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.
30 JH/02 Add ${listextract {number}{list}{success}{fail}}.
32 TL/03 Bugzilla 1433: Fix DMARC SEGV with specific From header contents.
33 Properly escape header and check for NULL return.
35 PP/01 Continue incomplete 4.82 PP/19 by fixing docs too: use dns_dnssec_ok
38 JH/03 Bugzilla 1157: support log_selector smtp_confirmation for lmtp.
40 TL/04 Add verify = header_names_ascii check to reject email with non-ASCII
41 characters in header names, implemented as a verify condition.
42 Contributed by Michael Fischer v. Mollard.
44 TL/05 Rename SPF condition results err_perm and err_temp to standardized
45 results permerror and temperror. Previous values are deprecated but
46 still accepted. In a future release, err_perm and err_temp will be
47 completely removed, which will be a backward incompatibility if the
48 ACL tests for either of these two old results. Patch contributed by
49 user bes-internal on the mailing list.
51 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
53 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
54 selectors, in both main and reject logs.
56 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
59 JH/07 Add malware type "sock" for talking to simple daemon.
61 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
63 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
64 routers/transports under cutthrough routing.
66 JH/10 Bugzilla 1005: ACL "condition =" should accept values which are negative
67 numbers. Touch up "bool" conditional to keep the same definition.
69 TL/06 Remove duplicated language in spec file from 4.82 TL/16.
71 JH/11 Add dnsdb tlsa lookup. From Todd Lyons.
73 JH/12 Expand items in router/transport headers_add or headers_remove lists
74 individually rather than the list as a whole. Bug 1452.
76 Required for reasonable handling of multiple headers_ options when
77 they may be empty; requires that headers_remove items with embedded
78 colons must have them doubled (or the list-separator changed).
80 TL/07 Add new dmarc expansion variable $dmarc_domain_policy to directly
81 view the policy declared in the DMARC record. Currently, $dmarc_status
82 is a combined value of both the record presence and the result of the
85 JH/13 Fix handling of $tls_cipher et.al. in (non-verify) transport. Bug 1455.
87 JH/14 New options dnssec_request_domains, dnssec_require_domains on the
88 dnslookup router and the smtp transport (applying to the forward
91 TL/08 Bugzilla 1453: New LDAP "SERVERS=" option allows admin to override list
92 of ldap servers used for a specific lookup. Patch provided by Heiko
95 JH/18 New options dnssec_lax, dnssec_strict on dnsdb lookups.
96 New variable $lookup_dnssec_authenticated for observability.
98 TL/09 Bugzilla 609: Add -C option to exiqgrep, specify which exim.conf to use.
99 Patch submitted by Lars Timman.
101 JH/19 EXPERIMENTAL_OCSP support under GnuTLS. Bug 1459.
107 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
109 PP/02 Make -n do something, by making it not do something.
110 When combined with -bP, the name of an option is not output.
112 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
115 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
116 $sender_host_name and config options to manage this, and basic check
119 PP/05 DSCP support for outbound connections and control modifier for inbound.
121 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
122 (Only plugin which currently uses this is kerberos4, which nobody should
123 be using, but we should make it available and other future plugins might
124 conceivably use it, even though it would break NAT; stuff *should* be
125 using channel bindings instead).
127 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
128 name; added for Sendmail compatibility; requires admin caller.
129 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
130 just ignore it); requires trusted caller.
131 Also parse but ignore: -Ac -Am -X<logfile>
134 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
136 TL/02 Add +smtp_confirmation as a default logging option.
138 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
139 Patch by Magnus Holmgren from 2007-02-20.
141 TL/04 Bugzilla 1281 - Spec typo.
142 Bugzilla 1283 - Spec typo.
143 Bugzilla 1290 - Spec grammar fixes.
145 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
147 TL/06 Add Experimental DMARC support using libopendmarc libraries.
149 TL/07 Fix an out of order global option causing a segfault. Reported to dev
150 mailing list by by Dmitry Isaikin.
152 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
154 JH/02 Support "G" suffix to numbers in ${if comparisons.
156 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
158 NM/01 Bugzilla 1197 - Spec typo
159 Bugzilla 1196 - Spec examples corrections
161 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
163 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
164 gnutls_enable_pkcs11, but renamed to more accurately indicate its
167 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
168 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
170 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
171 "acl {{name}{arg}...}", and optional args on acl condition
174 JH/05 Permit multiple router/transport headers_add/remove lines.
176 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
178 JH/07 Avoid using a waiting database for a single-message-only transport.
179 Performance patch from Paul Fisher. Bugzilla 1262.
181 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
184 JH/09 Add $headers_added variable, with content from use of ACL modifier
185 add_header (but not yet added to the message). Bugzilla 199.
187 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
188 Pulled from Bugzilla 817 by Wolfgang Breyha.
190 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
192 (nb: this is the same fix as in Exim 4.80.1)
194 JH/11 Add A= logging on delivery lines, and a client_set_id option on
197 JH/12 Add optional authenticated_sender logging to A= and a log_selector
200 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
202 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
203 advertise SMTP AUTH mechanism to us, instead of a generic
204 protocol violation error. Also, make Exim more robust to bad
205 data from the Dovecot auth socket.
207 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
209 When a queue runner is handling a message, Exim first routes the
210 recipient addresses, during which it prunes them based on the retry
211 hints database. After that it attempts to deliver the message to
212 any remaining recipients. It then updates the hints database using
215 So if a recipient address works intermittently, it can get repeatedly
216 deferred at routing time. The retry hints record remains fresh so the
217 address never reaches the final cutoff time.
219 This is a fairly common occurrence when a user is bumping up against
220 their storage quota. Exim had some logic in its local delivery code
221 to deal with this. However it did not apply to per-recipient defers
222 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
224 This change adds a proper retry rule check during routing so that the
225 final cutoff time is checked against the message's age. We only do
226 this check if there is an address retry record and there is not a
227 domain retry record; this implies that previous attempts to handle
228 the address had the retry_use_local_parts option turned on. We use
229 this as an approximation for the destination being like a local
230 delivery, as in LMTP.
232 I suspect this new check makes the old local delivery cutoff check
233 redundant, but I have not verified this so I left the code in place.
235 TF/02 Correct gecos expansion when From: is a prefix of the username.
237 Test 0254 submits a message to Exim with the header
241 When I ran the test suite under the user fanf2, Exim expanded
242 the header to contain my full name, whereas it should have added
243 a Resent-Sender: header. It erroneously treats any prefix of the
244 username as equal to the username.
246 This change corrects that bug.
248 GF/01 DCC debug and logging tidyup
249 Error conditions log to paniclog rather than rejectlog.
250 Debug lines prefixed by "DCC: " to remove any ambiguity.
252 TF/03 Avoid unnecessary rebuilds of lookup-related code.
254 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
255 Bug spotted by Jeremy Harris; was flawed since initial commit.
256 Would have resulted in OCSP responses post-SNI triggering an Exim
257 NULL dereference and crash.
259 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
261 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
262 Bug detection, analysis and fix by Samuel Thibault.
263 Bugzilla 1331, Debian bug #698092.
265 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
267 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
268 Server implementation by Todd Lyons, client by JH.
269 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
270 config variable "prdr_enable" controls whether the server
271 advertises the facility. If the client requests PRDR a new
272 acl_data_smtp_prdr ACL is called once for each recipient, after
273 the body content is received and before the acl_smtp_data ACL.
274 The client is controlled by bolth of: a hosts_try_prdr option
275 on the smtp transport, and the server advertisement.
276 Default client logging of deliveries and rejections involving
277 PRDR are flagged with the string "PRDR".
279 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
280 fclose(). Diagnosis by Todd Lyons.
282 PP/17 Update configure.default to handle IPv6 localhost better.
283 Patch by Alain Williams (plus minor tweaks).
286 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
287 This is now consistent with GnuTLS, and is now documented: the
288 previous undocumented portable approach to treating the option as
289 unset was to force an expansion failure. That still works, and
290 an empty string is now equivalent.
292 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
293 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
294 not performing validation itself.
296 PP/20 Added force_command boolean option to pipe transport.
297 Patch from Nick Koston, of cPanel Inc.
299 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
302 TF/04 Added udpsend ACL modifer and hexquote expansion operator
304 PP/21 Fix eximon continuous updating with timestamped log-files.
305 Broken in a format-string cleanup in 4.80, missed when I repaired the
306 other false fix of the same issue.
307 Report and fix from Heiko Schlichting.
310 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
311 Report from Prashanth Katuri.
313 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
314 It's SecureTransport, so affects any MacOS clients which use the
315 system-integrated TLS libraries, including email clients.
317 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
318 using a MIME ACL for non-SMTP local injection.
319 Report and assistance in diagnosis by Warren Baker.
321 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
323 JH/16 Fix comparisons for 64b. Bugzilla 1385.
325 TL/09 Add expansion variable $authenticated_fail_id to keep track of
326 last id that failed so it may be referenced in subsequent ACL's.
328 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
331 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
332 ldap library initialization, allowing self-signed CA's to be
333 used. Also properly sets require_cert option later in code by
334 using NULL (global ldap config) instead of ldap handle (per
335 session). Bug diagnosis and testing by alxgomz.
337 TL/12 Enhanced documentation in the ratelimit.pl script provided in
338 the src/util/ subdirectory.
340 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
341 renamed to Transport Post Delivery Action by Jeremy Harris, as
344 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
345 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
346 redis_servers = needs to be configured which will be used by the redis
347 lookup. Patch from Warren Baker, of The Packet Hub.
349 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
351 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
352 hostname or reverse DNS when processing a host list. Used suggestions
353 from multiple comments on this bug.
355 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
357 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
358 Missed a few lines, added it to make the runtest require no keyboard
361 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
362 contains upper case chars. Make router use caseful_local_part.
364 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
365 support when GnuTLS has been built with p11-kit.
371 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
373 This, or similar/improved, will also be change PP/11 of 4.82.
379 PP/01 Handle short writes when writing local log-files.
380 In practice, only affects FreeBSD (8 onwards).
381 Bugzilla 1053, with thanks to Dmitry Isaikin.
383 NM/01 Bugzilla 949 - Documentation tweak
385 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
388 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
390 PP/02 Implemented gsasl authenticator.
392 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
394 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
395 `pkg-config foo` for cflags/libs.
397 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
398 with rest of GSASL and with heimdal_gssapi.
400 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
401 `pkg-config foo` for cflags/libs for the TLS implementation.
403 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
404 properties get this fed in as external SSF. A number of robustness
405 and debugging improvements to the cyrus_sasl authenticator.
407 PP/08 cyrus_sasl server now expands the server_realm option.
409 PP/09 Bugzilla 1214 - Log authentication information in reject log.
410 Patch by Jeremy Harris.
412 PP/10 Added dbmjz lookup type.
414 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
416 PP/12 MAIL args handles TAB as well as SP, for better interop with
417 non-compliant senders.
418 Analysis and variant patch by Todd Lyons.
420 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
421 Bug report from Lars Müller <lars@samba.org> (via SUSE),
422 Patch from Dirk Mueller <dmueller@suse.com>
424 PP/13 tls_peerdn now print-escaped for spool files.
425 Observed some $tls_peerdn in wild which contained \n, which resulted
426 in spool file corruption.
428 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
429 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
430 or write after TLS renegotiation, which otherwise led to messages
433 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
434 as a tracking header (ie: a signed header comes before the signature).
435 Patch from Wolfgang Breyha.
437 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
438 comma-sep list; embedded commas doubled.
440 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
442 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
444 Report and patch from Dmitry Banschikov.
446 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
447 Removed SSL_clear() after SSL_new() which led to protocol negotiation
448 failures. We appear to now support TLS1.1+ with Exim.
450 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
451 lets Exim select keys and certificates based upon TLS SNI from client.
452 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
453 before an outbound SMTP session. New log_selector, +tls_sni.
455 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
456 NULL dereference. Report and patch from Alun Jones.
458 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
459 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
460 Not seeing resolver debug output on NetBSD, but suspect this is a
461 resolver implementation change.
463 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
464 Left warnings. Added "eximon gdb" invocation mode.
466 PP/21 Defaulting "accept_8bitmime" to true, not false.
468 PP/22 Added -bw for inetd wait mode support.
470 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
471 locate the relevant includes and libraries. Made this the default.
473 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
474 Bugzilla 1246, report and most of solution from Tomasz Kusy.
476 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
477 This may cause build issues on older platforms.
479 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
480 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
481 gnutls_require_mac & gnutls_require_protocols (no longer supported).
482 Added SNI support via GnuTLS too.
483 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
485 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
487 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
488 Applied second patch from Janne, implementing suggestion to default
489 multiple-strings-in-record handling to match SPF spec.
491 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
493 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
494 read-only, out of scope).
495 Patch from Wolfgang Breyha, report from Stuart Northfield.
497 PP/29 Fix three issues highlighted by clang analyser static analysis.
498 Only crash-plausible issue would require the Cambridge-specific
499 iplookup router and a misconfiguration.
500 Report from Marcin Mirosław.
502 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
504 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
505 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
506 As part of this, removing so much warning spew let me fix some minor
507 real issues in debug logging.
509 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
510 assignment on my part. Fixed.
512 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
513 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
514 Janne Snabb (who went above and beyond: thank you).
516 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
517 string otherwise requires a connection and a bunch more work and it's
518 relatively easy to get wrong. Should also expose TLS library linkage
521 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
522 64-bit ${eval} (JH/03).
524 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
525 GNU libc to support some of the 64-bit stuff, should not lead to
526 conflicts. Defined before os.h is pulled in, so if a given platform
527 needs to override this, it can.
529 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
530 protection layer was required, which is not implemented.
531 Bugzilla 1254, patch from Wolfgang Breyha.
533 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
534 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
535 tls_dhparam take prime identifiers. Also unbreak combination of
536 OpenSSL+DH_params+TLSSNI.
538 PP/39 Disable SSLv2 by default in OpenSSL support.
544 PP/01 Solaris build fix for Oracle's LDAP libraries.
545 Bugzilla 1109, patch from Stephen Usher.
547 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
549 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
552 TF/02 Fix a couple more cases where we did not log the error message
553 when unlink() failed. See also change 4.74-TF/03.
555 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
556 lock up or crash if it happened to be inside a call to libc when it
557 got a SIGUSR1 from exiwhat.
559 The SIGUSR1 handler appends the current process status to the process
560 log which is later printed by exiwhat. It used to use the general
561 purpose logging code to do this, but several functions it calls are
562 not safe for signals.
564 The new output code in the SIGUSR1 handler is specific to the process
565 log, and simple enough that it's easy to inspect for signal safety.
566 Removing some special cases also simplifies the general logging code.
567 Removing the spurious timestamps from the process log simplifies
570 TF/04 Improved ratelimit ACL condition.
572 The /noupdate option has been deprecated in favour of /readonly which
573 has clearer semantics. The /leaky, /strict, and /readonly update modes
574 are mutually exclusive. The update mode is no longer included in the
575 database key; it just determines when the database is updated. (This
576 means that when you upgrde Exim will forget old rate measurements.)
578 Exim now checks that the per_* options are used with an update mode that
579 makes sense for the current ACL. For example, when Exim is processing a
580 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
581 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
582 must specify per_mail/readonly. If you omit the update mode it defaults to
583 /leaky where that makes sense (as before) or /readonly where required.
585 The /noupdate option is now undocumented but still supported for
586 backwards compatibility. It is equivalent to /readonly except that in
587 ACLs where /readonly is required you may specify /leaky/noupdate or
588 /strict/noupdate which are treated the same as /readonly.
590 A useful new feature is the /count= option. This is a generalization
591 of the per_byte option, so that you can measure the throughput of other
592 aggregate values. For example, the per_byte option is now equivalent
593 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
595 The per_rcpt option has been generalized using the /count= mechanism
596 (though it's more complicated than the per_byte equivalence). When it is
597 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
598 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
599 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
600 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
601 Note that using per_rcpt with a non-readonly update mode in more than
602 one ACL will cause the recipients to be double-counted. (The per_mail
603 and per_byte options don't have this problem.)
605 The handling of very low rates has changed slightly. If the computed rate
606 is less than the event's count (usually one) then this event is the first
607 after a long gap. In this case the rate is set to the same as this event's
608 count, so that the first message of a spam run is counted properly.
610 The major new feature is a mechanism for counting the rate of unique
611 events. The new per_addr option counts the number of different
612 recipients that someone has sent messages to in the last time period. It
613 behaves like per_rcpt if all the recipient addresses are different, but
614 duplicate recipient addresses do not increase the measured rate. Like
615 the /count= option this is a general mechanism, so the per_addr option
616 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
617 example, measure the rate that a client uses different sender addresses
618 with the options per_mail/unique=$sender_address. There are further
619 details in the main documentation.
621 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
623 TF/06 Removed a few PCRE remnants.
625 TF/07 Automatically extract Exim's version number from tags in the git
626 repository when doing development or release builds.
628 PP/02 Raise smtp_cmd_buffer_size to 16kB.
629 Bugzilla 879. Patch from Paul Fisher.
631 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
632 Heavily based on revision 40f9a89a from Simon Arlott's tree.
635 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
637 PP/05 Variable $av_failed, true if the AV scanner deferred.
638 Bugzilla 1078. Patch from John Horne.
640 PP/06 Stop make process more reliably on build failure.
641 Bugzilla 1087. Patch from Heiko Schlittermann.
643 PP/07 Make maildir_use_size_file an _expandable_ boolean.
644 Bugzilla 1089. Patch from Heiko Schlittermann.
646 PP/08 Handle ${run} returning more data than OS pipe buffer size.
647 Bugzilla 1131. Patch from Holger Weiß.
649 PP/09 Handle IPv6 addresses with SPF.
650 Bugzilla 860. Patch from Wolfgang Breyha.
652 PP/10 GnuTLS: support TLS 1.2 & 1.1.
654 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
657 PP/11 match_* no longer expand right-hand-side by default.
658 New compile-time build option, EXPAND_LISTMATCH_RHS.
659 New expansion conditions, "inlist", "inlisti".
661 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
663 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
665 PP/14 fix log_write() format string regression from TF/03.
666 Bugzilla 1152. Patch from Dmitry Isaikin.
672 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
674 PP/02 Harmonised TLS library version reporting; only show if debugging.
675 Layout now matches that introduced for other libraries in 4.74 PP/03.
677 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
679 PP/04 New "dns_use_edns0" global option.
681 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
684 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
685 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
687 TK/01 Updated PolarSSL code to 0.14.2.
688 Bugzilla 1097. Patch from Andreas Metzler.
690 PP/07 Catch divide-by-zero in ${eval:...}.
693 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
696 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
697 format-string attack -- SECURITY: remote arbitrary code execution.
699 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
700 time unintentionally subject to list matching rules, letting the header
701 cause arbitrary Exim lookups (of items which can occur in lists, *not*
702 arbitrary string expansion). This allowed for information disclosure.
704 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
705 INT_MIN/-1 -- value coerced to INT_MAX.
711 NM/01 Workround for PCRE version dependency in version reporting
714 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
715 This fixes portability to compilers other than gcc, notably
716 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
718 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
719 makefiles for portability to HP-UX and POSIX correctness.
721 PP/01 Permit LOOKUP_foo enabling on the make command-line.
722 Also via indented variable definition in the Makefile.
723 (Debugging by Oliver Heesakkers).
725 PP/02 Restore caching of spamd results with expanded spamd_address.
726 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
728 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
729 Improves build reliability. Fix from: Frank Elsner
731 NM/02 Fix wide character breakage in the rfc2047 coding
732 Fixes bug 1064. Patch from Andrey N. Oktyabrski
734 NM/03 Allow underscore in dnslist lookups
735 Fixes bug 1026. Patch from Graeme Fowler
737 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
738 Code patches from Adam Ciarcinski of NetBSD.
740 NM/04 Fixed exiqgrep to cope with mailq missing size issue
743 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
744 is logged, to avoid truncation. Patch from John Horne.
746 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
747 Patch from Jakob Hirsch.
749 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
750 SQL string expansion failure details.
751 Patch from Andrey Oktyabrski.
753 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
754 Patch from Simon Arlott.
756 PP/09 New lookups functionality failed to compile on old gcc which rejects
757 extern declarations in function scope.
758 Patch from Oliver Fleischmann
760 PP/10 Use sig_atomic_t for flags set from signal handlers.
761 Check getgroups() return and improve debugging.
762 Fixed developed for diagnosis in bug 927 (which turned out to be
765 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
766 Patch from Mark Zealey.
768 PP/12 Bugzilla 1056: Improved spamd server selection.
769 Patch from Mark Zealey.
771 PP/13 Bugzilla 1086: Deal with maildir quota file races.
772 Based on patch from Heiko Schlittermann.
774 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
775 Patch from Uwe Doering, sign-off by Michael Haardt.
777 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
778 variable declaration deep within a block. Bug and patch from
781 PP/15 lookups-Makefile IRIX compatibilty coercion.
783 PP/16 Make DISABLE_DKIM build knob functional.
785 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
786 Patch by Simon Arlott
788 TF/03 Fix valgrind.h portability to C89 compilers that do not support
789 variable argument macros. Our copy now differs from upstream.
795 TF/01 Failure to get a lock on a hints database can have serious
796 consequences so log it to the panic log.
798 TF/02 Log LMTP confirmation messages in the same way as SMTP,
799 controlled using the smtp_confirmation log selector.
801 TF/03 Include the error message when we fail to unlink a spool file.
803 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
804 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
805 for maintaining out-of-tree patches for some time.
807 PP/01 Bugzilla 139: Documentation and portability issues.
808 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
809 Handle per-OS dynamic-module compilation flags.
811 PP/02 Let /dev/null have normal permissions.
812 The 4.73 fixes were a little too stringent and complained about the
813 permissions on /dev/null. Exempt it from some checks.
814 Reported by Andreas M. Kirchwitz.
816 PP/03 Report version information for many libraries, including
817 Exim version information for dynamically loaded libraries. Created
818 version.h, now support a version extension string for distributors
819 who patch heavily. Dynamic module ABI change.
821 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
822 privilege escalation vulnerability whereby the Exim run-time user
823 can cause root to append content of the attacker's choosing to
826 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
829 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
830 If dropping privileges for untrusted macros, we disabled normal logging
831 on the basis that it would fail; for the Exim run-time user, this is not
832 the case, and it resulted in successful deliveries going unlogged.
833 Fixed. Reported by Andreas Metzler.
839 PP/01 Date: & Message-Id: revert to normally being appended to a message,
840 only prepend for the Resent-* case. Fixes regression introduced in
841 Exim 4.70 by NM/22 for Bugzilla 607.
843 PP/02 Include check_rfc2047_length in configure.default because we're seeing
844 increasing numbers of administrators be bitten by this.
846 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
848 PP/03 Bugzilla 994: added openssl_options main configuration option.
850 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
852 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
854 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
856 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
857 without a peer certificate, leading to a segfault because of an
858 assumption that peers always have certificates. Be a little more
859 paranoid. Problem reported by Martin Tscholak.
861 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
862 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
863 NB: ClamAV planning to remove STREAM in "middle of 2010".
864 CL also introduces -bmalware, various -d+acl logging additions and
865 more caution in buffer sizes.
867 PP/09 Implemented reverse_ip expansion operator.
869 PP/10 Bugzilla 937: provide a "debug" ACL control.
871 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
873 PP/12 Bugzilla 973: Implement --version.
875 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
877 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
879 PP/15 Bugzilla 816: support multiple condition rules on Routers.
881 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
882 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
883 ignore trailing whitespace.
885 JJ/02 prevent non-panic DKIM error from being sent to paniclog
887 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
890 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
891 Notification from Dr Andrew Aitchison.
893 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
894 ExtendedDetectionInfo response format.
895 Notification from John Horne.
897 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
900 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
901 XSL and documented dependency on system catalogs, with examples of how
904 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
907 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
908 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
909 configuration file which is writeable by the Exim user or group.
911 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
912 of configuration files to cover files specified with the -C option if
913 they are going to be used with root privileges, not just the default
916 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
917 option (effectively making it always true).
919 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
920 files to be used while preserving root privileges.
922 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
923 that rogue child processes cannot use them.
925 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
926 run-time user, instead of root.
928 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
929 Exim run-time user without dropping privileges.
931 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
932 result string, instead of calling string_vformat() twice with the same
935 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
936 for other users. Others should always drop root privileges if they use
937 -C on the command line, even for a whitelisted configure file.
939 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
941 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
947 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
948 $data_path, and $header_path variables; fixed documentation bugs and
951 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
952 exipick to access non-standard spools, including the "frozen" queue
955 NM/01 Bugzilla 965: Support mysql stored procedures.
956 Patch from Alain Williams
958 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
960 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
961 Patch from Andreas Metzler
963 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
964 Patch from Kirill Miazine
966 NM/05 Bugzilla 671: Added umask to procmail example.
968 JJ/03 installed exipick 20100323.0, fixing doc bug
970 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
971 directory. Notification and patch from Dan Rosenberg.
973 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
975 TK/02 Improve log output when DKIM signing operation fails.
977 MH/01 Treat the transport option dkim_domain as a colon separated
978 list, not as a single string, and sign the message with each element,
979 omitting multiple occurences of the same signer.
981 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
982 Bugzilla 985, 986. Patch by Simon Arlott
984 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
985 Patch by Simon Arlott
987 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
988 MBX locking. Notification from Dan Rosenberg.
994 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
996 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
998 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
1000 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
1002 NM/04 Bugzilla 915: Flag broken perl installation during build.
1008 TK/01 Added patch by Johannes Berg that expands the main option
1009 "spamd_address" if it starts with a dollar sign.
1011 TK/02 Write list of recipients to X-Envelope-Sender header when building
1012 the mbox-format spool file for content scanning (suggested by Jakob
1015 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
1016 (http://www.dcc-servers.net/) support via dccifd. Activated by
1017 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
1019 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
1020 by Mark Daniel Reidel <mr@df.eu>.
1022 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
1023 When building exim an external PCRE library is now needed -
1024 PCRE is a system library on the majority of modern systems.
1025 See entry on PCRE_LIBS in EDITME file.
1027 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
1028 conversation. Added nologin parameter to request.
1029 Patch contributed by Kirill Miazine.
1031 TF/01 Do not log submission mode rewrites if they do not change the address.
1033 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
1035 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
1036 log files in place. Contributed by Roberto Lima.
1038 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1040 TF/03 Bugzilla 615: When checking the local_parts router precondition
1041 after a local_part_suffix or local_part_prefix option, Exim now
1042 does not use the address's named list lookup cache, since this
1043 contains cached lookups for the whole local part.
1045 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1046 Robert Millan. Documentation is in experimental-spec.txt.
1048 TF/04 Bugzilla 668: Fix parallel build (make -j).
1050 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1052 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1053 Patch provided by Jan Srzednicki.
1055 TF/05 Leading white space used to be stripped from $spam_report which
1056 wrecked the formatting. Now it is preserved.
1058 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1059 that they are available at delivery time.
1061 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1063 TF/08 TLS error reporting now respects the incoming_interface and
1064 incoming_port log selectors.
1066 TF/09 Produce a more useful error message if an SMTP transport's hosts
1067 setting expands to an empty string.
1069 NM/06 Bugzilla 744: EXPN did not work under TLS.
1070 Patch provided by Phil Pennock.
1072 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1073 Patch provided by Richard Godbee.
1075 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1076 acl_smtp_notquit, added index entry.
1078 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1079 Patch provided by Eugene Bujak.
1081 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1082 accept(). Patch provided by Maxim Dounin.
1084 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1085 Patch provided by Phil Pennock.
1087 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1089 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1090 Patch provided by Brad "anomie" Jorsch.
1092 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1093 Patch provided by Dean Brooks.
1095 TK/05 Add native DKIM support (does not depend on external libraries).
1097 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1098 Patch provided by Graeme Fowler.
1100 NM/16 Bugzilla 851: Documentation example syntax fix.
1102 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1104 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1107 NM/19 Bugzilla 745: TLS version reporting.
1108 Patch provided by Phil Pennock.
1110 NM/20 Bugzilla 167: bool: condition support.
1111 Patch provided by Phil Pennock.
1113 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1114 clients. Patch provided by Phil Pennock.
1116 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1117 Patch provided by Brad "anomie" Jorsch.
1119 NM/23 Bugzilla 687: Fix misparses in eximstats.
1120 Patch provided by Heiko Schlittermann.
1122 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1123 Patch provided by Heiko Schlittermann.
1125 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1126 plus update to original patch.
1128 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1130 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1131 Patch provided by David Brownlee.
1133 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1135 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1137 NM/30 Bugzilla 888: TLS documentation bugfixes.
1139 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1141 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1142 Unlike the original bugzilla I have changed all shell scripts in src tree.
1144 NM/33 Bugzilla 898: Transport filter timeout fix.
1145 Patch by Todd Rinaldo.
1147 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1148 Patch by Serge Demonchaux.
1150 NM/35 Bugzilla 39: Base64 decode bug fixes.
1151 Patch by Jakob Hirsch.
1153 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1155 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1157 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1159 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1165 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1166 ALT-N's libdkim that I have put here:
1167 http://duncanthrax.net/exim-experimental/
1169 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1170 were called 'true' and it seems that C99 defines that as a reserved
1171 keyword to be used with 'bool' variable types. That means you could
1172 not include C99-style headers which use bools without triggering
1173 build errors in sieve.c.
1175 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1176 as mailq or other aliases. Changed the --help handling significantly
1177 to do whats expected. exim_usage() emits usage/help information.
1179 SC/01 Added the -bylocaldomain option to eximstats.
1181 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1183 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1185 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1191 PH/01 Another patch from the Sieve maintainer.
1193 PH/02 When an IPv6 address is converted to a string for single-key lookup
1194 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1195 dots are used instead of colons so that keys in lsearch files need not
1196 contain colons. This was done some time before quoting was made available
1197 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1198 (notated using the quote facility) so as to distinguish them from IPv4
1199 keys. This meant that lookups for IP addresses in host lists did not work
1200 for iplsearch lookups.
1202 This has been fixed by arranging for IPv6 addresses to be expressed with
1203 colons if the lookup type is iplsearch. This is not incompatible, because
1204 previously such lookups could never work.
1206 The situation is now rather anomolous, since one *can* have colons in
1207 ordinary lsearch keys. However, making the change in all cases is
1208 incompatible and would probably break a number of configurations.
1210 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1213 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1214 conversion specification without a maximum field width, thereby enabling
1215 a rogue spamd server to cause a buffer overflow. While nobody in their
1216 right mind would setup Exim to query an untrusted spamd server, an
1217 attacker that gains access to a server running spamd could potentially
1218 exploit this vulnerability to run arbitrary code as the Exim user.
1220 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1221 $primary_hostname instead of what libspf2 thinks the hosts name is.
1223 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1224 a directory entry by the name of the lookup key. Previously, if a
1225 symlink pointed to a non-existing file or a file in a directory that
1226 Exim lacked permissions to read, a lookup for a key matching that
1227 symlink would fail. Now it is enough that a matching directory entry
1228 exists, symlink or not. (Bugzilla 503.)
1230 PH/03 The body_linecount and body_zerocount variables are now exported in the
1233 PH/04 Added the $dnslist_matched variable.
1235 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1236 This means they are set thereafter only if the connection becomes
1239 PH/06 Added the client_condition to authenticators so that some can be skipped
1240 by clients under certain conditions.
1242 PH/07 The error message for a badly-placed control=no_multiline_responses left
1243 "_responses" off the end of the name.
1245 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1247 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1248 (without spaces) instead of just copying the configuration text.
1250 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1252 PH/11 Added $max_received_linelength.
1254 PH/12 Added +ignore_defer and +include_defer to host lists.
1256 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1257 way in which PCRE > 7.0 is built.
1259 PH/14 Implemented queue_only_load_latch.
1261 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1262 MAIL command. The effect was to mangle the value on 64-bit systems.
1264 PH/16 Another patch from the Sieve maintainer.
1266 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1268 PH/18 If a system quota error occurred while trying to create the file for
1269 a maildir delivery, the message "Mailbox is full" was not appended to the
1270 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1271 only to a quota excession during the actual writing of the file.
1273 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1274 characters?) which causes problems in log lines. The DN values are now
1275 passed through string_printing() before being added to log lines.
1277 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1278 and InterBase are left for another time.)
1280 PH/21 Added message_body_newlines option.
1282 PH/22 Guard against possible overflow in moan_check_errorcopy().
1284 PH/23 POSIX allows open() to be a macro; guard against that.
1286 PH/24 If the recipient of an error message contained an @ in the local part
1287 (suitably quoted, of course), incorrect values were put in $domain and
1288 $local_part during the evaluation of errors_copy.
1294 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1295 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1298 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1299 issue a MAIL command.
1301 PH/02 In an ACL statement such as
1303 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1305 if a client was not listed at all, or was listed with a value other than
1306 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1307 the condition was not true (as it should be), so access was not denied.
1308 The bug was that the ! inversion was incorrectly passed on to the second
1309 item. This has been fixed.
1311 PH/03 Added additional dnslists conditions == and =& which are different from
1312 = and & when the dns lookup returns more than one IP address.
1314 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1315 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1317 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1318 FSYNC, which compiles an option called disable_fsync that allows for
1319 bypassing fsync(). The documentation is heavily laced with warnings.
1321 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1323 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1324 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1325 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1326 including adding "make clean"; (3) Added -fPIC when compiling the test
1327 dynamically loaded module, to get rid of a warning.
1329 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1330 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1331 = 0s. The bug is otherwise harmless.
1333 PH/07 There was a bug in the dovecot authenticator such that the value of
1334 $auth1 could be overwritten, and so not correctly preserved, after a
1335 successful authentication. This usually meant that the value preserved by
1336 the server_setid option was incorrect.
1338 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1340 PH/09 Installed PCRE release 7.0.
1342 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1343 run for batched SMTP input. It is now run at the start of every message
1344 in the batch. While fixing this I discovered that the process information
1345 (output by running exiwhat) was not always getting set for -bs and -bS
1346 input. This is fixed, and it now also says "batched" for BSMTP.
1348 PH/11 Added control=no_pipelining.
1350 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1351 patch, slightly modified), and move the expansion of helo_data till after
1352 the connection is made in the smtp transport (so it can use these
1355 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1357 PH/14 Added log_selector = +pid.
1359 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1361 PH/16 Add ${if forany and ${if forall.
1363 PH/17 Added dsn_from option to vary the From: line in DSNs.
1365 PH/18 Flush SMTP output before performing a callout, unless control =
1366 no_callout_flush is set.
1368 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1369 was true (the default) a successful delivery failed to delete the retry
1370 item, thus causing premature timeout of the address. The bug is now
1373 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1375 PH/21 Long custom messages for fakedefer and fakereject are now split up
1376 into multiline reponses in the same way that messages for "deny" and
1377 other ACL rejections are.
1379 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1380 with slight modification.
1382 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1383 draft, changing the syntax and factoring some duplicate code".
1385 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1386 for deliveries of the second and subsequent messages over the same SMTP
1389 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1390 ${reduce, with only minor "tidies".
1392 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1394 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1395 expansion side effects.
1397 PH/27 When a message times out after an over-quota error from an Exim-imposed
1398 quota, the bounce message says "mailbox is full". This message was not
1399 being given when it was a system quota that was exceeded. It now should
1402 MH/03 Made $recipients available in local_scan(). local_scan() already has
1403 better access to the recipient list through recipients_list[], but
1404 $recipients can be useful in postmaster-provided expansion strings.
1406 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1407 in the case of a MAIL command with additional options following the
1408 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1409 were accidentally chopped off.
1411 PH/29 SMTP synchronization checks are implemented when a command is read -
1412 there is a check that no more input is waiting when there shouldn't be
1413 any. However, for some commands, a delay in an ACL can mean that it is
1414 some time before the response is written. In this time, more input might
1415 arrive, invalidly. So now there are extra checks after an ACL has run for
1416 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1417 pipelining has not been advertised.
1419 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1421 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1422 returned in the SMTP response when smtp_return_error_details was false.
1423 This has been fixed.
1425 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1426 instead of the C I/O that was originally supplied, because problems were
1427 reported on Solaris.
1429 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1430 Exim which did not show up earlier: it was assuming that a call to
1431 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1432 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1433 was a macro that became an assignment, so it seemed to work. This has
1434 changed to a proper function call with a void return, hence the compile
1435 error. Exim's code has been fixed.
1437 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1440 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1442 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1444 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1445 because it turns out to be too big or there is a timeout) there is no
1446 "Completed" line in the log. When some messages of this type were
1447 selected by exigrep, they were listed as "not completed". Others were
1448 picked up by some special patterns. I have improved the selection
1449 criteria to be more general.
1451 PH/38 The host_find_failed option in the manualroute router can now be set
1452 to "ignore", to completely ignore a host whose IP address cannot be
1453 found. If all hosts are ignored, the behaviour is controlled by the new
1454 host_all_ignored option.
1456 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1457 homing or because of multiple MX records with /mx) generated more than
1458 one IP address, and the following item turned out to be the local host,
1459 all the secondary addresses of the first item were incorrectly removed
1460 from the list, along with the local host and any following hosts (which
1461 is what is supposed to happen).
1463 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1464 whoever called Exim into the -H file. In the case of the daemon it was
1465 behaving confusingly. When first started, it used values for whoever
1466 started the daemon, but after a SIGHUP it used the Exim user (because it
1467 calls itself on a restart). I have changed the code so that it now always
1470 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1471 message are rejected with the same error (e.g. no authentication or bad
1472 sender address), and a DATA command is nevertheless sent (as can happen
1473 with PIPELINING or a stupid MUA), the error message that was given to the
1474 RCPT commands is included in the rejection of the DATA command. This is
1475 intended to be helpful for MUAs that show only the final error to their
1478 PH/42 Another patch from the Sieve maintainer.
1480 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1481 Eximstats - Fixed some broken HTML links and added missing column headers
1483 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1484 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1486 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1488 PH/43 Yet another patch from the Sieve maintainer.
1490 PH/44 I found a way to check for a TCP/IP connection going away before sending
1491 the response to the final '.' that terminates a message, but only in the
1492 case where the client has not sent further data following the '.'
1493 (unfortunately, this is allowed). However, in many cases there won't be
1494 any further data because there won't be any more messages to send. A call
1495 to select() can be used: if it shows that the input is "ready", there is
1496 either input waiting, or the socket has been closed. An attempt to read
1497 the next input character can distinguish the two cases. Previously, Exim
1498 would have sent an OK response which the client would never have see.
1499 This could lead to message repetition. This fix should cure that, at
1500 least in a lot of common cases.
1502 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1503 advertised in response to EHLO.
1509 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1510 fixed by 4.65/MH/01 (is this a record?) are fixed:
1512 (i) An empty string was always treated as zero by the numeric comparison
1513 operators. This behaviour has been restored.
1515 (ii) It is documented that the numeric comparison operators always treat
1516 their arguments as decimal numbers. This was broken in that numbers
1517 starting with 0 were being interpreted as octal.
1519 While fixing these problems I realized that there was another issue that
1520 hadn't been noticed. Values of message_size_limit (both the global option
1521 and the transport option) were treated as octal if they started with 0.
1522 The documentation was vague. These values are now always treated as
1523 decimal, and I will make that clear in the documentation.
1529 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1530 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1533 MH/01 Don't check that the operands of numeric comparison operators are
1534 integers when their expansion is in "skipping" mode (fixes bug
1535 introduced by 4.64-PH/07).
1537 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1538 child addresses, Exim now panics and dies. Previously, because the count
1539 is held in a short int, deliveries were likely to be lost. As such a
1540 large number of recipients for a single message is ridiculous
1541 (performance will be very, very poor), I have chosen to impose a limit
1542 rather than extend the field.
1548 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1549 leftover -K file (the existence of which was triggered by #402).
1550 While we were at it, introduced process PID as part of the -K
1551 filename. This should rule out race conditions when creating
1554 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1555 processing considerably. Previous code took too long for large mails,
1556 triggering a timeout which in turn triggers #401.
1558 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1559 in the DK code in transports.c. sendfile() is not really portable,
1560 hence the _LINUX specificness.
1562 TF/01 In the add_headers option to the mail command in an Exim filter,
1563 there was a bug that Exim would claim a syntax error in any
1564 header after the first one which had an odd number of characters
1567 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1568 callout verification, Exim cached a "reject" for the entire domain. This
1569 is correct for most verifications, but it is not correct for a recipient
1570 verification with use_sender or use_postmaster set, because in that case
1571 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1572 case of MAIL FROM:<> rejection from other early rejections (e.g.
1573 rejection of HELO). When verifying a recipient using a non-null MAIL
1574 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1575 Whatever the result of the callout, the value of the domain cache is
1576 left unchanged (for any other kind of callout, getting as far as trying
1577 RCPT means that the domain itself is ok).
1579 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1582 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1583 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1584 session, the daemon ignores ECONNECT errors and logs others; it now
1585 ignores EPIPE as well.
1587 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1588 (quoted-printable decoding).
1590 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1591 later the small subsequent patch to fix an introduced bug.
1593 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1595 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1597 PH/08 An error is now given if message_size_limit is specified negative.
1599 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1600 to be given (somewhat) arbitrary names.
1602 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1605 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1606 miscellaneous code fixes
1608 PH/10 Added the log_reject_target ACL modifier to specify where to log
1611 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1612 hostname. This is wrong, because it relates to the incoming message (and
1613 probably the interface on which it is arriving) and not to the outgoing
1614 callout (which could be using a different interface). This has been
1615 changed to use the value of the helo_data option from the smtp transport
1616 instead - this is what is used when a message is actually being sent. If
1617 there is no remote transport (possible with a router that sets up host
1618 addresses), $smtp_active_hostname is used.
1620 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1621 tweaks were necessary in order to get it to work (see also 21 below):
1622 (a) The code assumed that strncpy() returns a negative number on buffer
1623 overflow, which isn't the case. Replaced with Exim's string_format()
1625 (b) There were several signed/unsigned issues. I just did the minimum
1626 hacking in of casts. There is scope for a larger refactoring.
1627 (c) The code used strcasecmp() which is not a standard C function.
1628 Replaced with Exim's strcmpic() function.
1629 (d) The code set only $1; it now sets $auth1 as well.
1630 (e) A simple test gave the error "authentication client didn't specify
1631 service in request". It would seem that Dovecot has changed its
1632 interface. Fortunately there's a specification; I followed it and
1633 changed what the client sends and it appears to be working now.
1635 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1638 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1639 address A is aliased to B and C, where B exists and C does not. Without
1640 -v the output is "A verified" because verification stops after a
1641 successful redirection if more than one address is generated. However,
1642 with -v the child addresses are also verified. Exim was outputting "A
1643 failed to verify" and then showing the successful verification for C,
1644 with its parentage. It now outputs "B failed to verify", showing B's
1645 parentage before showing the successful verification of C.
1647 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1648 look up a TXT record in a specific list after matching in a combined
1651 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1652 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1653 they consult the DNS. I had assumed they would set it the way they
1654 wanted; and indeed my experiments on Linux seem to show that in some
1655 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1656 To be on the safe side, however, I have now made the interface to
1657 host_find_byname() similar to host_find_bydns(), with an argument
1658 containing the DNS resolver options. The host_find_byname() function now
1659 sets these options at its start, just as host_find_bydns() does. The smtp
1660 transport options dns_qualify_single and dns_search_parents are passed to
1661 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1662 of host_find_byname() use the default settings of RES_DEFNAMES
1663 (qualify_single) but not RES_DNSRCH (search_parents).
1665 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1666 spool_read_header() do less string testing, by means of a preliminary
1667 switch on the second character of optional "-foo" lines. (This is
1668 overdue, caused by the large number of possibilities that now exist.
1669 Originally there were few.) While I was there, I also converted the
1670 str(n)cmp tests so they don't re-test the leading "-" and the first
1671 character, in the hope this might squeeze out yet more improvement.
1673 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1674 flag allowing group syntax was set by the header_syntax check but not
1675 turned off, possible causing trouble later; (2) The flag was not being
1676 set at all for the header_verify test, causing "group"-style headers to
1677 be rejected. I have now set it in this case, and also caused header_
1678 verify to ignore an empty address taken from a group. While doing this, I
1679 came across some other cases where the code for allowing group syntax
1680 while scanning a header line wasn't quite right (mostly, not resetting
1681 the flag correctly in the right place). These bugs could have caused
1682 trouble for malformed header lines. I hope it is now all correct.
1684 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1685 with the "reply" argument non-NULL. The code, however (which originally
1686 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1687 but it didn't always do it. This confused somebody who was copying the
1688 code for some other use. I have removed all the tests.
1690 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1691 feature that was used to support insecure browsers during the U.S. crypto
1692 embargo. It requires special client support, and Exim is probably the
1693 only MTA that supported it -- and would never use it because real RSA is
1694 always available. This code has been removed, because it had the bad
1695 effect of slowing Exim down by computing (never used) parameters for the
1696 RSA_EXPORT functionality.
1698 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1699 authenticator to fail if there's a tab character in the incoming data
1700 (there should never be unless someone is messing about, as it's supposed
1701 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1702 if the connection is using TLS or if the remote IP is the same as the
1703 local IP, and the "valid-client-cert option" if a client certificate has
1706 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1707 authenticators. This can be used for authorization after authentication
1708 succeeds. (In the case of plaintext, it servers for both authentication
1711 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1712 if any retry times were supplied.
1714 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1715 connection, where there is no client IP address to check. In this
1716 situation, the verify now always succeeds.
1718 PH/25 Applied John Jetmore's -Mset patch.
1720 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1722 PH/27 In a string expansion for a processed (not raw) header when multiple
1723 headers of the same name were present, leading whitespace was being
1724 removed from all of them, but trailing whitespace was being removed only
1725 from the last one. Now trailing whitespace is removed from each header
1726 before concatenation. Completely empty headers in a concatenation (as
1727 before) are ignored.
1729 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1730 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1732 PH/29 [Removed. This was a change that I later backed out, and forgot to
1733 correct the ChangeLog entry (that I had efficiently created) before
1734 committing the later change.]
1736 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1737 address errors (4xx response to RCPT) over the same connection as other
1738 messages routed to the same hosts. Such deliveries are always "forced",
1739 so retry times are not inspected. This resulted in far too many retries
1740 for the affected addresses. The effect occurred only when there were more
1741 hosts than the hosts_max_try setting in the smtp transport when it had
1742 the 4xx errors. Those hosts that it had tried were not added to the list
1743 of hosts for which the message was waiting, so if all were tried, there
1744 was no problem. Two fixes have been applied:
1746 (i) If there are any address or message errors in an SMTP delivery, none
1747 of the hosts (tried or untried) are now added to the list of hosts
1748 for which the message is waiting, so the message should not be a
1749 candidate for sending over the same connection that was used for a
1750 successful delivery of some other message. This seems entirely
1751 reasonable: after all the message is NOT "waiting for some host".
1752 This is so "obvious" that I'm not sure why it wasn't done
1753 previously. Hope I haven't missed anything, but it can't do any
1754 harm, as the worst effect is to miss an optimization.
1756 (ii) If, despite (i), such a delivery is accidentally attempted, the
1757 routing retry time is respected, so at least it doesn't keep
1758 hammering the server.
1760 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1761 in ${readsocket because some servers need this prod.
1763 PH/32 Added some extra debug output when updating a wait-xxx database.
1765 PH/33 The hint "could be header name not terminated by colon", which has been
1766 given for certain expansion errors for a long time, was not being given
1767 for the ${if def:h_colon_omitted{... case.
1769 PH/34 The spec says: "With one important exception, whenever a domain list is
1770 being scanned, $domain contains the subject domain." There was at least
1771 one case where this was not true.
1773 PH/35 The error "getsockname() failed: connection reset by peer" was being
1774 written to the panic log as well as the main log, but it isn't really
1775 panic-worthy as it just means the connection died rather early on. I have
1776 removed the panic log writing for the ECONNRESET error when getsockname()
1779 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1780 runs only) independently of the message's sender address. This meant
1781 that, if the 4xx error was in fact related to the sender, a different
1782 message to the same recipient with a different sender could confuse
1783 things. In particualar, this can happen when sending to a greylisting
1784 server, but other circumstances could also provoke similar problems.
1785 I have changed the default so that the retry time for these errors is now
1786 based a combination of the sender and recipient addresses. This change
1787 can be overridden by setting address_retry_include_sender=false in the
1790 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1791 remote server are returned as part of bounce messages. This was not
1792 happening for LMTP over a pipe (the lmtp transport), but now it is the
1793 same for both kinds of LMTP.
1795 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1796 in header lines that were in fact CNAMEs. This is no longer the case.
1798 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1799 and queue runs started by the daemon processed all messages. This has
1800 been fixed so that -R and -S can now usefully be given with -q<time>.
1802 PH/40 Import PCRE release 6.7 (fixes some bugs).
1804 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1806 PH/42 Give an error if -q is specified more than once.
1808 PH/43 Renamed the variables $interface_address and $interface_port as
1809 $received_ip_address and $received_port, to make it clear that these
1810 values apply to message reception, and not to the outgoing interface when
1811 a message is delivered. (The old names remain recognized, of course.)
1813 PH/44 There was no timeout on the connect() call when using a Unix domain
1814 socket in the ${readsocket expansion. There now is.
1816 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1817 be meaningful with "accept".
1819 SC/01 Eximstats V1.43
1820 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1822 SC/02 Eximstats V1.44
1823 Use a glob alias rather than an array ref in the generated
1824 parser. This improves both readability and performance.
1826 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1827 Collect SpamAssassin and rejection statistics.
1828 Don't display local sender or destination tables unless
1829 there is data to show.
1830 Added average volumes into the top table text output.
1832 SC/04 Eximstats V1.46
1833 Collect data on the number of addresses (recipients)
1834 as well as the number of messages.
1836 SC/05 Eximstats V1.47
1837 Added 'Message too big' to the list of mail rejection
1838 reasons (thanks to Marco Gaiarin).
1840 SC/06 Eximstats V1.48
1841 Mainlog lines which have GMT offsets and are too short to
1842 have a flag are now skipped.
1844 SC/07 Eximstats V1.49 (Alain Williams)
1845 Added the -emptyok flag.
1847 SC/08 Eximstats V1.50
1848 Fixes for obtaining the IP address from reject messages.
1850 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1851 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1852 whitesspace changes from 4.64-PH/27
1854 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1857 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1858 are found, allow negative numbers in numeric criteria)
1860 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1862 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1863 to match changes made in 4.64-PH/43
1865 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1867 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1868 "domain" is actually an IP address, and force a failure. This locks out
1869 those revolvers/nameservers that support "A-for-A" lookups, in
1870 contravention of the specifications.
1872 PH/48 When a host name was looked up from an IP address, and the subsequent
1873 forward lookup of the name timed out, the host name was left in
1874 $sender_host_name, contrary to the specification.
1876 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1877 restricted to single-key lookups, Exim was not diagnosing an error if
1878 * or *@ was used with a query-style lookup.
1880 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1882 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1883 long ago, but noone interested enough thought of it. Let's just say that
1884 the "1.1" means that there are some new functions that weren't there at
1885 some point in the past.
1887 PH/51 Error processing for expansion failure of helo_data from an smtp
1888 transport during callout processing was broken.
1890 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1891 tested/used via the -bh/-bhc/-bs options.
1893 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1894 bug, fixed in subsequent PCRE releases).
1896 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1897 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1899 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1905 SC/01 Use a glob alias rather than an array ref in eximstats generated
1906 parser. This improves both readability and performance.
1908 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1909 Don't display local sender or destination tables in eximstats unless
1910 there is data to show.
1911 Added average volumes into the eximstats top table text output.
1913 SC/03 Collect data on the number of addresses (recipients) as well
1914 as the number of messages in eximstats.
1916 TF/01 Correct an error in the documentation for the redirect router. Exim
1917 does (usually) call initgroups() when daemonizing.
1919 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1920 with consistent privilege compared to when running as a daemon.
1922 TF/03 Note in the spec that $authenticated_id is not set for local
1923 submissions from trusted users.
1925 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1926 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1928 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1929 by adding some example configuration directives to the default
1930 configuration file. A little bit of work is required to uncomment the
1931 directives and define how usernames and passwords are checked, but
1932 there is now a framework to start from.
1934 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1935 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1936 without this. I don't know how relevant this is to other LDAP libraries.
1938 PH/02 Add the verb name to the "unknown ACL verb" error.
1940 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1942 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1944 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1945 directory not expanded when it should be if an expanded home directory
1946 was set for the address (which is overridden by the transport).
1948 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1951 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1952 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1953 because it is too late at that time, and has no effect.
1955 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1956 security issue with \' (bugzilla #107). I could not use the
1957 PQescapeStringConn() function, because it needs a PGconn value as one of
1960 PH/08 When testing addresses using -bt, indicate those final addresses that
1961 are duplicates that would not cause an additional delivery. At least one
1962 person was confused, thinking that -bt output corresponded to deliveries.
1963 (Suppressing duplicates isn't a good idea as you lose the information
1964 about possibly different redirections that led to the duplicates.)
1966 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1967 systems where poll() doesn't work, in particular OS X.
1969 PH/10 Added more information to debugging output for retry time not reached.
1971 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1972 operations in malware.c.
1974 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1977 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1978 syslog_duplication set false, log lines that would normally be written
1979 both the the main log and to the reject log were not written to syslog at
1982 PH/14 In the default configuration, change the use of "message" in ACL warn
1983 statements to "add_header".
1985 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1986 not followed by a command (e.g. "seen endif").
1988 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1989 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1992 PH/17 Added extra conditions to the default value of delay_warning_condition
1996 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1997 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1998 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
2001 The Auto-Submitted: and various List- headers are standardised, whereas I
2002 don't think Precedence: ever was.
2004 PH/18 Refactored debugging code in route_finduser() to show more information,
2005 in particular, the error code if getpwnam() issues one.
2007 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
2008 This is apparently needed in addition to the PH/07 change above to avoid
2009 any possible encoding problems.
2011 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
2012 but not after initializing Perl.
2014 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
2015 output them only if debugging. By default they are written stderr,
2016 apparently, which is not desirable.
2018 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
2021 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
2024 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
2026 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
2027 authenticated or an ident call has been made. Suppress the default
2028 values for $authenticated_id and $authenticated_sender (but permit -oMai
2029 and -oMas) when testing with -bh.
2031 PH/24 Re-jigged the order of the tests in the default configuration so that the
2032 tests for valid domains and recipients precede the DNS black list and CSA
2033 tests, on the grounds that those ones are more expensive.
2035 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
2036 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
2037 This bug exists in every version of Exim that I still have, right back to
2040 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2041 However, an attempt to turn on case-sensitivity in a regex key by
2042 including (?-i) didn't work because the subject string was already
2043 lowercased, and the effects were non-intuitive. It turns out that a
2044 one-line patch can be used to allow (?-i) to work as expected.
2050 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2051 other effects) broke the use of negated acl sub-conditions.
2053 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2056 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2057 "Deny" causes Exim to reject the incoming connection with a 554 error.
2058 Unfortunately, if there is a major crisis, such as a disk failure,
2059 tcp-wrappers gives "deny", whereas what one would like would be some
2060 kind of temporary error. A kludge has been added to help with this.
2061 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2062 554 error is used if errno is still zero or contains ENOENT (which occurs
2063 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2066 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2068 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2069 errors. Otherwise a message that provokes a temporary error (when other
2070 messages do not) can cause a whole host to time out.
2072 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2073 addresses were routed directly to files or pipes from a redirect router.
2074 File deliveries just didn't batch; pipe deliveries might have suffered
2077 PH/06 A failure to get a lock for a hints database would erroneously always say
2078 "Failed to get write lock", even when it was really a read lock.
2080 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2081 of 0600. This has been changed to use the value of the lockfile_mode
2082 option (which defaults to 0600).
2084 PH/08 Applied small patch from the Sieve maintainer.
2086 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2087 folder from quota calculations, a direct delivery into this folder messed
2088 up the contents of the maildirsize file. This was because the regex was
2089 used only to exclude .Trash (or whatever) when the size of the mailbox
2090 was calculated. There was no check that a delivery was happening into an
2091 excluded directory. This bug has been fixed by ignoring all quota
2092 processing for deliveries into excluded directories.
2094 PH/10 Added the maildirfolder_create_regex option to appendfile.
2100 PH/01 The code for finding all the local interface addresses on a FreeBSD
2101 system running IPv6 was broken. This may well have applied to all BSD
2102 systems, as well as to others that have similar system calls. The broken
2103 code found IPv4 interfaces correctly, but gave incorrect values for the
2104 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2105 that it would not match correctly against @[] and not recognize the IPv6
2108 PH/02 The ipliteral router was not recognizing addresses of the form user@
2109 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2111 PH/03 Added disable_ipv6.
2113 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2114 decoded form, because it is most often used to construct To: headers
2115 lines in autoreplies, and the decoded form may well be syntactically
2116 invalid. However, $reply_address has leading white space removed, and all
2117 newlines turned into spaces so that the autoreply transport does not
2120 PH/05 If group was specified without a user on a router, and no group or user
2121 was specified on a transport, the group from the router was ignored.
2123 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2124 for visible compile-time settings that can be used to change these
2125 numbers, for those that want even more. Backwards compatibility with old
2126 spool files has been maintained. However, going back to a previous Exim
2127 release will lost any variables that are in spool files.
2129 PH/07 Two small changes when running in the test harness: increase delay when
2130 passing a TCP/IP connection to a new process, in case the original
2131 process has to generate a bounce, and remove special handling of
2132 127.0.0.2 (sic), which is no longer necessary.
2134 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2135 be the same on different OS.
2137 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2140 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2141 whether --show-vars was specified or not
2143 JJ/02 exipick: Added support for new ACL variable spool format introduced
2146 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2147 syntactically invalid From: or Reply-to: line, and a filter used this to
2148 generate an autoreply, and therefore failed to obtain an address for the
2149 autoreply, Exim could try to deliver to a non-existent relative file
2150 name, causing unrelated and misleading errors. What now happens is that
2151 it logs this as a hard delivery error, but does not attempt to create a
2154 PH/11 The exinext utility has a -C option for testing purposes, but although
2155 the given file was scanned by exinext itself; it wasn't being passed on
2156 when Exim was called.
2158 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2159 an end-of-file indication when reading a command response.
2161 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2162 compiled. In many other places in Exim, IPv6 addresses are always
2163 recognized, so I have changed this. It also means that IPv4 domain
2164 literals of the form [IPV4:n.n.n.n] are now always recognized.
2166 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2167 used if the router is not running as root, for example, when verifying at
2168 ACL time, or when using -bh. The debugging output from this situation was
2169 non-existent - all you got was a failure to exec. I have made two
2172 (a) Failures to set uid/gid, the current directory, or a process leader
2173 in a subprocess such as that created by queryprogram now generate
2174 suitable debugging ouput when -d is set.
2176 (b) The queryprogram router detects when it is not running as root,
2177 outputs suitable debugging information if -d is set, and then runs
2178 the subprocess without attempting to change uid/gid.
2180 PH/15 Minor change to Makefile for building test_host (undocumented testing
2183 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2184 additional section of a DNS packet that returns MX or SRV records.
2185 Instead, it always explicitly searches for A/AAAA records. This avoids
2186 major problems that occur when a DNS server includes only records of one
2187 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2188 fixed another bug: if SRV records were looked up and the corresponding
2189 address records were *not* found in the additional section, the port
2190 values from the SRV records were lost.
2192 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2193 using the correct key (the original address) when searching the retry
2194 rules in order to find which one to use for generating the retry hint.
2196 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2197 adding the default one. Similarly, if it contains a Reply-To: header, the
2198 errors_reply_to option, if set, is not used.
2200 PH/19 When calculating a retry time, Exim used to measure the "time since
2201 failure" by looking at the "first failed" field in the retry record. Now
2202 it does not use this if it is later than than the arrival time of the
2203 message. Instead it uses the arrival time. This makes for better
2204 behaviour in cases where some deliveries succeed, thus re-setting the
2205 "first failed" field. An example is a quota failure for a huge message
2206 when small messages continue to be delivered. Without this change, the
2207 "time since failure" will always be short, possible causing more frequent
2208 delivery attempts for the huge message than are intended.
2209 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2211 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2212 $1, $2, $3) because the numerical variables can be reset during some
2213 expansion items (e.g. "match"), thereby losing the authentication data.
2215 PH/21 Make -bV show the size of off_t variables so that the test suite can
2216 decide whether to run tests for quotas > 2G.
2218 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2219 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2220 filecount value is greater than 2G or if a quota value is greater than 2G
2221 on a system where the size of off_t is not greater than 4, a panic error
2224 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2225 never match. The debug and -bh output now contains an explicit error
2226 message indicating a malformed IPv4 address or mask.
2228 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2229 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2230 PH/23 above applies.
2232 PH/25 Do not write to syslog when running in the test harness. The only
2233 occasion when this arises is a failure to open the main or panic logs
2234 (for which there is an explicit test).
2236 PH/26 Added the /no_tell option to "control=freeze".
2238 PH/27 If a host name lookup failed very early in a connection, for example, if
2239 the IP address matched host_lookup and the reverse lookup yielded a name
2240 that did not have a forward lookup, an error message of the form "no IP
2241 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2242 could be logged. Now it outputs the IP address instead of "NULL".
2244 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2245 allows the sender and the authenticated sender to be set when
2246 submitting a message from within Exim. Since child_open_exim() is
2247 documented for local_scan(), the new function should be too.
2249 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2250 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2251 results in an empty string is now treated as unset.
2253 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2255 PH/31 Added sender_verify_fail logging option.
2257 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2258 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2259 tidied the source and removed it altogether.
2261 PH/33 When a queue run was abandoned because the load average was too high, a
2262 log line was always written; now it is written only if the queue_run log
2263 selector is set. In addition, the log line for abandonment now contains
2264 information about the queue run such as the pid. This is always present
2265 in "start" and "stop" lines but was omitted from the "abandon" line.
2267 PH/34 Omit spaces between a header name and the colon in the error message that
2268 is given when verify = headers_syntax fails (if there are lots of them,
2269 the message gets confusing).
2271 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2272 names, as there are now some PTR records that contain slashes. This check
2273 is only to protect against broken name servers that fall over on strange
2274 characters, so the fact that it applies to all lookups doesn't matter.
2276 PH/36 Now that the new test suite is complete, we can remove some of the
2277 special code in Exim that was needed for the old test suite. For example,
2278 sorting DNS records because real resolvers return them in an arbitrary
2279 order. The new test suite's fake resolver always returns records in the
2282 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2283 bounces) except when queue_only is set, to avoid logging races between
2284 the different processes.
2286 PH/38 Panic-die if .include specifies a non-absolute path.
2288 PH/39 A tweak to the "H" retry rule from its user.
2290 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2291 a label. They prevented compilation on older perls.
2293 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2294 a warning to be raised on newish perls.
2296 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2297 on queue. Changes to match documented behaviour of showing count of
2298 messages matching specified criteria.
2300 PH/40 Changed the default ident timeout from 30s to 5s.
2302 PH/41 Added support for the use of login_cap features, on those BSD systems
2303 that have them, for controlling the resources used by pipe deliveries.
2305 PH/42 The content-scanning code uses fopen() to create files in which to put
2306 message data. Previously it was not paying any attention to the mode of
2307 the files. Exim runs with umask(0) because the rest of the code creates
2308 files with open(), and sets the required mode explicitly. Thus, these
2309 files were ending up world-writeable. This was not a big issue, because,
2310 being within the spool directory, they were not world-accessible. I have
2311 created a function called modefopen, which takes an additional mode
2312 argument. It sets umask(777), creates the file, chmods it to the required
2313 mode, then resets the umask. All the relevant calls to fopen() in the
2314 content scanning code have been changed to use this function.
2316 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2317 to 24 hours. This avoids potential overflow problems when processing G
2318 and H retry rules. I suspect nobody ever tinkers with this value.
2320 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2322 PH/45 When the plaintext authenticator is running as a client, the server's
2323 challenges are checked to ensure they are valid base64 strings. By
2324 default, the authentication attempt is cancelled if an invalid string is
2325 received. Setting client_ignore_invalid_base64 true ignores these errors.
2326 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2327 they are received. Thus, the responses can be made to depend on the
2328 challenges. If an invalid string is ignored, an empty string is placed in
2331 PH/46 Messages that are created by the autoreply transport now contains a
2332 References: header, in accordance with RFCs 2822 and 3834.
2334 PH/47 Added authenticated_sender_force to the smtp transport.
2336 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2338 PH/49 Installed latest patch from the Sieve maintainer.
2340 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2341 was also set, the appendfile transport was unnecessarily scanning a
2342 directory of message files (e.g. for maildir delivery) to find the count
2343 of files (along with the size), even though it did not need this
2344 information. It now does the scan only if it needs to find either the
2345 size of the count of files.
2347 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2349 PH/52 Two bugs concerned with error handling when the smtp transport is
2352 (i) Exim was not creating retry information for temporary errors given
2353 for individual recipients after the DATA command when the smtp transport
2354 was used in LMTP mode. This meant that they could be retried too
2355 frequently, and not timed out correctly.
2357 (ii) Exim was setting the flag that allows error details to be returned
2358 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2359 recipients that were returned after the DATA command.
2361 PH/53 This is related to PH/52, but is more general: for any failing address,
2362 when detailed error information was permitted to be returned to the
2363 sender, but the error was temporary, then after the final timeout, only
2364 "retry timeout exceeded" was returned. Now it returns the full error as
2365 well as "retry timeout exceeded".
2367 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2368 do this, and (what is worse) MTAs that accept it.
2370 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2371 will now be deprecated.
2373 PH/56 New os.c-cygwin from the Cygwin maintainer.
2375 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2376 formats (previously only available in exim formats via -bpr, -bpru,
2377 and -bpra. Now also available in native and exiqgrep formats)
2379 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2380 with very large, slow to parse queues
2382 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2384 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2386 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2387 responses to authentication challenges, though it was showing the
2388 challenges; (ii) I've removed the CR characters from the debug output for
2391 PH/58 Allow for the insertion of a newline as well as a space when a string
2392 is turned into more than one encoded-word during RFC 2047 encoding. The
2393 Sieve code now uses this.
2395 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2396 data_4xx, lost_connection, tls_required.
2398 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2399 message was being sent as an SMTP response.
2401 PH/61 Add -l and -k options to exicyclog.
2403 PH/62 When verifying, if an address was redirected to one new address, so that
2404 verification continued, and the new address failed or deferred after
2405 having set something in $address_data, the value of $address_data was not
2406 passed back to the ACL. This was different to the case when no
2407 redirection occurred. The value is now passed back in both cases.
2409 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2410 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2411 use login_cap.h, so on its own it isn't the distinguishing feature. The
2412 new name refers directly to the setclassresources() function.
2414 PH/65 Added configuration files for NetBSD3.
2416 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2418 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2419 is preferred over IPv4.
2421 PH/68 The bounce_return_message and bounce_return_body options were not being
2422 honoured for bounces generated during the reception of non-SMTP messages.
2423 In particular, this applied to messages rejected by the ACL. This bug has
2424 been fixed. However, if bounce_return_message is true and bounce_return_
2425 body is false, the headers that are returned for a non-SMTP message
2426 include only those that have been read before the error was detected.
2427 (In the case of an ACL rejection, they have all been read.)
2429 PH/69 The HTML version of the specification is now built in a directory called
2430 spec_html instead of spec.html, because the latter looks like a path with
2431 a MIME-type, and this confuses some software.
2433 PH/70 Catch two compiler warnings in sieve.c.
2435 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2436 function verify_get_ident() calls ip_connect() to connect a socket, but
2437 if the "connect()" function timed out, ip_connect() used to close the
2438 socket. However, verify_get_ident() also closes the socket later, and in
2439 between Exim writes to the log, which may get opened at this point. When
2440 the socket was closed in ip_connect(), the log could get the same file
2441 descriptor number as the socket. This naturally causes chaos. The fix is
2442 not to close the socket in ip_connect(); the socket should be closed by
2443 the function that creates it. There was only one place in the code where
2444 this was missing, in the iplookup router, which I don't think anybody now
2445 uses, but I've fixed it anyway.
2447 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2448 well as to direct DNS lookups. Otherwise the handling of names in host
2449 lists is inconsistent and therefore confusing.
2455 PH/01 Two changes to the default runtime configuration:
2457 (1) Move the checks for relay_from_hosts and authenticated clients from
2458 after to before the (commented out) DNS black list checks.
2460 (2) Add control=submission to the relay_from_hosts and authenticated
2461 clients checks, on the grounds that messages accepted by these
2462 statements are most likely to be submissions.
2464 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2466 (1) Generate an error if the third argument for the ${prvs expansion is
2469 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2472 (3) Reset the variables that are obtained from the first argument of
2473 ${prvscheck and used in the second argument before leaving the code,
2474 because their memory is reclaimed, so using them afterwards may do
2477 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2478 one (it's much easier than Tom thought :-).
2480 (5) Because of (4), we can now allow for the use of $prvscheck_result
2481 inside the third argument.
2483 PH/03 For some reason, the default setting of PATH when running a command from
2484 a pipe transport was just "/usr/bin". I have changed it to
2487 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2488 anything to be listed in the output from -bV.
2490 PH/05 When a filter generated an autoreply, the entire To: header line was
2491 quoted in the delivery log line, like this:
2493 => >A.N.Other <ano@some.domain> <original@ddress> ...
2495 This has been changed so that it extracts the operative address. There
2496 may be more than one such address. If so, they are comma-separated, like
2499 => >ano@some.domain,ona@other.domain <original@ddress> ...
2501 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2502 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2503 not being looked up in the rDNS to get a host name, Exim was showing the
2504 IP address twice in Received: lines, even though the IP addresses were
2505 identical. For example:
2507 Received: from [1.2.3.4] (helo=[1.2.3.4])
2509 However, if the real host name was known, it was omitting the HELO data
2510 if it matched the actual IP address. This has been tidied up so that it
2511 doesn't show the same IP address twice.
2513 PH/07 When both +timestamp and +memory debugging was on, the value given by
2514 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2515 called by the debug printing, thereby overwriting the timestamp buffer.
2516 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2518 PH/08 When the original message was included in an autoreply transport, it
2519 always said "this is a copy of the message, including all the headers",
2520 even if body_only or headers_only was set. It now gives an appropriate
2523 PH/09 Applied a patch from the Sieve maintainer which:
2525 o fixes some comments
2526 o adds the (disabled) notify extension core
2527 o adds some debug output for the result of if/elsif tests
2528 o points to the current vacation draft in the documentation
2529 and documents the missing references header update
2533 o fixes a bug in processing the envelope test (when testing
2534 multiple envelope elements, the last element determinted the
2537 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2538 Electronic Mail") by including:
2540 Auto-submitted: auto-generated
2542 in the messages that it generates (bounce messages and others, such as
2543 warnings). In the case of bounce messages for non-SMTP mesages, there was
2544 also a typo: it was using "Auto_submitted" (underscore instead of
2545 hyphen). Since every message generated by Exim is necessarily in response
2546 to another message, thes have all been changed to:
2548 Auto-Submitted: auto-replied
2550 in accordance with these statements in the RFC:
2552 The auto-replied keyword:
2554 - SHOULD be used on messages sent in direct response to another
2555 message by an automatic process,
2557 - MUST NOT be used on manually-generated messages,
2559 - MAY be used on Delivery Status Notifications (DSNs) and Message
2560 Disposition Notifications (MDNs),
2562 - MUST NOT be used on messages generated by automatic or periodic
2563 processes, except for messages which are automatic responses to
2566 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2567 to the default Received: header definition.
2569 PH/12 Added log selector acl_warn_skipped (default on).
2571 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2572 variables because (a) they are in the wrong storage pool and (b) even if
2573 they were copied, it wouldn't work properly because of the caching.
2575 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2576 checking when decoding. Apparently there are clients that generate
2577 overlong encoded strings. Why am I not surprised?
2579 PH/15 If the first argument of "${if match_address" was not empty, but did not
2580 contain an "@" character, Exim crashed. Now it writes a panic log message
2581 and treats the condition as false.
2583 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2585 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2586 extension "envelope-auth". The code is finished and in agreement with
2587 other implementations, but there is no documentation so far and in fact,
2588 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2589 not changing the active code.
2591 Print executed "if" and "elsif" statements when debugging is used. This
2592 helps a great deal to understand what a filter does.
2594 Document more things not specified clearly in RFC3028. I had all this
2595 sorted out, when out of a sudden new issues came to my mind. Oops."
2597 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2600 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2601 canonical form (as documented). However, after a host name lookup from
2602 the IP address, check_host() was doing a simple string comparison with
2603 addresses acquired from the DNS when checking that the found name did
2604 have the original IP as one of its addresses. Since any found IPv6
2605 addresses are likely to be in abbreviated form, the comparison could
2606 fail. Luckily, there already exists a function for doing the comparison
2607 by converting both addresses to binary, so now that is used instead of
2608 the text comparison.
2610 PH/20 There was another similar case to PH/19, when a complete host name was
2611 given in a host list; looking up its IP address could give an abbreviated
2612 form, whereas the current host's name might or might not be abbreviated.
2613 The same fix has been applied.
2619 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2620 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2623 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2624 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2626 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2628 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2629 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2630 Also, for "Auto-Submitted", treat anything other than "no" as
2631 non-personal, in accordance with RFC 3834. (Previously it treated
2632 anything starting "auto-" as non-personal.)
2634 TF/01 The control=submission/name=... option had a problem with syntax
2635 errors if the name included a slash character. The /name= option
2636 now slurps the rest of the string, so it can include any characters
2637 but it must come last in the list of options (after /sender_retain
2640 PH/05 Some modifications to the interface to the fake nameserver for the new
2648 TK/01 Added the "success_on_redirect" address verification option. See
2649 NewStuff for rationale and an example.
2651 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2653 PH/02 Patch to exigrep to allow it to work on syslog lines.
2655 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2656 fread() to skip over the body file's header line, because in Cygwin the
2657 header line is locked and is inaccessible.
2659 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2660 co-exist for some time) to make it clear that it is the Exim ID that is
2661 referenced, not the Message-ID: header line.
2663 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2664 string_format() function, because snprintf() does not exist on all
2667 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2668 ${stat: expansion item.
2670 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2671 protocol synchronization error", to keep the pedants happy.
2673 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2674 well as for IRIX systems, when gcc is being used. See the host.c source
2677 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2679 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2682 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2683 transport and to the smtp transport in LMTP mode.
2685 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2687 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2688 run a filter in a subprocess. This could lead to confusion in subsequent
2689 lookups in the parent process. There should also be a search_tidyup() at
2690 the end of the subprocess.
2692 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2693 only if the host matched helo_try_verify_hosts, which caused the
2694 verification to occur when the EHLO/HELO command was issued. The ACL just
2695 tested the remembered result. Now, if a previous verification attempt has
2696 not happened, "verify = helo" does it there and then.
2698 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2700 TK/03 Fix log output including CR from clamd.
2702 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2703 exist provoked a memory error which could cause a segfault.
2705 PH/15 Installed PCRE 6.2
2707 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2709 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2710 of the problem. Specifically, suggested +O2 rather than +O1 for the
2713 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2715 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2716 record, the port was not correctly logged when the outgoing_port log
2717 selector was set (it logged the transort's default port).
2719 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2720 fallback_hosts, and "hosts" in the smtp transport.
2722 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2723 host errors such as "Connection refused".
2725 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2726 authentication with radiusclient 0.4.9:
2728 - Error returned from rc_read_config was caught wrongly
2729 - Username/password not passed on to radius server due to wrong length.
2731 The presumption is that some radiusclient API changes for 4.51/PH/17
2732 were not taken care of correctly. The code is still untested by me (my
2733 Linux distribution still has 0.3.2 of radiusclient), but it was
2734 contributed by a Radius user.
2736 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2737 expanding the "port" option of the smtp transport.
2739 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2740 while reading a MIME header. Thanks to Tom Hughes for a patch.
2742 PH/24 Include config.h inside local_scan.h so that configuration settings are
2745 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2746 that in an ACL for RCPT (for example), you can examine exactly what was
2749 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2750 commands, but it was not correctly comparing the address with the actual
2751 client host address. Thus, it would show the EHLO address in Received:
2752 header lines when this was not necessary.
2754 PH/27 Added the % operator to ${eval:}.
2756 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2757 it should be ignoring failures (because with -C, for example, it has lost
2758 privilege). It wasn't ignoring creation failures other than "already
2761 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2764 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2765 because an input line was too long, either on its own, or by virtue of
2766 too many continuations, the temporary file was not being removed, and the
2767 return code was incorrect.
2769 PH/31 Missing "BOOL" in function definition in filtertest.c.
2771 PH/32 Applied Sieve patches from the maintainer.
2773 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2775 PH/33 Added "verify = not_blind".
2777 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2778 Local/Makefile (with some defaults set). These are used in built scripts
2779 such as exicyclog, but they have never been used in the exim_install
2780 script (though there are many overriding facilities there). I have
2781 arranged that the exim_install script now takes note of these two
2784 PH/35 Installed configuration files for Dragonfly.
2786 PH/36 When a locally submitted message by a trusted user did not contain a
2787 From: header, and the sender address was obtained from -f or from an SMTP
2788 MAIL command, and the trusted user did not use -F to supply a sender
2789 name, $originator_name was incorrectly used when constructing a From:
2790 header. Furthermore, $originator_name was used for submission mode
2791 messages from external hosts without From: headers in a similar way,
2792 which is clearly wrong.
2794 PH/37 Added control=suppress_local_fixups.
2796 PH/38 When log_selector = +received_sender was set, and the addition of the
2797 sender made the log line's construction buffer exactly full, or one byte
2798 less than full, an overflow happened when the terminating "\n" was
2801 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2802 when the result of a list match is failure because a DNS lookup failed.
2804 PH/40 RM_COMMAND is now used in the building process.
2806 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2807 the "build-* directories that it finds.
2809 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2810 address in a domain literal was a prefix of an interface address.
2812 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2813 when verifying a sender address, unless rewrite_headers is false.
2815 PH/44 Wrote a long comment about why errors_to addresses are verified as
2816 recipients, not senders.
2818 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2819 the ratelimit ACL was added.
2821 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2823 PH/46 Added extra information about PostgreSQL errors to the error string.
2825 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2826 avoiding the need to install special zones in a real server. This is
2827 backwards compatible; if it can't find the fake resolver, it drops back.
2828 Thus, both old and new test suites can be run.
2830 TF/02 Added util/ratelimit.pl
2832 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2833 clock is set back in time.
2835 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2836 Candler <B.Candler@pobox.com>.
2838 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2839 OK for addresses that are the result of redirections.
2841 PH/48 A number of further additions for the benefit of the new test suite,
2842 including a fake gethostbyname() that interfaces to the fake DNS resolver
2845 TF/06 The fix for widen_domains has also been applied to qualify_single and
2846 search_parents which are the other dnslookup options that can cause
2849 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2852 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2854 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2855 service" if it encountered an MX record with an empty target hostname.
2856 The message is now "an MX or SRV record indicated no SMTP service".
2858 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2859 if the DNS of the sending site is misconfigured. This is quite a
2860 common situation. This change restores the behaviour of treating a
2861 helo verification defer as a failure.
2863 PH/51 If self=fail was set on a router, the bounce message did not include the
2864 actual error message.
2870 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2872 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2873 command itself ended in error, the underlying message about the transport
2874 filter timeout was being overwritten with the pipe command error. Now the
2875 underlying error message should be appended to the second error message.
2877 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2879 PH/02 Reduce the amount of output that "make" produces by default. Full output
2880 can still be requested.
2882 PH/03 The warning log line about a condition test deferring for a "warn" verb
2883 was being output only once per connection, rather than after each
2884 occurrence (because it was using the same function as for successful
2885 "warn" verbs). This seems wrong, so I have changed it.
2887 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2888 it should not have, which might have caused a crash in the right
2889 circumstances, but probably never did.
2891 PH/04 Installed a modified version of Tony Finch's patch to make submission
2892 mode fix the return path as well as the Sender: header line, and to
2893 add a /name= option so that you can make the user's friendly name appear
2896 TF/03 Added the control = fakedefer ACL modifier.
2898 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2899 Mark Lowes for thorough testing.
2901 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2903 TK/03 Merged latest SRS patch from Miles Wilton.
2905 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2906 with the definition in sysexits.h (which is #included earlier).
2907 Fortunately, Exim does not actually use EX_OK. The code used to try to
2908 preserve the sysexits.h value, by assumimg that macro definitions were
2909 scanned for macro replacements. I have been disabused of this notion,
2910 so now the code just undefines EX_OK before #including unistd.h.
2912 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2913 in the smtp transport. When a block could not be written in a single
2914 write() function, the timeout was being re-applied to each part-write.
2915 This seems wrong - if the receiver was accepting one byte at a time it
2916 would take for ever. The timeout is now adjusted when this happens. It
2917 doesn't have to be particularly precise.
2919 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2920 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2922 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2923 without a domain if the check to <postmaster@domain> fails.
2925 SC/01 Eximstats: added -xls and the ability to specify output files
2926 (patch written by Frank Heydlauf).
2928 SC/02 Eximstats: use FileHandles for outputing results.
2930 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2932 SC/04 Eximstats: fixed display of large numbers with -nvr option
2934 SC/05 Eximstats: fixed merging of reports with empty tables.
2936 SC/06 Eximstats: added the -include_original_destination flag
2938 SC/07 Eximstats: removed tabs and trailing whitespace.
2940 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2942 TK/06 MBOX spool code: Add real "From " MBOX separator line
2943 so the .eml file is really in mbox format (even though
2944 most programs do not really care). Patch from Alex Miller.
2946 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2947 The latter is generated from $received_to and is only set if the
2948 message has one envelope recipient. SA can use these headers,
2949 obviously out-of-the-box. Patch from Alex Miller.
2951 PH/08 The ${def test on a variable was returning false if the variable's
2952 value was "0", contrary to what the specification has always said!
2953 The result should be true unless the variable is empty.
2955 PH/09 The syntax error of a character other than { following "${if
2956 def:variable_name" (after optional whitespace) was not being diagnosed.
2957 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2958 accidental colon was present, for example, could give incorrect results.
2960 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2961 result is used (not including appendfile, where other changes are about
2964 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2965 This involved changing a lot of size variables from int to off_t. It
2966 should work with maildirs and everything.
2968 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2969 spamd dying while we are connected to it.
2971 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2974 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2975 function for BDB 4.3.
2977 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2979 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2980 the sources for PCRE have been re-arranged and more files are now
2983 PH/15 The code I had for printing potentially long long variables in PH/11
2984 above was not the best (it lost precision). The length of off_t variables
2985 is now inspected at build time, and an appropriate printing format (%ld
2986 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2987 to be "long long int" or "long int". This is needed for the internal
2988 formatting function string_vformat().
2990 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2991 the configuration file to be ":syslog", then the script "guesses" where
2992 the logs files are, rather than using the compiled in default. In our
2993 case the guess is not the same as the compiled default, so the script
2994 suddenly stopped working when I started to use syslog. The patch checks
2995 to see if log_file_path is "". If so, it attempts to read it from exim
2996 with no configuration file to get the compiled in version, before it
2997 falls back to the previous guessing code."
2999 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
3000 implementing BATV in an Exim configuration. See NewStuff for the gory
3003 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
3004 Makefile that are specific to HP-UX.
3006 PH/18 If the "use_postmaster" option was set for a recipient callout together
3007 with the "random" option, the postmaster address was used as the MAIL
3008 FROM address for the random test, but not for the subsequent recipient
3009 test. It is now used for both.
3011 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
3012 patch removes a few documentation additions to RFC 3028, because the
3013 latest draft now contains them. It adds the new en;ascii-case comparator
3014 and a new error check for 8bit text in MIME parts. Comparator and
3015 require names are now matched exactly. I enabled the subaddress
3016 extension, but it is not well tested yet (read: it works for me)."
3018 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
3019 rework some of the code of TK/09 above to avoid the hardwired use of
3020 "%lld" and "long long". Replaced the call to snprintf() with a call to
3023 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
3024 records point to non-existent hosts", "retry timeout exceeded", and
3025 "retry time not reached for any host after a long failure period".
3027 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
3028 experimental DomainKeys support:
3030 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
3031 (2) On an error such as an illegally used "control", the wrong name for
3032 the control was given.
3034 These problems did NOT occur unless DomainKeys support was compiled.
3036 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3038 PH/24 Added ${if match_ip condition.
3040 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3041 databases so that it will be absolutely obvious if a crash occurs in the
3042 DB library. This is a regular occurrence (often caused by mis-matched
3045 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3046 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3047 picked up on a user's system that detects such things. There doesn't seem
3048 to be a gcc warning option for this - only an attribute that has to be
3049 put on the function's prototype. It seems that in Fedora Core 4 they have
3050 set this on a number of new functions. No doubt there will be more in due
3053 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3054 specify a transport. However, if an address that was verified by such a
3055 router was the subject of a callout, Exim crashed because it tried to
3056 read the rcpt_include_affixes from the non-existent transport. Now it
3057 just assumes that the setting of that option is false. This bug was
3058 introduced by 4.51/PH/31.
3060 PH/28 Changed -d+all to exclude +memory, because that information is very
3061 rarely of interest, but it makes the output a lot bigger. People tend to
3062 do -d+all out of habit.
3064 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3065 code in os-type was giving problems when libc.so lives in lib64, like on
3068 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3069 aren't the modern standard, and it seems that some systems' include files
3070 don't always have them. Exim was already checking for some of the newer
3071 ones like T_AAAA, and defining it itself. I've added checks for all the
3072 record types that Exim uses.
3074 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3075 not automatically generating a new one, as it is supposed to. This
3076 prevented TLS from working. If the file did exist, but contained invalid
3077 data, a new version was generated, as expected. It was only the case of a
3078 non-existent file that was broken.
3080 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3081 with a change in libdomainkeys > 0.64.
3083 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3084 from DNS. If the selector record carries the flag, it now has
3085 precedence over the domain-wide flag.
3087 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3089 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3090 the use of a port name that isn't defined in /etc/services) occurred, the
3091 message was deferred as in a normal delivery, and thus remained on the
3092 spool, instead of being failed because of the mua_wrapper setting. This
3093 is now fixed, and I tidied up some of the mua_wrapper messages at the
3096 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3097 the messages in a hash of arrays rather than using individual hashes.
3098 This is a bit cleaner and results in dramatic memory savings, albeit
3099 at a slight CPU cost.
3101 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3102 as requested by Marc Sherman.
3104 SC/10 Eximstats: added histograms for user specified patterns as requested
3107 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3109 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3110 fopen() in the content-scanning modules that did not already have it.
3116 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3117 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3119 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3121 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3123 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3124 to test Sieve filters that use "vacation".
3126 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3127 that changes the way the GnuTLS parameters are stored in the cache file.
3128 The new format can be generated externally. For backward compatibility,
3129 if the data in the cache doesn't make sense, Exim assumes it has read an
3130 old-format file, and it generates new data and writes a new file. This
3131 means that you can't go back to an older release without removing the
3134 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3135 work if there are any delivery delays because "one_time" forces the
3136 parent to be marked "delivered", so its unseen clone is never tried
3137 again. For this reason, Exim now forbids the simultaneous setting of
3140 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3141 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3142 entry if you want to know the details. The fix, however, neglected to
3143 consider the case when local delivery batching is involved. The test for
3144 "previously delivered" was not happening when checking to see if an
3145 address could be batched with a previous (undelivered) one; under
3146 certain circumstances this could lead to multiple deliveries to the same
3149 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3150 in its include files, and this causes problems building Exim.
3152 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3153 header_syntax) but Exim was just ignoring anything given after a slash.
3154 In particular, this caused confusion with an attempt to use "verify =
3155 reverse_host_lookup/defer_ok". An error is now given when options are
3156 supplied for verify items that do not have them. (Maybe reverse_host_
3157 lookup should have a defer_ok option, but that's a different point.)
3159 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3160 defined by RFC 821) to 2048, because there were problems with some AUTH
3161 commands, and RFC 1869 says the size should be increased for extended
3162 SMTP commands that take arguments.
3164 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3167 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3168 "unknown" error; now it says that the functionality isn't in the binary.
3170 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3171 an address' error message when a string expansion fails (syntax or
3172 whatever). Otherwise the password may appear in the log. Following change
3173 PH/42 below, there is no longer a chance of it appearing in a bounce
3176 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3178 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3179 of its addresses was ever tried. (Bugzilla bug #2.)
3181 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3182 the result incorrectly in the debug output. (It correctly added a newline
3183 to what was transported.)
3185 TF/01 Added $received_time.
3187 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3188 commented out examples of how to interface to a virus scanner and to
3189 SpamAssassin. Also added commented examples of av_scanner and
3190 spamd_address settings.
3192 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3193 and controls are allowed in which ACLs. There were a couple of minor
3194 errors. Some of the entries in the conditions table (which is a table of
3195 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3196 negation of where the condition IS allowed.
3198 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3200 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3201 header file does not have a version number, so I've had to invent a new
3202 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3203 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3204 radiusclient), but it was contributed by a Radius user.
3206 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3207 files or directories, for OpenSSL.
3209 PH/19 When an Exim process that is running as root has to create an Exim log
3210 file, it does so in a subprocess that runs as exim:exim so as to get the
3211 ownership right at creation (otherwise, other Exim processes might see
3212 the file with the wrong ownership). There was no test for failure of this
3213 fork() call, which would lead to the process getting stuck as it waited
3214 for a non-existent subprocess. Forks do occasionally fail when resources
3215 run out. I reviewed all the other calls to fork(); they all seem to check
3218 PH/20 When checking for unexpected SMTP input at connect time (before writing
3219 the banner), Exim was not dealing correctly with a non-positive return
3220 from the read() function. If the client had disconnected by this time,
3221 the result was a log entry for a synchronization error with an empty
3222 string after "input=" when read() returned zero. If read() returned -1
3223 (an event I could not check), uninitialized data bytes were printed.
3224 There were reports of junk text (parts of files, etc) appearing after
3227 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3229 PH/22 Added support for macro redefinition, and (re)definition in between
3230 driver and ACL definitions.
3232 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3233 forgetting to use the resulting value; it was using the unexpanded value.
3235 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3236 hadn't been configured. The fix is from Juergen Kreileder, who
3237 understands it better than I do:
3239 "Here's what I see happening with three configured cyrus_sasl
3240 authenticators configured (plain, login, cram-md5):
3242 On startup auth_cyrus_sasl_init() gets called for each of these.
3243 This means three calls to sasl_listmech() without a specified mech_list.
3244 => SASL tests which mechs of all available mechs actually work
3245 => three warnings about OTP not working
3246 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3248 With the patch, sasl_listmech() also gets called three times. But now
3249 SASL's mech_list option is set to the server_mech specified in the the
3250 authenticator. Or in other words, the answer from sasl_listmech()
3251 gets limited to just the mech you're testing for (which is different
3253 => the return list contains just 'plain' or 'login', 'cram-md5' or
3254 nothing depending on the value of ob->server_mech.
3256 I've just tested the patch: Authentication still works fine,
3257 unavailable mechs specified in the exim configuration are still
3258 caught, and the auth.log warnings about OTP are gone."
3260 PH/25 When debugging is enabled, the contents of the command line are added
3261 to the debugging output, even when log_selector=+arguments is not
3264 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3265 answer is "GNU", and only if the return is "GNU/something" is the answer
3268 PH/27 $acl_verify_message is now set immediately after the failure of a
3269 verification in an ACL, and so is available in subsequent modifiers. In
3270 particular, the message can be preserved by coding like this:
3272 warn !verify = sender
3273 set acl_m0 = $acl_verify_message
3275 Previously, $acl_verify_message was set only while expanding "message"
3276 and "log_message" when a very denied access.
3278 PH/28 Modified OS/os.c-Linux with
3280 -#ifndef OS_LOAD_AVERAGE
3281 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3283 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3284 nomenclature these days.)
3286 PH/29 Installed patch from the Sieve maintainer that adds the options
3287 sieve_useraddress and sieve_subaddress to the redirect router.
3289 PH/30 In these circumstances:
3290 . Two addresses routed to the same list of hosts;
3291 . First host does not offer TLS;
3292 . First host accepts first address;
3293 . First host gives temporary error to second address;
3294 . Second host offers TLS and a TLS session is established;
3295 . Second host accepts second address.
3296 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3297 and peerdn, if requested) that were in fact used only for the second
3300 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3301 attention to any local part prefix or suffix that was matched by the
3302 router that accepted the address. It now behaves in the same way as it
3303 does for delivery: the affixes are removed from the local part unless
3304 rcpt_include_affixes is set on the transport.
3306 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3307 timeout during the DATA phase of an incoming message.
3309 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3310 applied a patch sanctioned by the Sieve maintainer.
3312 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3313 the uid or gid is negative. A case of a negative gid caused this to be
3314 noticed. The fix allows for either to be negative.
3316 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3317 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3320 PH/36 The change PH/12 above was broken. Fixed it.
3322 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3323 the grounds that routing the same address twice would always produce the
3324 same answer. This might have been true once, but it is certainly no
3325 longer true now. Routing a child address may depend on the previous
3326 routing that produced that child. Some complicated redirection strategies
3327 went wrong when messages had multiple recipients, and made Exim's
3328 behaviour dependent on the order in which the addresses were given.
3330 I have moved the duplicate checking until after the routing is complete.
3331 Exim scans the addresses that are assigned to local and remote
3332 transports, and removes any duplicates. This means that more work will be
3333 done, as duplicates will always all be routed, but duplicates are
3334 presumably rare, so I don't expect this is of any significance.
3336 For deliveries to pipes, files, and autoreplies, the duplicate checking
3337 still happens during the routing process, since they are not going to be
3340 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3341 It corrects a timeout issue with spamd. This is Ian's comment: "The
3342 background is that sometimes spamd either never reads data from a
3343 connection it has accepted, or it never writes response data. The exiscan
3344 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3345 blindly assumes that writes won't block so it may never time out."
3347 PH/39 Allow G after quota size as well as K and M.
3349 PH/40 The value set for $authenticated_id in an authenticator may not contain
3350 binary zeroes or newlines because the value is written to log lines and
3351 to spool files. There was no check on this. Now the value is run through
3352 the string_printing() function so that such characters are converted to
3353 printable escape sequences.
3355 PH/41 $message_linecount is a new variable that contains the total number of
3356 lines in the message. Compare $body_linecount, which is the count for the
3359 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3360 bounce and delay warning messages, except in certain special cases, which
3363 (a) An SMTP error message from a remote host;
3364 (b) A message specified in a :fail: redirection;
3365 (c) A message specified in a "fail" command in a system filter;
3366 (d) A message specified in a FAIL return from the queryprogram router;
3367 (e) A message specified by the cannot_route_message router option.
3369 In these cases only, Exim does include the error details in bounce and
3370 warning messages. There are also a few cases where bland messages such
3371 as "unrouteable address" or "local delivery error" are given.
3373 PH/43 $value is now also set for the "else" part of a ${run expansion.
3375 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3376 being worked on, but at least Exim now implements the latest version to
3379 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3380 process to complete was treated as a delivery failure, a timeout while
3381 writing the message to the pipe was logged, but erroneously treated as a
3382 successful delivery. Such timeouts include transport filter timeouts. For
3383 consistency with the overall process timeout, these timeouts are now
3384 treated as errors, giving rise to delivery failures by default. However,
3385 there is now a new Boolean option for the pipe transport called
3386 timeout_defer, which, if set TRUE, converts the failures into defers for
3387 both kinds of timeout. A transport filter timeout is now identified in
3390 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3391 systems where "make" and "gmake" are different, calling "gmake" at top
3392 level broke things. I've arranged for the value of $(MAKE) to be passed
3393 from the Makefile to this script so that it can call the same version of
3397 A note about Exim versions 4.44 and 4.50
3398 ----------------------------------------
3400 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3401 changes of various kinds. As a consequence, a big documentation update was
3402 needed. This delayed the release for rather longer than seemed good, especially
3403 in the light of a couple of (minor) security issues. Therefore, the changes
3404 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3405 release. So 4.44 and 4.50 are in effect two different branches that both start
3408 I have left the 4.50 change log unchanged; it contains all the changes since
3409 4.43. The change log for 4.44 is below; many of its items are identical to
3410 those for 4.50. This seems to be the most sensible way to preserve the
3411 historical information.
3417 1. Minor wording change to the doc/README.SIEVE file.
3419 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3420 computation of the current number of files was incorrect.
3422 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3423 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3426 4. Give more explanation in the error message when the command for a transport
3427 filter fails to execute.
3429 5. There are several places where Exim runs a non-Exim command in a
3430 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3431 was being done only for the command run by the queryprogram router. It is
3432 now done for all such subprocesses. The other cases are: ${run, transport
3433 filters, and the commands run by the lmtp and pipe transports.
3435 6. Added CONFIGURE_GROUP build-time option.
3437 7. Some older OS have a limit of 256 on the maximum number of file
3438 descriptors. Exim was using setrlimit() to set 1000 as a large value
3439 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3440 systems. I've change it so that if it can't get 1000, it tries for 256.
3442 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3443 was an oversight, and furthermore, ever since the addition of extra
3444 controls (e.g. 4.43/32), the checks on when to allow different forms of
3445 "control" were broken. There should now be diagnostics for all cases when a
3446 control that does not make sense is encountered.
3448 9. Added the /retain_sender option to "control=submission".
3450 10. $recipients is now available in the predata ACL (oversight).
3452 11. Tidy the search cache before the fork to do a delivery from a message
3453 received from the command line. Otherwise the child will trigger a lookup
3454 failure and thereby defer the delivery if it tries to use (for example) a
3455 cached ldap connection that the parent has called unbind on.
3457 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3458 of $address_data from the recipient verification was clobbered by the
3459 sender verification.
3461 13. The value of address_data from a sender verification is now available in
3462 $sender_address_data in subsequent conditions in the ACL statement.
3464 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3466 15. Added a new option "connect=<time>" to callout options, to set a different
3469 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3470 was its contents. (It was OK if the option was not defined at all.)
3472 17. A "Completed" log line is now written for messages that are removed from
3473 the spool by the -Mrm option.
3475 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3476 information about exactly what failed.
3478 19. Added -dd to debug only the daemon process.
3480 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3481 handles timeouts, both on the server side and network timeouts. Renamed the
3482 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3484 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3485 It is now set to "smtps".
3487 22. $host_address is now set to the target address during the checking of
3488 ignore_target_hosts.
3490 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3491 being passed; this would have caused $sender_host_name to have been used if
3492 matching the list had actually called for a host name (not very likely,
3493 since this list is usually IP addresses). A host name is now passed as
3496 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3497 code that specifies a non-restarting handler (typically sigaction() in
3498 modern systems) in an attempt to fix a rare and obscure crash bug.
3500 25. Narrowed the window for a race in the daemon that could cause it to ignore
3501 SIGCHLD signals. This is not a major problem, because they are used only to
3502 wake it up if nothing else does.
3504 26. A malformed maildirsize file could cause Exim to calculate negative values
3505 for the mailbox size or file count. Odd effects could occur as a result.
3506 The maildirsize information is now recalculated if the size or filecount
3509 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3510 support for a long time. Removed HAVE_SYS_VFS_H.
3512 28. Installed the latest version of exipick from John Jetmore.
3514 29. In an address list, if the pattern was not a regular expression, an empty
3515 subject address (from a bounce message) matched only if the pattern was an
3516 empty string. Non-empty patterns were not even tested. This was the wrong
3517 because it is perfectly reasonable to use an empty address as part of a
3518 database query. An empty address is now tested by patterns that are
3519 lookups. However, all the other forms of pattern expect the subject to
3520 contain a local part and a domain, and therefore, for them, an empty
3521 address still always fails if the pattern is not itself empty.
3523 30. Exim went into a mad DNS loop when attempting to do a callout where the
3524 host was specified on an smtp transport, and looking it up yielded more
3525 than one IP address.
3527 31. Re-factored the code for checking spool and log partition space into a
3528 function that finds that data and another that does the check. The former
3529 is then used to implement four new variables: $spool_space, $log_space,
3530 $spool_inodes, and $log_inodes.
3532 32. The RFC2047 encoding function was originally intended for short strings
3533 such as real names; it was not keeping to the 75-character limit for
3534 encoded words that the RFC imposes. It now respects the limit, and
3535 generates multiple encoded words if necessary. To be on the safe side, I
3536 have increased the buffer size for the ${rfc2047: expansion operator from
3539 33. It is now permitted to omit both strings after an "if" condition; if the
3540 condition is true, the result is "true". As before, when the second string
3541 is omitted, a false condition yields an empty string. This makes it less
3542 cumbersome to write custom ACL and router conditions.
3544 34. Failure to deliver a bounce message always caused it to be frozen, even if
3545 there was an errors_to setting on the router. The errors_to setting is now
3548 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3549 canonical form (fully expanded) before being placed in
3550 $sender_host_address.
3552 36. The table in the code that translates DNS record types into text (T_A to
3553 "A" for instance) was missing entries for NS and CNAME. It is just possible
3554 that this could have caused confusion if both these types were looked up
3555 for the same domain, because the text type is used as part of Exim's
3556 per-process caching. But the chance of anyone hitting this buglet seems
3559 37. The dnsdb lookup has been extended in a number of ways.
3561 (1) There is a new type, "zns", which walks up the domain tree until it
3562 finds some nameserver records. It should be used with care.
3564 (2) There is a new type, "mxh", which is like "mx" except that it returns
3565 just the host names, not the priorities.
3567 (3) It is now possible to give a list of domains (or IP addresses) to be
3568 looked up. The behaviour when one of the lookups defers can be
3569 controlled by a keyword.
3571 (4) It is now possible to specify the separator character for use when
3572 multiple records are returned.
3574 38. The dnslists ACL condition has been extended: it is now possible to supply
3575 a list of IP addresses and/or domains to be looked up in a particular DNS
3578 39. Added log_selector=+queue_time_overall.
3580 40. When running the queue in the test harness, wait just a tad after forking a
3581 delivery process, to get repeatability of debugging output.
3583 41. Include certificate and key file names in error message when GnuTLS fails
3584 to set them up, because the GnuTLS error message doesn't include the name
3585 of the failing file when there is a problem reading it.
3587 42. Allow both -bf and -bF in the same test run.
3589 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3591 44. The "Exiscan patch" is now merged into the mainline Exim source.
3593 45. Sometimes the final signoff response after QUIT could fail to get
3594 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3595 < 0 before doing a fflush(). This bug looks as though it goes back to the
3596 introduction of TLS in release 3.20, but "sometimes" must have been rare
3597 because the tests only now provoked it.
3599 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3600 (this can affect the format of dates).
3602 47. exim_tidydb, when checking for the continued existence of a message for
3603 which it has found a message-specific retry record, was not finding
3604 messages that were in split spool directories. Consequently, it was
3605 deleting retry records that should have stayed in existence.
3607 48. Steve fixed some bugs in eximstats.
3609 49. The SPA authentication driver was not abandoning authentication and moving
3610 on to the next authenticator when an expansion was forced to fail,
3611 contradicting the general specification for all authenticators. Instead it
3612 was generating a temporary error. It now behaves as specified.
3614 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3615 (the order specifies the preference for clients). The order is now AES256,
3616 AES128, 3DES, ARCFOUR128.
3618 51. Small patch to Sieve code - explicitly set From: when generating an
3621 52. Exim crashed if a remote delivery caused a very long error message to be
3622 recorded - for instance if somebody sent an entire SpamAssassin report back
3623 as a large number of 550 error lines. This bug was coincidentally fixed by
3624 increasing the size of one of Exim's internal buffers (big_buffer) that
3625 happened as part of the Exiscan merge. However, to be on the safe side, I
3626 have made the code more robust (and fixed the comments that describe what
3629 53. Now that there can be additional text after "Completed" in log lines (if
3630 the queue_time_overall log selector is set), a one-byte patch to exigrep
3631 was needed to allow it to recognize "Completed" as not the last thing in
3634 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3635 patch that reportedly fixes this has been added. I am not expert enough to
3636 create a test for it. This is what the patch creator wrote:
3638 "I found a little strange behaviour of ldap code when working with
3639 Windows 2003 AD Domain, where users was placed in more than one
3640 Organization Units. When I tried to give exim partial DN, the exit code
3641 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3642 But simultaneously result of request was absolutely normal ldap result,
3643 so I produce this patch..."
3645 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3646 REFERENCE, so I have modified the code to exclude the patch when that macro
3649 55. Some experimental protocols are using DNS PTR records for new purposes. The
3650 keys for these records are domain names, not reversed IP addresses. The
3651 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3652 leaves it alone. Component reversal etc. now happens only for IP addresses.
3655 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3657 57. Double the size of the debug message buffer (to 2048) so that more of very
3658 long debug lines gets shown.
3660 58. The exicyclog utility now does better if the number of log files to keep
3661 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3663 59. Two changes related to the smtp_active_hostname option:
3665 (1) $smtp_active_hostname is now available as a variable.
3666 (2) The default for smtp_banner uses $smtp_active_hostname instead
3667 of $primary_hostname.
3669 60. The host_aton() function is supposed to be passed a string that is known
3670 to be a valid IP address. However, in the case of IPv6 addresses, it was
3671 not checking this. This is a hostage to fortune. Exim now panics and dies
3672 if the condition is not met. A case was found where this could be provoked
3673 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3674 components; fortuitously, this particular loophole had already been fixed
3675 by change 4.50/55 above.
3677 If there are any other similar loopholes, the new check in host_aton()
3678 itself should stop them being exploited. The report I received stated that
3679 data on the command line could provoke the exploit when Exim was running as
3680 exim, but did not say which command line option was involved. All I could
3681 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3682 running as the user.
3685 61. There was a buffer overflow vulnerability in the SPA authentication code
3686 (which came originally from the Samba project). I have added a test to the
3687 spa_base64_to_bits() function which I hope fixes it.
3690 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3691 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3693 63. The daemon start-up calls getloadavg() while still root for those OS that
3694 need the first call to be done as root, but it missed one case: when
3695 deliver_queue_load_max is set with deliver_drop_privilege. This is
3696 necessary for the benefit of the queue runner, because there is no re-exec
3697 when deliver_drop_privilege is set.
3699 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3700 This has been fixed.
3702 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3703 was in use, was not putting the data itself into the right store pool;
3704 consequently, it could be overwritten for a subsequent message in the same
3705 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3708 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3710 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3711 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3712 the function were treating the return as a boolean value, which happened to
3713 work because 0=false and not-0=true, but is not correct code.
3715 68. The host_aton() function was not handling scoped IPv6 addresses (those
3716 with, for example, "%eth0" on the end) correctly.
3718 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3719 negated items (that is, ~something) in unsigned ints. Some compilers
3720 apparently mutter when there is no cast.
3722 70. If an address verification called from an ACL failed, and did not produce a
3723 user-specific message (i.e. there was only a "system" message), nothing was
3724 put in $acl_verify_message. In this situation, it now puts the system
3727 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3728 session; change 4.31/43 added the unwanted input to the log line - except
3729 that it did not do this in the start of session case. It now does.
3731 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3732 This is wrong and can cause the other end to generate a synchronization
3733 error if it is another Exim or anything else that does the synchronization
3734 check. A QUIT command is no longer sent after a timeout.
3736 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3737 during host lookups.
3739 74. The defer_ok option of callout verification was not working if it was used
3740 when verifying addresses in header lines, that is, for this case:
3742 verify = header_sender/callout=defer_ok
3744 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3745 those file descriptors could be used for SMTP connections. If anything
3746 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3747 could be sent to the SMTP client, causing chaos. The daemon now opens
3748 stdin, stdout, and stderr to /dev/null when it puts itself into the
3751 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3752 log by default. The user can override this with suitable Perl magic.
3754 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3755 the log message when discard triggers, was not working for the DATA ACL or
3756 for the non-SMTP ACL.
3758 78. Error message wording change in sieve.c.
3760 79. If smtp_accept_max_per_host was set, the number of connections could be
3761 restricted to fewer than expected, because the daemon was trying to set up
3762 a new connection before checking whether the processes handling previous
3763 connections had finished. The check for completed processes is now done
3764 earlier. On busy systems, this bug wouldn't be noticed because something
3765 else would have woken the daemon, and it would have reaped the completed
3768 80. If a message was submitted locally by a user whose login name contained one
3769 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3770 It caused a spool format error. I have fixed the spool reading code. A
3771 related problem was that the "from" clause in the Received: line became
3772 illegal because of the space(s). It is now covered by ${quote_local_part.
3774 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3777 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3779 83. Patch from Sieve maintainer to fix unterminated string problem in
3780 "vacation" handling.
3782 84. Some minor changes to the Linux configuration files to help with other
3783 OS variants using glibc.
3785 85. One more patch for Sieve to update vacation handling to latest spec.
3788 ----------------------------------------------------
3789 See the note above about the 4.44 and 4.50 releases.
3790 ----------------------------------------------------
3796 1. Change 4.43/35 introduced a bug that caused file counts to be
3797 incorrectly computed when quota_filecount was set in an appendfile
3800 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3801 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3804 3. Give more explanation in the error message when the command for a transport
3805 filter fails to execute.
3807 4. There are several places where Exim runs a non-Exim command in a
3808 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3809 was being done only for the command run by the queryprogram router. It is
3810 now done for all such subprocesses. The other cases are: ${run, transport
3811 filters, and the commands run by the lmtp and pipe transports.
3813 5. Some older OS have a limit of 256 on the maximum number of file
3814 descriptors. Exim was using setrlimit() to set 1000 as a large value
3815 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3816 systems. I've change it so that if it can't get 1000, it tries for 256.
3818 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3819 was an oversight, and furthermore, ever since the addition of extra
3820 controls (e.g. 4.43/32), the checks on when to allow different forms of
3821 "control" were broken. There should now be diagnostics for all cases when a
3822 control that does not make sense is encountered.
3824 7. $recipients is now available in the predata ACL (oversight).
3826 8. Tidy the search cache before the fork to do a delivery from a message
3827 received from the command line. Otherwise the child will trigger a lookup
3828 failure and thereby defer the delivery if it tries to use (for example) a
3829 cached ldap connection that the parent has called unbind on.
3831 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3832 of $address_data from the recipient verification was clobbered by the
3833 sender verification.
3835 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3836 was its contents. (It was OK if the option was not defined at all.)
3838 11. A "Completed" log line is now written for messages that are removed from
3839 the spool by the -Mrm option.
3841 12. $host_address is now set to the target address during the checking of
3842 ignore_target_hosts.
3844 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3845 being passed; this would have caused $sender_host_name to have been used if
3846 matching the list had actually called for a host name (not very likely,
3847 since this list is usually IP addresses). A host name is now passed as
3850 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3851 code that specifies a non-restarting handler (typically sigaction() in
3852 modern systems) in an attempt to fix a rare and obscure crash bug.
3854 15. Narrowed the window for a race in the daemon that could cause it to ignore
3855 SIGCHLD signals. This is not a major problem, because they are used only to
3856 wake it up if nothing else does.
3858 16. A malformed maildirsize file could cause Exim to calculate negative values
3859 for the mailbox size or file count. Odd effects could occur as a result.
3860 The maildirsize information is now recalculated if the size or filecount
3863 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3864 support for a long time. Removed HAVE_SYS_VFS_H.
3866 18. Updated exipick to current release from John Jetmore.
3868 19. Allow an empty sender to be matched against a lookup in an address list.
3869 Previously the only cases considered were a regular expression, or an
3872 20. Exim went into a mad DNS lookup loop when doing a callout where the
3873 host was specified on the transport, if the DNS lookup yielded more than
3876 21. The RFC2047 encoding function was originally intended for short strings
3877 such as real names; it was not keeping to the 75-character limit for
3878 encoded words that the RFC imposes. It now respects the limit, and
3879 generates multiple encoded words if necessary. To be on the safe side, I
3880 have increased the buffer size for the ${rfc2047: expansion operator from
3883 22. Failure to deliver a bounce message always caused it to be frozen, even if
3884 there was an errors_to setting on the router. The errors_to setting is now
3887 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3888 canonical form (fully expanded) before being placed in
3889 $sender_host_address.
3891 24. Updated eximstats to version 1.33
3893 25. Include certificate and key file names in error message when GnuTLS fails
3894 to set them up, because the GnuTLS error message doesn't include the name
3895 of the failing file when there is a problem reading it.
3897 26. Expand error message when OpenSSL has problems setting up cert/key files.
3900 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3901 (this can affect the format of dates).
3903 28. exim_tidydb, when checking for the continued existence of a message for
3904 which it has found a message-specific retry record, was not finding
3905 messages that were in split spool directories. Consequently, it was
3906 deleting retry records that should have stayed in existence.
3908 29. eximstats updated to version 1.35
3909 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3910 1.35 - bugfix such that pie charts by volume are generated correctly
3912 30. The SPA authentication driver was not abandoning authentication and moving
3913 on to the next authenticator when an expansion was forced to fail,
3914 contradicting the general specification for all authenticators. Instead it
3915 was generating a temporary error. It now behaves as specified.
3917 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3918 (the order specifies the preference for clients). The order is now AES256,
3919 AES128, 3DES, ARCFOUR128.
3921 31. Small patch to Sieve code - explicitly set From: when generating an
3924 32. Exim crashed if a remote delivery caused a very long error message to be
3925 recorded - for instance if somebody sent an entire SpamAssassin report back
3926 as a large number of 550 error lines. This bug was coincidentally fixed by
3927 increasing the size of one of Exim's internal buffers (big_buffer) that
3928 happened as part of the Exiscan merge. However, to be on the safe side, I
3929 have made the code more robust (and fixed the comments that describe what
3932 33. Some experimental protocols are using DNS PTR records for new purposes. The
3933 keys for these records are domain names, not reversed IP addresses. The
3934 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3935 leaves it alone. Component reversal etc. now happens only for IP addresses.
3938 34. The host_aton() function is supposed to be passed a string that is known
3939 to be a valid IP address. However, in the case of IPv6 addresses, it was
3940 not checking this. This is a hostage to fortune. Exim now panics and dies
3941 if the condition is not met. A case was found where this could be provoked
3942 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3943 components; fortuitously, this particular loophole had already been fixed
3944 by change 4.50/55 or 4.44/33 above.
3946 If there are any other similar loopholes, the new check in host_aton()
3947 itself should stop them being exploited. The report I received stated that
3948 data on the command line could provoke the exploit when Exim was running as
3949 exim, but did not say which command line option was involved. All I could
3950 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3951 running as the user.
3954 35. There was a buffer overflow vulnerability in the SPA authentication code
3955 (which came originally from the Samba project). I have added a test to the
3956 spa_base64_to_bits() function which I hope fixes it.
3959 36. The daemon start-up calls getloadavg() while still root for those OS that
3960 need the first call to be done as root, but it missed one case: when
3961 deliver_queue_load_max is set with deliver_drop_privilege. This is
3962 necessary for the benefit of the queue runner, because there is no re-exec
3963 when deliver_drop_privilege is set.
3965 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3966 was in use, was not putting the data itself into the right store pool;
3967 consequently, it could be overwritten for a subsequent message in the same
3968 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3971 38. Sometimes the final signoff response after QUIT could fail to get
3972 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3973 < 0 before doing a fflush(). This bug looks as though it goes back to the
3974 introduction of TLS in release 3.20, but "sometimes" must have been rare
3975 because the tests only now provoked it.
3981 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3982 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3983 Now it returns an integer. A number of calls were still expecting a T/F
3984 return. Fortuitously, in all cases, the tests worked in OK situations,
3985 which is the norm. However, things would have gone wrong on any write
3986 failures on the smtp file descriptor. This function is used when sending
3987 messages over SMTP and also when doing verify callouts.
3989 2. When Exim is called to do synchronous delivery of a locally submitted
3990 message (the -odf or -odi options), it no longer closes stderr before doing
3993 3. Implemented the mua_wrapper option.
3995 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3997 5. Implemented the functions header_remove(), header_testname(),
3998 header_add_at_position(), and receive_remove_recipient(), and exported them
4001 6. If an ACL "warn" statement specified the addition of headers, Exim already
4002 inserted X-ACL-Warn: at the start if there was no header name. However, it
4003 was not making this test for the second and subsequent header lines if
4004 there were newlines in the string. This meant that an invalid header could
4005 be inserted if Exim was badly configured.
4007 7. Allow an ACL "warn" statement to add header lines at the start or after all
4008 the Received: headers, as well as at the end.
4010 8. Added the rcpt_4xx retry error code.
4012 9. Added postmaster_mailfrom=xxx to callout verification option.
4014 10. Added mailfrom=xxxx to the callout verification option, for verify=
4017 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
4018 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
4020 12. Inserted some casts to stop certain compilers warning when using pointer
4021 differences as field lengths or precisions in printf-type calls (mostly
4022 affecting debugging statements).
4024 13. Added optional readline() support for -be (dynamically loaded).
4026 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
4027 same clock tick as a message's arrival, so that its received time was the
4028 same as the "first fail" time on the retry record, and that message
4029 remained on the queue past the ultimate address timeout, every queue runner
4030 would try a delivery (because it was past the ultimate address timeout) but
4031 after another failure, the ultimate address timeout, which should have then
4032 bounced the address, did not kick in. This was a "< instead of <=" error;
4033 in most cases the first failure would have been in the next clock tick
4034 after the received time, and all would be well.
4036 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
4037 being recognized when the domain list was tested by the match_domain
4038 condition in an expansion string.
4040 16. Added the ${str2b64: operator.
4042 17. Exim was always calling setrlimit() to set a large limit for the number of
4043 processes, without checking whether the existing limit was already
4044 adequate. (It did check for the limit on file descriptors.) Furthermore,
4045 errors from getrlimit() and setrlimit() were being ignored. Now they are
4046 logged to the main and panic logs, but Exim does carry on, to try to do its
4047 job under whatever limits there are.
4049 18. Imported PCRE 5.0.
4051 19. Trivial typo in log message " temporarily refused connection" (the leading
4054 20. If the log selector return_path_on_delivery was set and an address was
4055 redirected to /dev/null, the delivery process crashed because it assumed
4056 that a return path would always be set for a "successful" delivery. In this
4057 case, the whole delivery is bypassed as an optimization, and therefore no
4060 21. Internal re-arrangement: the function for sending a challenge and reading
4061 a response while authentication was assuming a zero-terminated challenge
4062 string. It's now changed to take a pointer and a length, to allow for
4063 binary data in such strings.
4065 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4067 23. Exim was not respecting finduser_retries when seeking the login of the
4068 uid under which it was called; it was always trying 10 times. (The default
4069 setting of finduser_retries is zero.) Also, it was sleeping after the final
4070 failure, which is pointless.
4072 24. Implemented tls_on_connect_ports.
4074 25. Implemented acl_smtp_predata.
4076 26. If the domain in control=submission is set empty, Exim assumes that the
4077 authenticated id is a complete email address when it generates From: or
4078 Sender: header lines.
4080 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4081 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4082 chown and chgrp in /bin and hostname in /usr/bin.
4084 28. Exim was keeping the "process log" file open after each use, just as it
4085 does for the main log. This opens the possibility of it remaining open for
4086 long periods when the USR1 signal hits a daemon. Occasional processlog
4087 errors were reported, that could have been caused by this. Anyway, it seems
4088 much more sensible not to leave this file open at all, so that is what now
4091 29. The long-running daemon process does not normally write to the log once it
4092 has entered its main loop, and it closes the log before doing so. This is
4093 so that log files can straightforwardly be renamed and moved. However,
4094 there are a couple of unusual error situations where the daemon does write
4095 log entries, and I had neglected to close the log afterwards.
4097 30. The text of an SMTP error response that was received during a remote
4098 delivery was being truncated at 512 bytes. This is too short for some of
4099 the long messages that one sometimes sees. I've increased the limit to
4102 31. It is now possible to make retry rules that apply only when a message has a
4103 specific sender, in particular, an empty sender.
4105 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4106 it possible to be selective about when SMTP synchronization is enforced.
4108 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4110 32. Implemented hosts_connection_nolog.
4112 33. Added an ACL for QUIT.
4114 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4117 35. Added mailbox_size and mailbox_filecount to appendfile.
4119 36. Added control = no_multiline_responses to ACLs.
4121 37. There was a bug in the logic of the code that waits for the clock to tick
4122 in the case where the clock went backwards by a substantial amount such
4123 that the microsecond fraction of "now" was more than the microsecond
4124 fraction of "then" (but the whole seconds number was less).
4126 38. Added support for the libradius Radius client library this is found on
4127 FreeBSD (previously only the radiusclient library was supported).
4133 1. When certain lookups returned multiple values in the form name=value, the
4134 quoting of the values was not always being done properly. Specifically:
4135 (a) If the value started with a double quote, but contained no whitespace,
4137 (b) If the value contained whitespace other than a space character (i.e.
4138 tabs or newlines or carriage returns) it was not quoted.
4139 This fix has been applied to the mysql and pgsql lookups by writing a
4140 separate quoting function and calling it from the lookup code. The fix
4141 should probably also be applied to nisplus, ibase and oracle lookups, but
4142 since I cannot test any of those, I have not disturbed their existing code.
4144 2. A hit in the callout cache for a specific address caused a log line with no
4145 reason for rejecting RCPT. Now it says "Previous (cached) callout
4146 verification failure".
4148 3. There was an off-by-one bug in the queryprogram router. An over-long
4149 return line was truncated at 256 instead of 255 characters, thereby
4150 overflowing its buffer with the terminating zero. As well as fixing this, I
4151 have increased the buffer size to 1024 (and made a note to document this).
4153 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4154 when Exim is waiting for an SMTP response from a remote server, Exim
4155 restarts its select() call on the socket, thereby resetting its timeout.
4156 This is not a problem when such interrupts are rare. Somebody set up a cron
4157 job to run exiwhat every 2 minutes, which is less than the normal select()
4158 timeout (5 or 10 minutes). This meant that the select() timeout never
4159 kicked in because it was always reset. I have fixed this by comparing the
4160 time when an interrupt arrives with the time at the start of the first call
4161 to select(). If more time than the timeout has elapsed, the interrupt is
4162 treated as a timeout.
4164 5. Some internal re-factoring in preparation for the addition of Sieve
4165 extensions (by MH). In particular, the "personal" test is moved to a
4166 separate function, and given an option for scanning Cc: and Bcc: (which is
4167 not set for Exim filters).
4169 6. When Exim created an email address using the login of the caller as the
4170 local part (e.g. when creating a From: or Sender: header line), it was not
4171 quoting the local part when it contained special characters such as @.
4173 7. Installed new OpenBSD configuration files.
4175 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4176 try to make them clearer.
4178 9. Callout options, other than the timeout value, were being ignored when
4179 verifying sender addresses in header lines. For example, when using
4181 verify = header_sender/callout=no_cache
4183 the cache was (incorrectly) being used.
4185 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4186 only the Cygwin environment.
4188 11. When return_path_on_delivery was set as a log selector, if different remote
4189 addresses in the same message used different return paths and parallel
4190 remote delivery occurred, the wrong values would sometimes be logged.
4191 (Whenever a remote delivery process finished, the return path value from
4192 the most recently started remote delivery process was logged.)
4194 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4195 lines when AUTH and/or TLS are in use. This is the "received protocol"
4196 field. Exim used to use "asmtp" for authenticated SMTP, without any
4197 indication (in the protocol name) for TLS use. Now it follows the RFC and
4198 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4199 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4200 names appear in log lines as well as in Received: header lines.
4202 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4203 extensions, and comparison tests, and to fix some bugs.
4205 14. Changes to the "personal" filter test:
4207 (1) The test was buggy in that it was just doing the equivalent of
4208 "contains" tests on header lines. For example, if a user's address was
4209 anne@some.where, the "personal" test would incorrectly be true for
4211 To: susanne@some.where
4213 This test is now done by extracting each address from the header in turn,
4214 and checking the entire address. Other tests that are part of "personal"
4215 are now done using regular expressions (for example, to check local parts
4216 of addresses in From: header lines).
4218 (2) The list of non-personal local parts in From: addresses has been
4219 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4220 taken from the Sieve specification recommendations.
4222 (3) If the message contains any header line starting with "List-" it is
4223 treated as non-personal.
4225 (4) The test for "circular" in the Subject: header line has been removed
4226 because it now seems ill-conceived.
4228 15. Minor typos in src/EDITME comments corrected.
4230 16. Installed latest exipick from John Jetmore.
4232 17. If headers_add on a router specified a text string that was too long for
4233 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4234 of string_sprintf() is now avoided.
4236 18. $message_body_size was not set (it was always zero) when running the DATA
4237 ACL and the local_scan() function.
4239 19. For the "mail" command in an Exim filter, no default was being set for
4240 the once_repeat time, causing a random time value to be used if "once" was
4241 specified. (If the value happened to be <= 0, no repeat happened.) The
4242 default is now 0s, meaning "never repeat". The "vacation" command was OK
4243 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4244 (I found it when inspecting the code).
4246 20. There is now an overall timeout for performing a callout verification. It
4247 defaults to 4 times the callout timeout, which applies to individual SMTP
4248 commands during the callout. The overall timeout applies when there is more
4249 than one host that can be tried. The timeout is checked before trying the
4250 next host. This prevents very long delays if there are a large number of
4251 hosts and all are timing out (e.g. when the network connections are timing
4252 out). The value of the overall timeout can be changed by specifying an
4253 additional sub-option for "callout", called "maxwait". For example:
4255 verify = sender/callout=5s,maxwait=20s
4257 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4258 to the end before writing, but this should make it even safer).
4260 22. Exim was forgetting that it had advertised PIPELINING for the second and
4261 subsequent messages on an SMTP connection. It was also not resetting its
4262 memory on STARTTLS and an internal HELO.
4264 23. When Exim logs an SMTP synchronization error within a session, it now
4265 records whether PIPELINING has been advertised or not.
4267 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4268 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4269 rather than long int.
4271 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4273 26. Added the never_mail option to autoreply.
4279 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4280 crash if the getsockname() call failed; this can happen if a connection is
4281 closed very soon after it is established. The problem was simply in the
4282 order in which certain operations were done, causing Exim to try to write
4283 to the SMTP stream before it had set up the file descriptor. The bug has
4284 been fixed by making things happen in the correct order.
4290 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4291 before the connection was closed, thus losing the rejection response.
4293 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4294 some early Solaris releases, but causes trouble in current releases where
4295 socklen_t is defined.
4297 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4300 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4303 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4304 that releasing the top of it at the end releases what was used for sub-
4305 expansions (unless the block got too big). However, discard this block if
4306 the first thing is a variable or header, so that we can use its block when
4307 it is dynamic (useful for very large $message_headers, for example).
4309 6. Lookups now cache *every* query, not just the most recent. A new, separate
4310 store pool is used for this. It can be recovered when all lookup caches are
4311 flushed. Lookups now release memory at the end of their result strings.
4312 This has involved some general refactoring of the lookup sources.
4314 7. Some code has been added to the store_xxx() functions to reduce the amount
4315 of flapping under certain conditions.
4317 8. log_incoming_interface used to affect only the <= reception log lines. Now
4318 it causes the local interface and port to be added to several more SMTP log
4319 lines, for example "SMTP connection from", and rejection lines.
4321 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4323 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4325 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4326 could be overwritten at the end of the current message (or the start of a
4327 new message if it was set in a HELO ACL). The value is now preserved for
4328 the duration of the SMTP connection.
4330 12. If a transport had a headers_rewrite setting, and a matching header line
4331 contained an unqualified address, that address was qualified, even if it
4332 did not match any rewriting rules. The underlying bug was that the values
4333 of the flags that permit the existence of unqualified sender and recipient
4334 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4335 non-local messages, and by -bnq for local messages) were not being
4336 preserved with the message after it was received.
4338 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4339 "next input=" as part of the text comprising the host identity instead of
4340 the correct text. The code was using the same buffer for two different
4341 strings. However, depending on which order the printing function evaluated
4342 its arguments, the bug did not always show up. Under Linux, for example, my
4343 test suite worked just fine.
4345 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4346 doesn't work with some older versions of Perl. It has been changed to "my",
4347 which in any case is probably the better facility to use.
4349 15. A really picky compiler found some instances of statements for creating
4350 error messages that either had too many or two few arguments for the format
4353 16. The size of the buffer for calls to the DNS resolver has been increased
4354 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4355 for addresses that have a lot of PTR records. This alleviates a problem; it
4356 does not fully solve it.
4358 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4359 buffer now truncates the list and logs the incident, which is the same
4360 action as happens when Exim is looking up a host name and its aliases.
4361 Previously in this situation something unpredictable would happen;
4362 sometimes it was "internal error: store_reset failed".
4364 18. If a server dropped the connection unexpectedly when an Exim client was
4365 using GnuTLS and trying to read a response, the client delivery process
4366 crashed while trying to generate an error log message.
4368 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4369 string, for example:
4371 warn message = H1: something\nH2: something
4373 the text was added as a single header line from Exim's point of view
4374 though it ended up OK in the delivered message. However, searching for the
4375 second and subsequent header lines using $h_h2: did not work. This has been
4376 fixed. Similarly, if a system filter added multiple headers in this way,
4377 the routers could not see them.
4379 20. Expanded the error message when iplsearch is called with an invalid key to
4380 suggest using net-iplsearch in a host list.
4382 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4383 ACLs are no longer actually imposed (and a message to that effect is
4386 22. If a "gecos" field in a passwd entry contained escaped characters, in
4387 particular, if it contained a \" sequence, Exim got it wrong when building
4388 a From: or a Sender: header from that name. A second bug also caused
4389 incorrect handling when an unquoted " was present following a character
4390 that needed quoting.
4392 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4393 was not being matched caselessly.
4395 24. Arranged for all hyphens in the exim.8 source to be escaped with
4398 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4399 properly. Recipient callout cache records were still being keyed to include
4400 the sender, even when use_sender was set false. This led to far more
4401 callouts that were necessary. The sender is no longer included in the key
4402 when use_sender is false.
4404 26. Added "control = submission" modifier to ACLs.
4406 27. Added the ${base62d: operator to decode base 62 numbers.
4408 28. dnsdb lookups can now access SRV records.
4410 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4411 the configuration file.
4413 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4414 (-v) mode. This makes the output for a verbose queue run more intelligible.
4416 31. Added a use_postmaster feature to recipient callouts.
4418 32. Added the $body_zerocount variable, containing the number of binary zero
4419 bytes in the message body.
4421 33. The time of last modification of the "new" subdirectory is now used as the
4422 "mailbox time last read" when there is a quota error for a maildir
4425 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4427 35. Added +ignore_unknown as a special item in host lists.
4429 36. Code for decoding IPv6 addresses in host lists is now included, even if
4430 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4431 address was recognized as an IP address, but was then not correctly decoded
4432 into binary, causing unexpected and incorrect effects when compared with
4439 1. Very minor rewording of debugging text in manualroute to say "list of
4440 hosts" instead of "hostlist".
4442 2. If verify=header_syntax was set, and a header line with an unqualified
4443 address (no domain) and a large number of spaces between the end of the
4444 name and the colon was received, the reception process suffered a buffer
4445 overflow, and (when I tested it) crashed. This was caused by some obsolete
4446 code that should have been removed. The fix is to remove it!
4448 3. When running in the test harness, delay a bit after writing a bounce
4449 message to get a bit more predictability in the log output.
4451 4. Added a call to search_tidyup() just before forking a reception process. In
4452 theory, someone could use a lookup in the expansion of smtp_accept_max_
4453 per_host which, without the tidyup, could leave open a database connection.
4455 5. Added the variables $recipient_data and $sender_data which get set from a
4456 lookup success in an ACL "recipients" or "senders" condition, or a router
4457 "senders" option, similar to $domain_data and $local_part_data.
4459 6. Moved the writing of debug_print from before to after the "senders" test
4462 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4463 problems for message scanning, either using a data ACL, or using
4464 local_scan() because the Received: header was not generated till after they
4465 were called (in order to set the time as the time of reception completion).
4466 I have revised the way this works. The header is now generated after the
4467 body is received, but before the ACL or local_scan() are called. After they
4468 are run, the timestamp in the header is updated.
4474 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4475 before starting a queue runner without re-exec. This happened only when
4476 deliver_drop_privilege was set or when the Exim user was set to root. The
4477 effect of the bug was that timeouts during subsequent deliveries caused
4478 crashes instead of being properly handled. The handler is now left at its
4479 default (and expected) setting.
4481 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4482 message, again when deliver_drop_privilege is set or Exim is run as root.
4483 The bug described in (1) was not present in this case, but the tidying up
4484 of the other signals was missing. I have made the two cases consistent.
4486 3. The ignore_target_hosts setting on a manualroute router was being ignored
4487 for hosts that were looked up using the /MX notation.
4489 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4492 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4493 operated on the sender address. After changing the $sender_address to <>
4494 for the sender address verify, Exim was re-instated it as the original
4495 (before rewriting) address, but remembering that it had rewritten it, so it
4496 wasn't rewriting it again. This bug also had the effect of breaking the
4497 sender address verification caching when the sender address was rewritten.
4499 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4500 This has been changed so that if the ip literal address matches
4501 ignore_target_hosts, the router declines.
4503 7. Added expansion conditions match_domain, match_address, and match_local_
4504 part (NOT match_host).
4506 8. The placeholder for the Received: header didn't have a length field set.
4508 9. Added code to Exim itself and to exim_lock to test for a specific race
4509 condition that could lead to file corruption when using MBX delivery. The
4510 issue is with the lockfile that is created in /tmp. If this file is removed
4511 after a process has opened it but before that process has acquired a lock,
4512 there is the potential for a second process to recreate the file and also
4513 acquire a lock. This could lead to two Exim processes writing to the file
4514 at the same time. The added code performs the same test as UW imapd; it
4515 checks after acquiring the lock that its file descriptor still refers to
4516 the same named file.
4518 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4519 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4522 11. Added the smtp_active_hostname option. If used, this will typically be made
4523 to depend on the incoming interface address. Because $interface_address is
4524 not set up until the daemon has forked a reception process, error responses
4525 that can happen earlier (such as "too many connections") no longer contain
4528 12. If an expansion in a condition on a "warn" statement fails because a lookup
4529 defers, the "warn" statement is abandoned, and the next ACL statement is
4530 processed. Previously this caused the whole ACL to be aborted.
4532 13. Added the iplsearch lookup type.
4534 14. Added ident_timeout as a log selector.
4536 15. Added tls_certificate_verified as a log selector.
4538 16. Added a global option tls_require_ciphers (compare the smtp transport
4539 option of the same name). This controls incoming TLS connections.
4541 17. I finally figured out how to make tls_require_ciphers do a similar thing
4542 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4543 before starting the TLS session.
4545 18. Tabs are now shown as \t in -bP output.
4547 19. If the log selector return_path_on_delivery was set, Exim crashed when
4548 bouncing a message because it had too many Received: header lines.
4550 20. If two routers both had headers_remove settings, and the first one included
4551 a superfluous trailing colon, the final name in the first list and the
4552 first name in the second list were incorrectly joined into one item (with a
4553 colon in the middle).
4559 1. Added -C and -D options to the exinext utility, mainly to make it easier
4560 to include in the automated testing, but these could be helpful when
4561 multiple configurations are in use.
4563 2. The exinext utility was not formatting the output nicely when there was
4564 an alternate port involved in the retry record key, nor when there was a
4565 message id as well (for retries that were specific to a specific message
4566 and a specific host). It was also confused by IPv6 addresses, because of
4567 the additional colons they contain. I have fixed the IPv4 problem, and
4568 patched it up to do a reasonable job for IPv6.
4570 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4571 delivery, the log line now contains "pipelined" if PIPELINING was used.
4573 4. An SMTP transport process used to panic and die if the bind() call to set
4574 an explicit outgoing interface failed. This has been changed; it is now
4575 treated in the same way as a connect() failure.
4577 5. A reference to $sender_host_name in the part of a conditional expansion
4578 that was being skipped was still causing a DNS lookup. This no longer
4581 6. The def: expansion condition was not recognizing references to header lines
4582 that used bh_ and bheader_.
4584 7. Added the _cache feature to named lists.
4586 8. The code for checking quota_filecount in the appendfile transport was
4587 allowing one more file than it should have been.
4589 9. For compatibility with Sendmail, the command line option
4597 and sets the incoming protocol and host name (for trusted callers). The
4598 host name and its colon can be omitted when only the protocol is to be set.
4599 Note the Exim already has two private options, -pd and -ps, that refer to
4600 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4601 "s", but I don't think that's a major issue.
4603 10. A number of refactoring changes to the code, none of which should affect
4606 (a) The number of logging options was getting close to filling up the
4607 32-bit word that was used as a bit map. I have split them into two classes:
4608 those that are passed in the argument to log_write(), and those that are
4609 only ever tested independently outside of that function. These are now in
4610 separate 32-bit words, so there is plenty of room for expansion again.
4611 There is no change in the user interface or the logging behaviour.
4613 (b) When building, for example, log lines, the code previously used a
4614 macro that called string_cat() twice, in order to add two strings. This is
4615 not really sufficiently general. Furthermore, there was one instance where
4616 it was actually wrong because one of the argument was used twice, and in
4617 one call a function was used. (As it happened, calling the function twice
4618 did not affect the overall behaviour.) The macro has been replaced by a
4619 function that can join an arbitrary number of extra strings onto a growing
4622 (c) The code for expansion conditions now uses a table and a binary chop
4623 instead of a serial search (which was left over from when there were very
4624 few conditions). Also, it now recognizes conditions like "pam" even when
4625 the relevant support is not compiled in: a suitably worded error message is
4626 given if an attempt is made to use such a condition.
4628 11. Added ${time_interval:xxxxx}.
4630 12. A bug was causing one of the ddress fields not to be passed back correctly
4631 from remote delivery subprocesses. The field in question was not being
4632 subsequently used, so this caused to problems in practice.
4634 13. Added new log selectors queue_time and deliver_time.
4636 14. Might have fixed a bug in maildirsizefile handling that threw up
4637 "unexpected character" debug warnings, and recalculated the data
4638 unnecessarily. In any case, I expanded the warning message to give more
4641 15. Added the message "Restricted characters in address" to the statements in
4642 the default ACL that block characters like @ and % in local parts.
4644 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4645 Three changes have been made:
4647 (a) There was a serious bug; a negative response to MAIL caused the whole
4648 recipient domain to be cached as invalid, thereby blocking all messages
4649 to all local parts at the same domain, from all senders. This bug has
4650 been fixed. The domain is no longer cached after a negative response to
4651 MAIL if the sender used is not empty.
4653 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4656 (c) A new callout option, "use_sender" has been added for people who want
4657 the modified behaviour.
4663 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4666 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4667 indeed breaks things for older releases.
4669 3. Added additional logging to the case where there is a problem reading data
4670 from a filter that is running in a subprocess using a pipe, in order to
4671 try to track down a specific problem.
4673 4. Testing facility fudge: when running in the test harness and attempting
4674 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4675 getting "No route to host". Convert this to a timeout.
4677 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4680 6. Some OS don't have socklen_t but use size_t instead. This affects the
4681 fifth argument of getsockopt() amongst other things. This is now
4682 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4683 can be set for individual OS. I have set it for SunOS5, OSF1, and
4684 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4685 some earlier ones do not.
4687 7. Change 4.30/15 was not doing the test caselessly.
4689 8. The standard form for an IPv6 address literal was being rejected by address
4690 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4691 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4692 this, as well as the form without the "IPv6" on the front (but only when
4693 address literals are enabled, of course).
4695 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4697 10. Exim crashed if a message with an empty sender address specified by -f
4698 encountered a router with an errors_to setting. This could be provoked only
4699 by a command such as
4703 where an empty string was supplied; "<>" did not hit this bug.
4705 11. Installed PCRE release 4.5.
4707 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4708 remained set. It is now erased.
4710 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4711 times from message ids (which are base 36 rather than the normal 62).
4713 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4714 were being counted as actual protocol errors, and logged if the log
4715 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4716 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4717 and DATA following a set of rejected RCPTs do not count as protocol errors.
4718 In other words, Exim assumes they were pipelined, though this may not
4719 actually be the case. Of course, in all cases the client gets an
4720 appropriate error code.
4722 15. If a lookup fails in an ACL condition, a message about the failure may
4723 be available; it is used if testing the ACL cannot continue, because most
4724 such messages specify what the cause of the deferral is. However, some
4725 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4726 that caused an old message to be retained and used if a later statement
4727 caused a defer, replacing the real cause of the deferral.
4729 16. If an IP address had so many PTR records that the DNS lookup buffer
4730 was not large enough to hold them, Exim could crash while trying to process
4731 the truncated data. It now detects and logs this case.
4733 17. Further to 4.21/58, another change has been made: if (and only if) the
4734 first line of a message (the first header line) ends with CRLF, a bare LF
4735 in a subsequent header line has a space inserted after it, so as not to
4736 terminate the header.
4738 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4739 unnecessarily, used atoi() instead of strtol(), and didn't check the
4740 termination when getting file sizes from file names by regex.
4742 19. Completely re-implemented the support for maildirsize files, in the light
4743 of a number of problems with the previous contributed implementation
4744 (4.30/29). In particular:
4746 . If the quota is zero, the maildirsize file is maintained, but no quota is
4749 . If the maildir directory does not exist, it is created before any attempt
4750 to write a maildirsize file.
4752 . The quota value in the file is just a cache; if the quota is changed in
4753 the transport, the new value overrides.
4755 . A regular expression is available for excluding directories from the
4758 20. The autoreply transport checks the characters in options that define the
4759 message's headers; it allows continued headers, but it was checking with
4760 isspace() after an embedded newline instead of explicitly looking for a
4763 21. If all the "regular" hosts to which an address was routed had passed their
4764 expiry times, and had not reached their retry times, the address was
4765 bounced, even if fallback hosts were defined. Now Exim should go on to try
4768 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4769 equivalent code in the SMTP transport. Some hosts send humungous responses
4770 to HELO/EHLO, more than 1024 it seems.
4772 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4773 gives compiler warnings in some environments. I've now done it "properly",
4776 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4777 (because of problems with the built-in one) was declared to return uschar *
4778 instead of char *, causing compiler failure.
4780 25. Fixed a file descriptor leak when processing alias/forward files.
4782 26. Fixed a minor format string issue in dbfn.c.
4784 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4786 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4787 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4790 29. When a sender address is verified, it is cached, to save repeating the test
4791 when there is more than one recipient in a message. However, when the
4792 verification involves a callout, it is possible for different callout
4793 options to be set for different recipients. It is too complicated to keep
4794 track of this in the cache, so now Exim always runs a verification when a
4795 callout is required, relying on the callout cache for the optimization.
4796 The overhead is duplication of the address routing, but this should not be
4799 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4800 to be verified with callout=postmaster, and the main callout worked but the
4801 postmaster check failed, the verification correctly failed. However, if a
4802 subsequent RCPT command asked for sender verification *without* the
4803 postmaster check, incorrect caching caused this verification also to fail,
4806 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4807 it was not caching the DNS options (qualify_single, search_parents) that
4808 were used when the lookup failed. A subsequent lookup with different
4809 options therefore always gave the same answer, though there were cases
4810 where it should not have. (Example: a "domains = !$mx_any" option on a
4811 dnslookup router: the "domains" option is always processed without any
4812 widening, but the router might have qualify_single set.) Now Exim uses the
4813 cached value only when the same options are set.
4815 32. Added John Jetmore's "exipick" utility to the distribution.
4817 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4818 than a timeout (e.g. a certificate is required, and is not provided), an
4819 Exim server now closes the connection immediately. Previously it waited for
4820 the client to close - but if the client is SSL, it seems that they each
4821 wait for each other, leading to a delay before one of them times out.
4823 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4824 maintained 0.8.x compatibility because I don't think many are using it, and
4825 it is clearly obsolete.
4827 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4830 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4831 client certificate was expired. A simple patch fixes this, though I don't
4832 understand the full logic of why the verify callback is called multiple
4835 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4836 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4837 which causes problems with some clients (such as the Certicom SSL Plus
4838 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4839 disables the coutermeasure allowing Eudora to connect."
4841 38. Exim was not checking that a write() to a log file succeeded. This could
4842 lead to Bad Things if a log got too big, in particular if it hit a file
4843 size limit. Exim now panics and dies if it cannot write to a log file, just
4844 as it does if it cannot open a log file.
4846 39. Modified OS/Makefile-Linux so that it now contains
4848 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4850 The two -D definitions ensure that Exim is compiled with large file
4851 support, which makes it possible to handle log files that are bigger than
4854 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4855 instance) a domain was checked against a named list that involved a lookup,
4856 causing $domain_data to be set, then another domain was checked against the
4857 same list, then the first domain was re-checked, the value of $domain_data
4858 after the final check could be wrong. In particular, if the second check
4859 failed, it could be set empty. This bug probably also applied to
4862 41. The strip_trailing_dot option was not being applied to the address given
4863 with the -f command-line option.
4865 42. The code for reading a message's header from the spool was incrementing
4866 $received_count, but never initializing it. This meant that the value was
4867 incorrect (doubled) while delivering a message in the same process in which
4868 it was received. In the most common configuration of Exim, this never
4869 happens - a fresh exec is done - but it can happen when
4870 deliver_drop_privilege is set.
4872 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4873 it now includes up to 150 characters of the unexpected data in the log
4876 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4877 and building data strings. The size of both of these buffers was 10 000
4878 bytes - far larger than anybody would *ever* want, thought I. Needless to
4879 say, somebody hit the limit. I have increased the maximum line length to
4880 20 000 and the maximum data length of concatenated lines to 100 000. I have
4881 also fixed two bugs, because there was no checking on these buffers. Tsk,
4882 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4883 buffer is too small.
4885 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4886 lsearch lookups. Now it does.
4888 46. When parsing a route_list item in a manualroute router, a fixed-length
4889 buffer was used for the list of hosts. I made this 1024 bytes long,
4890 thinking that nobody would ever have a list of hosts that long. Wrong.
4891 Somebody had a whole pile of complicated expansion conditions, and the
4892 string was silently truncated, leading to an expansion error. It turns out
4893 that it is easier to change to an unlimited length (owing to other changes
4894 that have happened since this code was originally written) than to build
4895 structure for giving a limitation error. The length of the item that
4896 expands into the list of hosts is now unlimited.
4898 47. The lsearch lookup could not handle data where the length of text line was
4899 more than 4095 characters. Such lines were truncated, leading to shortened
4900 data being returned. It should now handle lines of any length.
4902 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4903 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4906 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4907 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4908 no longer happens in comment lines. A list of replacements is now placed
4909 at the head of all of the source files, except those whose only change is
4910 to replace PERL_COMMAND in the very first #! line.
4912 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4913 messages on the queue, with a bottom-up merge sort, using code contributed
4914 by Michael Haardt. This should make operations like -bp somewhat faster on
4915 large queues. It won't affect queue runners, except when queue_run_in_order
4918 51. Installed eximstats 1.31 in the distribution.
4920 52. Added support for SRV lookups to the dnslookup router.
4922 53. If an ACL referred to $message_body or $message_body_end, the value was not
4923 reset for any messages that followed in the same SMTP session.
4925 54. The store-handling optimization for building very long strings was not
4926 differentiating between the different store pools. I don't think this
4927 actually made any difference in practice, but I've tidied it.
4929 55. While running the routers to verify a sender address, $sender_address
4930 was still set to the sender address. This is wrong, because when routing to
4931 send a bounce to the sender, it would be empty. Therefore, I have changed
4932 it so that, while verifying a sender address, $sender_address is set to <>.
4933 (There is no change to what happens when verifying a recipient address.)
4935 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4936 target A or AAAA records (if not already returned) without resetting the
4937 qualify_single or search_parents options of the DNS resolver. These are
4938 inappropriate in this case because the targets of MX and SRV records must
4939 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4940 target that, when qualified, matched something in the local domain. These
4941 two options are now turned off when doing these lookups.
4943 57. It seems that at least some releases of Reiserfs (which does not have the
4944 concept of a fixed number of inodes) returns zero and not -1 for the
4945 number of available inodes. This interacted badly with check_spool_inodes,
4946 which assumed that -1 was the "no such thing" setting. What I have done is
4947 to check that the total number of inodes is greater than zero before doing
4948 the test of how many are available.
4950 58. When a "warn" ACL statement has a log_message modifier, the message is
4951 remembered, and not repeated. This is to avoid a lot of repetition when a
4952 message has many recipients that cause the same warning to be written.
4953 Howewer, Exim was preserving the list of already written lines for an
4954 entire SMTP session, which doesn't seem right. The memory is now reset if a
4955 new message is started.
4957 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4958 debugging output unless log_rewrite was also set.
4960 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4961 of (int)(handle) when we know that handle contains (void *)(-1).
4963 61. The Exim daemon panic-logs an error return when it closes the incoming
4964 connection. However "connection reset by peer" seems to be common, and
4965 isn't really an error worthy of noting specially, so that particular error
4968 62. When Exim is trying to find all the local interfaces, it used to panic and
4969 die if the ioctl to get the interface flags failed. However, it seems that
4970 on at least one OS (Solaris 9) it is possible to have an interface that is
4971 included in the list of interfaces, but for which you get a failure error
4972 for this call. This happens when the interface is not "plumbed" into a
4973 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4974 failure of the "get flags" call assumes that the interface is down.
4976 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4977 makes life easier for people who are doing arithmetic on fields extracted
4978 from dates, where you often get leading zeros that should not be
4979 interpreted as octal.
4981 64. Added qualify_domain to the redirect router, to override the global
4984 65. If a pathologically long header line contained very many addresses (the
4985 report of this problem mentioned 10 000) and each of them was rewritten,
4986 Exim could use up a very large amount of memory. (It kept on making new
4987 copies of the header line as it rewrote, and never released the old ones.)
4988 At the expense of a bit more processing, the header rewriting function has
4989 been changed so that it no longer eats memory in this way.
4991 66. The generation of the Received: header has been moved from the time that a
4992 message starts to be received, to the time that it finishes. The timestamp
4993 in the Received: header should now be very close to that of the <= log
4994 line. There are two side-effects of this change:
4996 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4997 logged header lines no longer include the local Received: line, because
4998 it has not yet been created. The same applies to a copy of the message
4999 that is returned to a non-SMTP sender when a message is rejected.
5001 (b) When a filter file is tested using -bf, no additional Received: header
5002 is added to the test message. After some thought, I decided that this
5005 This change does not affect the value of $received_for. It is still set
5006 after address rewriting, but before local_scan() is called.
5008 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
5010 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
5011 gave an unhelpful panic error message, and a defer error. I have managed to
5012 change this behaviour so that it now rejects any supplied certificate,
5013 which seems right, as the list of acceptable certificates is empty.
5015 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
5016 gave an unhelpful defer error. I have not managed to make this reject any
5017 supplied certificates, but the error message it gives is "no certificate
5018 supplied", which is not helpful.
5020 70. exigrep's output now also includes lines that are not associated with any
5021 message, but which match the given pattern. Implemented by a patch from
5022 Martin Sluka, which also tidied up the Perl a bit.
5024 71. Recipient callout verification, like sender verification, was using <> in
5025 the MAIL FROM command. This isn't really the right thing, since the actual
5026 sender may affect whether the remote host accepts the recipient or not. I
5027 have changed it to use the actual sender in the callout; this means that
5028 the cache record is now keyed on a recipient/sender pair, not just the
5029 recipient address. There doesn't seem to be a real danger of callout loops,
5030 since a callout by the remote host to check the sender would use <>.
5031 [SEE ABOVE: changed after hitting problems.]
5033 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
5034 temporary errors. However, in the case of such a code being given after
5035 the end of a data transmission (i.e. after ".") Exim was failing to write
5036 a retry record for the message. (Yes, there was some broken host that was
5037 actually sending 8xx at this point.)
5039 73. An unknown lookup type in a host list could cause Exim to panic-die when
5040 the list was checked. (An example that provoked this was putting <; in the
5041 middle of a list instead of at the start.) If this happened during a DATA
5042 ACL check, a -D file could be left lying around. This kind of configuration
5043 error no longer causes Exim to die; instead it causes a defer errror. The
5044 incident is still logged to the main and panic logs.
5046 74. Buglet left over from Exim 3 conversion. The message "too many messages
5047 in one connection" was written to the rejectlog but not the mainlog, except
5048 when address rewriting (yes!) was being logged.
5050 75. Added write_rejectlog option.
5052 76. When a system filter was run not as root (that is, when system_filter_user
5053 was set), the values of the $n variables were not being returned to the
5054 main process; thus, they were not subsequently available in the $sn
5057 77. Added +return_path_on_delivery log selector.
5059 78. A connection timeout was being treated differently from recipients deferred
5060 when testing hosts_max_try with a message that was older than the host's
5061 retry timeout. (The host should not be counted, thus allowing all hosts to
5062 be tried at least once before bouncing.) This may have been the cause of an
5063 occasionally reported bug whereby a message would remain on the queue
5064 longer than the retry timeout, but would be bounced if a delivery was
5065 forced. I say "may" because I never totally pinned down the problem;
5066 setting up timeout/retry tests is difficult. See also the next item.
5068 79. The ultimate address timeout was not being applied to errors that involved
5069 a combination of host plus message (for example, a timeout on a MAIL
5070 command). When an address resolved to a number of possible hosts, and they
5071 were not all tried for each delivery (e.g. because of hosts_max_try), a
5072 message could remain on the queue longer than the retry timeout.
5074 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5077 81. Fixed an obscure SMTP outgoing bug which required at least the following
5078 conditions: (a) there was another message waiting for the same server;
5079 (b) the server returned 5xx to all RCPT commands in the first message so
5080 that the message was not completed; (c) the server dropped the connection
5081 or gave a negative response to the RSET that Exim sends to abort the
5082 transaction. The observed case was a dropped connection after DATA that had
5083 been sent in pipelining mode. That is, the server had advertised PIPELINING
5084 but was not implementing it correctly. The effect of the bug was incorrect
5085 behaviour, such as trying another host, and this could lead to a crash.
5091 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5092 and daemon.c were passed as pointers to ints; they should have been
5093 pointers to socklen_t variables (which are typically unsigned ints).
5095 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5098 3. Fixed a really odd bug that affected only the testing scheme; patching a
5099 certain fixed string in the binary changed the value of another string that
5100 happened to be identical to the end of the original first string.
5102 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5103 name", it returns that address as the IP address. On some operating
5104 systems (e.g. Solaris), it also passes back the IP address string as the
5105 "host name". However, on others (e.g. Linux), it passes back an empty
5106 string. Exim wasn't checking for this, and was changing the host name to an
5107 empty string, assuming it had been canonicized.
5109 5. Although rare, it is permitted to have more than one PTR record for a given
5110 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5111 all the names associated with an address, because they do in Solaris.
5112 However, it seems that they do not in Linux for data that comes from the
5113 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5114 I found this out when I moved to a new Linux workstation and tried to run
5115 the Exim test suite.
5117 To get round this problem I have changed the code so that it now does its
5118 own call to the DNS to look up PTR records when searching for a host name.
5119 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5120 addresses that are only in /etc/hosts are still found.
5122 This behaviour is, however, controlled by an option called host_lookup_
5123 order, which defaults to "bydns:byaddr". If people want to use the other
5124 order, or indeed, just use one or the other means of lookup, they can
5125 specify it in this variable.
5127 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5128 some operating systems, this comes back from gethostbyaddr() as an empty
5129 string, and this is what Exim used to test for. However, it seems that in
5130 other systems, "." is yielded. Exim now tests for this case too.
5132 7. The values of check_spool_space and check_log_space are now held internally
5133 as a number of kilobytes instead of an absolute number of bytes. If a
5134 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5135 kilobyte. This means that much larger values can be stored.
5137 8. Exim monitor: an attempt to get the action menu when not actually pointing
5138 at a message produces an empty menu entitled "No message selected". This
5139 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5140 no entries in it ("Shell widget menu has zero width and/or height"). So I
5141 have added a single, blank menu entry in this case.
5143 9. Added ${quote_local_part.
5145 10. MIME decoding is now applied to the contents of Subject: header lines when
5148 11. Now that a reference to $sender_host_address automatically causes a reverse
5149 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5150 host lookup before query-style lookups in lists that might use this
5151 variable. This has therefore been abolished, and the "net-" prefix is no
5152 longer necessary for query-style lookups.
5154 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5155 have been a typo for LFLAGS, so it has been changed.
5157 13. The install script calls Exim with "-C /dev/null" in order to find the
5158 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5159 to be output. Howeve, since Exim outputs its version number before the
5160 error, it didn't break the script. It just looked ugly. I fixed this by
5161 always allowing "-C /dev/null" if the caller is root.
5163 14. Ignore overlarge ACL variable number when reading spool file - insurance
5164 against a later release with more variables having written the file.
5166 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5167 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5168 this, as well as the form without the "IPv6" on the front.
5170 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5171 OS/Makefile-Darwin file.
5173 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5174 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5176 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5177 variable. However, some OS use ints for inodes. I've added cast to long int
5178 to get rid of the compiler warning.
5180 19. I had forgotten to lock out "/../" in configuration file names when
5181 ALT_CONFIG_PREFIX was set.
5183 20. Routers used for verification do not need to specify transports. However,
5184 if such a router generated a host list, and callout was configured, Exim
5185 crashed, because it could not find a port number from the (non-existent)
5186 transport. It now assumes port 25 in this circumstance.
5188 21. Added the -t option to exigrep.
5190 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5191 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5192 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5193 them is set, LOOKUP_LSEARCH is forced.
5195 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5197 24. Added sender and host information to the "rejected by local_scan()" log
5198 line; previously there was no indication of these.
5200 25. Added .include_if_exists.
5202 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5203 Linux. It turns out that not all file systems support this. Apparently some
5204 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5205 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5206 not supported on the file descriptor, is now ignored when Exim is trying to
5207 sync a directory. This applies only to Linux.
5209 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5211 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5212 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5215 29. Added support for maildirsize files from supplied patch (modified a bit).
5217 30. The use of :fail: followed by an empty string could lead Exim to respond to
5218 sender verification failures with (e.g.):
5220 550 Verification failed for <xxx>
5221 550 Sender verify failed
5223 where the first response line was missing the '-' that indicates it is not
5224 the final line of the response.
5226 31. The loop for finding the name of the user that called Exim had a hardwired
5227 limit of 10; it now uses the value of finduser_retries, which is used for
5228 all other user lookups.
5230 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5233 33. Exim was neglecting to zero errno before one call of strtol() when
5234 expanding a string and expecting an integer value. On some systems this
5235 resulted in spurious "integer overflow" errors. Also, it was casting the
5236 result into an int without checking.
5238 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5239 did not work. The code looks as if it has *never* worked, though it appears
5240 to have been documented since at least releast 1.62. I have made it work.
5242 35. The "timeout_DNS" error in retry rules, also documented since at least
5243 1.62, also never worked. As it isn't clear exactly what this means, and
5244 clearly it isn't a major issue, I have abolished the feature by treating it
5245 as "timeout", and writing a warning to the main and panic logs.
5247 36. The display of retry rules for -brt wasn't always showing the error code
5250 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5251 timeout_connect_A, timeout_connect_MX.
5253 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5254 to the empty sender.
5256 39. The daemon was not analysing the content of -oX till after it had closed
5257 stderr and disconnected from the controlling terminal. This meant that any
5258 syntax errors were only noted on the panic log, and the return code from
5259 the command was 0. By re-arranging the code a little, I've made the
5260 decoding happen first, so such errors now appear on stderr, and the return
5261 code is 1. However, the actual setting up of the sockets still happens in
5262 the disconnected process, so errors there are still only recorded on the
5265 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5266 connections (as happens on some IP stacks) was logged at start up time as
5267 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5268 it from "IPv6 and IPv4", which means that two separate sockets are being
5271 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5272 says whether AF_INET or AF_INET6 was passed as an argument.
5274 42. Exiwhat output was messed up when time zones were included in log
5277 43. Exiwhat now gives more information about the daemon's listening ports,
5278 and whether -tls-on-connect was used.
5280 44. The "port" option of the smtp transport is now expanded.
5282 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5283 silently ignored. Now an error message is written to the main and panic
5286 46. There's a new ACL modifier called "logwrite" which writes to a log file
5287 as soon as it is encountered.
5289 47. Added $local_user_uid and $local_user_gid at routing time.
5291 48. Exim crashed when trying to verify a sender address that was being
5294 49. Exim was recognizing only a space character after ".include". It now also
5295 recognizes a tab character.
5297 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5298 extracting the relevant information from the specification. The man page no
5299 longer contains scrambled data for the -d option, and I've added a section
5300 at the front about calling Exim under different names.
5302 51. Added "extra_headers" argument to the "mail" command in filter files.
5304 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5307 53. Installed eximstats 1.29.
5309 54. Added transport_filter_timeout as a generic transport option.
5311 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5312 Cc: header lines. This was required by RFC 822, but it not required by RFC
5315 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5316 incoming messages that did not have them. Now it does so only if the
5317 message originates locally, that is, if there is no associated remote host
5318 address. When Resent- header lines are present, this applies to the Resent-
5319 lines rather than the non-Resent- lines.
5321 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5322 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5324 58. Messages for configuration errors now include the name of the main
5325 configuration file - useful now that there may be more than one file in a
5326 list (.included file names were always shown).
5328 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5329 for those rare installations that do not start the daemon as root or run it
5330 setuid root. I've cut out the call to initgroups() if the daemon is not
5333 60. The Exim user and group can now be bound into the binary as text strings
5334 that are looked up at the start of Exim's processing.
5336 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5338 62. Added $mailstore_basename variable.
5340 63. Installed patch to sieve.c from Michael Haardt.
5342 64. When Exim failed to open the panic log after failing to open the main log,
5343 the original message it was trying to log was written to stderr and debug
5344 output, but if they were not available (the usual case in production), it
5345 was lost. Now it is written to syslog before the two lines that record the
5346 failures to open the logs.
5348 65. Users' Exim filters run in subprocesses under the user's uid. It is
5349 possible for a "deliver" command or an alias in a "personal" command to
5350 provoke an address rewrite. If logging of address rewriting is configured,
5351 this fails because the process is not running as root or exim. There may be
5352 a better way of dealing with this, but for the moment (because 4.30 needs
5353 to be released), I have disabled address rewrite logging when running a
5354 filter in a non-root, non-exim process.
5360 1. The buildconfig auxiliary program wasn't quoting the value set for
5361 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5362 not defined. This bug was masked in 4.22 by the effect that was fixed in
5365 2. Some messages that were rejected after a message id was allocated were
5366 shown as "incomplete" by exigrep. It no longer does this for messages that
5367 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5369 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5370 have allow_domain_literals set, the ID did not get logged in the <= line.
5371 Domain literals are now always recognized in Message-ID: header lines.
5373 4. The first argument for a ${extract expansion item is the key name or field
5374 number. Leading and trailing spaces in this item were not being ignored,
5375 causing some misleading effects.
5377 5. When deliver_drop_privilege was set, single queue runner processes started
5378 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5379 same command in the process it spins off) were not dropping privilege.
5381 6. When the daemon running as "exim" started a queue runner, it always
5382 re-executed Exim in the spun-off process. This is a waste of effort when
5383 deliver_drop_privilege is set. The new process now just calls the
5384 queue-runner function directly.
5390 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5393 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5394 ignored. Though the use of -oP was forcing the writing of a pid file, it
5395 was always written to the default place.
5397 3. If the message "no IP address found for host xxxx" is generated during
5398 incoming verification, it is now followed by identification of the incoming
5399 connection (so you can more easily find what provoked it).
5401 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5403 5. Added some features to "harden" Exim a bit more against certain attacks:
5405 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5406 be put in Local/Makefile. This is like the never_users runtime option,
5407 but it cannot be overridden. The default setting is "root".
5409 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5410 prefix string with which any file named in a -C command line option
5413 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5414 is retained for -C and -D only if the caller of Exim is root. Without
5415 it, the exim user may also use -C and -D and retain privilege.
5417 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5418 command line option is disabled.
5420 6. Macro names set by the -D option must start with an upper case letter, just
5421 like macro names defined in the configuration file.
5423 7. Added "dereference=" facility to LDAP.
5425 8. Two instances of the typo "uknown" in the source files are fixed.
5427 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5428 the Configure-Makefile script screwed up while processing it.
5430 10. Incorporated PCRE 4.4.
5432 11. The SMTP synchronization check was not operating right at the start of an
5433 SMTP session. For example, it could not catch a HELO sent before the client
5434 waited for the greeting. There is now a check for outstanding input at the
5435 point when the greeting is written. Because of the duplex, asynchronous
5436 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5437 way, but not yet received, when the check is performed.
5439 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5440 on TCP/IP sockets, because this apparently causes some broken clients to
5443 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5444 unchanged) from the Cygwin maintainer.
5446 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5447 was supported instead of testing for mbx. Effectively a typo.
5449 15. The spa authenticator server code was not checking that the input it
5450 received was valid base64.
5452 16. The debug output line for the "set" modifier in ACLs was not showing the
5453 name of the variable that was being set.
5455 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5457 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5458 lookup on its own. Something else was needed to trigger the lookup. For
5459 example, a match in host_lookup or the need for a host name in a host list.
5460 Now, if $sender_host_name is referenced and the host name has not yet been
5461 looked up, a lookup is performed. If the lookup fails, the variable remains
5462 empty, and $host_lookup_failed is set to "1".
5464 19. Added "eqi" as a case-independent comparison operator.
5466 20. The saslauthd authentication condition could segfault if neither service
5467 nor realm was specified.
5469 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5470 error message that was logged was misleading, and incoming SMTP
5471 connections were dropped. The message is now more accurate, and temporary
5472 errors are given to SMTP connections.
5474 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5475 (except RSET) with a 421 error, until QUIT is received. However, it was
5476 failing to send a response to QUIT.
5478 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5479 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5480 ACL is now run after the helo_try_verify_hosts code.
5482 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5483 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5484 case-independent, so other case variants are also recognized). Apparently
5485 some systems use these upper case variants.
5487 25. If more than two messages were waiting for the same host, and a transport
5488 filter was specified for the transport, Exim sent two messages over the
5489 same TCP/IP connection, and then failed with "socket operation on non-
5490 socket" when it tried to send the third.
5492 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5494 27. The extern definition of crypt16() in expand.c was not being excluded when
5495 the OS had its own crypt16() function.
5497 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5498 as a preferred synonym for return_size_limit, both as an option and as an
5501 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5503 30. Changed the default configuration ACL to relax the local part checking rule
5504 for addresses that are not in any local domains. For these addresses,
5505 slashes and pipe symbols are allowed within local parts, but the sequence
5506 /../ is explicitly forbidden.
5508 31. SPA server authentication was not clearing the challenge buffer before
5511 32. log_message in a "warn" ACL statement was writing to the reject log as
5512 well as to the main log, which contradicts the documentation and doesn't
5513 seem right (because no rejection is happening). So I have stopped it.
5515 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5516 However, I am unable to do any testing of this.
5518 34. Fixed an infelicity in the appendfile transport. When checking directories
5519 for a mailbox, to see if any needed to be created, it was accidentally
5520 using path names with one or more superfluous leading slashes; tracing
5521 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5523 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5524 contents are added to the log line that is written for every discarded
5525 recipient. (Previously a log_message setting was ignored.)
5527 36. The ${quote: operator now quotes the string if it is empty.
5529 37. The install script runs exim in order to find its version number. If for
5530 some reason other than non-existence or emptiness, which it checks, it
5531 could not run './exim', it was installing it with an empty version number,
5532 i.e. as "exim-". This error state is now caught, and the installation is
5535 38. An argument was missing from the function that creates an error message
5536 when Exim fails to connect to the socket for saslauthd authentication.
5537 This could cause Exim to crash, or give a corrupted message.
5539 39. Added isip, isip4, and isip6 to ${if conditions.
5541 40. The ACL variables $acl_xx are now saved with the message, and can be
5542 accessed later in routers, transports, and filters.
5544 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5545 strings in the file are not string-expanded.
5547 42. If a MAIL command specified a SIZE value that was too large to fit into an
5548 int variable, the check against message_size_limit failed. Such values are
5549 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5550 day this will have to be increased, but I don't think I want to be around
5551 when emails are that large.
5558 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5559 iconv() is not standard in FreeBSD.
5561 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5562 IPv6 enabled. The observed symptom was a segmentation fault on return from
5563 the function os_common_find_running_interfaces() in src/os.c.
5565 3. In the check_special_case() function in daemon.c I had used "errno" as an
5566 argument name, which causes warnings on some systems. This was basically a
5567 typo, since it was named "eno" in the comments!
5569 4. The code that waits for the clock to tick (at a resolution of some fraction
5570 of a second) so as to ensure message-id uniqueness was always waiting for
5571 at least one whole tick, when it could have waited for less. [This is
5572 almost certainly not relevant at current processor speeds, where it is
5573 unlikely to ever wait at all. But we try to future-proof.]
5575 5. The function that sleeps for a time interval that includes fractions of a
5576 second contained a race. It did not block SIGALRM between setting the
5577 timer, and suspending (a couple of lines later). If the interval was short
5578 and the sigsuspend() was delayed until after it had expired, the suspension
5579 never ended. On busy systems this could lead to processes getting stuck for
5582 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5583 process, before it forks any delivery processes. The open lookup caching
5584 mechanism meant that the open file or database connection was passed into
5585 the delivery process. The problem was that delivery processes always tidy
5586 up cached lookup data. This could cause a problem for the next delivery
5587 process started by the queue runner, because the external queue runner
5588 process does not know about the closure. So the next delivery process
5589 still has data in the lookup cache. In the case of a file lookup, there was
5590 no problem because closing a file descriptor in a subprocess doesn't affect
5591 the parent. However, if the lookup was caching a connection to a database,
5592 the connection was closed, and the second delivery process was likely to
5593 see errors such as "PGSQL: query failed: server closed the connection
5594 unexpectedly". The problem has been fixed by closing all cached lookups
5595 in a queue runner before running a delivery process.
5597 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5598 seem to have the "const" qualifier which it has on other OS. I've
5601 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5602 *of the same type* (client or server) with the same public name that an
5603 error should be diagnosed.
5605 9. When Exim looked up a host name for an IP address, but failed to find the
5606 original IP address when looking up the host name (a safety check), it
5607 output the message "<ip address> does not match any IP for NULL", which was
5608 confusing, to say the least. The bug was that the host name should have
5609 appeared instead of "NULL".
5611 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5612 user that is built into the binary, and the -C or -D options is used, root
5613 privilege is dropped before the configuration file is read. In addition,
5614 logging is switched to stderr instead of the normal log files. If the
5615 configuration then re-defines the Exim user, the unprivileged environment
5616 is probably not what is expected, so Exim logs a panic warning message (but
5619 However, if deliver_drop_privilege is set, the unprivileged state may well
5620 be exactly what is intended, so the warning has been cut out in that case,
5621 and Exim is allowed to try to write to its normal log files.
5627 1. smtp_return_error_details was not giving details for temporary sender
5628 or receiver verification errors.
5630 2. Diagnose a configuration error if two authenticators have the same public
5633 3. Exim used not to create the message log file for a message until the first
5634 delivery attempt. This could be confusing when incoming messages were held
5635 for policy or load reasons. The message log file is now created at the time
5636 the message is received, and an initial "Received" line is written to it.
5638 4. The automatically generated man page for command line options had a minor
5639 bug that caused no ill effects; however, a more serious problem was that
5640 the procedure for building the man page automatically didn't always
5641 operate. Consequently, release 4.20 contains an out-of-date version. This
5642 shouldn't happen again.
5644 5. When building Exim with embedded Perl support, the script that builds the
5645 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5646 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5648 6. The freeze_tell option was not being used for messages that were frozen on
5649 arrival, either by an ACL or by local_scan().
5651 7. Added the smtp_incomplete_transaction log selector.
5653 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5654 was accepting AUTH without a new EHLO.
5656 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5657 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5660 10. Logging of TCP/IP connections (when configured) now happens in the main
5661 daemon process instead of the child process, so that the TCP/IP connection
5662 count is more accurate (but it can never be perfect).
5664 11. The use of "drop" in a nested ACL was not being handled correctly in the
5665 outer ACL. Now, if condition failure induced by the nested "drop" causes
5666 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5667 or "require"), the connection is dropped.
5669 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5670 that yield "discard" can now be used with an "accept" or a "discard" verb,
5671 but an error is generated for any others (because I can't see a useful way
5672 to define what should happen).
5674 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5675 are now processed in the same way as lines in the Exim configuration file.
5676 In particular, continuation lines are supported.
5678 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5680 15. Added -ti meaning -t -i.
5682 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5683 domains, and warn by logging if others are found.
5685 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5686 returned from GIFCONF when Exim is trying to find the list of interfaces on
5687 a host. The code in os.c has been modified to copy each ifreq to an aligned
5688 structure in all cases.
5690 Also, in some cases, the returned ifreq's were being copied to a 'struct
5691 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5692 means the last couple of bytes of an IPv6 address could be chopped if the
5693 ifreq contained only a normal sockaddr (14 bytes storage).
5695 18. Named domain lists were not supported in the hosts_treat_as_local option.
5696 An entry such as +xxxx was not recognized, and was treated as a literal
5699 19. Ensure that header lines added by a DATA ACL are included in the reject log
5700 if the ACL subsequently rejects the message.
5702 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5703 MD5 (which is deprecated).
5705 21. When testing a filter file using -bf, Exim was writing a message when it
5706 took the sender from a "From " line in the message, but it was not doing so
5707 when it took $return_path from a Return-Path: header line. It now does.
5709 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5710 with a valid header line field name (a series of printing characters
5711 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5713 23. Changed "disc" in the source to "disk" to conform to the documentation and
5714 the book and for uniformity.
5716 24. Ignore Sendmail's -Ooption=value command line item.
5718 25. When execve() failed while trying to run a command in a pipe transport,
5719 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5720 could be confused with a return value of 69 from the command itself. This
5721 has been changed to 127, the value the shell returns if it is asked to run
5722 a non-existent command. The wording for the related log line suggests a
5723 non-existent command as the problem.
5725 26. If received_header_text expands to an empty string, do not add a Received:
5726 header line to the message. (Well, it adds a token one on the spool, but
5727 marks it "old" so that it doesn't get used or transmitted.)
5729 27. Installed eximstats 1.28 (addition of -nt option).
5731 28. There was no check for failure on the call to getsockname() in the daemon
5732 code. This can fail if there is a shortage of resources on the system, with
5733 ENOMEM, for example. A temporary error is now given on failure.
5735 29. Contrary to the C standard, it seems that in some environments, the
5736 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5737 program. Exim now does this explicitly; it affects the formatting of
5738 timestamps using strftime().
5740 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5741 complaints. I've now initialized all the variables, to avoid this.
5743 32. Header lines added by a system filter were not being "seen" during
5744 transport-time rewrites.
5746 33. The info_callback() function passed to OpenSSL is set up with type void
5747 (*)(SSL *, int, int), as described somewhere. However, when calling the
5748 function (actually a macro) that sets it up, the type void(*)() is
5749 expected. I've put in a cast to prevent warnings from picky compilers.
5751 34. If a DNS black list lookup found a CNAME record, but there were no A
5752 records associated with the domain it pointed at, Exim crashed.
5754 35. If a DNS black list lookup returned more than one A record, Exim ignored
5755 all but the first. It now scans all returned addresses if a particular IP
5756 value is being sought. In this situation, the contents of the
5757 $dnslist_value variable are a list of all the addresses, separated by a
5760 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5761 to accept a host name and all its aliases if the forward lookup for any of
5762 them yielded the IP address of the incoming connection. Now it accepts only
5763 those names whose forward lookup yields the correct IP address. Any other
5764 names are discarded. This closes a loophole whereby a rogue DNS
5765 administrator could create reverse DNS records to break through a
5766 wildcarded host restriction in an ACL.
5768 37. If a user filter or a system filter that ran in a subprocess used any of
5769 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5770 the wrong values were passed to the pipe command ($thisaddress had the
5771 value of $0, $0 had the value of $1, etc). This bug was introduced by
5772 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5774 38. Improved the line breaking for long SMTP error messages from ACLs.
5775 Previously, if there was no break point between 40 and 75 characters, Exim
5776 left the rest of the message alone. Two changes have been made: (a) I've
5777 reduced the minimum length to 35 characters; (b) if it can't find a break
5778 point between 35 and 75 characters, it looks ahead and uses the first one
5779 that it finds. This may give the occasional overlong line, but at least the
5780 remaining text gets split now.
5782 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5783 file descriptors might be low, and that setting 1000 would always raise it.
5784 It turns out that in some environments, the limit is already over 1000 and
5785 that lowering it causes trouble. So now Exim takes care not to decrease it.
5787 40. When delivering a message, the value of $return_path is set to $sender_
5788 address at the start of routing (routers may change the value). By an
5789 oversight, this default was not being set up when an address was tested by
5790 -bt or -bv, which affected the outcome if any router or filter referred to
5793 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5794 log only when "message" or "log_message" are set. However, if one of the
5795 conditions was an address verification, or a call to a nested ACL, the
5796 messages generated by the underlying test were being passed through. This
5797 no longer happens. The underlying message is available in $acl_verify_
5798 message for both "message" and "log_message" expansions, so it can be
5799 passed through if needed.
5801 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5802 new expansion $bh_ to give the encoded byte string without charset
5803 translation. Translation happens only if iconv() is available; HAVE_ICONV
5804 indicates this at build time. HEADERS_CHARSET gives the charset to
5805 translate to; headers_charset can change it in the configuration, and
5806 "headers charset" can change it in an individual filter file.
5808 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5809 that creates RFC 2047 encoded "words" labels them as that charset instead
5810 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5811 expansion operator; (ii) when Exim creates a From: line for a local
5812 message; (iii) when a header line is rewritten to include a "phrase" part.
5814 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5815 buggy, causing it to skip the first lines of messages whose message ID
5816 ended in 'D'. This would not have bitten before Exim release 4.14, because
5817 message IDs were unlikely to end in 'D' before then. The effect was to have
5818 incorrect size information for certain domains.
5820 45. #include "config.h" was missing at the start of the crypt16.c module. This
5821 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5824 46. If there was a timeout during a "random" callout check, Exim treated it as
5825 a failure of the random address, and carried on sending RSET and the real
5826 address. If the delay was just some slowness somewhere, the response to the
5827 original RCPT would be taken as a response to RSET and so on, causing
5828 mayhem of various kinds.
5830 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5831 when I implemented it. It didn't allow for the fact that some option values
5832 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5833 the right test for positive values.
5835 48. Domain names in DNS records are case-independent. Exim always looks them up
5836 in lower case. Some resolvers return domain names in exactly the case they
5837 appear in the zone file, that is, they may contain uppercase letters. Not
5838 all resolvers do this - some return always lower case. Exim was treating a
5839 change of case by a resolver as a change of domain, similar to a widening
5840 of a domain abbreviation. This triggered its re-routing code and so it was
5841 trying to route what was effectively the same domain again. This normally
5842 caused routing to fail (because the router wouldn't handle the domain
5843 twice). Now Exim checks for this case specially, and just changes the
5844 casing of the domain that it ultimately uses when it transmits the message
5847 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5850 50. If a filter generated a file delivery with a non-absolute name (possible if
5851 no home directory exists for the router), the forbid_file option was not
5854 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5855 the existing equality matching.
5857 52. Exim was using ints instead of ino_t variables in some places where it was
5858 dealing with inode numbers.
5860 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5861 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5862 called TMPDIR, and if it finds it is different, it changes its value.
5864 54. The smtp_printf() function is now made available to local_scan() so
5865 additional output lines can be written before returning. There is also an
5866 smtp_fflush() function to enable the detection of a dropped connection.
5867 The variables smtp_input and smtp_batched_input are exported to
5870 55. Changed the default runtime configuration: the message "Unknown user"
5871 has been removed from the ACL, and instead placed on the localuser router,
5872 using the cannot_route_message feature. This means that any verification
5873 failures that generate their own messages won't get overridden. Similarly,
5874 the "Unrouteable address" message that was in the ACL for unverifiable
5875 relay addresses has also been removed.
5877 56. Added hosts_avoid_esmtp to the smtp transport.
5879 57. The exicyclog script was not checking for the esoteric option
5880 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5881 will work only if exicyclog is run under the appropriate euid.
5883 58. Following a discussion on the list, the rules by which Exim recognises line
5884 endings on incoming messages have been changed. The -dropcr and drop_cr
5885 options are now no-ops, retained only for backwards compatibility. The
5886 following line terminators are recognized: LF CRLF CR. However, special
5887 processing applies to CR:
5889 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5890 nor a local message in the state where . is a terminator.
5892 (ii) If a bare CR is encountered in a header line, an extra space is added
5893 after the line terminator so as not to end the header. The reasoning
5894 behind this is that bare CRs in header lines are most likely either
5895 to be mistakes, or people trying to play silly games.
5897 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5898 was being incorrectly given as 18 bytes larger than it should have been.
5899 This is a VOB (very old bug).
5901 60. This may never have affected anything current, but just in case it has:
5902 When the local host is found other than at the start of a list of hosts,
5903 the local host, those with the same MX, and any that follow, are discarded.
5904 When the list in question was part of a longer list of hosts, the following
5905 hosts (not currently being processed) were also being discarded. This no
5906 longer happens. I'm not sure if this situation could ever has previously
5909 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5912 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5913 header. This is something that is recommended in a new Internet Draft, and
5914 is something that is documented as being done by Sendmail. There are two
5915 possible values. For messages generated by the autoreply transport, Exim
5918 Auto-Submitted: auto-replied
5920 whereas for all other generated messages (e.g. bounces) it adds
5922 Auto-Submitted: auto-generated
5924 63. The "personal" condition in filters now includes a test for the
5925 Auto-Submitted: header. If it contains the string "auto-" the message it
5926 not considered personal.
5928 64. Added rcpt_include_affixes as a generic transport option.
5930 65. Added queue_only_override (default true).
5932 66. Added the syslog_duplication option.
5934 67. If what should have been the first header line of a message consisted of
5935 a space followed by a colon, Exim was mis-interpreting it as a header line.
5936 It isn't of course - it is syntactically invalid and should therefore be
5937 treated as the start of the message body. The misbehaviour could have
5938 caused a number of strange effects, including loss of data in subsequent
5939 header lines, and spool format errors.
5941 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5942 client host had authenticated. This control can now be exercised by an ACL
5943 for more flexibility.
5945 69. By default, callouts do not happen when testing with -bh. There is now a
5946 variant, -bhc, which does actually run the callout code, including
5947 consulting and updating the callout cache.
5949 70. Added support for saslauthd authentication, courtesy of Alexander
5952 71. If statvfs() failed on the spool or log directories while checking their
5953 size for availability, Exim confusingly gave the error "space shortage".
5954 Furthermore, in debugging mode it crashed with a floating point exception.
5955 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5956 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5957 serious problem, Exim now writes to the main and panic logs when this
5958 happens, with details of the failure. It then refuses to accept the
5959 incoming message, giving the message "spool directory problem" or "log
5960 directory problem" with a 421 code for SMTP messages.
5962 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5963 0, 1, and 2 exist, because some OS complain for execs without them (see
5964 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5965 descriptors. However, the code omitted to check that the open succeeded,
5966 causing mysterious errors if for some reason the permissions on /dev/null
5967 got screwed. Now Exim writes a message to the main and panic logs, and
5968 bombs out if it can't open /dev/null.
5970 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5971 interact so that it is all more flexible. It is supposed to remain
5972 backwards compatible. Also added extra_local_interfaces.
5974 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5975 to bomb out with an assertion failure - to the client this appears as a
5976 connection drop. This problem occurs in the part of the code that was taken
5977 from the Samba project. Fortunately, the assertion is in a very simple
5978 function, so I have fixed this by reproducing the function inline in the
5979 one place where it is called, and arranging for authentication to fail
5980 instead of killing the process with assert().
5982 75. The SPA client code was not working when the server requested OEM rather
5983 than Unicode encoding.
5985 76. Added code to make require_files with a specific uid setting more usable in
5986 the case where statting the file as root fails - usually a non-root-mounted
5987 NFS file system. When this happens and the failure is EACCES, Exim now
5988 forks a subprocess and does the per-uid checking as the relevant uid.
5990 77. Added process_log_path.
5992 78. If log_file_path was not explicitly set, a setting of check_log_space or
5993 check_log_inodes was ignored.
5995 79. If a space check for the spool or log partitions fails, the incident is now
5996 logged. Of course, in the latter case the data may get lost...
5998 80. Added the %p formatting code to string_format() so that it can be used to
5999 print addresses in debug_print(). Adjusted all the address printing in the
6000 debugging in store.c to use %p rather than %d.
6002 81. There was a concern that a line of code in smtp_in.c could overflow a
6003 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
6004 initially expressed, the concern was not well-founded, because trailing
6005 spaces are removed early. However, if the trailing spaces were followed by
6006 a NULL, they did not get removed, so the overflow was possible. Two fixes
6009 (a) I re-wrote the offending code in a cleaner fashion.
6010 (b) If an incoming SMTP command contains a NULL character, it is rejected
6013 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
6014 runs initgroups(), so that if the Exim user is in any additional groups,
6015 they will be used during message reception.
6021 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob