthy as it just means the connection died rather early on. I have
1824 removed the panic log writing for the ECONNRESET error when getsockname()
1827 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1828 runs only) independently of the message's sender address. This meant
1829 that, if the 4xx error was in fact related to the sender, a different
1830 message to the same recipient with a different sender could confuse
1831 things. In particualar, this can happen when sending to a greylisting
1832 server, but other circumstances could also provoke similar problems.
1833 I have changed the default so that the retry time for these errors is now
1834 based a combination of the sender and recipient addresses. This change
1835 can be overridden by setting address_retry_include_sender=false in the
1838 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1839 remote server are returned as part of bounce messages. This was not
1840 happening for LMTP over a pipe (the lmtp transport), but now it is the
1841 same for both kinds of LMTP.
1843 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1844 in header lines that were in fact CNAMEs. This is no longer the case.
1846 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1847 and queue runs started by the daemon processed all messages. This has
1848 been fixed so that -R and -S can now usefully be given with -q<time>.
1850 PH/40 Import PCRE release 6.7 (fixes some bugs).
1852 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1854 PH/42 Give an error if -q is specified more than once.
1856 PH/43 Renamed the variables $interface_address and $interface_port as
1857 $received_ip_address and $received_port, to make it clear that these
1858 values apply to message reception, and not to the outgoing interface when
1859 a message is delivered. (The old names remain recognized, of course.)
1861 PH/44 There was no timeout on the connect() call when using a Unix domain
1862 socket in the ${readsocket expansion. There now is.
1864 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1865 be meaningful with "accept".
1867 SC/01 Eximstats V1.43
1868 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1870 SC/02 Eximstats V1.44
1871 Use a glob alias rather than an array ref in the generated
1872 parser. This improves both readability and performance.
1874 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1875 Collect SpamAssassin and rejection statistics.
1876 Don't display local sender or destination tables unless
1877 there is data to show.
1878 Added average volumes into the top table text output.
1880 SC/04 Eximstats V1.46
1881 Collect data on the number of addresses (recipients)
1882 as well as the number of messages.
1884 SC/05 Eximstats V1.47
1885 Added 'Message too big' to the list of mail rejection
1886 reasons (thanks to Marco Gaiarin).
1888 SC/06 Eximstats V1.48
1889 Mainlog lines which have GMT offsets and are too short to
1890 have a flag are now skipped.
1892 SC/07 Eximstats V1.49 (Alain Williams)
1893 Added the -emptyok flag.
1895 SC/08 Eximstats V1.50
1896 Fixes for obtaining the IP address from reject messages.
1898 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1899 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1900 whitesspace changes from 4.64-PH/27
1902 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1905 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1906 are found, allow negative numbers in numeric criteria)
1908 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1910 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1911 to match changes made in 4.64-PH/43
1913 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1915 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1916 "domain" is actually an IP address, and force a failure. This locks out
1917 those revolvers/nameservers that support "A-for-A" lookups, in
1918 contravention of the specifications.
1920 PH/48 When a host name was looked up from an IP address, and the subsequent
1921 forward lookup of the name timed out, the host name was left in
1922 $sender_host_name, contrary to the specification.
1924 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1925 restricted to single-key lookups, Exim was not diagnosing an error if
1926 * or *@ was used with a query-style lookup.
1928 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1930 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1931 long ago, but noone interested enough thought of it. Let's just say that
1932 the "1.1" means that there are some new functions that weren't there at
1933 some point in the past.
1935 PH/51 Error processing for expansion failure of helo_data from an smtp
1936 transport during callout processing was broken.
1938 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1939 tested/used via the -bh/-bhc/-bs options.
1941 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1942 bug, fixed in subsequent PCRE releases).
1944 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1945 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1947 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1953 SC/01 Use a glob alias rather than an array ref in eximstats generated
1954 parser. This improves both readability and performance.
1956 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1957 Don't display local sender or destination tables in eximstats unless
1958 there is data to show.
1959 Added average volumes into the eximstats top table text output.
1961 SC/03 Collect data on the number of addresses (recipients) as well
1962 as the number of messages in eximstats.
1964 TF/01 Correct an error in the documentation for the redirect router. Exim
1965 does (usually) call initgroups() when daemonizing.
1967 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1968 with consistent privilege compared to when running as a daemon.
1970 TF/03 Note in the spec that $authenticated_id is not set for local
1971 submissions from trusted users.
1973 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1974 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1976 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1977 by adding some example configuration directives to the default
1978 configuration file. A little bit of work is required to uncomment the
1979 directives and define how usernames and passwords are checked, but
1980 there is now a framework to start from.
1982 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1983 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1984 without this. I don't know how relevant this is to other LDAP libraries.
1986 PH/02 Add the verb name to the "unknown ACL verb" error.
1988 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1990 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1992 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1993 directory not expanded when it should be if an expanded home directory
1994 was set for the address (which is overridden by the transport).
1996 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1999 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
2000 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
2001 because it is too late at that time, and has no effect.
2003 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
2004 security issue with \' (bugzilla #107). I could not use the
2005 PQescapeStringConn() function, because it needs a PGconn value as one of
2008 PH/08 When testing addresses using -bt, indicate those final addresses that
2009 are duplicates that would not cause an additional delivery. At least one
2010 person was confused, thinking that -bt output corresponded to deliveries.
2011 (Suppressing duplicates isn't a good idea as you lose the information
2012 about possibly different redirections that led to the duplicates.)
2014 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
2015 systems where poll() doesn't work, in particular OS X.
2017 PH/10 Added more information to debugging output for retry time not reached.
2019 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
2020 operations in malware.c.
2022 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
2025 PH/13 If write_rejectlog was set false when logging was sent to syslog with
2026 syslog_duplication set false, log lines that would normally be written
2027 both the the main log and to the reject log were not written to syslog at
2030 PH/14 In the default configuration, change the use of "message" in ACL warn
2031 statements to "add_header".
2033 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
2034 not followed by a command (e.g. "seen endif").
2036 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
2037 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
2040 PH/17 Added extra conditions to the default value of delay_warning_condition
2044 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
2045 { match{$h_precedence:}{(?i)bulk|list|junk} } \
2046 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2049 The Auto-Submitted: and various List- headers are standardised, whereas I
2050 don't think Precedence: ever was.
2052 PH/18 Refactored debugging code in route_finduser() to show more information,
2053 in particular, the error code if getpwnam() issues one.
2055 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2056 This is apparently needed in addition to the PH/07 change above to avoid
2057 any possible encoding problems.
2059 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2060 but not after initializing Perl.
2062 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2063 output them only if debugging. By default they are written stderr,
2064 apparently, which is not desirable.
2066 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2069 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2072 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2074 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2075 authenticated or an ident call has been made. Suppress the default
2076 values for $authenticated_id and $authenticated_sender (but permit -oMai
2077 and -oMas) when testing with -bh.
2079 PH/24 Re-jigged the order of the tests in the default configuration so that the
2080 tests for valid domains and recipients precede the DNS black list and CSA
2081 tests, on the grounds that those ones are more expensive.
2083 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2084 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2085 This bug exists in every version of Exim that I still have, right back to
2088 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2089 However, an attempt to turn on case-sensitivity in a regex key by
2090 including (?-i) didn't work because the subject string was already
2091 lowercased, and the effects were non-intuitive. It turns out that a
2092 one-line patch can be used to allow (?-i) to work as expected.
2098 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2099 other effects) broke the use of negated acl sub-conditions.
2101 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2104 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2105 "Deny" causes Exim to reject the incoming connection with a 554 error.
2106 Unfortunately, if there is a major crisis, such as a disk failure,
2107 tcp-wrappers gives "deny", whereas what one would like would be some
2108 kind of temporary error. A kludge has been added to help with this.
2109 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2110 554 error is used if errno is still zero or contains ENOENT (which occurs
2111 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2114 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2116 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2117 errors. Otherwise a message that provokes a temporary error (when other
2118 messages do not) can cause a whole host to time out.
2120 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2121 addresses were routed directly to files or pipes from a redirect router.
2122 File deliveries just didn't batch; pipe deliveries might have suffered
2125 PH/06 A failure to get a lock for a hints database would erroneously always say
2126 "Failed to get write lock", even when it was really a read lock.
2128 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2129 of 0600. This has been changed to use the value of the lockfile_mode
2130 option (which defaults to 0600).
2132 PH/08 Applied small patch from the Sieve maintainer.
2134 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2135 folder from quota calculations, a direct delivery into this folder messed
2136 up the contents of the maildirsize file. This was because the regex was
2137 used only to exclude .Trash (or whatever) when the size of the mailbox
2138 was calculated. There was no check that a delivery was happening into an
2139 excluded directory. This bug has been fixed by ignoring all quota
2140 processing for deliveries into excluded directories.
2142 PH/10 Added the maildirfolder_create_regex option to appendfile.
2148 PH/01 The code for finding all the local interface addresses on a FreeBSD
2149 system running IPv6 was broken. This may well have applied to all BSD
2150 systems, as well as to others that have similar system calls. The broken
2151 code found IPv4 interfaces correctly, but gave incorrect values for the
2152 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2153 that it would not match correctly against @[] and not recognize the IPv6
2156 PH/02 The ipliteral router was not recognizing addresses of the form user@
2157 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2159 PH/03 Added disable_ipv6.
2161 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2162 decoded form, because it is most often used to construct To: headers
2163 lines in autoreplies, and the decoded form may well be syntactically
2164 invalid. However, $reply_address has leading white space removed, and all
2165 newlines turned into spaces so that the autoreply transport does not
2168 PH/05 If group was specified without a user on a router, and no group or user
2169 was specified on a transport, the group from the router was ignored.
2171 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2172 for visible compile-time settings that can be used to change these
2173 numbers, for those that want even more. Backwards compatibility with old
2174 spool files has been maintained. However, going back to a previous Exim
2175 release will lost any variables that are in spool files.
2177 PH/07 Two small changes when running in the test harness: increase delay when
2178 passing a TCP/IP connection to a new process, in case the original
2179 process has to generate a bounce, and remove special handling of
2180 127.0.0.2 (sic), which is no longer necessary.
2182 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2183 be the same on different OS.
2185 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2188 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2189 whether --show-vars was specified or not
2191 JJ/02 exipick: Added support for new ACL variable spool format introduced
2194 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2195 syntactically invalid From: or Reply-to: line, and a filter used this to
2196 generate an autoreply, and therefore failed to obtain an address for the
2197 autoreply, Exim could try to deliver to a non-existent relative file
2198 name, causing unrelated and misleading errors. What now happens is that
2199 it logs this as a hard delivery error, but does not attempt to create a
2202 PH/11 The exinext utility has a -C option for testing purposes, but although
2203 the given file was scanned by exinext itself; it wasn't being passed on
2204 when Exim was called.
2206 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2207 an end-of-file indication when reading a command response.
2209 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2210 compiled. In many other places in Exim, IPv6 addresses are always
2211 recognized, so I have changed this. It also means that IPv4 domain
2212 literals of the form [IPV4:n.n.n.n] are now always recognized.
2214 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2215 used if the router is not running as root, for example, when verifying at
2216 ACL time, or when using -bh. The debugging output from this situation was
2217 non-existent - all you got was a failure to exec. I have made two
2220 (a) Failures to set uid/gid, the current directory, or a process leader
2221 in a subprocess such as that created by queryprogram now generate
2222 suitable debugging ouput when -d is set.
2224 (b) The queryprogram router detects when it is not running as root,
2225 outputs suitable debugging information if -d is set, and then runs
2226 the subprocess without attempting to change uid/gid.
2228 PH/15 Minor change to Makefile for building test_host (undocumented testing
2231 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2232 additional section of a DNS packet that returns MX or SRV records.
2233 Instead, it always explicitly searches for A/AAAA records. This avoids
2234 major problems that occur when a DNS server includes only records of one
2235 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2236 fixed another bug: if SRV records were looked up and the corresponding
2237 address records were *not* found in the additional section, the port
2238 values from the SRV records were lost.
2240 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2241 using the correct key (the original address) when searching the retry
2242 rules in order to find which one to use for generating the retry hint.
2244 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2245 adding the default one. Similarly, if it contains a Reply-To: header, the
2246 errors_reply_to option, if set, is not used.
2248 PH/19 When calculating a retry time, Exim used to measure the "time since
2249 failure" by looking at the "first failed" field in the retry record. Now
2250 it does not use this if it is later than than the arrival time of the
2251 message. Instead it uses the arrival time. This makes for better
2252 behaviour in cases where some deliveries succeed, thus re-setting the
2253 "first failed" field. An example is a quota failure for a huge message
2254 when small messages continue to be delivered. Without this change, the
2255 "time since failure" will always be short, possible causing more frequent
2256 delivery attempts for the huge message than are intended.
2257 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2259 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2260 $1, $2, $3) because the numerical variables can be reset during some
2261 expansion items (e.g. "match"), thereby losing the authentication data.
2263 PH/21 Make -bV show the size of off_t variables so that the test suite can
2264 decide whether to run tests for quotas > 2G.
2266 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2267 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2268 filecount value is greater than 2G or if a quota value is greater than 2G
2269 on a system where the size of off_t is not greater than 4, a panic error
2272 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2273 never match. The debug and -bh output now contains an explicit error
2274 message indicating a malformed IPv4 address or mask.
2276 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2277 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2278 PH/23 above applies.
2280 PH/25 Do not write to syslog when running in the test harness. The only
2281 occasion when this arises is a failure to open the main or panic logs
2282 (for which there is an explicit test).
2284 PH/26 Added the /no_tell option to "control=freeze".
2286 PH/27 If a host name lookup failed very early in a connection, for example, if
2287 the IP address matched host_lookup and the reverse lookup yielded a name
2288 that did not have a forward lookup, an error message of the form "no IP
2289 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2290 could be logged. Now it outputs the IP address instead of "NULL".
2292 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2293 allows the sender and the authenticated sender to be set when
2294 submitting a message from within Exim. Since child_open_exim() is
2295 documented for local_scan(), the new function should be too.
2297 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2298 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2299 results in an empty string is now treated as unset.
2301 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2303 PH/31 Added sender_verify_fail logging option.
2305 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2306 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2307 tidied the source and removed it altogether.
2309 PH/33 When a queue run was abandoned because the load average was too high, a
2310 log line was always written; now it is written only if the queue_run log
2311 selector is set. In addition, the log line for abandonment now contains
2312 information about the queue run such as the pid. This is always present
2313 in "start" and "stop" lines but was omitted from the "abandon" line.
2315 PH/34 Omit spaces between a header name and the colon in the error message that
2316 is given when verify = headers_syntax fails (if there are lots of them,
2317 the message gets confusing).
2319 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2320 names, as there are now some PTR records that contain slashes. This check
2321 is only to protect against broken name servers that fall over on strange
2322 characters, so the fact that it applies to all lookups doesn't matter.
2324 PH/36 Now that the new test suite is complete, we can remove some of the
2325 special code in Exim that was needed for the old test suite. For example,
2326 sorting DNS records because real resolvers return them in an arbitrary
2327 order. The new test suite's fake resolver always returns records in the
2330 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2331 bounces) except when queue_only is set, to avoid logging races between
2332 the different processes.
2334 PH/38 Panic-die if .include specifies a non-absolute path.
2336 PH/39 A tweak to the "H" retry rule from its user.
2338 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2339 a label. They prevented compilation on older perls.
2341 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2342 a warning to be raised on newish perls.
2344 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2345 on queue. Changes to match documented behaviour of showing count of
2346 messages matching specified criteria.
2348 PH/40 Changed the default ident timeout from 30s to 5s.
2350 PH/41 Added support for the use of login_cap features, on those BSD systems
2351 that have them, for controlling the resources used by pipe deliveries.
2353 PH/42 The content-scanning code uses fopen() to create files in which to put
2354 message data. Previously it was not paying any attention to the mode of
2355 the files. Exim runs with umask(0) because the rest of the code creates
2356 files with open(), and sets the required mode explicitly. Thus, these
2357 files were ending up world-writeable. This was not a big issue, because,
2358 being within the spool directory, they were not world-accessible. I have
2359 created a function called modefopen, which takes an additional mode
2360 argument. It sets umask(777), creates the file, chmods it to the required
2361 mode, then resets the umask. All the relevant calls to fopen() in the
2362 content scanning code have been changed to use this function.
2364 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2365 to 24 hours. This avoids potential overflow problems when processing G
2366 and H retry rules. I suspect nobody ever tinkers with this value.
2368 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2370 PH/45 When the plaintext authenticator is running as a client, the server's
2371 challenges are checked to ensure they are valid base64 strings. By
2372 default, the authentication attempt is cancelled if an invalid string is
2373 received. Setting client_ignore_invalid_base64 true ignores these errors.
2374 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2375 they are received. Thus, the responses can be made to depend on the
2376 challenges. If an invalid string is ignored, an empty string is placed in
2379 PH/46 Messages that are created by the autoreply transport now contains a
2380 References: header, in accordance with RFCs 2822 and 3834.
2382 PH/47 Added authenticated_sender_force to the smtp transport.
2384 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2386 PH/49 Installed latest patch from the Sieve maintainer.
2388 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2389 was also set, the appendfile transport was unnecessarily scanning a
2390 directory of message files (e.g. for maildir delivery) to find the count
2391 of files (along with the size), even though it did not need this
2392 information. It now does the scan only if it needs to find either the
2393 size of the count of files.
2395 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2397 PH/52 Two bugs concerned with error handling when the smtp transport is
2400 (i) Exim was not creating retry information for temporary errors given
2401 for individual recipients after the DATA command when the smtp transport
2402 was used in LMTP mode. This meant that they could be retried too
2403 frequently, and not timed out correctly.
2405 (ii) Exim was setting the flag that allows error details to be returned
2406 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2407 recipients that were returned after the DATA command.
2409 PH/53 This is related to PH/52, but is more general: for any failing address,
2410 when detailed error information was permitted to be returned to the
2411 sender, but the error was temporary, then after the final timeout, only
2412 "retry timeout exceeded" was returned. Now it returns the full error as
2413 well as "retry timeout exceeded".
2415 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2416 do this, and (what is worse) MTAs that accept it.
2418 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2419 will now be deprecated.
2421 PH/56 New os.c-cygwin from the Cygwin maintainer.
2423 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2424 formats (previously only available in exim formats via -bpr, -bpru,
2425 and -bpra. Now also available in native and exiqgrep formats)
2427 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2428 with very large, slow to parse queues
2430 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2432 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2434 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2435 responses to authentication challenges, though it was showing the
2436 challenges; (ii) I've removed the CR characters from the debug output for
2439 PH/58 Allow for the insertion of a newline as well as a space when a string
2440 is turned into more than one encoded-word during RFC 2047 encoding. The
2441 Sieve code now uses this.
2443 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2444 data_4xx, lost_connection, tls_required.
2446 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2447 message was being sent as an SMTP response.
2449 PH/61 Add -l and -k options to exicyclog.
2451 PH/62 When verifying, if an address was redirected to one new address, so that
2452 verification continued, and the new address failed or deferred after
2453 having set something in $address_data, the value of $address_data was not
2454 passed back to the ACL. This was different to the case when no
2455 redirection occurred. The value is now passed back in both cases.
2457 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2458 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2459 use login_cap.h, so on its own it isn't the distinguishing feature. The
2460 new name refers directly to the setclassresources() function.
2462 PH/65 Added configuration files for NetBSD3.
2464 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2466 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2467 is preferred over IPv4.
2469 PH/68 The bounce_return_message and bounce_return_body options were not being
2470 honoured for bounces generated during the reception of non-SMTP messages.
2471 In particular, this applied to messages rejected by the ACL. This bug has
2472 been fixed. However, if bounce_return_message is true and bounce_return_
2473 body is false, the headers that are returned for a non-SMTP message
2474 include only those that have been read before the error was detected.
2475 (In the case of an ACL rejection, they have all been read.)
2477 PH/69 The HTML version of the specification is now built in a directory called
2478 spec_html instead of spec.html, because the latter looks like a path with
2479 a MIME-type, and this confuses some software.
2481 PH/70 Catch two compiler warnings in sieve.c.
2483 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2484 function verify_get_ident() calls ip_connect() to connect a socket, but
2485 if the "connect()" function timed out, ip_connect() used to close the
2486 socket. However, verify_get_ident() also closes the socket later, and in
2487 between Exim writes to the log, which may get opened at this point. When
2488 the socket was closed in ip_connect(), the log could get the same file
2489 descriptor number as the socket. This naturally causes chaos. The fix is
2490 not to close the socket in ip_connect(); the socket should be closed by
2491 the function that creates it. There was only one place in the code where
2492 this was missing, in the iplookup router, which I don't think anybody now
2493 uses, but I've fixed it anyway.
2495 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2496 well as to direct DNS lookups. Otherwise the handling of names in host
2497 lists is inconsistent and therefore confusing.
2503 PH/01 Two changes to the default runtime configuration:
2505 (1) Move the checks for relay_from_hosts and authenticated clients from
2506 after to before the (commented out) DNS black list checks.
2508 (2) Add control=submission to the relay_from_hosts and authenticated
2509 clients checks, on the grounds that messages accepted by these
2510 statements are most likely to be submissions.
2512 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2514 (1) Generate an error if the third argument for the ${prvs expansion is
2517 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2520 (3) Reset the variables that are obtained from the first argument of
2521 ${prvscheck and used in the second argument before leaving the code,
2522 because their memory is reclaimed, so using them afterwards may do
2525 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2526 one (it's much easier than Tom thought :-).
2528 (5) Because of (4), we can now allow for the use of $prvscheck_result
2529 inside the third argument.
2531 PH/03 For some reason, the default setting of PATH when running a command from
2532 a pipe transport was just "/usr/bin". I have changed it to
2535 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2536 anything to be listed in the output from -bV.
2538 PH/05 When a filter generated an autoreply, the entire To: header line was
2539 quoted in the delivery log line, like this:
2541 => >A.N.Other <ano@some.domain> <original@ddress> ...
2543 This has been changed so that it extracts the operative address. There
2544 may be more than one such address. If so, they are comma-separated, like
2547 => >ano@some.domain,ona@other.domain <original@ddress> ...
2549 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2550 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2551 not being looked up in the rDNS to get a host name, Exim was showing the
2552 IP address twice in Received: lines, even though the IP addresses were
2553 identical. For example:
2555 Received: from [1.2.3.4] (helo=[1.2.3.4])
2557 However, if the real host name was known, it was omitting the HELO data
2558 if it matched the actual IP address. This has been tidied up so that it
2559 doesn't show the same IP address twice.
2561 PH/07 When both +timestamp and +memory debugging was on, the value given by
2562 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2563 called by the debug printing, thereby overwriting the timestamp buffer.
2564 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2566 PH/08 When the original message was included in an autoreply transport, it
2567 always said "this is a copy of the message, including all the headers",
2568 even if body_only or headers_only was set. It now gives an appropriate
2571 PH/09 Applied a patch from the Sieve maintainer which:
2573 o fixes some comments
2574 o adds the (disabled) notify extension core
2575 o adds some debug output for the result of if/elsif tests
2576 o points to the current vacation draft in the documentation
2577 and documents the missing references header update
2581 o fixes a bug in processing the envelope test (when testing
2582 multiple envelope elements, the last element determinted the
2585 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2586 Electronic Mail") by including:
2588 Auto-submitted: auto-generated
2590 in the messages that it generates (bounce messages and others, such as
2591 warnings). In the case of bounce messages for non-SMTP mesages, there was
2592 also a typo: it was using "Auto_submitted" (underscore instead of
2593 hyphen). Since every message generated by Exim is necessarily in response
2594 to another message, thes have all been changed to:
2596 Auto-Submitted: auto-replied
2598 in accordance with these statements in the RFC:
2600 The auto-replied keyword:
2602 - SHOULD be used on messages sent in direct response to another
2603 message by an automatic process,
2605 - MUST NOT be used on manually-generated messages,
2607 - MAY be used on Delivery Status Notifications (DSNs) and Message
2608 Disposition Notifications (MDNs),
2610 - MUST NOT be used on messages generated by automatic or periodic
2611 processes, except for messages which are automatic responses to
2614 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2615 to the default Received: header definition.
2617 PH/12 Added log selector acl_warn_skipped (default on).
2619 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2620 variables because (a) they are in the wrong storage pool and (b) even if
2621 they were copied, it wouldn't work properly because of the caching.
2623 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2624 checking when decoding. Apparently there are clients that generate
2625 overlong encoded strings. Why am I not surprised?
2627 PH/15 If the first argument of "${if match_address" was not empty, but did not
2628 contain an "@" character, Exim crashed. Now it writes a panic log message
2629 and treats the condition as false.
2631 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2633 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2634 extension "envelope-auth". The code is finished and in agreement with
2635 other implementations, but there is no documentation so far and in fact,
2636 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2637 not changing the active code.
2639 Print executed "if" and "elsif" statements when debugging is used. This
2640 helps a great deal to understand what a filter does.
2642 Document more things not specified clearly in RFC3028. I had all this
2643 sorted out, when out of a sudden new issues came to my mind. Oops."
2645 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2648 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2649 canonical form (as documented). However, after a host name lookup from
2650 the IP address, check_host() was doing a simple string comparison with
2651 addresses acquired from the DNS when checking that the found name did
2652 have the original IP as one of its addresses. Since any found IPv6
2653 addresses are likely to be in abbreviated form, the comparison could
2654 fail. Luckily, there already exists a function for doing the comparison
2655 by converting both addresses to binary, so now that is used instead of
2656 the text comparison.
2658 PH/20 There was another similar case to PH/19, when a complete host name was
2659 given in a host list; looking up its IP address could give an abbreviated
2660 form, whereas the current host's name might or might not be abbreviated.
2661 The same fix has been applied.
2667 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2668 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2671 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2672 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2674 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2676 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2677 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2678 Also, for "Auto-Submitted", treat anything other than "no" as
2679 non-personal, in accordance with RFC 3834. (Previously it treated
2680 anything starting "auto-" as non-personal.)
2682 TF/01 The control=submission/name=... option had a problem with syntax
2683 errors if the name included a slash character. The /name= option
2684 now slurps the rest of the string, so it can include any characters
2685 but it must come last in the list of options (after /sender_retain
2688 PH/05 Some modifications to the interface to the fake nameserver for the new
2696 TK/01 Added the "success_on_redirect" address verification option. See
2697 NewStuff for rationale and an example.
2699 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2701 PH/02 Patch to exigrep to allow it to work on syslog lines.
2703 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2704 fread() to skip over the body file's header line, because in Cygwin the
2705 header line is locked and is inaccessible.
2707 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2708 co-exist for some time) to make it clear that it is the Exim ID that is
2709 referenced, not the Message-ID: header line.
2711 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2712 string_format() function, because snprintf() does not exist on all
2715 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2716 ${stat: expansion item.
2718 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2719 protocol synchronization error", to keep the pedants happy.
2721 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2722 well as for IRIX systems, when gcc is being used. See the host.c source
2725 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2727 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2730 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2731 transport and to the smtp transport in LMTP mode.
2733 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2735 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2736 run a filter in a subprocess. This could lead to confusion in subsequent
2737 lookups in the parent process. There should also be a search_tidyup() at
2738 the end of the subprocess.
2740 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2741 only if the host matched helo_try_verify_hosts, which caused the
2742 verification to occur when the EHLO/HELO command was issued. The ACL just
2743 tested the remembered result. Now, if a previous verification attempt has
2744 not happened, "verify = helo" does it there and then.
2746 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2748 TK/03 Fix log output including CR from clamd.
2750 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2751 exist provoked a memory error which could cause a segfault.
2753 PH/15 Installed PCRE 6.2
2755 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2757 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2758 of the problem. Specifically, suggested +O2 rather than +O1 for the
2761 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2763 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2764 record, the port was not correctly logged when the outgoing_port log
2765 selector was set (it logged the transort's default port).
2767 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2768 fallback_hosts, and "hosts" in the smtp transport.
2770 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2771 host errors such as "Connection refused".
2773 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2774 authentication with radiusclient 0.4.9:
2776 - Error returned from rc_read_config was caught wrongly
2777 - Username/password not passed on to radius server due to wrong length.
2779 The presumption is that some radiusclient API changes for 4.51/PH/17
2780 were not taken care of correctly. The code is still untested by me (my
2781 Linux distribution still has 0.3.2 of radiusclient), but it was
2782 contributed by a Radius user.
2784 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2785 expanding the "port" option of the smtp transport.
2787 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2788 while reading a MIME header. Thanks to Tom Hughes for a patch.
2790 PH/24 Include config.h inside local_scan.h so that configuration settings are
2793 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2794 that in an ACL for RCPT (for example), you can examine exactly what was
2797 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2798 commands, but it was not correctly comparing the address with the actual
2799 client host address. Thus, it would show the EHLO address in Received:
2800 header lines when this was not necessary.
2802 PH/27 Added the % operator to ${eval:}.
2804 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2805 it should be ignoring failures (because with -C, for example, it has lost
2806 privilege). It wasn't ignoring creation failures other than "already
2809 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2812 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2813 because an input line was too long, either on its own, or by virtue of
2814 too many continuations, the temporary file was not being removed, and the
2815 return code was incorrect.
2817 PH/31 Missing "BOOL" in function definition in filtertest.c.
2819 PH/32 Applied Sieve patches from the maintainer.
2821 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2823 PH/33 Added "verify = not_blind".
2825 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2826 Local/Makefile (with some defaults set). These are used in built scripts
2827 such as exicyclog, but they have never been used in the exim_install
2828 script (though there are many overriding facilities there). I have
2829 arranged that the exim_install script now takes note of these two
2832 PH/35 Installed configuration files for Dragonfly.
2834 PH/36 When a locally submitted message by a trusted user did not contain a
2835 From: header, and the sender address was obtained from -f or from an SMTP
2836 MAIL command, and the trusted user did not use -F to supply a sender
2837 name, $originator_name was incorrectly used when constructing a From:
2838 header. Furthermore, $originator_name was used for submission mode
2839 messages from external hosts without From: headers in a similar way,
2840 which is clearly wrong.
2842 PH/37 Added control=suppress_local_fixups.
2844 PH/38 When log_selector = +received_sender was set, and the addition of the
2845 sender made the log line's construction buffer exactly full, or one byte
2846 less than full, an overflow happened when the terminating "\n" was
2849 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2850 when the result of a list match is failure because a DNS lookup failed.
2852 PH/40 RM_COMMAND is now used in the building process.
2854 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2855 the "build-* directories that it finds.
2857 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2858 address in a domain literal was a prefix of an interface address.
2860 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2861 when verifying a sender address, unless rewrite_headers is false.
2863 PH/44 Wrote a long comment about why errors_to addresses are verified as
2864 recipients, not senders.
2866 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2867 the ratelimit ACL was added.
2869 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2871 PH/46 Added extra information about PostgreSQL errors to the error string.
2873 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2874 avoiding the need to install special zones in a real server. This is
2875 backwards compatible; if it can't find the fake resolver, it drops back.
2876 Thus, both old and new test suites can be run.
2878 TF/02 Added util/ratelimit.pl
2880 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2881 clock is set back in time.
2883 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2884 Candler <B.Candler@pobox.com>.
2886 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2887 OK for addresses that are the result of redirections.
2889 PH/48 A number of further additions for the benefit of the new test suite,
2890 including a fake gethostbyname() that interfaces to the fake DNS resolver
2893 TF/06 The fix for widen_domains has also been applied to qualify_single and
2894 search_parents which are the other dnslookup options that can cause
2897 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2900 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2902 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2903 service" if it encountered an MX record with an empty target hostname.
2904 The message is now "an MX or SRV record indicated no SMTP service".
2906 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2907 if the DNS of the sending site is misconfigured. This is quite a
2908 common situation. This change restores the behaviour of treating a
2909 helo verification defer as a failure.
2911 PH/51 If self=fail was set on a router, the bounce message did not include the
2912 actual error message.
2918 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2920 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2921 command itself ended in error, the underlying message about the transport
2922 filter timeout was being overwritten with the pipe command error. Now the
2923 underlying error message should be appended to the second error message.
2925 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2927 PH/02 Reduce the amount of output that "make" produces by default. Full output
2928 can still be requested.
2930 PH/03 The warning log line about a condition test deferring for a "warn" verb
2931 was being output only once per connection, rather than after each
2932 occurrence (because it was using the same function as for successful
2933 "warn" verbs). This seems wrong, so I have changed it.
2935 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2936 it should not have, which might have caused a crash in the right
2937 circumstances, but probably never did.
2939 PH/04 Installed a modified version of Tony Finch's patch to make submission
2940 mode fix the return path as well as the Sender: header line, and to
2941 add a /name= option so that you can make the user's friendly name appear
2944 TF/03 Added the control = fakedefer ACL modifier.
2946 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2947 Mark Lowes for thorough testing.
2949 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2951 TK/03 Merged latest SRS patch from Miles Wilton.
2953 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2954 with the definition in sysexits.h (which is #included earlier).
2955 Fortunately, Exim does not actually use EX_OK. The code used to try to
2956 preserve the sysexits.h value, by assumimg that macro definitions were
2957 scanned for macro replacements. I have been disabused of this notion,
2958 so now the code just undefines EX_OK before #including unistd.h.
2960 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2961 in the smtp transport. When a block could not be written in a single
2962 write() function, the timeout was being re-applied to each part-write.
2963 This seems wrong - if the receiver was accepting one byte at a time it
2964 would take for ever. The timeout is now adjusted when this happens. It
2965 doesn't have to be particularly precise.
2967 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2968 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2970 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2971 without a domain if the check to <postmaster@domain> fails.
2973 SC/01 Eximstats: added -xls and the ability to specify output files
2974 (patch written by Frank Heydlauf).
2976 SC/02 Eximstats: use FileHandles for outputing results.
2978 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2980 SC/04 Eximstats: fixed display of large numbers with -nvr option
2982 SC/05 Eximstats: fixed merging of reports with empty tables.
2984 SC/06 Eximstats: added the -include_original_destination flag
2986 SC/07 Eximstats: removed tabs and trailing whitespace.
2988 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2990 TK/06 MBOX spool code: Add real "From " MBOX separator line
2991 so the .eml file is really in mbox format (even though
2992 most programs do not really care). Patch from Alex Miller.
2994 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2995 The latter is generated from $received_to and is only set if the
2996 message has one envelope recipient. SA can use these headers,
2997 obviously out-of-the-box. Patch from Alex Miller.
2999 PH/08 The ${def test on a variable was returning false if the variable's
3000 value was "0", contrary to what the specification has always said!
3001 The result should be true unless the variable is empty.
3003 PH/09 The syntax error of a character other than { following "${if
3004 def:variable_name" (after optional whitespace) was not being diagnosed.
3005 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
3006 accidental colon was present, for example, could give incorrect results.
3008 PH/10 Tidied the code in a number of places where the st_size field of a stat()
3009 result is used (not including appendfile, where other changes are about
3012 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
3013 This involved changing a lot of size variables from int to off_t. It
3014 should work with maildirs and everything.
3016 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
3017 spamd dying while we are connected to it.
3019 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
3022 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
3023 function for BDB 4.3.
3025 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
3027 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
3028 the sources for PCRE have been re-arranged and more files are now
3031 PH/15 The code I had for printing potentially long long variables in PH/11
3032 above was not the best (it lost precision). The length of off_t variables
3033 is now inspected at build time, and an appropriate printing format (%ld
3034 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
3035 to be "long long int" or "long int". This is needed for the internal
3036 formatting function string_vformat().
3038 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
3039 the configuration file to be ":syslog", then the script "guesses" where
3040 the logs files are, rather than using the compiled in default. In our
3041 case the guess is not the same as the compiled default, so the script
3042 suddenly stopped working when I started to use syslog. The patch checks
3043 to see if log_file_path is "". If so, it attempts to read it from exim
3044 with no configuration file to get the compiled in version, before it
3045 falls back to the previous guessing code."
3047 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3048 implementing BATV in an Exim configuration. See NewStuff for the gory
3051 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3052 Makefile that are specific to HP-UX.
3054 PH/18 If the "use_postmaster" option was set for a recipient callout together
3055 with the "random" option, the postmaster address was used as the MAIL
3056 FROM address for the random test, but not for the subsequent recipient
3057 test. It is now used for both.
3059 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3060 patch removes a few documentation additions to RFC 3028, because the
3061 latest draft now contains them. It adds the new en;ascii-case comparator
3062 and a new error check for 8bit text in MIME parts. Comparator and
3063 require names are now matched exactly. I enabled the subaddress
3064 extension, but it is not well tested yet (read: it works for me)."
3066 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3067 rework some of the code of TK/09 above to avoid the hardwired use of
3068 "%lld" and "long long". Replaced the call to snprintf() with a call to
3071 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3072 records point to non-existent hosts", "retry timeout exceeded", and
3073 "retry time not reached for any host after a long failure period".
3075 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3076 experimental DomainKeys support:
3078 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3079 (2) On an error such as an illegally used "control", the wrong name for
3080 the control was given.
3082 These problems did NOT occur unless DomainKeys support was compiled.
3084 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3086 PH/24 Added ${if match_ip condition.
3088 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3089 databases so that it will be absolutely obvious if a crash occurs in the
3090 DB library. This is a regular occurrence (often caused by mis-matched
3093 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3094 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3095 picked up on a user's system that detects such things. There doesn't seem
3096 to be a gcc warning option for this - only an attribute that has to be
3097 put on the function's prototype. It seems that in Fedora Core 4 they have
3098 set this on a number of new functions. No doubt there will be more in due
3101 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3102 specify a transport. However, if an address that was verified by such a
3103 router was the subject of a callout, Exim crashed because it tried to
3104 read the rcpt_include_affixes from the non-existent transport. Now it
3105 just assumes that the setting of that option is false. This bug was
3106 introduced by 4.51/PH/31.
3108 PH/28 Changed -d+all to exclude +memory, because that information is very
3109 rarely of interest, but it makes the output a lot bigger. People tend to
3110 do -d+all out of habit.
3112 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3113 code in os-type was giving problems when libc.so lives in lib64, like on
3116 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3117 aren't the modern standard, and it seems that some systems' include files
3118 don't always have them. Exim was already checking for some of the newer
3119 ones like T_AAAA, and defining it itself. I've added checks for all the
3120 record types that Exim uses.
3122 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3123 not automatically generating a new one, as it is supposed to. This
3124 prevented TLS from working. If the file did exist, but contained invalid
3125 data, a new version was generated, as expected. It was only the case of a
3126 non-existent file that was broken.
3128 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3129 with a change in libdomainkeys > 0.64.
3131 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3132 from DNS. If the selector record carries the flag, it now has
3133 precedence over the domain-wide flag.
3135 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3137 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3138 the use of a port name that isn't defined in /etc/services) occurred, the
3139 message was deferred as in a normal delivery, and thus remained on the
3140 spool, instead of being failed because of the mua_wrapper setting. This
3141 is now fixed, and I tidied up some of the mua_wrapper messages at the
3144 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3145 the messages in a hash of arrays rather than using individual hashes.
3146 This is a bit cleaner and results in dramatic memory savings, albeit
3147 at a slight CPU cost.
3149 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3150 as requested by Marc Sherman.
3152 SC/10 Eximstats: added histograms for user specified patterns as requested
3155 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3157 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3158 fopen() in the content-scanning modules that did not already have it.
3164 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3165 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3167 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3169 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3171 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3172 to test Sieve filters that use "vacation".
3174 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3175 that changes the way the GnuTLS parameters are stored in the cache file.
3176 The new format can be generated externally. For backward compatibility,
3177 if the data in the cache doesn't make sense, Exim assumes it has read an
3178 old-format file, and it generates new data and writes a new file. This
3179 means that you can't go back to an older release without removing the
3182 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3183 work if there are any delivery delays because "one_time" forces the
3184 parent to be marked "delivered", so its unseen clone is never tried
3185 again. For this reason, Exim now forbids the simultaneous setting of
3188 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3189 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3190 entry if you want to know the details. The fix, however, neglected to
3191 consider the case when local delivery batching is involved. The test for
3192 "previously delivered" was not happening when checking to see if an
3193 address could be batched with a previous (undelivered) one; under
3194 certain circumstances this could lead to multiple deliveries to the same
3197 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3198 in its include files, and this causes problems building Exim.
3200 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3201 header_syntax) but Exim was just ignoring anything given after a slash.
3202 In particular, this caused confusion with an attempt to use "verify =
3203 reverse_host_lookup/defer_ok". An error is now given when options are
3204 supplied for verify items that do not have them. (Maybe reverse_host_
3205 lookup should have a defer_ok option, but that's a different point.)
3207 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3208 defined by RFC 821) to 2048, because there were problems with some AUTH
3209 commands, and RFC 1869 says the size should be increased for extended
3210 SMTP commands that take arguments.
3212 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3215 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3216 "unknown" error; now it says that the functionality isn't in the binary.
3218 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3219 an address' error message when a string expansion fails (syntax or
3220 whatever). Otherwise the password may appear in the log. Following change
3221 PH/42 below, there is no longer a chance of it appearing in a bounce
3224 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3226 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3227 of its addresses was ever tried. (Bugzilla bug #2.)
3229 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3230 the result incorrectly in the debug output. (It correctly added a newline
3231 to what was transported.)
3233 TF/01 Added $received_time.
3235 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3236 commented out examples of how to interface to a virus scanner and to
3237 SpamAssassin. Also added commented examples of av_scanner and
3238 spamd_address settings.
3240 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3241 and controls are allowed in which ACLs. There were a couple of minor
3242 errors. Some of the entries in the conditions table (which is a table of
3243 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3244 negation of where the condition IS allowed.
3246 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3248 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3249 header file does not have a version number, so I've had to invent a new
3250 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3251 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3252 radiusclient), but it was contributed by a Radius user.
3254 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3255 files or directories, for OpenSSL.
3257 PH/19 When an Exim process that is running as root has to create an Exim log
3258 file, it does so in a subprocess that runs as exim:exim so as to get the
3259 ownership right at creation (otherwise, other Exim processes might see
3260 the file with the wrong ownership). There was no test for failure of this
3261 fork() call, which would lead to the process getting stuck as it waited
3262 for a non-existent subprocess. Forks do occasionally fail when resources
3263 run out. I reviewed all the other calls to fork(); they all seem to check
3266 PH/20 When checking for unexpected SMTP input at connect time (before writing
3267 the banner), Exim was not dealing correctly with a non-positive return
3268 from the read() function. If the client had disconnected by this time,
3269 the result was a log entry for a synchronization error with an empty
3270 string after "input=" when read() returned zero. If read() returned -1
3271 (an event I could not check), uninitialized data bytes were printed.
3272 There were reports of junk text (parts of files, etc) appearing after
3275 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3277 PH/22 Added support for macro redefinition, and (re)definition in between
3278 driver and ACL definitions.
3280 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3281 forgetting to use the resulting value; it was using the unexpanded value.
3283 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3284 hadn't been configured. The fix is from Juergen Kreileder, who
3285 understands it better than I do:
3287 "Here's what I see happening with three configured cyrus_sasl
3288 authenticators configured (plain, login, cram-md5):
3290 On startup auth_cyrus_sasl_init() gets called for each of these.
3291 This means three calls to sasl_listmech() without a specified mech_list.
3292 => SASL tests which mechs of all available mechs actually work
3293 => three warnings about OTP not working
3294 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3296 With the patch, sasl_listmech() also gets called three times. But now
3297 SASL's mech_list option is set to the server_mech specified in the the
3298 authenticator. Or in other words, the answer from sasl_listmech()
3299 gets limited to just the mech you're testing for (which is different
3301 => the return list contains just 'plain' or 'login', 'cram-md5' or
3302 nothing depending on the value of ob->server_mech.
3304 I've just tested the patch: Authentication still works fine,
3305 unavailable mechs specified in the exim configuration are still
3306 caught, and the auth.log warnings about OTP are gone."
3308 PH/25 When debugging is enabled, the contents of the command line are added
3309 to the debugging output, even when log_selector=+arguments is not
3312 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3313 answer is "GNU", and only if the return is "GNU/something" is the answer
3316 PH/27 $acl_verify_message is now set immediately after the failure of a
3317 verification in an ACL, and so is available in subsequent modifiers. In
3318 particular, the message can be preserved by coding like this:
3320 warn !verify = sender
3321 set acl_m0 = $acl_verify_message
3323 Previously, $acl_verify_message was set only while expanding "message"
3324 and "log_message" when a very denied access.
3326 PH/28 Modified OS/os.c-Linux with
3328 -#ifndef OS_LOAD_AVERAGE
3329 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3331 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3332 nomenclature these days.)
3334 PH/29 Installed patch from the Sieve maintainer that adds the options
3335 sieve_useraddress and sieve_subaddress to the redirect router.
3337 PH/30 In these circumstances:
3338 . Two addresses routed to the same list of hosts;
3339 . First host does not offer TLS;
3340 . First host accepts first address;
3341 . First host gives temporary error to second address;
3342 . Second host offers TLS and a TLS session is established;
3343 . Second host accepts second address.
3344 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3345 and peerdn, if requested) that were in fact used only for the second
3348 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3349 attention to any local part prefix or suffix that was matched by the
3350 router that accepted the address. It now behaves in the same way as it
3351 does for delivery: the affixes are removed from the local part unless
3352 rcpt_include_affixes is set on the transport.
3354 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3355 timeout during the DATA phase of an incoming message.
3357 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3358 applied a patch sanctioned by the Sieve maintainer.
3360 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3361 the uid or gid is negative. A case of a negative gid caused this to be
3362 noticed. The fix allows for either to be negative.
3364 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3365 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3368 PH/36 The change PH/12 above was broken. Fixed it.
3370 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3371 the grounds that routing the same address twice would always produce the
3372 same answer. This might have been true once, but it is certainly no
3373 longer true now. Routing a child address may depend on the previous
3374 routing that produced that child. Some complicated redirection strategies
3375 went wrong when messages had multiple recipients, and made Exim's
3376 behaviour dependent on the order in which the addresses were given.
3378 I have moved the duplicate checking until after the routing is complete.
3379 Exim scans the addresses that are assigned to local and remote
3380 transports, and removes any duplicates. This means that more work will be
3381 done, as duplicates will always all be routed, but duplicates are
3382 presumably rare, so I don't expect this is of any significance.
3384 For deliveries to pipes, files, and autoreplies, the duplicate checking
3385 still happens during the routing process, since they are not going to be
3388 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3389 It corrects a timeout issue with spamd. This is Ian's comment: "The
3390 background is that sometimes spamd either never reads data from a
3391 connection it has accepted, or it never writes response data. The exiscan
3392 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3393 blindly assumes that writes won't block so it may never time out."
3395 PH/39 Allow G after quota size as well as K and M.
3397 PH/40 The value set for $authenticated_id in an authenticator may not contain
3398 binary zeroes or newlines because the value is written to log lines and
3399 to spool files. There was no check on this. Now the value is run through
3400 the string_printing() function so that such characters are converted to
3401 printable escape sequences.
3403 PH/41 $message_linecount is a new variable that contains the total number of
3404 lines in the message. Compare $body_linecount, which is the count for the
3407 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3408 bounce and delay warning messages, except in certain special cases, which
3411 (a) An SMTP error message from a remote host;
3412 (b) A message specified in a :fail: redirection;
3413 (c) A message specified in a "fail" command in a system filter;
3414 (d) A message specified in a FAIL return from the queryprogram router;
3415 (e) A message specified by the cannot_route_message router option.
3417 In these cases only, Exim does include the error details in bounce and
3418 warning messages. There are also a few cases where bland messages such
3419 as "unrouteable address" or "local delivery error" are given.
3421 PH/43 $value is now also set for the "else" part of a ${run expansion.
3423 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3424 being worked on, but at least Exim now implements the latest version to
3427 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3428 process to complete was treated as a delivery failure, a timeout while
3429 writing the message to the pipe was logged, but erroneously treated as a
3430 successful delivery. Such timeouts include transport filter timeouts. For
3431 consistency with the overall process timeout, these timeouts are now
3432 treated as errors, giving rise to delivery failures by default. However,
3433 there is now a new Boolean option for the pipe transport called
3434 timeout_defer, which, if set TRUE, converts the failures into defers for
3435 both kinds of timeout. A transport filter timeout is now identified in
3438 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3439 systems where "make" and "gmake" are different, calling "gmake" at top
3440 level broke things. I've arranged for the value of $(MAKE) to be passed
3441 from the Makefile to this script so that it can call the same version of
3445 A note about Exim versions 4.44 and 4.50
3446 ----------------------------------------
3448 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3449 changes of various kinds. As a consequence, a big documentation update was
3450 needed. This delayed the release for rather longer than seemed good, especially
3451 in the light of a couple of (minor) security issues. Therefore, the changes
3452 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3453 release. So 4.44 and 4.50 are in effect two different branches that both start
3456 I have left the 4.50 change log unchanged; it contains all the changes since
3457 4.43. The change log for 4.44 is below; many of its items are identical to
3458 those for 4.50. This seems to be the most sensible way to preserve the
3459 historical information.
3465 1. Minor wording change to the doc/README.SIEVE file.
3467 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3468 computation of the current number of files was incorrect.
3470 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3471 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3474 4. Give more explanation in the error message when the command for a transport
3475 filter fails to execute.
3477 5. There are several places where Exim runs a non-Exim command in a
3478 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3479 was being done only for the command run by the queryprogram router. It is
3480 now done for all such subprocesses. The other cases are: ${run, transport
3481 filters, and the commands run by the lmtp and pipe transports.
3483 6. Added CONFIGURE_GROUP build-time option.
3485 7. Some older OS have a limit of 256 on the maximum number of file
3486 descriptors. Exim was using setrlimit() to set 1000 as a large value
3487 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3488 systems. I've change it so that if it can't get 1000, it tries for 256.
3490 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3491 was an oversight, and furthermore, ever since the addition of extra
3492 controls (e.g. 4.43/32), the checks on when to allow different forms of
3493 "control" were broken. There should now be diagnostics for all cases when a
3494 control that does not make sense is encountered.
3496 9. Added the /retain_sender option to "control=submission".
3498 10. $recipients is now available in the predata ACL (oversight).
3500 11. Tidy the search cache before the fork to do a delivery from a message
3501 received from the command line. Otherwise the child will trigger a lookup
3502 failure and thereby defer the delivery if it tries to use (for example) a
3503 cached ldap connection that the parent has called unbind on.
3505 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3506 of $address_data from the recipient verification was clobbered by the
3507 sender verification.
3509 13. The value of address_data from a sender verification is now available in
3510 $sender_address_data in subsequent conditions in the ACL statement.
3512 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3514 15. Added a new option "connect=<time>" to callout options, to set a different
3517 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3518 was its contents. (It was OK if the option was not defined at all.)
3520 17. A "Completed" log line is now written for messages that are removed from
3521 the spool by the -Mrm option.
3523 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3524 information about exactly what failed.
3526 19. Added -dd to debug only the daemon process.
3528 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3529 handles timeouts, both on the server side and network timeouts. Renamed the
3530 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3532 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3533 It is now set to "smtps".
3535 22. $host_address is now set to the target address during the checking of
3536 ignore_target_hosts.
3538 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3539 being passed; this would have caused $sender_host_name to have been used if
3540 matching the list had actually called for a host name (not very likely,
3541 since this list is usually IP addresses). A host name is now passed as
3544 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3545 code that specifies a non-restarting handler (typically sigaction() in
3546 modern systems) in an attempt to fix a rare and obscure crash bug.
3548 25. Narrowed the window for a race in the daemon that could cause it to ignore
3549 SIGCHLD signals. This is not a major problem, because they are used only to
3550 wake it up if nothing else does.
3552 26. A malformed maildirsize file could cause Exim to calculate negative values
3553 for the mailbox size or file count. Odd effects could occur as a result.
3554 The maildirsize information is now recalculated if the size or filecount
3557 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3558 support for a long time. Removed HAVE_SYS_VFS_H.
3560 28. Installed the latest version of exipick from John Jetmore.
3562 29. In an address list, if the pattern was not a regular expression, an empty
3563 subject address (from a bounce message) matched only if the pattern was an
3564 empty string. Non-empty patterns were not even tested. This was the wrong
3565 because it is perfectly reasonable to use an empty address as part of a
3566 database query. An empty address is now tested by patterns that are
3567 lookups. However, all the other forms of pattern expect the subject to
3568 contain a local part and a domain, and therefore, for them, an empty
3569 address still always fails if the pattern is not itself empty.
3571 30. Exim went into a mad DNS loop when attempting to do a callout where the
3572 host was specified on an smtp transport, and looking it up yielded more
3573 than one IP address.
3575 31. Re-factored the code for checking spool and log partition space into a
3576 function that finds that data and another that does the check. The former
3577 is then used to implement four new variables: $spool_space, $log_space,
3578 $spool_inodes, and $log_inodes.
3580 32. The RFC2047 encoding function was originally intended for short strings
3581 such as real names; it was not keeping to the 75-character limit for
3582 encoded words that the RFC imposes. It now respects the limit, and
3583 generates multiple encoded words if necessary. To be on the safe side, I
3584 have increased the buffer size for the ${rfc2047: expansion operator from
3587 33. It is now permitted to omit both strings after an "if" condition; if the
3588 condition is true, the result is "true". As before, when the second string
3589 is omitted, a false condition yields an empty string. This makes it less
3590 cumbersome to write custom ACL and router conditions.
3592 34. Failure to deliver a bounce message always caused it to be frozen, even if
3593 there was an errors_to setting on the router. The errors_to setting is now
3596 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3597 canonical form (fully expanded) before being placed in
3598 $sender_host_address.
3600 36. The table in the code that translates DNS record types into text (T_A to
3601 "A" for instance) was missing entries for NS and CNAME. It is just possible
3602 that this could have caused confusion if both these types were looked up
3603 for the same domain, because the text type is used as part of Exim's
3604 per-process caching. But the chance of anyone hitting this buglet seems
3607 37. The dnsdb lookup has been extended in a number of ways.
3609 (1) There is a new type, "zns", which walks up the domain tree until it
3610 finds some nameserver records. It should be used with care.
3612 (2) There is a new type, "mxh", which is like "mx" except that it returns
3613 just the host names, not the priorities.
3615 (3) It is now possible to give a list of domains (or IP addresses) to be
3616 looked up. The behaviour when one of the lookups defers can be
3617 controlled by a keyword.
3619 (4) It is now possible to specify the separator character for use when
3620 multiple records are returned.
3622 38. The dnslists ACL condition has been extended: it is now possible to supply
3623 a list of IP addresses and/or domains to be looked up in a particular DNS
3626 39. Added log_selector=+queue_time_overall.
3628 40. When running the queue in the test harness, wait just a tad after forking a
3629 delivery process, to get repeatability of debugging output.
3631 41. Include certificate and key file names in error message when GnuTLS fails
3632 to set them up, because the GnuTLS error message doesn't include the name
3633 of the failing file when there is a problem reading it.
3635 42. Allow both -bf and -bF in the same test run.
3637 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3639 44. The "Exiscan patch" is now merged into the mainline Exim source.
3641 45. Sometimes the final signoff response after QUIT could fail to get
3642 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3643 < 0 before doing a fflush(). This bug looks as though it goes back to the
3644 introduction of TLS in release 3.20, but "sometimes" must have been rare
3645 because the tests only now provoked it.
3647 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3648 (this can affect the format of dates).
3650 47. exim_tidydb, when checking for the continued existence of a message for
3651 which it has found a message-specific retry record, was not finding
3652 messages that were in split spool directories. Consequently, it was
3653 deleting retry records that should have stayed in existence.
3655 48. Steve fixed some bugs in eximstats.
3657 49. The SPA authentication driver was not abandoning authentication and moving
3658 on to the next authenticator when an expansion was forced to fail,
3659 contradicting the general specification for all authenticators. Instead it
3660 was generating a temporary error. It now behaves as specified.
3662 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3663 (the order specifies the preference for clients). The order is now AES256,
3664 AES128, 3DES, ARCFOUR128.
3666 51. Small patch to Sieve code - explicitly set From: when generating an
3669 52. Exim crashed if a remote delivery caused a very long error message to be
3670 recorded - for instance if somebody sent an entire SpamAssassin report back
3671 as a large number of 550 error lines. This bug was coincidentally fixed by
3672 increasing the size of one of Exim's internal buffers (big_buffer) that
3673 happened as part of the Exiscan merge. However, to be on the safe side, I
3674 have made the code more robust (and fixed the comments that describe what
3677 53. Now that there can be additional text after "Completed" in log lines (if
3678 the queue_time_overall log selector is set), a one-byte patch to exigrep
3679 was needed to allow it to recognize "Completed" as not the last thing in
3682 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3683 patch that reportedly fixes this has been added. I am not expert enough to
3684 create a test for it. This is what the patch creator wrote:
3686 "I found a little strange behaviour of ldap code when working with
3687 Windows 2003 AD Domain, where users was placed in more than one
3688 Organization Units. When I tried to give exim partial DN, the exit code
3689 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3690 But simultaneously result of request was absolutely normal ldap result,
3691 so I produce this patch..."
3693 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3694 REFERENCE, so I have modified the code to exclude the patch when that macro
3697 55. Some experimental protocols are using DNS PTR records for new purposes. The
3698 keys for these records are domain names, not reversed IP addresses. The
3699 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3700 leaves it alone. Component reversal etc. now happens only for IP addresses.
3703 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3705 57. Double the size of the debug message buffer (to 2048) so that more of very
3706 long debug lines gets shown.
3708 58. The exicyclog utility now does better if the number of log files to keep
3709 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3711 59. Two changes related to the smtp_active_hostname option:
3713 (1) $smtp_active_hostname is now available as a variable.
3714 (2) The default for smtp_banner uses $smtp_active_hostname instead
3715 of $primary_hostname.
3717 60. The host_aton() function is supposed to be passed a string that is known
3718 to be a valid IP address. However, in the case of IPv6 addresses, it was
3719 not checking this. This is a hostage to fortune. Exim now panics and dies
3720 if the condition is not met. A case was found where this could be provoked
3721 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3722 components; fortuitously, this particular loophole had already been fixed
3723 by change 4.50/55 above.
3725 If there are any other similar loopholes, the new check in host_aton()
3726 itself should stop them being exploited. The report I received stated that
3727 data on the command line could provoke the exploit when Exim was running as
3728 exim, but did not say which command line option was involved. All I could
3729 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3730 running as the user.
3733 61. There was a buffer overflow vulnerability in the SPA authentication code
3734 (which came originally from the Samba project). I have added a test to the
3735 spa_base64_to_bits() function which I hope fixes it.
3738 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3739 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3741 63. The daemon start-up calls getloadavg() while still root for those OS that
3742 need the first call to be done as root, but it missed one case: when
3743 deliver_queue_load_max is set with deliver_drop_privilege. This is
3744 necessary for the benefit of the queue runner, because there is no re-exec
3745 when deliver_drop_privilege is set.
3747 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3748 This has been fixed.
3750 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3751 was in use, was not putting the data itself into the right store pool;
3752 consequently, it could be overwritten for a subsequent message in the same
3753 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3756 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3758 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3759 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3760 the function were treating the return as a boolean value, which happened to
3761 work because 0=false and not-0=true, but is not correct code.
3763 68. The host_aton() function was not handling scoped IPv6 addresses (those
3764 with, for example, "%eth0" on the end) correctly.
3766 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3767 negated items (that is, ~something) in unsigned ints. Some compilers
3768 apparently mutter when there is no cast.
3770 70. If an address verification called from an ACL failed, and did not produce a
3771 user-specific message (i.e. there was only a "system" message), nothing was
3772 put in $acl_verify_message. In this situation, it now puts the system
3775 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3776 session; change 4.31/43 added the unwanted input to the log line - except
3777 that it did not do this in the start of session case. It now does.
3779 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3780 This is wrong and can cause the other end to generate a synchronization
3781 error if it is another Exim or anything else that does the synchronization
3782 check. A QUIT command is no longer sent after a timeout.
3784 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3785 during host lookups.
3787 74. The defer_ok option of callout verification was not working if it was used
3788 when verifying addresses in header lines, that is, for this case:
3790 verify = header_sender/callout=defer_ok
3792 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3793 those file descriptors could be used for SMTP connections. If anything
3794 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3795 could be sent to the SMTP client, causing chaos. The daemon now opens
3796 stdin, stdout, and stderr to /dev/null when it puts itself into the
3799 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3800 log by default. The user can override this with suitable Perl magic.
3802 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3803 the log message when discard triggers, was not working for the DATA ACL or
3804 for the non-SMTP ACL.
3806 78. Error message wording change in sieve.c.
3808 79. If smtp_accept_max_per_host was set, the number of connections could be
3809 restricted to fewer than expected, because the daemon was trying to set up
3810 a new connection before checking whether the processes handling previous
3811 connections had finished. The check for completed processes is now done
3812 earlier. On busy systems, this bug wouldn't be noticed because something
3813 else would have woken the daemon, and it would have reaped the completed
3816 80. If a message was submitted locally by a user whose login name contained one
3817 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3818 It caused a spool format error. I have fixed the spool reading code. A
3819 related problem was that the "from" clause in the Received: line became
3820 illegal because of the space(s). It is now covered by ${quote_local_part.
3822 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3825 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3827 83. Patch from Sieve maintainer to fix unterminated string problem in
3828 "vacation" handling.
3830 84. Some minor changes to the Linux configuration files to help with other
3831 OS variants using glibc.
3833 85. One more patch for Sieve to update vacation handling to latest spec.
3836 ----------------------------------------------------
3837 See the note above about the 4.44 and 4.50 releases.
3838 ----------------------------------------------------
3844 1. Change 4.43/35 introduced a bug that caused file counts to be
3845 incorrectly computed when quota_filecount was set in an appendfile
3848 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3849 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3852 3. Give more explanation in the error message when the command for a transport
3853 filter fails to execute.
3855 4. There are several places where Exim runs a non-Exim command in a
3856 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3857 was being done only for the command run by the queryprogram router. It is
3858 now done for all such subprocesses. The other cases are: ${run, transport
3859 filters, and the commands run by the lmtp and pipe transports.
3861 5. Some older OS have a limit of 256 on the maximum number of file
3862 descriptors. Exim was using setrlimit() to set 1000 as a large value
3863 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3864 systems. I've change it so that if it can't get 1000, it tries for 256.
3866 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3867 was an oversight, and furthermore, ever since the addition of extra
3868 controls (e.g. 4.43/32), the checks on when to allow different forms of
3869 "control" were broken. There should now be diagnostics for all cases when a
3870 control that does not make sense is encountered.
3872 7. $recipients is now available in the predata ACL (oversight).
3874 8. Tidy the search cache before the fork to do a delivery from a message
3875 received from the command line. Otherwise the child will trigger a lookup
3876 failure and thereby defer the delivery if it tries to use (for example) a
3877 cached ldap connection that the parent has called unbind on.
3879 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3880 of $address_data from the recipient verification was clobbered by the
3881 sender verification.
3883 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3884 was its contents. (It was OK if the option was not defined at all.)
3886 11. A "Completed" log line is now written for messages that are removed from
3887 the spool by the -Mrm option.
3889 12. $host_address is now set to the target address during the checking of
3890 ignore_target_hosts.
3892 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3893 being passed; this would have caused $sender_host_name to have been used if
3894 matching the list had actually called for a host name (not very likely,
3895 since this list is usually IP addresses). A host name is now passed as
3898 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3899 code that specifies a non-restarting handler (typically sigaction() in
3900 modern systems) in an attempt to fix a rare and obscure crash bug.
3902 15. Narrowed the window for a race in the daemon that could cause it to ignore
3903 SIGCHLD signals. This is not a major problem, because they are used only to
3904 wake it up if nothing else does.
3906 16. A malformed maildirsize file could cause Exim to calculate negative values
3907 for the mailbox size or file count. Odd effects could occur as a result.
3908 The maildirsize information is now recalculated if the size or filecount
3911 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3912 support for a long time. Removed HAVE_SYS_VFS_H.
3914 18. Updated exipick to current release from John Jetmore.
3916 19. Allow an empty sender to be matched against a lookup in an address list.
3917 Previously the only cases considered were a regular expression, or an
3920 20. Exim went into a mad DNS lookup loop when doing a callout where the
3921 host was specified on the transport, if the DNS lookup yielded more than
3924 21. The RFC2047 encoding function was originally intended for short strings
3925 such as real names; it was not keeping to the 75-character limit for
3926 encoded words that the RFC imposes. It now respects the limit, and
3927 generates multiple encoded words if necessary. To be on the safe side, I
3928 have increased the buffer size for the ${rfc2047: expansion operator from
3931 22. Failure to deliver a bounce message always caused it to be frozen, even if
3932 there was an errors_to setting on the router. The errors_to setting is now
3935 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3936 canonical form (fully expanded) before being placed in
3937 $sender_host_address.
3939 24. Updated eximstats to version 1.33
3941 25. Include certificate and key file names in error message when GnuTLS fails
3942 to set them up, because the GnuTLS error message doesn't include the name
3943 of the failing file when there is a problem reading it.
3945 26. Expand error message when OpenSSL has problems setting up cert/key files.
3948 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3949 (this can affect the format of dates).
3951 28. exim_tidydb, when checking for the continued existence of a message for
3952 which it has found a message-specific retry record, was not finding
3953 messages that were in split spool directories. Consequently, it was
3954 deleting retry records that should have stayed in existence.
3956 29. eximstats updated to version 1.35
3957 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3958 1.35 - bugfix such that pie charts by volume are generated correctly
3960 30. The SPA authentication driver was not abandoning authentication and moving
3961 on to the next authenticator when an expansion was forced to fail,
3962 contradicting the general specification for all authenticators. Instead it
3963 was generating a temporary error. It now behaves as specified.
3965 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3966 (the order specifies the preference for clients). The order is now AES256,
3967 AES128, 3DES, ARCFOUR128.
3969 31. Small patch to Sieve code - explicitly set From: when generating an
3972 32. Exim crashed if a remote delivery caused a very long error message to be
3973 recorded - for instance if somebody sent an entire SpamAssassin report back
3974 as a large number of 550 error lines. This bug was coincidentally fixed by
3975 increasing the size of one of Exim's internal buffers (big_buffer) that
3976 happened as part of the Exiscan merge. However, to be on the safe side, I
3977 have made the code more robust (and fixed the comments that describe what
3980 33. Some experimental protocols are using DNS PTR records for new purposes. The
3981 keys for these records are domain names, not reversed IP addresses. The
3982 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3983 leaves it alone. Component reversal etc. now happens only for IP addresses.
3986 34. The host_aton() function is supposed to be passed a string that is known
3987 to be a valid IP address. However, in the case of IPv6 addresses, it was
3988 not checking this. This is a hostage to fortune. Exim now panics and dies
3989 if the condition is not met. A case was found where this could be provoked
3990 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3991 components; fortuitously, this particular loophole had already been fixed
3992 by change 4.50/55 or 4.44/33 above.
3994 If there are any other similar loopholes, the new check in host_aton()
3995 itself should stop them being exploited. The report I received stated that
3996 data on the command line could provoke the exploit when Exim was running as
3997 exim, but did not say which command line option was involved. All I could
3998 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3999 running as the user.
4002 35. There was a buffer overflow vulnerability in the SPA authentication code
4003 (which came originally from the Samba project). I have added a test to the
4004 spa_base64_to_bits() function which I hope fixes it.
4007 36. The daemon start-up calls getloadavg() while still root for those OS that
4008 need the first call to be done as root, but it missed one case: when
4009 deliver_queue_load_max is set with deliver_drop_privilege. This is
4010 necessary for the benefit of the queue runner, because there is no re-exec
4011 when deliver_drop_privilege is set.
4013 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
4014 was in use, was not putting the data itself into the right store pool;
4015 consequently, it could be overwritten for a subsequent message in the same
4016 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
4019 38. Sometimes the final signoff response after QUIT could fail to get
4020 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
4021 < 0 before doing a fflush(). This bug looks as though it goes back to the
4022 introduction of TLS in release 3.20, but "sometimes" must have been rare
4023 because the tests only now provoked it.
4029 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
4030 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
4031 Now it returns an integer. A number of calls were still expecting a T/F
4032 return. Fortuitously, in all cases, the tests worked in OK situations,
4033 which is the norm. However, things would have gone wrong on any write
4034 failures on the smtp file descriptor. This function is used when sending
4035 messages over SMTP and also when doing verify callouts.
4037 2. When Exim is called to do synchronous delivery of a locally submitted
4038 message (the -odf or -odi options), it no longer closes stderr before doing
4041 3. Implemented the mua_wrapper option.
4043 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
4045 5. Implemented the functions header_remove(), header_testname(),
4046 header_add_at_position(), and receive_remove_recipient(), and exported them
4049 6. If an ACL "warn" statement specified the addition of headers, Exim already
4050 inserted X-ACL-Warn: at the start if there was no header name. However, it
4051 was not making this test for the second and subsequent header lines if
4052 there were newlines in the string. This meant that an invalid header could
4053 be inserted if Exim was badly configured.
4055 7. Allow an ACL "warn" statement to add header lines at the start or after all
4056 the Received: headers, as well as at the end.
4058 8. Added the rcpt_4xx retry error code.
4060 9. Added postmaster_mailfrom=xxx to callout verification option.
4062 10. Added mailfrom=xxxx to the callout verification option, for verify=
4065 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4066 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4068 12. Inserted some casts to stop certain compilers warning when using pointer
4069 differences as field lengths or precisions in printf-type calls (mostly
4070 affecting debugging statements).
4072 13. Added optional readline() support for -be (dynamically loaded).
4074 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4075 same clock tick as a message's arrival, so that its received time was the
4076 same as the "first fail" time on the retry record, and that message
4077 remained on the queue past the ultimate address timeout, every queue runner
4078 would try a delivery (because it was past the ultimate address timeout) but
4079 after another failure, the ultimate address timeout, which should have then
4080 bounced the address, did not kick in. This was a "< instead of <=" error;
4081 in most cases the first failure would have been in the next clock tick
4082 after the received time, and all would be well.
4084 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4085 being recognized when the domain list was tested by the match_domain
4086 condition in an expansion string.
4088 16. Added the ${str2b64: operator.
4090 17. Exim was always calling setrlimit() to set a large limit for the number of
4091 processes, without checking whether the existing limit was already
4092 adequate. (It did check for the limit on file descriptors.) Furthermore,
4093 errors from getrlimit() and setrlimit() were being ignored. Now they are
4094 logged to the main and panic logs, but Exim does carry on, to try to do its
4095 job under whatever limits there are.
4097 18. Imported PCRE 5.0.
4099 19. Trivial typo in log message " temporarily refused connection" (the leading
4102 20. If the log selector return_path_on_delivery was set and an address was
4103 redirected to /dev/null, the delivery process crashed because it assumed
4104 that a return path would always be set for a "successful" delivery. In this
4105 case, the whole delivery is bypassed as an optimization, and therefore no
4108 21. Internal re-arrangement: the function for sending a challenge and reading
4109 a response while authentication was assuming a zero-terminated challenge
4110 string. It's now changed to take a pointer and a length, to allow for
4111 binary data in such strings.
4113 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4115 23. Exim was not respecting finduser_retries when seeking the login of the
4116 uid under which it was called; it was always trying 10 times. (The default
4117 setting of finduser_retries is zero.) Also, it was sleeping after the final
4118 failure, which is pointless.
4120 24. Implemented tls_on_connect_ports.
4122 25. Implemented acl_smtp_predata.
4124 26. If the domain in control=submission is set empty, Exim assumes that the
4125 authenticated id is a complete email address when it generates From: or
4126 Sender: header lines.
4128 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4129 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4130 chown and chgrp in /bin and hostname in /usr/bin.
4132 28. Exim was keeping the "process log" file open after each use, just as it
4133 does for the main log. This opens the possibility of it remaining open for
4134 long periods when the USR1 signal hits a daemon. Occasional processlog
4135 errors were reported, that could have been caused by this. Anyway, it seems
4136 much more sensible not to leave this file open at all, so that is what now
4139 29. The long-running daemon process does not normally write to the log once it
4140 has entered its main loop, and it closes the log before doing so. This is
4141 so that log files can straightforwardly be renamed and moved. However,
4142 there are a couple of unusual error situations where the daemon does write
4143 log entries, and I had neglected to close the log afterwards.
4145 30. The text of an SMTP error response that was received during a remote
4146 delivery was being truncated at 512 bytes. This is too short for some of
4147 the long messages that one sometimes sees. I've increased the limit to
4150 31. It is now possible to make retry rules that apply only when a message has a
4151 specific sender, in particular, an empty sender.
4153 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4154 it possible to be selective about when SMTP synchronization is enforced.
4156 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4158 32. Implemented hosts_connection_nolog.
4160 33. Added an ACL for QUIT.
4162 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4165 35. Added mailbox_size and mailbox_filecount to appendfile.
4167 36. Added control = no_multiline_responses to ACLs.
4169 37. There was a bug in the logic of the code that waits for the clock to tick
4170 in the case where the clock went backwards by a substantial amount such
4171 that the microsecond fraction of "now" was more than the microsecond
4172 fraction of "then" (but the whole seconds number was less).
4174 38. Added support for the libradius Radius client library this is found on
4175 FreeBSD (previously only the radiusclient library was supported).
4181 1. When certain lookups returned multiple values in the form name=value, the
4182 quoting of the values was not always being done properly. Specifically:
4183 (a) If the value started with a double quote, but contained no whitespace,
4185 (b) If the value contained whitespace other than a space character (i.e.
4186 tabs or newlines or carriage returns) it was not quoted.
4187 This fix has been applied to the mysql and pgsql lookups by writing a
4188 separate quoting function and calling it from the lookup code. The fix
4189 should probably also be applied to nisplus, ibase and oracle lookups, but
4190 since I cannot test any of those, I have not disturbed their existing code.
4192 2. A hit in the callout cache for a specific address caused a log line with no
4193 reason for rejecting RCPT. Now it says "Previous (cached) callout
4194 verification failure".
4196 3. There was an off-by-one bug in the queryprogram router. An over-long
4197 return line was truncated at 256 instead of 255 characters, thereby
4198 overflowing its buffer with the terminating zero. As well as fixing this, I
4199 have increased the buffer size to 1024 (and made a note to document this).
4201 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4202 when Exim is waiting for an SMTP response from a remote server, Exim
4203 restarts its select() call on the socket, thereby resetting its timeout.
4204 This is not a problem when such interrupts are rare. Somebody set up a cron
4205 job to run exiwhat every 2 minutes, which is less than the normal select()
4206 timeout (5 or 10 minutes). This meant that the select() timeout never
4207 kicked in because it was always reset. I have fixed this by comparing the
4208 time when an interrupt arrives with the time at the start of the first call
4209 to select(). If more time than the timeout has elapsed, the interrupt is
4210 treated as a timeout.
4212 5. Some internal re-factoring in preparation for the addition of Sieve
4213 extensions (by MH). In particular, the "personal" test is moved to a
4214 separate function, and given an option for scanning Cc: and Bcc: (which is
4215 not set for Exim filters).
4217 6. When Exim created an email address using the login of the caller as the
4218 local part (e.g. when creating a From: or Sender: header line), it was not
4219 quoting the local part when it contained special characters such as @.
4221 7. Installed new OpenBSD configuration files.
4223 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4224 try to make them clearer.
4226 9. Callout options, other than the timeout value, were being ignored when
4227 verifying sender addresses in header lines. For example, when using
4229 verify = header_sender/callout=no_cache
4231 the cache was (incorrectly) being used.
4233 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4234 only the Cygwin environment.
4236 11. When return_path_on_delivery was set as a log selector, if different remote
4237 addresses in the same message used different return paths and parallel
4238 remote delivery occurred, the wrong values would sometimes be logged.
4239 (Whenever a remote delivery process finished, the return path value from
4240 the most recently started remote delivery process was logged.)
4242 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4243 lines when AUTH and/or TLS are in use. This is the "received protocol"
4244 field. Exim used to use "asmtp" for authenticated SMTP, without any
4245 indication (in the protocol name) for TLS use. Now it follows the RFC and
4246 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4247 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4248 names appear in log lines as well as in Received: header lines.
4250 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4251 extensions, and comparison tests, and to fix some bugs.
4253 14. Changes to the "personal" filter test:
4255 (1) The test was buggy in that it was just doing the equivalent of
4256 "contains" tests on header lines. For example, if a user's address was
4257 anne@some.where, the "personal" test would incorrectly be true for
4259 To: susanne@some.where
4261 This test is now done by extracting each address from the header in turn,
4262 and checking the entire address. Other tests that are part of "personal"
4263 are now done using regular expressions (for example, to check local parts
4264 of addresses in From: header lines).
4266 (2) The list of non-personal local parts in From: addresses has been
4267 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4268 taken from the Sieve specification recommendations.
4270 (3) If the message contains any header line starting with "List-" it is
4271 treated as non-personal.
4273 (4) The test for "circular" in the Subject: header line has been removed
4274 because it now seems ill-conceived.
4276 15. Minor typos in src/EDITME comments corrected.
4278 16. Installed latest exipick from John Jetmore.
4280 17. If headers_add on a router specified a text string that was too long for
4281 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4282 of string_sprintf() is now avoided.
4284 18. $message_body_size was not set (it was always zero) when running the DATA
4285 ACL and the local_scan() function.
4287 19. For the "mail" command in an Exim filter, no default was being set for
4288 the once_repeat time, causing a random time value to be used if "once" was
4289 specified. (If the value happened to be <= 0, no repeat happened.) The
4290 default is now 0s, meaning "never repeat". The "vacation" command was OK
4291 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4292 (I found it when inspecting the code).
4294 20. There is now an overall timeout for performing a callout verification. It
4295 defaults to 4 times the callout timeout, which applies to individual SMTP
4296 commands during the callout. The overall timeout applies when there is more
4297 than one host that can be tried. The timeout is checked before trying the
4298 next host. This prevents very long delays if there are a large number of
4299 hosts and all are timing out (e.g. when the network connections are timing
4300 out). The value of the overall timeout can be changed by specifying an
4301 additional sub-option for "callout", called "maxwait". For example:
4303 verify = sender/callout=5s,maxwait=20s
4305 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4306 to the end before writing, but this should make it even safer).
4308 22. Exim was forgetting that it had advertised PIPELINING for the second and
4309 subsequent messages on an SMTP connection. It was also not resetting its
4310 memory on STARTTLS and an internal HELO.
4312 23. When Exim logs an SMTP synchronization error within a session, it now
4313 records whether PIPELINING has been advertised or not.
4315 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4316 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4317 rather than long int.
4319 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4321 26. Added the never_mail option to autoreply.
4327 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4328 crash if the getsockname() call failed; this can happen if a connection is
4329 closed very soon after it is established. The problem was simply in the
4330 order in which certain operations were done, causing Exim to try to write
4331 to the SMTP stream before it had set up the file descriptor. The bug has
4332 been fixed by making things happen in the correct order.
4338 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4339 before the connection was closed, thus losing the rejection response.
4341 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4342 some early Solaris releases, but causes trouble in current releases where
4343 socklen_t is defined.
4345 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4348 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4351 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4352 that releasing the top of it at the end releases what was used for sub-
4353 expansions (unless the block got too big). However, discard this block if
4354 the first thing is a variable or header, so that we can use its block when
4355 it is dynamic (useful for very large $message_headers, for example).
4357 6. Lookups now cache *every* query, not just the most recent. A new, separate
4358 store pool is used for this. It can be recovered when all lookup caches are
4359 flushed. Lookups now release memory at the end of their result strings.
4360 This has involved some general refactoring of the lookup sources.
4362 7. Some code has been added to the store_xxx() functions to reduce the amount
4363 of flapping under certain conditions.
4365 8. log_incoming_interface used to affect only the <= reception log lines. Now
4366 it causes the local interface and port to be added to several more SMTP log
4367 lines, for example "SMTP connection from", and rejection lines.
4369 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4371 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4373 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4374 could be overwritten at the end of the current message (or the start of a
4375 new message if it was set in a HELO ACL). The value is now preserved for
4376 the duration of the SMTP connection.
4378 12. If a transport had a headers_rewrite setting, and a matching header line
4379 contained an unqualified address, that address was qualified, even if it
4380 did not match any rewriting rules. The underlying bug was that the values
4381 of the flags that permit the existence of unqualified sender and recipient
4382 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4383 non-local messages, and by -bnq for local messages) were not being
4384 preserved with the message after it was received.
4386 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4387 "next input=" as part of the text comprising the host identity instead of
4388 the correct text. The code was using the same buffer for two different
4389 strings. However, depending on which order the printing function evaluated
4390 its arguments, the bug did not always show up. Under Linux, for example, my
4391 test suite worked just fine.
4393 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4394 doesn't work with some older versions of Perl. It has been changed to "my",
4395 which in any case is probably the better facility to use.
4397 15. A really picky compiler found some instances of statements for creating
4398 error messages that either had too many or two few arguments for the format
4401 16. The size of the buffer for calls to the DNS resolver has been increased
4402 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4403 for addresses that have a lot of PTR records. This alleviates a problem; it
4404 does not fully solve it.
4406 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4407 buffer now truncates the list and logs the incident, which is the same
4408 action as happens when Exim is looking up a host name and its aliases.
4409 Previously in this situation something unpredictable would happen;
4410 sometimes it was "internal error: store_reset failed".
4412 18. If a server dropped the connection unexpectedly when an Exim client was
4413 using GnuTLS and trying to read a response, the client delivery process
4414 crashed while trying to generate an error log message.
4416 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4417 string, for example:
4419 warn message = H1: something\nH2: something
4421 the text was added as a single header line from Exim's point of view
4422 though it ended up OK in the delivered message. However, searching for the
4423 second and subsequent header lines using $h_h2: did not work. This has been
4424 fixed. Similarly, if a system filter added multiple headers in this way,
4425 the routers could not see them.
4427 20. Expanded the error message when iplsearch is called with an invalid key to
4428 suggest using net-iplsearch in a host list.
4430 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4431 ACLs are no longer actually imposed (and a message to that effect is
4434 22. If a "gecos" field in a passwd entry contained escaped characters, in
4435 particular, if it contained a \" sequence, Exim got it wrong when building
4436 a From: or a Sender: header from that name. A second bug also caused
4437 incorrect handling when an unquoted " was present following a character
4438 that needed quoting.
4440 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4441 was not being matched caselessly.
4443 24. Arranged for all hyphens in the exim.8 source to be escaped with
4446 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4447 properly. Recipient callout cache records were still being keyed to include
4448 the sender, even when use_sender was set false. This led to far more
4449 callouts that were necessary. The sender is no longer included in the key
4450 when use_sender is false.
4452 26. Added "control = submission" modifier to ACLs.
4454 27. Added the ${base62d: operator to decode base 62 numbers.
4456 28. dnsdb lookups can now access SRV records.
4458 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4459 the configuration file.
4461 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4462 (-v) mode. This makes the output for a verbose queue run more intelligible.
4464 31. Added a use_postmaster feature to recipient callouts.
4466 32. Added the $body_zerocount variable, containing the number of binary zero
4467 bytes in the message body.
4469 33. The time of last modification of the "new" subdirectory is now used as the
4470 "mailbox time last read" when there is a quota error for a maildir
4473 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4475 35. Added +ignore_unknown as a special item in host lists.
4477 36. Code for decoding IPv6 addresses in host lists is now included, even if
4478 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4479 address was recognized as an IP address, but was then not correctly decoded
4480 into binary, causing unexpected and incorrect effects when compared with
4487 1. Very minor rewording of debugging text in manualroute to say "list of
4488 hosts" instead of "hostlist".
4490 2. If verify=header_syntax was set, and a header line with an unqualified
4491 address (no domain) and a large number of spaces between the end of the
4492 name and the colon was received, the reception process suffered a buffer
4493 overflow, and (when I tested it) crashed. This was caused by some obsolete
4494 code that should have been removed. The fix is to remove it!
4496 3. When running in the test harness, delay a bit after writing a bounce
4497 message to get a bit more predictability in the log output.
4499 4. Added a call to search_tidyup() just before forking a reception process. In
4500 theory, someone could use a lookup in the expansion of smtp_accept_max_
4501 per_host which, without the tidyup, could leave open a database connection.
4503 5. Added the variables $recipient_data and $sender_data which get set from a
4504 lookup success in an ACL "recipients" or "senders" condition, or a router
4505 "senders" option, similar to $domain_data and $local_part_data.
4507 6. Moved the writing of debug_print from before to after the "senders" test
4510 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4511 problems for message scanning, either using a data ACL, or using
4512 local_scan() because the Received: header was not generated till after they
4513 were called (in order to set the time as the time of reception completion).
4514 I have revised the way this works. The header is now generated after the
4515 body is received, but before the ACL or local_scan() are called. After they
4516 are run, the timestamp in the header is updated.
4522 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4523 before starting a queue runner without re-exec. This happened only when
4524 deliver_drop_privilege was set or when the Exim user was set to root. The
4525 effect of the bug was that timeouts during subsequent deliveries caused
4526 crashes instead of being properly handled. The handler is now left at its
4527 default (and expected) setting.
4529 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4530 message, again when deliver_drop_privilege is set or Exim is run as root.
4531 The bug described in (1) was not present in this case, but the tidying up
4532 of the other signals was missing. I have made the two cases consistent.
4534 3. The ignore_target_hosts setting on a manualroute router was being ignored
4535 for hosts that were looked up using the /MX notation.
4537 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4540 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4541 operated on the sender address. After changing the $sender_address to <>
4542 for the sender address verify, Exim was re-instated it as the original
4543 (before rewriting) address, but remembering that it had rewritten it, so it
4544 wasn't rewriting it again. This bug also had the effect of breaking the
4545 sender address verification caching when the sender address was rewritten.
4547 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4548 This has been changed so that if the ip literal address matches
4549 ignore_target_hosts, the router declines.
4551 7. Added expansion conditions match_domain, match_address, and match_local_
4552 part (NOT match_host).
4554 8. The placeholder for the Received: header didn't have a length field set.
4556 9. Added code to Exim itself and to exim_lock to test for a specific race
4557 condition that could lead to file corruption when using MBX delivery. The
4558 issue is with the lockfile that is created in /tmp. If this file is removed
4559 after a process has opened it but before that process has acquired a lock,
4560 there is the potential for a second process to recreate the file and also
4561 acquire a lock. This could lead to two Exim processes writing to the file
4562 at the same time. The added code performs the same test as UW imapd; it
4563 checks after acquiring the lock that its file descriptor still refers to
4564 the same named file.
4566 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4567 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4570 11. Added the smtp_active_hostname option. If used, this will typically be made
4571 to depend on the incoming interface address. Because $interface_address is
4572 not set up until the daemon has forked a reception process, error responses
4573 that can happen earlier (such as "too many connections") no longer contain
4576 12. If an expansion in a condition on a "warn" statement fails because a lookup
4577 defers, the "warn" statement is abandoned, and the next ACL statement is
4578 processed. Previously this caused the whole ACL to be aborted.
4580 13. Added the iplsearch lookup type.
4582 14. Added ident_timeout as a log selector.
4584 15. Added tls_certificate_verified as a log selector.
4586 16. Added a global option tls_require_ciphers (compare the smtp transport
4587 option of the same name). This controls incoming TLS connections.
4589 17. I finally figured out how to make tls_require_ciphers do a similar thing
4590 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4591 before starting the TLS session.
4593 18. Tabs are now shown as \t in -bP output.
4595 19. If the log selector return_path_on_delivery was set, Exim crashed when
4596 bouncing a message because it had too many Received: header lines.
4598 20. If two routers both had headers_remove settings, and the first one included
4599 a superfluous trailing colon, the final name in the first list and the
4600 first name in the second list were incorrectly joined into one item (with a
4601 colon in the middle).
4607 1. Added -C and -D options to the exinext utility, mainly to make it easier
4608 to include in the automated testing, but these could be helpful when
4609 multiple configurations are in use.
4611 2. The exinext utility was not formatting the output nicely when there was
4612 an alternate port involved in the retry record key, nor when there was a
4613 message id as well (for retries that were specific to a specific message
4614 and a specific host). It was also confused by IPv6 addresses, because of
4615 the additional colons they contain. I have fixed the IPv4 problem, and
4616 patched it up to do a reasonable job for IPv6.
4618 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4619 delivery, the log line now contains "pipelined" if PIPELINING was used.
4621 4. An SMTP transport process used to panic and die if the bind() call to set
4622 an explicit outgoing interface failed. This has been changed; it is now
4623 treated in the same way as a connect() failure.
4625 5. A reference to $sender_host_name in the part of a conditional expansion
4626 that was being skipped was still causing a DNS lookup. This no longer
4629 6. The def: expansion condition was not recognizing references to header lines
4630 that used bh_ and bheader_.
4632 7. Added the _cache feature to named lists.
4634 8. The code for checking quota_filecount in the appendfile transport was
4635 allowing one more file than it should have been.
4637 9. For compatibility with Sendmail, the command line option
4645 and sets the incoming protocol and host name (for trusted callers). The
4646 host name and its colon can be omitted when only the protocol is to be set.
4647 Note the Exim already has two private options, -pd and -ps, that refer to
4648 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4649 "s", but I don't think that's a major issue.
4651 10. A number of refactoring changes to the code, none of which should affect
4654 (a) The number of logging options was getting close to filling up the
4655 32-bit word that was used as a bit map. I have split them into two classes:
4656 those that are passed in the argument to log_write(), and those that are
4657 only ever tested independently outside of that function. These are now in
4658 separate 32-bit words, so there is plenty of room for expansion again.
4659 There is no change in the user interface or the logging behaviour.
4661 (b) When building, for example, log lines, the code previously used a
4662 macro that called string_cat() twice, in order to add two strings. This is
4663 not really sufficiently general. Furthermore, there was one instance where
4664 it was actually wrong because one of the argument was used twice, and in
4665 one call a function was used. (As it happened, calling the function twice
4666 did not affect the overall behaviour.) The macro has been replaced by a
4667 function that can join an arbitrary number of extra strings onto a growing
4670 (c) The code for expansion conditions now uses a table and a binary chop
4671 instead of a serial search (which was left over from when there were very
4672 few conditions). Also, it now recognizes conditions like "pam" even when
4673 the relevant support is not compiled in: a suitably worded error message is
4674 given if an attempt is made to use such a condition.
4676 11. Added ${time_interval:xxxxx}.
4678 12. A bug was causing one of the ddress fields not to be passed back correctly
4679 from remote delivery subprocesses. The field in question was not being
4680 subsequently used, so this caused to problems in practice.
4682 13. Added new log selectors queue_time and deliver_time.
4684 14. Might have fixed a bug in maildirsizefile handling that threw up
4685 "unexpected character" debug warnings, and recalculated the data
4686 unnecessarily. In any case, I expanded the warning message to give more
4689 15. Added the message "Restricted characters in address" to the statements in
4690 the default ACL that block characters like @ and % in local parts.
4692 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4693 Three changes have been made:
4695 (a) There was a serious bug; a negative response to MAIL caused the whole
4696 recipient domain to be cached as invalid, thereby blocking all messages
4697 to all local parts at the same domain, from all senders. This bug has
4698 been fixed. The domain is no longer cached after a negative response to
4699 MAIL if the sender used is not empty.
4701 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4704 (c) A new callout option, "use_sender" has been added for people who want
4705 the modified behaviour.
4711 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4714 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4715 indeed breaks things for older releases.
4717 3. Added additional logging to the case where there is a problem reading data
4718 from a filter that is running in a subprocess using a pipe, in order to
4719 try to track down a specific problem.
4721 4. Testing facility fudge: when running in the test harness and attempting
4722 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4723 getting "No route to host". Convert this to a timeout.
4725 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4728 6. Some OS don't have socklen_t but use size_t instead. This affects the
4729 fifth argument of getsockopt() amongst other things. This is now
4730 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4731 can be set for individual OS. I have set it for SunOS5, OSF1, and
4732 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4733 some earlier ones do not.
4735 7. Change 4.30/15 was not doing the test caselessly.
4737 8. The standard form for an IPv6 address literal was being rejected by address
4738 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4739 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4740 this, as well as the form without the "IPv6" on the front (but only when
4741 address literals are enabled, of course).
4743 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4745 10. Exim crashed if a message with an empty sender address specified by -f
4746 encountered a router with an errors_to setting. This could be provoked only
4747 by a command such as
4751 where an empty string was supplied; "<>" did not hit this bug.
4753 11. Installed PCRE release 4.5.
4755 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4756 remained set. It is now erased.
4758 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4759 times from message ids (which are base 36 rather than the normal 62).
4761 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4762 were being counted as actual protocol errors, and logged if the log
4763 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4764 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4765 and DATA following a set of rejected RCPTs do not count as protocol errors.
4766 In other words, Exim assumes they were pipelined, though this may not
4767 actually be the case. Of course, in all cases the client gets an
4768 appropriate error code.
4770 15. If a lookup fails in an ACL condition, a message about the failure may
4771 be available; it is used if testing the ACL cannot continue, because most
4772 such messages specify what the cause of the deferral is. However, some
4773 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4774 that caused an old message to be retained and used if a later statement
4775 caused a defer, replacing the real cause of the deferral.
4777 16. If an IP address had so many PTR records that the DNS lookup buffer
4778 was not large enough to hold them, Exim could crash while trying to process
4779 the truncated data. It now detects and logs this case.
4781 17. Further to 4.21/58, another change has been made: if (and only if) the
4782 first line of a message (the first header line) ends with CRLF, a bare LF
4783 in a subsequent header line has a space inserted after it, so as not to
4784 terminate the header.
4786 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4787 unnecessarily, used atoi() instead of strtol(), and didn't check the
4788 termination when getting file sizes from file names by regex.
4790 19. Completely re-implemented the support for maildirsize files, in the light
4791 of a number of problems with the previous contributed implementation
4792 (4.30/29). In particular:
4794 . If the quota is zero, the maildirsize file is maintained, but no quota is
4797 . If the maildir directory does not exist, it is created before any attempt
4798 to write a maildirsize file.
4800 . The quota value in the file is just a cache; if the quota is changed in
4801 the transport, the new value overrides.
4803 . A regular expression is available for excluding directories from the
4806 20. The autoreply transport checks the characters in options that define the
4807 message's headers; it allows continued headers, but it was checking with
4808 isspace() after an embedded newline instead of explicitly looking for a
4811 21. If all the "regular" hosts to which an address was routed had passed their
4812 expiry times, and had not reached their retry times, the address was
4813 bounced, even if fallback hosts were defined. Now Exim should go on to try
4816 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4817 equivalent code in the SMTP transport. Some hosts send humungous responses
4818 to HELO/EHLO, more than 1024 it seems.
4820 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4821 gives compiler warnings in some environments. I've now done it "properly",
4824 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4825 (because of problems with the built-in one) was declared to return uschar *
4826 instead of char *, causing compiler failure.
4828 25. Fixed a file descriptor leak when processing alias/forward files.
4830 26. Fixed a minor format string issue in dbfn.c.
4832 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4834 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4835 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4838 29. When a sender address is verified, it is cached, to save repeating the test
4839 when there is more than one recipient in a message. However, when the
4840 verification involves a callout, it is possible for different callout
4841 options to be set for different recipients. It is too complicated to keep
4842 track of this in the cache, so now Exim always runs a verification when a
4843 callout is required, relying on the callout cache for the optimization.
4844 The overhead is duplication of the address routing, but this should not be
4847 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4848 to be verified with callout=postmaster, and the main callout worked but the
4849 postmaster check failed, the verification correctly failed. However, if a
4850 subsequent RCPT command asked for sender verification *without* the
4851 postmaster check, incorrect caching caused this verification also to fail,
4854 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4855 it was not caching the DNS options (qualify_single, search_parents) that
4856 were used when the lookup failed. A subsequent lookup with different
4857 options therefore always gave the same answer, though there were cases
4858 where it should not have. (Example: a "domains = !$mx_any" option on a
4859 dnslookup router: the "domains" option is always processed without any
4860 widening, but the router might have qualify_single set.) Now Exim uses the
4861 cached value only when the same options are set.
4863 32. Added John Jetmore's "exipick" utility to the distribution.
4865 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4866 than a timeout (e.g. a certificate is required, and is not provided), an
4867 Exim server now closes the connection immediately. Previously it waited for
4868 the client to close - but if the client is SSL, it seems that they each
4869 wait for each other, leading to a delay before one of them times out.
4871 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4872 maintained 0.8.x compatibility because I don't think many are using it, and
4873 it is clearly obsolete.
4875 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4878 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4879 client certificate was expired. A simple patch fixes this, though I don't
4880 understand the full logic of why the verify callback is called multiple
4883 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4884 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4885 which causes problems with some clients (such as the Certicom SSL Plus
4886 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4887 disables the coutermeasure allowing Eudora to connect."
4889 38. Exim was not checking that a write() to a log file succeeded. This could
4890 lead to Bad Things if a log got too big, in particular if it hit a file
4891 size limit. Exim now panics and dies if it cannot write to a log file, just
4892 as it does if it cannot open a log file.
4894 39. Modified OS/Makefile-Linux so that it now contains
4896 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4898 The two -D definitions ensure that Exim is compiled with large file
4899 support, which makes it possible to handle log files that are bigger than
4902 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4903 instance) a domain was checked against a named list that involved a lookup,
4904 causing $domain_data to be set, then another domain was checked against the
4905 same list, then the first domain was re-checked, the value of $domain_data
4906 after the final check could be wrong. In particular, if the second check
4907 failed, it could be set empty. This bug probably also applied to
4910 41. The strip_trailing_dot option was not being applied to the address given
4911 with the -f command-line option.
4913 42. The code for reading a message's header from the spool was incrementing
4914 $received_count, but never initializing it. This meant that the value was
4915 incorrect (doubled) while delivering a message in the same process in which
4916 it was received. In the most common configuration of Exim, this never
4917 happens - a fresh exec is done - but it can happen when
4918 deliver_drop_privilege is set.
4920 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4921 it now includes up to 150 characters of the unexpected data in the log
4924 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4925 and building data strings. The size of both of these buffers was 10 000
4926 bytes - far larger than anybody would *ever* want, thought I. Needless to
4927 say, somebody hit the limit. I have increased the maximum line length to
4928 20 000 and the maximum data length of concatenated lines to 100 000. I have
4929 also fixed two bugs, because there was no checking on these buffers. Tsk,
4930 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4931 buffer is too small.
4933 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4934 lsearch lookups. Now it does.
4936 46. When parsing a route_list item in a manualroute router, a fixed-length
4937 buffer was used for the list of hosts. I made this 1024 bytes long,
4938 thinking that nobody would ever have a list of hosts that long. Wrong.
4939 Somebody had a whole pile of complicated expansion conditions, and the
4940 string was silently truncated, leading to an expansion error. It turns out
4941 that it is easier to change to an unlimited length (owing to other changes
4942 that have happened since this code was originally written) than to build
4943 structure for giving a limitation error. The length of the item that
4944 expands into the list of hosts is now unlimited.
4946 47. The lsearch lookup could not handle data where the length of text line was
4947 more than 4095 characters. Such lines were truncated, leading to shortened
4948 data being returned. It should now handle lines of any length.
4950 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4951 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4954 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4955 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4956 no longer happens in comment lines. A list of replacements is now placed
4957 at the head of all of the source files, except those whose only change is
4958 to replace PERL_COMMAND in the very first #! line.
4960 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4961 messages on the queue, with a bottom-up merge sort, using code contributed
4962 by Michael Haardt. This should make operations like -bp somewhat faster on
4963 large queues. It won't affect queue runners, except when queue_run_in_order
4966 51. Installed eximstats 1.31 in the distribution.
4968 52. Added support for SRV lookups to the dnslookup router.
4970 53. If an ACL referred to $message_body or $message_body_end, the value was not
4971 reset for any messages that followed in the same SMTP session.
4973 54. The store-handling optimization for building very long strings was not
4974 differentiating between the different store pools. I don't think this
4975 actually made any difference in practice, but I've tidied it.
4977 55. While running the routers to verify a sender address, $sender_address
4978 was still set to the sender address. This is wrong, because when routing to
4979 send a bounce to the sender, it would be empty. Therefore, I have changed
4980 it so that, while verifying a sender address, $sender_address is set to <>.
4981 (There is no change to what happens when verifying a recipient address.)
4983 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4984 target A or AAAA records (if not already returned) without resetting the
4985 qualify_single or search_parents options of the DNS resolver. These are
4986 inappropriate in this case because the targets of MX and SRV records must
4987 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4988 target that, when qualified, matched something in the local domain. These
4989 two options are now turned off when doing these lookups.
4991 57. It seems that at least some releases of Reiserfs (which does not have the
4992 concept of a fixed number of inodes) returns zero and not -1 for the
4993 number of available inodes. This interacted badly with check_spool_inodes,
4994 which assumed that -1 was the "no such thing" setting. What I have done is
4995 to check that the total number of inodes is greater than zero before doing
4996 the test of how many are available.
4998 58. When a "warn" ACL statement has a log_message modifier, the message is
4999 remembered, and not repeated. This is to avoid a lot of repetition when a
5000 message has many recipients that cause the same warning to be written.
5001 Howewer, Exim was preserving the list of already written lines for an
5002 entire SMTP session, which doesn't seem right. The memory is now reset if a
5003 new message is started.
5005 59. The "rewrite" debugging flag was not showing the result of rewriting in the
5006 debugging output unless log_rewrite was also set.
5008 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
5009 of (int)(handle) when we know that handle contains (void *)(-1).
5011 61. The Exim daemon panic-logs an error return when it closes the incoming
5012 connection. However "connection reset by peer" seems to be common, and
5013 isn't really an error worthy of noting specially, so that particular error
5016 62. When Exim is trying to find all the local interfaces, it used to panic and
5017 die if the ioctl to get the interface flags failed. However, it seems that
5018 on at least one OS (Solaris 9) it is possible to have an interface that is
5019 included in the list of interfaces, but for which you get a failure error
5020 for this call. This happens when the interface is not "plumbed" into a
5021 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
5022 failure of the "get flags" call assumes that the interface is down.
5024 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
5025 makes life easier for people who are doing arithmetic on fields extracted
5026 from dates, where you often get leading zeros that should not be
5027 interpreted as octal.
5029 64. Added qualify_domain to the redirect router, to override the global
5032 65. If a pathologically long header line contained very many addresses (the
5033 report of this problem mentioned 10 000) and each of them was rewritten,
5034 Exim could use up a very large amount of memory. (It kept on making new
5035 copies of the header line as it rewrote, and never released the old ones.)
5036 At the expense of a bit more processing, the header rewriting function has
5037 been changed so that it no longer eats memory in this way.
5039 66. The generation of the Received: header has been moved from the time that a
5040 message starts to be received, to the time that it finishes. The timestamp
5041 in the Received: header should now be very close to that of the <= log
5042 line. There are two side-effects of this change:
5044 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
5045 logged header lines no longer include the local Received: line, because
5046 it has not yet been created. The same applies to a copy of the message
5047 that is returned to a non-SMTP sender when a message is rejected.
5049 (b) When a filter file is tested using -bf, no additional Received: header
5050 is added to the test message. After some thought, I decided that this
5053 This change does not affect the value of $received_for. It is still set
5054 after address rewriting, but before local_scan() is called.
5056 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5058 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5059 gave an unhelpful panic error message, and a defer error. I have managed to
5060 change this behaviour so that it now rejects any supplied certificate,
5061 which seems right, as the list of acceptable certificates is empty.
5063 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5064 gave an unhelpful defer error. I have not managed to make this reject any
5065 supplied certificates, but the error message it gives is "no certificate
5066 supplied", which is not helpful.
5068 70. exigrep's output now also includes lines that are not associated with any
5069 message, but which match the given pattern. Implemented by a patch from
5070 Martin Sluka, which also tidied up the Perl a bit.
5072 71. Recipient callout verification, like sender verification, was using <> in
5073 the MAIL FROM command. This isn't really the right thing, since the actual
5074 sender may affect whether the remote host accepts the recipient or not. I
5075 have changed it to use the actual sender in the callout; this means that
5076 the cache record is now keyed on a recipient/sender pair, not just the
5077 recipient address. There doesn't seem to be a real danger of callout loops,
5078 since a callout by the remote host to check the sender would use <>.
5079 [SEE ABOVE: changed after hitting problems.]
5081 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5082 temporary errors. However, in the case of such a code being given after
5083 the end of a data transmission (i.e. after ".") Exim was failing to write
5084 a retry record for the message. (Yes, there was some broken host that was
5085 actually sending 8xx at this point.)
5087 73. An unknown lookup type in a host list could cause Exim to panic-die when
5088 the list was checked. (An example that provoked this was putting <; in the
5089 middle of a list instead of at the start.) If this happened during a DATA
5090 ACL check, a -D file could be left lying around. This kind of configuration
5091 error no longer causes Exim to die; instead it causes a defer errror. The
5092 incident is still logged to the main and panic logs.
5094 74. Buglet left over from Exim 3 conversion. The message "too many messages
5095 in one connection" was written to the rejectlog but not the mainlog, except
5096 when address rewriting (yes!) was being logged.
5098 75. Added write_rejectlog option.
5100 76. When a system filter was run not as root (that is, when system_filter_user
5101 was set), the values of the $n variables were not being returned to the
5102 main process; thus, they were not subsequently available in the $sn
5105 77. Added +return_path_on_delivery log selector.
5107 78. A connection timeout was being treated differently from recipients deferred
5108 when testing hosts_max_try with a message that was older than the host's
5109 retry timeout. (The host should not be counted, thus allowing all hosts to
5110 be tried at least once before bouncing.) This may have been the cause of an
5111 occasionally reported bug whereby a message would remain on the queue
5112 longer than the retry timeout, but would be bounced if a delivery was
5113 forced. I say "may" because I never totally pinned down the problem;
5114 setting up timeout/retry tests is difficult. See also the next item.
5116 79. The ultimate address timeout was not being applied to errors that involved
5117 a combination of host plus message (for example, a timeout on a MAIL
5118 command). When an address resolved to a number of possible hosts, and they
5119 were not all tried for each delivery (e.g. because of hosts_max_try), a
5120 message could remain on the queue longer than the retry timeout.
5122 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5125 81. Fixed an obscure SMTP outgoing bug which required at least the following
5126 conditions: (a) there was another message waiting for the same server;
5127 (b) the server returned 5xx to all RCPT commands in the first message so
5128 that the message was not completed; (c) the server dropped the connection
5129 or gave a negative response to the RSET that Exim sends to abort the
5130 transaction. The observed case was a dropped connection after DATA that had
5131 been sent in pipelining mode. That is, the server had advertised PIPELINING
5132 but was not implementing it correctly. The effect of the bug was incorrect
5133 behaviour, such as trying another host, and this could lead to a crash.
5139 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5140 and daemon.c were passed as pointers to ints; they should have been
5141 pointers to socklen_t variables (which are typically unsigned ints).
5143 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5146 3. Fixed a really odd bug that affected only the testing scheme; patching a
5147 certain fixed string in the binary changed the value of another string that
5148 happened to be identical to the end of the original first string.
5150 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5151 name", it returns that address as the IP address. On some operating
5152 systems (e.g. Solaris), it also passes back the IP address string as the
5153 "host name". However, on others (e.g. Linux), it passes back an empty
5154 string. Exim wasn't checking for this, and was changing the host name to an
5155 empty string, assuming it had been canonicized.
5157 5. Although rare, it is permitted to have more than one PTR record for a given
5158 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5159 all the names associated with an address, because they do in Solaris.
5160 However, it seems that they do not in Linux for data that comes from the
5161 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5162 I found this out when I moved to a new Linux workstation and tried to run
5163 the Exim test suite.
5165 To get round this problem I have changed the code so that it now does its
5166 own call to the DNS to look up PTR records when searching for a host name.
5167 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5168 addresses that are only in /etc/hosts are still found.
5170 This behaviour is, however, controlled by an option called host_lookup_
5171 order, which defaults to "bydns:byaddr". If people want to use the other
5172 order, or indeed, just use one or the other means of lookup, they can
5173 specify it in this variable.
5175 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5176 some operating systems, this comes back from gethostbyaddr() as an empty
5177 string, and this is what Exim used to test for. However, it seems that in
5178 other systems, "." is yielded. Exim now tests for this case too.
5180 7. The values of check_spool_space and check_log_space are now held internally
5181 as a number of kilobytes instead of an absolute number of bytes. If a
5182 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5183 kilobyte. This means that much larger values can be stored.
5185 8. Exim monitor: an attempt to get the action menu when not actually pointing
5186 at a message produces an empty menu entitled "No message selected". This
5187 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5188 no entries in it ("Shell widget menu has zero width and/or height"). So I
5189 have added a single, blank menu entry in this case.
5191 9. Added ${quote_local_part.
5193 10. MIME decoding is now applied to the contents of Subject: header lines when
5196 11. Now that a reference to $sender_host_address automatically causes a reverse
5197 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5198 host lookup before query-style lookups in lists that might use this
5199 variable. This has therefore been abolished, and the "net-" prefix is no
5200 longer necessary for query-style lookups.
5202 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5203 have been a typo for LFLAGS, so it has been changed.
5205 13. The install script calls Exim with "-C /dev/null" in order to find the
5206 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5207 to be output. Howeve, since Exim outputs its version number before the
5208 error, it didn't break the script. It just looked ugly. I fixed this by
5209 always allowing "-C /dev/null" if the caller is root.
5211 14. Ignore overlarge ACL variable number when reading spool file - insurance
5212 against a later release with more variables having written the file.
5214 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5215 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5216 this, as well as the form without the "IPv6" on the front.
5218 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5219 OS/Makefile-Darwin file.
5221 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5222 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5224 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5225 variable. However, some OS use ints for inodes. I've added cast to long int
5226 to get rid of the compiler warning.
5228 19. I had forgotten to lock out "/../" in configuration file names when
5229 ALT_CONFIG_PREFIX was set.
5231 20. Routers used for verification do not need to specify transports. However,
5232 if such a router generated a host list, and callout was configured, Exim
5233 crashed, because it could not find a port number from the (non-existent)
5234 transport. It now assumes port 25 in this circumstance.
5236 21. Added the -t option to exigrep.
5238 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5239 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5240 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5241 them is set, LOOKUP_LSEARCH is forced.
5243 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5245 24. Added sender and host information to the "rejected by local_scan()" log
5246 line; previously there was no indication of these.
5248 25. Added .include_if_exists.
5250 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5251 Linux. It turns out that not all file systems support this. Apparently some
5252 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5253 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5254 not supported on the file descriptor, is now ignored when Exim is trying to
5255 sync a directory. This applies only to Linux.
5257 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5259 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5260 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5263 29. Added support for maildirsize files from supplied patch (modified a bit).
5265 30. The use of :fail: followed by an empty string could lead Exim to respond to
5266 sender verification failures with (e.g.):
5268 550 Verification failed for <xxx>
5269 550 Sender verify failed
5271 where the first response line was missing the '-' that indicates it is not
5272 the final line of the response.
5274 31. The loop for finding the name of the user that called Exim had a hardwired
5275 limit of 10; it now uses the value of finduser_retries, which is used for
5276 all other user lookups.
5278 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5281 33. Exim was neglecting to zero errno before one call of strtol() when
5282 expanding a string and expecting an integer value. On some systems this
5283 resulted in spurious "integer overflow" errors. Also, it was casting the
5284 result into an int without checking.
5286 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5287 did not work. The code looks as if it has *never* worked, though it appears
5288 to have been documented since at least releast 1.62. I have made it work.
5290 35. The "timeout_DNS" error in retry rules, also documented since at least
5291 1.62, also never worked. As it isn't clear exactly what this means, and
5292 clearly it isn't a major issue, I have abolished the feature by treating it
5293 as "timeout", and writing a warning to the main and panic logs.
5295 36. The display of retry rules for -brt wasn't always showing the error code
5298 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5299 timeout_connect_A, timeout_connect_MX.
5301 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5302 to the empty sender.
5304 39. The daemon was not analysing the content of -oX till after it had closed
5305 stderr and disconnected from the controlling terminal. This meant that any
5306 syntax errors were only noted on the panic log, and the return code from
5307 the command was 0. By re-arranging the code a little, I've made the
5308 decoding happen first, so such errors now appear on stderr, and the return
5309 code is 1. However, the actual setting up of the sockets still happens in
5310 the disconnected process, so errors there are still only recorded on the
5313 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5314 connections (as happens on some IP stacks) was logged at start up time as
5315 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5316 it from "IPv6 and IPv4", which means that two separate sockets are being
5319 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5320 says whether AF_INET or AF_INET6 was passed as an argument.
5322 42. Exiwhat output was messed up when time zones were included in log
5325 43. Exiwhat now gives more information about the daemon's listening ports,
5326 and whether -tls-on-connect was used.
5328 44. The "port" option of the smtp transport is now expanded.
5330 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5331 silently ignored. Now an error message is written to the main and panic
5334 46. There's a new ACL modifier called "logwrite" which writes to a log file
5335 as soon as it is encountered.
5337 47. Added $local_user_uid and $local_user_gid at routing time.
5339 48. Exim crashed when trying to verify a sender address that was being
5342 49. Exim was recognizing only a space character after ".include". It now also
5343 recognizes a tab character.
5345 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5346 extracting the relevant information from the specification. The man page no
5347 longer contains scrambled data for the -d option, and I've added a section
5348 at the front about calling Exim under different names.
5350 51. Added "extra_headers" argument to the "mail" command in filter files.
5352 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5355 53. Installed eximstats 1.29.
5357 54. Added transport_filter_timeout as a generic transport option.
5359 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5360 Cc: header lines. This was required by RFC 822, but it not required by RFC
5363 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5364 incoming messages that did not have them. Now it does so only if the
5365 message originates locally, that is, if there is no associated remote host
5366 address. When Resent- header lines are present, this applies to the Resent-
5367 lines rather than the non-Resent- lines.
5369 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5370 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5372 58. Messages for configuration errors now include the name of the main
5373 configuration file - useful now that there may be more than one file in a
5374 list (.included file names were always shown).
5376 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5377 for those rare installations that do not start the daemon as root or run it
5378 setuid root. I've cut out the call to initgroups() if the daemon is not
5381 60. The Exim user and group can now be bound into the binary as text strings
5382 that are looked up at the start of Exim's processing.
5384 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5386 62. Added $mailstore_basename variable.
5388 63. Installed patch to sieve.c from Michael Haardt.
5390 64. When Exim failed to open the panic log after failing to open the main log,
5391 the original message it was trying to log was written to stderr and debug
5392 output, but if they were not available (the usual case in production), it
5393 was lost. Now it is written to syslog before the two lines that record the
5394 failures to open the logs.
5396 65. Users' Exim filters run in subprocesses under the user's uid. It is
5397 possible for a "deliver" command or an alias in a "personal" command to
5398 provoke an address rewrite. If logging of address rewriting is configured,
5399 this fails because the process is not running as root or exim. There may be
5400 a better way of dealing with this, but for the moment (because 4.30 needs
5401 to be released), I have disabled address rewrite logging when running a
5402 filter in a non-root, non-exim process.
5408 1. The buildconfig auxiliary program wasn't quoting the value set for
5409 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5410 not defined. This bug was masked in 4.22 by the effect that was fixed in
5413 2. Some messages that were rejected after a message id was allocated were
5414 shown as "incomplete" by exigrep. It no longer does this for messages that
5415 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5417 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5418 have allow_domain_literals set, the ID did not get logged in the <= line.
5419 Domain literals are now always recognized in Message-ID: header lines.
5421 4. The first argument for a ${extract expansion item is the key name or field
5422 number. Leading and trailing spaces in this item were not being ignored,
5423 causing some misleading effects.
5425 5. When deliver_drop_privilege was set, single queue runner processes started
5426 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5427 same command in the process it spins off) were not dropping privilege.
5429 6. When the daemon running as "exim" started a queue runner, it always
5430 re-executed Exim in the spun-off process. This is a waste of effort when
5431 deliver_drop_privilege is set. The new process now just calls the
5432 queue-runner function directly.
5438 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5441 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5442 ignored. Though the use of -oP was forcing the writing of a pid file, it
5443 was always written to the default place.
5445 3. If the message "no IP address found for host xxxx" is generated during
5446 incoming verification, it is now followed by identification of the incoming
5447 connection (so you can more easily find what provoked it).
5449 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5451 5. Added some features to "harden" Exim a bit more against certain attacks:
5453 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5454 be put in Local/Makefile. This is like the never_users runtime option,
5455 but it cannot be overridden. The default setting is "root".
5457 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5458 prefix string with which any file named in a -C command line option
5461 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5462 is retained for -C and -D only if the caller of Exim is root. Without
5463 it, the exim user may also use -C and -D and retain privilege.
5465 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5466 command line option is disabled.
5468 6. Macro names set by the -D option must start with an upper case letter, just
5469 like macro names defined in the configuration file.
5471 7. Added "dereference=" facility to LDAP.
5473 8. Two instances of the typo "uknown" in the source files are fixed.
5475 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5476 the Configure-Makefile script screwed up while processing it.
5478 10. Incorporated PCRE 4.4.
5480 11. The SMTP synchronization check was not operating right at the start of an
5481 SMTP session. For example, it could not catch a HELO sent before the client
5482 waited for the greeting. There is now a check for outstanding input at the
5483 point when the greeting is written. Because of the duplex, asynchronous
5484 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5485 way, but not yet received, when the check is performed.
5487 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5488 on TCP/IP sockets, because this apparently causes some broken clients to
5491 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5492 unchanged) from the Cygwin maintainer.
5494 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5495 was supported instead of testing for mbx. Effectively a typo.
5497 15. The spa authenticator server code was not checking that the input it
5498 received was valid base64.
5500 16. The debug output line for the "set" modifier in ACLs was not showing the
5501 name of the variable that was being set.
5503 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5505 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5506 lookup on its own. Something else was needed to trigger the lookup. For
5507 example, a match in host_lookup or the need for a host name in a host list.
5508 Now, if $sender_host_name is referenced and the host name has not yet been
5509 looked up, a lookup is performed. If the lookup fails, the variable remains
5510 empty, and $host_lookup_failed is set to "1".
5512 19. Added "eqi" as a case-independent comparison operator.
5514 20. The saslauthd authentication condition could segfault if neither service
5515 nor realm was specified.
5517 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5518 error message that was logged was misleading, and incoming SMTP
5519 connections were dropped. The message is now more accurate, and temporary
5520 errors are given to SMTP connections.
5522 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5523 (except RSET) with a 421 error, until QUIT is received. However, it was
5524 failing to send a response to QUIT.
5526 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5527 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5528 ACL is now run after the helo_try_verify_hosts code.
5530 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5531 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5532 case-independent, so other case variants are also recognized). Apparently
5533 some systems use these upper case variants.
5535 25. If more than two messages were waiting for the same host, and a transport
5536 filter was specified for the transport, Exim sent two messages over the
5537 same TCP/IP connection, and then failed with "socket operation on non-
5538 socket" when it tried to send the third.
5540 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5542 27. The extern definition of crypt16() in expand.c was not being excluded when
5543 the OS had its own crypt16() function.
5545 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5546 as a preferred synonym for return_size_limit, both as an option and as an
5549 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5551 30. Changed the default configuration ACL to relax the local part checking rule
5552 for addresses that are not in any local domains. For these addresses,
5553 slashes and pipe symbols are allowed within local parts, but the sequence
5554 /../ is explicitly forbidden.
5556 31. SPA server authentication was not clearing the challenge buffer before
5559 32. log_message in a "warn" ACL statement was writing to the reject log as
5560 well as to the main log, which contradicts the documentation and doesn't
5561 seem right (because no rejection is happening). So I have stopped it.
5563 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5564 However, I am unable to do any testing of this.
5566 34. Fixed an infelicity in the appendfile transport. When checking directories
5567 for a mailbox, to see if any needed to be created, it was accidentally
5568 using path names with one or more superfluous leading slashes; tracing
5569 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5571 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5572 contents are added to the log line that is written for every discarded
5573 recipient. (Previously a log_message setting was ignored.)
5575 36. The ${quote: operator now quotes the string if it is empty.
5577 37. The install script runs exim in order to find its version number. If for
5578 some reason other than non-existence or emptiness, which it checks, it
5579 could not run './exim', it was installing it with an empty version number,
5580 i.e. as "exim-". This error state is now caught, and the installation is
5583 38. An argument was missing from the function that creates an error message
5584 when Exim fails to connect to the socket for saslauthd authentication.
5585 This could cause Exim to crash, or give a corrupted message.
5587 39. Added isip, isip4, and isip6 to ${if conditions.
5589 40. The ACL variables $acl_xx are now saved with the message, and can be
5590 accessed later in routers, transports, and filters.
5592 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5593 strings in the file are not string-expanded.
5595 42. If a MAIL command specified a SIZE value that was too large to fit into an
5596 int variable, the check against message_size_limit failed. Such values are
5597 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5598 day this will have to be increased, but I don't think I want to be around
5599 when emails are that large.
5606 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5607 iconv() is not standard in FreeBSD.
5609 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5610 IPv6 enabled. The observed symptom was a segmentation fault on return from
5611 the function os_common_find_running_interfaces() in src/os.c.
5613 3. In the check_special_case() function in daemon.c I had used "errno" as an
5614 argument name, which causes warnings on some systems. This was basically a
5615 typo, since it was named "eno" in the comments!
5617 4. The code that waits for the clock to tick (at a resolution of some fraction
5618 of a second) so as to ensure message-id uniqueness was always waiting for
5619 at least one whole tick, when it could have waited for less. [This is
5620 almost certainly not relevant at current processor speeds, where it is
5621 unlikely to ever wait at all. But we try to future-proof.]
5623 5. The function that sleeps for a time interval that includes fractions of a
5624 second contained a race. It did not block SIGALRM between setting the
5625 timer, and suspending (a couple of lines later). If the interval was short
5626 and the sigsuspend() was delayed until after it had expired, the suspension
5627 never ended. On busy systems this could lead to processes getting stuck for
5630 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5631 process, before it forks any delivery processes. The open lookup caching
5632 mechanism meant that the open file or database connection was passed into
5633 the delivery process. The problem was that delivery processes always tidy
5634 up cached lookup data. This could cause a problem for the next delivery
5635 process started by the queue runner, because the external queue runner
5636 process does not know about the closure. So the next delivery process
5637 still has data in the lookup cache. In the case of a file lookup, there was
5638 no problem because closing a file descriptor in a subprocess doesn't affect
5639 the parent. However, if the lookup was caching a connection to a database,
5640 the connection was closed, and the second delivery process was likely to
5641 see errors such as "PGSQL: query failed: server closed the connection
5642 unexpectedly". The problem has been fixed by closing all cached lookups
5643 in a queue runner before running a delivery process.
5645 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5646 seem to have the "const" qualifier which it has on other OS. I've
5649 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5650 *of the same type* (client or server) with the same public name that an
5651 error should be diagnosed.
5653 9. When Exim looked up a host name for an IP address, but failed to find the
5654 original IP address when looking up the host name (a safety check), it
5655 output the message "<ip address> does not match any IP for NULL", which was
5656 confusing, to say the least. The bug was that the host name should have
5657 appeared instead of "NULL".
5659 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5660 user that is built into the binary, and the -C or -D options is used, root
5661 privilege is dropped before the configuration file is read. In addition,
5662 logging is switched to stderr instead of the normal log files. If the
5663 configuration then re-defines the Exim user, the unprivileged environment
5664 is probably not what is expected, so Exim logs a panic warning message (but
5667 However, if deliver_drop_privilege is set, the unprivileged state may well
5668 be exactly what is intended, so the warning has been cut out in that case,
5669 and Exim is allowed to try to write to its normal log files.
5675 1. smtp_return_error_details was not giving details for temporary sender
5676 or receiver verification errors.
5678 2. Diagnose a configuration error if two authenticators have the same public
5681 3. Exim used not to create the message log file for a message until the first
5682 delivery attempt. This could be confusing when incoming messages were held
5683 for policy or load reasons. The message log file is now created at the time
5684 the message is received, and an initial "Received" line is written to it.
5686 4. The automatically generated man page for command line options had a minor
5687 bug that caused no ill effects; however, a more serious problem was that
5688 the procedure for building the man page automatically didn't always
5689 operate. Consequently, release 4.20 contains an out-of-date version. This
5690 shouldn't happen again.
5692 5. When building Exim with embedded Perl support, the script that builds the
5693 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5694 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5696 6. The freeze_tell option was not being used for messages that were frozen on
5697 arrival, either by an ACL or by local_scan().
5699 7. Added the smtp_incomplete_transaction log selector.
5701 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5702 was accepting AUTH without a new EHLO.
5704 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5705 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5708 10. Logging of TCP/IP connections (when configured) now happens in the main
5709 daemon process instead of the child process, so that the TCP/IP connection
5710 count is more accurate (but it can never be perfect).
5712 11. The use of "drop" in a nested ACL was not being handled correctly in the
5713 outer ACL. Now, if condition failure induced by the nested "drop" causes
5714 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5715 or "require"), the connection is dropped.
5717 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5718 that yield "discard" can now be used with an "accept" or a "discard" verb,
5719 but an error is generated for any others (because I can't see a useful way
5720 to define what should happen).
5722 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5723 are now processed in the same way as lines in the Exim configuration file.
5724 In particular, continuation lines are supported.
5726 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5728 15. Added -ti meaning -t -i.
5730 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5731 domains, and warn by logging if others are found.
5733 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5734 returned from GIFCONF when Exim is trying to find the list of interfaces on
5735 a host. The code in os.c has been modified to copy each ifreq to an aligned
5736 structure in all cases.
5738 Also, in some cases, the returned ifreq's were being copied to a 'struct
5739 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5740 means the last couple of bytes of an IPv6 address could be chopped if the
5741 ifreq contained only a normal sockaddr (14 bytes storage).
5743 18. Named domain lists were not supported in the hosts_treat_as_local option.
5744 An entry such as +xxxx was not recognized, and was treated as a literal
5747 19. Ensure that header lines added by a DATA ACL are included in the reject log
5748 if the ACL subsequently rejects the message.
5750 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5751 MD5 (which is deprecated).
5753 21. When testing a filter file using -bf, Exim was writing a message when it
5754 took the sender from a "From " line in the message, but it was not doing so
5755 when it took $return_path from a Return-Path: header line. It now does.
5757 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5758 with a valid header line field name (a series of printing characters
5759 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5761 23. Changed "disc" in the source to "disk" to conform to the documentation and
5762 the book and for uniformity.
5764 24. Ignore Sendmail's -Ooption=value command line item.
5766 25. When execve() failed while trying to run a command in a pipe transport,
5767 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5768 could be confused with a return value of 69 from the command itself. This
5769 has been changed to 127, the value the shell returns if it is asked to run
5770 a non-existent command. The wording for the related log line suggests a
5771 non-existent command as the problem.
5773 26. If received_header_text expands to an empty string, do not add a Received:
5774 header line to the message. (Well, it adds a token one on the spool, but
5775 marks it "old" so that it doesn't get used or transmitted.)
5777 27. Installed eximstats 1.28 (addition of -nt option).
5779 28. There was no check for failure on the call to getsockname() in the daemon
5780 code. This can fail if there is a shortage of resources on the system, with
5781 ENOMEM, for example. A temporary error is now given on failure.
5783 29. Contrary to the C standard, it seems that in some environments, the
5784 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5785 program. Exim now does this explicitly; it affects the formatting of
5786 timestamps using strftime().
5788 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5789 complaints. I've now initialized all the variables, to avoid this.
5791 32. Header lines added by a system filter were not being "seen" during
5792 transport-time rewrites.
5794 33. The info_callback() function passed to OpenSSL is set up with type void
5795 (*)(SSL *, int, int), as described somewhere. However, when calling the
5796 function (actually a macro) that sets it up, the type void(*)() is
5797 expected. I've put in a cast to prevent warnings from picky compilers.
5799 34. If a DNS black list lookup found a CNAME record, but there were no A
5800 records associated with the domain it pointed at, Exim crashed.
5802 35. If a DNS black list lookup returned more than one A record, Exim ignored
5803 all but the first. It now scans all returned addresses if a particular IP
5804 value is being sought. In this situation, the contents of the
5805 $dnslist_value variable are a list of all the addresses, separated by a
5808 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5809 to accept a host name and all its aliases if the forward lookup for any of
5810 them yielded the IP address of the incoming connection. Now it accepts only
5811 those names whose forward lookup yields the correct IP address. Any other
5812 names are discarded. This closes a loophole whereby a rogue DNS
5813 administrator could create reverse DNS records to break through a
5814 wildcarded host restriction in an ACL.
5816 37. If a user filter or a system filter that ran in a subprocess used any of
5817 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5818 the wrong values were passed to the pipe command ($thisaddress had the
5819 value of $0, $0 had the value of $1, etc). This bug was introduced by
5820 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5822 38. Improved the line breaking for long SMTP error messages from ACLs.
5823 Previously, if there was no break point between 40 and 75 characters, Exim
5824 left the rest of the message alone. Two changes have been made: (a) I've
5825 reduced the minimum length to 35 characters; (b) if it can't find a break
5826 point between 35 and 75 characters, it looks ahead and uses the first one
5827 that it finds. This may give the occasional overlong line, but at least the
5828 remaining text gets split now.
5830 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5831 file descriptors might be low, and that setting 1000 would always raise it.
5832 It turns out that in some environments, the limit is already over 1000 and
5833 that lowering it causes trouble. So now Exim takes care not to decrease it.
5835 40. When delivering a message, the value of $return_path is set to $sender_
5836 address at the start of routing (routers may change the value). By an
5837 oversight, this default was not being set up when an address was tested by
5838 -bt or -bv, which affected the outcome if any router or filter referred to
5841 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5842 log only when "message" or "log_message" are set. However, if one of the
5843 conditions was an address verification, or a call to a nested ACL, the
5844 messages generated by the underlying test were being passed through. This
5845 no longer happens. The underlying message is available in $acl_verify_
5846 message for both "message" and "log_message" expansions, so it can be
5847 passed through if needed.
5849 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5850 new expansion $bh_ to give the encoded byte string without charset
5851 translation. Translation happens only if iconv() is available; HAVE_ICONV
5852 indicates this at build time. HEADERS_CHARSET gives the charset to
5853 translate to; headers_charset can change it in the configuration, and
5854 "headers charset" can change it in an individual filter file.
5856 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5857 that creates RFC 2047 encoded "words" labels them as that charset instead
5858 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5859 expansion operator; (ii) when Exim creates a From: line for a local
5860 message; (iii) when a header line is rewritten to include a "phrase" part.
5862 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5863 buggy, causing it to skip the first lines of messages whose message ID
5864 ended in 'D'. This would not have bitten before Exim release 4.14, because
5865 message IDs were unlikely to end in 'D' before then. The effect was to have
5866 incorrect size information for certain domains.
5868 45. #include "config.h" was missing at the start of the crypt16.c module. This
5869 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5872 46. If there was a timeout during a "random" callout check, Exim treated it as
5873 a failure of the random address, and carried on sending RSET and the real
5874 address. If the delay was just some slowness somewhere, the response to the
5875 original RCPT would be taken as a response to RSET and so on, causing
5876 mayhem of various kinds.
5878 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5879 when I implemented it. It didn't allow for the fact that some option values
5880 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5881 the right test for positive values.
5883 48. Domain names in DNS records are case-independent. Exim always looks them up
5884 in lower case. Some resolvers return domain names in exactly the case they
5885 appear in the zone file, that is, they may contain uppercase letters. Not
5886 all resolvers do this - some return always lower case. Exim was treating a
5887 change of case by a resolver as a change of domain, similar to a widening
5888 of a domain abbreviation. This triggered its re-routing code and so it was
5889 trying to route what was effectively the same domain again. This normally
5890 caused routing to fail (because the router wouldn't handle the domain
5891 twice). Now Exim checks for this case specially, and just changes the
5892 casing of the domain that it ultimately uses when it transmits the message
5895 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5898 50. If a filter generated a file delivery with a non-absolute name (possible if
5899 no home directory exists for the router), the forbid_file option was not
5902 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5903 the existing equality matching.
5905 52. Exim was using ints instead of ino_t variables in some places where it was
5906 dealing with inode numbers.
5908 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5909 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5910 called TMPDIR, and if it finds it is different, it changes its value.
5912 54. The smtp_printf() function is now made available to local_scan() so
5913 additional output lines can be written before returning. There is also an
5914 smtp_fflush() function to enable the detection of a dropped connection.
5915 The variables smtp_input and smtp_batched_input are exported to
5918 55. Changed the default runtime configuration: the message "Unknown user"
5919 has been removed from the ACL, and instead placed on the localuser router,
5920 using the cannot_route_message feature. This means that any verification
5921 failures that generate their own messages won't get overridden. Similarly,
5922 the "Unrouteable address" message that was in the ACL for unverifiable
5923 relay addresses has also been removed.
5925 56. Added hosts_avoid_esmtp to the smtp transport.
5927 57. The exicyclog script was not checking for the esoteric option
5928 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5929 will work only if exicyclog is run under the appropriate euid.
5931 58. Following a discussion on the list, the rules by which Exim recognises line
5932 endings on incoming messages have been changed. The -dropcr and drop_cr
5933 options are now no-ops, retained only for backwards compatibility. The
5934 following line terminators are recognized: LF CRLF CR. However, special
5935 processing applies to CR:
5937 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5938 nor a local message in the state where . is a terminator.
5940 (ii) If a bare CR is encountered in a header line, an extra space is added
5941 after the line terminator so as not to end the header. The reasoning
5942 behind this is that bare CRs in header lines are most likely either
5943 to be mistakes, or people trying to play silly games.
5945 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5946 was being incorrectly given as 18 bytes larger than it should have been.
5947 This is a VOB (very old bug).
5949 60. This may never have affected anything current, but just in case it has:
5950 When the local host is found other than at the start of a list of hosts,
5951 the local host, those with the same MX, and any that follow, are discarded.
5952 When the list in question was part of a longer list of hosts, the following
5953 hosts (not currently being processed) were also being discarded. This no
5954 longer happens. I'm not sure if this situation could ever has previously
5957 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5960 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5961 header. This is something that is recommended in a new Internet Draft, and
5962 is something that is documented as being done by Sendmail. There are two
5963 possible values. For messages generated by the autoreply transport, Exim
5966 Auto-Submitted: auto-replied
5968 whereas for all other generated messages (e.g. bounces) it adds
5970 Auto-Submitted: auto-generated
5972 63. The "personal" condition in filters now includes a test for the
5973 Auto-Submitted: header. If it contains the string "auto-" the message it
5974 not considered personal.
5976 64. Added rcpt_include_affixes as a generic transport option.
5978 65. Added queue_only_override (default true).
5980 66. Added the syslog_duplication option.
5982 67. If what should have been the first header line of a message consisted of
5983 a space followed by a colon, Exim was mis-interpreting it as a header line.
5984 It isn't of course - it is syntactically invalid and should therefore be
5985 treated as the start of the message body. The misbehaviour could have
5986 caused a number of strange effects, including loss of data in subsequent
5987 header lines, and spool format errors.
5989 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5990 client host had authenticated. This control can now be exercised by an ACL
5991 for more flexibility.
5993 69. By default, callouts do not happen when testing with -bh. There is now a
5994 variant, -bhc, which does actually run the callout code, including
5995 consulting and updating the callout cache.
5997 70. Added support for saslauthd authentication, courtesy of Alexander
6000 71. If statvfs() failed on the spool or log directories while checking their
6001 size for availability, Exim confusingly gave the error "space shortage".
6002 Furthermore, in debugging mode it crashed with a floating point exception.
6003 These checks are done if check_{spool,log}_{space,inodes} are set, and when
6004 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
6005 serious problem, Exim now writes to the main and panic logs when this
6006 happens, with details of the failure. It then refuses to accept the
6007 incoming message, giving the message "spool directory problem" or "log
6008 directory problem" with a 421 code for SMTP messages.
6010 72. When Exim is about to re-exec itself, it ensures that the file descriptors
6011 0, 1, and 2 exist, because some OS complain for execs without them (see
6012 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
6013 descriptors. However, the code omitted to check that the open succeeded,
6014 causing mysterious errors if for some reason the permissions on /dev/null
6015 got screwed. Now Exim writes a message to the main and panic logs, and
6016 bombs out if it can't open /dev/null.
6018 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
6019 interact so that it is all more flexible. It is supposed to remain
6020 backwards compatible. Also added extra_local_interfaces.
6022 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
6023 to bomb out with an assertion failure - to the client this appears as a
6024 connection drop. This problem occurs in the part of the code that was taken
6025 from the Samba project. Fortunately, the assertion is in a very simple
6026 function, so I have fixed this by reproducing the function inline in the
6027 one place where it is called, and arranging for authentication to fail
6028 instead of killing the process with assert().
6030 75. The SPA client code was not working when the server requested OEM rather
6031 than Unicode encoding.
6033 76. Added code to make require_files with a specific uid setting more usable in
6034 the case where statting the file as root fails - usually a non-root-mounted
6035 NFS file system. When this happens and the failure is EACCES, Exim now
6036 forks a subprocess and does the per-uid checking as the relevant uid.
6038 77. Added process_log_path.
6040 78. If log_file_path was not explicitly set, a setting of check_log_space or
6041 check_log_inodes was ignored.
6043 79. If a space check for the spool or log partitions fails, the incident is now
6044 logged. Of course, in the latter case the data may get lost...
6046 80. Added the %p formatting code to string_format() so that it can be used to
6047 print addresses in debug_print(). Adjusted all the address printing in the
6048 debugging in store.c to use %p rather than %d.
6050 81. There was a concern that a line of code in smtp_in.c could overflow a
6051 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6052 initially expressed, the concern was not well-founded, because trailing
6053 spaces are removed early. However, if the trailing spaces were followed by
6054 a NULL, they did not get removed, so the overflow was possible. Two fixes
6057 (a) I re-wrote the offending code in a cleaner fashion.
6058 (b) If an incoming SMTP command contains a NULL character, it is rejected
6061 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6062 runs initgroups(), so that if the Exim user is in any additional groups,
6063 they will be used during message reception.
6069 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob