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. Is a backward incompatibility if
46 the ACL tests for either of these two results. Patch contributed by
47 user bes-internal on the mailing list.
49 JH/04 Add ${utf8clean:} operator. Contributed by Alex Rau.
51 JH/05 Bugzilla 305: Log incoming-TLS details on rejects, subject to log
52 selectors, in both main and reject logs.
54 JH/06 Log outbound-TLS and port details, subject to log selectors, for a
57 JH/07 Add malware type "sock" for talking to simple daemon.
59 JH/08 Bugzilla 1371: Add tls_{,try_}verify_hosts to smtp transport.
62 JH/09 Bugzilla 1431: Support (with limitations) headers_add/headers_remove in
63 routers/transports under cutthrough routing.
69 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
71 PP/02 Make -n do something, by making it not do something.
72 When combined with -bP, the name of an option is not output.
74 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
77 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
78 $sender_host_name and config options to manage this, and basic check
81 PP/05 DSCP support for outbound connections and control modifier for inbound.
83 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
84 (Only plugin which currently uses this is kerberos4, which nobody should
85 be using, but we should make it available and other future plugins might
86 conceivably use it, even though it would break NAT; stuff *should* be
87 using channel bindings instead).
89 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
90 name; added for Sendmail compatibility; requires admin caller.
91 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
92 just ignore it); requires trusted caller.
93 Also parse but ignore: -Ac -Am -X<logfile>
96 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
98 TL/02 Add +smtp_confirmation as a default logging option.
100 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
101 Patch by Magnus Holmgren from 2007-02-20.
103 TL/04 Bugzilla 1281 - Spec typo.
104 Bugzilla 1283 - Spec typo.
105 Bugzilla 1290 - Spec grammar fixes.
107 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
109 TL/06 Add Experimental DMARC support using libopendmarc libraries.
111 TL/07 Fix an out of order global option causing a segfault. Reported to dev
112 mailing list by by Dmitry Isaikin.
114 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
116 JH/02 Support "G" suffix to numbers in ${if comparisons.
118 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
120 NM/01 Bugzilla 1197 - Spec typo
121 Bugzilla 1196 - Spec examples corrections
123 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
125 PP/09 Add gnutls_allow_auto_pkcs11 option (was originally called
126 gnutls_enable_pkcs11, but renamed to more accurately indicate its
129 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
130 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
132 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
133 "acl {{name}{arg}...}", and optional args on acl condition
136 JH/05 Permit multiple router/transport headers_add/remove lines.
138 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
140 JH/07 Avoid using a waiting database for a single-message-only transport.
141 Performance patch from Paul Fisher. Bugzilla 1262.
143 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
146 JH/09 Add $headers_added variable, with content from use of ACL modifier
147 add_header (but not yet added to the message). Bugzilla 199.
149 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
150 Pulled from Bugzilla 817 by Wolfgang Breyha.
152 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
154 (nb: this is the same fix as in Exim 4.80.1)
156 JH/11 Add A= logging on delivery lines, and a client_set_id option on
159 JH/12 Add optional authenticated_sender logging to A= and a log_selector
162 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
164 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
165 advertise SMTP AUTH mechanism to us, instead of a generic
166 protocol violation error. Also, make Exim more robust to bad
167 data from the Dovecot auth socket.
169 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
171 When a queue runner is handling a message, Exim first routes the
172 recipient addresses, during which it prunes them based on the retry
173 hints database. After that it attempts to deliver the message to
174 any remaining recipients. It then updates the hints database using
177 So if a recipient address works intermittently, it can get repeatedly
178 deferred at routing time. The retry hints record remains fresh so the
179 address never reaches the final cutoff time.
181 This is a fairly common occurrence when a user is bumping up against
182 their storage quota. Exim had some logic in its local delivery code
183 to deal with this. However it did not apply to per-recipient defers
184 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
186 This change adds a proper retry rule check during routing so that the
187 final cutoff time is checked against the message's age. We only do
188 this check if there is an address retry record and there is not a
189 domain retry record; this implies that previous attempts to handle
190 the address had the retry_use_local_parts option turned on. We use
191 this as an approximation for the destination being like a local
192 delivery, as in LMTP.
194 I suspect this new check makes the old local delivery cutoff check
195 redundant, but I have not verified this so I left the code in place.
197 TF/02 Correct gecos expansion when From: is a prefix of the username.
199 Test 0254 submits a message to Exim with the header
203 When I ran the test suite under the user fanf2, Exim expanded
204 the header to contain my full name, whereas it should have added
205 a Resent-Sender: header. It erroneously treats any prefix of the
206 username as equal to the username.
208 This change corrects that bug.
210 GF/01 DCC debug and logging tidyup
211 Error conditions log to paniclog rather than rejectlog.
212 Debug lines prefixed by "DCC: " to remove any ambiguity.
214 TF/03 Avoid unnecessary rebuilds of lookup-related code.
216 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
217 Bug spotted by Jeremy Harris; was flawed since initial commit.
218 Would have resulted in OCSP responses post-SNI triggering an Exim
219 NULL dereference and crash.
221 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
223 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
224 Bug detection, analysis and fix by Samuel Thibault.
225 Bugzilla 1331, Debian bug #698092.
227 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
229 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
230 Server implementation by Todd Lyons, client by JH.
231 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
232 config variable "prdr_enable" controls whether the server
233 advertises the facility. If the client requests PRDR a new
234 acl_data_smtp_prdr ACL is called once for each recipient, after
235 the body content is received and before the acl_smtp_data ACL.
236 The client is controlled by bolth of: a hosts_try_prdr option
237 on the smtp transport, and the server advertisement.
238 Default client logging of deliveries and rejections involving
239 PRDR are flagged with the string "PRDR".
241 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
242 fclose(). Diagnosis by Todd Lyons.
244 PP/17 Update configure.default to handle IPv6 localhost better.
245 Patch by Alain Williams (plus minor tweaks).
248 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
249 This is now consistent with GnuTLS, and is now documented: the
250 previous undocumented portable approach to treating the option as
251 unset was to force an expansion failure. That still works, and
252 an empty string is now equivalent.
254 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
255 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
256 not performing validation itself.
258 PP/20 Added force_command boolean option to pipe transport.
259 Patch from Nick Koston, of cPanel Inc.
261 JH/15 AUTH support on callouts (and hence cutthrough-deliveries).
264 TF/04 Added udpsend ACL modifer and hexquote expansion operator
266 PP/21 Fix eximon continuous updating with timestamped log-files.
267 Broken in a format-string cleanup in 4.80, missed when I repaired the
268 other false fix of the same issue.
269 Report and fix from Heiko Schlichting.
272 PP/22 Guard LDAP TLS usage against Solaris LDAP variant.
273 Report from Prashanth Katuri.
275 PP/23 Support safari_ecdhe_ecdsa_bug for openssl_options.
276 It's SecureTransport, so affects any MacOS clients which use the
277 system-integrated TLS libraries, including email clients.
279 PP/24 Fix segfault from trying to fprintf() to a NULL stdio FILE* if
280 using a MIME ACL for non-SMTP local injection.
281 Report and assistance in diagnosis by Warren Baker.
283 TL/08 Adjust exiqgrep to be case-insensitive for sender/receiver.
285 JH/16 Fix comparisons for 64b. Bugzilla 1385.
287 TL/09 Add expansion variable $authenticated_fail_id to keep track of
288 last id that failed so it may be referenced in subsequent ACL's.
290 TL/10 Bugzilla 1375 - Prevent TLS rebinding in ldap. Patch provided by
293 TL/11 Bugzilla 1382 - Option ldap_require_cert overrides start_tls
294 ldap library initialization, allowing self-signed CA's to be
295 used. Also properly sets require_cert option later in code by
296 using NULL (global ldap config) instead of ldap handle (per
297 session). Bug diagnosis and testing by alxgomz.
299 TL/12 Enhanced documentation in the ratelimit.pl script provided in
300 the src/util/ subdirectory.
302 TL/13 Bug 1301 - Imported transport SQL logging patch from Axel Rau
303 renamed to Transport Post Delivery Action by Jeremy Harris, as
306 TL/14 Bugzilla 1217 - Redis lookup support has been added. It is only enabled
307 when Exim is compiled with EXPERIMENTAL_REDIS. A new config variable
308 redis_servers = needs to be configured which will be used by the redis
309 lookup. Patch from Warren Baker, of The Packet Hub.
311 TL/15 Fix exiqsumm summary for corner case. Patch provided by Richard Hall.
313 TL/16 Bugzilla 1289 - Clarify host/ip processing when have errors looking up a
314 hostname or reverse DNS when processing a host list. Used suggestions
315 from multiple comments on this bug.
317 TL/17 Bugzilla 1057 - Multiple clamd TCP targets patch from Mark Zealey.
319 TL/18 Had previously added a -CONTINUE option to runtest in the test suite.
320 Missed a few lines, added it to make the runtest require no keyboard
323 TL/19 Bugzilla 1402 - Test 533 fails if any part of the path to the test suite
324 contains upper case chars. Make router use caseful_local_part.
326 TL/20 Bugzilla 1400 - Add AVOID_GNUTLS_PKCS11 build option. Allows GnuTLS
327 support when GnuTLS has been built with p11-kit.
333 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
335 This, or similar/improved, will also be change PP/11 of 4.82.
341 PP/01 Handle short writes when writing local log-files.
342 In practice, only affects FreeBSD (8 onwards).
343 Bugzilla 1053, with thanks to Dmitry Isaikin.
345 NM/01 Bugzilla 949 - Documentation tweak
347 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
350 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
352 PP/02 Implemented gsasl authenticator.
354 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
356 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
357 `pkg-config foo` for cflags/libs.
359 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
360 with rest of GSASL and with heimdal_gssapi.
362 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
363 `pkg-config foo` for cflags/libs for the TLS implementation.
365 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
366 properties get this fed in as external SSF. A number of robustness
367 and debugging improvements to the cyrus_sasl authenticator.
369 PP/08 cyrus_sasl server now expands the server_realm option.
371 PP/09 Bugzilla 1214 - Log authentication information in reject log.
372 Patch by Jeremy Harris.
374 PP/10 Added dbmjz lookup type.
376 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
378 PP/12 MAIL args handles TAB as well as SP, for better interop with
379 non-compliant senders.
380 Analysis and variant patch by Todd Lyons.
382 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
383 Bug report from Lars Müller <lars@samba.org> (via SUSE),
384 Patch from Dirk Mueller <dmueller@suse.com>
386 PP/13 tls_peerdn now print-escaped for spool files.
387 Observed some $tls_peerdn in wild which contained \n, which resulted
388 in spool file corruption.
390 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
391 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
392 or write after TLS renegotiation, which otherwise led to messages
395 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
396 as a tracking header (ie: a signed header comes before the signature).
397 Patch from Wolfgang Breyha.
399 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
400 comma-sep list; embedded commas doubled.
402 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
404 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
406 Report and patch from Dmitry Banschikov.
408 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
409 Removed SSL_clear() after SSL_new() which led to protocol negotiation
410 failures. We appear to now support TLS1.1+ with Exim.
412 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
413 lets Exim select keys and certificates based upon TLS SNI from client.
414 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
415 before an outbound SMTP session. New log_selector, +tls_sni.
417 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
418 NULL dereference. Report and patch from Alun Jones.
420 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
421 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
422 Not seeing resolver debug output on NetBSD, but suspect this is a
423 resolver implementation change.
425 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
426 Left warnings. Added "eximon gdb" invocation mode.
428 PP/21 Defaulting "accept_8bitmime" to true, not false.
430 PP/22 Added -bw for inetd wait mode support.
432 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
433 locate the relevant includes and libraries. Made this the default.
435 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
436 Bugzilla 1246, report and most of solution from Tomasz Kusy.
438 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
439 This may cause build issues on older platforms.
441 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
442 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
443 gnutls_require_mac & gnutls_require_protocols (no longer supported).
444 Added SNI support via GnuTLS too.
445 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
447 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
449 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
450 Applied second patch from Janne, implementing suggestion to default
451 multiple-strings-in-record handling to match SPF spec.
453 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
455 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
456 read-only, out of scope).
457 Patch from Wolfgang Breyha, report from Stuart Northfield.
459 PP/29 Fix three issues highlighted by clang analyser static analysis.
460 Only crash-plausible issue would require the Cambridge-specific
461 iplookup router and a misconfiguration.
462 Report from Marcin Mirosław.
464 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
466 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
467 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
468 As part of this, removing so much warning spew let me fix some minor
469 real issues in debug logging.
471 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
472 assignment on my part. Fixed.
474 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
475 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
476 Janne Snabb (who went above and beyond: thank you).
478 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
479 string otherwise requires a connection and a bunch more work and it's
480 relatively easy to get wrong. Should also expose TLS library linkage
483 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
484 64-bit ${eval} (JH/03).
486 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
487 GNU libc to support some of the 64-bit stuff, should not lead to
488 conflicts. Defined before os.h is pulled in, so if a given platform
489 needs to override this, it can.
491 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
492 protection layer was required, which is not implemented.
493 Bugzilla 1254, patch from Wolfgang Breyha.
495 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
496 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
497 tls_dhparam take prime identifiers. Also unbreak combination of
498 OpenSSL+DH_params+TLSSNI.
500 PP/39 Disable SSLv2 by default in OpenSSL support.
506 PP/01 Solaris build fix for Oracle's LDAP libraries.
507 Bugzilla 1109, patch from Stephen Usher.
509 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
511 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
514 TF/02 Fix a couple more cases where we did not log the error message
515 when unlink() failed. See also change 4.74-TF/03.
517 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
518 lock up or crash if it happened to be inside a call to libc when it
519 got a SIGUSR1 from exiwhat.
521 The SIGUSR1 handler appends the current process status to the process
522 log which is later printed by exiwhat. It used to use the general
523 purpose logging code to do this, but several functions it calls are
524 not safe for signals.
526 The new output code in the SIGUSR1 handler is specific to the process
527 log, and simple enough that it's easy to inspect for signal safety.
528 Removing some special cases also simplifies the general logging code.
529 Removing the spurious timestamps from the process log simplifies
532 TF/04 Improved ratelimit ACL condition.
534 The /noupdate option has been deprecated in favour of /readonly which
535 has clearer semantics. The /leaky, /strict, and /readonly update modes
536 are mutually exclusive. The update mode is no longer included in the
537 database key; it just determines when the database is updated. (This
538 means that when you upgrde Exim will forget old rate measurements.)
540 Exim now checks that the per_* options are used with an update mode that
541 makes sense for the current ACL. For example, when Exim is processing a
542 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
543 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
544 must specify per_mail/readonly. If you omit the update mode it defaults to
545 /leaky where that makes sense (as before) or /readonly where required.
547 The /noupdate option is now undocumented but still supported for
548 backwards compatibility. It is equivalent to /readonly except that in
549 ACLs where /readonly is required you may specify /leaky/noupdate or
550 /strict/noupdate which are treated the same as /readonly.
552 A useful new feature is the /count= option. This is a generalization
553 of the per_byte option, so that you can measure the throughput of other
554 aggregate values. For example, the per_byte option is now equivalent
555 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
557 The per_rcpt option has been generalized using the /count= mechanism
558 (though it's more complicated than the per_byte equivalence). When it is
559 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
560 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
561 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
562 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
563 Note that using per_rcpt with a non-readonly update mode in more than
564 one ACL will cause the recipients to be double-counted. (The per_mail
565 and per_byte options don't have this problem.)
567 The handling of very low rates has changed slightly. If the computed rate
568 is less than the event's count (usually one) then this event is the first
569 after a long gap. In this case the rate is set to the same as this event's
570 count, so that the first message of a spam run is counted properly.
572 The major new feature is a mechanism for counting the rate of unique
573 events. The new per_addr option counts the number of different
574 recipients that someone has sent messages to in the last time period. It
575 behaves like per_rcpt if all the recipient addresses are different, but
576 duplicate recipient addresses do not increase the measured rate. Like
577 the /count= option this is a general mechanism, so the per_addr option
578 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
579 example, measure the rate that a client uses different sender addresses
580 with the options per_mail/unique=$sender_address. There are further
581 details in the main documentation.
583 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
585 TF/06 Removed a few PCRE remnants.
587 TF/07 Automatically extract Exim's version number from tags in the git
588 repository when doing development or release builds.
590 PP/02 Raise smtp_cmd_buffer_size to 16kB.
591 Bugzilla 879. Patch from Paul Fisher.
593 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
594 Heavily based on revision 40f9a89a from Simon Arlott's tree.
597 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
599 PP/05 Variable $av_failed, true if the AV scanner deferred.
600 Bugzilla 1078. Patch from John Horne.
602 PP/06 Stop make process more reliably on build failure.
603 Bugzilla 1087. Patch from Heiko Schlittermann.
605 PP/07 Make maildir_use_size_file an _expandable_ boolean.
606 Bugzilla 1089. Patch from Heiko Schlittermann.
608 PP/08 Handle ${run} returning more data than OS pipe buffer size.
609 Bugzilla 1131. Patch from Holger Weiß.
611 PP/09 Handle IPv6 addresses with SPF.
612 Bugzilla 860. Patch from Wolfgang Breyha.
614 PP/10 GnuTLS: support TLS 1.2 & 1.1.
616 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
619 PP/11 match_* no longer expand right-hand-side by default.
620 New compile-time build option, EXPAND_LISTMATCH_RHS.
621 New expansion conditions, "inlist", "inlisti".
623 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
625 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
627 PP/14 fix log_write() format string regression from TF/03.
628 Bugzilla 1152. Patch from Dmitry Isaikin.
634 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
636 PP/02 Harmonised TLS library version reporting; only show if debugging.
637 Layout now matches that introduced for other libraries in 4.74 PP/03.
639 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
641 PP/04 New "dns_use_edns0" global option.
643 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
646 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
647 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
649 TK/01 Updated PolarSSL code to 0.14.2.
650 Bugzilla 1097. Patch from Andreas Metzler.
652 PP/07 Catch divide-by-zero in ${eval:...}.
655 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
658 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
659 format-string attack -- SECURITY: remote arbitrary code execution.
661 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
662 time unintentionally subject to list matching rules, letting the header
663 cause arbitrary Exim lookups (of items which can occur in lists, *not*
664 arbitrary string expansion). This allowed for information disclosure.
666 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
667 INT_MIN/-1 -- value coerced to INT_MAX.
673 NM/01 Workround for PCRE version dependency in version reporting
676 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
677 This fixes portability to compilers other than gcc, notably
678 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
680 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
681 makefiles for portability to HP-UX and POSIX correctness.
683 PP/01 Permit LOOKUP_foo enabling on the make command-line.
684 Also via indented variable definition in the Makefile.
685 (Debugging by Oliver Heesakkers).
687 PP/02 Restore caching of spamd results with expanded spamd_address.
688 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
690 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
691 Improves build reliability. Fix from: Frank Elsner
693 NM/02 Fix wide character breakage in the rfc2047 coding
694 Fixes bug 1064. Patch from Andrey N. Oktyabrski
696 NM/03 Allow underscore in dnslist lookups
697 Fixes bug 1026. Patch from Graeme Fowler
699 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
700 Code patches from Adam Ciarcinski of NetBSD.
702 NM/04 Fixed exiqgrep to cope with mailq missing size issue
705 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
706 is logged, to avoid truncation. Patch from John Horne.
708 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
709 Patch from Jakob Hirsch.
711 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
712 SQL string expansion failure details.
713 Patch from Andrey Oktyabrski.
715 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
716 Patch from Simon Arlott.
718 PP/09 New lookups functionality failed to compile on old gcc which rejects
719 extern declarations in function scope.
720 Patch from Oliver Fleischmann
722 PP/10 Use sig_atomic_t for flags set from signal handlers.
723 Check getgroups() return and improve debugging.
724 Fixed developed for diagnosis in bug 927 (which turned out to be
727 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
728 Patch from Mark Zealey.
730 PP/12 Bugzilla 1056: Improved spamd server selection.
731 Patch from Mark Zealey.
733 PP/13 Bugzilla 1086: Deal with maildir quota file races.
734 Based on patch from Heiko Schlittermann.
736 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
737 Patch from Uwe Doering, sign-off by Michael Haardt.
739 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
740 variable declaration deep within a block. Bug and patch from
743 PP/15 lookups-Makefile IRIX compatibilty coercion.
745 PP/16 Make DISABLE_DKIM build knob functional.
747 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
748 Patch by Simon Arlott
750 TF/03 Fix valgrind.h portability to C89 compilers that do not support
751 variable argument macros. Our copy now differs from upstream.
757 TF/01 Failure to get a lock on a hints database can have serious
758 consequences so log it to the panic log.
760 TF/02 Log LMTP confirmation messages in the same way as SMTP,
761 controlled using the smtp_confirmation log selector.
763 TF/03 Include the error message when we fail to unlink a spool file.
765 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
766 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
767 for maintaining out-of-tree patches for some time.
769 PP/01 Bugzilla 139: Documentation and portability issues.
770 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
771 Handle per-OS dynamic-module compilation flags.
773 PP/02 Let /dev/null have normal permissions.
774 The 4.73 fixes were a little too stringent and complained about the
775 permissions on /dev/null. Exempt it from some checks.
776 Reported by Andreas M. Kirchwitz.
778 PP/03 Report version information for many libraries, including
779 Exim version information for dynamically loaded libraries. Created
780 version.h, now support a version extension string for distributors
781 who patch heavily. Dynamic module ABI change.
783 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
784 privilege escalation vulnerability whereby the Exim run-time user
785 can cause root to append content of the attacker's choosing to
788 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
791 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
792 If dropping privileges for untrusted macros, we disabled normal logging
793 on the basis that it would fail; for the Exim run-time user, this is not
794 the case, and it resulted in successful deliveries going unlogged.
795 Fixed. Reported by Andreas Metzler.
801 PP/01 Date: & Message-Id: revert to normally being appended to a message,
802 only prepend for the Resent-* case. Fixes regression introduced in
803 Exim 4.70 by NM/22 for Bugzilla 607.
805 PP/02 Include check_rfc2047_length in configure.default because we're seeing
806 increasing numbers of administrators be bitten by this.
808 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
810 PP/03 Bugzilla 994: added openssl_options main configuration option.
812 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
814 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
816 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
818 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
819 without a peer certificate, leading to a segfault because of an
820 assumption that peers always have certificates. Be a little more
821 paranoid. Problem reported by Martin Tscholak.
823 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
824 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
825 NB: ClamAV planning to remove STREAM in "middle of 2010".
826 CL also introduces -bmalware, various -d+acl logging additions and
827 more caution in buffer sizes.
829 PP/09 Implemented reverse_ip expansion operator.
831 PP/10 Bugzilla 937: provide a "debug" ACL control.
833 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
835 PP/12 Bugzilla 973: Implement --version.
837 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
839 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
841 PP/15 Bugzilla 816: support multiple condition rules on Routers.
843 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
844 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
845 ignore trailing whitespace.
847 JJ/02 prevent non-panic DKIM error from being sent to paniclog
849 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
852 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
853 Notification from Dr Andrew Aitchison.
855 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
856 ExtendedDetectionInfo response format.
857 Notification from John Horne.
859 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
862 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
863 XSL and documented dependency on system catalogs, with examples of how
866 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
869 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
870 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
871 configuration file which is writeable by the Exim user or group.
873 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
874 of configuration files to cover files specified with the -C option if
875 they are going to be used with root privileges, not just the default
878 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
879 option (effectively making it always true).
881 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
882 files to be used while preserving root privileges.
884 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
885 that rogue child processes cannot use them.
887 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
888 run-time user, instead of root.
890 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
891 Exim run-time user without dropping privileges.
893 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
894 result string, instead of calling string_vformat() twice with the same
897 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
898 for other users. Others should always drop root privileges if they use
899 -C on the command line, even for a whitelisted configure file.
901 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
903 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
909 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
910 $data_path, and $header_path variables; fixed documentation bugs and
913 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
914 exipick to access non-standard spools, including the "frozen" queue
917 NM/01 Bugzilla 965: Support mysql stored procedures.
918 Patch from Alain Williams
920 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
922 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
923 Patch from Andreas Metzler
925 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
926 Patch from Kirill Miazine
928 NM/05 Bugzilla 671: Added umask to procmail example.
930 JJ/03 installed exipick 20100323.0, fixing doc bug
932 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
933 directory. Notification and patch from Dan Rosenberg.
935 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
937 TK/02 Improve log output when DKIM signing operation fails.
939 MH/01 Treat the transport option dkim_domain as a colon separated
940 list, not as a single string, and sign the message with each element,
941 omitting multiple occurences of the same signer.
943 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
944 Bugzilla 985, 986. Patch by Simon Arlott
946 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
947 Patch by Simon Arlott
949 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
950 MBX locking. Notification from Dan Rosenberg.
956 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
958 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
960 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
962 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
964 NM/04 Bugzilla 915: Flag broken perl installation during build.
970 TK/01 Added patch by Johannes Berg that expands the main option
971 "spamd_address" if it starts with a dollar sign.
973 TK/02 Write list of recipients to X-Envelope-Sender header when building
974 the mbox-format spool file for content scanning (suggested by Jakob
977 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
978 (http://www.dcc-servers.net/) support via dccifd. Activated by
979 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
981 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
982 by Mark Daniel Reidel <mr@df.eu>.
984 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
985 When building exim an external PCRE library is now needed -
986 PCRE is a system library on the majority of modern systems.
987 See entry on PCRE_LIBS in EDITME file.
989 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
990 conversation. Added nologin parameter to request.
991 Patch contributed by Kirill Miazine.
993 TF/01 Do not log submission mode rewrites if they do not change the address.
995 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
997 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
998 log files in place. Contributed by Roberto Lima.
1000 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
1002 TF/03 Bugzilla 615: When checking the local_parts router precondition
1003 after a local_part_suffix or local_part_prefix option, Exim now
1004 does not use the address's named list lookup cache, since this
1005 contains cached lookups for the whole local part.
1007 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
1008 Robert Millan. Documentation is in experimental-spec.txt.
1010 TF/04 Bugzilla 668: Fix parallel build (make -j).
1012 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
1014 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
1015 Patch provided by Jan Srzednicki.
1017 TF/05 Leading white space used to be stripped from $spam_report which
1018 wrecked the formatting. Now it is preserved.
1020 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
1021 that they are available at delivery time.
1023 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
1025 TF/08 TLS error reporting now respects the incoming_interface and
1026 incoming_port log selectors.
1028 TF/09 Produce a more useful error message if an SMTP transport's hosts
1029 setting expands to an empty string.
1031 NM/06 Bugzilla 744: EXPN did not work under TLS.
1032 Patch provided by Phil Pennock.
1034 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
1035 Patch provided by Richard Godbee.
1037 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
1038 acl_smtp_notquit, added index entry.
1040 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
1041 Patch provided by Eugene Bujak.
1043 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
1044 accept(). Patch provided by Maxim Dounin.
1046 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
1047 Patch provided by Phil Pennock.
1049 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
1051 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
1052 Patch provided by Brad "anomie" Jorsch.
1054 NM/14 Bugzilla 622: Added timeout setting to transport filter.
1055 Patch provided by Dean Brooks.
1057 TK/05 Add native DKIM support (does not depend on external libraries).
1059 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
1060 Patch provided by Graeme Fowler.
1062 NM/16 Bugzilla 851: Documentation example syntax fix.
1064 NM/17 Changed NOTICE file to remove references to embedded PCRE.
1066 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
1069 NM/19 Bugzilla 745: TLS version reporting.
1070 Patch provided by Phil Pennock.
1072 NM/20 Bugzilla 167: bool: condition support.
1073 Patch provided by Phil Pennock.
1075 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
1076 clients. Patch provided by Phil Pennock.
1078 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
1079 Patch provided by Brad "anomie" Jorsch.
1081 NM/23 Bugzilla 687: Fix misparses in eximstats.
1082 Patch provided by Heiko Schlittermann.
1084 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
1085 Patch provided by Heiko Schlittermann.
1087 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
1088 plus update to original patch.
1090 NM/26 Bugzilla 799: Documentation correction for ratelimit.
1092 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
1093 Patch provided by David Brownlee.
1095 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
1097 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
1099 NM/30 Bugzilla 888: TLS documentation bugfixes.
1101 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
1103 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
1104 Unlike the original bugzilla I have changed all shell scripts in src tree.
1106 NM/33 Bugzilla 898: Transport filter timeout fix.
1107 Patch by Todd Rinaldo.
1109 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
1110 Patch by Serge Demonchaux.
1112 NM/35 Bugzilla 39: Base64 decode bug fixes.
1113 Patch by Jakob Hirsch.
1115 NM/36 Bugzilla 909: Correct connect() call in dcc code.
1117 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
1119 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
1121 NM/39 Bugzilla 911: Fixed MakeLinks build script.
1127 TK/01 Add preliminary DKIM support. Currently requires a forked version of
1128 ALT-N's libdkim that I have put here:
1129 http://duncanthrax.net/exim-experimental/
1131 Note to Michael Haardt: I had to rename some vars in sieve.c. They
1132 were called 'true' and it seems that C99 defines that as a reserved
1133 keyword to be used with 'bool' variable types. That means you could
1134 not include C99-style headers which use bools without triggering
1135 build errors in sieve.c.
1137 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
1138 as mailq or other aliases. Changed the --help handling significantly
1139 to do whats expected. exim_usage() emits usage/help information.
1141 SC/01 Added the -bylocaldomain option to eximstats.
1143 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1145 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1147 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1153 PH/01 Another patch from the Sieve maintainer.
1155 PH/02 When an IPv6 address is converted to a string for single-key lookup
1156 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1157 dots are used instead of colons so that keys in lsearch files need not
1158 contain colons. This was done some time before quoting was made available
1159 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1160 (notated using the quote facility) so as to distinguish them from IPv4
1161 keys. This meant that lookups for IP addresses in host lists did not work
1162 for iplsearch lookups.
1164 This has been fixed by arranging for IPv6 addresses to be expressed with
1165 colons if the lookup type is iplsearch. This is not incompatible, because
1166 previously such lookups could never work.
1168 The situation is now rather anomolous, since one *can* have colons in
1169 ordinary lsearch keys. However, making the change in all cases is
1170 incompatible and would probably break a number of configurations.
1172 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1175 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1176 conversion specification without a maximum field width, thereby enabling
1177 a rogue spamd server to cause a buffer overflow. While nobody in their
1178 right mind would setup Exim to query an untrusted spamd server, an
1179 attacker that gains access to a server running spamd could potentially
1180 exploit this vulnerability to run arbitrary code as the Exim user.
1182 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1183 $primary_hostname instead of what libspf2 thinks the hosts name is.
1185 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1186 a directory entry by the name of the lookup key. Previously, if a
1187 symlink pointed to a non-existing file or a file in a directory that
1188 Exim lacked permissions to read, a lookup for a key matching that
1189 symlink would fail. Now it is enough that a matching directory entry
1190 exists, symlink or not. (Bugzilla 503.)
1192 PH/03 The body_linecount and body_zerocount variables are now exported in the
1195 PH/04 Added the $dnslist_matched variable.
1197 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1198 This means they are set thereafter only if the connection becomes
1201 PH/06 Added the client_condition to authenticators so that some can be skipped
1202 by clients under certain conditions.
1204 PH/07 The error message for a badly-placed control=no_multiline_responses left
1205 "_responses" off the end of the name.
1207 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1209 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1210 (without spaces) instead of just copying the configuration text.
1212 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1214 PH/11 Added $max_received_linelength.
1216 PH/12 Added +ignore_defer and +include_defer to host lists.
1218 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1219 way in which PCRE > 7.0 is built.
1221 PH/14 Implemented queue_only_load_latch.
1223 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1224 MAIL command. The effect was to mangle the value on 64-bit systems.
1226 PH/16 Another patch from the Sieve maintainer.
1228 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1230 PH/18 If a system quota error occurred while trying to create the file for
1231 a maildir delivery, the message "Mailbox is full" was not appended to the
1232 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1233 only to a quota excession during the actual writing of the file.
1235 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1236 characters?) which causes problems in log lines. The DN values are now
1237 passed through string_printing() before being added to log lines.
1239 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1240 and InterBase are left for another time.)
1242 PH/21 Added message_body_newlines option.
1244 PH/22 Guard against possible overflow in moan_check_errorcopy().
1246 PH/23 POSIX allows open() to be a macro; guard against that.
1248 PH/24 If the recipient of an error message contained an @ in the local part
1249 (suitably quoted, of course), incorrect values were put in $domain and
1250 $local_part during the evaluation of errors_copy.
1256 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1257 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1260 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1261 issue a MAIL command.
1263 PH/02 In an ACL statement such as
1265 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1267 if a client was not listed at all, or was listed with a value other than
1268 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1269 the condition was not true (as it should be), so access was not denied.
1270 The bug was that the ! inversion was incorrectly passed on to the second
1271 item. This has been fixed.
1273 PH/03 Added additional dnslists conditions == and =& which are different from
1274 = and & when the dns lookup returns more than one IP address.
1276 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1277 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1279 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1280 FSYNC, which compiles an option called disable_fsync that allows for
1281 bypassing fsync(). The documentation is heavily laced with warnings.
1283 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1285 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1286 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1287 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1288 including adding "make clean"; (3) Added -fPIC when compiling the test
1289 dynamically loaded module, to get rid of a warning.
1291 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1292 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1293 = 0s. The bug is otherwise harmless.
1295 PH/07 There was a bug in the dovecot authenticator such that the value of
1296 $auth1 could be overwritten, and so not correctly preserved, after a
1297 successful authentication. This usually meant that the value preserved by
1298 the server_setid option was incorrect.
1300 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1302 PH/09 Installed PCRE release 7.0.
1304 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1305 run for batched SMTP input. It is now run at the start of every message
1306 in the batch. While fixing this I discovered that the process information
1307 (output by running exiwhat) was not always getting set for -bs and -bS
1308 input. This is fixed, and it now also says "batched" for BSMTP.
1310 PH/11 Added control=no_pipelining.
1312 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1313 patch, slightly modified), and move the expansion of helo_data till after
1314 the connection is made in the smtp transport (so it can use these
1317 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1319 PH/14 Added log_selector = +pid.
1321 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1323 PH/16 Add ${if forany and ${if forall.
1325 PH/17 Added dsn_from option to vary the From: line in DSNs.
1327 PH/18 Flush SMTP output before performing a callout, unless control =
1328 no_callout_flush is set.
1330 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1331 was true (the default) a successful delivery failed to delete the retry
1332 item, thus causing premature timeout of the address. The bug is now
1335 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1337 PH/21 Long custom messages for fakedefer and fakereject are now split up
1338 into multiline reponses in the same way that messages for "deny" and
1339 other ACL rejections are.
1341 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1342 with slight modification.
1344 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1345 draft, changing the syntax and factoring some duplicate code".
1347 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1348 for deliveries of the second and subsequent messages over the same SMTP
1351 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1352 ${reduce, with only minor "tidies".
1354 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1356 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1357 expansion side effects.
1359 PH/27 When a message times out after an over-quota error from an Exim-imposed
1360 quota, the bounce message says "mailbox is full". This message was not
1361 being given when it was a system quota that was exceeded. It now should
1364 MH/03 Made $recipients available in local_scan(). local_scan() already has
1365 better access to the recipient list through recipients_list[], but
1366 $recipients can be useful in postmaster-provided expansion strings.
1368 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1369 in the case of a MAIL command with additional options following the
1370 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1371 were accidentally chopped off.
1373 PH/29 SMTP synchronization checks are implemented when a command is read -
1374 there is a check that no more input is waiting when there shouldn't be
1375 any. However, for some commands, a delay in an ACL can mean that it is
1376 some time before the response is written. In this time, more input might
1377 arrive, invalidly. So now there are extra checks after an ACL has run for
1378 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1379 pipelining has not been advertised.
1381 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1383 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1384 returned in the SMTP response when smtp_return_error_details was false.
1385 This has been fixed.
1387 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1388 instead of the C I/O that was originally supplied, because problems were
1389 reported on Solaris.
1391 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1392 Exim which did not show up earlier: it was assuming that a call to
1393 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1394 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1395 was a macro that became an assignment, so it seemed to work. This has
1396 changed to a proper function call with a void return, hence the compile
1397 error. Exim's code has been fixed.
1399 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1402 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1404 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1406 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1407 because it turns out to be too big or there is a timeout) there is no
1408 "Completed" line in the log. When some messages of this type were
1409 selected by exigrep, they were listed as "not completed". Others were
1410 picked up by some special patterns. I have improved the selection
1411 criteria to be more general.
1413 PH/38 The host_find_failed option in the manualroute router can now be set
1414 to "ignore", to completely ignore a host whose IP address cannot be
1415 found. If all hosts are ignored, the behaviour is controlled by the new
1416 host_all_ignored option.
1418 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1419 homing or because of multiple MX records with /mx) generated more than
1420 one IP address, and the following item turned out to be the local host,
1421 all the secondary addresses of the first item were incorrectly removed
1422 from the list, along with the local host and any following hosts (which
1423 is what is supposed to happen).
1425 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1426 whoever called Exim into the -H file. In the case of the daemon it was
1427 behaving confusingly. When first started, it used values for whoever
1428 started the daemon, but after a SIGHUP it used the Exim user (because it
1429 calls itself on a restart). I have changed the code so that it now always
1432 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1433 message are rejected with the same error (e.g. no authentication or bad
1434 sender address), and a DATA command is nevertheless sent (as can happen
1435 with PIPELINING or a stupid MUA), the error message that was given to the
1436 RCPT commands is included in the rejection of the DATA command. This is
1437 intended to be helpful for MUAs that show only the final error to their
1440 PH/42 Another patch from the Sieve maintainer.
1442 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1443 Eximstats - Fixed some broken HTML links and added missing column headers
1445 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1446 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1448 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1450 PH/43 Yet another patch from the Sieve maintainer.
1452 PH/44 I found a way to check for a TCP/IP connection going away before sending
1453 the response to the final '.' that terminates a message, but only in the
1454 case where the client has not sent further data following the '.'
1455 (unfortunately, this is allowed). However, in many cases there won't be
1456 any further data because there won't be any more messages to send. A call
1457 to select() can be used: if it shows that the input is "ready", there is
1458 either input waiting, or the socket has been closed. An attempt to read
1459 the next input character can distinguish the two cases. Previously, Exim
1460 would have sent an OK response which the client would never have see.
1461 This could lead to message repetition. This fix should cure that, at
1462 least in a lot of common cases.
1464 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1465 advertised in response to EHLO.
1471 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1472 fixed by 4.65/MH/01 (is this a record?) are fixed:
1474 (i) An empty string was always treated as zero by the numeric comparison
1475 operators. This behaviour has been restored.
1477 (ii) It is documented that the numeric comparison operators always treat
1478 their arguments as decimal numbers. This was broken in that numbers
1479 starting with 0 were being interpreted as octal.
1481 While fixing these problems I realized that there was another issue that
1482 hadn't been noticed. Values of message_size_limit (both the global option
1483 and the transport option) were treated as octal if they started with 0.
1484 The documentation was vague. These values are now always treated as
1485 decimal, and I will make that clear in the documentation.
1491 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1492 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1495 MH/01 Don't check that the operands of numeric comparison operators are
1496 integers when their expansion is in "skipping" mode (fixes bug
1497 introduced by 4.64-PH/07).
1499 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1500 child addresses, Exim now panics and dies. Previously, because the count
1501 is held in a short int, deliveries were likely to be lost. As such a
1502 large number of recipients for a single message is ridiculous
1503 (performance will be very, very poor), I have chosen to impose a limit
1504 rather than extend the field.
1510 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1511 leftover -K file (the existence of which was triggered by #402).
1512 While we were at it, introduced process PID as part of the -K
1513 filename. This should rule out race conditions when creating
1516 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1517 processing considerably. Previous code took too long for large mails,
1518 triggering a timeout which in turn triggers #401.
1520 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1521 in the DK code in transports.c. sendfile() is not really portable,
1522 hence the _LINUX specificness.
1524 TF/01 In the add_headers option to the mail command in an Exim filter,
1525 there was a bug that Exim would claim a syntax error in any
1526 header after the first one which had an odd number of characters
1529 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1530 callout verification, Exim cached a "reject" for the entire domain. This
1531 is correct for most verifications, but it is not correct for a recipient
1532 verification with use_sender or use_postmaster set, because in that case
1533 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1534 case of MAIL FROM:<> rejection from other early rejections (e.g.
1535 rejection of HELO). When verifying a recipient using a non-null MAIL
1536 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1537 Whatever the result of the callout, the value of the domain cache is
1538 left unchanged (for any other kind of callout, getting as far as trying
1539 RCPT means that the domain itself is ok).
1541 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1544 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1545 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1546 session, the daemon ignores ECONNECT errors and logs others; it now
1547 ignores EPIPE as well.
1549 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1550 (quoted-printable decoding).
1552 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1553 later the small subsequent patch to fix an introduced bug.
1555 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1557 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1559 PH/08 An error is now given if message_size_limit is specified negative.
1561 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1562 to be given (somewhat) arbitrary names.
1564 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1567 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1568 miscellaneous code fixes
1570 PH/10 Added the log_reject_target ACL modifier to specify where to log
1573 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1574 hostname. This is wrong, because it relates to the incoming message (and
1575 probably the interface on which it is arriving) and not to the outgoing
1576 callout (which could be using a different interface). This has been
1577 changed to use the value of the helo_data option from the smtp transport
1578 instead - this is what is used when a message is actually being sent. If
1579 there is no remote transport (possible with a router that sets up host
1580 addresses), $smtp_active_hostname is used.
1582 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1583 tweaks were necessary in order to get it to work (see also 21 below):
1584 (a) The code assumed that strncpy() returns a negative number on buffer
1585 overflow, which isn't the case. Replaced with Exim's string_format()
1587 (b) There were several signed/unsigned issues. I just did the minimum
1588 hacking in of casts. There is scope for a larger refactoring.
1589 (c) The code used strcasecmp() which is not a standard C function.
1590 Replaced with Exim's strcmpic() function.
1591 (d) The code set only $1; it now sets $auth1 as well.
1592 (e) A simple test gave the error "authentication client didn't specify
1593 service in request". It would seem that Dovecot has changed its
1594 interface. Fortunately there's a specification; I followed it and
1595 changed what the client sends and it appears to be working now.
1597 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1600 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1601 address A is aliased to B and C, where B exists and C does not. Without
1602 -v the output is "A verified" because verification stops after a
1603 successful redirection if more than one address is generated. However,
1604 with -v the child addresses are also verified. Exim was outputting "A
1605 failed to verify" and then showing the successful verification for C,
1606 with its parentage. It now outputs "B failed to verify", showing B's
1607 parentage before showing the successful verification of C.
1609 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1610 look up a TXT record in a specific list after matching in a combined
1613 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1614 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1615 they consult the DNS. I had assumed they would set it the way they
1616 wanted; and indeed my experiments on Linux seem to show that in some
1617 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1618 To be on the safe side, however, I have now made the interface to
1619 host_find_byname() similar to host_find_bydns(), with an argument
1620 containing the DNS resolver options. The host_find_byname() function now
1621 sets these options at its start, just as host_find_bydns() does. The smtp
1622 transport options dns_qualify_single and dns_search_parents are passed to
1623 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1624 of host_find_byname() use the default settings of RES_DEFNAMES
1625 (qualify_single) but not RES_DNSRCH (search_parents).
1627 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1628 spool_read_header() do less string testing, by means of a preliminary
1629 switch on the second character of optional "-foo" lines. (This is
1630 overdue, caused by the large number of possibilities that now exist.
1631 Originally there were few.) While I was there, I also converted the
1632 str(n)cmp tests so they don't re-test the leading "-" and the first
1633 character, in the hope this might squeeze out yet more improvement.
1635 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1636 flag allowing group syntax was set by the header_syntax check but not
1637 turned off, possible causing trouble later; (2) The flag was not being
1638 set at all for the header_verify test, causing "group"-style headers to
1639 be rejected. I have now set it in this case, and also caused header_
1640 verify to ignore an empty address taken from a group. While doing this, I
1641 came across some other cases where the code for allowing group syntax
1642 while scanning a header line wasn't quite right (mostly, not resetting
1643 the flag correctly in the right place). These bugs could have caused
1644 trouble for malformed header lines. I hope it is now all correct.
1646 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1647 with the "reply" argument non-NULL. The code, however (which originally
1648 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1649 but it didn't always do it. This confused somebody who was copying the
1650 code for some other use. I have removed all the tests.
1652 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1653 feature that was used to support insecure browsers during the U.S. crypto
1654 embargo. It requires special client support, and Exim is probably the
1655 only MTA that supported it -- and would never use it because real RSA is
1656 always available. This code has been removed, because it had the bad
1657 effect of slowing Exim down by computing (never used) parameters for the
1658 RSA_EXPORT functionality.
1660 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1661 authenticator to fail if there's a tab character in the incoming data
1662 (there should never be unless someone is messing about, as it's supposed
1663 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1664 if the connection is using TLS or if the remote IP is the same as the
1665 local IP, and the "valid-client-cert option" if a client certificate has
1668 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1669 authenticators. This can be used for authorization after authentication
1670 succeeds. (In the case of plaintext, it servers for both authentication
1673 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1674 if any retry times were supplied.
1676 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1677 connection, where there is no client IP address to check. In this
1678 situation, the verify now always succeeds.
1680 PH/25 Applied John Jetmore's -Mset patch.
1682 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1684 PH/27 In a string expansion for a processed (not raw) header when multiple
1685 headers of the same name were present, leading whitespace was being
1686 removed from all of them, but trailing whitespace was being removed only
1687 from the last one. Now trailing whitespace is removed from each header
1688 before concatenation. Completely empty headers in a concatenation (as
1689 before) are ignored.
1691 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1692 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1694 PH/29 [Removed. This was a change that I later backed out, and forgot to
1695 correct the ChangeLog entry (that I had efficiently created) before
1696 committing the later change.]
1698 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1699 address errors (4xx response to RCPT) over the same connection as other
1700 messages routed to the same hosts. Such deliveries are always "forced",
1701 so retry times are not inspected. This resulted in far too many retries
1702 for the affected addresses. The effect occurred only when there were more
1703 hosts than the hosts_max_try setting in the smtp transport when it had
1704 the 4xx errors. Those hosts that it had tried were not added to the list
1705 of hosts for which the message was waiting, so if all were tried, there
1706 was no problem. Two fixes have been applied:
1708 (i) If there are any address or message errors in an SMTP delivery, none
1709 of the hosts (tried or untried) are now added to the list of hosts
1710 for which the message is waiting, so the message should not be a
1711 candidate for sending over the same connection that was used for a
1712 successful delivery of some other message. This seems entirely
1713 reasonable: after all the message is NOT "waiting for some host".
1714 This is so "obvious" that I'm not sure why it wasn't done
1715 previously. Hope I haven't missed anything, but it can't do any
1716 harm, as the worst effect is to miss an optimization.
1718 (ii) If, despite (i), such a delivery is accidentally attempted, the
1719 routing retry time is respected, so at least it doesn't keep
1720 hammering the server.
1722 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1723 in ${readsocket because some servers need this prod.
1725 PH/32 Added some extra debug output when updating a wait-xxx database.
1727 PH/33 The hint "could be header name not terminated by colon", which has been
1728 given for certain expansion errors for a long time, was not being given
1729 for the ${if def:h_colon_omitted{... case.
1731 PH/34 The spec says: "With one important exception, whenever a domain list is
1732 being scanned, $domain contains the subject domain." There was at least
1733 one case where this was not true.
1735 PH/35 The error "getsockname() failed: connection reset by peer" was being
1736 written to the panic log as well as the main log, but it isn't really
1737 panic-worthy as it just means the connection died rather early on. I have
1738 removed the panic log writing for the ECONNRESET error when getsockname()
1741 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1742 runs only) independently of the message's sender address. This meant
1743 that, if the 4xx error was in fact related to the sender, a different
1744 message to the same recipient with a different sender could confuse
1745 things. In particualar, this can happen when sending to a greylisting
1746 server, but other circumstances could also provoke similar problems.
1747 I have changed the default so that the retry time for these errors is now
1748 based a combination of the sender and recipient addresses. This change
1749 can be overridden by setting address_retry_include_sender=false in the
1752 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1753 remote server are returned as part of bounce messages. This was not
1754 happening for LMTP over a pipe (the lmtp transport), but now it is the
1755 same for both kinds of LMTP.
1757 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1758 in header lines that were in fact CNAMEs. This is no longer the case.
1760 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1761 and queue runs started by the daemon processed all messages. This has
1762 been fixed so that -R and -S can now usefully be given with -q<time>.
1764 PH/40 Import PCRE release 6.7 (fixes some bugs).
1766 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1768 PH/42 Give an error if -q is specified more than once.
1770 PH/43 Renamed the variables $interface_address and $interface_port as
1771 $received_ip_address and $received_port, to make it clear that these
1772 values apply to message reception, and not to the outgoing interface when
1773 a message is delivered. (The old names remain recognized, of course.)
1775 PH/44 There was no timeout on the connect() call when using a Unix domain
1776 socket in the ${readsocket expansion. There now is.
1778 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1779 be meaningful with "accept".
1781 SC/01 Eximstats V1.43
1782 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1784 SC/02 Eximstats V1.44
1785 Use a glob alias rather than an array ref in the generated
1786 parser. This improves both readability and performance.
1788 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1789 Collect SpamAssassin and rejection statistics.
1790 Don't display local sender or destination tables unless
1791 there is data to show.
1792 Added average volumes into the top table text output.
1794 SC/04 Eximstats V1.46
1795 Collect data on the number of addresses (recipients)
1796 as well as the number of messages.
1798 SC/05 Eximstats V1.47
1799 Added 'Message too big' to the list of mail rejection
1800 reasons (thanks to Marco Gaiarin).
1802 SC/06 Eximstats V1.48
1803 Mainlog lines which have GMT offsets and are too short to
1804 have a flag are now skipped.
1806 SC/07 Eximstats V1.49 (Alain Williams)
1807 Added the -emptyok flag.
1809 SC/08 Eximstats V1.50
1810 Fixes for obtaining the IP address from reject messages.
1812 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1813 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1814 whitesspace changes from 4.64-PH/27
1816 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1819 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1820 are found, allow negative numbers in numeric criteria)
1822 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1824 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1825 to match changes made in 4.64-PH/43
1827 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1829 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1830 "domain" is actually an IP address, and force a failure. This locks out
1831 those revolvers/nameservers that support "A-for-A" lookups, in
1832 contravention of the specifications.
1834 PH/48 When a host name was looked up from an IP address, and the subsequent
1835 forward lookup of the name timed out, the host name was left in
1836 $sender_host_name, contrary to the specification.
1838 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1839 restricted to single-key lookups, Exim was not diagnosing an error if
1840 * or *@ was used with a query-style lookup.
1842 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1844 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1845 long ago, but noone interested enough thought of it. Let's just say that
1846 the "1.1" means that there are some new functions that weren't there at
1847 some point in the past.
1849 PH/51 Error processing for expansion failure of helo_data from an smtp
1850 transport during callout processing was broken.
1852 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1853 tested/used via the -bh/-bhc/-bs options.
1855 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1856 bug, fixed in subsequent PCRE releases).
1858 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1859 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1861 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1867 SC/01 Use a glob alias rather than an array ref in eximstats generated
1868 parser. This improves both readability and performance.
1870 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1871 Don't display local sender or destination tables in eximstats unless
1872 there is data to show.
1873 Added average volumes into the eximstats top table text output.
1875 SC/03 Collect data on the number of addresses (recipients) as well
1876 as the number of messages in eximstats.
1878 TF/01 Correct an error in the documentation for the redirect router. Exim
1879 does (usually) call initgroups() when daemonizing.
1881 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1882 with consistent privilege compared to when running as a daemon.
1884 TF/03 Note in the spec that $authenticated_id is not set for local
1885 submissions from trusted users.
1887 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1888 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1890 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1891 by adding some example configuration directives to the default
1892 configuration file. A little bit of work is required to uncomment the
1893 directives and define how usernames and passwords are checked, but
1894 there is now a framework to start from.
1896 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1897 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1898 without this. I don't know how relevant this is to other LDAP libraries.
1900 PH/02 Add the verb name to the "unknown ACL verb" error.
1902 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1904 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1906 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1907 directory not expanded when it should be if an expanded home directory
1908 was set for the address (which is overridden by the transport).
1910 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1913 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1914 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1915 because it is too late at that time, and has no effect.
1917 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1918 security issue with \' (bugzilla #107). I could not use the
1919 PQescapeStringConn() function, because it needs a PGconn value as one of
1922 PH/08 When testing addresses using -bt, indicate those final addresses that
1923 are duplicates that would not cause an additional delivery. At least one
1924 person was confused, thinking that -bt output corresponded to deliveries.
1925 (Suppressing duplicates isn't a good idea as you lose the information
1926 about possibly different redirections that led to the duplicates.)
1928 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1929 systems where poll() doesn't work, in particular OS X.
1931 PH/10 Added more information to debugging output for retry time not reached.
1933 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1934 operations in malware.c.
1936 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1939 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1940 syslog_duplication set false, log lines that would normally be written
1941 both the the main log and to the reject log were not written to syslog at
1944 PH/14 In the default configuration, change the use of "message" in ACL warn
1945 statements to "add_header".
1947 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1948 not followed by a command (e.g. "seen endif").
1950 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1951 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1954 PH/17 Added extra conditions to the default value of delay_warning_condition
1958 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1959 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1960 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1963 The Auto-Submitted: and various List- headers are standardised, whereas I
1964 don't think Precedence: ever was.
1966 PH/18 Refactored debugging code in route_finduser() to show more information,
1967 in particular, the error code if getpwnam() issues one.
1969 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1970 This is apparently needed in addition to the PH/07 change above to avoid
1971 any possible encoding problems.
1973 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1974 but not after initializing Perl.
1976 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1977 output them only if debugging. By default they are written stderr,
1978 apparently, which is not desirable.
1980 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1983 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1986 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1988 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1989 authenticated or an ident call has been made. Suppress the default
1990 values for $authenticated_id and $authenticated_sender (but permit -oMai
1991 and -oMas) when testing with -bh.
1993 PH/24 Re-jigged the order of the tests in the default configuration so that the
1994 tests for valid domains and recipients precede the DNS black list and CSA
1995 tests, on the grounds that those ones are more expensive.
1997 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1998 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1999 This bug exists in every version of Exim that I still have, right back to
2002 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
2003 However, an attempt to turn on case-sensitivity in a regex key by
2004 including (?-i) didn't work because the subject string was already
2005 lowercased, and the effects were non-intuitive. It turns out that a
2006 one-line patch can be used to allow (?-i) to work as expected.
2012 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
2013 other effects) broke the use of negated acl sub-conditions.
2015 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
2018 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
2019 "Deny" causes Exim to reject the incoming connection with a 554 error.
2020 Unfortunately, if there is a major crisis, such as a disk failure,
2021 tcp-wrappers gives "deny", whereas what one would like would be some
2022 kind of temporary error. A kludge has been added to help with this.
2023 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
2024 554 error is used if errno is still zero or contains ENOENT (which occurs
2025 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
2028 PH/03 Add -lutil to the default FreeBSD LIBS setting.
2030 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
2031 errors. Otherwise a message that provokes a temporary error (when other
2032 messages do not) can cause a whole host to time out.
2034 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
2035 addresses were routed directly to files or pipes from a redirect router.
2036 File deliveries just didn't batch; pipe deliveries might have suffered
2039 PH/06 A failure to get a lock for a hints database would erroneously always say
2040 "Failed to get write lock", even when it was really a read lock.
2042 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
2043 of 0600. This has been changed to use the value of the lockfile_mode
2044 option (which defaults to 0600).
2046 PH/08 Applied small patch from the Sieve maintainer.
2048 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
2049 folder from quota calculations, a direct delivery into this folder messed
2050 up the contents of the maildirsize file. This was because the regex was
2051 used only to exclude .Trash (or whatever) when the size of the mailbox
2052 was calculated. There was no check that a delivery was happening into an
2053 excluded directory. This bug has been fixed by ignoring all quota
2054 processing for deliveries into excluded directories.
2056 PH/10 Added the maildirfolder_create_regex option to appendfile.
2062 PH/01 The code for finding all the local interface addresses on a FreeBSD
2063 system running IPv6 was broken. This may well have applied to all BSD
2064 systems, as well as to others that have similar system calls. The broken
2065 code found IPv4 interfaces correctly, but gave incorrect values for the
2066 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
2067 that it would not match correctly against @[] and not recognize the IPv6
2070 PH/02 The ipliteral router was not recognizing addresses of the form user@
2071 [ipv6:....] because it didn't know about the "ipv6:" prefix.
2073 PH/03 Added disable_ipv6.
2075 PH/04 Changed $reply_address to use the raw form of the headers instead of the
2076 decoded form, because it is most often used to construct To: headers
2077 lines in autoreplies, and the decoded form may well be syntactically
2078 invalid. However, $reply_address has leading white space removed, and all
2079 newlines turned into spaces so that the autoreply transport does not
2082 PH/05 If group was specified without a user on a router, and no group or user
2083 was specified on a transport, the group from the router was ignored.
2085 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
2086 for visible compile-time settings that can be used to change these
2087 numbers, for those that want even more. Backwards compatibility with old
2088 spool files has been maintained. However, going back to a previous Exim
2089 release will lost any variables that are in spool files.
2091 PH/07 Two small changes when running in the test harness: increase delay when
2092 passing a TCP/IP connection to a new process, in case the original
2093 process has to generate a bounce, and remove special handling of
2094 127.0.0.2 (sic), which is no longer necessary.
2096 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
2097 be the same on different OS.
2099 PH/09 Moved a debug statement in filter processing to avoid a race problem when
2102 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
2103 whether --show-vars was specified or not
2105 JJ/02 exipick: Added support for new ACL variable spool format introduced
2108 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
2109 syntactically invalid From: or Reply-to: line, and a filter used this to
2110 generate an autoreply, and therefore failed to obtain an address for the
2111 autoreply, Exim could try to deliver to a non-existent relative file
2112 name, causing unrelated and misleading errors. What now happens is that
2113 it logs this as a hard delivery error, but does not attempt to create a
2116 PH/11 The exinext utility has a -C option for testing purposes, but although
2117 the given file was scanned by exinext itself; it wasn't being passed on
2118 when Exim was called.
2120 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
2121 an end-of-file indication when reading a command response.
2123 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
2124 compiled. In many other places in Exim, IPv6 addresses are always
2125 recognized, so I have changed this. It also means that IPv4 domain
2126 literals of the form [IPV4:n.n.n.n] are now always recognized.
2128 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
2129 used if the router is not running as root, for example, when verifying at
2130 ACL time, or when using -bh. The debugging output from this situation was
2131 non-existent - all you got was a failure to exec. I have made two
2134 (a) Failures to set uid/gid, the current directory, or a process leader
2135 in a subprocess such as that created by queryprogram now generate
2136 suitable debugging ouput when -d is set.
2138 (b) The queryprogram router detects when it is not running as root,
2139 outputs suitable debugging information if -d is set, and then runs
2140 the subprocess without attempting to change uid/gid.
2142 PH/15 Minor change to Makefile for building test_host (undocumented testing
2145 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2146 additional section of a DNS packet that returns MX or SRV records.
2147 Instead, it always explicitly searches for A/AAAA records. This avoids
2148 major problems that occur when a DNS server includes only records of one
2149 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2150 fixed another bug: if SRV records were looked up and the corresponding
2151 address records were *not* found in the additional section, the port
2152 values from the SRV records were lost.
2154 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2155 using the correct key (the original address) when searching the retry
2156 rules in order to find which one to use for generating the retry hint.
2158 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2159 adding the default one. Similarly, if it contains a Reply-To: header, the
2160 errors_reply_to option, if set, is not used.
2162 PH/19 When calculating a retry time, Exim used to measure the "time since
2163 failure" by looking at the "first failed" field in the retry record. Now
2164 it does not use this if it is later than than the arrival time of the
2165 message. Instead it uses the arrival time. This makes for better
2166 behaviour in cases where some deliveries succeed, thus re-setting the
2167 "first failed" field. An example is a quota failure for a huge message
2168 when small messages continue to be delivered. Without this change, the
2169 "time since failure" will always be short, possible causing more frequent
2170 delivery attempts for the huge message than are intended.
2171 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2173 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2174 $1, $2, $3) because the numerical variables can be reset during some
2175 expansion items (e.g. "match"), thereby losing the authentication data.
2177 PH/21 Make -bV show the size of off_t variables so that the test suite can
2178 decide whether to run tests for quotas > 2G.
2180 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2181 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2182 filecount value is greater than 2G or if a quota value is greater than 2G
2183 on a system where the size of off_t is not greater than 4, a panic error
2186 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2187 never match. The debug and -bh output now contains an explicit error
2188 message indicating a malformed IPv4 address or mask.
2190 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2191 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2192 PH/23 above applies.
2194 PH/25 Do not write to syslog when running in the test harness. The only
2195 occasion when this arises is a failure to open the main or panic logs
2196 (for which there is an explicit test).
2198 PH/26 Added the /no_tell option to "control=freeze".
2200 PH/27 If a host name lookup failed very early in a connection, for example, if
2201 the IP address matched host_lookup and the reverse lookup yielded a name
2202 that did not have a forward lookup, an error message of the form "no IP
2203 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2204 could be logged. Now it outputs the IP address instead of "NULL".
2206 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2207 allows the sender and the authenticated sender to be set when
2208 submitting a message from within Exim. Since child_open_exim() is
2209 documented for local_scan(), the new function should be too.
2211 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2212 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2213 results in an empty string is now treated as unset.
2215 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2217 PH/31 Added sender_verify_fail logging option.
2219 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2220 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2221 tidied the source and removed it altogether.
2223 PH/33 When a queue run was abandoned because the load average was too high, a
2224 log line was always written; now it is written only if the queue_run log
2225 selector is set. In addition, the log line for abandonment now contains
2226 information about the queue run such as the pid. This is always present
2227 in "start" and "stop" lines but was omitted from the "abandon" line.
2229 PH/34 Omit spaces between a header name and the colon in the error message that
2230 is given when verify = headers_syntax fails (if there are lots of them,
2231 the message gets confusing).
2233 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2234 names, as there are now some PTR records that contain slashes. This check
2235 is only to protect against broken name servers that fall over on strange
2236 characters, so the fact that it applies to all lookups doesn't matter.
2238 PH/36 Now that the new test suite is complete, we can remove some of the
2239 special code in Exim that was needed for the old test suite. For example,
2240 sorting DNS records because real resolvers return them in an arbitrary
2241 order. The new test suite's fake resolver always returns records in the
2244 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2245 bounces) except when queue_only is set, to avoid logging races between
2246 the different processes.
2248 PH/38 Panic-die if .include specifies a non-absolute path.
2250 PH/39 A tweak to the "H" retry rule from its user.
2252 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2253 a label. They prevented compilation on older perls.
2255 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2256 a warning to be raised on newish perls.
2258 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2259 on queue. Changes to match documented behaviour of showing count of
2260 messages matching specified criteria.
2262 PH/40 Changed the default ident timeout from 30s to 5s.
2264 PH/41 Added support for the use of login_cap features, on those BSD systems
2265 that have them, for controlling the resources used by pipe deliveries.
2267 PH/42 The content-scanning code uses fopen() to create files in which to put
2268 message data. Previously it was not paying any attention to the mode of
2269 the files. Exim runs with umask(0) because the rest of the code creates
2270 files with open(), and sets the required mode explicitly. Thus, these
2271 files were ending up world-writeable. This was not a big issue, because,
2272 being within the spool directory, they were not world-accessible. I have
2273 created a function called modefopen, which takes an additional mode
2274 argument. It sets umask(777), creates the file, chmods it to the required
2275 mode, then resets the umask. All the relevant calls to fopen() in the
2276 content scanning code have been changed to use this function.
2278 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2279 to 24 hours. This avoids potential overflow problems when processing G
2280 and H retry rules. I suspect nobody ever tinkers with this value.
2282 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2284 PH/45 When the plaintext authenticator is running as a client, the server's
2285 challenges are checked to ensure they are valid base64 strings. By
2286 default, the authentication attempt is cancelled if an invalid string is
2287 received. Setting client_ignore_invalid_base64 true ignores these errors.
2288 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2289 they are received. Thus, the responses can be made to depend on the
2290 challenges. If an invalid string is ignored, an empty string is placed in
2293 PH/46 Messages that are created by the autoreply transport now contains a
2294 References: header, in accordance with RFCs 2822 and 3834.
2296 PH/47 Added authenticated_sender_force to the smtp transport.
2298 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2300 PH/49 Installed latest patch from the Sieve maintainer.
2302 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2303 was also set, the appendfile transport was unnecessarily scanning a
2304 directory of message files (e.g. for maildir delivery) to find the count
2305 of files (along with the size), even though it did not need this
2306 information. It now does the scan only if it needs to find either the
2307 size of the count of files.
2309 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2311 PH/52 Two bugs concerned with error handling when the smtp transport is
2314 (i) Exim was not creating retry information for temporary errors given
2315 for individual recipients after the DATA command when the smtp transport
2316 was used in LMTP mode. This meant that they could be retried too
2317 frequently, and not timed out correctly.
2319 (ii) Exim was setting the flag that allows error details to be returned
2320 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2321 recipients that were returned after the DATA command.
2323 PH/53 This is related to PH/52, but is more general: for any failing address,
2324 when detailed error information was permitted to be returned to the
2325 sender, but the error was temporary, then after the final timeout, only
2326 "retry timeout exceeded" was returned. Now it returns the full error as
2327 well as "retry timeout exceeded".
2329 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2330 do this, and (what is worse) MTAs that accept it.
2332 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2333 will now be deprecated.
2335 PH/56 New os.c-cygwin from the Cygwin maintainer.
2337 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2338 formats (previously only available in exim formats via -bpr, -bpru,
2339 and -bpra. Now also available in native and exiqgrep formats)
2341 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2342 with very large, slow to parse queues
2344 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2346 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2348 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2349 responses to authentication challenges, though it was showing the
2350 challenges; (ii) I've removed the CR characters from the debug output for
2353 PH/58 Allow for the insertion of a newline as well as a space when a string
2354 is turned into more than one encoded-word during RFC 2047 encoding. The
2355 Sieve code now uses this.
2357 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2358 data_4xx, lost_connection, tls_required.
2360 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2361 message was being sent as an SMTP response.
2363 PH/61 Add -l and -k options to exicyclog.
2365 PH/62 When verifying, if an address was redirected to one new address, so that
2366 verification continued, and the new address failed or deferred after
2367 having set something in $address_data, the value of $address_data was not
2368 passed back to the ACL. This was different to the case when no
2369 redirection occurred. The value is now passed back in both cases.
2371 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2372 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2373 use login_cap.h, so on its own it isn't the distinguishing feature. The
2374 new name refers directly to the setclassresources() function.
2376 PH/65 Added configuration files for NetBSD3.
2378 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2380 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2381 is preferred over IPv4.
2383 PH/68 The bounce_return_message and bounce_return_body options were not being
2384 honoured for bounces generated during the reception of non-SMTP messages.
2385 In particular, this applied to messages rejected by the ACL. This bug has
2386 been fixed. However, if bounce_return_message is true and bounce_return_
2387 body is false, the headers that are returned for a non-SMTP message
2388 include only those that have been read before the error was detected.
2389 (In the case of an ACL rejection, they have all been read.)
2391 PH/69 The HTML version of the specification is now built in a directory called
2392 spec_html instead of spec.html, because the latter looks like a path with
2393 a MIME-type, and this confuses some software.
2395 PH/70 Catch two compiler warnings in sieve.c.
2397 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2398 function verify_get_ident() calls ip_connect() to connect a socket, but
2399 if the "connect()" function timed out, ip_connect() used to close the
2400 socket. However, verify_get_ident() also closes the socket later, and in
2401 between Exim writes to the log, which may get opened at this point. When
2402 the socket was closed in ip_connect(), the log could get the same file
2403 descriptor number as the socket. This naturally causes chaos. The fix is
2404 not to close the socket in ip_connect(); the socket should be closed by
2405 the function that creates it. There was only one place in the code where
2406 this was missing, in the iplookup router, which I don't think anybody now
2407 uses, but I've fixed it anyway.
2409 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2410 well as to direct DNS lookups. Otherwise the handling of names in host
2411 lists is inconsistent and therefore confusing.
2417 PH/01 Two changes to the default runtime configuration:
2419 (1) Move the checks for relay_from_hosts and authenticated clients from
2420 after to before the (commented out) DNS black list checks.
2422 (2) Add control=submission to the relay_from_hosts and authenticated
2423 clients checks, on the grounds that messages accepted by these
2424 statements are most likely to be submissions.
2426 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2428 (1) Generate an error if the third argument for the ${prvs expansion is
2431 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2434 (3) Reset the variables that are obtained from the first argument of
2435 ${prvscheck and used in the second argument before leaving the code,
2436 because their memory is reclaimed, so using them afterwards may do
2439 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2440 one (it's much easier than Tom thought :-).
2442 (5) Because of (4), we can now allow for the use of $prvscheck_result
2443 inside the third argument.
2445 PH/03 For some reason, the default setting of PATH when running a command from
2446 a pipe transport was just "/usr/bin". I have changed it to
2449 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2450 anything to be listed in the output from -bV.
2452 PH/05 When a filter generated an autoreply, the entire To: header line was
2453 quoted in the delivery log line, like this:
2455 => >A.N.Other <ano@some.domain> <original@ddress> ...
2457 This has been changed so that it extracts the operative address. There
2458 may be more than one such address. If so, they are comma-separated, like
2461 => >ano@some.domain,ona@other.domain <original@ddress> ...
2463 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2464 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2465 not being looked up in the rDNS to get a host name, Exim was showing the
2466 IP address twice in Received: lines, even though the IP addresses were
2467 identical. For example:
2469 Received: from [1.2.3.4] (helo=[1.2.3.4])
2471 However, if the real host name was known, it was omitting the HELO data
2472 if it matched the actual IP address. This has been tidied up so that it
2473 doesn't show the same IP address twice.
2475 PH/07 When both +timestamp and +memory debugging was on, the value given by
2476 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2477 called by the debug printing, thereby overwriting the timestamp buffer.
2478 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2480 PH/08 When the original message was included in an autoreply transport, it
2481 always said "this is a copy of the message, including all the headers",
2482 even if body_only or headers_only was set. It now gives an appropriate
2485 PH/09 Applied a patch from the Sieve maintainer which:
2487 o fixes some comments
2488 o adds the (disabled) notify extension core
2489 o adds some debug output for the result of if/elsif tests
2490 o points to the current vacation draft in the documentation
2491 and documents the missing references header update
2495 o fixes a bug in processing the envelope test (when testing
2496 multiple envelope elements, the last element determinted the
2499 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2500 Electronic Mail") by including:
2502 Auto-submitted: auto-generated
2504 in the messages that it generates (bounce messages and others, such as
2505 warnings). In the case of bounce messages for non-SMTP mesages, there was
2506 also a typo: it was using "Auto_submitted" (underscore instead of
2507 hyphen). Since every message generated by Exim is necessarily in response
2508 to another message, thes have all been changed to:
2510 Auto-Submitted: auto-replied
2512 in accordance with these statements in the RFC:
2514 The auto-replied keyword:
2516 - SHOULD be used on messages sent in direct response to another
2517 message by an automatic process,
2519 - MUST NOT be used on manually-generated messages,
2521 - MAY be used on Delivery Status Notifications (DSNs) and Message
2522 Disposition Notifications (MDNs),
2524 - MUST NOT be used on messages generated by automatic or periodic
2525 processes, except for messages which are automatic responses to
2528 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2529 to the default Received: header definition.
2531 PH/12 Added log selector acl_warn_skipped (default on).
2533 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2534 variables because (a) they are in the wrong storage pool and (b) even if
2535 they were copied, it wouldn't work properly because of the caching.
2537 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2538 checking when decoding. Apparently there are clients that generate
2539 overlong encoded strings. Why am I not surprised?
2541 PH/15 If the first argument of "${if match_address" was not empty, but did not
2542 contain an "@" character, Exim crashed. Now it writes a panic log message
2543 and treats the condition as false.
2545 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2547 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2548 extension "envelope-auth". The code is finished and in agreement with
2549 other implementations, but there is no documentation so far and in fact,
2550 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2551 not changing the active code.
2553 Print executed "if" and "elsif" statements when debugging is used. This
2554 helps a great deal to understand what a filter does.
2556 Document more things not specified clearly in RFC3028. I had all this
2557 sorted out, when out of a sudden new issues came to my mind. Oops."
2559 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2562 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2563 canonical form (as documented). However, after a host name lookup from
2564 the IP address, check_host() was doing a simple string comparison with
2565 addresses acquired from the DNS when checking that the found name did
2566 have the original IP as one of its addresses. Since any found IPv6
2567 addresses are likely to be in abbreviated form, the comparison could
2568 fail. Luckily, there already exists a function for doing the comparison
2569 by converting both addresses to binary, so now that is used instead of
2570 the text comparison.
2572 PH/20 There was another similar case to PH/19, when a complete host name was
2573 given in a host list; looking up its IP address could give an abbreviated
2574 form, whereas the current host's name might or might not be abbreviated.
2575 The same fix has been applied.
2581 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2582 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2585 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2586 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2588 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2590 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2591 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2592 Also, for "Auto-Submitted", treat anything other than "no" as
2593 non-personal, in accordance with RFC 3834. (Previously it treated
2594 anything starting "auto-" as non-personal.)
2596 TF/01 The control=submission/name=... option had a problem with syntax
2597 errors if the name included a slash character. The /name= option
2598 now slurps the rest of the string, so it can include any characters
2599 but it must come last in the list of options (after /sender_retain
2602 PH/05 Some modifications to the interface to the fake nameserver for the new
2610 TK/01 Added the "success_on_redirect" address verification option. See
2611 NewStuff for rationale and an example.
2613 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2615 PH/02 Patch to exigrep to allow it to work on syslog lines.
2617 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2618 fread() to skip over the body file's header line, because in Cygwin the
2619 header line is locked and is inaccessible.
2621 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2622 co-exist for some time) to make it clear that it is the Exim ID that is
2623 referenced, not the Message-ID: header line.
2625 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2626 string_format() function, because snprintf() does not exist on all
2629 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2630 ${stat: expansion item.
2632 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2633 protocol synchronization error", to keep the pedants happy.
2635 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2636 well as for IRIX systems, when gcc is being used. See the host.c source
2639 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2641 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2644 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2645 transport and to the smtp transport in LMTP mode.
2647 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2649 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2650 run a filter in a subprocess. This could lead to confusion in subsequent
2651 lookups in the parent process. There should also be a search_tidyup() at
2652 the end of the subprocess.
2654 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2655 only if the host matched helo_try_verify_hosts, which caused the
2656 verification to occur when the EHLO/HELO command was issued. The ACL just
2657 tested the remembered result. Now, if a previous verification attempt has
2658 not happened, "verify = helo" does it there and then.
2660 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2662 TK/03 Fix log output including CR from clamd.
2664 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2665 exist provoked a memory error which could cause a segfault.
2667 PH/15 Installed PCRE 6.2
2669 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2671 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2672 of the problem. Specifically, suggested +O2 rather than +O1 for the
2675 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2677 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2678 record, the port was not correctly logged when the outgoing_port log
2679 selector was set (it logged the transort's default port).
2681 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2682 fallback_hosts, and "hosts" in the smtp transport.
2684 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2685 host errors such as "Connection refused".
2687 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2688 authentication with radiusclient 0.4.9:
2690 - Error returned from rc_read_config was caught wrongly
2691 - Username/password not passed on to radius server due to wrong length.
2693 The presumption is that some radiusclient API changes for 4.51/PH/17
2694 were not taken care of correctly. The code is still untested by me (my
2695 Linux distribution still has 0.3.2 of radiusclient), but it was
2696 contributed by a Radius user.
2698 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2699 expanding the "port" option of the smtp transport.
2701 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2702 while reading a MIME header. Thanks to Tom Hughes for a patch.
2704 PH/24 Include config.h inside local_scan.h so that configuration settings are
2707 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2708 that in an ACL for RCPT (for example), you can examine exactly what was
2711 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2712 commands, but it was not correctly comparing the address with the actual
2713 client host address. Thus, it would show the EHLO address in Received:
2714 header lines when this was not necessary.
2716 PH/27 Added the % operator to ${eval:}.
2718 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2719 it should be ignoring failures (because with -C, for example, it has lost
2720 privilege). It wasn't ignoring creation failures other than "already
2723 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2726 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2727 because an input line was too long, either on its own, or by virtue of
2728 too many continuations, the temporary file was not being removed, and the
2729 return code was incorrect.
2731 PH/31 Missing "BOOL" in function definition in filtertest.c.
2733 PH/32 Applied Sieve patches from the maintainer.
2735 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2737 PH/33 Added "verify = not_blind".
2739 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2740 Local/Makefile (with some defaults set). These are used in built scripts
2741 such as exicyclog, but they have never been used in the exim_install
2742 script (though there are many overriding facilities there). I have
2743 arranged that the exim_install script now takes note of these two
2746 PH/35 Installed configuration files for Dragonfly.
2748 PH/36 When a locally submitted message by a trusted user did not contain a
2749 From: header, and the sender address was obtained from -f or from an SMTP
2750 MAIL command, and the trusted user did not use -F to supply a sender
2751 name, $originator_name was incorrectly used when constructing a From:
2752 header. Furthermore, $originator_name was used for submission mode
2753 messages from external hosts without From: headers in a similar way,
2754 which is clearly wrong.
2756 PH/37 Added control=suppress_local_fixups.
2758 PH/38 When log_selector = +received_sender was set, and the addition of the
2759 sender made the log line's construction buffer exactly full, or one byte
2760 less than full, an overflow happened when the terminating "\n" was
2763 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2764 when the result of a list match is failure because a DNS lookup failed.
2766 PH/40 RM_COMMAND is now used in the building process.
2768 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2769 the "build-* directories that it finds.
2771 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2772 address in a domain literal was a prefix of an interface address.
2774 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2775 when verifying a sender address, unless rewrite_headers is false.
2777 PH/44 Wrote a long comment about why errors_to addresses are verified as
2778 recipients, not senders.
2780 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2781 the ratelimit ACL was added.
2783 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2785 PH/46 Added extra information about PostgreSQL errors to the error string.
2787 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2788 avoiding the need to install special zones in a real server. This is
2789 backwards compatible; if it can't find the fake resolver, it drops back.
2790 Thus, both old and new test suites can be run.
2792 TF/02 Added util/ratelimit.pl
2794 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2795 clock is set back in time.
2797 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2798 Candler <B.Candler@pobox.com>.
2800 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2801 OK for addresses that are the result of redirections.
2803 PH/48 A number of further additions for the benefit of the new test suite,
2804 including a fake gethostbyname() that interfaces to the fake DNS resolver
2807 TF/06 The fix for widen_domains has also been applied to qualify_single and
2808 search_parents which are the other dnslookup options that can cause
2811 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2814 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2816 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2817 service" if it encountered an MX record with an empty target hostname.
2818 The message is now "an MX or SRV record indicated no SMTP service".
2820 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2821 if the DNS of the sending site is misconfigured. This is quite a
2822 common situation. This change restores the behaviour of treating a
2823 helo verification defer as a failure.
2825 PH/51 If self=fail was set on a router, the bounce message did not include the
2826 actual error message.
2832 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2834 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2835 command itself ended in error, the underlying message about the transport
2836 filter timeout was being overwritten with the pipe command error. Now the
2837 underlying error message should be appended to the second error message.
2839 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2841 PH/02 Reduce the amount of output that "make" produces by default. Full output
2842 can still be requested.
2844 PH/03 The warning log line about a condition test deferring for a "warn" verb
2845 was being output only once per connection, rather than after each
2846 occurrence (because it was using the same function as for successful
2847 "warn" verbs). This seems wrong, so I have changed it.
2849 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2850 it should not have, which might have caused a crash in the right
2851 circumstances, but probably never did.
2853 PH/04 Installed a modified version of Tony Finch's patch to make submission
2854 mode fix the return path as well as the Sender: header line, and to
2855 add a /name= option so that you can make the user's friendly name appear
2858 TF/03 Added the control = fakedefer ACL modifier.
2860 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2861 Mark Lowes for thorough testing.
2863 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2865 TK/03 Merged latest SRS patch from Miles Wilton.
2867 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2868 with the definition in sysexits.h (which is #included earlier).
2869 Fortunately, Exim does not actually use EX_OK. The code used to try to
2870 preserve the sysexits.h value, by assumimg that macro definitions were
2871 scanned for macro replacements. I have been disabused of this notion,
2872 so now the code just undefines EX_OK before #including unistd.h.
2874 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2875 in the smtp transport. When a block could not be written in a single
2876 write() function, the timeout was being re-applied to each part-write.
2877 This seems wrong - if the receiver was accepting one byte at a time it
2878 would take for ever. The timeout is now adjusted when this happens. It
2879 doesn't have to be particularly precise.
2881 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2882 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2884 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2885 without a domain if the check to <postmaster@domain> fails.
2887 SC/01 Eximstats: added -xls and the ability to specify output files
2888 (patch written by Frank Heydlauf).
2890 SC/02 Eximstats: use FileHandles for outputing results.
2892 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2894 SC/04 Eximstats: fixed display of large numbers with -nvr option
2896 SC/05 Eximstats: fixed merging of reports with empty tables.
2898 SC/06 Eximstats: added the -include_original_destination flag
2900 SC/07 Eximstats: removed tabs and trailing whitespace.
2902 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2904 TK/06 MBOX spool code: Add real "From " MBOX separator line
2905 so the .eml file is really in mbox format (even though
2906 most programs do not really care). Patch from Alex Miller.
2908 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2909 The latter is generated from $received_to and is only set if the
2910 message has one envelope recipient. SA can use these headers,
2911 obviously out-of-the-box. Patch from Alex Miller.
2913 PH/08 The ${def test on a variable was returning false if the variable's
2914 value was "0", contrary to what the specification has always said!
2915 The result should be true unless the variable is empty.
2917 PH/09 The syntax error of a character other than { following "${if
2918 def:variable_name" (after optional whitespace) was not being diagnosed.
2919 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2920 accidental colon was present, for example, could give incorrect results.
2922 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2923 result is used (not including appendfile, where other changes are about
2926 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2927 This involved changing a lot of size variables from int to off_t. It
2928 should work with maildirs and everything.
2930 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2931 spamd dying while we are connected to it.
2933 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2936 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2937 function for BDB 4.3.
2939 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2941 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2942 the sources for PCRE have been re-arranged and more files are now
2945 PH/15 The code I had for printing potentially long long variables in PH/11
2946 above was not the best (it lost precision). The length of off_t variables
2947 is now inspected at build time, and an appropriate printing format (%ld
2948 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2949 to be "long long int" or "long int". This is needed for the internal
2950 formatting function string_vformat().
2952 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2953 the configuration file to be ":syslog", then the script "guesses" where
2954 the logs files are, rather than using the compiled in default. In our
2955 case the guess is not the same as the compiled default, so the script
2956 suddenly stopped working when I started to use syslog. The patch checks
2957 to see if log_file_path is "". If so, it attempts to read it from exim
2958 with no configuration file to get the compiled in version, before it
2959 falls back to the previous guessing code."
2961 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2962 implementing BATV in an Exim configuration. See NewStuff for the gory
2965 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2966 Makefile that are specific to HP-UX.
2968 PH/18 If the "use_postmaster" option was set for a recipient callout together
2969 with the "random" option, the postmaster address was used as the MAIL
2970 FROM address for the random test, but not for the subsequent recipient
2971 test. It is now used for both.
2973 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2974 patch removes a few documentation additions to RFC 3028, because the
2975 latest draft now contains them. It adds the new en;ascii-case comparator
2976 and a new error check for 8bit text in MIME parts. Comparator and
2977 require names are now matched exactly. I enabled the subaddress
2978 extension, but it is not well tested yet (read: it works for me)."
2980 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2981 rework some of the code of TK/09 above to avoid the hardwired use of
2982 "%lld" and "long long". Replaced the call to snprintf() with a call to
2985 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2986 records point to non-existent hosts", "retry timeout exceeded", and
2987 "retry time not reached for any host after a long failure period".
2989 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2990 experimental DomainKeys support:
2992 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2993 (2) On an error such as an illegally used "control", the wrong name for
2994 the control was given.
2996 These problems did NOT occur unless DomainKeys support was compiled.
2998 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
3000 PH/24 Added ${if match_ip condition.
3002 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
3003 databases so that it will be absolutely obvious if a crash occurs in the
3004 DB library. This is a regular occurrence (often caused by mis-matched
3007 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
3008 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
3009 picked up on a user's system that detects such things. There doesn't seem
3010 to be a gcc warning option for this - only an attribute that has to be
3011 put on the function's prototype. It seems that in Fedora Core 4 they have
3012 set this on a number of new functions. No doubt there will be more in due
3015 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
3016 specify a transport. However, if an address that was verified by such a
3017 router was the subject of a callout, Exim crashed because it tried to
3018 read the rcpt_include_affixes from the non-existent transport. Now it
3019 just assumes that the setting of that option is false. This bug was
3020 introduced by 4.51/PH/31.
3022 PH/28 Changed -d+all to exclude +memory, because that information is very
3023 rarely of interest, but it makes the output a lot bigger. People tend to
3024 do -d+all out of habit.
3026 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
3027 code in os-type was giving problems when libc.so lives in lib64, like on
3030 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
3031 aren't the modern standard, and it seems that some systems' include files
3032 don't always have them. Exim was already checking for some of the newer
3033 ones like T_AAAA, and defining it itself. I've added checks for all the
3034 record types that Exim uses.
3036 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
3037 not automatically generating a new one, as it is supposed to. This
3038 prevented TLS from working. If the file did exist, but contained invalid
3039 data, a new version was generated, as expected. It was only the case of a
3040 non-existent file that was broken.
3042 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
3043 with a change in libdomainkeys > 0.64.
3045 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
3046 from DNS. If the selector record carries the flag, it now has
3047 precedence over the domain-wide flag.
3049 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
3051 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
3052 the use of a port name that isn't defined in /etc/services) occurred, the
3053 message was deferred as in a normal delivery, and thus remained on the
3054 spool, instead of being failed because of the mua_wrapper setting. This
3055 is now fixed, and I tidied up some of the mua_wrapper messages at the
3058 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
3059 the messages in a hash of arrays rather than using individual hashes.
3060 This is a bit cleaner and results in dramatic memory savings, albeit
3061 at a slight CPU cost.
3063 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
3064 as requested by Marc Sherman.
3066 SC/10 Eximstats: added histograms for user specified patterns as requested
3069 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
3071 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
3072 fopen() in the content-scanning modules that did not already have it.
3078 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
3079 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
3081 TK/02 Fix ACL "control" statement not being available in MIME ACL.
3083 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
3085 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
3086 to test Sieve filters that use "vacation".
3088 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
3089 that changes the way the GnuTLS parameters are stored in the cache file.
3090 The new format can be generated externally. For backward compatibility,
3091 if the data in the cache doesn't make sense, Exim assumes it has read an
3092 old-format file, and it generates new data and writes a new file. This
3093 means that you can't go back to an older release without removing the
3096 PH/03 A redirect router that has both "unseen" and "one_time" set does not
3097 work if there are any delivery delays because "one_time" forces the
3098 parent to be marked "delivered", so its unseen clone is never tried
3099 again. For this reason, Exim now forbids the simultaneous setting of
3102 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
3103 redirected to themselves ("homonym" addresses). Read the long ChangeLog
3104 entry if you want to know the details. The fix, however, neglected to
3105 consider the case when local delivery batching is involved. The test for
3106 "previously delivered" was not happening when checking to see if an
3107 address could be batched with a previous (undelivered) one; under
3108 certain circumstances this could lead to multiple deliveries to the same
3111 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
3112 in its include files, and this causes problems building Exim.
3114 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
3115 header_syntax) but Exim was just ignoring anything given after a slash.
3116 In particular, this caused confusion with an attempt to use "verify =
3117 reverse_host_lookup/defer_ok". An error is now given when options are
3118 supplied for verify items that do not have them. (Maybe reverse_host_
3119 lookup should have a defer_ok option, but that's a different point.)
3121 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
3122 defined by RFC 821) to 2048, because there were problems with some AUTH
3123 commands, and RFC 1869 says the size should be increased for extended
3124 SMTP commands that take arguments.
3126 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
3129 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
3130 "unknown" error; now it says that the functionality isn't in the binary.
3132 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
3133 an address' error message when a string expansion fails (syntax or
3134 whatever). Otherwise the password may appear in the log. Following change
3135 PH/42 below, there is no longer a chance of it appearing in a bounce
3138 PH/11 Installed exipick version 20050225.0 from John Jetmore.
3140 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3141 of its addresses was ever tried. (Bugzilla bug #2.)
3143 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3144 the result incorrectly in the debug output. (It correctly added a newline
3145 to what was transported.)
3147 TF/01 Added $received_time.
3149 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3150 commented out examples of how to interface to a virus scanner and to
3151 SpamAssassin. Also added commented examples of av_scanner and
3152 spamd_address settings.
3154 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3155 and controls are allowed in which ACLs. There were a couple of minor
3156 errors. Some of the entries in the conditions table (which is a table of
3157 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3158 negation of where the condition IS allowed.
3160 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3162 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3163 header file does not have a version number, so I've had to invent a new
3164 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3165 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3166 radiusclient), but it was contributed by a Radius user.
3168 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3169 files or directories, for OpenSSL.
3171 PH/19 When an Exim process that is running as root has to create an Exim log
3172 file, it does so in a subprocess that runs as exim:exim so as to get the
3173 ownership right at creation (otherwise, other Exim processes might see
3174 the file with the wrong ownership). There was no test for failure of this
3175 fork() call, which would lead to the process getting stuck as it waited
3176 for a non-existent subprocess. Forks do occasionally fail when resources
3177 run out. I reviewed all the other calls to fork(); they all seem to check
3180 PH/20 When checking for unexpected SMTP input at connect time (before writing
3181 the banner), Exim was not dealing correctly with a non-positive return
3182 from the read() function. If the client had disconnected by this time,
3183 the result was a log entry for a synchronization error with an empty
3184 string after "input=" when read() returned zero. If read() returned -1
3185 (an event I could not check), uninitialized data bytes were printed.
3186 There were reports of junk text (parts of files, etc) appearing after
3189 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3191 PH/22 Added support for macro redefinition, and (re)definition in between
3192 driver and ACL definitions.
3194 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3195 forgetting to use the resulting value; it was using the unexpanded value.
3197 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3198 hadn't been configured. The fix is from Juergen Kreileder, who
3199 understands it better than I do:
3201 "Here's what I see happening with three configured cyrus_sasl
3202 authenticators configured (plain, login, cram-md5):
3204 On startup auth_cyrus_sasl_init() gets called for each of these.
3205 This means three calls to sasl_listmech() without a specified mech_list.
3206 => SASL tests which mechs of all available mechs actually work
3207 => three warnings about OTP not working
3208 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3210 With the patch, sasl_listmech() also gets called three times. But now
3211 SASL's mech_list option is set to the server_mech specified in the the
3212 authenticator. Or in other words, the answer from sasl_listmech()
3213 gets limited to just the mech you're testing for (which is different
3215 => the return list contains just 'plain' or 'login', 'cram-md5' or
3216 nothing depending on the value of ob->server_mech.
3218 I've just tested the patch: Authentication still works fine,
3219 unavailable mechs specified in the exim configuration are still
3220 caught, and the auth.log warnings about OTP are gone."
3222 PH/25 When debugging is enabled, the contents of the command line are added
3223 to the debugging output, even when log_selector=+arguments is not
3226 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3227 answer is "GNU", and only if the return is "GNU/something" is the answer
3230 PH/27 $acl_verify_message is now set immediately after the failure of a
3231 verification in an ACL, and so is available in subsequent modifiers. In
3232 particular, the message can be preserved by coding like this:
3234 warn !verify = sender
3235 set acl_m0 = $acl_verify_message
3237 Previously, $acl_verify_message was set only while expanding "message"
3238 and "log_message" when a very denied access.
3240 PH/28 Modified OS/os.c-Linux with
3242 -#ifndef OS_LOAD_AVERAGE
3243 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3245 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3246 nomenclature these days.)
3248 PH/29 Installed patch from the Sieve maintainer that adds the options
3249 sieve_useraddress and sieve_subaddress to the redirect router.
3251 PH/30 In these circumstances:
3252 . Two addresses routed to the same list of hosts;
3253 . First host does not offer TLS;
3254 . First host accepts first address;
3255 . First host gives temporary error to second address;
3256 . Second host offers TLS and a TLS session is established;
3257 . Second host accepts second address.
3258 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3259 and peerdn, if requested) that were in fact used only for the second
3262 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3263 attention to any local part prefix or suffix that was matched by the
3264 router that accepted the address. It now behaves in the same way as it
3265 does for delivery: the affixes are removed from the local part unless
3266 rcpt_include_affixes is set on the transport.
3268 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3269 timeout during the DATA phase of an incoming message.
3271 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3272 applied a patch sanctioned by the Sieve maintainer.
3274 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3275 the uid or gid is negative. A case of a negative gid caused this to be
3276 noticed. The fix allows for either to be negative.
3278 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3279 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3282 PH/36 The change PH/12 above was broken. Fixed it.
3284 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3285 the grounds that routing the same address twice would always produce the
3286 same answer. This might have been true once, but it is certainly no
3287 longer true now. Routing a child address may depend on the previous
3288 routing that produced that child. Some complicated redirection strategies
3289 went wrong when messages had multiple recipients, and made Exim's
3290 behaviour dependent on the order in which the addresses were given.
3292 I have moved the duplicate checking until after the routing is complete.
3293 Exim scans the addresses that are assigned to local and remote
3294 transports, and removes any duplicates. This means that more work will be
3295 done, as duplicates will always all be routed, but duplicates are
3296 presumably rare, so I don't expect this is of any significance.
3298 For deliveries to pipes, files, and autoreplies, the duplicate checking
3299 still happens during the routing process, since they are not going to be
3302 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3303 It corrects a timeout issue with spamd. This is Ian's comment: "The
3304 background is that sometimes spamd either never reads data from a
3305 connection it has accepted, or it never writes response data. The exiscan
3306 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3307 blindly assumes that writes won't block so it may never time out."
3309 PH/39 Allow G after quota size as well as K and M.
3311 PH/40 The value set for $authenticated_id in an authenticator may not contain
3312 binary zeroes or newlines because the value is written to log lines and
3313 to spool files. There was no check on this. Now the value is run through
3314 the string_printing() function so that such characters are converted to
3315 printable escape sequences.
3317 PH/41 $message_linecount is a new variable that contains the total number of
3318 lines in the message. Compare $body_linecount, which is the count for the
3321 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3322 bounce and delay warning messages, except in certain special cases, which
3325 (a) An SMTP error message from a remote host;
3326 (b) A message specified in a :fail: redirection;
3327 (c) A message specified in a "fail" command in a system filter;
3328 (d) A message specified in a FAIL return from the queryprogram router;
3329 (e) A message specified by the cannot_route_message router option.
3331 In these cases only, Exim does include the error details in bounce and
3332 warning messages. There are also a few cases where bland messages such
3333 as "unrouteable address" or "local delivery error" are given.
3335 PH/43 $value is now also set for the "else" part of a ${run expansion.
3337 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3338 being worked on, but at least Exim now implements the latest version to
3341 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3342 process to complete was treated as a delivery failure, a timeout while
3343 writing the message to the pipe was logged, but erroneously treated as a
3344 successful delivery. Such timeouts include transport filter timeouts. For
3345 consistency with the overall process timeout, these timeouts are now
3346 treated as errors, giving rise to delivery failures by default. However,
3347 there is now a new Boolean option for the pipe transport called
3348 timeout_defer, which, if set TRUE, converts the failures into defers for
3349 both kinds of timeout. A transport filter timeout is now identified in
3352 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3353 systems where "make" and "gmake" are different, calling "gmake" at top
3354 level broke things. I've arranged for the value of $(MAKE) to be passed
3355 from the Makefile to this script so that it can call the same version of
3359 A note about Exim versions 4.44 and 4.50
3360 ----------------------------------------
3362 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3363 changes of various kinds. As a consequence, a big documentation update was
3364 needed. This delayed the release for rather longer than seemed good, especially
3365 in the light of a couple of (minor) security issues. Therefore, the changes
3366 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3367 release. So 4.44 and 4.50 are in effect two different branches that both start
3370 I have left the 4.50 change log unchanged; it contains all the changes since
3371 4.43. The change log for 4.44 is below; many of its items are identical to
3372 those for 4.50. This seems to be the most sensible way to preserve the
3373 historical information.
3379 1. Minor wording change to the doc/README.SIEVE file.
3381 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3382 computation of the current number of files was incorrect.
3384 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3385 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3388 4. Give more explanation in the error message when the command for a transport
3389 filter fails to execute.
3391 5. There are several places where Exim runs a non-Exim command in a
3392 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3393 was being done only for the command run by the queryprogram router. It is
3394 now done for all such subprocesses. The other cases are: ${run, transport
3395 filters, and the commands run by the lmtp and pipe transports.
3397 6. Added CONFIGURE_GROUP build-time option.
3399 7. Some older OS have a limit of 256 on the maximum number of file
3400 descriptors. Exim was using setrlimit() to set 1000 as a large value
3401 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3402 systems. I've change it so that if it can't get 1000, it tries for 256.
3404 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3405 was an oversight, and furthermore, ever since the addition of extra
3406 controls (e.g. 4.43/32), the checks on when to allow different forms of
3407 "control" were broken. There should now be diagnostics for all cases when a
3408 control that does not make sense is encountered.
3410 9. Added the /retain_sender option to "control=submission".
3412 10. $recipients is now available in the predata ACL (oversight).
3414 11. Tidy the search cache before the fork to do a delivery from a message
3415 received from the command line. Otherwise the child will trigger a lookup
3416 failure and thereby defer the delivery if it tries to use (for example) a
3417 cached ldap connection that the parent has called unbind on.
3419 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3420 of $address_data from the recipient verification was clobbered by the
3421 sender verification.
3423 13. The value of address_data from a sender verification is now available in
3424 $sender_address_data in subsequent conditions in the ACL statement.
3426 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3428 15. Added a new option "connect=<time>" to callout options, to set a different
3431 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3432 was its contents. (It was OK if the option was not defined at all.)
3434 17. A "Completed" log line is now written for messages that are removed from
3435 the spool by the -Mrm option.
3437 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3438 information about exactly what failed.
3440 19. Added -dd to debug only the daemon process.
3442 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3443 handles timeouts, both on the server side and network timeouts. Renamed the
3444 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3446 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3447 It is now set to "smtps".
3449 22. $host_address is now set to the target address during the checking of
3450 ignore_target_hosts.
3452 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3453 being passed; this would have caused $sender_host_name to have been used if
3454 matching the list had actually called for a host name (not very likely,
3455 since this list is usually IP addresses). A host name is now passed as
3458 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3459 code that specifies a non-restarting handler (typically sigaction() in
3460 modern systems) in an attempt to fix a rare and obscure crash bug.
3462 25. Narrowed the window for a race in the daemon that could cause it to ignore
3463 SIGCHLD signals. This is not a major problem, because they are used only to
3464 wake it up if nothing else does.
3466 26. A malformed maildirsize file could cause Exim to calculate negative values
3467 for the mailbox size or file count. Odd effects could occur as a result.
3468 The maildirsize information is now recalculated if the size or filecount
3471 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3472 support for a long time. Removed HAVE_SYS_VFS_H.
3474 28. Installed the latest version of exipick from John Jetmore.
3476 29. In an address list, if the pattern was not a regular expression, an empty
3477 subject address (from a bounce message) matched only if the pattern was an
3478 empty string. Non-empty patterns were not even tested. This was the wrong
3479 because it is perfectly reasonable to use an empty address as part of a
3480 database query. An empty address is now tested by patterns that are
3481 lookups. However, all the other forms of pattern expect the subject to
3482 contain a local part and a domain, and therefore, for them, an empty
3483 address still always fails if the pattern is not itself empty.
3485 30. Exim went into a mad DNS loop when attempting to do a callout where the
3486 host was specified on an smtp transport, and looking it up yielded more
3487 than one IP address.
3489 31. Re-factored the code for checking spool and log partition space into a
3490 function that finds that data and another that does the check. The former
3491 is then used to implement four new variables: $spool_space, $log_space,
3492 $spool_inodes, and $log_inodes.
3494 32. The RFC2047 encoding function was originally intended for short strings
3495 such as real names; it was not keeping to the 75-character limit for
3496 encoded words that the RFC imposes. It now respects the limit, and
3497 generates multiple encoded words if necessary. To be on the safe side, I
3498 have increased the buffer size for the ${rfc2047: expansion operator from
3501 33. It is now permitted to omit both strings after an "if" condition; if the
3502 condition is true, the result is "true". As before, when the second string
3503 is omitted, a false condition yields an empty string. This makes it less
3504 cumbersome to write custom ACL and router conditions.
3506 34. Failure to deliver a bounce message always caused it to be frozen, even if
3507 there was an errors_to setting on the router. The errors_to setting is now
3510 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3511 canonical form (fully expanded) before being placed in
3512 $sender_host_address.
3514 36. The table in the code that translates DNS record types into text (T_A to
3515 "A" for instance) was missing entries for NS and CNAME. It is just possible
3516 that this could have caused confusion if both these types were looked up
3517 for the same domain, because the text type is used as part of Exim's
3518 per-process caching. But the chance of anyone hitting this buglet seems
3521 37. The dnsdb lookup has been extended in a number of ways.
3523 (1) There is a new type, "zns", which walks up the domain tree until it
3524 finds some nameserver records. It should be used with care.
3526 (2) There is a new type, "mxh", which is like "mx" except that it returns
3527 just the host names, not the priorities.
3529 (3) It is now possible to give a list of domains (or IP addresses) to be
3530 looked up. The behaviour when one of the lookups defers can be
3531 controlled by a keyword.
3533 (4) It is now possible to specify the separator character for use when
3534 multiple records are returned.
3536 38. The dnslists ACL condition has been extended: it is now possible to supply
3537 a list of IP addresses and/or domains to be looked up in a particular DNS
3540 39. Added log_selector=+queue_time_overall.
3542 40. When running the queue in the test harness, wait just a tad after forking a
3543 delivery process, to get repeatability of debugging output.
3545 41. Include certificate and key file names in error message when GnuTLS fails
3546 to set them up, because the GnuTLS error message doesn't include the name
3547 of the failing file when there is a problem reading it.
3549 42. Allow both -bf and -bF in the same test run.
3551 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3553 44. The "Exiscan patch" is now merged into the mainline Exim source.
3555 45. Sometimes the final signoff response after QUIT could fail to get
3556 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3557 < 0 before doing a fflush(). This bug looks as though it goes back to the
3558 introduction of TLS in release 3.20, but "sometimes" must have been rare
3559 because the tests only now provoked it.
3561 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3562 (this can affect the format of dates).
3564 47. exim_tidydb, when checking for the continued existence of a message for
3565 which it has found a message-specific retry record, was not finding
3566 messages that were in split spool directories. Consequently, it was
3567 deleting retry records that should have stayed in existence.
3569 48. Steve fixed some bugs in eximstats.
3571 49. The SPA authentication driver was not abandoning authentication and moving
3572 on to the next authenticator when an expansion was forced to fail,
3573 contradicting the general specification for all authenticators. Instead it
3574 was generating a temporary error. It now behaves as specified.
3576 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3577 (the order specifies the preference for clients). The order is now AES256,
3578 AES128, 3DES, ARCFOUR128.
3580 51. Small patch to Sieve code - explicitly set From: when generating an
3583 52. Exim crashed if a remote delivery caused a very long error message to be
3584 recorded - for instance if somebody sent an entire SpamAssassin report back
3585 as a large number of 550 error lines. This bug was coincidentally fixed by
3586 increasing the size of one of Exim's internal buffers (big_buffer) that
3587 happened as part of the Exiscan merge. However, to be on the safe side, I
3588 have made the code more robust (and fixed the comments that describe what
3591 53. Now that there can be additional text after "Completed" in log lines (if
3592 the queue_time_overall log selector is set), a one-byte patch to exigrep
3593 was needed to allow it to recognize "Completed" as not the last thing in
3596 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3597 patch that reportedly fixes this has been added. I am not expert enough to
3598 create a test for it. This is what the patch creator wrote:
3600 "I found a little strange behaviour of ldap code when working with
3601 Windows 2003 AD Domain, where users was placed in more than one
3602 Organization Units. When I tried to give exim partial DN, the exit code
3603 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3604 But simultaneously result of request was absolutely normal ldap result,
3605 so I produce this patch..."
3607 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3608 REFERENCE, so I have modified the code to exclude the patch when that macro
3611 55. Some experimental protocols are using DNS PTR records for new purposes. The
3612 keys for these records are domain names, not reversed IP addresses. The
3613 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3614 leaves it alone. Component reversal etc. now happens only for IP addresses.
3617 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3619 57. Double the size of the debug message buffer (to 2048) so that more of very
3620 long debug lines gets shown.
3622 58. The exicyclog utility now does better if the number of log files to keep
3623 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3625 59. Two changes related to the smtp_active_hostname option:
3627 (1) $smtp_active_hostname is now available as a variable.
3628 (2) The default for smtp_banner uses $smtp_active_hostname instead
3629 of $primary_hostname.
3631 60. The host_aton() function is supposed to be passed a string that is known
3632 to be a valid IP address. However, in the case of IPv6 addresses, it was
3633 not checking this. This is a hostage to fortune. Exim now panics and dies
3634 if the condition is not met. A case was found where this could be provoked
3635 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3636 components; fortuitously, this particular loophole had already been fixed
3637 by change 4.50/55 above.
3639 If there are any other similar loopholes, the new check in host_aton()
3640 itself should stop them being exploited. The report I received stated that
3641 data on the command line could provoke the exploit when Exim was running as
3642 exim, but did not say which command line option was involved. All I could
3643 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3644 running as the user.
3647 61. There was a buffer overflow vulnerability in the SPA authentication code
3648 (which came originally from the Samba project). I have added a test to the
3649 spa_base64_to_bits() function which I hope fixes it.
3652 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3653 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3655 63. The daemon start-up calls getloadavg() while still root for those OS that
3656 need the first call to be done as root, but it missed one case: when
3657 deliver_queue_load_max is set with deliver_drop_privilege. This is
3658 necessary for the benefit of the queue runner, because there is no re-exec
3659 when deliver_drop_privilege is set.
3661 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3662 This has been fixed.
3664 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3665 was in use, was not putting the data itself into the right store pool;
3666 consequently, it could be overwritten for a subsequent message in the same
3667 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3670 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3672 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3673 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3674 the function were treating the return as a boolean value, which happened to
3675 work because 0=false and not-0=true, but is not correct code.
3677 68. The host_aton() function was not handling scoped IPv6 addresses (those
3678 with, for example, "%eth0" on the end) correctly.
3680 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3681 negated items (that is, ~something) in unsigned ints. Some compilers
3682 apparently mutter when there is no cast.
3684 70. If an address verification called from an ACL failed, and did not produce a
3685 user-specific message (i.e. there was only a "system" message), nothing was
3686 put in $acl_verify_message. In this situation, it now puts the system
3689 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3690 session; change 4.31/43 added the unwanted input to the log line - except
3691 that it did not do this in the start of session case. It now does.
3693 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3694 This is wrong and can cause the other end to generate a synchronization
3695 error if it is another Exim or anything else that does the synchronization
3696 check. A QUIT command is no longer sent after a timeout.
3698 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3699 during host lookups.
3701 74. The defer_ok option of callout verification was not working if it was used
3702 when verifying addresses in header lines, that is, for this case:
3704 verify = header_sender/callout=defer_ok
3706 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3707 those file descriptors could be used for SMTP connections. If anything
3708 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3709 could be sent to the SMTP client, causing chaos. The daemon now opens
3710 stdin, stdout, and stderr to /dev/null when it puts itself into the
3713 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3714 log by default. The user can override this with suitable Perl magic.
3716 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3717 the log message when discard triggers, was not working for the DATA ACL or
3718 for the non-SMTP ACL.
3720 78. Error message wording change in sieve.c.
3722 79. If smtp_accept_max_per_host was set, the number of connections could be
3723 restricted to fewer than expected, because the daemon was trying to set up
3724 a new connection before checking whether the processes handling previous
3725 connections had finished. The check for completed processes is now done
3726 earlier. On busy systems, this bug wouldn't be noticed because something
3727 else would have woken the daemon, and it would have reaped the completed
3730 80. If a message was submitted locally by a user whose login name contained one
3731 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3732 It caused a spool format error. I have fixed the spool reading code. A
3733 related problem was that the "from" clause in the Received: line became
3734 illegal because of the space(s). It is now covered by ${quote_local_part.
3736 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3739 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3741 83. Patch from Sieve maintainer to fix unterminated string problem in
3742 "vacation" handling.
3744 84. Some minor changes to the Linux configuration files to help with other
3745 OS variants using glibc.
3747 85. One more patch for Sieve to update vacation handling to latest spec.
3750 ----------------------------------------------------
3751 See the note above about the 4.44 and 4.50 releases.
3752 ----------------------------------------------------
3758 1. Change 4.43/35 introduced a bug that caused file counts to be
3759 incorrectly computed when quota_filecount was set in an appendfile
3762 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3763 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3766 3. Give more explanation in the error message when the command for a transport
3767 filter fails to execute.
3769 4. There are several places where Exim runs a non-Exim command in a
3770 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3771 was being done only for the command run by the queryprogram router. It is
3772 now done for all such subprocesses. The other cases are: ${run, transport
3773 filters, and the commands run by the lmtp and pipe transports.
3775 5. Some older OS have a limit of 256 on the maximum number of file
3776 descriptors. Exim was using setrlimit() to set 1000 as a large value
3777 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3778 systems. I've change it so that if it can't get 1000, it tries for 256.
3780 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3781 was an oversight, and furthermore, ever since the addition of extra
3782 controls (e.g. 4.43/32), the checks on when to allow different forms of
3783 "control" were broken. There should now be diagnostics for all cases when a
3784 control that does not make sense is encountered.
3786 7. $recipients is now available in the predata ACL (oversight).
3788 8. Tidy the search cache before the fork to do a delivery from a message
3789 received from the command line. Otherwise the child will trigger a lookup
3790 failure and thereby defer the delivery if it tries to use (for example) a
3791 cached ldap connection that the parent has called unbind on.
3793 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3794 of $address_data from the recipient verification was clobbered by the
3795 sender verification.
3797 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3798 was its contents. (It was OK if the option was not defined at all.)
3800 11. A "Completed" log line is now written for messages that are removed from
3801 the spool by the -Mrm option.
3803 12. $host_address is now set to the target address during the checking of
3804 ignore_target_hosts.
3806 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3807 being passed; this would have caused $sender_host_name to have been used if
3808 matching the list had actually called for a host name (not very likely,
3809 since this list is usually IP addresses). A host name is now passed as
3812 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3813 code that specifies a non-restarting handler (typically sigaction() in
3814 modern systems) in an attempt to fix a rare and obscure crash bug.
3816 15. Narrowed the window for a race in the daemon that could cause it to ignore
3817 SIGCHLD signals. This is not a major problem, because they are used only to
3818 wake it up if nothing else does.
3820 16. A malformed maildirsize file could cause Exim to calculate negative values
3821 for the mailbox size or file count. Odd effects could occur as a result.
3822 The maildirsize information is now recalculated if the size or filecount
3825 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3826 support for a long time. Removed HAVE_SYS_VFS_H.
3828 18. Updated exipick to current release from John Jetmore.
3830 19. Allow an empty sender to be matched against a lookup in an address list.
3831 Previously the only cases considered were a regular expression, or an
3834 20. Exim went into a mad DNS lookup loop when doing a callout where the
3835 host was specified on the transport, if the DNS lookup yielded more than
3838 21. The RFC2047 encoding function was originally intended for short strings
3839 such as real names; it was not keeping to the 75-character limit for
3840 encoded words that the RFC imposes. It now respects the limit, and
3841 generates multiple encoded words if necessary. To be on the safe side, I
3842 have increased the buffer size for the ${rfc2047: expansion operator from
3845 22. Failure to deliver a bounce message always caused it to be frozen, even if
3846 there was an errors_to setting on the router. The errors_to setting is now
3849 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3850 canonical form (fully expanded) before being placed in
3851 $sender_host_address.
3853 24. Updated eximstats to version 1.33
3855 25. Include certificate and key file names in error message when GnuTLS fails
3856 to set them up, because the GnuTLS error message doesn't include the name
3857 of the failing file when there is a problem reading it.
3859 26. Expand error message when OpenSSL has problems setting up cert/key files.
3862 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3863 (this can affect the format of dates).
3865 28. exim_tidydb, when checking for the continued existence of a message for
3866 which it has found a message-specific retry record, was not finding
3867 messages that were in split spool directories. Consequently, it was
3868 deleting retry records that should have stayed in existence.
3870 29. eximstats updated to version 1.35
3871 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3872 1.35 - bugfix such that pie charts by volume are generated correctly
3874 30. The SPA authentication driver was not abandoning authentication and moving
3875 on to the next authenticator when an expansion was forced to fail,
3876 contradicting the general specification for all authenticators. Instead it
3877 was generating a temporary error. It now behaves as specified.
3879 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3880 (the order specifies the preference for clients). The order is now AES256,
3881 AES128, 3DES, ARCFOUR128.
3883 31. Small patch to Sieve code - explicitly set From: when generating an
3886 32. Exim crashed if a remote delivery caused a very long error message to be
3887 recorded - for instance if somebody sent an entire SpamAssassin report back
3888 as a large number of 550 error lines. This bug was coincidentally fixed by
3889 increasing the size of one of Exim's internal buffers (big_buffer) that
3890 happened as part of the Exiscan merge. However, to be on the safe side, I
3891 have made the code more robust (and fixed the comments that describe what
3894 33. Some experimental protocols are using DNS PTR records for new purposes. The
3895 keys for these records are domain names, not reversed IP addresses. The
3896 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3897 leaves it alone. Component reversal etc. now happens only for IP addresses.
3900 34. The host_aton() function is supposed to be passed a string that is known
3901 to be a valid IP address. However, in the case of IPv6 addresses, it was
3902 not checking this. This is a hostage to fortune. Exim now panics and dies
3903 if the condition is not met. A case was found where this could be provoked
3904 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3905 components; fortuitously, this particular loophole had already been fixed
3906 by change 4.50/55 or 4.44/33 above.
3908 If there are any other similar loopholes, the new check in host_aton()
3909 itself should stop them being exploited. The report I received stated that
3910 data on the command line could provoke the exploit when Exim was running as
3911 exim, but did not say which command line option was involved. All I could
3912 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3913 running as the user.
3916 35. There was a buffer overflow vulnerability in the SPA authentication code
3917 (which came originally from the Samba project). I have added a test to the
3918 spa_base64_to_bits() function which I hope fixes it.
3921 36. The daemon start-up calls getloadavg() while still root for those OS that
3922 need the first call to be done as root, but it missed one case: when
3923 deliver_queue_load_max is set with deliver_drop_privilege. This is
3924 necessary for the benefit of the queue runner, because there is no re-exec
3925 when deliver_drop_privilege is set.
3927 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3928 was in use, was not putting the data itself into the right store pool;
3929 consequently, it could be overwritten for a subsequent message in the same
3930 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3933 38. Sometimes the final signoff response after QUIT could fail to get
3934 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3935 < 0 before doing a fflush(). This bug looks as though it goes back to the
3936 introduction of TLS in release 3.20, but "sometimes" must have been rare
3937 because the tests only now provoked it.
3943 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3944 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3945 Now it returns an integer. A number of calls were still expecting a T/F
3946 return. Fortuitously, in all cases, the tests worked in OK situations,
3947 which is the norm. However, things would have gone wrong on any write
3948 failures on the smtp file descriptor. This function is used when sending
3949 messages over SMTP and also when doing verify callouts.
3951 2. When Exim is called to do synchronous delivery of a locally submitted
3952 message (the -odf or -odi options), it no longer closes stderr before doing
3955 3. Implemented the mua_wrapper option.
3957 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3959 5. Implemented the functions header_remove(), header_testname(),
3960 header_add_at_position(), and receive_remove_recipient(), and exported them
3963 6. If an ACL "warn" statement specified the addition of headers, Exim already
3964 inserted X-ACL-Warn: at the start if there was no header name. However, it
3965 was not making this test for the second and subsequent header lines if
3966 there were newlines in the string. This meant that an invalid header could
3967 be inserted if Exim was badly configured.
3969 7. Allow an ACL "warn" statement to add header lines at the start or after all
3970 the Received: headers, as well as at the end.
3972 8. Added the rcpt_4xx retry error code.
3974 9. Added postmaster_mailfrom=xxx to callout verification option.
3976 10. Added mailfrom=xxxx to the callout verification option, for verify=
3979 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3980 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3982 12. Inserted some casts to stop certain compilers warning when using pointer
3983 differences as field lengths or precisions in printf-type calls (mostly
3984 affecting debugging statements).
3986 13. Added optional readline() support for -be (dynamically loaded).
3988 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3989 same clock tick as a message's arrival, so that its received time was the
3990 same as the "first fail" time on the retry record, and that message
3991 remained on the queue past the ultimate address timeout, every queue runner
3992 would try a delivery (because it was past the ultimate address timeout) but
3993 after another failure, the ultimate address timeout, which should have then
3994 bounced the address, did not kick in. This was a "< instead of <=" error;
3995 in most cases the first failure would have been in the next clock tick
3996 after the received time, and all would be well.
3998 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3999 being recognized when the domain list was tested by the match_domain
4000 condition in an expansion string.
4002 16. Added the ${str2b64: operator.
4004 17. Exim was always calling setrlimit() to set a large limit for the number of
4005 processes, without checking whether the existing limit was already
4006 adequate. (It did check for the limit on file descriptors.) Furthermore,
4007 errors from getrlimit() and setrlimit() were being ignored. Now they are
4008 logged to the main and panic logs, but Exim does carry on, to try to do its
4009 job under whatever limits there are.
4011 18. Imported PCRE 5.0.
4013 19. Trivial typo in log message " temporarily refused connection" (the leading
4016 20. If the log selector return_path_on_delivery was set and an address was
4017 redirected to /dev/null, the delivery process crashed because it assumed
4018 that a return path would always be set for a "successful" delivery. In this
4019 case, the whole delivery is bypassed as an optimization, and therefore no
4022 21. Internal re-arrangement: the function for sending a challenge and reading
4023 a response while authentication was assuming a zero-terminated challenge
4024 string. It's now changed to take a pointer and a length, to allow for
4025 binary data in such strings.
4027 22. Added the cyrus_sasl authenticator (code supplied by MBM).
4029 23. Exim was not respecting finduser_retries when seeking the login of the
4030 uid under which it was called; it was always trying 10 times. (The default
4031 setting of finduser_retries is zero.) Also, it was sleeping after the final
4032 failure, which is pointless.
4034 24. Implemented tls_on_connect_ports.
4036 25. Implemented acl_smtp_predata.
4038 26. If the domain in control=submission is set empty, Exim assumes that the
4039 authenticated id is a complete email address when it generates From: or
4040 Sender: header lines.
4042 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
4043 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
4044 chown and chgrp in /bin and hostname in /usr/bin.
4046 28. Exim was keeping the "process log" file open after each use, just as it
4047 does for the main log. This opens the possibility of it remaining open for
4048 long periods when the USR1 signal hits a daemon. Occasional processlog
4049 errors were reported, that could have been caused by this. Anyway, it seems
4050 much more sensible not to leave this file open at all, so that is what now
4053 29. The long-running daemon process does not normally write to the log once it
4054 has entered its main loop, and it closes the log before doing so. This is
4055 so that log files can straightforwardly be renamed and moved. However,
4056 there are a couple of unusual error situations where the daemon does write
4057 log entries, and I had neglected to close the log afterwards.
4059 30. The text of an SMTP error response that was received during a remote
4060 delivery was being truncated at 512 bytes. This is too short for some of
4061 the long messages that one sometimes sees. I've increased the limit to
4064 31. It is now possible to make retry rules that apply only when a message has a
4065 specific sender, in particular, an empty sender.
4067 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
4068 it possible to be selective about when SMTP synchronization is enforced.
4070 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
4072 32. Implemented hosts_connection_nolog.
4074 33. Added an ACL for QUIT.
4076 34. Setting "delay_warning=" to disable warnings was not working; it gave a
4079 35. Added mailbox_size and mailbox_filecount to appendfile.
4081 36. Added control = no_multiline_responses to ACLs.
4083 37. There was a bug in the logic of the code that waits for the clock to tick
4084 in the case where the clock went backwards by a substantial amount such
4085 that the microsecond fraction of "now" was more than the microsecond
4086 fraction of "then" (but the whole seconds number was less).
4088 38. Added support for the libradius Radius client library this is found on
4089 FreeBSD (previously only the radiusclient library was supported).
4095 1. When certain lookups returned multiple values in the form name=value, the
4096 quoting of the values was not always being done properly. Specifically:
4097 (a) If the value started with a double quote, but contained no whitespace,
4099 (b) If the value contained whitespace other than a space character (i.e.
4100 tabs or newlines or carriage returns) it was not quoted.
4101 This fix has been applied to the mysql and pgsql lookups by writing a
4102 separate quoting function and calling it from the lookup code. The fix
4103 should probably also be applied to nisplus, ibase and oracle lookups, but
4104 since I cannot test any of those, I have not disturbed their existing code.
4106 2. A hit in the callout cache for a specific address caused a log line with no
4107 reason for rejecting RCPT. Now it says "Previous (cached) callout
4108 verification failure".
4110 3. There was an off-by-one bug in the queryprogram router. An over-long
4111 return line was truncated at 256 instead of 255 characters, thereby
4112 overflowing its buffer with the terminating zero. As well as fixing this, I
4113 have increased the buffer size to 1024 (and made a note to document this).
4115 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
4116 when Exim is waiting for an SMTP response from a remote server, Exim
4117 restarts its select() call on the socket, thereby resetting its timeout.
4118 This is not a problem when such interrupts are rare. Somebody set up a cron
4119 job to run exiwhat every 2 minutes, which is less than the normal select()
4120 timeout (5 or 10 minutes). This meant that the select() timeout never
4121 kicked in because it was always reset. I have fixed this by comparing the
4122 time when an interrupt arrives with the time at the start of the first call
4123 to select(). If more time than the timeout has elapsed, the interrupt is
4124 treated as a timeout.
4126 5. Some internal re-factoring in preparation for the addition of Sieve
4127 extensions (by MH). In particular, the "personal" test is moved to a
4128 separate function, and given an option for scanning Cc: and Bcc: (which is
4129 not set for Exim filters).
4131 6. When Exim created an email address using the login of the caller as the
4132 local part (e.g. when creating a From: or Sender: header line), it was not
4133 quoting the local part when it contained special characters such as @.
4135 7. Installed new OpenBSD configuration files.
4137 8. Reworded some messages for syntax errors in "and" and "or" conditions to
4138 try to make them clearer.
4140 9. Callout options, other than the timeout value, were being ignored when
4141 verifying sender addresses in header lines. For example, when using
4143 verify = header_sender/callout=no_cache
4145 the cache was (incorrectly) being used.
4147 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4148 only the Cygwin environment.
4150 11. When return_path_on_delivery was set as a log selector, if different remote
4151 addresses in the same message used different return paths and parallel
4152 remote delivery occurred, the wrong values would sometimes be logged.
4153 (Whenever a remote delivery process finished, the return path value from
4154 the most recently started remote delivery process was logged.)
4156 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4157 lines when AUTH and/or TLS are in use. This is the "received protocol"
4158 field. Exim used to use "asmtp" for authenticated SMTP, without any
4159 indication (in the protocol name) for TLS use. Now it follows the RFC and
4160 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4161 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4162 names appear in log lines as well as in Received: header lines.
4164 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4165 extensions, and comparison tests, and to fix some bugs.
4167 14. Changes to the "personal" filter test:
4169 (1) The test was buggy in that it was just doing the equivalent of
4170 "contains" tests on header lines. For example, if a user's address was
4171 anne@some.where, the "personal" test would incorrectly be true for
4173 To: susanne@some.where
4175 This test is now done by extracting each address from the header in turn,
4176 and checking the entire address. Other tests that are part of "personal"
4177 are now done using regular expressions (for example, to check local parts
4178 of addresses in From: header lines).
4180 (2) The list of non-personal local parts in From: addresses has been
4181 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4182 taken from the Sieve specification recommendations.
4184 (3) If the message contains any header line starting with "List-" it is
4185 treated as non-personal.
4187 (4) The test for "circular" in the Subject: header line has been removed
4188 because it now seems ill-conceived.
4190 15. Minor typos in src/EDITME comments corrected.
4192 16. Installed latest exipick from John Jetmore.
4194 17. If headers_add on a router specified a text string that was too long for
4195 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4196 of string_sprintf() is now avoided.
4198 18. $message_body_size was not set (it was always zero) when running the DATA
4199 ACL and the local_scan() function.
4201 19. For the "mail" command in an Exim filter, no default was being set for
4202 the once_repeat time, causing a random time value to be used if "once" was
4203 specified. (If the value happened to be <= 0, no repeat happened.) The
4204 default is now 0s, meaning "never repeat". The "vacation" command was OK
4205 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4206 (I found it when inspecting the code).
4208 20. There is now an overall timeout for performing a callout verification. It
4209 defaults to 4 times the callout timeout, which applies to individual SMTP
4210 commands during the callout. The overall timeout applies when there is more
4211 than one host that can be tried. The timeout is checked before trying the
4212 next host. This prevents very long delays if there are a large number of
4213 hosts and all are timing out (e.g. when the network connections are timing
4214 out). The value of the overall timeout can be changed by specifying an
4215 additional sub-option for "callout", called "maxwait". For example:
4217 verify = sender/callout=5s,maxwait=20s
4219 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4220 to the end before writing, but this should make it even safer).
4222 22. Exim was forgetting that it had advertised PIPELINING for the second and
4223 subsequent messages on an SMTP connection. It was also not resetting its
4224 memory on STARTTLS and an internal HELO.
4226 23. When Exim logs an SMTP synchronization error within a session, it now
4227 records whether PIPELINING has been advertised or not.
4229 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4230 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4231 rather than long int.
4233 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4235 26. Added the never_mail option to autoreply.
4241 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4242 crash if the getsockname() call failed; this can happen if a connection is
4243 closed very soon after it is established. The problem was simply in the
4244 order in which certain operations were done, causing Exim to try to write
4245 to the SMTP stream before it had set up the file descriptor. The bug has
4246 been fixed by making things happen in the correct order.
4252 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4253 before the connection was closed, thus losing the rejection response.
4255 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4256 some early Solaris releases, but causes trouble in current releases where
4257 socklen_t is defined.
4259 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4262 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4265 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4266 that releasing the top of it at the end releases what was used for sub-
4267 expansions (unless the block got too big). However, discard this block if
4268 the first thing is a variable or header, so that we can use its block when
4269 it is dynamic (useful for very large $message_headers, for example).
4271 6. Lookups now cache *every* query, not just the most recent. A new, separate
4272 store pool is used for this. It can be recovered when all lookup caches are
4273 flushed. Lookups now release memory at the end of their result strings.
4274 This has involved some general refactoring of the lookup sources.
4276 7. Some code has been added to the store_xxx() functions to reduce the amount
4277 of flapping under certain conditions.
4279 8. log_incoming_interface used to affect only the <= reception log lines. Now
4280 it causes the local interface and port to be added to several more SMTP log
4281 lines, for example "SMTP connection from", and rejection lines.
4283 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4285 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4287 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4288 could be overwritten at the end of the current message (or the start of a
4289 new message if it was set in a HELO ACL). The value is now preserved for
4290 the duration of the SMTP connection.
4292 12. If a transport had a headers_rewrite setting, and a matching header line
4293 contained an unqualified address, that address was qualified, even if it
4294 did not match any rewriting rules. The underlying bug was that the values
4295 of the flags that permit the existence of unqualified sender and recipient
4296 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4297 non-local messages, and by -bnq for local messages) were not being
4298 preserved with the message after it was received.
4300 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4301 "next input=" as part of the text comprising the host identity instead of
4302 the correct text. The code was using the same buffer for two different
4303 strings. However, depending on which order the printing function evaluated
4304 its arguments, the bug did not always show up. Under Linux, for example, my
4305 test suite worked just fine.
4307 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4308 doesn't work with some older versions of Perl. It has been changed to "my",
4309 which in any case is probably the better facility to use.
4311 15. A really picky compiler found some instances of statements for creating
4312 error messages that either had too many or two few arguments for the format
4315 16. The size of the buffer for calls to the DNS resolver has been increased
4316 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4317 for addresses that have a lot of PTR records. This alleviates a problem; it
4318 does not fully solve it.
4320 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4321 buffer now truncates the list and logs the incident, which is the same
4322 action as happens when Exim is looking up a host name and its aliases.
4323 Previously in this situation something unpredictable would happen;
4324 sometimes it was "internal error: store_reset failed".
4326 18. If a server dropped the connection unexpectedly when an Exim client was
4327 using GnuTLS and trying to read a response, the client delivery process
4328 crashed while trying to generate an error log message.
4330 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4331 string, for example:
4333 warn message = H1: something\nH2: something
4335 the text was added as a single header line from Exim's point of view
4336 though it ended up OK in the delivered message. However, searching for the
4337 second and subsequent header lines using $h_h2: did not work. This has been
4338 fixed. Similarly, if a system filter added multiple headers in this way,
4339 the routers could not see them.
4341 20. Expanded the error message when iplsearch is called with an invalid key to
4342 suggest using net-iplsearch in a host list.
4344 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4345 ACLs are no longer actually imposed (and a message to that effect is
4348 22. If a "gecos" field in a passwd entry contained escaped characters, in
4349 particular, if it contained a \" sequence, Exim got it wrong when building
4350 a From: or a Sender: header from that name. A second bug also caused
4351 incorrect handling when an unquoted " was present following a character
4352 that needed quoting.
4354 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4355 was not being matched caselessly.
4357 24. Arranged for all hyphens in the exim.8 source to be escaped with
4360 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4361 properly. Recipient callout cache records were still being keyed to include
4362 the sender, even when use_sender was set false. This led to far more
4363 callouts that were necessary. The sender is no longer included in the key
4364 when use_sender is false.
4366 26. Added "control = submission" modifier to ACLs.
4368 27. Added the ${base62d: operator to decode base 62 numbers.
4370 28. dnsdb lookups can now access SRV records.
4372 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4373 the configuration file.
4375 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4376 (-v) mode. This makes the output for a verbose queue run more intelligible.
4378 31. Added a use_postmaster feature to recipient callouts.
4380 32. Added the $body_zerocount variable, containing the number of binary zero
4381 bytes in the message body.
4383 33. The time of last modification of the "new" subdirectory is now used as the
4384 "mailbox time last read" when there is a quota error for a maildir
4387 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4389 35. Added +ignore_unknown as a special item in host lists.
4391 36. Code for decoding IPv6 addresses in host lists is now included, even if
4392 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4393 address was recognized as an IP address, but was then not correctly decoded
4394 into binary, causing unexpected and incorrect effects when compared with
4401 1. Very minor rewording of debugging text in manualroute to say "list of
4402 hosts" instead of "hostlist".
4404 2. If verify=header_syntax was set, and a header line with an unqualified
4405 address (no domain) and a large number of spaces between the end of the
4406 name and the colon was received, the reception process suffered a buffer
4407 overflow, and (when I tested it) crashed. This was caused by some obsolete
4408 code that should have been removed. The fix is to remove it!
4410 3. When running in the test harness, delay a bit after writing a bounce
4411 message to get a bit more predictability in the log output.
4413 4. Added a call to search_tidyup() just before forking a reception process. In
4414 theory, someone could use a lookup in the expansion of smtp_accept_max_
4415 per_host which, without the tidyup, could leave open a database connection.
4417 5. Added the variables $recipient_data and $sender_data which get set from a
4418 lookup success in an ACL "recipients" or "senders" condition, or a router
4419 "senders" option, similar to $domain_data and $local_part_data.
4421 6. Moved the writing of debug_print from before to after the "senders" test
4424 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4425 problems for message scanning, either using a data ACL, or using
4426 local_scan() because the Received: header was not generated till after they
4427 were called (in order to set the time as the time of reception completion).
4428 I have revised the way this works. The header is now generated after the
4429 body is received, but before the ACL or local_scan() are called. After they
4430 are run, the timestamp in the header is updated.
4436 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4437 before starting a queue runner without re-exec. This happened only when
4438 deliver_drop_privilege was set or when the Exim user was set to root. The
4439 effect of the bug was that timeouts during subsequent deliveries caused
4440 crashes instead of being properly handled. The handler is now left at its
4441 default (and expected) setting.
4443 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4444 message, again when deliver_drop_privilege is set or Exim is run as root.
4445 The bug described in (1) was not present in this case, but the tidying up
4446 of the other signals was missing. I have made the two cases consistent.
4448 3. The ignore_target_hosts setting on a manualroute router was being ignored
4449 for hosts that were looked up using the /MX notation.
4451 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4454 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4455 operated on the sender address. After changing the $sender_address to <>
4456 for the sender address verify, Exim was re-instated it as the original
4457 (before rewriting) address, but remembering that it had rewritten it, so it
4458 wasn't rewriting it again. This bug also had the effect of breaking the
4459 sender address verification caching when the sender address was rewritten.
4461 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4462 This has been changed so that if the ip literal address matches
4463 ignore_target_hosts, the router declines.
4465 7. Added expansion conditions match_domain, match_address, and match_local_
4466 part (NOT match_host).
4468 8. The placeholder for the Received: header didn't have a length field set.
4470 9. Added code to Exim itself and to exim_lock to test for a specific race
4471 condition that could lead to file corruption when using MBX delivery. The
4472 issue is with the lockfile that is created in /tmp. If this file is removed
4473 after a process has opened it but before that process has acquired a lock,
4474 there is the potential for a second process to recreate the file and also
4475 acquire a lock. This could lead to two Exim processes writing to the file
4476 at the same time. The added code performs the same test as UW imapd; it
4477 checks after acquiring the lock that its file descriptor still refers to
4478 the same named file.
4480 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4481 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4484 11. Added the smtp_active_hostname option. If used, this will typically be made
4485 to depend on the incoming interface address. Because $interface_address is
4486 not set up until the daemon has forked a reception process, error responses
4487 that can happen earlier (such as "too many connections") no longer contain
4490 12. If an expansion in a condition on a "warn" statement fails because a lookup
4491 defers, the "warn" statement is abandoned, and the next ACL statement is
4492 processed. Previously this caused the whole ACL to be aborted.
4494 13. Added the iplsearch lookup type.
4496 14. Added ident_timeout as a log selector.
4498 15. Added tls_certificate_verified as a log selector.
4500 16. Added a global option tls_require_ciphers (compare the smtp transport
4501 option of the same name). This controls incoming TLS connections.
4503 17. I finally figured out how to make tls_require_ciphers do a similar thing
4504 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4505 before starting the TLS session.
4507 18. Tabs are now shown as \t in -bP output.
4509 19. If the log selector return_path_on_delivery was set, Exim crashed when
4510 bouncing a message because it had too many Received: header lines.
4512 20. If two routers both had headers_remove settings, and the first one included
4513 a superfluous trailing colon, the final name in the first list and the
4514 first name in the second list were incorrectly joined into one item (with a
4515 colon in the middle).
4521 1. Added -C and -D options to the exinext utility, mainly to make it easier
4522 to include in the automated testing, but these could be helpful when
4523 multiple configurations are in use.
4525 2. The exinext utility was not formatting the output nicely when there was
4526 an alternate port involved in the retry record key, nor when there was a
4527 message id as well (for retries that were specific to a specific message
4528 and a specific host). It was also confused by IPv6 addresses, because of
4529 the additional colons they contain. I have fixed the IPv4 problem, and
4530 patched it up to do a reasonable job for IPv6.
4532 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4533 delivery, the log line now contains "pipelined" if PIPELINING was used.
4535 4. An SMTP transport process used to panic and die if the bind() call to set
4536 an explicit outgoing interface failed. This has been changed; it is now
4537 treated in the same way as a connect() failure.
4539 5. A reference to $sender_host_name in the part of a conditional expansion
4540 that was being skipped was still causing a DNS lookup. This no longer
4543 6. The def: expansion condition was not recognizing references to header lines
4544 that used bh_ and bheader_.
4546 7. Added the _cache feature to named lists.
4548 8. The code for checking quota_filecount in the appendfile transport was
4549 allowing one more file than it should have been.
4551 9. For compatibility with Sendmail, the command line option
4559 and sets the incoming protocol and host name (for trusted callers). The
4560 host name and its colon can be omitted when only the protocol is to be set.
4561 Note the Exim already has two private options, -pd and -ps, that refer to
4562 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4563 "s", but I don't think that's a major issue.
4565 10. A number of refactoring changes to the code, none of which should affect
4568 (a) The number of logging options was getting close to filling up the
4569 32-bit word that was used as a bit map. I have split them into two classes:
4570 those that are passed in the argument to log_write(), and those that are
4571 only ever tested independently outside of that function. These are now in
4572 separate 32-bit words, so there is plenty of room for expansion again.
4573 There is no change in the user interface or the logging behaviour.
4575 (b) When building, for example, log lines, the code previously used a
4576 macro that called string_cat() twice, in order to add two strings. This is
4577 not really sufficiently general. Furthermore, there was one instance where
4578 it was actually wrong because one of the argument was used twice, and in
4579 one call a function was used. (As it happened, calling the function twice
4580 did not affect the overall behaviour.) The macro has been replaced by a
4581 function that can join an arbitrary number of extra strings onto a growing
4584 (c) The code for expansion conditions now uses a table and a binary chop
4585 instead of a serial search (which was left over from when there were very
4586 few conditions). Also, it now recognizes conditions like "pam" even when
4587 the relevant support is not compiled in: a suitably worded error message is
4588 given if an attempt is made to use such a condition.
4590 11. Added ${time_interval:xxxxx}.
4592 12. A bug was causing one of the ddress fields not to be passed back correctly
4593 from remote delivery subprocesses. The field in question was not being
4594 subsequently used, so this caused to problems in practice.
4596 13. Added new log selectors queue_time and deliver_time.
4598 14. Might have fixed a bug in maildirsizefile handling that threw up
4599 "unexpected character" debug warnings, and recalculated the data
4600 unnecessarily. In any case, I expanded the warning message to give more
4603 15. Added the message "Restricted characters in address" to the statements in
4604 the default ACL that block characters like @ and % in local parts.
4606 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4607 Three changes have been made:
4609 (a) There was a serious bug; a negative response to MAIL caused the whole
4610 recipient domain to be cached as invalid, thereby blocking all messages
4611 to all local parts at the same domain, from all senders. This bug has
4612 been fixed. The domain is no longer cached after a negative response to
4613 MAIL if the sender used is not empty.
4615 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4618 (c) A new callout option, "use_sender" has been added for people who want
4619 the modified behaviour.
4625 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4628 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4629 indeed breaks things for older releases.
4631 3. Added additional logging to the case where there is a problem reading data
4632 from a filter that is running in a subprocess using a pipe, in order to
4633 try to track down a specific problem.
4635 4. Testing facility fudge: when running in the test harness and attempting
4636 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4637 getting "No route to host". Convert this to a timeout.
4639 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4642 6. Some OS don't have socklen_t but use size_t instead. This affects the
4643 fifth argument of getsockopt() amongst other things. This is now
4644 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4645 can be set for individual OS. I have set it for SunOS5, OSF1, and
4646 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4647 some earlier ones do not.
4649 7. Change 4.30/15 was not doing the test caselessly.
4651 8. The standard form for an IPv6 address literal was being rejected by address
4652 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4653 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4654 this, as well as the form without the "IPv6" on the front (but only when
4655 address literals are enabled, of course).
4657 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4659 10. Exim crashed if a message with an empty sender address specified by -f
4660 encountered a router with an errors_to setting. This could be provoked only
4661 by a command such as
4665 where an empty string was supplied; "<>" did not hit this bug.
4667 11. Installed PCRE release 4.5.
4669 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4670 remained set. It is now erased.
4672 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4673 times from message ids (which are base 36 rather than the normal 62).
4675 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4676 were being counted as actual protocol errors, and logged if the log
4677 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4678 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4679 and DATA following a set of rejected RCPTs do not count as protocol errors.
4680 In other words, Exim assumes they were pipelined, though this may not
4681 actually be the case. Of course, in all cases the client gets an
4682 appropriate error code.
4684 15. If a lookup fails in an ACL condition, a message about the failure may
4685 be available; it is used if testing the ACL cannot continue, because most
4686 such messages specify what the cause of the deferral is. However, some
4687 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4688 that caused an old message to be retained and used if a later statement
4689 caused a defer, replacing the real cause of the deferral.
4691 16. If an IP address had so many PTR records that the DNS lookup buffer
4692 was not large enough to hold them, Exim could crash while trying to process
4693 the truncated data. It now detects and logs this case.
4695 17. Further to 4.21/58, another change has been made: if (and only if) the
4696 first line of a message (the first header line) ends with CRLF, a bare LF
4697 in a subsequent header line has a space inserted after it, so as not to
4698 terminate the header.
4700 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4701 unnecessarily, used atoi() instead of strtol(), and didn't check the
4702 termination when getting file sizes from file names by regex.
4704 19. Completely re-implemented the support for maildirsize files, in the light
4705 of a number of problems with the previous contributed implementation
4706 (4.30/29). In particular:
4708 . If the quota is zero, the maildirsize file is maintained, but no quota is
4711 . If the maildir directory does not exist, it is created before any attempt
4712 to write a maildirsize file.
4714 . The quota value in the file is just a cache; if the quota is changed in
4715 the transport, the new value overrides.
4717 . A regular expression is available for excluding directories from the
4720 20. The autoreply transport checks the characters in options that define the
4721 message's headers; it allows continued headers, but it was checking with
4722 isspace() after an embedded newline instead of explicitly looking for a
4725 21. If all the "regular" hosts to which an address was routed had passed their
4726 expiry times, and had not reached their retry times, the address was
4727 bounced, even if fallback hosts were defined. Now Exim should go on to try
4730 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4731 equivalent code in the SMTP transport. Some hosts send humungous responses
4732 to HELO/EHLO, more than 1024 it seems.
4734 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4735 gives compiler warnings in some environments. I've now done it "properly",
4738 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4739 (because of problems with the built-in one) was declared to return uschar *
4740 instead of char *, causing compiler failure.
4742 25. Fixed a file descriptor leak when processing alias/forward files.
4744 26. Fixed a minor format string issue in dbfn.c.
4746 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4748 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4749 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4752 29. When a sender address is verified, it is cached, to save repeating the test
4753 when there is more than one recipient in a message. However, when the
4754 verification involves a callout, it is possible for different callout
4755 options to be set for different recipients. It is too complicated to keep
4756 track of this in the cache, so now Exim always runs a verification when a
4757 callout is required, relying on the callout cache for the optimization.
4758 The overhead is duplication of the address routing, but this should not be
4761 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4762 to be verified with callout=postmaster, and the main callout worked but the
4763 postmaster check failed, the verification correctly failed. However, if a
4764 subsequent RCPT command asked for sender verification *without* the
4765 postmaster check, incorrect caching caused this verification also to fail,
4768 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4769 it was not caching the DNS options (qualify_single, search_parents) that
4770 were used when the lookup failed. A subsequent lookup with different
4771 options therefore always gave the same answer, though there were cases
4772 where it should not have. (Example: a "domains = !$mx_any" option on a
4773 dnslookup router: the "domains" option is always processed without any
4774 widening, but the router might have qualify_single set.) Now Exim uses the
4775 cached value only when the same options are set.
4777 32. Added John Jetmore's "exipick" utility to the distribution.
4779 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4780 than a timeout (e.g. a certificate is required, and is not provided), an
4781 Exim server now closes the connection immediately. Previously it waited for
4782 the client to close - but if the client is SSL, it seems that they each
4783 wait for each other, leading to a delay before one of them times out.
4785 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4786 maintained 0.8.x compatibility because I don't think many are using it, and
4787 it is clearly obsolete.
4789 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4792 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4793 client certificate was expired. A simple patch fixes this, though I don't
4794 understand the full logic of why the verify callback is called multiple
4797 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4798 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4799 which causes problems with some clients (such as the Certicom SSL Plus
4800 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4801 disables the coutermeasure allowing Eudora to connect."
4803 38. Exim was not checking that a write() to a log file succeeded. This could
4804 lead to Bad Things if a log got too big, in particular if it hit a file
4805 size limit. Exim now panics and dies if it cannot write to a log file, just
4806 as it does if it cannot open a log file.
4808 39. Modified OS/Makefile-Linux so that it now contains
4810 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4812 The two -D definitions ensure that Exim is compiled with large file
4813 support, which makes it possible to handle log files that are bigger than
4816 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4817 instance) a domain was checked against a named list that involved a lookup,
4818 causing $domain_data to be set, then another domain was checked against the
4819 same list, then the first domain was re-checked, the value of $domain_data
4820 after the final check could be wrong. In particular, if the second check
4821 failed, it could be set empty. This bug probably also applied to
4824 41. The strip_trailing_dot option was not being applied to the address given
4825 with the -f command-line option.
4827 42. The code for reading a message's header from the spool was incrementing
4828 $received_count, but never initializing it. This meant that the value was
4829 incorrect (doubled) while delivering a message in the same process in which
4830 it was received. In the most common configuration of Exim, this never
4831 happens - a fresh exec is done - but it can happen when
4832 deliver_drop_privilege is set.
4834 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4835 it now includes up to 150 characters of the unexpected data in the log
4838 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4839 and building data strings. The size of both of these buffers was 10 000
4840 bytes - far larger than anybody would *ever* want, thought I. Needless to
4841 say, somebody hit the limit. I have increased the maximum line length to
4842 20 000 and the maximum data length of concatenated lines to 100 000. I have
4843 also fixed two bugs, because there was no checking on these buffers. Tsk,
4844 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4845 buffer is too small.
4847 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4848 lsearch lookups. Now it does.
4850 46. When parsing a route_list item in a manualroute router, a fixed-length
4851 buffer was used for the list of hosts. I made this 1024 bytes long,
4852 thinking that nobody would ever have a list of hosts that long. Wrong.
4853 Somebody had a whole pile of complicated expansion conditions, and the
4854 string was silently truncated, leading to an expansion error. It turns out
4855 that it is easier to change to an unlimited length (owing to other changes
4856 that have happened since this code was originally written) than to build
4857 structure for giving a limitation error. The length of the item that
4858 expands into the list of hosts is now unlimited.
4860 47. The lsearch lookup could not handle data where the length of text line was
4861 more than 4095 characters. Such lines were truncated, leading to shortened
4862 data being returned. It should now handle lines of any length.
4864 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4865 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4868 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4869 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4870 no longer happens in comment lines. A list of replacements is now placed
4871 at the head of all of the source files, except those whose only change is
4872 to replace PERL_COMMAND in the very first #! line.
4874 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4875 messages on the queue, with a bottom-up merge sort, using code contributed
4876 by Michael Haardt. This should make operations like -bp somewhat faster on
4877 large queues. It won't affect queue runners, except when queue_run_in_order
4880 51. Installed eximstats 1.31 in the distribution.
4882 52. Added support for SRV lookups to the dnslookup router.
4884 53. If an ACL referred to $message_body or $message_body_end, the value was not
4885 reset for any messages that followed in the same SMTP session.
4887 54. The store-handling optimization for building very long strings was not
4888 differentiating between the different store pools. I don't think this
4889 actually made any difference in practice, but I've tidied it.
4891 55. While running the routers to verify a sender address, $sender_address
4892 was still set to the sender address. This is wrong, because when routing to
4893 send a bounce to the sender, it would be empty. Therefore, I have changed
4894 it so that, while verifying a sender address, $sender_address is set to <>.
4895 (There is no change to what happens when verifying a recipient address.)
4897 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4898 target A or AAAA records (if not already returned) without resetting the
4899 qualify_single or search_parents options of the DNS resolver. These are
4900 inappropriate in this case because the targets of MX and SRV records must
4901 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4902 target that, when qualified, matched something in the local domain. These
4903 two options are now turned off when doing these lookups.
4905 57. It seems that at least some releases of Reiserfs (which does not have the
4906 concept of a fixed number of inodes) returns zero and not -1 for the
4907 number of available inodes. This interacted badly with check_spool_inodes,
4908 which assumed that -1 was the "no such thing" setting. What I have done is
4909 to check that the total number of inodes is greater than zero before doing
4910 the test of how many are available.
4912 58. When a "warn" ACL statement has a log_message modifier, the message is
4913 remembered, and not repeated. This is to avoid a lot of repetition when a
4914 message has many recipients that cause the same warning to be written.
4915 Howewer, Exim was preserving the list of already written lines for an
4916 entire SMTP session, which doesn't seem right. The memory is now reset if a
4917 new message is started.
4919 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4920 debugging output unless log_rewrite was also set.
4922 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4923 of (int)(handle) when we know that handle contains (void *)(-1).
4925 61. The Exim daemon panic-logs an error return when it closes the incoming
4926 connection. However "connection reset by peer" seems to be common, and
4927 isn't really an error worthy of noting specially, so that particular error
4930 62. When Exim is trying to find all the local interfaces, it used to panic and
4931 die if the ioctl to get the interface flags failed. However, it seems that
4932 on at least one OS (Solaris 9) it is possible to have an interface that is
4933 included in the list of interfaces, but for which you get a failure error
4934 for this call. This happens when the interface is not "plumbed" into a
4935 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4936 failure of the "get flags" call assumes that the interface is down.
4938 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4939 makes life easier for people who are doing arithmetic on fields extracted
4940 from dates, where you often get leading zeros that should not be
4941 interpreted as octal.
4943 64. Added qualify_domain to the redirect router, to override the global
4946 65. If a pathologically long header line contained very many addresses (the
4947 report of this problem mentioned 10 000) and each of them was rewritten,
4948 Exim could use up a very large amount of memory. (It kept on making new
4949 copies of the header line as it rewrote, and never released the old ones.)
4950 At the expense of a bit more processing, the header rewriting function has
4951 been changed so that it no longer eats memory in this way.
4953 66. The generation of the Received: header has been moved from the time that a
4954 message starts to be received, to the time that it finishes. The timestamp
4955 in the Received: header should now be very close to that of the <= log
4956 line. There are two side-effects of this change:
4958 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4959 logged header lines no longer include the local Received: line, because
4960 it has not yet been created. The same applies to a copy of the message
4961 that is returned to a non-SMTP sender when a message is rejected.
4963 (b) When a filter file is tested using -bf, no additional Received: header
4964 is added to the test message. After some thought, I decided that this
4967 This change does not affect the value of $received_for. It is still set
4968 after address rewriting, but before local_scan() is called.
4970 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4972 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4973 gave an unhelpful panic error message, and a defer error. I have managed to
4974 change this behaviour so that it now rejects any supplied certificate,
4975 which seems right, as the list of acceptable certificates is empty.
4977 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4978 gave an unhelpful defer error. I have not managed to make this reject any
4979 supplied certificates, but the error message it gives is "no certificate
4980 supplied", which is not helpful.
4982 70. exigrep's output now also includes lines that are not associated with any
4983 message, but which match the given pattern. Implemented by a patch from
4984 Martin Sluka, which also tidied up the Perl a bit.
4986 71. Recipient callout verification, like sender verification, was using <> in
4987 the MAIL FROM command. This isn't really the right thing, since the actual
4988 sender may affect whether the remote host accepts the recipient or not. I
4989 have changed it to use the actual sender in the callout; this means that
4990 the cache record is now keyed on a recipient/sender pair, not just the
4991 recipient address. There doesn't seem to be a real danger of callout loops,
4992 since a callout by the remote host to check the sender would use <>.
4993 [SEE ABOVE: changed after hitting problems.]
4995 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4996 temporary errors. However, in the case of such a code being given after
4997 the end of a data transmission (i.e. after ".") Exim was failing to write
4998 a retry record for the message. (Yes, there was some broken host that was
4999 actually sending 8xx at this point.)
5001 73. An unknown lookup type in a host list could cause Exim to panic-die when
5002 the list was checked. (An example that provoked this was putting <; in the
5003 middle of a list instead of at the start.) If this happened during a DATA
5004 ACL check, a -D file could be left lying around. This kind of configuration
5005 error no longer causes Exim to die; instead it causes a defer errror. The
5006 incident is still logged to the main and panic logs.
5008 74. Buglet left over from Exim 3 conversion. The message "too many messages
5009 in one connection" was written to the rejectlog but not the mainlog, except
5010 when address rewriting (yes!) was being logged.
5012 75. Added write_rejectlog option.
5014 76. When a system filter was run not as root (that is, when system_filter_user
5015 was set), the values of the $n variables were not being returned to the
5016 main process; thus, they were not subsequently available in the $sn
5019 77. Added +return_path_on_delivery log selector.
5021 78. A connection timeout was being treated differently from recipients deferred
5022 when testing hosts_max_try with a message that was older than the host's
5023 retry timeout. (The host should not be counted, thus allowing all hosts to
5024 be tried at least once before bouncing.) This may have been the cause of an
5025 occasionally reported bug whereby a message would remain on the queue
5026 longer than the retry timeout, but would be bounced if a delivery was
5027 forced. I say "may" because I never totally pinned down the problem;
5028 setting up timeout/retry tests is difficult. See also the next item.
5030 79. The ultimate address timeout was not being applied to errors that involved
5031 a combination of host plus message (for example, a timeout on a MAIL
5032 command). When an address resolved to a number of possible hosts, and they
5033 were not all tried for each delivery (e.g. because of hosts_max_try), a
5034 message could remain on the queue longer than the retry timeout.
5036 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
5039 81. Fixed an obscure SMTP outgoing bug which required at least the following
5040 conditions: (a) there was another message waiting for the same server;
5041 (b) the server returned 5xx to all RCPT commands in the first message so
5042 that the message was not completed; (c) the server dropped the connection
5043 or gave a negative response to the RSET that Exim sends to abort the
5044 transaction. The observed case was a dropped connection after DATA that had
5045 been sent in pipelining mode. That is, the server had advertised PIPELINING
5046 but was not implementing it correctly. The effect of the bug was incorrect
5047 behaviour, such as trying another host, and this could lead to a crash.
5053 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
5054 and daemon.c were passed as pointers to ints; they should have been
5055 pointers to socklen_t variables (which are typically unsigned ints).
5057 2. Some signed/unsigned type warnings in the os.c file for Linux have been
5060 3. Fixed a really odd bug that affected only the testing scheme; patching a
5061 certain fixed string in the binary changed the value of another string that
5062 happened to be identical to the end of the original first string.
5064 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
5065 name", it returns that address as the IP address. On some operating
5066 systems (e.g. Solaris), it also passes back the IP address string as the
5067 "host name". However, on others (e.g. Linux), it passes back an empty
5068 string. Exim wasn't checking for this, and was changing the host name to an
5069 empty string, assuming it had been canonicized.
5071 5. Although rare, it is permitted to have more than one PTR record for a given
5072 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
5073 all the names associated with an address, because they do in Solaris.
5074 However, it seems that they do not in Linux for data that comes from the
5075 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
5076 I found this out when I moved to a new Linux workstation and tried to run
5077 the Exim test suite.
5079 To get round this problem I have changed the code so that it now does its
5080 own call to the DNS to look up PTR records when searching for a host name.
5081 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
5082 addresses that are only in /etc/hosts are still found.
5084 This behaviour is, however, controlled by an option called host_lookup_
5085 order, which defaults to "bydns:byaddr". If people want to use the other
5086 order, or indeed, just use one or the other means of lookup, they can
5087 specify it in this variable.
5089 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
5090 some operating systems, this comes back from gethostbyaddr() as an empty
5091 string, and this is what Exim used to test for. However, it seems that in
5092 other systems, "." is yielded. Exim now tests for this case too.
5094 7. The values of check_spool_space and check_log_space are now held internally
5095 as a number of kilobytes instead of an absolute number of bytes. If a
5096 numbers is specified without 'K' or 'M', it is rounded up to the nearest
5097 kilobyte. This means that much larger values can be stored.
5099 8. Exim monitor: an attempt to get the action menu when not actually pointing
5100 at a message produces an empty menu entitled "No message selected". This
5101 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
5102 no entries in it ("Shell widget menu has zero width and/or height"). So I
5103 have added a single, blank menu entry in this case.
5105 9. Added ${quote_local_part.
5107 10. MIME decoding is now applied to the contents of Subject: header lines when
5110 11. Now that a reference to $sender_host_address automatically causes a reverse
5111 lookup to occur if necessary (4.13/18), there is no need to arrange for a
5112 host lookup before query-style lookups in lists that might use this
5113 variable. This has therefore been abolished, and the "net-" prefix is no
5114 longer necessary for query-style lookups.
5116 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
5117 have been a typo for LFLAGS, so it has been changed.
5119 13. The install script calls Exim with "-C /dev/null" in order to find the
5120 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
5121 to be output. Howeve, since Exim outputs its version number before the
5122 error, it didn't break the script. It just looked ugly. I fixed this by
5123 always allowing "-C /dev/null" if the caller is root.
5125 14. Ignore overlarge ACL variable number when reading spool file - insurance
5126 against a later release with more variables having written the file.
5128 15. The standard form for an IPv6 address literal was being rejected by EHLO.
5129 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
5130 this, as well as the form without the "IPv6" on the front.
5132 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
5133 OS/Makefile-Darwin file.
5135 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
5136 only to LDAP libraries that do not have LDAP_OPT_DEREF.
5138 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
5139 variable. However, some OS use ints for inodes. I've added cast to long int
5140 to get rid of the compiler warning.
5142 19. I had forgotten to lock out "/../" in configuration file names when
5143 ALT_CONFIG_PREFIX was set.
5145 20. Routers used for verification do not need to specify transports. However,
5146 if such a router generated a host list, and callout was configured, Exim
5147 crashed, because it could not find a port number from the (non-existent)
5148 transport. It now assumes port 25 in this circumstance.
5150 21. Added the -t option to exigrep.
5152 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5153 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5154 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5155 them is set, LOOKUP_LSEARCH is forced.
5157 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5159 24. Added sender and host information to the "rejected by local_scan()" log
5160 line; previously there was no indication of these.
5162 25. Added .include_if_exists.
5164 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5165 Linux. It turns out that not all file systems support this. Apparently some
5166 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5167 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5168 not supported on the file descriptor, is now ignored when Exim is trying to
5169 sync a directory. This applies only to Linux.
5171 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5173 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5174 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5177 29. Added support for maildirsize files from supplied patch (modified a bit).
5179 30. The use of :fail: followed by an empty string could lead Exim to respond to
5180 sender verification failures with (e.g.):
5182 550 Verification failed for <xxx>
5183 550 Sender verify failed
5185 where the first response line was missing the '-' that indicates it is not
5186 the final line of the response.
5188 31. The loop for finding the name of the user that called Exim had a hardwired
5189 limit of 10; it now uses the value of finduser_retries, which is used for
5190 all other user lookups.
5192 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5195 33. Exim was neglecting to zero errno before one call of strtol() when
5196 expanding a string and expecting an integer value. On some systems this
5197 resulted in spurious "integer overflow" errors. Also, it was casting the
5198 result into an int without checking.
5200 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5201 did not work. The code looks as if it has *never* worked, though it appears
5202 to have been documented since at least releast 1.62. I have made it work.
5204 35. The "timeout_DNS" error in retry rules, also documented since at least
5205 1.62, also never worked. As it isn't clear exactly what this means, and
5206 clearly it isn't a major issue, I have abolished the feature by treating it
5207 as "timeout", and writing a warning to the main and panic logs.
5209 36. The display of retry rules for -brt wasn't always showing the error code
5212 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5213 timeout_connect_A, timeout_connect_MX.
5215 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5216 to the empty sender.
5218 39. The daemon was not analysing the content of -oX till after it had closed
5219 stderr and disconnected from the controlling terminal. This meant that any
5220 syntax errors were only noted on the panic log, and the return code from
5221 the command was 0. By re-arranging the code a little, I've made the
5222 decoding happen first, so such errors now appear on stderr, and the return
5223 code is 1. However, the actual setting up of the sockets still happens in
5224 the disconnected process, so errors there are still only recorded on the
5227 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5228 connections (as happens on some IP stacks) was logged at start up time as
5229 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5230 it from "IPv6 and IPv4", which means that two separate sockets are being
5233 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5234 says whether AF_INET or AF_INET6 was passed as an argument.
5236 42. Exiwhat output was messed up when time zones were included in log
5239 43. Exiwhat now gives more information about the daemon's listening ports,
5240 and whether -tls-on-connect was used.
5242 44. The "port" option of the smtp transport is now expanded.
5244 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5245 silently ignored. Now an error message is written to the main and panic
5248 46. There's a new ACL modifier called "logwrite" which writes to a log file
5249 as soon as it is encountered.
5251 47. Added $local_user_uid and $local_user_gid at routing time.
5253 48. Exim crashed when trying to verify a sender address that was being
5256 49. Exim was recognizing only a space character after ".include". It now also
5257 recognizes a tab character.
5259 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5260 extracting the relevant information from the specification. The man page no
5261 longer contains scrambled data for the -d option, and I've added a section
5262 at the front about calling Exim under different names.
5264 51. Added "extra_headers" argument to the "mail" command in filter files.
5266 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5269 53. Installed eximstats 1.29.
5271 54. Added transport_filter_timeout as a generic transport option.
5273 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5274 Cc: header lines. This was required by RFC 822, but it not required by RFC
5277 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5278 incoming messages that did not have them. Now it does so only if the
5279 message originates locally, that is, if there is no associated remote host
5280 address. When Resent- header lines are present, this applies to the Resent-
5281 lines rather than the non-Resent- lines.
5283 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5284 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5286 58. Messages for configuration errors now include the name of the main
5287 configuration file - useful now that there may be more than one file in a
5288 list (.included file names were always shown).
5290 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5291 for those rare installations that do not start the daemon as root or run it
5292 setuid root. I've cut out the call to initgroups() if the daemon is not
5295 60. The Exim user and group can now be bound into the binary as text strings
5296 that are looked up at the start of Exim's processing.
5298 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5300 62. Added $mailstore_basename variable.
5302 63. Installed patch to sieve.c from Michael Haardt.
5304 64. When Exim failed to open the panic log after failing to open the main log,
5305 the original message it was trying to log was written to stderr and debug
5306 output, but if they were not available (the usual case in production), it
5307 was lost. Now it is written to syslog before the two lines that record the
5308 failures to open the logs.
5310 65. Users' Exim filters run in subprocesses under the user's uid. It is
5311 possible for a "deliver" command or an alias in a "personal" command to
5312 provoke an address rewrite. If logging of address rewriting is configured,
5313 this fails because the process is not running as root or exim. There may be
5314 a better way of dealing with this, but for the moment (because 4.30 needs
5315 to be released), I have disabled address rewrite logging when running a
5316 filter in a non-root, non-exim process.
5322 1. The buildconfig auxiliary program wasn't quoting the value set for
5323 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5324 not defined. This bug was masked in 4.22 by the effect that was fixed in
5327 2. Some messages that were rejected after a message id was allocated were
5328 shown as "incomplete" by exigrep. It no longer does this for messages that
5329 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5331 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5332 have allow_domain_literals set, the ID did not get logged in the <= line.
5333 Domain literals are now always recognized in Message-ID: header lines.
5335 4. The first argument for a ${extract expansion item is the key name or field
5336 number. Leading and trailing spaces in this item were not being ignored,
5337 causing some misleading effects.
5339 5. When deliver_drop_privilege was set, single queue runner processes started
5340 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5341 same command in the process it spins off) were not dropping privilege.
5343 6. When the daemon running as "exim" started a queue runner, it always
5344 re-executed Exim in the spun-off process. This is a waste of effort when
5345 deliver_drop_privilege is set. The new process now just calls the
5346 queue-runner function directly.
5352 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5355 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5356 ignored. Though the use of -oP was forcing the writing of a pid file, it
5357 was always written to the default place.
5359 3. If the message "no IP address found for host xxxx" is generated during
5360 incoming verification, it is now followed by identification of the incoming
5361 connection (so you can more easily find what provoked it).
5363 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5365 5. Added some features to "harden" Exim a bit more against certain attacks:
5367 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5368 be put in Local/Makefile. This is like the never_users runtime option,
5369 but it cannot be overridden. The default setting is "root".
5371 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5372 prefix string with which any file named in a -C command line option
5375 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5376 is retained for -C and -D only if the caller of Exim is root. Without
5377 it, the exim user may also use -C and -D and retain privilege.
5379 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5380 command line option is disabled.
5382 6. Macro names set by the -D option must start with an upper case letter, just
5383 like macro names defined in the configuration file.
5385 7. Added "dereference=" facility to LDAP.
5387 8. Two instances of the typo "uknown" in the source files are fixed.
5389 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5390 the Configure-Makefile script screwed up while processing it.
5392 10. Incorporated PCRE 4.4.
5394 11. The SMTP synchronization check was not operating right at the start of an
5395 SMTP session. For example, it could not catch a HELO sent before the client
5396 waited for the greeting. There is now a check for outstanding input at the
5397 point when the greeting is written. Because of the duplex, asynchronous
5398 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5399 way, but not yet received, when the check is performed.
5401 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5402 on TCP/IP sockets, because this apparently causes some broken clients to
5405 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5406 unchanged) from the Cygwin maintainer.
5408 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5409 was supported instead of testing for mbx. Effectively a typo.
5411 15. The spa authenticator server code was not checking that the input it
5412 received was valid base64.
5414 16. The debug output line for the "set" modifier in ACLs was not showing the
5415 name of the variable that was being set.
5417 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5419 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5420 lookup on its own. Something else was needed to trigger the lookup. For
5421 example, a match in host_lookup or the need for a host name in a host list.
5422 Now, if $sender_host_name is referenced and the host name has not yet been
5423 looked up, a lookup is performed. If the lookup fails, the variable remains
5424 empty, and $host_lookup_failed is set to "1".
5426 19. Added "eqi" as a case-independent comparison operator.
5428 20. The saslauthd authentication condition could segfault if neither service
5429 nor realm was specified.
5431 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5432 error message that was logged was misleading, and incoming SMTP
5433 connections were dropped. The message is now more accurate, and temporary
5434 errors are given to SMTP connections.
5436 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5437 (except RSET) with a 421 error, until QUIT is received. However, it was
5438 failing to send a response to QUIT.
5440 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5441 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5442 ACL is now run after the helo_try_verify_hosts code.
5444 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5445 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5446 case-independent, so other case variants are also recognized). Apparently
5447 some systems use these upper case variants.
5449 25. If more than two messages were waiting for the same host, and a transport
5450 filter was specified for the transport, Exim sent two messages over the
5451 same TCP/IP connection, and then failed with "socket operation on non-
5452 socket" when it tried to send the third.
5454 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5456 27. The extern definition of crypt16() in expand.c was not being excluded when
5457 the OS had its own crypt16() function.
5459 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5460 as a preferred synonym for return_size_limit, both as an option and as an
5463 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5465 30. Changed the default configuration ACL to relax the local part checking rule
5466 for addresses that are not in any local domains. For these addresses,
5467 slashes and pipe symbols are allowed within local parts, but the sequence
5468 /../ is explicitly forbidden.
5470 31. SPA server authentication was not clearing the challenge buffer before
5473 32. log_message in a "warn" ACL statement was writing to the reject log as
5474 well as to the main log, which contradicts the documentation and doesn't
5475 seem right (because no rejection is happening). So I have stopped it.
5477 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5478 However, I am unable to do any testing of this.
5480 34. Fixed an infelicity in the appendfile transport. When checking directories
5481 for a mailbox, to see if any needed to be created, it was accidentally
5482 using path names with one or more superfluous leading slashes; tracing
5483 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5485 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5486 contents are added to the log line that is written for every discarded
5487 recipient. (Previously a log_message setting was ignored.)
5489 36. The ${quote: operator now quotes the string if it is empty.
5491 37. The install script runs exim in order to find its version number. If for
5492 some reason other than non-existence or emptiness, which it checks, it
5493 could not run './exim', it was installing it with an empty version number,
5494 i.e. as "exim-". This error state is now caught, and the installation is
5497 38. An argument was missing from the function that creates an error message
5498 when Exim fails to connect to the socket for saslauthd authentication.
5499 This could cause Exim to crash, or give a corrupted message.
5501 39. Added isip, isip4, and isip6 to ${if conditions.
5503 40. The ACL variables $acl_xx are now saved with the message, and can be
5504 accessed later in routers, transports, and filters.
5506 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5507 strings in the file are not string-expanded.
5509 42. If a MAIL command specified a SIZE value that was too large to fit into an
5510 int variable, the check against message_size_limit failed. Such values are
5511 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5512 day this will have to be increased, but I don't think I want to be around
5513 when emails are that large.
5520 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5521 iconv() is not standard in FreeBSD.
5523 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5524 IPv6 enabled. The observed symptom was a segmentation fault on return from
5525 the function os_common_find_running_interfaces() in src/os.c.
5527 3. In the check_special_case() function in daemon.c I had used "errno" as an
5528 argument name, which causes warnings on some systems. This was basically a
5529 typo, since it was named "eno" in the comments!
5531 4. The code that waits for the clock to tick (at a resolution of some fraction
5532 of a second) so as to ensure message-id uniqueness was always waiting for
5533 at least one whole tick, when it could have waited for less. [This is
5534 almost certainly not relevant at current processor speeds, where it is
5535 unlikely to ever wait at all. But we try to future-proof.]
5537 5. The function that sleeps for a time interval that includes fractions of a
5538 second contained a race. It did not block SIGALRM between setting the
5539 timer, and suspending (a couple of lines later). If the interval was short
5540 and the sigsuspend() was delayed until after it had expired, the suspension
5541 never ended. On busy systems this could lead to processes getting stuck for
5544 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5545 process, before it forks any delivery processes. The open lookup caching
5546 mechanism meant that the open file or database connection was passed into
5547 the delivery process. The problem was that delivery processes always tidy
5548 up cached lookup data. This could cause a problem for the next delivery
5549 process started by the queue runner, because the external queue runner
5550 process does not know about the closure. So the next delivery process
5551 still has data in the lookup cache. In the case of a file lookup, there was
5552 no problem because closing a file descriptor in a subprocess doesn't affect
5553 the parent. However, if the lookup was caching a connection to a database,
5554 the connection was closed, and the second delivery process was likely to
5555 see errors such as "PGSQL: query failed: server closed the connection
5556 unexpectedly". The problem has been fixed by closing all cached lookups
5557 in a queue runner before running a delivery process.
5559 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5560 seem to have the "const" qualifier which it has on other OS. I've
5563 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5564 *of the same type* (client or server) with the same public name that an
5565 error should be diagnosed.
5567 9. When Exim looked up a host name for an IP address, but failed to find the
5568 original IP address when looking up the host name (a safety check), it
5569 output the message "<ip address> does not match any IP for NULL", which was
5570 confusing, to say the least. The bug was that the host name should have
5571 appeared instead of "NULL".
5573 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5574 user that is built into the binary, and the -C or -D options is used, root
5575 privilege is dropped before the configuration file is read. In addition,
5576 logging is switched to stderr instead of the normal log files. If the
5577 configuration then re-defines the Exim user, the unprivileged environment
5578 is probably not what is expected, so Exim logs a panic warning message (but
5581 However, if deliver_drop_privilege is set, the unprivileged state may well
5582 be exactly what is intended, so the warning has been cut out in that case,
5583 and Exim is allowed to try to write to its normal log files.
5589 1. smtp_return_error_details was not giving details for temporary sender
5590 or receiver verification errors.
5592 2. Diagnose a configuration error if two authenticators have the same public
5595 3. Exim used not to create the message log file for a message until the first
5596 delivery attempt. This could be confusing when incoming messages were held
5597 for policy or load reasons. The message log file is now created at the time
5598 the message is received, and an initial "Received" line is written to it.
5600 4. The automatically generated man page for command line options had a minor
5601 bug that caused no ill effects; however, a more serious problem was that
5602 the procedure for building the man page automatically didn't always
5603 operate. Consequently, release 4.20 contains an out-of-date version. This
5604 shouldn't happen again.
5606 5. When building Exim with embedded Perl support, the script that builds the
5607 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5608 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5610 6. The freeze_tell option was not being used for messages that were frozen on
5611 arrival, either by an ACL or by local_scan().
5613 7. Added the smtp_incomplete_transaction log selector.
5615 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5616 was accepting AUTH without a new EHLO.
5618 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5619 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5622 10. Logging of TCP/IP connections (when configured) now happens in the main
5623 daemon process instead of the child process, so that the TCP/IP connection
5624 count is more accurate (but it can never be perfect).
5626 11. The use of "drop" in a nested ACL was not being handled correctly in the
5627 outer ACL. Now, if condition failure induced by the nested "drop" causes
5628 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5629 or "require"), the connection is dropped.
5631 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5632 that yield "discard" can now be used with an "accept" or a "discard" verb,
5633 but an error is generated for any others (because I can't see a useful way
5634 to define what should happen).
5636 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5637 are now processed in the same way as lines in the Exim configuration file.
5638 In particular, continuation lines are supported.
5640 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5642 15. Added -ti meaning -t -i.
5644 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5645 domains, and warn by logging if others are found.
5647 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5648 returned from GIFCONF when Exim is trying to find the list of interfaces on
5649 a host. The code in os.c has been modified to copy each ifreq to an aligned
5650 structure in all cases.
5652 Also, in some cases, the returned ifreq's were being copied to a 'struct
5653 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5654 means the last couple of bytes of an IPv6 address could be chopped if the
5655 ifreq contained only a normal sockaddr (14 bytes storage).
5657 18. Named domain lists were not supported in the hosts_treat_as_local option.
5658 An entry such as +xxxx was not recognized, and was treated as a literal
5661 19. Ensure that header lines added by a DATA ACL are included in the reject log
5662 if the ACL subsequently rejects the message.
5664 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5665 MD5 (which is deprecated).
5667 21. When testing a filter file using -bf, Exim was writing a message when it
5668 took the sender from a "From " line in the message, but it was not doing so
5669 when it took $return_path from a Return-Path: header line. It now does.
5671 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5672 with a valid header line field name (a series of printing characters
5673 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5675 23. Changed "disc" in the source to "disk" to conform to the documentation and
5676 the book and for uniformity.
5678 24. Ignore Sendmail's -Ooption=value command line item.
5680 25. When execve() failed while trying to run a command in a pipe transport,
5681 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5682 could be confused with a return value of 69 from the command itself. This
5683 has been changed to 127, the value the shell returns if it is asked to run
5684 a non-existent command. The wording for the related log line suggests a
5685 non-existent command as the problem.
5687 26. If received_header_text expands to an empty string, do not add a Received:
5688 header line to the message. (Well, it adds a token one on the spool, but
5689 marks it "old" so that it doesn't get used or transmitted.)
5691 27. Installed eximstats 1.28 (addition of -nt option).
5693 28. There was no check for failure on the call to getsockname() in the daemon
5694 code. This can fail if there is a shortage of resources on the system, with
5695 ENOMEM, for example. A temporary error is now given on failure.
5697 29. Contrary to the C standard, it seems that in some environments, the
5698 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5699 program. Exim now does this explicitly; it affects the formatting of
5700 timestamps using strftime().
5702 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5703 complaints. I've now initialized all the variables, to avoid this.
5705 32. Header lines added by a system filter were not being "seen" during
5706 transport-time rewrites.
5708 33. The info_callback() function passed to OpenSSL is set up with type void
5709 (*)(SSL *, int, int), as described somewhere. However, when calling the
5710 function (actually a macro) that sets it up, the type void(*)() is
5711 expected. I've put in a cast to prevent warnings from picky compilers.
5713 34. If a DNS black list lookup found a CNAME record, but there were no A
5714 records associated with the domain it pointed at, Exim crashed.
5716 35. If a DNS black list lookup returned more than one A record, Exim ignored
5717 all but the first. It now scans all returned addresses if a particular IP
5718 value is being sought. In this situation, the contents of the
5719 $dnslist_value variable are a list of all the addresses, separated by a
5722 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5723 to accept a host name and all its aliases if the forward lookup for any of
5724 them yielded the IP address of the incoming connection. Now it accepts only
5725 those names whose forward lookup yields the correct IP address. Any other
5726 names are discarded. This closes a loophole whereby a rogue DNS
5727 administrator could create reverse DNS records to break through a
5728 wildcarded host restriction in an ACL.
5730 37. If a user filter or a system filter that ran in a subprocess used any of
5731 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5732 the wrong values were passed to the pipe command ($thisaddress had the
5733 value of $0, $0 had the value of $1, etc). This bug was introduced by
5734 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5736 38. Improved the line breaking for long SMTP error messages from ACLs.
5737 Previously, if there was no break point between 40 and 75 characters, Exim
5738 left the rest of the message alone. Two changes have been made: (a) I've
5739 reduced the minimum length to 35 characters; (b) if it can't find a break
5740 point between 35 and 75 characters, it looks ahead and uses the first one
5741 that it finds. This may give the occasional overlong line, but at least the
5742 remaining text gets split now.
5744 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5745 file descriptors might be low, and that setting 1000 would always raise it.
5746 It turns out that in some environments, the limit is already over 1000 and
5747 that lowering it causes trouble. So now Exim takes care not to decrease it.
5749 40. When delivering a message, the value of $return_path is set to $sender_
5750 address at the start of routing (routers may change the value). By an
5751 oversight, this default was not being set up when an address was tested by
5752 -bt or -bv, which affected the outcome if any router or filter referred to
5755 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5756 log only when "message" or "log_message" are set. However, if one of the
5757 conditions was an address verification, or a call to a nested ACL, the
5758 messages generated by the underlying test were being passed through. This
5759 no longer happens. The underlying message is available in $acl_verify_
5760 message for both "message" and "log_message" expansions, so it can be
5761 passed through if needed.
5763 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5764 new expansion $bh_ to give the encoded byte string without charset
5765 translation. Translation happens only if iconv() is available; HAVE_ICONV
5766 indicates this at build time. HEADERS_CHARSET gives the charset to
5767 translate to; headers_charset can change it in the configuration, and
5768 "headers charset" can change it in an individual filter file.
5770 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5771 that creates RFC 2047 encoded "words" labels them as that charset instead
5772 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5773 expansion operator; (ii) when Exim creates a From: line for a local
5774 message; (iii) when a header line is rewritten to include a "phrase" part.
5776 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5777 buggy, causing it to skip the first lines of messages whose message ID
5778 ended in 'D'. This would not have bitten before Exim release 4.14, because
5779 message IDs were unlikely to end in 'D' before then. The effect was to have
5780 incorrect size information for certain domains.
5782 45. #include "config.h" was missing at the start of the crypt16.c module. This
5783 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5786 46. If there was a timeout during a "random" callout check, Exim treated it as
5787 a failure of the random address, and carried on sending RSET and the real
5788 address. If the delay was just some slowness somewhere, the response to the
5789 original RCPT would be taken as a response to RSET and so on, causing
5790 mayhem of various kinds.
5792 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5793 when I implemented it. It didn't allow for the fact that some option values
5794 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5795 the right test for positive values.
5797 48. Domain names in DNS records are case-independent. Exim always looks them up
5798 in lower case. Some resolvers return domain names in exactly the case they
5799 appear in the zone file, that is, they may contain uppercase letters. Not
5800 all resolvers do this - some return always lower case. Exim was treating a
5801 change of case by a resolver as a change of domain, similar to a widening
5802 of a domain abbreviation. This triggered its re-routing code and so it was
5803 trying to route what was effectively the same domain again. This normally
5804 caused routing to fail (because the router wouldn't handle the domain
5805 twice). Now Exim checks for this case specially, and just changes the
5806 casing of the domain that it ultimately uses when it transmits the message
5809 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5812 50. If a filter generated a file delivery with a non-absolute name (possible if
5813 no home directory exists for the router), the forbid_file option was not
5816 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5817 the existing equality matching.
5819 52. Exim was using ints instead of ino_t variables in some places where it was
5820 dealing with inode numbers.
5822 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5823 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5824 called TMPDIR, and if it finds it is different, it changes its value.
5826 54. The smtp_printf() function is now made available to local_scan() so
5827 additional output lines can be written before returning. There is also an
5828 smtp_fflush() function to enable the detection of a dropped connection.
5829 The variables smtp_input and smtp_batched_input are exported to
5832 55. Changed the default runtime configuration: the message "Unknown user"
5833 has been removed from the ACL, and instead placed on the localuser router,
5834 using the cannot_route_message feature. This means that any verification
5835 failures that generate their own messages won't get overridden. Similarly,
5836 the "Unrouteable address" message that was in the ACL for unverifiable
5837 relay addresses has also been removed.
5839 56. Added hosts_avoid_esmtp to the smtp transport.
5841 57. The exicyclog script was not checking for the esoteric option
5842 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5843 will work only if exicyclog is run under the appropriate euid.
5845 58. Following a discussion on the list, the rules by which Exim recognises line
5846 endings on incoming messages have been changed. The -dropcr and drop_cr
5847 options are now no-ops, retained only for backwards compatibility. The
5848 following line terminators are recognized: LF CRLF CR. However, special
5849 processing applies to CR:
5851 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5852 nor a local message in the state where . is a terminator.
5854 (ii) If a bare CR is encountered in a header line, an extra space is added
5855 after the line terminator so as not to end the header. The reasoning
5856 behind this is that bare CRs in header lines are most likely either
5857 to be mistakes, or people trying to play silly games.
5859 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5860 was being incorrectly given as 18 bytes larger than it should have been.
5861 This is a VOB (very old bug).
5863 60. This may never have affected anything current, but just in case it has:
5864 When the local host is found other than at the start of a list of hosts,
5865 the local host, those with the same MX, and any that follow, are discarded.
5866 When the list in question was part of a longer list of hosts, the following
5867 hosts (not currently being processed) were also being discarded. This no
5868 longer happens. I'm not sure if this situation could ever has previously
5871 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5874 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5875 header. This is something that is recommended in a new Internet Draft, and
5876 is something that is documented as being done by Sendmail. There are two
5877 possible values. For messages generated by the autoreply transport, Exim
5880 Auto-Submitted: auto-replied
5882 whereas for all other generated messages (e.g. bounces) it adds
5884 Auto-Submitted: auto-generated
5886 63. The "personal" condition in filters now includes a test for the
5887 Auto-Submitted: header. If it contains the string "auto-" the message it
5888 not considered personal.
5890 64. Added rcpt_include_affixes as a generic transport option.
5892 65. Added queue_only_override (default true).
5894 66. Added the syslog_duplication option.
5896 67. If what should have been the first header line of a message consisted of
5897 a space followed by a colon, Exim was mis-interpreting it as a header line.
5898 It isn't of course - it is syntactically invalid and should therefore be
5899 treated as the start of the message body. The misbehaviour could have
5900 caused a number of strange effects, including loss of data in subsequent
5901 header lines, and spool format errors.
5903 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5904 client host had authenticated. This control can now be exercised by an ACL
5905 for more flexibility.
5907 69. By default, callouts do not happen when testing with -bh. There is now a
5908 variant, -bhc, which does actually run the callout code, including
5909 consulting and updating the callout cache.
5911 70. Added support for saslauthd authentication, courtesy of Alexander
5914 71. If statvfs() failed on the spool or log directories while checking their
5915 size for availability, Exim confusingly gave the error "space shortage".
5916 Furthermore, in debugging mode it crashed with a floating point exception.
5917 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5918 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5919 serious problem, Exim now writes to the main and panic logs when this
5920 happens, with details of the failure. It then refuses to accept the
5921 incoming message, giving the message "spool directory problem" or "log
5922 directory problem" with a 421 code for SMTP messages.
5924 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5925 0, 1, and 2 exist, because some OS complain for execs without them (see
5926 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5927 descriptors. However, the code omitted to check that the open succeeded,
5928 causing mysterious errors if for some reason the permissions on /dev/null
5929 got screwed. Now Exim writes a message to the main and panic logs, and
5930 bombs out if it can't open /dev/null.
5932 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5933 interact so that it is all more flexible. It is supposed to remain
5934 backwards compatible. Also added extra_local_interfaces.
5936 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5937 to bomb out with an assertion failure - to the client this appears as a
5938 connection drop. This problem occurs in the part of the code that was taken
5939 from the Samba project. Fortunately, the assertion is in a very simple
5940 function, so I have fixed this by reproducing the function inline in the
5941 one place where it is called, and arranging for authentication to fail
5942 instead of killing the process with assert().
5944 75. The SPA client code was not working when the server requested OEM rather
5945 than Unicode encoding.
5947 76. Added code to make require_files with a specific uid setting more usable in
5948 the case where statting the file as root fails - usually a non-root-mounted
5949 NFS file system. When this happens and the failure is EACCES, Exim now
5950 forks a subprocess and does the per-uid checking as the relevant uid.
5952 77. Added process_log_path.
5954 78. If log_file_path was not explicitly set, a setting of check_log_space or
5955 check_log_inodes was ignored.
5957 79. If a space check for the spool or log partitions fails, the incident is now
5958 logged. Of course, in the latter case the data may get lost...
5960 80. Added the %p formatting code to string_format() so that it can be used to
5961 print addresses in debug_print(). Adjusted all the address printing in the
5962 debugging in store.c to use %p rather than %d.
5964 81. There was a concern that a line of code in smtp_in.c could overflow a
5965 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5966 initially expressed, the concern was not well-founded, because trailing
5967 spaces are removed early. However, if the trailing spaces were followed by
5968 a NULL, they did not get removed, so the overflow was possible. Two fixes
5971 (a) I re-wrote the offending code in a cleaner fashion.
5972 (b) If an incoming SMTP command contains a NULL character, it is rejected
5975 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5976 runs initgroups(), so that if the Exim user is in any additional groups,
5977 they will be used during message reception.
5983 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob