1 Change log file for Exim from version 4.21
2 -------------------------------------------
7 PP/01 Add -bI: framework, and -bI:sieve for querying sieve capabilities.
9 PP/02 Make -n do something, by making it not do something.
10 When combined with -bP, the name of an option is not output.
12 PP/03 Added tls_dh_min_bits SMTP transport driver option, only honoured
15 PP/04 First step towards DNSSEC, provide $sender_host_dnssec for
16 $sender_host_name and config options to manage this, and basic check
19 PP/05 DSCP support for outbound connections and control modifier for inbound.
21 PP/06 Cyrus SASL: set local and remote IP;port properties for driver.
22 (Only plugin which currently uses this is kerberos4, which nobody should
23 be using, but we should make it available and other future plugins might
24 conceivably use it, even though it would break NAT; stuff *should* be
25 using channel bindings instead).
27 PP/07 Handle "exim -L <tag>" to indicate to use syslog with tag as the process
28 name; added for Sendmail compatibility; requires admin caller.
29 Handle -G as equivalent to "control = suppress_local_fixups" (we used to
30 just ignore it); requires trusted caller.
31 Also parse but ignore: -Ac -Am -X<logfile>
34 TL/01 Bugzilla 1258 - Refactor MAIL FROM optional args processing.
36 TL/02 Add +smtp_confirmation as a default logging option.
38 TL/03 Bugzilla 198 - Implement remove_header ACL modifier.
39 Patch by Magnus Holmgren from 2007-02-20.
41 TL/04 Bugzilla 1281 - Spec typo.
42 Bugzilla 1283 - Spec typo.
43 Bugzilla 1290 - Spec grammar fixes.
45 TL/05 Bugzilla 1285 - Spec omission, fix docbook errors for spec.txt creation.
47 JH/01 Bugzilla 1201 & 304 - New cutthrough-delivery feature, with TLS support.
49 JH/02 Support "G" suffix to numbers in ${if comparisons.
51 PP/08 Handle smtp transport tls_sni option forced-fail for OpenSSL.
53 NM/01 Bugzilla 1197 - Spec typo
54 Bugzilla 1196 - Spec examples corrections
56 JH/03 Add expansion operators ${listnamed:name} and ${listcount:string}
58 PP/09 Add gnutls_enable_pkcs11 option.
60 PP/10 Let Linux makefile inherit CFLAGS/CFLAGS_DYNAMIC.
61 Pulled from Debian 30_dontoverridecflags.dpatch by Andreas Metzler.
63 JH/04 Add expansion item ${acl {name}{arg}...}, expansion condition
64 "acl {{name}{arg}...}", and optional args on acl condition
67 JH/05 Permit multiple router/transport headers_add/remove lines.
69 JH/06 Add dnsdb pseudo-lookup "a+" to do an "aaaa" + "a" combination.
71 JH/07 Avoid using a waiting database for a single-message-only transport.
72 Performance patch from Paul Fisher. Bugzilla 1262.
74 JH/08 Strip leading/trailing newlines from add_header ACL modifier data.
77 JH/09 Add $headers_added variable, with content from use of ACL modifier
78 add_header (but not yet added to the message). Bugzilla 199.
80 JH/10 Add 8bitmime log_selector, for 8bitmime status on the received line.
81 Pulled from Bugzilla 817 by Wolfgang Breyha.
83 PP/11 SECURITY: protect DKIM DNS decoding from remote exploit.
86 JH/11 Add A= logging on delivery lines, and a client_set_id option on
89 JH/12 Add optional authenticated_sender logging to A= and a log_selector
92 PP/12 Unbreak server_set_id for NTLM/SPA auth, broken by 4.80 PP/29.
94 PP/13 Dovecot auth: log better reason to rejectlog if Dovecot did not
95 advertise SMTP AUTH mechanism to us, instead of a generic
96 protocol violation error. Also, make Exim more robust to bad
97 data from the Dovecot auth socket.
99 TF/01 Fix ultimate retry timeouts for intermittently deliverable recipients.
101 When a queue runner is handling a message, Exim first routes the
102 recipient addresses, during which it prunes them based on the retry
103 hints database. After that it attempts to deliver the message to
104 any remaining recipients. It then updates the hints database using
107 So if a recipient address works intermittently, it can get repeatedly
108 deferred at routing time. The retry hints record remains fresh so the
109 address never reaches the final cutoff time.
111 This is a fairly common occurrence when a user is bumping up against
112 their storage quota. Exim had some logic in its local delivery code
113 to deal with this. However it did not apply to per-recipient defers
114 in remote deliveries, e.g. over LMTP to a separate IMAP message store.
116 This change adds a proper retry rule check during routing so that the
117 final cutoff time is checked against the message's age. We only do
118 this check if there is an address retry record and there is not a
119 domain retry record; this implies that previous attempts to handle
120 the address had the retry_use_local_parts option turned on. We use
121 this as an approximation for the destination being like a local
122 delivery, as in LMTP.
124 I suspect this new check makes the old local delivery cutoff check
125 redundant, but I have not verified this so I left the code in place.
127 TF/02 Correct gecos expansion when From: is a prefix of the username.
129 Test 0254 submits a message to Exim with the header
133 When I ran the test suite under the user fanf2, Exim expanded
134 the header to contain my full name, whereas it should have added
135 a Resent-Sender: header. It erroneously treats any prefix of the
136 username as equal to the username.
138 This change corrects that bug.
140 GF/01 DCC debug and logging tidyup
141 Error conditions log to paniclog rather than rejectlog.
142 Debug lines prefixed by "DCC: " to remove any ambiguity.
144 TF/03 Avoid unnecessary rebuilds of lookup-related code.
146 PP/14 Fix OCSP reinitialisation in SNI handling for Exim/TLS as server.
147 Bug spotted by Jeremy Harris; was flawed since initial commit.
148 Would have resulted in OCSP responses post-SNI triggering an Exim
149 NULL dereference and crash.
151 JH/13 Add $router_name and $transport_name variables. Bugzilla 308.
153 PP/15 Define SIOCGIFCONF_GIVES_ADDR for GNU Hurd.
154 Bug detection, analysis and fix by Samuel Thibault.
155 Bugzilla 1331, Debian bug #698092.
157 SC/01 Update eximstats to watch out for senders sending 'HELO [IpAddr]'
159 JH/14 SMTP PRDR (http://www.eric-a-hall.com/specs/draft-hall-prdr-00.txt).
160 Server implementation by Todd Lyons, client by JH.
161 Only enabled when compiled with EXPERIMENTAL_PRDR. A new
162 config variable "prdr_enable" controls whether the server
163 advertises the facility. If the client requests PRDR a new
164 acl_data_smtp_prdr ACL is called once for each recipient, after
165 the body content is received and before the acl_smtp_data ACL.
166 The client is controlled by bolth of: a hosts_try_prdr option
167 on the smtp transport, and the server advertisement.
168 Default client logging of deliveries and rejections involving
169 PRDR are flagged with the string "PRDR".
171 PP/16 Fix problems caused by timeouts during quit ACLs trying to double
172 fclose(). Diagnosis by Todd Lyons.
174 PP/17 Update configure.default to handle IPv6 localhost better.
175 Patch by Alain Williams (plus minor tweaks).
178 PP/18 OpenSSL made graceful with empty tls_verify_certificates setting.
179 This is now consistent with GnuTLS, and is now documented: the
180 previous undocumented portable approach to treating the option as
181 unset was to force an expansion failure. That still works, and
182 an empty string is now equivalent.
184 PP/19 Renamed DNSSEC-enabling option to "dns_dnssec_ok", to make it
185 clearer that Exim is using the DO (DNSSEC OK) EDNS0 resolver flag,
186 not performing validation itself.
192 PP/01 SECURITY: protect DKIM DNS decoding from remote exploit.
194 This, or similar/improved, will also be change PP/11 of 4.82.
200 PP/01 Handle short writes when writing local log-files.
201 In practice, only affects FreeBSD (8 onwards).
202 Bugzilla 1053, with thanks to Dmitry Isaikin.
204 NM/01 Bugzilla 949 - Documentation tweak
206 NM/02 Bugzilla 1093 - eximstats DATA reject detection regexps
209 NM/03 Bugzilla 1169 - primary_hostname spelling was incorrect in docs.
211 PP/02 Implemented gsasl authenticator.
213 PP/03 Implemented heimdal_gssapi authenticator with "server_keytab" option.
215 PP/04 Local/Makefile support for (AUTH|LOOKUP)_*_PC=foo to use
216 `pkg-config foo` for cflags/libs.
218 PP/05 Swapped $auth1/$auth2 for gsasl GSSAPI mechanism, to be more consistent
219 with rest of GSASL and with heimdal_gssapi.
221 PP/06 Local/Makefile support for USE_(GNUTLS|OPENSSL)_PC=foo to use
222 `pkg-config foo` for cflags/libs for the TLS implementation.
224 PP/07 New expansion variable $tls_bits; Cyrus SASL server connection
225 properties get this fed in as external SSF. A number of robustness
226 and debugging improvements to the cyrus_sasl authenticator.
228 PP/08 cyrus_sasl server now expands the server_realm option.
230 PP/09 Bugzilla 1214 - Log authentication information in reject log.
231 Patch by Jeremy Harris.
233 PP/10 Added dbmjz lookup type.
235 PP/11 Let heimdal_gssapi authenticator take a SASL message without an authzid.
237 PP/12 MAIL args handles TAB as well as SP, for better interop with
238 non-compliant senders.
239 Analysis and variant patch by Todd Lyons.
241 NM/04 Bugzilla 1237 - fix cases where printf format usage not indicated
242 Bug report from Lars Müller <lars@samba.org> (via SUSE),
243 Patch from Dirk Mueller <dmueller@suse.com>
245 PP/13 tls_peerdn now print-escaped for spool files.
246 Observed some $tls_peerdn in wild which contained \n, which resulted
247 in spool file corruption.
249 PP/14 TLS fixes for OpenSSL: support TLS 1.1 & 1.2; new "openssl_options"
250 values; set SSL_MODE_AUTO_RETRY so that OpenSSL will retry a read
251 or write after TLS renegotiation, which otherwise led to messages
254 TK/01 Bugzilla 1239 - fix DKIM verification when signature was not inserted
255 as a tracking header (ie: a signed header comes before the signature).
256 Patch from Wolfgang Breyha.
258 JH/01 Bugzilla 660 - Multi-valued attributes from ldap now parseable as a
259 comma-sep list; embedded commas doubled.
261 JH/02 Refactored ACL "verify =" logic to table-driven dispatch.
263 PP/15 LDAP: Check for errors of TLS initialisation, to give correct
265 Report and patch from Dmitry Banschikov.
267 PP/16 Removed "dont_insert_empty_fragments" fron "openssl_options".
268 Removed SSL_clear() after SSL_new() which led to protocol negotiation
269 failures. We appear to now support TLS1.1+ with Exim.
271 PP/17 OpenSSL: new expansion var $tls_sni, which if used in tls_certificate
272 lets Exim select keys and certificates based upon TLS SNI from client.
273 Also option tls_sni on SMTP Transports. Also clear $tls_bits correctly
274 before an outbound SMTP session. New log_selector, +tls_sni.
276 PP/18 Bugzilla 1122 - check localhost_number expansion for failure, avoid
277 NULL dereference. Report and patch from Alun Jones.
279 PP/19 DNS resolver init changes for NetBSD compatibility. (Risk of breakage
280 on less well tested platforms). Obviates NetBSD pkgsrc patch-ac.
281 Not seeing resolver debug output on NetBSD, but suspect this is a
282 resolver implementation change.
284 PP/20 Revert part of NM/04, it broke log_path containing %D expansions.
285 Left warnings. Added "eximon gdb" invocation mode.
287 PP/21 Defaulting "accept_8bitmime" to true, not false.
289 PP/22 Added -bw for inetd wait mode support.
291 PP/23 Added PCRE_CONFIG=yes support to Makefile for using pcre-config to
292 locate the relevant includes and libraries. Made this the default.
294 PP/24 Fixed headers_only on smtp transports (was not sending trailing dot).
295 Bugzilla 1246, report and most of solution from Tomasz Kusy.
297 JH/03 ${eval } now uses 64-bit and supports a "g" suffix (like to "k" and "m").
298 This may cause build issues on older platforms.
300 PP/25 Revamped GnuTLS support, passing tls_require_ciphers to
301 gnutls_priority_init, ignoring Exim options gnutls_require_kx,
302 gnutls_require_mac & gnutls_require_protocols (no longer supported).
303 Added SNI support via GnuTLS too.
304 Made ${randint:..} supplier available, if using not-too-old GnuTLS.
306 PP/26 Added EXPERIMENTAL_OCSP for OpenSSL.
308 PP/27 Applied dnsdb SPF support patch from Janne Snabb.
309 Applied second patch from Janne, implementing suggestion to default
310 multiple-strings-in-record handling to match SPF spec.
312 JH/04 Added expansion variable $tod_epoch_l for a higher-precision time.
314 PP/28 Fix DCC dcc_header content corruption (stack memory referenced,
315 read-only, out of scope).
316 Patch from Wolfgang Breyha, report from Stuart Northfield.
318 PP/29 Fix three issues highlighted by clang analyser static analysis.
319 Only crash-plausible issue would require the Cambridge-specific
320 iplookup router and a misconfiguration.
321 Report from Marcin Mirosław.
323 PP/30 Another attempt to deal with PCRE_PRERELEASE, this one less buggy.
325 PP/31 %D in printf continues to cause issues (-Wformat=security), so for
326 now guard some of the printf checks behind WANT_DEEPER_PRINTF_CHECKS.
327 As part of this, removing so much warning spew let me fix some minor
328 real issues in debug logging.
330 PP/32 GnuTLS was always using default tls_require_ciphers, due to a missing
331 assignment on my part. Fixed.
333 PP/33 Added tls_dh_max_bits option, defaulting to current hard-coded limit
334 of NSS, for GnuTLS/NSS interop. Problem root cause diagnosis by
335 Janne Snabb (who went above and beyond: thank you).
337 PP/34 Validate tls_require_ciphers on startup, since debugging an invalid
338 string otherwise requires a connection and a bunch more work and it's
339 relatively easy to get wrong. Should also expose TLS library linkage
342 PP/35 Pull in <features.h> on Linux, for some portability edge-cases of
343 64-bit ${eval} (JH/03).
345 PP/36 Define _GNU_SOURCE in exim.h; it's needed for some releases of
346 GNU libc to support some of the 64-bit stuff, should not lead to
347 conflicts. Defined before os.h is pulled in, so if a given platform
348 needs to override this, it can.
350 PP/37 Unbreak Cyrus SASL auth: SSF retrieval was incorrect, Exim thought
351 protection layer was required, which is not implemented.
352 Bugzilla 1254, patch from Wolfgang Breyha.
354 PP/38 Overhaul DH prime handling, supply RFC-specified DH primes as built
355 into Exim, default to IKE id 23 from RFC 5114 (2048 bit). Make
356 tls_dhparam take prime identifiers. Also unbreak combination of
357 OpenSSL+DH_params+TLSSNI.
359 PP/39 Disable SSLv2 by default in OpenSSL support.
365 PP/01 Solaris build fix for Oracle's LDAP libraries.
366 Bugzilla 1109, patch from Stephen Usher.
368 TF/01 HP/UX build fix: avoid arithmetic on a void pointer.
370 TK/01 DKIM Verification: Fix relaxed canon for empty headers w/o
373 TF/02 Fix a couple more cases where we did not log the error message
374 when unlink() failed. See also change 4.74-TF/03.
376 TF/03 Make the exiwhat support code safe for signals. Previously Exim might
377 lock up or crash if it happened to be inside a call to libc when it
378 got a SIGUSR1 from exiwhat.
380 The SIGUSR1 handler appends the current process status to the process
381 log which is later printed by exiwhat. It used to use the general
382 purpose logging code to do this, but several functions it calls are
383 not safe for signals.
385 The new output code in the SIGUSR1 handler is specific to the process
386 log, and simple enough that it's easy to inspect for signal safety.
387 Removing some special cases also simplifies the general logging code.
388 Removing the spurious timestamps from the process log simplifies
391 TF/04 Improved ratelimit ACL condition.
393 The /noupdate option has been deprecated in favour of /readonly which
394 has clearer semantics. The /leaky, /strict, and /readonly update modes
395 are mutually exclusive. The update mode is no longer included in the
396 database key; it just determines when the database is updated. (This
397 means that when you upgrde Exim will forget old rate measurements.)
399 Exim now checks that the per_* options are used with an update mode that
400 makes sense for the current ACL. For example, when Exim is processing a
401 message (e.g. acl_smtp_rcpt or acl_smtp_data, etc.) you can specify
402 per_mail/leaky or per_mail/strict; otherwise (e.g. in acl_smtp_helo) you
403 must specify per_mail/readonly. If you omit the update mode it defaults to
404 /leaky where that makes sense (as before) or /readonly where required.
406 The /noupdate option is now undocumented but still supported for
407 backwards compatibility. It is equivalent to /readonly except that in
408 ACLs where /readonly is required you may specify /leaky/noupdate or
409 /strict/noupdate which are treated the same as /readonly.
411 A useful new feature is the /count= option. This is a generalization
412 of the per_byte option, so that you can measure the throughput of other
413 aggregate values. For example, the per_byte option is now equivalent
414 to per_mail/count=${if >{0}{$message_size} {0} {$message_size} }.
416 The per_rcpt option has been generalized using the /count= mechanism
417 (though it's more complicated than the per_byte equivalence). When it is
418 used in acl_smtp_rcpt, the per_rcpt option adds recipients to the
419 measured rate one at a time; if it is used later (e.g. in acl_smtp_data)
420 or in a non-SMTP ACL it adds all the recipients in one go. (The latter
421 /count=$recipients_count behaviour used to work only in non-SMTP ACLs.)
422 Note that using per_rcpt with a non-readonly update mode in more than
423 one ACL will cause the recipients to be double-counted. (The per_mail
424 and per_byte options don't have this problem.)
426 The handling of very low rates has changed slightly. If the computed rate
427 is less than the event's count (usually one) then this event is the first
428 after a long gap. In this case the rate is set to the same as this event's
429 count, so that the first message of a spam run is counted properly.
431 The major new feature is a mechanism for counting the rate of unique
432 events. The new per_addr option counts the number of different
433 recipients that someone has sent messages to in the last time period. It
434 behaves like per_rcpt if all the recipient addresses are different, but
435 duplicate recipient addresses do not increase the measured rate. Like
436 the /count= option this is a general mechanism, so the per_addr option
437 is equivalent to per_rcpt/unique=$local_part@$domain. You can, for
438 example, measure the rate that a client uses different sender addresses
439 with the options per_mail/unique=$sender_address. There are further
440 details in the main documentation.
442 TF/05 Removed obsolete $Cambridge$ CVS revision strings.
444 TF/06 Removed a few PCRE remnants.
446 TF/07 Automatically extract Exim's version number from tags in the git
447 repository when doing development or release builds.
449 PP/02 Raise smtp_cmd_buffer_size to 16kB.
450 Bugzilla 879. Patch from Paul Fisher.
452 PP/03 Implement SSL-on-connect outbound with protocol=smtps on smtp transport.
453 Heavily based on revision 40f9a89a from Simon Arlott's tree.
456 PP/04 Use .dylib instead of .so for dynamic library loading on MacOS.
458 PP/05 Variable $av_failed, true if the AV scanner deferred.
459 Bugzilla 1078. Patch from John Horne.
461 PP/06 Stop make process more reliably on build failure.
462 Bugzilla 1087. Patch from Heiko Schlittermann.
464 PP/07 Make maildir_use_size_file an _expandable_ boolean.
465 Bugzilla 1089. Patch from Heiko Schlittermann.
467 PP/08 Handle ${run} returning more data than OS pipe buffer size.
468 Bugzilla 1131. Patch from Holger Weiß.
470 PP/09 Handle IPv6 addresses with SPF.
471 Bugzilla 860. Patch from Wolfgang Breyha.
473 PP/10 GnuTLS: support TLS 1.2 & 1.1.
475 Use gnutls_certificate_verify_peers2() [patch from Andreas Metzler].
478 PP/11 match_* no longer expand right-hand-side by default.
479 New compile-time build option, EXPAND_LISTMATCH_RHS.
480 New expansion conditions, "inlist", "inlisti".
482 PP/12 fix uninitialised greeting string from PP/03 (smtps client support).
484 PP/13 shell and compiler warnings fixes for RC1-RC4 changes.
486 PP/14 fix log_write() format string regression from TF/03.
487 Bugzilla 1152. Patch from Dmitry Isaikin.
493 PP/01 The new ldap_require_cert option would segfault if used. Fixed.
495 PP/02 Harmonised TLS library version reporting; only show if debugging.
496 Layout now matches that introduced for other libraries in 4.74 PP/03.
498 PP/03 New openssl_options items: no_sslv2 no_sslv3 no_ticket no_tlsv1
500 PP/04 New "dns_use_edns0" global option.
502 PP/05 Don't segfault on misconfiguration of ref:name exim-user as uid.
505 PP/06 Extra paranoia around buffer usage at the STARTTLS transition.
506 nb: Exim is not vulnerable to http://www.kb.cert.org/vuls/id/555316
508 TK/01 Updated PolarSSL code to 0.14.2.
509 Bugzilla 1097. Patch from Andreas Metzler.
511 PP/07 Catch divide-by-zero in ${eval:...}.
514 PP/08 Condition negation of bool{}/bool_lax{} did not negate. Fixed.
517 TK/02 Bugzilla 1106: CVE-2011-1764 - DKIM log line was subject to a
518 format-string attack -- SECURITY: remote arbitrary code execution.
520 TK/03 SECURITY - DKIM signature header parsing was double-expanded, second
521 time unintentionally subject to list matching rules, letting the header
522 cause arbitrary Exim lookups (of items which can occur in lists, *not*
523 arbitrary string expansion). This allowed for information disclosure.
525 PP/09 Fix another SIGFPE (x86) in ${eval:...} expansion, this time related to
526 INT_MIN/-1 -- value coerced to INT_MAX.
532 NM/01 Workround for PCRE version dependency in version reporting
535 TF/01 Update valgrind.h and memcheck.h to copies from valgrind-3.6.0.
536 This fixes portability to compilers other than gcc, notably
537 Solaris CC and HP-UX CC. Fixes Bugzilla 1050.
539 TF/02 Bugzilla 139: Avoid using the += operator in the modular lookup
540 makefiles for portability to HP-UX and POSIX correctness.
542 PP/01 Permit LOOKUP_foo enabling on the make command-line.
543 Also via indented variable definition in the Makefile.
544 (Debugging by Oliver Heesakkers).
546 PP/02 Restore caching of spamd results with expanded spamd_address.
547 Patch from author of expandable spamd_address patch, Wolfgang Breyha.
549 PP/03 Build issue: lookups-Makefile now exports LC_ALL=C
550 Improves build reliability. Fix from: Frank Elsner
552 NM/02 Fix wide character breakage in the rfc2047 coding
553 Fixes bug 1064. Patch from Andrey N. Oktyabrski
555 NM/03 Allow underscore in dnslist lookups
556 Fixes bug 1026. Patch from Graeme Fowler
558 PP/04 Bugzilla 230: Support TLS-enabled LDAP (in addition to ldaps).
559 Code patches from Adam Ciarcinski of NetBSD.
561 NM/04 Fixed exiqgrep to cope with mailq missing size issue
564 PP/05 Bugzilla 1083: when lookup expansion defers, escape the output which
565 is logged, to avoid truncation. Patch from John Horne.
567 PP/06 Bugzilla 1042: implement freeze_signal on pipe transports.
568 Patch from Jakob Hirsch.
570 PP/07 Bugzilla 1061: restrict error messages sent over SMTP to not reveal
571 SQL string expansion failure details.
572 Patch from Andrey Oktyabrski.
574 PP/08 Bugzilla 486: implement %M datestamping in log filenames.
575 Patch from Simon Arlott.
577 PP/09 New lookups functionality failed to compile on old gcc which rejects
578 extern declarations in function scope.
579 Patch from Oliver Fleischmann
581 PP/10 Use sig_atomic_t for flags set from signal handlers.
582 Check getgroups() return and improve debugging.
583 Fixed developed for diagnosis in bug 927 (which turned out to be
586 PP/11 Bugzilla 1055: Update $message_linecount for maildir_tag.
587 Patch from Mark Zealey.
589 PP/12 Bugzilla 1056: Improved spamd server selection.
590 Patch from Mark Zealey.
592 PP/13 Bugzilla 1086: Deal with maildir quota file races.
593 Based on patch from Heiko Schlittermann.
595 PP/14 Bugzilla 1019: DKIM multiple signature generation fix.
596 Patch from Uwe Doering, sign-off by Michael Haardt.
598 NM/05 Fix to spam.c to accommodate older gcc versions which dislike
599 variable declaration deep within a block. Bug and patch from
602 PP/15 lookups-Makefile IRIX compatibilty coercion.
604 PP/16 Make DISABLE_DKIM build knob functional.
606 NM/06 Bugzilla 968: child_open_uid: restore default SIGPIPE handler
607 Patch by Simon Arlott
609 TF/03 Fix valgrind.h portability to C89 compilers that do not support
610 variable argument macros. Our copy now differs from upstream.
616 TF/01 Failure to get a lock on a hints database can have serious
617 consequences so log it to the panic log.
619 TF/02 Log LMTP confirmation messages in the same way as SMTP,
620 controlled using the smtp_confirmation log selector.
622 TF/03 Include the error message when we fail to unlink a spool file.
624 DW/01 Bugzilla 139: Support dynamically loaded lookups as modules.
625 With thanks to Steve Haslam, Johannes Berg & Serge Demonchaux
626 for maintaining out-of-tree patches for some time.
628 PP/01 Bugzilla 139: Documentation and portability issues.
629 Avoid GNU Makefile-isms, let Exim continue to build on BSD.
630 Handle per-OS dynamic-module compilation flags.
632 PP/02 Let /dev/null have normal permissions.
633 The 4.73 fixes were a little too stringent and complained about the
634 permissions on /dev/null. Exempt it from some checks.
635 Reported by Andreas M. Kirchwitz.
637 PP/03 Report version information for many libraries, including
638 Exim version information for dynamically loaded libraries. Created
639 version.h, now support a version extension string for distributors
640 who patch heavily. Dynamic module ABI change.
642 PP/04 CVE-2011-0017 - check return value of setuid/setgid. This is a
643 privilege escalation vulnerability whereby the Exim run-time user
644 can cause root to append content of the attacker's choosing to
647 PP/05 Bugzilla 1041: merged DCC maintainer's fixes for return code.
650 PP/06 Bugzilla 1071: fix delivery logging with untrusted macros.
651 If dropping privileges for untrusted macros, we disabled normal logging
652 on the basis that it would fail; for the Exim run-time user, this is not
653 the case, and it resulted in successful deliveries going unlogged.
654 Fixed. Reported by Andreas Metzler.
660 PP/01 Date: & Message-Id: revert to normally being appended to a message,
661 only prepend for the Resent-* case. Fixes regression introduced in
662 Exim 4.70 by NM/22 for Bugzilla 607.
664 PP/02 Include check_rfc2047_length in configure.default because we're seeing
665 increasing numbers of administrators be bitten by this.
667 JJ/01 Added DISABLE_DKIM and comment to src/EDITME
669 PP/03 Bugzilla 994: added openssl_options main configuration option.
671 PP/04 Bugzilla 995: provide better SSL diagnostics on failed reads.
673 PP/05 Bugzilla 834: provide a permit_coredump option for pipe transports.
675 PP/06 Adjust NTLM authentication to handle SASL Initial Response.
677 PP/07 If TLS negotiated an anonymous cipher, we could end up with SSL but
678 without a peer certificate, leading to a segfault because of an
679 assumption that peers always have certificates. Be a little more
680 paranoid. Problem reported by Martin Tscholak.
682 PP/08 Bugzilla 926: switch ClamAV to use the new zINSTREAM API for content
683 filtering; old API available if built with WITH_OLD_CLAMAV_STREAM=yes
684 NB: ClamAV planning to remove STREAM in "middle of 2010".
685 CL also introduces -bmalware, various -d+acl logging additions and
686 more caution in buffer sizes.
688 PP/09 Implemented reverse_ip expansion operator.
690 PP/10 Bugzilla 937: provide a "debug" ACL control.
692 PP/11 Bugzilla 922: Documentation dusting, patch provided by John Horne.
694 PP/12 Bugzilla 973: Implement --version.
696 PP/13 Bugzilla 752: Refuse to build/run if Exim user is root/0.
698 PP/14 Build without WITH_CONTENT_SCAN. Path from Andreas Metzler.
700 PP/15 Bugzilla 816: support multiple condition rules on Routers.
702 PP/16 Add bool_lax{} expansion operator and use that for combining multiple
703 condition rules, instead of bool{}. Make both bool{} and bool_lax{}
704 ignore trailing whitespace.
706 JJ/02 prevent non-panic DKIM error from being sent to paniclog
708 JJ/03 added tcp_wrappers_daemon_name to allow host entries other than
711 PP/17 Fix malware regression for cmdline scanner introduced in PP/08.
712 Notification from Dr Andrew Aitchison.
714 PP/18 Change ClamAV response parsing to be more robust and to handle ClamAV's
715 ExtendedDetectionInfo response format.
716 Notification from John Horne.
718 PP/19 OpenSSL 1.0.0a compatibility const-ness change, should be backwards
721 PP/20 Added a CONTRIBUTING file. Fixed the documentation build to use http:
722 XSL and documented dependency on system catalogs, with examples of how
725 DW/21 Added Valgrind hooks in store.c to help it capture out-of-bounds store
728 DW/22 Bugzilla 1044: CVE-2010-4345 - partial fix: restrict default behaviour
729 of CONFIGURE_OWNER and CONFIGURE_GROUP options to no longer allow a
730 configuration file which is writeable by the Exim user or group.
732 DW/23 Bugzilla 1044: CVE-2010-4345 - part two: extend checks for writeability
733 of configuration files to cover files specified with the -C option if
734 they are going to be used with root privileges, not just the default
737 DW/24 Bugzilla 1044: CVE-2010-4345 - part three: remove ALT_CONFIG_ROOT_ONLY
738 option (effectively making it always true).
740 DW/25 Add TRUSTED_CONFIG_PREFIX_FILE option to allow alternative configuration
741 files to be used while preserving root privileges.
743 DW/26 Set FD_CLOEXEC on SMTP sockets after forking in the daemon, to ensure
744 that rogue child processes cannot use them.
746 PP/27 Bugzilla 1047: change the default for system_filter_user to be the Exim
747 run-time user, instead of root.
749 PP/28 Add WHITELIST_D_MACROS option to let some macros be overridden by the
750 Exim run-time user without dropping privileges.
752 DW/29 Remove use of va_copy() which breaks pre-C99 systems. Duplicate the
753 result string, instead of calling string_vformat() twice with the same
756 DW/30 Allow TRUSTED_CONFIG_PREFIX_FILE only for Exim or CONFIGURE_OWNER, not
757 for other users. Others should always drop root privileges if they use
758 -C on the command line, even for a whitelisted configure file.
760 DW/31 Turn TRUSTED_CONFIG_PREFIX_FILE into TRUSTED_CONFIG_FILE. No prefixes.
762 NM/01 Fixed bug #1002 - Message loss when using multiple deliveries
768 JJ/01 installed exipick 20100104.1, adding $max_received_linelength,
769 $data_path, and $header_path variables; fixed documentation bugs and
772 JJ/02 installed exipick 20100222.0, added --input-dir and --finput to allow
773 exipick to access non-standard spools, including the "frozen" queue
776 NM/01 Bugzilla 965: Support mysql stored procedures.
777 Patch from Alain Williams
779 NM/02 Bugzilla 961: Spacing fix (syntax error) on Makefile directives for NetBSD
781 NM/03 Bugzilla 955: Documentation fix for max_rcpts.
782 Patch from Andreas Metzler
784 NM/04 Bugzilla 954: Fix for unknown responses from Dovecot authenticator.
785 Patch from Kirill Miazine
787 NM/05 Bugzilla 671: Added umask to procmail example.
789 JJ/03 installed exipick 20100323.0, fixing doc bug
791 NM/06 Bugzilla 988: CVE-2010-2023 - prevent hardlink attack on sticky mail
792 directory. Notification and patch from Dan Rosenberg.
794 TK/01 PDKIM: Upgrade PolarSSL files to upstream version 0.12.1.
796 TK/02 Improve log output when DKIM signing operation fails.
798 MH/01 Treat the transport option dkim_domain as a colon separated
799 list, not as a single string, and sign the message with each element,
800 omitting multiple occurences of the same signer.
802 NM/07 Null terminate DKIM strings, Null initialise DKIM variable
803 Bugzilla 985, 986. Patch by Simon Arlott
805 NM/08 Bugzilla 967. dnsdb DNS TXT record bug fix (DKIM-related)
806 Patch by Simon Arlott
808 PP/01 Bugzilla 989: CVE-2010-2024 - work round race condition on
809 MBX locking. Notification from Dan Rosenberg.
815 TK/01 Bugzilla 912: Fix DKIM segfault on empty headers/body.
817 NM/01 Bugzilla 913: Documentation fix for gnutls_* options.
819 NM/02 Bugzilla 722: Documentation for randint. Better randomness defaults.
821 NM/03 Bugzilla 847: Enable DNSDB lookup by default.
823 NM/04 Bugzilla 915: Flag broken perl installation during build.
829 TK/01 Added patch by Johannes Berg that expands the main option
830 "spamd_address" if it starts with a dollar sign.
832 TK/02 Write list of recipients to X-Envelope-Sender header when building
833 the mbox-format spool file for content scanning (suggested by Jakob
836 TK/03 Added patch by Wolfgang Breyha that adds experimental DCC
837 (http://www.dcc-servers.net/) support via dccifd. Activated by
838 setting EXPERIMENTAL_DCC=yes in Local/Makefile.
840 TK/04 Bugzilla 673: Add f-protd malware scanner support. Patch submitted
841 by Mark Daniel Reidel <mr@df.eu>.
843 NM/01 Bugzilla 657: Embedded PCRE removed from the exim source tree.
844 When building exim an external PCRE library is now needed -
845 PCRE is a system library on the majority of modern systems.
846 See entry on PCRE_LIBS in EDITME file.
848 NM/02 Bugzilla 646: Removed unwanted C/R in Dovecot authenticator
849 conversation. Added nologin parameter to request.
850 Patch contributed by Kirill Miazine.
852 TF/01 Do not log submission mode rewrites if they do not change the address.
854 TF/02 Bugzilla 662: Fix stack corruption before exec() in daemon.c.
856 NM/03 Bugzilla 602: exicyclog now handles panic log, and creates empty
857 log files in place. Contributed by Roberto Lima.
859 NM/04 Bugzilla 667: Close socket used by dovecot authenticator.
861 TF/03 Bugzilla 615: When checking the local_parts router precondition
862 after a local_part_suffix or local_part_prefix option, Exim now
863 does not use the address's named list lookup cache, since this
864 contains cached lookups for the whole local part.
866 NM/05 Bugzilla 521: Integrated SPF Best Guess support contributed by
867 Robert Millan. Documentation is in experimental-spec.txt.
869 TF/04 Bugzilla 668: Fix parallel build (make -j).
871 NM/05.2 Bugzilla 437: Prevent Maildir aux files being created with mode 000.
873 NM/05.3 Bugzilla 598: Improvement to Dovecot authenticator handling.
874 Patch provided by Jan Srzednicki.
876 TF/05 Leading white space used to be stripped from $spam_report which
877 wrecked the formatting. Now it is preserved.
879 TF/06 Save $spam_score, $spam_bar, and $spam_report in spool files, so
880 that they are available at delivery time.
882 TF/07 Fix the way ${extract is skipped in the untaken branch of a conditional.
884 TF/08 TLS error reporting now respects the incoming_interface and
885 incoming_port log selectors.
887 TF/09 Produce a more useful error message if an SMTP transport's hosts
888 setting expands to an empty string.
890 NM/06 Bugzilla 744: EXPN did not work under TLS.
891 Patch provided by Phil Pennock.
893 NM/07 Bugzilla 769: Extraneous comma in usage fprintf
894 Patch provided by Richard Godbee.
896 NM/08 Fixed erroneous documentation references to smtp_notquit_acl to be
897 acl_smtp_notquit, added index entry.
899 NM/09 Bugzilla 787: Potential buffer overflow in string_format.
900 Patch provided by Eugene Bujak.
902 NM/10 Bugzilla 770: Problem on some platforms modifying the len parameter to
903 accept(). Patch provided by Maxim Dounin.
905 NM/11 Bugzilla 749: Preserve old behaviour of blanks comparing equal to zero.
906 Patch provided by Phil Pennock.
908 NM/12 Bugzilla 497: Correct behaviour of exiwhat when no config exists.
910 NM/13 Bugzilla 590: Correct handling of Resent-Date headers.
911 Patch provided by Brad "anomie" Jorsch.
913 NM/14 Bugzilla 622: Added timeout setting to transport filter.
914 Patch provided by Dean Brooks.
916 TK/05 Add native DKIM support (does not depend on external libraries).
918 NM/15 Bugzilla 854: Removed code that symlinks to pcre as its no longer useful.
919 Patch provided by Graeme Fowler.
921 NM/16 Bugzilla 851: Documentation example syntax fix.
923 NM/17 Changed NOTICE file to remove references to embedded PCRE.
925 NM/18 Bugzilla 894: Fix issue with very long lines including comments in
928 NM/19 Bugzilla 745: TLS version reporting.
929 Patch provided by Phil Pennock.
931 NM/20 Bugzilla 167: bool: condition support.
932 Patch provided by Phil Pennock.
934 NM/21 Bugzilla 665: gnutls_compat_mode to allow compatibility with broken
935 clients. Patch provided by Phil Pennock.
937 NM/22 Bugzilla 607: prepend (not append) Resent-Message-ID and Resent-Date.
938 Patch provided by Brad "anomie" Jorsch.
940 NM/23 Bugzilla 687: Fix misparses in eximstats.
941 Patch provided by Heiko Schlittermann.
943 NM/24 Bugzilla 688: Fix exiwhat to handle log_selector = +pid.
944 Patch provided by Heiko Schlittermann.
946 NM/25 Bugzilla 727: Use transport mode as default mode for maildirsize file.
947 plus update to original patch.
949 NM/26 Bugzilla 799: Documentation correction for ratelimit.
951 NM/27 Bugzilla 802: Improvements to local interface IP addr detection.
952 Patch provided by David Brownlee.
954 NM/28 Bugzilla 807: Improvements to LMTP delivery logging.
956 NM/29 Bugzilla 862, 866, 875: Documentation bugfixes.
958 NM/30 Bugzilla 888: TLS documentation bugfixes.
960 NM/31 Bugzilla 896: Dovecot buffer overrun fix.
962 NM/32 Bugzilla 889: Change all instances of "expr" in shell scripts to "expr --"
963 Unlike the original bugzilla I have changed all shell scripts in src tree.
965 NM/33 Bugzilla 898: Transport filter timeout fix.
966 Patch by Todd Rinaldo.
968 NM/34 Bugzilla 901: Fix sign/unsigned and UTF mismatches.
969 Patch by Serge Demonchaux.
971 NM/35 Bugzilla 39: Base64 decode bug fixes.
972 Patch by Jakob Hirsch.
974 NM/36 Bugzilla 909: Correct connect() call in dcc code.
976 NM/37 Bugzilla 910: Correct issue with relaxed/simple handling.
978 NM/38 Bugzilla 908: Removed NetBSD3 support as no longer needed.
980 NM/39 Bugzilla 911: Fixed MakeLinks build script.
986 TK/01 Add preliminary DKIM support. Currently requires a forked version of
987 ALT-N's libdkim that I have put here:
988 http://duncanthrax.net/exim-experimental/
990 Note to Michael Haardt: I had to rename some vars in sieve.c. They
991 were called 'true' and it seems that C99 defines that as a reserved
992 keyword to be used with 'bool' variable types. That means you could
993 not include C99-style headers which use bools without triggering
994 build errors in sieve.c.
996 NM/01 Bugzilla 592: --help option is handled incorrectly if exim is invoked
997 as mailq or other aliases. Changed the --help handling significantly
998 to do whats expected. exim_usage() emits usage/help information.
1000 SC/01 Added the -bylocaldomain option to eximstats.
1002 NM/02 Bugzilla 619: Defended against bad data coming back from gethostbyaddr.
1004 NM/03 Bugzilla 613: Documentation fix for acl_not_smtp.
1006 NM/04 Bugzilla 628: PCRE update to 7.4 (work done by John Hall).
1012 PH/01 Another patch from the Sieve maintainer.
1014 PH/02 When an IPv6 address is converted to a string for single-key lookup
1015 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
1016 dots are used instead of colons so that keys in lsearch files need not
1017 contain colons. This was done some time before quoting was made available
1018 in lsearch files. However, iplsearch files do require colons in IPv6 keys
1019 (notated using the quote facility) so as to distinguish them from IPv4
1020 keys. This meant that lookups for IP addresses in host lists did not work
1021 for iplsearch lookups.
1023 This has been fixed by arranging for IPv6 addresses to be expressed with
1024 colons if the lookup type is iplsearch. This is not incompatible, because
1025 previously such lookups could never work.
1027 The situation is now rather anomolous, since one *can* have colons in
1028 ordinary lsearch keys. However, making the change in all cases is
1029 incompatible and would probably break a number of configurations.
1031 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
1034 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
1035 conversion specification without a maximum field width, thereby enabling
1036 a rogue spamd server to cause a buffer overflow. While nobody in their
1037 right mind would setup Exim to query an untrusted spamd server, an
1038 attacker that gains access to a server running spamd could potentially
1039 exploit this vulnerability to run arbitrary code as the Exim user.
1041 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
1042 $primary_hostname instead of what libspf2 thinks the hosts name is.
1044 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
1045 a directory entry by the name of the lookup key. Previously, if a
1046 symlink pointed to a non-existing file or a file in a directory that
1047 Exim lacked permissions to read, a lookup for a key matching that
1048 symlink would fail. Now it is enough that a matching directory entry
1049 exists, symlink or not. (Bugzilla 503.)
1051 PH/03 The body_linecount and body_zerocount variables are now exported in the
1054 PH/04 Added the $dnslist_matched variable.
1056 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
1057 This means they are set thereafter only if the connection becomes
1060 PH/06 Added the client_condition to authenticators so that some can be skipped
1061 by clients under certain conditions.
1063 PH/07 The error message for a badly-placed control=no_multiline_responses left
1064 "_responses" off the end of the name.
1066 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
1068 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
1069 (without spaces) instead of just copying the configuration text.
1071 PH/10 Added the /noupdate option to the ratelimit ACL condition.
1073 PH/11 Added $max_received_linelength.
1075 PH/12 Added +ignore_defer and +include_defer to host lists.
1077 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
1078 way in which PCRE > 7.0 is built.
1080 PH/14 Implemented queue_only_load_latch.
1082 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
1083 MAIL command. The effect was to mangle the value on 64-bit systems.
1085 PH/16 Another patch from the Sieve maintainer.
1087 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
1089 PH/18 If a system quota error occurred while trying to create the file for
1090 a maildir delivery, the message "Mailbox is full" was not appended to the
1091 bounce if the delivery eventually timed out. Change 4.67/27 below applied
1092 only to a quota excession during the actual writing of the file.
1094 PH/19 It seems that peer DN values may contain newlines (and other non-printing
1095 characters?) which causes problems in log lines. The DN values are now
1096 passed through string_printing() before being added to log lines.
1098 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
1099 and InterBase are left for another time.)
1101 PH/21 Added message_body_newlines option.
1103 PH/22 Guard against possible overflow in moan_check_errorcopy().
1105 PH/23 POSIX allows open() to be a macro; guard against that.
1107 PH/24 If the recipient of an error message contained an @ in the local part
1108 (suitably quoted, of course), incorrect values were put in $domain and
1109 $local_part during the evaluation of errors_copy.
1115 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
1116 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
1119 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
1120 issue a MAIL command.
1122 PH/02 In an ACL statement such as
1124 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
1126 if a client was not listed at all, or was listed with a value other than
1127 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
1128 the condition was not true (as it should be), so access was not denied.
1129 The bug was that the ! inversion was incorrectly passed on to the second
1130 item. This has been fixed.
1132 PH/03 Added additional dnslists conditions == and =& which are different from
1133 = and & when the dns lookup returns more than one IP address.
1135 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
1136 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
1138 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
1139 FSYNC, which compiles an option called disable_fsync that allows for
1140 bypassing fsync(). The documentation is heavily laced with warnings.
1142 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
1144 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
1145 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
1146 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
1147 including adding "make clean"; (3) Added -fPIC when compiling the test
1148 dynamically loaded module, to get rid of a warning.
1150 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
1151 message fails, move_frozen_messages = true and ignore_bounce_errors_after
1152 = 0s. The bug is otherwise harmless.
1154 PH/07 There was a bug in the dovecot authenticator such that the value of
1155 $auth1 could be overwritten, and so not correctly preserved, after a
1156 successful authentication. This usually meant that the value preserved by
1157 the server_setid option was incorrect.
1159 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
1161 PH/09 Installed PCRE release 7.0.
1163 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
1164 run for batched SMTP input. It is now run at the start of every message
1165 in the batch. While fixing this I discovered that the process information
1166 (output by running exiwhat) was not always getting set for -bs and -bS
1167 input. This is fixed, and it now also says "batched" for BSMTP.
1169 PH/11 Added control=no_pipelining.
1171 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
1172 patch, slightly modified), and move the expansion of helo_data till after
1173 the connection is made in the smtp transport (so it can use these
1176 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
1178 PH/14 Added log_selector = +pid.
1180 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
1182 PH/16 Add ${if forany and ${if forall.
1184 PH/17 Added dsn_from option to vary the From: line in DSNs.
1186 PH/18 Flush SMTP output before performing a callout, unless control =
1187 no_callout_flush is set.
1189 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
1190 was true (the default) a successful delivery failed to delete the retry
1191 item, thus causing premature timeout of the address. The bug is now
1194 PH/20 Added hosts_avoid_pipelining to the smtp transport.
1196 PH/21 Long custom messages for fakedefer and fakereject are now split up
1197 into multiline reponses in the same way that messages for "deny" and
1198 other ACL rejections are.
1200 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
1201 with slight modification.
1203 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
1204 draft, changing the syntax and factoring some duplicate code".
1206 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
1207 for deliveries of the second and subsequent messages over the same SMTP
1210 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
1211 ${reduce, with only minor "tidies".
1213 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
1215 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
1216 expansion side effects.
1218 PH/27 When a message times out after an over-quota error from an Exim-imposed
1219 quota, the bounce message says "mailbox is full". This message was not
1220 being given when it was a system quota that was exceeded. It now should
1223 MH/03 Made $recipients available in local_scan(). local_scan() already has
1224 better access to the recipient list through recipients_list[], but
1225 $recipients can be useful in postmaster-provided expansion strings.
1227 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
1228 in the case of a MAIL command with additional options following the
1229 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
1230 were accidentally chopped off.
1232 PH/29 SMTP synchronization checks are implemented when a command is read -
1233 there is a check that no more input is waiting when there shouldn't be
1234 any. However, for some commands, a delay in an ACL can mean that it is
1235 some time before the response is written. In this time, more input might
1236 arrive, invalidly. So now there are extra checks after an ACL has run for
1237 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
1238 pipelining has not been advertised.
1240 PH/30 MH's patch to allow iscntrl() characters to be list separators.
1242 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
1243 returned in the SMTP response when smtp_return_error_details was false.
1244 This has been fixed.
1246 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
1247 instead of the C I/O that was originally supplied, because problems were
1248 reported on Solaris.
1250 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
1251 Exim which did not show up earlier: it was assuming that a call to
1252 SSL_CTX_set_info_callback() might give an error value. In fact, there is
1253 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
1254 was a macro that became an assignment, so it seemed to work. This has
1255 changed to a proper function call with a void return, hence the compile
1256 error. Exim's code has been fixed.
1258 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
1261 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
1263 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
1265 PH/37 If a message is not accepted after it has had an id assigned (e.g.
1266 because it turns out to be too big or there is a timeout) there is no
1267 "Completed" line in the log. When some messages of this type were
1268 selected by exigrep, they were listed as "not completed". Others were
1269 picked up by some special patterns. I have improved the selection
1270 criteria to be more general.
1272 PH/38 The host_find_failed option in the manualroute router can now be set
1273 to "ignore", to completely ignore a host whose IP address cannot be
1274 found. If all hosts are ignored, the behaviour is controlled by the new
1275 host_all_ignored option.
1277 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
1278 homing or because of multiple MX records with /mx) generated more than
1279 one IP address, and the following item turned out to be the local host,
1280 all the secondary addresses of the first item were incorrectly removed
1281 from the list, along with the local host and any following hosts (which
1282 is what is supposed to happen).
1284 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
1285 whoever called Exim into the -H file. In the case of the daemon it was
1286 behaving confusingly. When first started, it used values for whoever
1287 started the daemon, but after a SIGHUP it used the Exim user (because it
1288 calls itself on a restart). I have changed the code so that it now always
1291 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
1292 message are rejected with the same error (e.g. no authentication or bad
1293 sender address), and a DATA command is nevertheless sent (as can happen
1294 with PIPELINING or a stupid MUA), the error message that was given to the
1295 RCPT commands is included in the rejection of the DATA command. This is
1296 intended to be helpful for MUAs that show only the final error to their
1299 PH/42 Another patch from the Sieve maintainer.
1301 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
1302 Eximstats - Fixed some broken HTML links and added missing column headers
1304 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
1305 columns for Rejects, Temp Rejects, Ham, and Spam rows.
1307 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
1309 PH/43 Yet another patch from the Sieve maintainer.
1311 PH/44 I found a way to check for a TCP/IP connection going away before sending
1312 the response to the final '.' that terminates a message, but only in the
1313 case where the client has not sent further data following the '.'
1314 (unfortunately, this is allowed). However, in many cases there won't be
1315 any further data because there won't be any more messages to send. A call
1316 to select() can be used: if it shows that the input is "ready", there is
1317 either input waiting, or the socket has been closed. An attempt to read
1318 the next input character can distinguish the two cases. Previously, Exim
1319 would have sent an OK response which the client would never have see.
1320 This could lead to message repetition. This fix should cure that, at
1321 least in a lot of common cases.
1323 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
1324 advertised in response to EHLO.
1330 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
1331 fixed by 4.65/MH/01 (is this a record?) are fixed:
1333 (i) An empty string was always treated as zero by the numeric comparison
1334 operators. This behaviour has been restored.
1336 (ii) It is documented that the numeric comparison operators always treat
1337 their arguments as decimal numbers. This was broken in that numbers
1338 starting with 0 were being interpreted as octal.
1340 While fixing these problems I realized that there was another issue that
1341 hadn't been noticed. Values of message_size_limit (both the global option
1342 and the transport option) were treated as octal if they started with 0.
1343 The documentation was vague. These values are now always treated as
1344 decimal, and I will make that clear in the documentation.
1350 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
1351 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
1354 MH/01 Don't check that the operands of numeric comparison operators are
1355 integers when their expansion is in "skipping" mode (fixes bug
1356 introduced by 4.64-PH/07).
1358 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
1359 child addresses, Exim now panics and dies. Previously, because the count
1360 is held in a short int, deliveries were likely to be lost. As such a
1361 large number of recipients for a single message is ridiculous
1362 (performance will be very, very poor), I have chosen to impose a limit
1363 rather than extend the field.
1369 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
1370 leftover -K file (the existence of which was triggered by #402).
1371 While we were at it, introduced process PID as part of the -K
1372 filename. This should rule out race conditions when creating
1375 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
1376 processing considerably. Previous code took too long for large mails,
1377 triggering a timeout which in turn triggers #401.
1379 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
1380 in the DK code in transports.c. sendfile() is not really portable,
1381 hence the _LINUX specificness.
1383 TF/01 In the add_headers option to the mail command in an Exim filter,
1384 there was a bug that Exim would claim a syntax error in any
1385 header after the first one which had an odd number of characters
1388 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
1389 callout verification, Exim cached a "reject" for the entire domain. This
1390 is correct for most verifications, but it is not correct for a recipient
1391 verification with use_sender or use_postmaster set, because in that case
1392 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
1393 case of MAIL FROM:<> rejection from other early rejections (e.g.
1394 rejection of HELO). When verifying a recipient using a non-null MAIL
1395 address, the cache is ignored if it shows MAIL FROM:<> rejection.
1396 Whatever the result of the callout, the value of the domain cache is
1397 left unchanged (for any other kind of callout, getting as far as trying
1398 RCPT means that the domain itself is ok).
1400 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
1403 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
1404 manifest itself as EPIPE rather than ECONNECT. When tidying away a
1405 session, the daemon ignores ECONNECT errors and logs others; it now
1406 ignores EPIPE as well.
1408 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
1409 (quoted-printable decoding).
1411 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
1412 later the small subsequent patch to fix an introduced bug.
1414 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
1416 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
1418 PH/08 An error is now given if message_size_limit is specified negative.
1420 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
1421 to be given (somewhat) arbitrary names.
1423 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
1426 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
1427 miscellaneous code fixes
1429 PH/10 Added the log_reject_target ACL modifier to specify where to log
1432 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
1433 hostname. This is wrong, because it relates to the incoming message (and
1434 probably the interface on which it is arriving) and not to the outgoing
1435 callout (which could be using a different interface). This has been
1436 changed to use the value of the helo_data option from the smtp transport
1437 instead - this is what is used when a message is actually being sent. If
1438 there is no remote transport (possible with a router that sets up host
1439 addresses), $smtp_active_hostname is used.
1441 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
1442 tweaks were necessary in order to get it to work (see also 21 below):
1443 (a) The code assumed that strncpy() returns a negative number on buffer
1444 overflow, which isn't the case. Replaced with Exim's string_format()
1446 (b) There were several signed/unsigned issues. I just did the minimum
1447 hacking in of casts. There is scope for a larger refactoring.
1448 (c) The code used strcasecmp() which is not a standard C function.
1449 Replaced with Exim's strcmpic() function.
1450 (d) The code set only $1; it now sets $auth1 as well.
1451 (e) A simple test gave the error "authentication client didn't specify
1452 service in request". It would seem that Dovecot has changed its
1453 interface. Fortunately there's a specification; I followed it and
1454 changed what the client sends and it appears to be working now.
1456 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
1459 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
1460 address A is aliased to B and C, where B exists and C does not. Without
1461 -v the output is "A verified" because verification stops after a
1462 successful redirection if more than one address is generated. However,
1463 with -v the child addresses are also verified. Exim was outputting "A
1464 failed to verify" and then showing the successful verification for C,
1465 with its parentage. It now outputs "B failed to verify", showing B's
1466 parentage before showing the successful verification of C.
1468 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
1469 look up a TXT record in a specific list after matching in a combined
1472 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
1473 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
1474 they consult the DNS. I had assumed they would set it the way they
1475 wanted; and indeed my experiments on Linux seem to show that in some
1476 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
1477 To be on the safe side, however, I have now made the interface to
1478 host_find_byname() similar to host_find_bydns(), with an argument
1479 containing the DNS resolver options. The host_find_byname() function now
1480 sets these options at its start, just as host_find_bydns() does. The smtp
1481 transport options dns_qualify_single and dns_search_parents are passed to
1482 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
1483 of host_find_byname() use the default settings of RES_DEFNAMES
1484 (qualify_single) but not RES_DNSRCH (search_parents).
1486 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
1487 spool_read_header() do less string testing, by means of a preliminary
1488 switch on the second character of optional "-foo" lines. (This is
1489 overdue, caused by the large number of possibilities that now exist.
1490 Originally there were few.) While I was there, I also converted the
1491 str(n)cmp tests so they don't re-test the leading "-" and the first
1492 character, in the hope this might squeeze out yet more improvement.
1494 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
1495 flag allowing group syntax was set by the header_syntax check but not
1496 turned off, possible causing trouble later; (2) The flag was not being
1497 set at all for the header_verify test, causing "group"-style headers to
1498 be rejected. I have now set it in this case, and also caused header_
1499 verify to ignore an empty address taken from a group. While doing this, I
1500 came across some other cases where the code for allowing group syntax
1501 while scanning a header line wasn't quite right (mostly, not resetting
1502 the flag correctly in the right place). These bugs could have caused
1503 trouble for malformed header lines. I hope it is now all correct.
1505 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
1506 with the "reply" argument non-NULL. The code, however (which originally
1507 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
1508 but it didn't always do it. This confused somebody who was copying the
1509 code for some other use. I have removed all the tests.
1511 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
1512 feature that was used to support insecure browsers during the U.S. crypto
1513 embargo. It requires special client support, and Exim is probably the
1514 only MTA that supported it -- and would never use it because real RSA is
1515 always available. This code has been removed, because it had the bad
1516 effect of slowing Exim down by computing (never used) parameters for the
1517 RSA_EXPORT functionality.
1519 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
1520 authenticator to fail if there's a tab character in the incoming data
1521 (there should never be unless someone is messing about, as it's supposed
1522 to be base64-encoded). Also added, on Timo's advice, the "secured" option
1523 if the connection is using TLS or if the remote IP is the same as the
1524 local IP, and the "valid-client-cert option" if a client certificate has
1527 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
1528 authenticators. This can be used for authorization after authentication
1529 succeeds. (In the case of plaintext, it servers for both authentication
1532 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
1533 if any retry times were supplied.
1535 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
1536 connection, where there is no client IP address to check. In this
1537 situation, the verify now always succeeds.
1539 PH/25 Applied John Jetmore's -Mset patch.
1541 PH/26 Added -bem to be like -Mset, but loading a message from a file.
1543 PH/27 In a string expansion for a processed (not raw) header when multiple
1544 headers of the same name were present, leading whitespace was being
1545 removed from all of them, but trailing whitespace was being removed only
1546 from the last one. Now trailing whitespace is removed from each header
1547 before concatenation. Completely empty headers in a concatenation (as
1548 before) are ignored.
1550 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
1551 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
1553 PH/29 [Removed. This was a change that I later backed out, and forgot to
1554 correct the ChangeLog entry (that I had efficiently created) before
1555 committing the later change.]
1557 PH/30 Exim was sometimes attempting to deliver messages that had suffered
1558 address errors (4xx response to RCPT) over the same connection as other
1559 messages routed to the same hosts. Such deliveries are always "forced",
1560 so retry times are not inspected. This resulted in far too many retries
1561 for the affected addresses. The effect occurred only when there were more
1562 hosts than the hosts_max_try setting in the smtp transport when it had
1563 the 4xx errors. Those hosts that it had tried were not added to the list
1564 of hosts for which the message was waiting, so if all were tried, there
1565 was no problem. Two fixes have been applied:
1567 (i) If there are any address or message errors in an SMTP delivery, none
1568 of the hosts (tried or untried) are now added to the list of hosts
1569 for which the message is waiting, so the message should not be a
1570 candidate for sending over the same connection that was used for a
1571 successful delivery of some other message. This seems entirely
1572 reasonable: after all the message is NOT "waiting for some host".
1573 This is so "obvious" that I'm not sure why it wasn't done
1574 previously. Hope I haven't missed anything, but it can't do any
1575 harm, as the worst effect is to miss an optimization.
1577 (ii) If, despite (i), such a delivery is accidentally attempted, the
1578 routing retry time is respected, so at least it doesn't keep
1579 hammering the server.
1581 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
1582 in ${readsocket because some servers need this prod.
1584 PH/32 Added some extra debug output when updating a wait-xxx database.
1586 PH/33 The hint "could be header name not terminated by colon", which has been
1587 given for certain expansion errors for a long time, was not being given
1588 for the ${if def:h_colon_omitted{... case.
1590 PH/34 The spec says: "With one important exception, whenever a domain list is
1591 being scanned, $domain contains the subject domain." There was at least
1592 one case where this was not true.
1594 PH/35 The error "getsockname() failed: connection reset by peer" was being
1595 written to the panic log as well as the main log, but it isn't really
1596 panic-worthy as it just means the connection died rather early on. I have
1597 removed the panic log writing for the ECONNRESET error when getsockname()
1600 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
1601 runs only) independently of the message's sender address. This meant
1602 that, if the 4xx error was in fact related to the sender, a different
1603 message to the same recipient with a different sender could confuse
1604 things. In particualar, this can happen when sending to a greylisting
1605 server, but other circumstances could also provoke similar problems.
1606 I have changed the default so that the retry time for these errors is now
1607 based a combination of the sender and recipient addresses. This change
1608 can be overridden by setting address_retry_include_sender=false in the
1611 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
1612 remote server are returned as part of bounce messages. This was not
1613 happening for LMTP over a pipe (the lmtp transport), but now it is the
1614 same for both kinds of LMTP.
1616 PH/38 Despite being documented as not happening, Exim was rewriting addresses
1617 in header lines that were in fact CNAMEs. This is no longer the case.
1619 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
1620 and queue runs started by the daemon processed all messages. This has
1621 been fixed so that -R and -S can now usefully be given with -q<time>.
1623 PH/40 Import PCRE release 6.7 (fixes some bugs).
1625 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
1627 PH/42 Give an error if -q is specified more than once.
1629 PH/43 Renamed the variables $interface_address and $interface_port as
1630 $received_ip_address and $received_port, to make it clear that these
1631 values apply to message reception, and not to the outgoing interface when
1632 a message is delivered. (The old names remain recognized, of course.)
1634 PH/44 There was no timeout on the connect() call when using a Unix domain
1635 socket in the ${readsocket expansion. There now is.
1637 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
1638 be meaningful with "accept".
1640 SC/01 Eximstats V1.43
1641 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
1643 SC/02 Eximstats V1.44
1644 Use a glob alias rather than an array ref in the generated
1645 parser. This improves both readability and performance.
1647 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
1648 Collect SpamAssassin and rejection statistics.
1649 Don't display local sender or destination tables unless
1650 there is data to show.
1651 Added average volumes into the top table text output.
1653 SC/04 Eximstats V1.46
1654 Collect data on the number of addresses (recipients)
1655 as well as the number of messages.
1657 SC/05 Eximstats V1.47
1658 Added 'Message too big' to the list of mail rejection
1659 reasons (thanks to Marco Gaiarin).
1661 SC/06 Eximstats V1.48
1662 Mainlog lines which have GMT offsets and are too short to
1663 have a flag are now skipped.
1665 SC/07 Eximstats V1.49 (Alain Williams)
1666 Added the -emptyok flag.
1668 SC/08 Eximstats V1.50
1669 Fixes for obtaining the IP address from reject messages.
1671 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
1672 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
1673 whitesspace changes from 4.64-PH/27
1675 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
1678 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
1679 are found, allow negative numbers in numeric criteria)
1681 JJ/06 exipick.20061117.2, added new $message_body_missing variable
1683 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
1684 to match changes made in 4.64-PH/43
1686 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
1688 PH/47 Put in an explicit test for a DNS lookup of an address record where the
1689 "domain" is actually an IP address, and force a failure. This locks out
1690 those revolvers/nameservers that support "A-for-A" lookups, in
1691 contravention of the specifications.
1693 PH/48 When a host name was looked up from an IP address, and the subsequent
1694 forward lookup of the name timed out, the host name was left in
1695 $sender_host_name, contrary to the specification.
1697 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
1698 restricted to single-key lookups, Exim was not diagnosing an error if
1699 * or *@ was used with a query-style lookup.
1701 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
1703 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
1704 long ago, but noone interested enough thought of it. Let's just say that
1705 the "1.1" means that there are some new functions that weren't there at
1706 some point in the past.
1708 PH/51 Error processing for expansion failure of helo_data from an smtp
1709 transport during callout processing was broken.
1711 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
1712 tested/used via the -bh/-bhc/-bs options.
1714 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
1715 bug, fixed in subsequent PCRE releases).
1717 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
1718 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
1720 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
1726 SC/01 Use a glob alias rather than an array ref in eximstats generated
1727 parser. This improves both readability and performance.
1729 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
1730 Don't display local sender or destination tables in eximstats unless
1731 there is data to show.
1732 Added average volumes into the eximstats top table text output.
1734 SC/03 Collect data on the number of addresses (recipients) as well
1735 as the number of messages in eximstats.
1737 TF/01 Correct an error in the documentation for the redirect router. Exim
1738 does (usually) call initgroups() when daemonizing.
1740 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
1741 with consistent privilege compared to when running as a daemon.
1743 TF/03 Note in the spec that $authenticated_id is not set for local
1744 submissions from trusted users.
1746 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
1747 Thanks to Dean Brooks <dean@iglou.com> for the patch.
1749 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
1750 by adding some example configuration directives to the default
1751 configuration file. A little bit of work is required to uncomment the
1752 directives and define how usernames and passwords are checked, but
1753 there is now a framework to start from.
1755 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
1756 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
1757 without this. I don't know how relevant this is to other LDAP libraries.
1759 PH/02 Add the verb name to the "unknown ACL verb" error.
1761 PH/03 Magnus Holmgren's patch for filter_prepend_home.
1763 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
1765 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
1766 directory not expanded when it should be if an expanded home directory
1767 was set for the address (which is overridden by the transport).
1769 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
1772 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
1773 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
1774 because it is too late at that time, and has no effect.
1776 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
1777 security issue with \' (bugzilla #107). I could not use the
1778 PQescapeStringConn() function, because it needs a PGconn value as one of
1781 PH/08 When testing addresses using -bt, indicate those final addresses that
1782 are duplicates that would not cause an additional delivery. At least one
1783 person was confused, thinking that -bt output corresponded to deliveries.
1784 (Suppressing duplicates isn't a good idea as you lose the information
1785 about possibly different redirections that led to the duplicates.)
1787 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
1788 systems where poll() doesn't work, in particular OS X.
1790 PH/10 Added more information to debugging output for retry time not reached.
1792 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
1793 operations in malware.c.
1795 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
1798 PH/13 If write_rejectlog was set false when logging was sent to syslog with
1799 syslog_duplication set false, log lines that would normally be written
1800 both the the main log and to the reject log were not written to syslog at
1803 PH/14 In the default configuration, change the use of "message" in ACL warn
1804 statements to "add_header".
1806 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
1807 not followed by a command (e.g. "seen endif").
1809 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
1810 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
1813 PH/17 Added extra conditions to the default value of delay_warning_condition
1817 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
1818 { match{$h_precedence:}{(?i)bulk|list|junk} } \
1819 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
1822 The Auto-Submitted: and various List- headers are standardised, whereas I
1823 don't think Precedence: ever was.
1825 PH/18 Refactored debugging code in route_finduser() to show more information,
1826 in particular, the error code if getpwnam() issues one.
1828 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
1829 This is apparently needed in addition to the PH/07 change above to avoid
1830 any possible encoding problems.
1832 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
1833 but not after initializing Perl.
1835 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
1836 output them only if debugging. By default they are written stderr,
1837 apparently, which is not desirable.
1839 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
1842 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
1845 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
1847 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
1848 authenticated or an ident call has been made. Suppress the default
1849 values for $authenticated_id and $authenticated_sender (but permit -oMai
1850 and -oMas) when testing with -bh.
1852 PH/24 Re-jigged the order of the tests in the default configuration so that the
1853 tests for valid domains and recipients precede the DNS black list and CSA
1854 tests, on the grounds that those ones are more expensive.
1856 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
1857 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
1858 This bug exists in every version of Exim that I still have, right back to
1861 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
1862 However, an attempt to turn on case-sensitivity in a regex key by
1863 including (?-i) didn't work because the subject string was already
1864 lowercased, and the effects were non-intuitive. It turns out that a
1865 one-line patch can be used to allow (?-i) to work as expected.
1871 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
1872 other effects) broke the use of negated acl sub-conditions.
1874 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
1877 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
1878 "Deny" causes Exim to reject the incoming connection with a 554 error.
1879 Unfortunately, if there is a major crisis, such as a disk failure,
1880 tcp-wrappers gives "deny", whereas what one would like would be some
1881 kind of temporary error. A kludge has been added to help with this.
1882 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
1883 554 error is used if errno is still zero or contains ENOENT (which occurs
1884 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
1887 PH/03 Add -lutil to the default FreeBSD LIBS setting.
1889 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
1890 errors. Otherwise a message that provokes a temporary error (when other
1891 messages do not) can cause a whole host to time out.
1893 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
1894 addresses were routed directly to files or pipes from a redirect router.
1895 File deliveries just didn't batch; pipe deliveries might have suffered
1898 PH/06 A failure to get a lock for a hints database would erroneously always say
1899 "Failed to get write lock", even when it was really a read lock.
1901 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
1902 of 0600. This has been changed to use the value of the lockfile_mode
1903 option (which defaults to 0600).
1905 PH/08 Applied small patch from the Sieve maintainer.
1907 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
1908 folder from quota calculations, a direct delivery into this folder messed
1909 up the contents of the maildirsize file. This was because the regex was
1910 used only to exclude .Trash (or whatever) when the size of the mailbox
1911 was calculated. There was no check that a delivery was happening into an
1912 excluded directory. This bug has been fixed by ignoring all quota
1913 processing for deliveries into excluded directories.
1915 PH/10 Added the maildirfolder_create_regex option to appendfile.
1921 PH/01 The code for finding all the local interface addresses on a FreeBSD
1922 system running IPv6 was broken. This may well have applied to all BSD
1923 systems, as well as to others that have similar system calls. The broken
1924 code found IPv4 interfaces correctly, but gave incorrect values for the
1925 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
1926 that it would not match correctly against @[] and not recognize the IPv6
1929 PH/02 The ipliteral router was not recognizing addresses of the form user@
1930 [ipv6:....] because it didn't know about the "ipv6:" prefix.
1932 PH/03 Added disable_ipv6.
1934 PH/04 Changed $reply_address to use the raw form of the headers instead of the
1935 decoded form, because it is most often used to construct To: headers
1936 lines in autoreplies, and the decoded form may well be syntactically
1937 invalid. However, $reply_address has leading white space removed, and all
1938 newlines turned into spaces so that the autoreply transport does not
1941 PH/05 If group was specified without a user on a router, and no group or user
1942 was specified on a transport, the group from the router was ignored.
1944 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
1945 for visible compile-time settings that can be used to change these
1946 numbers, for those that want even more. Backwards compatibility with old
1947 spool files has been maintained. However, going back to a previous Exim
1948 release will lost any variables that are in spool files.
1950 PH/07 Two small changes when running in the test harness: increase delay when
1951 passing a TCP/IP connection to a new process, in case the original
1952 process has to generate a bounce, and remove special handling of
1953 127.0.0.2 (sic), which is no longer necessary.
1955 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
1956 be the same on different OS.
1958 PH/09 Moved a debug statement in filter processing to avoid a race problem when
1961 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
1962 whether --show-vars was specified or not
1964 JJ/02 exipick: Added support for new ACL variable spool format introduced
1967 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
1968 syntactically invalid From: or Reply-to: line, and a filter used this to
1969 generate an autoreply, and therefore failed to obtain an address for the
1970 autoreply, Exim could try to deliver to a non-existent relative file
1971 name, causing unrelated and misleading errors. What now happens is that
1972 it logs this as a hard delivery error, but does not attempt to create a
1975 PH/11 The exinext utility has a -C option for testing purposes, but although
1976 the given file was scanned by exinext itself; it wasn't being passed on
1977 when Exim was called.
1979 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
1980 an end-of-file indication when reading a command response.
1982 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
1983 compiled. In many other places in Exim, IPv6 addresses are always
1984 recognized, so I have changed this. It also means that IPv4 domain
1985 literals of the form [IPV4:n.n.n.n] are now always recognized.
1987 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
1988 used if the router is not running as root, for example, when verifying at
1989 ACL time, or when using -bh. The debugging output from this situation was
1990 non-existent - all you got was a failure to exec. I have made two
1993 (a) Failures to set uid/gid, the current directory, or a process leader
1994 in a subprocess such as that created by queryprogram now generate
1995 suitable debugging ouput when -d is set.
1997 (b) The queryprogram router detects when it is not running as root,
1998 outputs suitable debugging information if -d is set, and then runs
1999 the subprocess without attempting to change uid/gid.
2001 PH/15 Minor change to Makefile for building test_host (undocumented testing
2004 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
2005 additional section of a DNS packet that returns MX or SRV records.
2006 Instead, it always explicitly searches for A/AAAA records. This avoids
2007 major problems that occur when a DNS server includes only records of one
2008 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
2009 fixed another bug: if SRV records were looked up and the corresponding
2010 address records were *not* found in the additional section, the port
2011 values from the SRV records were lost.
2013 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
2014 using the correct key (the original address) when searching the retry
2015 rules in order to find which one to use for generating the retry hint.
2017 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
2018 adding the default one. Similarly, if it contains a Reply-To: header, the
2019 errors_reply_to option, if set, is not used.
2021 PH/19 When calculating a retry time, Exim used to measure the "time since
2022 failure" by looking at the "first failed" field in the retry record. Now
2023 it does not use this if it is later than than the arrival time of the
2024 message. Instead it uses the arrival time. This makes for better
2025 behaviour in cases where some deliveries succeed, thus re-setting the
2026 "first failed" field. An example is a quota failure for a huge message
2027 when small messages continue to be delivered. Without this change, the
2028 "time since failure" will always be short, possible causing more frequent
2029 delivery attempts for the huge message than are intended.
2030 [Note: This change was subsequently modified - see PH/04 for 4.62.]
2032 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
2033 $1, $2, $3) because the numerical variables can be reset during some
2034 expansion items (e.g. "match"), thereby losing the authentication data.
2036 PH/21 Make -bV show the size of off_t variables so that the test suite can
2037 decide whether to run tests for quotas > 2G.
2039 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
2040 mailbox_size, and mailbox_filecount in the appendfile transport. If a
2041 filecount value is greater than 2G or if a quota value is greater than 2G
2042 on a system where the size of off_t is not greater than 4, a panic error
2045 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
2046 never match. The debug and -bh output now contains an explicit error
2047 message indicating a malformed IPv4 address or mask.
2049 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
2050 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
2051 PH/23 above applies.
2053 PH/25 Do not write to syslog when running in the test harness. The only
2054 occasion when this arises is a failure to open the main or panic logs
2055 (for which there is an explicit test).
2057 PH/26 Added the /no_tell option to "control=freeze".
2059 PH/27 If a host name lookup failed very early in a connection, for example, if
2060 the IP address matched host_lookup and the reverse lookup yielded a name
2061 that did not have a forward lookup, an error message of the form "no IP
2062 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
2063 could be logged. Now it outputs the IP address instead of "NULL".
2065 PH/28 An enabling patch from MH: add new function child_open_exim2() which
2066 allows the sender and the authenticated sender to be set when
2067 submitting a message from within Exim. Since child_open_exim() is
2068 documented for local_scan(), the new function should be too.
2070 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
2071 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
2072 results in an empty string is now treated as unset.
2074 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
2076 PH/31 Added sender_verify_fail logging option.
2078 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
2079 needed by RFC 822 but not by RFC 2822 was commented out. I have now
2080 tidied the source and removed it altogether.
2082 PH/33 When a queue run was abandoned because the load average was too high, a
2083 log line was always written; now it is written only if the queue_run log
2084 selector is set. In addition, the log line for abandonment now contains
2085 information about the queue run such as the pid. This is always present
2086 in "start" and "stop" lines but was omitted from the "abandon" line.
2088 PH/34 Omit spaces between a header name and the colon in the error message that
2089 is given when verify = headers_syntax fails (if there are lots of them,
2090 the message gets confusing).
2092 PH/35 Change the default for dns_check_names_pattern to allow slashes within
2093 names, as there are now some PTR records that contain slashes. This check
2094 is only to protect against broken name servers that fall over on strange
2095 characters, so the fact that it applies to all lookups doesn't matter.
2097 PH/36 Now that the new test suite is complete, we can remove some of the
2098 special code in Exim that was needed for the old test suite. For example,
2099 sorting DNS records because real resolvers return them in an arbitrary
2100 order. The new test suite's fake resolver always returns records in the
2103 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
2104 bounces) except when queue_only is set, to avoid logging races between
2105 the different processes.
2107 PH/38 Panic-die if .include specifies a non-absolute path.
2109 PH/39 A tweak to the "H" retry rule from its user.
2111 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
2112 a label. They prevented compilation on older perls.
2114 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
2115 a warning to be raised on newish perls.
2117 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
2118 on queue. Changes to match documented behaviour of showing count of
2119 messages matching specified criteria.
2121 PH/40 Changed the default ident timeout from 30s to 5s.
2123 PH/41 Added support for the use of login_cap features, on those BSD systems
2124 that have them, for controlling the resources used by pipe deliveries.
2126 PH/42 The content-scanning code uses fopen() to create files in which to put
2127 message data. Previously it was not paying any attention to the mode of
2128 the files. Exim runs with umask(0) because the rest of the code creates
2129 files with open(), and sets the required mode explicitly. Thus, these
2130 files were ending up world-writeable. This was not a big issue, because,
2131 being within the spool directory, they were not world-accessible. I have
2132 created a function called modefopen, which takes an additional mode
2133 argument. It sets umask(777), creates the file, chmods it to the required
2134 mode, then resets the umask. All the relevant calls to fopen() in the
2135 content scanning code have been changed to use this function.
2137 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
2138 to 24 hours. This avoids potential overflow problems when processing G
2139 and H retry rules. I suspect nobody ever tinkers with this value.
2141 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
2143 PH/45 When the plaintext authenticator is running as a client, the server's
2144 challenges are checked to ensure they are valid base64 strings. By
2145 default, the authentication attempt is cancelled if an invalid string is
2146 received. Setting client_ignore_invalid_base64 true ignores these errors.
2147 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
2148 they are received. Thus, the responses can be made to depend on the
2149 challenges. If an invalid string is ignored, an empty string is placed in
2152 PH/46 Messages that are created by the autoreply transport now contains a
2153 References: header, in accordance with RFCs 2822 and 3834.
2155 PH/47 Added authenticated_sender_force to the smtp transport.
2157 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
2159 PH/49 Installed latest patch from the Sieve maintainer.
2161 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
2162 was also set, the appendfile transport was unnecessarily scanning a
2163 directory of message files (e.g. for maildir delivery) to find the count
2164 of files (along with the size), even though it did not need this
2165 information. It now does the scan only if it needs to find either the
2166 size of the count of files.
2168 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
2170 PH/52 Two bugs concerned with error handling when the smtp transport is
2173 (i) Exim was not creating retry information for temporary errors given
2174 for individual recipients after the DATA command when the smtp transport
2175 was used in LMTP mode. This meant that they could be retried too
2176 frequently, and not timed out correctly.
2178 (ii) Exim was setting the flag that allows error details to be returned
2179 for LMTP errors on RCPT commands, but not for LMTP errors for individual
2180 recipients that were returned after the DATA command.
2182 PH/53 This is related to PH/52, but is more general: for any failing address,
2183 when detailed error information was permitted to be returned to the
2184 sender, but the error was temporary, then after the final timeout, only
2185 "retry timeout exceeded" was returned. Now it returns the full error as
2186 well as "retry timeout exceeded".
2188 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
2189 do this, and (what is worse) MTAs that accept it.
2191 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
2192 will now be deprecated.
2194 PH/56 New os.c-cygwin from the Cygwin maintainer.
2196 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
2197 formats (previously only available in exim formats via -bpr, -bpru,
2198 and -bpra. Now also available in native and exiqgrep formats)
2200 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
2201 with very large, slow to parse queues
2203 JJ/08 exipick: added ! as generic prefix to negate any criteria format
2205 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
2207 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
2208 responses to authentication challenges, though it was showing the
2209 challenges; (ii) I've removed the CR characters from the debug output for
2212 PH/58 Allow for the insertion of a newline as well as a space when a string
2213 is turned into more than one encoded-word during RFC 2047 encoding. The
2214 Sieve code now uses this.
2216 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
2217 data_4xx, lost_connection, tls_required.
2219 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
2220 message was being sent as an SMTP response.
2222 PH/61 Add -l and -k options to exicyclog.
2224 PH/62 When verifying, if an address was redirected to one new address, so that
2225 verification continued, and the new address failed or deferred after
2226 having set something in $address_data, the value of $address_data was not
2227 passed back to the ACL. This was different to the case when no
2228 redirection occurred. The value is now passed back in both cases.
2230 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
2231 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
2232 use login_cap.h, so on its own it isn't the distinguishing feature. The
2233 new name refers directly to the setclassresources() function.
2235 PH/65 Added configuration files for NetBSD3.
2237 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
2239 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
2240 is preferred over IPv4.
2242 PH/68 The bounce_return_message and bounce_return_body options were not being
2243 honoured for bounces generated during the reception of non-SMTP messages.
2244 In particular, this applied to messages rejected by the ACL. This bug has
2245 been fixed. However, if bounce_return_message is true and bounce_return_
2246 body is false, the headers that are returned for a non-SMTP message
2247 include only those that have been read before the error was detected.
2248 (In the case of an ACL rejection, they have all been read.)
2250 PH/69 The HTML version of the specification is now built in a directory called
2251 spec_html instead of spec.html, because the latter looks like a path with
2252 a MIME-type, and this confuses some software.
2254 PH/70 Catch two compiler warnings in sieve.c.
2256 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
2257 function verify_get_ident() calls ip_connect() to connect a socket, but
2258 if the "connect()" function timed out, ip_connect() used to close the
2259 socket. However, verify_get_ident() also closes the socket later, and in
2260 between Exim writes to the log, which may get opened at this point. When
2261 the socket was closed in ip_connect(), the log could get the same file
2262 descriptor number as the socket. This naturally causes chaos. The fix is
2263 not to close the socket in ip_connect(); the socket should be closed by
2264 the function that creates it. There was only one place in the code where
2265 this was missing, in the iplookup router, which I don't think anybody now
2266 uses, but I've fixed it anyway.
2268 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
2269 well as to direct DNS lookups. Otherwise the handling of names in host
2270 lists is inconsistent and therefore confusing.
2276 PH/01 Two changes to the default runtime configuration:
2278 (1) Move the checks for relay_from_hosts and authenticated clients from
2279 after to before the (commented out) DNS black list checks.
2281 (2) Add control=submission to the relay_from_hosts and authenticated
2282 clients checks, on the grounds that messages accepted by these
2283 statements are most likely to be submissions.
2285 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
2287 (1) Generate an error if the third argument for the ${prvs expansion is
2290 (2) Treat a missing third argument of ${prvscheck as if it were an empty
2293 (3) Reset the variables that are obtained from the first argument of
2294 ${prvscheck and used in the second argument before leaving the code,
2295 because their memory is reclaimed, so using them afterwards may do
2298 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
2299 one (it's much easier than Tom thought :-).
2301 (5) Because of (4), we can now allow for the use of $prvscheck_result
2302 inside the third argument.
2304 PH/03 For some reason, the default setting of PATH when running a command from
2305 a pipe transport was just "/usr/bin". I have changed it to
2308 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
2309 anything to be listed in the output from -bV.
2311 PH/05 When a filter generated an autoreply, the entire To: header line was
2312 quoted in the delivery log line, like this:
2314 => >A.N.Other <ano@some.domain> <original@ddress> ...
2316 This has been changed so that it extracts the operative address. There
2317 may be more than one such address. If so, they are comma-separated, like
2320 => >ano@some.domain,ona@other.domain <original@ddress> ...
2322 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
2323 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
2324 not being looked up in the rDNS to get a host name, Exim was showing the
2325 IP address twice in Received: lines, even though the IP addresses were
2326 identical. For example:
2328 Received: from [1.2.3.4] (helo=[1.2.3.4])
2330 However, if the real host name was known, it was omitting the HELO data
2331 if it matched the actual IP address. This has been tidied up so that it
2332 doesn't show the same IP address twice.
2334 PH/07 When both +timestamp and +memory debugging was on, the value given by
2335 $tod_xxx expansions could be wrong, because the tod_stamp() function was
2336 called by the debug printing, thereby overwriting the timestamp buffer.
2337 Debugging no longer uses the tod_stamp() function when +timestamp is set.
2339 PH/08 When the original message was included in an autoreply transport, it
2340 always said "this is a copy of the message, including all the headers",
2341 even if body_only or headers_only was set. It now gives an appropriate
2344 PH/09 Applied a patch from the Sieve maintainer which:
2346 o fixes some comments
2347 o adds the (disabled) notify extension core
2348 o adds some debug output for the result of if/elsif tests
2349 o points to the current vacation draft in the documentation
2350 and documents the missing references header update
2354 o fixes a bug in processing the envelope test (when testing
2355 multiple envelope elements, the last element determinted the
2358 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
2359 Electronic Mail") by including:
2361 Auto-submitted: auto-generated
2363 in the messages that it generates (bounce messages and others, such as
2364 warnings). In the case of bounce messages for non-SMTP mesages, there was
2365 also a typo: it was using "Auto_submitted" (underscore instead of
2366 hyphen). Since every message generated by Exim is necessarily in response
2367 to another message, thes have all been changed to:
2369 Auto-Submitted: auto-replied
2371 in accordance with these statements in the RFC:
2373 The auto-replied keyword:
2375 - SHOULD be used on messages sent in direct response to another
2376 message by an automatic process,
2378 - MUST NOT be used on manually-generated messages,
2380 - MAY be used on Delivery Status Notifications (DSNs) and Message
2381 Disposition Notifications (MDNs),
2383 - MUST NOT be used on messages generated by automatic or periodic
2384 processes, except for messages which are automatic responses to
2387 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
2388 to the default Received: header definition.
2390 PH/12 Added log selector acl_warn_skipped (default on).
2392 PH/13 After a successful wildlsearch lookup, discard the values of numeric
2393 variables because (a) they are in the wrong storage pool and (b) even if
2394 they were copied, it wouldn't work properly because of the caching.
2396 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
2397 checking when decoding. Apparently there are clients that generate
2398 overlong encoded strings. Why am I not surprised?
2400 PH/15 If the first argument of "${if match_address" was not empty, but did not
2401 contain an "@" character, Exim crashed. Now it writes a panic log message
2402 and treats the condition as false.
2404 PH/16 In autoreply, treat an empty string for "once" the same as unset.
2406 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
2407 extension "envelope-auth". The code is finished and in agreement with
2408 other implementations, but there is no documentation so far and in fact,
2409 nobody wrote the draft yet. This extension is currently #undef'ed, thus
2410 not changing the active code.
2412 Print executed "if" and "elsif" statements when debugging is used. This
2413 helps a great deal to understand what a filter does.
2415 Document more things not specified clearly in RFC3028. I had all this
2416 sorted out, when out of a sudden new issues came to my mind. Oops."
2418 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
2421 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
2422 canonical form (as documented). However, after a host name lookup from
2423 the IP address, check_host() was doing a simple string comparison with
2424 addresses acquired from the DNS when checking that the found name did
2425 have the original IP as one of its addresses. Since any found IPv6
2426 addresses are likely to be in abbreviated form, the comparison could
2427 fail. Luckily, there already exists a function for doing the comparison
2428 by converting both addresses to binary, so now that is used instead of
2429 the text comparison.
2431 PH/20 There was another similar case to PH/19, when a complete host name was
2432 given in a host list; looking up its IP address could give an abbreviated
2433 form, whereas the current host's name might or might not be abbreviated.
2434 The same fix has been applied.
2440 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
2441 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
2444 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
2445 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
2447 PH/03 Typo: missing ".o" in src/pcre/Makefile.
2449 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
2450 header line, restrict the check to what is listed in RFCs 2369 and 2929.
2451 Also, for "Auto-Submitted", treat anything other than "no" as
2452 non-personal, in accordance with RFC 3834. (Previously it treated
2453 anything starting "auto-" as non-personal.)
2455 TF/01 The control=submission/name=... option had a problem with syntax
2456 errors if the name included a slash character. The /name= option
2457 now slurps the rest of the string, so it can include any characters
2458 but it must come last in the list of options (after /sender_retain
2461 PH/05 Some modifications to the interface to the fake nameserver for the new
2469 TK/01 Added the "success_on_redirect" address verification option. See
2470 NewStuff for rationale and an example.
2472 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
2474 PH/02 Patch to exigrep to allow it to work on syslog lines.
2476 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
2477 fread() to skip over the body file's header line, because in Cygwin the
2478 header line is locked and is inaccessible.
2480 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
2481 co-exist for some time) to make it clear that it is the Exim ID that is
2482 referenced, not the Message-ID: header line.
2484 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
2485 string_format() function, because snprintf() does not exist on all
2488 PH/06 The use of forbid_filter_existstest now also locks out the use of the
2489 ${stat: expansion item.
2491 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
2492 protocol synchronization error", to keep the pedants happy.
2494 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
2495 well as for IRIX systems, when gcc is being used. See the host.c source
2498 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
2500 PH/10 Named domain lists were not working if used in a queue_smtp_domains
2503 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
2504 transport and to the smtp transport in LMTP mode.
2506 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
2508 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
2509 run a filter in a subprocess. This could lead to confusion in subsequent
2510 lookups in the parent process. There should also be a search_tidyup() at
2511 the end of the subprocess.
2513 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
2514 only if the host matched helo_try_verify_hosts, which caused the
2515 verification to occur when the EHLO/HELO command was issued. The ACL just
2516 tested the remembered result. Now, if a previous verification attempt has
2517 not happened, "verify = helo" does it there and then.
2519 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
2521 TK/03 Fix log output including CR from clamd.
2523 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
2524 exist provoked a memory error which could cause a segfault.
2526 PH/15 Installed PCRE 6.2
2528 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
2530 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
2531 of the problem. Specifically, suggested +O2 rather than +O1 for the
2534 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
2536 PH/20 If a delivery was routed to a non-standard port by means of an SRV
2537 record, the port was not correctly logged when the outgoing_port log
2538 selector was set (it logged the transort's default port).
2540 PH/21 Added support for host-specific ports to manualroute, queryprogram,
2541 fallback_hosts, and "hosts" in the smtp transport.
2543 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
2544 host errors such as "Connection refused".
2546 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
2547 authentication with radiusclient 0.4.9:
2549 - Error returned from rc_read_config was caught wrongly
2550 - Username/password not passed on to radius server due to wrong length.
2552 The presumption is that some radiusclient API changes for 4.51/PH/17
2553 were not taken care of correctly. The code is still untested by me (my
2554 Linux distribution still has 0.3.2 of radiusclient), but it was
2555 contributed by a Radius user.
2557 PH/24 When doing a callout, the value of $domain wasn't set correctly when
2558 expanding the "port" option of the smtp transport.
2560 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
2561 while reading a MIME header. Thanks to Tom Hughes for a patch.
2563 PH/24 Include config.h inside local_scan.h so that configuration settings are
2566 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
2567 that in an ACL for RCPT (for example), you can examine exactly what was
2570 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
2571 commands, but it was not correctly comparing the address with the actual
2572 client host address. Thus, it would show the EHLO address in Received:
2573 header lines when this was not necessary.
2575 PH/27 Added the % operator to ${eval:}.
2577 PH/28 Exim tries to create and chdir to its spool directory when it starts;
2578 it should be ignoring failures (because with -C, for example, it has lost
2579 privilege). It wasn't ignoring creation failures other than "already
2582 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
2585 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
2586 because an input line was too long, either on its own, or by virtue of
2587 too many continuations, the temporary file was not being removed, and the
2588 return code was incorrect.
2590 PH/31 Missing "BOOL" in function definition in filtertest.c.
2592 PH/32 Applied Sieve patches from the maintainer.
2594 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
2596 PH/33 Added "verify = not_blind".
2598 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
2599 Local/Makefile (with some defaults set). These are used in built scripts
2600 such as exicyclog, but they have never been used in the exim_install
2601 script (though there are many overriding facilities there). I have
2602 arranged that the exim_install script now takes note of these two
2605 PH/35 Installed configuration files for Dragonfly.
2607 PH/36 When a locally submitted message by a trusted user did not contain a
2608 From: header, and the sender address was obtained from -f or from an SMTP
2609 MAIL command, and the trusted user did not use -F to supply a sender
2610 name, $originator_name was incorrectly used when constructing a From:
2611 header. Furthermore, $originator_name was used for submission mode
2612 messages from external hosts without From: headers in a similar way,
2613 which is clearly wrong.
2615 PH/37 Added control=suppress_local_fixups.
2617 PH/38 When log_selector = +received_sender was set, and the addition of the
2618 sender made the log line's construction buffer exactly full, or one byte
2619 less than full, an overflow happened when the terminating "\n" was
2622 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
2623 when the result of a list match is failure because a DNS lookup failed.
2625 PH/40 RM_COMMAND is now used in the building process.
2627 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
2628 the "build-* directories that it finds.
2630 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
2631 address in a domain literal was a prefix of an interface address.
2633 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
2634 when verifying a sender address, unless rewrite_headers is false.
2636 PH/44 Wrote a long comment about why errors_to addresses are verified as
2637 recipients, not senders.
2639 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
2640 the ratelimit ACL was added.
2642 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
2644 PH/46 Added extra information about PostgreSQL errors to the error string.
2646 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
2647 avoiding the need to install special zones in a real server. This is
2648 backwards compatible; if it can't find the fake resolver, it drops back.
2649 Thus, both old and new test suites can be run.
2651 TF/02 Added util/ratelimit.pl
2653 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
2654 clock is set back in time.
2656 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
2657 Candler <B.Candler@pobox.com>.
2659 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
2660 OK for addresses that are the result of redirections.
2662 PH/48 A number of further additions for the benefit of the new test suite,
2663 including a fake gethostbyname() that interfaces to the fake DNS resolver
2666 TF/06 The fix for widen_domains has also been applied to qualify_single and
2667 search_parents which are the other dnslookup options that can cause
2670 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
2673 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
2675 TF/07 Exim produced the error message "an SRV record indicated no SMTP
2676 service" if it encountered an MX record with an empty target hostname.
2677 The message is now "an MX or SRV record indicated no SMTP service".
2679 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
2680 if the DNS of the sending site is misconfigured. This is quite a
2681 common situation. This change restores the behaviour of treating a
2682 helo verification defer as a failure.
2684 PH/51 If self=fail was set on a router, the bounce message did not include the
2685 actual error message.
2691 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
2693 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
2694 command itself ended in error, the underlying message about the transport
2695 filter timeout was being overwritten with the pipe command error. Now the
2696 underlying error message should be appended to the second error message.
2698 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
2700 PH/02 Reduce the amount of output that "make" produces by default. Full output
2701 can still be requested.
2703 PH/03 The warning log line about a condition test deferring for a "warn" verb
2704 was being output only once per connection, rather than after each
2705 occurrence (because it was using the same function as for successful
2706 "warn" verbs). This seems wrong, so I have changed it.
2708 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
2709 it should not have, which might have caused a crash in the right
2710 circumstances, but probably never did.
2712 PH/04 Installed a modified version of Tony Finch's patch to make submission
2713 mode fix the return path as well as the Sender: header line, and to
2714 add a /name= option so that you can make the user's friendly name appear
2717 TF/03 Added the control = fakedefer ACL modifier.
2719 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
2720 Mark Lowes for thorough testing.
2722 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
2724 TK/03 Merged latest SRS patch from Miles Wilton.
2726 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
2727 with the definition in sysexits.h (which is #included earlier).
2728 Fortunately, Exim does not actually use EX_OK. The code used to try to
2729 preserve the sysexits.h value, by assumimg that macro definitions were
2730 scanned for macro replacements. I have been disabused of this notion,
2731 so now the code just undefines EX_OK before #including unistd.h.
2733 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
2734 in the smtp transport. When a block could not be written in a single
2735 write() function, the timeout was being re-applied to each part-write.
2736 This seems wrong - if the receiver was accepting one byte at a time it
2737 would take for ever. The timeout is now adjusted when this happens. It
2738 doesn't have to be particularly precise.
2740 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
2741 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
2743 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
2744 without a domain if the check to <postmaster@domain> fails.
2746 SC/01 Eximstats: added -xls and the ability to specify output files
2747 (patch written by Frank Heydlauf).
2749 SC/02 Eximstats: use FileHandles for outputing results.
2751 SC/03 Eximstats: allow any combination of xls, txt, and html output.
2753 SC/04 Eximstats: fixed display of large numbers with -nvr option
2755 SC/05 Eximstats: fixed merging of reports with empty tables.
2757 SC/06 Eximstats: added the -include_original_destination flag
2759 SC/07 Eximstats: removed tabs and trailing whitespace.
2761 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
2763 TK/06 MBOX spool code: Add real "From " MBOX separator line
2764 so the .eml file is really in mbox format (even though
2765 most programs do not really care). Patch from Alex Miller.
2767 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
2768 The latter is generated from $received_to and is only set if the
2769 message has one envelope recipient. SA can use these headers,
2770 obviously out-of-the-box. Patch from Alex Miller.
2772 PH/08 The ${def test on a variable was returning false if the variable's
2773 value was "0", contrary to what the specification has always said!
2774 The result should be true unless the variable is empty.
2776 PH/09 The syntax error of a character other than { following "${if
2777 def:variable_name" (after optional whitespace) was not being diagnosed.
2778 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
2779 accidental colon was present, for example, could give incorrect results.
2781 PH/10 Tidied the code in a number of places where the st_size field of a stat()
2782 result is used (not including appendfile, where other changes are about
2785 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
2786 This involved changing a lot of size variables from int to off_t. It
2787 should work with maildirs and everything.
2789 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
2790 spamd dying while we are connected to it.
2792 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
2795 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
2796 function for BDB 4.3.
2798 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
2800 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
2801 the sources for PCRE have been re-arranged and more files are now
2804 PH/15 The code I had for printing potentially long long variables in PH/11
2805 above was not the best (it lost precision). The length of off_t variables
2806 is now inspected at build time, and an appropriate printing format (%ld
2807 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
2808 to be "long long int" or "long int". This is needed for the internal
2809 formatting function string_vformat().
2811 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
2812 the configuration file to be ":syslog", then the script "guesses" where
2813 the logs files are, rather than using the compiled in default. In our
2814 case the guess is not the same as the compiled default, so the script
2815 suddenly stopped working when I started to use syslog. The patch checks
2816 to see if log_file_path is "". If so, it attempts to read it from exim
2817 with no configuration file to get the compiled in version, before it
2818 falls back to the previous guessing code."
2820 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
2821 implementing BATV in an Exim configuration. See NewStuff for the gory
2824 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
2825 Makefile that are specific to HP-UX.
2827 PH/18 If the "use_postmaster" option was set for a recipient callout together
2828 with the "random" option, the postmaster address was used as the MAIL
2829 FROM address for the random test, but not for the subsequent recipient
2830 test. It is now used for both.
2832 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
2833 patch removes a few documentation additions to RFC 3028, because the
2834 latest draft now contains them. It adds the new en;ascii-case comparator
2835 and a new error check for 8bit text in MIME parts. Comparator and
2836 require names are now matched exactly. I enabled the subaddress
2837 extension, but it is not well tested yet (read: it works for me)."
2839 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
2840 rework some of the code of TK/09 above to avoid the hardwired use of
2841 "%lld" and "long long". Replaced the call to snprintf() with a call to
2844 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
2845 records point to non-existent hosts", "retry timeout exceeded", and
2846 "retry time not reached for any host after a long failure period".
2848 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
2849 experimental DomainKeys support:
2851 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
2852 (2) On an error such as an illegally used "control", the wrong name for
2853 the control was given.
2855 These problems did NOT occur unless DomainKeys support was compiled.
2857 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
2859 PH/24 Added ${if match_ip condition.
2861 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
2862 databases so that it will be absolutely obvious if a crash occurs in the
2863 DB library. This is a regular occurrence (often caused by mis-matched
2866 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
2867 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
2868 picked up on a user's system that detects such things. There doesn't seem
2869 to be a gcc warning option for this - only an attribute that has to be
2870 put on the function's prototype. It seems that in Fedora Core 4 they have
2871 set this on a number of new functions. No doubt there will be more in due
2874 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
2875 specify a transport. However, if an address that was verified by such a
2876 router was the subject of a callout, Exim crashed because it tried to
2877 read the rcpt_include_affixes from the non-existent transport. Now it
2878 just assumes that the setting of that option is false. This bug was
2879 introduced by 4.51/PH/31.
2881 PH/28 Changed -d+all to exclude +memory, because that information is very
2882 rarely of interest, but it makes the output a lot bigger. People tend to
2883 do -d+all out of habit.
2885 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
2886 code in os-type was giving problems when libc.so lives in lib64, like on
2889 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
2890 aren't the modern standard, and it seems that some systems' include files
2891 don't always have them. Exim was already checking for some of the newer
2892 ones like T_AAAA, and defining it itself. I've added checks for all the
2893 record types that Exim uses.
2895 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
2896 not automatically generating a new one, as it is supposed to. This
2897 prevented TLS from working. If the file did exist, but contained invalid
2898 data, a new version was generated, as expected. It was only the case of a
2899 non-existent file that was broken.
2901 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
2902 with a change in libdomainkeys > 0.64.
2904 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
2905 from DNS. If the selector record carries the flag, it now has
2906 precedence over the domain-wide flag.
2908 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
2910 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
2911 the use of a port name that isn't defined in /etc/services) occurred, the
2912 message was deferred as in a normal delivery, and thus remained on the
2913 spool, instead of being failed because of the mua_wrapper setting. This
2914 is now fixed, and I tidied up some of the mua_wrapper messages at the
2917 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
2918 the messages in a hash of arrays rather than using individual hashes.
2919 This is a bit cleaner and results in dramatic memory savings, albeit
2920 at a slight CPU cost.
2922 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
2923 as requested by Marc Sherman.
2925 SC/10 Eximstats: added histograms for user specified patterns as requested
2928 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
2930 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
2931 fopen() in the content-scanning modules that did not already have it.
2937 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
2938 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
2940 TK/02 Fix ACL "control" statement not being available in MIME ACL.
2942 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
2944 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
2945 to test Sieve filters that use "vacation".
2947 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
2948 that changes the way the GnuTLS parameters are stored in the cache file.
2949 The new format can be generated externally. For backward compatibility,
2950 if the data in the cache doesn't make sense, Exim assumes it has read an
2951 old-format file, and it generates new data and writes a new file. This
2952 means that you can't go back to an older release without removing the
2955 PH/03 A redirect router that has both "unseen" and "one_time" set does not
2956 work if there are any delivery delays because "one_time" forces the
2957 parent to be marked "delivered", so its unseen clone is never tried
2958 again. For this reason, Exim now forbids the simultaneous setting of
2961 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
2962 redirected to themselves ("homonym" addresses). Read the long ChangeLog
2963 entry if you want to know the details. The fix, however, neglected to
2964 consider the case when local delivery batching is involved. The test for
2965 "previously delivered" was not happening when checking to see if an
2966 address could be batched with a previous (undelivered) one; under
2967 certain circumstances this could lead to multiple deliveries to the same
2970 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
2971 in its include files, and this causes problems building Exim.
2973 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
2974 header_syntax) but Exim was just ignoring anything given after a slash.
2975 In particular, this caused confusion with an attempt to use "verify =
2976 reverse_host_lookup/defer_ok". An error is now given when options are
2977 supplied for verify items that do not have them. (Maybe reverse_host_
2978 lookup should have a defer_ok option, but that's a different point.)
2980 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
2981 defined by RFC 821) to 2048, because there were problems with some AUTH
2982 commands, and RFC 1869 says the size should be increased for extended
2983 SMTP commands that take arguments.
2985 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
2988 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
2989 "unknown" error; now it says that the functionality isn't in the binary.
2991 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
2992 an address' error message when a string expansion fails (syntax or
2993 whatever). Otherwise the password may appear in the log. Following change
2994 PH/42 below, there is no longer a chance of it appearing in a bounce
2997 PH/11 Installed exipick version 20050225.0 from John Jetmore.
2999 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
3000 of its addresses was ever tried. (Bugzilla bug #2.)
3002 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
3003 the result incorrectly in the debug output. (It correctly added a newline
3004 to what was transported.)
3006 TF/01 Added $received_time.
3008 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
3009 commented out examples of how to interface to a virus scanner and to
3010 SpamAssassin. Also added commented examples of av_scanner and
3011 spamd_address settings.
3013 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
3014 and controls are allowed in which ACLs. There were a couple of minor
3015 errors. Some of the entries in the conditions table (which is a table of
3016 where they are NOT allowed) were getting very unwieldy; rewrote them as a
3017 negation of where the condition IS allowed.
3019 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
3021 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
3022 header file does not have a version number, so I've had to invent a new
3023 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
3024 API. The code is untested by me (my Linux distribution still has 0.3.2 of
3025 radiusclient), but it was contributed by a Radius user.
3027 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
3028 files or directories, for OpenSSL.
3030 PH/19 When an Exim process that is running as root has to create an Exim log
3031 file, it does so in a subprocess that runs as exim:exim so as to get the
3032 ownership right at creation (otherwise, other Exim processes might see
3033 the file with the wrong ownership). There was no test for failure of this
3034 fork() call, which would lead to the process getting stuck as it waited
3035 for a non-existent subprocess. Forks do occasionally fail when resources
3036 run out. I reviewed all the other calls to fork(); they all seem to check
3039 PH/20 When checking for unexpected SMTP input at connect time (before writing
3040 the banner), Exim was not dealing correctly with a non-positive return
3041 from the read() function. If the client had disconnected by this time,
3042 the result was a log entry for a synchronization error with an empty
3043 string after "input=" when read() returned zero. If read() returned -1
3044 (an event I could not check), uninitialized data bytes were printed.
3045 There were reports of junk text (parts of files, etc) appearing after
3048 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
3050 PH/22 Added support for macro redefinition, and (re)definition in between
3051 driver and ACL definitions.
3053 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
3054 forgetting to use the resulting value; it was using the unexpanded value.
3056 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
3057 hadn't been configured. The fix is from Juergen Kreileder, who
3058 understands it better than I do:
3060 "Here's what I see happening with three configured cyrus_sasl
3061 authenticators configured (plain, login, cram-md5):
3063 On startup auth_cyrus_sasl_init() gets called for each of these.
3064 This means three calls to sasl_listmech() without a specified mech_list.
3065 => SASL tests which mechs of all available mechs actually work
3066 => three warnings about OTP not working
3067 => the returned list contains: plain, login, cram-md5, digest-md5, ...
3069 With the patch, sasl_listmech() also gets called three times. But now
3070 SASL's mech_list option is set to the server_mech specified in the the
3071 authenticator. Or in other words, the answer from sasl_listmech()
3072 gets limited to just the mech you're testing for (which is different
3074 => the return list contains just 'plain' or 'login', 'cram-md5' or
3075 nothing depending on the value of ob->server_mech.
3077 I've just tested the patch: Authentication still works fine,
3078 unavailable mechs specified in the exim configuration are still
3079 caught, and the auth.log warnings about OTP are gone."
3081 PH/25 When debugging is enabled, the contents of the command line are added
3082 to the debugging output, even when log_selector=+arguments is not
3085 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
3086 answer is "GNU", and only if the return is "GNU/something" is the answer
3089 PH/27 $acl_verify_message is now set immediately after the failure of a
3090 verification in an ACL, and so is available in subsequent modifiers. In
3091 particular, the message can be preserved by coding like this:
3093 warn !verify = sender
3094 set acl_m0 = $acl_verify_message
3096 Previously, $acl_verify_message was set only while expanding "message"
3097 and "log_message" when a very denied access.
3099 PH/28 Modified OS/os.c-Linux with
3101 -#ifndef OS_LOAD_AVERAGE
3102 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
3104 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
3105 nomenclature these days.)
3107 PH/29 Installed patch from the Sieve maintainer that adds the options
3108 sieve_useraddress and sieve_subaddress to the redirect router.
3110 PH/30 In these circumstances:
3111 . Two addresses routed to the same list of hosts;
3112 . First host does not offer TLS;
3113 . First host accepts first address;
3114 . First host gives temporary error to second address;
3115 . Second host offers TLS and a TLS session is established;
3116 . Second host accepts second address.
3117 Exim incorrectly logged both deliveries with the TLS parameters (cipher
3118 and peerdn, if requested) that were in fact used only for the second
3121 PH/31 When doing a callout as part of verifying an address, Exim was not paying
3122 attention to any local part prefix or suffix that was matched by the
3123 router that accepted the address. It now behaves in the same way as it
3124 does for delivery: the affixes are removed from the local part unless
3125 rcpt_include_affixes is set on the transport.
3127 PH/32 Add the sender address, as F=<...>, to the log line when logging a
3128 timeout during the DATA phase of an incoming message.
3130 PH/33 Sieve envelope tests were broken for match types other than :is. I have
3131 applied a patch sanctioned by the Sieve maintainer.
3133 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
3134 the uid or gid is negative. A case of a negative gid caused this to be
3135 noticed. The fix allows for either to be negative.
3137 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
3138 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
3141 PH/36 The change PH/12 above was broken. Fixed it.
3143 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
3144 the grounds that routing the same address twice would always produce the
3145 same answer. This might have been true once, but it is certainly no
3146 longer true now. Routing a child address may depend on the previous
3147 routing that produced that child. Some complicated redirection strategies
3148 went wrong when messages had multiple recipients, and made Exim's
3149 behaviour dependent on the order in which the addresses were given.
3151 I have moved the duplicate checking until after the routing is complete.
3152 Exim scans the addresses that are assigned to local and remote
3153 transports, and removes any duplicates. This means that more work will be
3154 done, as duplicates will always all be routed, but duplicates are
3155 presumably rare, so I don't expect this is of any significance.
3157 For deliveries to pipes, files, and autoreplies, the duplicate checking
3158 still happens during the routing process, since they are not going to be
3161 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
3162 It corrects a timeout issue with spamd. This is Ian's comment: "The
3163 background is that sometimes spamd either never reads data from a
3164 connection it has accepted, or it never writes response data. The exiscan
3165 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
3166 blindly assumes that writes won't block so it may never time out."
3168 PH/39 Allow G after quota size as well as K and M.
3170 PH/40 The value set for $authenticated_id in an authenticator may not contain
3171 binary zeroes or newlines because the value is written to log lines and
3172 to spool files. There was no check on this. Now the value is run through
3173 the string_printing() function so that such characters are converted to
3174 printable escape sequences.
3176 PH/41 $message_linecount is a new variable that contains the total number of
3177 lines in the message. Compare $body_linecount, which is the count for the
3180 PH/42 Exim no longer gives details of delivery errors for specific addresses in
3181 bounce and delay warning messages, except in certain special cases, which
3184 (a) An SMTP error message from a remote host;
3185 (b) A message specified in a :fail: redirection;
3186 (c) A message specified in a "fail" command in a system filter;
3187 (d) A message specified in a FAIL return from the queryprogram router;
3188 (e) A message specified by the cannot_route_message router option.
3190 In these cases only, Exim does include the error details in bounce and
3191 warning messages. There are also a few cases where bland messages such
3192 as "unrouteable address" or "local delivery error" are given.
3194 PH/43 $value is now also set for the "else" part of a ${run expansion.
3196 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
3197 being worked on, but at least Exim now implements the latest version to
3200 PH/45 In a pipe transport, although a timeout while waiting for the pipe
3201 process to complete was treated as a delivery failure, a timeout while
3202 writing the message to the pipe was logged, but erroneously treated as a
3203 successful delivery. Such timeouts include transport filter timeouts. For
3204 consistency with the overall process timeout, these timeouts are now
3205 treated as errors, giving rise to delivery failures by default. However,
3206 there is now a new Boolean option for the pipe transport called
3207 timeout_defer, which, if set TRUE, converts the failures into defers for
3208 both kinds of timeout. A transport filter timeout is now identified in
3211 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
3212 systems where "make" and "gmake" are different, calling "gmake" at top
3213 level broke things. I've arranged for the value of $(MAKE) to be passed
3214 from the Makefile to this script so that it can call the same version of
3218 A note about Exim versions 4.44 and 4.50
3219 ----------------------------------------
3221 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
3222 changes of various kinds. As a consequence, a big documentation update was
3223 needed. This delayed the release for rather longer than seemed good, especially
3224 in the light of a couple of (minor) security issues. Therefore, the changes
3225 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
3226 release. So 4.44 and 4.50 are in effect two different branches that both start
3229 I have left the 4.50 change log unchanged; it contains all the changes since
3230 4.43. The change log for 4.44 is below; many of its items are identical to
3231 those for 4.50. This seems to be the most sensible way to preserve the
3232 historical information.
3238 1. Minor wording change to the doc/README.SIEVE file.
3240 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
3241 computation of the current number of files was incorrect.
3243 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3244 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3247 4. Give more explanation in the error message when the command for a transport
3248 filter fails to execute.
3250 5. There are several places where Exim runs a non-Exim command in a
3251 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3252 was being done only for the command run by the queryprogram router. It is
3253 now done for all such subprocesses. The other cases are: ${run, transport
3254 filters, and the commands run by the lmtp and pipe transports.
3256 6. Added CONFIGURE_GROUP build-time option.
3258 7. Some older OS have a limit of 256 on the maximum number of file
3259 descriptors. Exim was using setrlimit() to set 1000 as a large value
3260 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3261 systems. I've change it so that if it can't get 1000, it tries for 256.
3263 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3264 was an oversight, and furthermore, ever since the addition of extra
3265 controls (e.g. 4.43/32), the checks on when to allow different forms of
3266 "control" were broken. There should now be diagnostics for all cases when a
3267 control that does not make sense is encountered.
3269 9. Added the /retain_sender option to "control=submission".
3271 10. $recipients is now available in the predata ACL (oversight).
3273 11. Tidy the search cache before the fork to do a delivery from a message
3274 received from the command line. Otherwise the child will trigger a lookup
3275 failure and thereby defer the delivery if it tries to use (for example) a
3276 cached ldap connection that the parent has called unbind on.
3278 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3279 of $address_data from the recipient verification was clobbered by the
3280 sender verification.
3282 13. The value of address_data from a sender verification is now available in
3283 $sender_address_data in subsequent conditions in the ACL statement.
3285 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
3287 15. Added a new option "connect=<time>" to callout options, to set a different
3290 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3291 was its contents. (It was OK if the option was not defined at all.)
3293 17. A "Completed" log line is now written for messages that are removed from
3294 the spool by the -Mrm option.
3296 18. New variables $sender_verify_failure and $recipient_verify_failure contain
3297 information about exactly what failed.
3299 19. Added -dd to debug only the daemon process.
3301 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
3302 handles timeouts, both on the server side and network timeouts. Renamed the
3303 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
3305 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
3306 It is now set to "smtps".
3308 22. $host_address is now set to the target address during the checking of
3309 ignore_target_hosts.
3311 23. When checking ignore_target_hosts for an ipliteral router, no host name was
3312 being passed; this would have caused $sender_host_name to have been used if
3313 matching the list had actually called for a host name (not very likely,
3314 since this list is usually IP addresses). A host name is now passed as
3317 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3318 code that specifies a non-restarting handler (typically sigaction() in
3319 modern systems) in an attempt to fix a rare and obscure crash bug.
3321 25. Narrowed the window for a race in the daemon that could cause it to ignore
3322 SIGCHLD signals. This is not a major problem, because they are used only to
3323 wake it up if nothing else does.
3325 26. A malformed maildirsize file could cause Exim to calculate negative values
3326 for the mailbox size or file count. Odd effects could occur as a result.
3327 The maildirsize information is now recalculated if the size or filecount
3330 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3331 support for a long time. Removed HAVE_SYS_VFS_H.
3333 28. Installed the latest version of exipick from John Jetmore.
3335 29. In an address list, if the pattern was not a regular expression, an empty
3336 subject address (from a bounce message) matched only if the pattern was an
3337 empty string. Non-empty patterns were not even tested. This was the wrong
3338 because it is perfectly reasonable to use an empty address as part of a
3339 database query. An empty address is now tested by patterns that are
3340 lookups. However, all the other forms of pattern expect the subject to
3341 contain a local part and a domain, and therefore, for them, an empty
3342 address still always fails if the pattern is not itself empty.
3344 30. Exim went into a mad DNS loop when attempting to do a callout where the
3345 host was specified on an smtp transport, and looking it up yielded more
3346 than one IP address.
3348 31. Re-factored the code for checking spool and log partition space into a
3349 function that finds that data and another that does the check. The former
3350 is then used to implement four new variables: $spool_space, $log_space,
3351 $spool_inodes, and $log_inodes.
3353 32. The RFC2047 encoding function was originally intended for short strings
3354 such as real names; it was not keeping to the 75-character limit for
3355 encoded words that the RFC imposes. It now respects the limit, and
3356 generates multiple encoded words if necessary. To be on the safe side, I
3357 have increased the buffer size for the ${rfc2047: expansion operator from
3360 33. It is now permitted to omit both strings after an "if" condition; if the
3361 condition is true, the result is "true". As before, when the second string
3362 is omitted, a false condition yields an empty string. This makes it less
3363 cumbersome to write custom ACL and router conditions.
3365 34. Failure to deliver a bounce message always caused it to be frozen, even if
3366 there was an errors_to setting on the router. The errors_to setting is now
3369 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3370 canonical form (fully expanded) before being placed in
3371 $sender_host_address.
3373 36. The table in the code that translates DNS record types into text (T_A to
3374 "A" for instance) was missing entries for NS and CNAME. It is just possible
3375 that this could have caused confusion if both these types were looked up
3376 for the same domain, because the text type is used as part of Exim's
3377 per-process caching. But the chance of anyone hitting this buglet seems
3380 37. The dnsdb lookup has been extended in a number of ways.
3382 (1) There is a new type, "zns", which walks up the domain tree until it
3383 finds some nameserver records. It should be used with care.
3385 (2) There is a new type, "mxh", which is like "mx" except that it returns
3386 just the host names, not the priorities.
3388 (3) It is now possible to give a list of domains (or IP addresses) to be
3389 looked up. The behaviour when one of the lookups defers can be
3390 controlled by a keyword.
3392 (4) It is now possible to specify the separator character for use when
3393 multiple records are returned.
3395 38. The dnslists ACL condition has been extended: it is now possible to supply
3396 a list of IP addresses and/or domains to be looked up in a particular DNS
3399 39. Added log_selector=+queue_time_overall.
3401 40. When running the queue in the test harness, wait just a tad after forking a
3402 delivery process, to get repeatability of debugging output.
3404 41. Include certificate and key file names in error message when GnuTLS fails
3405 to set them up, because the GnuTLS error message doesn't include the name
3406 of the failing file when there is a problem reading it.
3408 42. Allow both -bf and -bF in the same test run.
3410 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
3412 44. The "Exiscan patch" is now merged into the mainline Exim source.
3414 45. Sometimes the final signoff response after QUIT could fail to get
3415 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3416 < 0 before doing a fflush(). This bug looks as though it goes back to the
3417 introduction of TLS in release 3.20, but "sometimes" must have been rare
3418 because the tests only now provoked it.
3420 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
3421 (this can affect the format of dates).
3423 47. exim_tidydb, when checking for the continued existence of a message for
3424 which it has found a message-specific retry record, was not finding
3425 messages that were in split spool directories. Consequently, it was
3426 deleting retry records that should have stayed in existence.
3428 48. Steve fixed some bugs in eximstats.
3430 49. The SPA authentication driver was not abandoning authentication and moving
3431 on to the next authenticator when an expansion was forced to fail,
3432 contradicting the general specification for all authenticators. Instead it
3433 was generating a temporary error. It now behaves as specified.
3435 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
3436 (the order specifies the preference for clients). The order is now AES256,
3437 AES128, 3DES, ARCFOUR128.
3439 51. Small patch to Sieve code - explicitly set From: when generating an
3442 52. Exim crashed if a remote delivery caused a very long error message to be
3443 recorded - for instance if somebody sent an entire SpamAssassin report back
3444 as a large number of 550 error lines. This bug was coincidentally fixed by
3445 increasing the size of one of Exim's internal buffers (big_buffer) that
3446 happened as part of the Exiscan merge. However, to be on the safe side, I
3447 have made the code more robust (and fixed the comments that describe what
3450 53. Now that there can be additional text after "Completed" in log lines (if
3451 the queue_time_overall log selector is set), a one-byte patch to exigrep
3452 was needed to allow it to recognize "Completed" as not the last thing in
3455 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
3456 patch that reportedly fixes this has been added. I am not expert enough to
3457 create a test for it. This is what the patch creator wrote:
3459 "I found a little strange behaviour of ldap code when working with
3460 Windows 2003 AD Domain, where users was placed in more than one
3461 Organization Units. When I tried to give exim partial DN, the exit code
3462 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
3463 But simultaneously result of request was absolutely normal ldap result,
3464 so I produce this patch..."
3466 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
3467 REFERENCE, so I have modified the code to exclude the patch when that macro
3470 55. Some experimental protocols are using DNS PTR records for new purposes. The
3471 keys for these records are domain names, not reversed IP addresses. The
3472 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3473 leaves it alone. Component reversal etc. now happens only for IP addresses.
3476 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
3478 57. Double the size of the debug message buffer (to 2048) so that more of very
3479 long debug lines gets shown.
3481 58. The exicyclog utility now does better if the number of log files to keep
3482 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
3484 59. Two changes related to the smtp_active_hostname option:
3486 (1) $smtp_active_hostname is now available as a variable.
3487 (2) The default for smtp_banner uses $smtp_active_hostname instead
3488 of $primary_hostname.
3490 60. The host_aton() function is supposed to be passed a string that is known
3491 to be a valid IP address. However, in the case of IPv6 addresses, it was
3492 not checking this. This is a hostage to fortune. Exim now panics and dies
3493 if the condition is not met. A case was found where this could be provoked
3494 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3495 components; fortuitously, this particular loophole had already been fixed
3496 by change 4.50/55 above.
3498 If there are any other similar loopholes, the new check in host_aton()
3499 itself should stop them being exploited. The report I received stated that
3500 data on the command line could provoke the exploit when Exim was running as
3501 exim, but did not say which command line option was involved. All I could
3502 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3503 running as the user.
3506 61. There was a buffer overflow vulnerability in the SPA authentication code
3507 (which came originally from the Samba project). I have added a test to the
3508 spa_base64_to_bits() function which I hope fixes it.
3511 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
3512 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
3514 63. The daemon start-up calls getloadavg() while still root for those OS that
3515 need the first call to be done as root, but it missed one case: when
3516 deliver_queue_load_max is set with deliver_drop_privilege. This is
3517 necessary for the benefit of the queue runner, because there is no re-exec
3518 when deliver_drop_privilege is set.
3520 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
3521 This has been fixed.
3523 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3524 was in use, was not putting the data itself into the right store pool;
3525 consequently, it could be overwritten for a subsequent message in the same
3526 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3529 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
3531 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
3532 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
3533 the function were treating the return as a boolean value, which happened to
3534 work because 0=false and not-0=true, but is not correct code.
3536 68. The host_aton() function was not handling scoped IPv6 addresses (those
3537 with, for example, "%eth0" on the end) correctly.
3539 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
3540 negated items (that is, ~something) in unsigned ints. Some compilers
3541 apparently mutter when there is no cast.
3543 70. If an address verification called from an ACL failed, and did not produce a
3544 user-specific message (i.e. there was only a "system" message), nothing was
3545 put in $acl_verify_message. In this situation, it now puts the system
3548 71. Change 4.23/11 added synchronization checking at the start of an SMTP
3549 session; change 4.31/43 added the unwanted input to the log line - except
3550 that it did not do this in the start of session case. It now does.
3552 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
3553 This is wrong and can cause the other end to generate a synchronization
3554 error if it is another Exim or anything else that does the synchronization
3555 check. A QUIT command is no longer sent after a timeout.
3557 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
3558 during host lookups.
3560 74. The defer_ok option of callout verification was not working if it was used
3561 when verifying addresses in header lines, that is, for this case:
3563 verify = header_sender/callout=defer_ok
3565 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
3566 those file descriptors could be used for SMTP connections. If anything
3567 wrote to stderr (the example that came up was "warn" in embedded Perl), it
3568 could be sent to the SMTP client, causing chaos. The daemon now opens
3569 stdin, stdout, and stderr to /dev/null when it puts itself into the
3572 76. Arrange for output from Perl's "warn" command to be written to Exim's main
3573 log by default. The user can override this with suitable Perl magic.
3575 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
3576 the log message when discard triggers, was not working for the DATA ACL or
3577 for the non-SMTP ACL.
3579 78. Error message wording change in sieve.c.
3581 79. If smtp_accept_max_per_host was set, the number of connections could be
3582 restricted to fewer than expected, because the daemon was trying to set up
3583 a new connection before checking whether the processes handling previous
3584 connections had finished. The check for completed processes is now done
3585 earlier. On busy systems, this bug wouldn't be noticed because something
3586 else would have woken the daemon, and it would have reaped the completed
3589 80. If a message was submitted locally by a user whose login name contained one
3590 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
3591 It caused a spool format error. I have fixed the spool reading code. A
3592 related problem was that the "from" clause in the Received: line became
3593 illegal because of the space(s). It is now covered by ${quote_local_part.
3595 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
3598 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
3600 83. Patch from Sieve maintainer to fix unterminated string problem in
3601 "vacation" handling.
3603 84. Some minor changes to the Linux configuration files to help with other
3604 OS variants using glibc.
3606 85. One more patch for Sieve to update vacation handling to latest spec.
3609 ----------------------------------------------------
3610 See the note above about the 4.44 and 4.50 releases.
3611 ----------------------------------------------------
3617 1. Change 4.43/35 introduced a bug that caused file counts to be
3618 incorrectly computed when quota_filecount was set in an appendfile
3621 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
3622 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
3625 3. Give more explanation in the error message when the command for a transport
3626 filter fails to execute.
3628 4. There are several places where Exim runs a non-Exim command in a
3629 subprocess. The SIGUSR1 signal should be disabled for these processes. This
3630 was being done only for the command run by the queryprogram router. It is
3631 now done for all such subprocesses. The other cases are: ${run, transport
3632 filters, and the commands run by the lmtp and pipe transports.
3634 5. Some older OS have a limit of 256 on the maximum number of file
3635 descriptors. Exim was using setrlimit() to set 1000 as a large value
3636 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
3637 systems. I've change it so that if it can't get 1000, it tries for 256.
3639 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
3640 was an oversight, and furthermore, ever since the addition of extra
3641 controls (e.g. 4.43/32), the checks on when to allow different forms of
3642 "control" were broken. There should now be diagnostics for all cases when a
3643 control that does not make sense is encountered.
3645 7. $recipients is now available in the predata ACL (oversight).
3647 8. Tidy the search cache before the fork to do a delivery from a message
3648 received from the command line. Otherwise the child will trigger a lookup
3649 failure and thereby defer the delivery if it tries to use (for example) a
3650 cached ldap connection that the parent has called unbind on.
3652 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
3653 of $address_data from the recipient verification was clobbered by the
3654 sender verification.
3656 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
3657 was its contents. (It was OK if the option was not defined at all.)
3659 11. A "Completed" log line is now written for messages that are removed from
3660 the spool by the -Mrm option.
3662 12. $host_address is now set to the target address during the checking of
3663 ignore_target_hosts.
3665 13. When checking ignore_target_hosts for an ipliteral router, no host name was
3666 being passed; this would have caused $sender_host_name to have been used if
3667 matching the list had actually called for a host name (not very likely,
3668 since this list is usually IP addresses). A host name is now passed as
3671 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
3672 code that specifies a non-restarting handler (typically sigaction() in
3673 modern systems) in an attempt to fix a rare and obscure crash bug.
3675 15. Narrowed the window for a race in the daemon that could cause it to ignore
3676 SIGCHLD signals. This is not a major problem, because they are used only to
3677 wake it up if nothing else does.
3679 16. A malformed maildirsize file could cause Exim to calculate negative values
3680 for the mailbox size or file count. Odd effects could occur as a result.
3681 The maildirsize information is now recalculated if the size or filecount
3684 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
3685 support for a long time. Removed HAVE_SYS_VFS_H.
3687 18. Updated exipick to current release from John Jetmore.
3689 19. Allow an empty sender to be matched against a lookup in an address list.
3690 Previously the only cases considered were a regular expression, or an
3693 20. Exim went into a mad DNS lookup loop when doing a callout where the
3694 host was specified on the transport, if the DNS lookup yielded more than
3697 21. The RFC2047 encoding function was originally intended for short strings
3698 such as real names; it was not keeping to the 75-character limit for
3699 encoded words that the RFC imposes. It now respects the limit, and
3700 generates multiple encoded words if necessary. To be on the safe side, I
3701 have increased the buffer size for the ${rfc2047: expansion operator from
3704 22. Failure to deliver a bounce message always caused it to be frozen, even if
3705 there was an errors_to setting on the router. The errors_to setting is now
3708 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
3709 canonical form (fully expanded) before being placed in
3710 $sender_host_address.
3712 24. Updated eximstats to version 1.33
3714 25. Include certificate and key file names in error message when GnuTLS fails
3715 to set them up, because the GnuTLS error message doesn't include the name
3716 of the failing file when there is a problem reading it.
3718 26. Expand error message when OpenSSL has problems setting up cert/key files.
3721 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
3722 (this can affect the format of dates).
3724 28. exim_tidydb, when checking for the continued existence of a message for
3725 which it has found a message-specific retry record, was not finding
3726 messages that were in split spool directories. Consequently, it was
3727 deleting retry records that should have stayed in existence.
3729 29. eximstats updated to version 1.35
3730 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
3731 1.35 - bugfix such that pie charts by volume are generated correctly
3733 30. The SPA authentication driver was not abandoning authentication and moving
3734 on to the next authenticator when an expansion was forced to fail,
3735 contradicting the general specification for all authenticators. Instead it
3736 was generating a temporary error. It now behaves as specified.
3738 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
3739 (the order specifies the preference for clients). The order is now AES256,
3740 AES128, 3DES, ARCFOUR128.
3742 31. Small patch to Sieve code - explicitly set From: when generating an
3745 32. Exim crashed if a remote delivery caused a very long error message to be
3746 recorded - for instance if somebody sent an entire SpamAssassin report back
3747 as a large number of 550 error lines. This bug was coincidentally fixed by
3748 increasing the size of one of Exim's internal buffers (big_buffer) that
3749 happened as part of the Exiscan merge. However, to be on the safe side, I
3750 have made the code more robust (and fixed the comments that describe what
3753 33. Some experimental protocols are using DNS PTR records for new purposes. The
3754 keys for these records are domain names, not reversed IP addresses. The
3755 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
3756 leaves it alone. Component reversal etc. now happens only for IP addresses.
3759 34. The host_aton() function is supposed to be passed a string that is known
3760 to be a valid IP address. However, in the case of IPv6 addresses, it was
3761 not checking this. This is a hostage to fortune. Exim now panics and dies
3762 if the condition is not met. A case was found where this could be provoked
3763 from a dnsdb PTR lookup with an IPv6 address that had more than 8
3764 components; fortuitously, this particular loophole had already been fixed
3765 by change 4.50/55 or 4.44/33 above.
3767 If there are any other similar loopholes, the new check in host_aton()
3768 itself should stop them being exploited. The report I received stated that
3769 data on the command line could provoke the exploit when Exim was running as
3770 exim, but did not say which command line option was involved. All I could
3771 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
3772 running as the user.
3775 35. There was a buffer overflow vulnerability in the SPA authentication code
3776 (which came originally from the Samba project). I have added a test to the
3777 spa_base64_to_bits() function which I hope fixes it.
3780 36. The daemon start-up calls getloadavg() while still root for those OS that
3781 need the first call to be done as root, but it missed one case: when
3782 deliver_queue_load_max is set with deliver_drop_privilege. This is
3783 necessary for the benefit of the queue runner, because there is no re-exec
3784 when deliver_drop_privilege is set.
3786 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
3787 was in use, was not putting the data itself into the right store pool;
3788 consequently, it could be overwritten for a subsequent message in the same
3789 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
3792 38. Sometimes the final signoff response after QUIT could fail to get
3793 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
3794 < 0 before doing a fflush(). This bug looks as though it goes back to the
3795 introduction of TLS in release 3.20, but "sometimes" must have been rare
3796 because the tests only now provoked it.
3802 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
3803 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
3804 Now it returns an integer. A number of calls were still expecting a T/F
3805 return. Fortuitously, in all cases, the tests worked in OK situations,
3806 which is the norm. However, things would have gone wrong on any write
3807 failures on the smtp file descriptor. This function is used when sending
3808 messages over SMTP and also when doing verify callouts.
3810 2. When Exim is called to do synchronous delivery of a locally submitted
3811 message (the -odf or -odi options), it no longer closes stderr before doing
3814 3. Implemented the mua_wrapper option.
3816 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
3818 5. Implemented the functions header_remove(), header_testname(),
3819 header_add_at_position(), and receive_remove_recipient(), and exported them
3822 6. If an ACL "warn" statement specified the addition of headers, Exim already
3823 inserted X-ACL-Warn: at the start if there was no header name. However, it
3824 was not making this test for the second and subsequent header lines if
3825 there were newlines in the string. This meant that an invalid header could
3826 be inserted if Exim was badly configured.
3828 7. Allow an ACL "warn" statement to add header lines at the start or after all
3829 the Received: headers, as well as at the end.
3831 8. Added the rcpt_4xx retry error code.
3833 9. Added postmaster_mailfrom=xxx to callout verification option.
3835 10. Added mailfrom=xxxx to the callout verification option, for verify=
3838 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
3839 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
3841 12. Inserted some casts to stop certain compilers warning when using pointer
3842 differences as field lengths or precisions in printf-type calls (mostly
3843 affecting debugging statements).
3845 13. Added optional readline() support for -be (dynamically loaded).
3847 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
3848 same clock tick as a message's arrival, so that its received time was the
3849 same as the "first fail" time on the retry record, and that message
3850 remained on the queue past the ultimate address timeout, every queue runner
3851 would try a delivery (because it was past the ultimate address timeout) but
3852 after another failure, the ultimate address timeout, which should have then
3853 bounced the address, did not kick in. This was a "< instead of <=" error;
3854 in most cases the first failure would have been in the next clock tick
3855 after the received time, and all would be well.
3857 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
3858 being recognized when the domain list was tested by the match_domain
3859 condition in an expansion string.
3861 16. Added the ${str2b64: operator.
3863 17. Exim was always calling setrlimit() to set a large limit for the number of
3864 processes, without checking whether the existing limit was already
3865 adequate. (It did check for the limit on file descriptors.) Furthermore,
3866 errors from getrlimit() and setrlimit() were being ignored. Now they are
3867 logged to the main and panic logs, but Exim does carry on, to try to do its
3868 job under whatever limits there are.
3870 18. Imported PCRE 5.0.
3872 19. Trivial typo in log message " temporarily refused connection" (the leading
3875 20. If the log selector return_path_on_delivery was set and an address was
3876 redirected to /dev/null, the delivery process crashed because it assumed
3877 that a return path would always be set for a "successful" delivery. In this
3878 case, the whole delivery is bypassed as an optimization, and therefore no
3881 21. Internal re-arrangement: the function for sending a challenge and reading
3882 a response while authentication was assuming a zero-terminated challenge
3883 string. It's now changed to take a pointer and a length, to allow for
3884 binary data in such strings.
3886 22. Added the cyrus_sasl authenticator (code supplied by MBM).
3888 23. Exim was not respecting finduser_retries when seeking the login of the
3889 uid under which it was called; it was always trying 10 times. (The default
3890 setting of finduser_retries is zero.) Also, it was sleeping after the final
3891 failure, which is pointless.
3893 24. Implemented tls_on_connect_ports.
3895 25. Implemented acl_smtp_predata.
3897 26. If the domain in control=submission is set empty, Exim assumes that the
3898 authenticated id is a complete email address when it generates From: or
3899 Sender: header lines.
3901 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
3902 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
3903 chown and chgrp in /bin and hostname in /usr/bin.
3905 28. Exim was keeping the "process log" file open after each use, just as it
3906 does for the main log. This opens the possibility of it remaining open for
3907 long periods when the USR1 signal hits a daemon. Occasional processlog
3908 errors were reported, that could have been caused by this. Anyway, it seems
3909 much more sensible not to leave this file open at all, so that is what now
3912 29. The long-running daemon process does not normally write to the log once it
3913 has entered its main loop, and it closes the log before doing so. This is
3914 so that log files can straightforwardly be renamed and moved. However,
3915 there are a couple of unusual error situations where the daemon does write
3916 log entries, and I had neglected to close the log afterwards.
3918 30. The text of an SMTP error response that was received during a remote
3919 delivery was being truncated at 512 bytes. This is too short for some of
3920 the long messages that one sometimes sees. I've increased the limit to
3923 31. It is now possible to make retry rules that apply only when a message has a
3924 specific sender, in particular, an empty sender.
3926 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
3927 it possible to be selective about when SMTP synchronization is enforced.
3929 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
3931 32. Implemented hosts_connection_nolog.
3933 33. Added an ACL for QUIT.
3935 34. Setting "delay_warning=" to disable warnings was not working; it gave a
3938 35. Added mailbox_size and mailbox_filecount to appendfile.
3940 36. Added control = no_multiline_responses to ACLs.
3942 37. There was a bug in the logic of the code that waits for the clock to tick
3943 in the case where the clock went backwards by a substantial amount such
3944 that the microsecond fraction of "now" was more than the microsecond
3945 fraction of "then" (but the whole seconds number was less).
3947 38. Added support for the libradius Radius client library this is found on
3948 FreeBSD (previously only the radiusclient library was supported).
3954 1. When certain lookups returned multiple values in the form name=value, the
3955 quoting of the values was not always being done properly. Specifically:
3956 (a) If the value started with a double quote, but contained no whitespace,
3958 (b) If the value contained whitespace other than a space character (i.e.
3959 tabs or newlines or carriage returns) it was not quoted.
3960 This fix has been applied to the mysql and pgsql lookups by writing a
3961 separate quoting function and calling it from the lookup code. The fix
3962 should probably also be applied to nisplus, ibase and oracle lookups, but
3963 since I cannot test any of those, I have not disturbed their existing code.
3965 2. A hit in the callout cache for a specific address caused a log line with no
3966 reason for rejecting RCPT. Now it says "Previous (cached) callout
3967 verification failure".
3969 3. There was an off-by-one bug in the queryprogram router. An over-long
3970 return line was truncated at 256 instead of 255 characters, thereby
3971 overflowing its buffer with the terminating zero. As well as fixing this, I
3972 have increased the buffer size to 1024 (and made a note to document this).
3974 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
3975 when Exim is waiting for an SMTP response from a remote server, Exim
3976 restarts its select() call on the socket, thereby resetting its timeout.
3977 This is not a problem when such interrupts are rare. Somebody set up a cron
3978 job to run exiwhat every 2 minutes, which is less than the normal select()
3979 timeout (5 or 10 minutes). This meant that the select() timeout never
3980 kicked in because it was always reset. I have fixed this by comparing the
3981 time when an interrupt arrives with the time at the start of the first call
3982 to select(). If more time than the timeout has elapsed, the interrupt is
3983 treated as a timeout.
3985 5. Some internal re-factoring in preparation for the addition of Sieve
3986 extensions (by MH). In particular, the "personal" test is moved to a
3987 separate function, and given an option for scanning Cc: and Bcc: (which is
3988 not set for Exim filters).
3990 6. When Exim created an email address using the login of the caller as the
3991 local part (e.g. when creating a From: or Sender: header line), it was not
3992 quoting the local part when it contained special characters such as @.
3994 7. Installed new OpenBSD configuration files.
3996 8. Reworded some messages for syntax errors in "and" and "or" conditions to
3997 try to make them clearer.
3999 9. Callout options, other than the timeout value, were being ignored when
4000 verifying sender addresses in header lines. For example, when using
4002 verify = header_sender/callout=no_cache
4004 the cache was (incorrectly) being used.
4006 10. Added a missing instance of ${EXE} to the exim_install script; this affects
4007 only the Cygwin environment.
4009 11. When return_path_on_delivery was set as a log selector, if different remote
4010 addresses in the same message used different return paths and parallel
4011 remote delivery occurred, the wrong values would sometimes be logged.
4012 (Whenever a remote delivery process finished, the return path value from
4013 the most recently started remote delivery process was logged.)
4015 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
4016 lines when AUTH and/or TLS are in use. This is the "received protocol"
4017 field. Exim used to use "asmtp" for authenticated SMTP, without any
4018 indication (in the protocol name) for TLS use. Now it follows the RFC and
4019 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
4020 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
4021 names appear in log lines as well as in Received: header lines.
4023 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
4024 extensions, and comparison tests, and to fix some bugs.
4026 14. Changes to the "personal" filter test:
4028 (1) The test was buggy in that it was just doing the equivalent of
4029 "contains" tests on header lines. For example, if a user's address was
4030 anne@some.where, the "personal" test would incorrectly be true for
4032 To: susanne@some.where
4034 This test is now done by extracting each address from the header in turn,
4035 and checking the entire address. Other tests that are part of "personal"
4036 are now done using regular expressions (for example, to check local parts
4037 of addresses in From: header lines).
4039 (2) The list of non-personal local parts in From: addresses has been
4040 extended to include "listserv", "majordomo", "*-request", and "owner-*",
4041 taken from the Sieve specification recommendations.
4043 (3) If the message contains any header line starting with "List-" it is
4044 treated as non-personal.
4046 (4) The test for "circular" in the Subject: header line has been removed
4047 because it now seems ill-conceived.
4049 15. Minor typos in src/EDITME comments corrected.
4051 16. Installed latest exipick from John Jetmore.
4053 17. If headers_add on a router specified a text string that was too long for
4054 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
4055 of string_sprintf() is now avoided.
4057 18. $message_body_size was not set (it was always zero) when running the DATA
4058 ACL and the local_scan() function.
4060 19. For the "mail" command in an Exim filter, no default was being set for
4061 the once_repeat time, causing a random time value to be used if "once" was
4062 specified. (If the value happened to be <= 0, no repeat happened.) The
4063 default is now 0s, meaning "never repeat". The "vacation" command was OK
4064 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
4065 (I found it when inspecting the code).
4067 20. There is now an overall timeout for performing a callout verification. It
4068 defaults to 4 times the callout timeout, which applies to individual SMTP
4069 commands during the callout. The overall timeout applies when there is more
4070 than one host that can be tried. The timeout is checked before trying the
4071 next host. This prevents very long delays if there are a large number of
4072 hosts and all are timing out (e.g. when the network connections are timing
4073 out). The value of the overall timeout can be changed by specifying an
4074 additional sub-option for "callout", called "maxwait". For example:
4076 verify = sender/callout=5s,maxwait=20s
4078 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
4079 to the end before writing, but this should make it even safer).
4081 22. Exim was forgetting that it had advertised PIPELINING for the second and
4082 subsequent messages on an SMTP connection. It was also not resetting its
4083 memory on STARTTLS and an internal HELO.
4085 23. When Exim logs an SMTP synchronization error within a session, it now
4086 records whether PIPELINING has been advertised or not.
4088 24. Added 3 instances of "(long int)" casts to time_t variables that were being
4089 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
4090 rather than long int.
4092 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
4094 26. Added the never_mail option to autoreply.
4100 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
4101 crash if the getsockname() call failed; this can happen if a connection is
4102 closed very soon after it is established. The problem was simply in the
4103 order in which certain operations were done, causing Exim to try to write
4104 to the SMTP stream before it had set up the file descriptor. The bug has
4105 been fixed by making things happen in the correct order.
4111 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
4112 before the connection was closed, thus losing the rejection response.
4114 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
4115 some early Solaris releases, but causes trouble in current releases where
4116 socklen_t is defined.
4118 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
4121 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
4124 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
4125 that releasing the top of it at the end releases what was used for sub-
4126 expansions (unless the block got too big). However, discard this block if
4127 the first thing is a variable or header, so that we can use its block when
4128 it is dynamic (useful for very large $message_headers, for example).
4130 6. Lookups now cache *every* query, not just the most recent. A new, separate
4131 store pool is used for this. It can be recovered when all lookup caches are
4132 flushed. Lookups now release memory at the end of their result strings.
4133 This has involved some general refactoring of the lookup sources.
4135 7. Some code has been added to the store_xxx() functions to reduce the amount
4136 of flapping under certain conditions.
4138 8. log_incoming_interface used to affect only the <= reception log lines. Now
4139 it causes the local interface and port to be added to several more SMTP log
4140 lines, for example "SMTP connection from", and rejection lines.
4142 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
4144 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
4146 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
4147 could be overwritten at the end of the current message (or the start of a
4148 new message if it was set in a HELO ACL). The value is now preserved for
4149 the duration of the SMTP connection.
4151 12. If a transport had a headers_rewrite setting, and a matching header line
4152 contained an unqualified address, that address was qualified, even if it
4153 did not match any rewriting rules. The underlying bug was that the values
4154 of the flags that permit the existence of unqualified sender and recipient
4155 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
4156 non-local messages, and by -bnq for local messages) were not being
4157 preserved with the message after it was received.
4159 13. When Exim was logging an SMTP synchronization error, it could sometimes log
4160 "next input=" as part of the text comprising the host identity instead of
4161 the correct text. The code was using the same buffer for two different
4162 strings. However, depending on which order the printing function evaluated
4163 its arguments, the bug did not always show up. Under Linux, for example, my
4164 test suite worked just fine.
4166 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
4167 doesn't work with some older versions of Perl. It has been changed to "my",
4168 which in any case is probably the better facility to use.
4170 15. A really picky compiler found some instances of statements for creating
4171 error messages that either had too many or two few arguments for the format
4174 16. The size of the buffer for calls to the DNS resolver has been increased
4175 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
4176 for addresses that have a lot of PTR records. This alleviates a problem; it
4177 does not fully solve it.
4179 17. A dnsdb lookup for PTR records that receives more data than will fit in the
4180 buffer now truncates the list and logs the incident, which is the same
4181 action as happens when Exim is looking up a host name and its aliases.
4182 Previously in this situation something unpredictable would happen;
4183 sometimes it was "internal error: store_reset failed".
4185 18. If a server dropped the connection unexpectedly when an Exim client was
4186 using GnuTLS and trying to read a response, the client delivery process
4187 crashed while trying to generate an error log message.
4189 19. If a "warn" verb in an ACL added multiple headers to a message in a single
4190 string, for example:
4192 warn message = H1: something\nH2: something
4194 the text was added as a single header line from Exim's point of view
4195 though it ended up OK in the delivered message. However, searching for the
4196 second and subsequent header lines using $h_h2: did not work. This has been
4197 fixed. Similarly, if a system filter added multiple headers in this way,
4198 the routers could not see them.
4200 20. Expanded the error message when iplsearch is called with an invalid key to
4201 suggest using net-iplsearch in a host list.
4203 21. When running tests using -bh, any delays imposed by "delay" modifiers in
4204 ACLs are no longer actually imposed (and a message to that effect is
4207 22. If a "gecos" field in a passwd entry contained escaped characters, in
4208 particular, if it contained a \" sequence, Exim got it wrong when building
4209 a From: or a Sender: header from that name. A second bug also caused
4210 incorrect handling when an unquoted " was present following a character
4211 that needed quoting.
4213 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
4214 was not being matched caselessly.
4216 24. Arranged for all hyphens in the exim.8 source to be escaped with
4219 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
4220 properly. Recipient callout cache records were still being keyed to include
4221 the sender, even when use_sender was set false. This led to far more
4222 callouts that were necessary. The sender is no longer included in the key
4223 when use_sender is false.
4225 26. Added "control = submission" modifier to ACLs.
4227 27. Added the ${base62d: operator to decode base 62 numbers.
4229 28. dnsdb lookups can now access SRV records.
4231 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
4232 the configuration file.
4234 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
4235 (-v) mode. This makes the output for a verbose queue run more intelligible.
4237 31. Added a use_postmaster feature to recipient callouts.
4239 32. Added the $body_zerocount variable, containing the number of binary zero
4240 bytes in the message body.
4242 33. The time of last modification of the "new" subdirectory is now used as the
4243 "mailbox time last read" when there is a quota error for a maildir
4246 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
4248 35. Added +ignore_unknown as a special item in host lists.
4250 36. Code for decoding IPv6 addresses in host lists is now included, even if
4251 IPv6 support is not being compiled. This fixes a bug in which an IPv6
4252 address was recognized as an IP address, but was then not correctly decoded
4253 into binary, causing unexpected and incorrect effects when compared with
4260 1. Very minor rewording of debugging text in manualroute to say "list of
4261 hosts" instead of "hostlist".
4263 2. If verify=header_syntax was set, and a header line with an unqualified
4264 address (no domain) and a large number of spaces between the end of the
4265 name and the colon was received, the reception process suffered a buffer
4266 overflow, and (when I tested it) crashed. This was caused by some obsolete
4267 code that should have been removed. The fix is to remove it!
4269 3. When running in the test harness, delay a bit after writing a bounce
4270 message to get a bit more predictability in the log output.
4272 4. Added a call to search_tidyup() just before forking a reception process. In
4273 theory, someone could use a lookup in the expansion of smtp_accept_max_
4274 per_host which, without the tidyup, could leave open a database connection.
4276 5. Added the variables $recipient_data and $sender_data which get set from a
4277 lookup success in an ACL "recipients" or "senders" condition, or a router
4278 "senders" option, similar to $domain_data and $local_part_data.
4280 6. Moved the writing of debug_print from before to after the "senders" test
4283 7. Change 4.31/66 (moving the time when the Received: is generated) caused
4284 problems for message scanning, either using a data ACL, or using
4285 local_scan() because the Received: header was not generated till after they
4286 were called (in order to set the time as the time of reception completion).
4287 I have revised the way this works. The header is now generated after the
4288 body is received, but before the ACL or local_scan() are called. After they
4289 are run, the timestamp in the header is updated.
4295 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
4296 before starting a queue runner without re-exec. This happened only when
4297 deliver_drop_privilege was set or when the Exim user was set to root. The
4298 effect of the bug was that timeouts during subsequent deliveries caused
4299 crashes instead of being properly handled. The handler is now left at its
4300 default (and expected) setting.
4302 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
4303 message, again when deliver_drop_privilege is set or Exim is run as root.
4304 The bug described in (1) was not present in this case, but the tidying up
4305 of the other signals was missing. I have made the two cases consistent.
4307 3. The ignore_target_hosts setting on a manualroute router was being ignored
4308 for hosts that were looked up using the /MX notation.
4310 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
4313 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
4314 operated on the sender address. After changing the $sender_address to <>
4315 for the sender address verify, Exim was re-instated it as the original
4316 (before rewriting) address, but remembering that it had rewritten it, so it
4317 wasn't rewriting it again. This bug also had the effect of breaking the
4318 sender address verification caching when the sender address was rewritten.
4320 6. The ignore_target_hosts option was being ignored by the ipliteral router.
4321 This has been changed so that if the ip literal address matches
4322 ignore_target_hosts, the router declines.
4324 7. Added expansion conditions match_domain, match_address, and match_local_
4325 part (NOT match_host).
4327 8. The placeholder for the Received: header didn't have a length field set.
4329 9. Added code to Exim itself and to exim_lock to test for a specific race
4330 condition that could lead to file corruption when using MBX delivery. The
4331 issue is with the lockfile that is created in /tmp. If this file is removed
4332 after a process has opened it but before that process has acquired a lock,
4333 there is the potential for a second process to recreate the file and also
4334 acquire a lock. This could lead to two Exim processes writing to the file
4335 at the same time. The added code performs the same test as UW imapd; it
4336 checks after acquiring the lock that its file descriptor still refers to
4337 the same named file.
4339 10. The buffer for building added header lines was of fixed size, 8192 bytes.
4340 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
4343 11. Added the smtp_active_hostname option. If used, this will typically be made
4344 to depend on the incoming interface address. Because $interface_address is
4345 not set up until the daemon has forked a reception process, error responses
4346 that can happen earlier (such as "too many connections") no longer contain
4349 12. If an expansion in a condition on a "warn" statement fails because a lookup
4350 defers, the "warn" statement is abandoned, and the next ACL statement is
4351 processed. Previously this caused the whole ACL to be aborted.
4353 13. Added the iplsearch lookup type.
4355 14. Added ident_timeout as a log selector.
4357 15. Added tls_certificate_verified as a log selector.
4359 16. Added a global option tls_require_ciphers (compare the smtp transport
4360 option of the same name). This controls incoming TLS connections.
4362 17. I finally figured out how to make tls_require_ciphers do a similar thing
4363 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
4364 before starting the TLS session.
4366 18. Tabs are now shown as \t in -bP output.
4368 19. If the log selector return_path_on_delivery was set, Exim crashed when
4369 bouncing a message because it had too many Received: header lines.
4371 20. If two routers both had headers_remove settings, and the first one included
4372 a superfluous trailing colon, the final name in the first list and the
4373 first name in the second list were incorrectly joined into one item (with a
4374 colon in the middle).
4380 1. Added -C and -D options to the exinext utility, mainly to make it easier
4381 to include in the automated testing, but these could be helpful when
4382 multiple configurations are in use.
4384 2. The exinext utility was not formatting the output nicely when there was
4385 an alternate port involved in the retry record key, nor when there was a
4386 message id as well (for retries that were specific to a specific message
4387 and a specific host). It was also confused by IPv6 addresses, because of
4388 the additional colons they contain. I have fixed the IPv4 problem, and
4389 patched it up to do a reasonable job for IPv6.
4391 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
4392 delivery, the log line now contains "pipelined" if PIPELINING was used.
4394 4. An SMTP transport process used to panic and die if the bind() call to set
4395 an explicit outgoing interface failed. This has been changed; it is now
4396 treated in the same way as a connect() failure.
4398 5. A reference to $sender_host_name in the part of a conditional expansion
4399 that was being skipped was still causing a DNS lookup. This no longer
4402 6. The def: expansion condition was not recognizing references to header lines
4403 that used bh_ and bheader_.
4405 7. Added the _cache feature to named lists.
4407 8. The code for checking quota_filecount in the appendfile transport was
4408 allowing one more file than it should have been.
4410 9. For compatibility with Sendmail, the command line option
4418 and sets the incoming protocol and host name (for trusted callers). The
4419 host name and its colon can be omitted when only the protocol is to be set.
4420 Note the Exim already has two private options, -pd and -ps, that refer to
4421 embedded Perl. It is therefore impossible to set a protocol value of "d" or
4422 "s", but I don't think that's a major issue.
4424 10. A number of refactoring changes to the code, none of which should affect
4427 (a) The number of logging options was getting close to filling up the
4428 32-bit word that was used as a bit map. I have split them into two classes:
4429 those that are passed in the argument to log_write(), and those that are
4430 only ever tested independently outside of that function. These are now in
4431 separate 32-bit words, so there is plenty of room for expansion again.
4432 There is no change in the user interface or the logging behaviour.
4434 (b) When building, for example, log lines, the code previously used a
4435 macro that called string_cat() twice, in order to add two strings. This is
4436 not really sufficiently general. Furthermore, there was one instance where
4437 it was actually wrong because one of the argument was used twice, and in
4438 one call a function was used. (As it happened, calling the function twice
4439 did not affect the overall behaviour.) The macro has been replaced by a
4440 function that can join an arbitrary number of extra strings onto a growing
4443 (c) The code for expansion conditions now uses a table and a binary chop
4444 instead of a serial search (which was left over from when there were very
4445 few conditions). Also, it now recognizes conditions like "pam" even when
4446 the relevant support is not compiled in: a suitably worded error message is
4447 given if an attempt is made to use such a condition.
4449 11. Added ${time_interval:xxxxx}.
4451 12. A bug was causing one of the ddress fields not to be passed back correctly
4452 from remote delivery subprocesses. The field in question was not being
4453 subsequently used, so this caused to problems in practice.
4455 13. Added new log selectors queue_time and deliver_time.
4457 14. Might have fixed a bug in maildirsizefile handling that threw up
4458 "unexpected character" debug warnings, and recalculated the data
4459 unnecessarily. In any case, I expanded the warning message to give more
4462 15. Added the message "Restricted characters in address" to the statements in
4463 the default ACL that block characters like @ and % in local parts.
4465 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
4466 Three changes have been made:
4468 (a) There was a serious bug; a negative response to MAIL caused the whole
4469 recipient domain to be cached as invalid, thereby blocking all messages
4470 to all local parts at the same domain, from all senders. This bug has
4471 been fixed. The domain is no longer cached after a negative response to
4472 MAIL if the sender used is not empty.
4474 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
4477 (c) A new callout option, "use_sender" has been added for people who want
4478 the modified behaviour.
4484 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
4487 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
4488 indeed breaks things for older releases.
4490 3. Added additional logging to the case where there is a problem reading data
4491 from a filter that is running in a subprocess using a pipe, in order to
4492 try to track down a specific problem.
4494 4. Testing facility fudge: when running in the test harness and attempting
4495 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
4496 getting "No route to host". Convert this to a timeout.
4498 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
4501 6. Some OS don't have socklen_t but use size_t instead. This affects the
4502 fifth argument of getsockopt() amongst other things. This is now
4503 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
4504 can be set for individual OS. I have set it for SunOS5, OSF1, and
4505 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
4506 some earlier ones do not.
4508 7. Change 4.30/15 was not doing the test caselessly.
4510 8. The standard form for an IPv6 address literal was being rejected by address
4511 parsing in, for example, MAIL and RCPT commands. An example of this kind of
4512 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4513 this, as well as the form without the "IPv6" on the front (but only when
4514 address literals are enabled, of course).
4516 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
4518 10. Exim crashed if a message with an empty sender address specified by -f
4519 encountered a router with an errors_to setting. This could be provoked only
4520 by a command such as
4524 where an empty string was supplied; "<>" did not hit this bug.
4526 11. Installed PCRE release 4.5.
4528 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
4529 remained set. It is now erased.
4531 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
4532 times from message ids (which are base 36 rather than the normal 62).
4534 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
4535 were being counted as actual protocol errors, and logged if the log
4536 selector +smtp_protocol_error was set. One cannot be perfect in this test,
4537 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
4538 and DATA following a set of rejected RCPTs do not count as protocol errors.
4539 In other words, Exim assumes they were pipelined, though this may not
4540 actually be the case. Of course, in all cases the client gets an
4541 appropriate error code.
4543 15. If a lookup fails in an ACL condition, a message about the failure may
4544 be available; it is used if testing the ACL cannot continue, because most
4545 such messages specify what the cause of the deferral is. However, some
4546 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
4547 that caused an old message to be retained and used if a later statement
4548 caused a defer, replacing the real cause of the deferral.
4550 16. If an IP address had so many PTR records that the DNS lookup buffer
4551 was not large enough to hold them, Exim could crash while trying to process
4552 the truncated data. It now detects and logs this case.
4554 17. Further to 4.21/58, another change has been made: if (and only if) the
4555 first line of a message (the first header line) ends with CRLF, a bare LF
4556 in a subsequent header line has a space inserted after it, so as not to
4557 terminate the header.
4559 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
4560 unnecessarily, used atoi() instead of strtol(), and didn't check the
4561 termination when getting file sizes from file names by regex.
4563 19. Completely re-implemented the support for maildirsize files, in the light
4564 of a number of problems with the previous contributed implementation
4565 (4.30/29). In particular:
4567 . If the quota is zero, the maildirsize file is maintained, but no quota is
4570 . If the maildir directory does not exist, it is created before any attempt
4571 to write a maildirsize file.
4573 . The quota value in the file is just a cache; if the quota is changed in
4574 the transport, the new value overrides.
4576 . A regular expression is available for excluding directories from the
4579 20. The autoreply transport checks the characters in options that define the
4580 message's headers; it allows continued headers, but it was checking with
4581 isspace() after an embedded newline instead of explicitly looking for a
4584 21. If all the "regular" hosts to which an address was routed had passed their
4585 expiry times, and had not reached their retry times, the address was
4586 bounced, even if fallback hosts were defined. Now Exim should go on to try
4589 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
4590 equivalent code in the SMTP transport. Some hosts send humungous responses
4591 to HELO/EHLO, more than 1024 it seems.
4593 23. Refactoring: code in filter.c used (void *) for "any old type" but this
4594 gives compiler warnings in some environments. I've now done it "properly",
4597 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
4598 (because of problems with the built-in one) was declared to return uschar *
4599 instead of char *, causing compiler failure.
4601 25. Fixed a file descriptor leak when processing alias/forward files.
4603 26. Fixed a minor format string issue in dbfn.c.
4605 27. Typo in exim.c: ("dmbnz" for "dbmnz").
4607 28. If a filter file refered to $h_xxx or $message_headers, and the headers
4608 contained RFC 2047 "words", Exim's memory could, under certain conditions,
4611 29. When a sender address is verified, it is cached, to save repeating the test
4612 when there is more than one recipient in a message. However, when the
4613 verification involves a callout, it is possible for different callout
4614 options to be set for different recipients. It is too complicated to keep
4615 track of this in the cache, so now Exim always runs a verification when a
4616 callout is required, relying on the callout cache for the optimization.
4617 The overhead is duplication of the address routing, but this should not be
4620 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
4621 to be verified with callout=postmaster, and the main callout worked but the
4622 postmaster check failed, the verification correctly failed. However, if a
4623 subsequent RCPT command asked for sender verification *without* the
4624 postmaster check, incorrect caching caused this verification also to fail,
4627 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
4628 it was not caching the DNS options (qualify_single, search_parents) that
4629 were used when the lookup failed. A subsequent lookup with different
4630 options therefore always gave the same answer, though there were cases
4631 where it should not have. (Example: a "domains = !$mx_any" option on a
4632 dnslookup router: the "domains" option is always processed without any
4633 widening, but the router might have qualify_single set.) Now Exim uses the
4634 cached value only when the same options are set.
4636 32. Added John Jetmore's "exipick" utility to the distribution.
4638 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
4639 than a timeout (e.g. a certificate is required, and is not provided), an
4640 Exim server now closes the connection immediately. Previously it waited for
4641 the client to close - but if the client is SSL, it seems that they each
4642 wait for each other, leading to a delay before one of them times out.
4644 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
4645 maintained 0.8.x compatibility because I don't think many are using it, and
4646 it is clearly obsolete.
4648 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
4651 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
4652 client certificate was expired. A simple patch fixes this, though I don't
4653 understand the full logic of why the verify callback is called multiple
4656 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
4657 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
4658 which causes problems with some clients (such as the Certicom SSL Plus
4659 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
4660 disables the coutermeasure allowing Eudora to connect."
4662 38. Exim was not checking that a write() to a log file succeeded. This could
4663 lead to Bad Things if a log got too big, in particular if it hit a file
4664 size limit. Exim now panics and dies if it cannot write to a log file, just
4665 as it does if it cannot open a log file.
4667 39. Modified OS/Makefile-Linux so that it now contains
4669 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
4671 The two -D definitions ensure that Exim is compiled with large file
4672 support, which makes it possible to handle log files that are bigger than
4675 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
4676 instance) a domain was checked against a named list that involved a lookup,
4677 causing $domain_data to be set, then another domain was checked against the
4678 same list, then the first domain was re-checked, the value of $domain_data
4679 after the final check could be wrong. In particular, if the second check
4680 failed, it could be set empty. This bug probably also applied to
4683 41. The strip_trailing_dot option was not being applied to the address given
4684 with the -f command-line option.
4686 42. The code for reading a message's header from the spool was incrementing
4687 $received_count, but never initializing it. This meant that the value was
4688 incorrect (doubled) while delivering a message in the same process in which
4689 it was received. In the most common configuration of Exim, this never
4690 happens - a fresh exec is done - but it can happen when
4691 deliver_drop_privilege is set.
4693 43. When Exim logs an SMTP synchronization error - client data sent too soon -
4694 it now includes up to 150 characters of the unexpected data in the log
4697 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
4698 and building data strings. The size of both of these buffers was 10 000
4699 bytes - far larger than anybody would *ever* want, thought I. Needless to
4700 say, somebody hit the limit. I have increased the maximum line length to
4701 20 000 and the maximum data length of concatenated lines to 100 000. I have
4702 also fixed two bugs, because there was no checking on these buffers. Tsk,
4703 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
4704 buffer is too small.
4706 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
4707 lsearch lookups. Now it does.
4709 46. When parsing a route_list item in a manualroute router, a fixed-length
4710 buffer was used for the list of hosts. I made this 1024 bytes long,
4711 thinking that nobody would ever have a list of hosts that long. Wrong.
4712 Somebody had a whole pile of complicated expansion conditions, and the
4713 string was silently truncated, leading to an expansion error. It turns out
4714 that it is easier to change to an unlimited length (owing to other changes
4715 that have happened since this code was originally written) than to build
4716 structure for giving a limitation error. The length of the item that
4717 expands into the list of hosts is now unlimited.
4719 47. The lsearch lookup could not handle data where the length of text line was
4720 more than 4095 characters. Such lines were truncated, leading to shortened
4721 data being returned. It should now handle lines of any length.
4723 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
4724 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
4727 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
4728 replacing strings such as BIN_DIRECTORY in a source file: the replacement
4729 no longer happens in comment lines. A list of replacements is now placed
4730 at the head of all of the source files, except those whose only change is
4731 to replace PERL_COMMAND in the very first #! line.
4733 50. Replaced the slow insertion sort in queue.c, for sorting the list of
4734 messages on the queue, with a bottom-up merge sort, using code contributed
4735 by Michael Haardt. This should make operations like -bp somewhat faster on
4736 large queues. It won't affect queue runners, except when queue_run_in_order
4739 51. Installed eximstats 1.31 in the distribution.
4741 52. Added support for SRV lookups to the dnslookup router.
4743 53. If an ACL referred to $message_body or $message_body_end, the value was not
4744 reset for any messages that followed in the same SMTP session.
4746 54. The store-handling optimization for building very long strings was not
4747 differentiating between the different store pools. I don't think this
4748 actually made any difference in practice, but I've tidied it.
4750 55. While running the routers to verify a sender address, $sender_address
4751 was still set to the sender address. This is wrong, because when routing to
4752 send a bounce to the sender, it would be empty. Therefore, I have changed
4753 it so that, while verifying a sender address, $sender_address is set to <>.
4754 (There is no change to what happens when verifying a recipient address.)
4756 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
4757 target A or AAAA records (if not already returned) without resetting the
4758 qualify_single or search_parents options of the DNS resolver. These are
4759 inappropriate in this case because the targets of MX and SRV records must
4760 be FQDNs. A broken DNS record could cause trouble if it happened to have a
4761 target that, when qualified, matched something in the local domain. These
4762 two options are now turned off when doing these lookups.
4764 57. It seems that at least some releases of Reiserfs (which does not have the
4765 concept of a fixed number of inodes) returns zero and not -1 for the
4766 number of available inodes. This interacted badly with check_spool_inodes,
4767 which assumed that -1 was the "no such thing" setting. What I have done is
4768 to check that the total number of inodes is greater than zero before doing
4769 the test of how many are available.
4771 58. When a "warn" ACL statement has a log_message modifier, the message is
4772 remembered, and not repeated. This is to avoid a lot of repetition when a
4773 message has many recipients that cause the same warning to be written.
4774 Howewer, Exim was preserving the list of already written lines for an
4775 entire SMTP session, which doesn't seem right. The memory is now reset if a
4776 new message is started.
4778 59. The "rewrite" debugging flag was not showing the result of rewriting in the
4779 debugging output unless log_rewrite was also set.
4781 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
4782 of (int)(handle) when we know that handle contains (void *)(-1).
4784 61. The Exim daemon panic-logs an error return when it closes the incoming
4785 connection. However "connection reset by peer" seems to be common, and
4786 isn't really an error worthy of noting specially, so that particular error
4789 62. When Exim is trying to find all the local interfaces, it used to panic and
4790 die if the ioctl to get the interface flags failed. However, it seems that
4791 on at least one OS (Solaris 9) it is possible to have an interface that is
4792 included in the list of interfaces, but for which you get a failure error
4793 for this call. This happens when the interface is not "plumbed" into a
4794 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
4795 failure of the "get flags" call assumes that the interface is down.
4797 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
4798 makes life easier for people who are doing arithmetic on fields extracted
4799 from dates, where you often get leading zeros that should not be
4800 interpreted as octal.
4802 64. Added qualify_domain to the redirect router, to override the global
4805 65. If a pathologically long header line contained very many addresses (the
4806 report of this problem mentioned 10 000) and each of them was rewritten,
4807 Exim could use up a very large amount of memory. (It kept on making new
4808 copies of the header line as it rewrote, and never released the old ones.)
4809 At the expense of a bit more processing, the header rewriting function has
4810 been changed so that it no longer eats memory in this way.
4812 66. The generation of the Received: header has been moved from the time that a
4813 message starts to be received, to the time that it finishes. The timestamp
4814 in the Received: header should now be very close to that of the <= log
4815 line. There are two side-effects of this change:
4817 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
4818 logged header lines no longer include the local Received: line, because
4819 it has not yet been created. The same applies to a copy of the message
4820 that is returned to a non-SMTP sender when a message is rejected.
4822 (b) When a filter file is tested using -bf, no additional Received: header
4823 is added to the test message. After some thought, I decided that this
4826 This change does not affect the value of $received_for. It is still set
4827 after address rewriting, but before local_scan() is called.
4829 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
4831 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
4832 gave an unhelpful panic error message, and a defer error. I have managed to
4833 change this behaviour so that it now rejects any supplied certificate,
4834 which seems right, as the list of acceptable certificates is empty.
4836 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
4837 gave an unhelpful defer error. I have not managed to make this reject any
4838 supplied certificates, but the error message it gives is "no certificate
4839 supplied", which is not helpful.
4841 70. exigrep's output now also includes lines that are not associated with any
4842 message, but which match the given pattern. Implemented by a patch from
4843 Martin Sluka, which also tidied up the Perl a bit.
4845 71. Recipient callout verification, like sender verification, was using <> in
4846 the MAIL FROM command. This isn't really the right thing, since the actual
4847 sender may affect whether the remote host accepts the recipient or not. I
4848 have changed it to use the actual sender in the callout; this means that
4849 the cache record is now keyed on a recipient/sender pair, not just the
4850 recipient address. There doesn't seem to be a real danger of callout loops,
4851 since a callout by the remote host to check the sender would use <>.
4852 [SEE ABOVE: changed after hitting problems.]
4854 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
4855 temporary errors. However, in the case of such a code being given after
4856 the end of a data transmission (i.e. after ".") Exim was failing to write
4857 a retry record for the message. (Yes, there was some broken host that was
4858 actually sending 8xx at this point.)
4860 73. An unknown lookup type in a host list could cause Exim to panic-die when
4861 the list was checked. (An example that provoked this was putting <; in the
4862 middle of a list instead of at the start.) If this happened during a DATA
4863 ACL check, a -D file could be left lying around. This kind of configuration
4864 error no longer causes Exim to die; instead it causes a defer errror. The
4865 incident is still logged to the main and panic logs.
4867 74. Buglet left over from Exim 3 conversion. The message "too many messages
4868 in one connection" was written to the rejectlog but not the mainlog, except
4869 when address rewriting (yes!) was being logged.
4871 75. Added write_rejectlog option.
4873 76. When a system filter was run not as root (that is, when system_filter_user
4874 was set), the values of the $n variables were not being returned to the
4875 main process; thus, they were not subsequently available in the $sn
4878 77. Added +return_path_on_delivery log selector.
4880 78. A connection timeout was being treated differently from recipients deferred
4881 when testing hosts_max_try with a message that was older than the host's
4882 retry timeout. (The host should not be counted, thus allowing all hosts to
4883 be tried at least once before bouncing.) This may have been the cause of an
4884 occasionally reported bug whereby a message would remain on the queue
4885 longer than the retry timeout, but would be bounced if a delivery was
4886 forced. I say "may" because I never totally pinned down the problem;
4887 setting up timeout/retry tests is difficult. See also the next item.
4889 79. The ultimate address timeout was not being applied to errors that involved
4890 a combination of host plus message (for example, a timeout on a MAIL
4891 command). When an address resolved to a number of possible hosts, and they
4892 were not all tried for each delivery (e.g. because of hosts_max_try), a
4893 message could remain on the queue longer than the retry timeout.
4895 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
4898 81. Fixed an obscure SMTP outgoing bug which required at least the following
4899 conditions: (a) there was another message waiting for the same server;
4900 (b) the server returned 5xx to all RCPT commands in the first message so
4901 that the message was not completed; (c) the server dropped the connection
4902 or gave a negative response to the RSET that Exim sends to abort the
4903 transaction. The observed case was a dropped connection after DATA that had
4904 been sent in pipelining mode. That is, the server had advertised PIPELINING
4905 but was not implementing it correctly. The effect of the bug was incorrect
4906 behaviour, such as trying another host, and this could lead to a crash.
4912 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
4913 and daemon.c were passed as pointers to ints; they should have been
4914 pointers to socklen_t variables (which are typically unsigned ints).
4916 2. Some signed/unsigned type warnings in the os.c file for Linux have been
4919 3. Fixed a really odd bug that affected only the testing scheme; patching a
4920 certain fixed string in the binary changed the value of another string that
4921 happened to be identical to the end of the original first string.
4923 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
4924 name", it returns that address as the IP address. On some operating
4925 systems (e.g. Solaris), it also passes back the IP address string as the
4926 "host name". However, on others (e.g. Linux), it passes back an empty
4927 string. Exim wasn't checking for this, and was changing the host name to an
4928 empty string, assuming it had been canonicized.
4930 5. Although rare, it is permitted to have more than one PTR record for a given
4931 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
4932 all the names associated with an address, because they do in Solaris.
4933 However, it seems that they do not in Linux for data that comes from the
4934 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
4935 I found this out when I moved to a new Linux workstation and tried to run
4936 the Exim test suite.
4938 To get round this problem I have changed the code so that it now does its
4939 own call to the DNS to look up PTR records when searching for a host name.
4940 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
4941 addresses that are only in /etc/hosts are still found.
4943 This behaviour is, however, controlled by an option called host_lookup_
4944 order, which defaults to "bydns:byaddr". If people want to use the other
4945 order, or indeed, just use one or the other means of lookup, they can
4946 specify it in this variable.
4948 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
4949 some operating systems, this comes back from gethostbyaddr() as an empty
4950 string, and this is what Exim used to test for. However, it seems that in
4951 other systems, "." is yielded. Exim now tests for this case too.
4953 7. The values of check_spool_space and check_log_space are now held internally
4954 as a number of kilobytes instead of an absolute number of bytes. If a
4955 numbers is specified without 'K' or 'M', it is rounded up to the nearest
4956 kilobyte. This means that much larger values can be stored.
4958 8. Exim monitor: an attempt to get the action menu when not actually pointing
4959 at a message produces an empty menu entitled "No message selected". This
4960 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
4961 no entries in it ("Shell widget menu has zero width and/or height"). So I
4962 have added a single, blank menu entry in this case.
4964 9. Added ${quote_local_part.
4966 10. MIME decoding is now applied to the contents of Subject: header lines when
4969 11. Now that a reference to $sender_host_address automatically causes a reverse
4970 lookup to occur if necessary (4.13/18), there is no need to arrange for a
4971 host lookup before query-style lookups in lists that might use this
4972 variable. This has therefore been abolished, and the "net-" prefix is no
4973 longer necessary for query-style lookups.
4975 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
4976 have been a typo for LFLAGS, so it has been changed.
4978 13. The install script calls Exim with "-C /dev/null" in order to find the
4979 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
4980 to be output. Howeve, since Exim outputs its version number before the
4981 error, it didn't break the script. It just looked ugly. I fixed this by
4982 always allowing "-C /dev/null" if the caller is root.
4984 14. Ignore overlarge ACL variable number when reading spool file - insurance
4985 against a later release with more variables having written the file.
4987 15. The standard form for an IPv6 address literal was being rejected by EHLO.
4988 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
4989 this, as well as the form without the "IPv6" on the front.
4991 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
4992 OS/Makefile-Darwin file.
4994 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
4995 only to LDAP libraries that do not have LDAP_OPT_DEREF.
4997 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
4998 variable. However, some OS use ints for inodes. I've added cast to long int
4999 to get rid of the compiler warning.
5001 19. I had forgotten to lock out "/../" in configuration file names when
5002 ALT_CONFIG_PREFIX was set.
5004 20. Routers used for verification do not need to specify transports. However,
5005 if such a router generated a host list, and callout was configured, Exim
5006 crashed, because it could not find a port number from the (non-existent)
5007 transport. It now assumes port 25 in this circumstance.
5009 21. Added the -t option to exigrep.
5011 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
5012 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
5013 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
5014 them is set, LOOKUP_LSEARCH is forced.
5016 23. "exim -bV" now outputs a list of lookups that are included in the binary.
5018 24. Added sender and host information to the "rejected by local_scan()" log
5019 line; previously there was no indication of these.
5021 25. Added .include_if_exists.
5023 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
5024 Linux. It turns out that not all file systems support this. Apparently some
5025 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
5026 do it.) To cope with this, the error EINVAL, which means that sync-ing is
5027 not supported on the file descriptor, is now ignored when Exim is trying to
5028 sync a directory. This applies only to Linux.
5030 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
5032 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
5033 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
5036 29. Added support for maildirsize files from supplied patch (modified a bit).
5038 30. The use of :fail: followed by an empty string could lead Exim to respond to
5039 sender verification failures with (e.g.):
5041 550 Verification failed for <xxx>
5042 550 Sender verify failed
5044 where the first response line was missing the '-' that indicates it is not
5045 the final line of the response.
5047 31. The loop for finding the name of the user that called Exim had a hardwired
5048 limit of 10; it now uses the value of finduser_retries, which is used for
5049 all other user lookups.
5051 32. Added $received_count variable, available in data and not_smtp ACLs, and at
5054 33. Exim was neglecting to zero errno before one call of strtol() when
5055 expanding a string and expecting an integer value. On some systems this
5056 resulted in spurious "integer overflow" errors. Also, it was casting the
5057 result into an int without checking.
5059 34. Testing for a connection timeout using "timeout_connect" in the retry rules
5060 did not work. The code looks as if it has *never* worked, though it appears
5061 to have been documented since at least releast 1.62. I have made it work.
5063 35. The "timeout_DNS" error in retry rules, also documented since at least
5064 1.62, also never worked. As it isn't clear exactly what this means, and
5065 clearly it isn't a major issue, I have abolished the feature by treating it
5066 as "timeout", and writing a warning to the main and panic logs.
5068 36. The display of retry rules for -brt wasn't always showing the error code
5071 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
5072 timeout_connect_A, timeout_connect_MX.
5074 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
5075 to the empty sender.
5077 39. The daemon was not analysing the content of -oX till after it had closed
5078 stderr and disconnected from the controlling terminal. This meant that any
5079 syntax errors were only noted on the panic log, and the return code from
5080 the command was 0. By re-arranging the code a little, I've made the
5081 decoding happen first, so such errors now appear on stderr, and the return
5082 code is 1. However, the actual setting up of the sockets still happens in
5083 the disconnected process, so errors there are still only recorded on the
5086 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
5087 connections (as happens on some IP stacks) was logged at start up time as
5088 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
5089 it from "IPv6 and IPv4", which means that two separate sockets are being
5092 41. The debug output for gethostbyname2() or getipnodebyname() failures now
5093 says whether AF_INET or AF_INET6 was passed as an argument.
5095 42. Exiwhat output was messed up when time zones were included in log
5098 43. Exiwhat now gives more information about the daemon's listening ports,
5099 and whether -tls-on-connect was used.
5101 44. The "port" option of the smtp transport is now expanded.
5103 45. A "message" modifier in a "warn" statement in a non-message ACL was being
5104 silently ignored. Now an error message is written to the main and panic
5107 46. There's a new ACL modifier called "logwrite" which writes to a log file
5108 as soon as it is encountered.
5110 47. Added $local_user_uid and $local_user_gid at routing time.
5112 48. Exim crashed when trying to verify a sender address that was being
5115 49. Exim was recognizing only a space character after ".include". It now also
5116 recognizes a tab character.
5118 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
5119 extracting the relevant information from the specification. The man page no
5120 longer contains scrambled data for the -d option, and I've added a section
5121 at the front about calling Exim under different names.
5123 51. Added "extra_headers" argument to the "mail" command in filter files.
5125 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
5128 53. Installed eximstats 1.29.
5130 54. Added transport_filter_timeout as a generic transport option.
5132 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
5133 Cc: header lines. This was required by RFC 822, but it not required by RFC
5136 56. Exim used to add From:, Date:, and Message-Id: header lines to any
5137 incoming messages that did not have them. Now it does so only if the
5138 message originates locally, that is, if there is no associated remote host
5139 address. When Resent- header lines are present, this applies to the Resent-
5140 lines rather than the non-Resent- lines.
5142 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
5143 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
5145 58. Messages for configuration errors now include the name of the main
5146 configuration file - useful now that there may be more than one file in a
5147 list (.included file names were always shown).
5149 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
5150 for those rare installations that do not start the daemon as root or run it
5151 setuid root. I've cut out the call to initgroups() if the daemon is not
5154 60. The Exim user and group can now be bound into the binary as text strings
5155 that are looked up at the start of Exim's processing.
5157 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
5159 62. Added $mailstore_basename variable.
5161 63. Installed patch to sieve.c from Michael Haardt.
5163 64. When Exim failed to open the panic log after failing to open the main log,
5164 the original message it was trying to log was written to stderr and debug
5165 output, but if they were not available (the usual case in production), it
5166 was lost. Now it is written to syslog before the two lines that record the
5167 failures to open the logs.
5169 65. Users' Exim filters run in subprocesses under the user's uid. It is
5170 possible for a "deliver" command or an alias in a "personal" command to
5171 provoke an address rewrite. If logging of address rewriting is configured,
5172 this fails because the process is not running as root or exim. There may be
5173 a better way of dealing with this, but for the moment (because 4.30 needs
5174 to be released), I have disabled address rewrite logging when running a
5175 filter in a non-root, non-exim process.
5181 1. The buildconfig auxiliary program wasn't quoting the value set for
5182 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
5183 not defined. This bug was masked in 4.22 by the effect that was fixed in
5186 2. Some messages that were rejected after a message id was allocated were
5187 shown as "incomplete" by exigrep. It no longer does this for messages that
5188 are rejected by local_scan() or the DATA or non-SMTP ACLs.
5190 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
5191 have allow_domain_literals set, the ID did not get logged in the <= line.
5192 Domain literals are now always recognized in Message-ID: header lines.
5194 4. The first argument for a ${extract expansion item is the key name or field
5195 number. Leading and trailing spaces in this item were not being ignored,
5196 causing some misleading effects.
5198 5. When deliver_drop_privilege was set, single queue runner processes started
5199 manually (i.e. by the command "exim -q") or by the daemon (which uses the
5200 same command in the process it spins off) were not dropping privilege.
5202 6. When the daemon running as "exim" started a queue runner, it always
5203 re-executed Exim in the spun-off process. This is a waste of effort when
5204 deliver_drop_privilege is set. The new process now just calls the
5205 queue-runner function directly.
5211 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
5214 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
5215 ignored. Though the use of -oP was forcing the writing of a pid file, it
5216 was always written to the default place.
5218 3. If the message "no IP address found for host xxxx" is generated during
5219 incoming verification, it is now followed by identification of the incoming
5220 connection (so you can more easily find what provoked it).
5222 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
5224 5. Added some features to "harden" Exim a bit more against certain attacks:
5226 (a) There is now a build-time option called FIXED_NEVER_USERS that can
5227 be put in Local/Makefile. This is like the never_users runtime option,
5228 but it cannot be overridden. The default setting is "root".
5230 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
5231 prefix string with which any file named in a -C command line option
5234 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
5235 is retained for -C and -D only if the caller of Exim is root. Without
5236 it, the exim user may also use -C and -D and retain privilege.
5238 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
5239 command line option is disabled.
5241 6. Macro names set by the -D option must start with an upper case letter, just
5242 like macro names defined in the configuration file.
5244 7. Added "dereference=" facility to LDAP.
5246 8. Two instances of the typo "uknown" in the source files are fixed.
5248 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
5249 the Configure-Makefile script screwed up while processing it.
5251 10. Incorporated PCRE 4.4.
5253 11. The SMTP synchronization check was not operating right at the start of an
5254 SMTP session. For example, it could not catch a HELO sent before the client
5255 waited for the greeting. There is now a check for outstanding input at the
5256 point when the greeting is written. Because of the duplex, asynchronous
5257 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
5258 way, but not yet received, when the check is performed.
5260 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
5261 on TCP/IP sockets, because this apparently causes some broken clients to
5264 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
5265 unchanged) from the Cygwin maintainer.
5267 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
5268 was supported instead of testing for mbx. Effectively a typo.
5270 15. The spa authenticator server code was not checking that the input it
5271 received was valid base64.
5273 16. The debug output line for the "set" modifier in ACLs was not showing the
5274 name of the variable that was being set.
5276 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
5278 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
5279 lookup on its own. Something else was needed to trigger the lookup. For
5280 example, a match in host_lookup or the need for a host name in a host list.
5281 Now, if $sender_host_name is referenced and the host name has not yet been
5282 looked up, a lookup is performed. If the lookup fails, the variable remains
5283 empty, and $host_lookup_failed is set to "1".
5285 19. Added "eqi" as a case-independent comparison operator.
5287 20. The saslauthd authentication condition could segfault if neither service
5288 nor realm was specified.
5290 21. If an overflowing value such as "2048M" was set for message_size_limit, the
5291 error message that was logged was misleading, and incoming SMTP
5292 connections were dropped. The message is now more accurate, and temporary
5293 errors are given to SMTP connections.
5295 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
5296 (except RSET) with a 421 error, until QUIT is received. However, it was
5297 failing to send a response to QUIT.
5299 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
5300 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
5301 ACL is now run after the helo_try_verify_hosts code.
5303 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
5304 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
5305 case-independent, so other case variants are also recognized). Apparently
5306 some systems use these upper case variants.
5308 25. If more than two messages were waiting for the same host, and a transport
5309 filter was specified for the transport, Exim sent two messages over the
5310 same TCP/IP connection, and then failed with "socket operation on non-
5311 socket" when it tried to send the third.
5313 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
5315 27. The extern definition of crypt16() in expand.c was not being excluded when
5316 the OS had its own crypt16() function.
5318 28. Added bounce_return_body as a new option, and bounce_return_size_limit
5319 as a preferred synonym for return_size_limit, both as an option and as an
5322 29. Added LIBS=-liconv to OS/Makefile-OSF1.
5324 30. Changed the default configuration ACL to relax the local part checking rule
5325 for addresses that are not in any local domains. For these addresses,
5326 slashes and pipe symbols are allowed within local parts, but the sequence
5327 /../ is explicitly forbidden.
5329 31. SPA server authentication was not clearing the challenge buffer before
5332 32. log_message in a "warn" ACL statement was writing to the reject log as
5333 well as to the main log, which contradicts the documentation and doesn't
5334 seem right (because no rejection is happening). So I have stopped it.
5336 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
5337 However, I am unable to do any testing of this.
5339 34. Fixed an infelicity in the appendfile transport. When checking directories
5340 for a mailbox, to see if any needed to be created, it was accidentally
5341 using path names with one or more superfluous leading slashes; tracing
5342 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
5344 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
5345 contents are added to the log line that is written for every discarded
5346 recipient. (Previously a log_message setting was ignored.)
5348 36. The ${quote: operator now quotes the string if it is empty.
5350 37. The install script runs exim in order to find its version number. If for
5351 some reason other than non-existence or emptiness, which it checks, it
5352 could not run './exim', it was installing it with an empty version number,
5353 i.e. as "exim-". This error state is now caught, and the installation is
5356 38. An argument was missing from the function that creates an error message
5357 when Exim fails to connect to the socket for saslauthd authentication.
5358 This could cause Exim to crash, or give a corrupted message.
5360 39. Added isip, isip4, and isip6 to ${if conditions.
5362 40. The ACL variables $acl_xx are now saved with the message, and can be
5363 accessed later in routers, transports, and filters.
5365 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
5366 strings in the file are not string-expanded.
5368 42. If a MAIL command specified a SIZE value that was too large to fit into an
5369 int variable, the check against message_size_limit failed. Such values are
5370 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
5371 day this will have to be increased, but I don't think I want to be around
5372 when emails are that large.
5379 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
5380 iconv() is not standard in FreeBSD.
5382 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
5383 IPv6 enabled. The observed symptom was a segmentation fault on return from
5384 the function os_common_find_running_interfaces() in src/os.c.
5386 3. In the check_special_case() function in daemon.c I had used "errno" as an
5387 argument name, which causes warnings on some systems. This was basically a
5388 typo, since it was named "eno" in the comments!
5390 4. The code that waits for the clock to tick (at a resolution of some fraction
5391 of a second) so as to ensure message-id uniqueness was always waiting for
5392 at least one whole tick, when it could have waited for less. [This is
5393 almost certainly not relevant at current processor speeds, where it is
5394 unlikely to ever wait at all. But we try to future-proof.]
5396 5. The function that sleeps for a time interval that includes fractions of a
5397 second contained a race. It did not block SIGALRM between setting the
5398 timer, and suspending (a couple of lines later). If the interval was short
5399 and the sigsuspend() was delayed until after it had expired, the suspension
5400 never ended. On busy systems this could lead to processes getting stuck for
5403 6. Some uncommon configurations may cause a lookup to happen in a queue runner
5404 process, before it forks any delivery processes. The open lookup caching
5405 mechanism meant that the open file or database connection was passed into
5406 the delivery process. The problem was that delivery processes always tidy
5407 up cached lookup data. This could cause a problem for the next delivery
5408 process started by the queue runner, because the external queue runner
5409 process does not know about the closure. So the next delivery process
5410 still has data in the lookup cache. In the case of a file lookup, there was
5411 no problem because closing a file descriptor in a subprocess doesn't affect
5412 the parent. However, if the lookup was caching a connection to a database,
5413 the connection was closed, and the second delivery process was likely to
5414 see errors such as "PGSQL: query failed: server closed the connection
5415 unexpectedly". The problem has been fixed by closing all cached lookups
5416 in a queue runner before running a delivery process.
5418 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
5419 seem to have the "const" qualifier which it has on other OS. I've
5422 8. Change 4.21/2 was too strict. It is only if there are two authenticators
5423 *of the same type* (client or server) with the same public name that an
5424 error should be diagnosed.
5426 9. When Exim looked up a host name for an IP address, but failed to find the
5427 original IP address when looking up the host name (a safety check), it
5428 output the message "<ip address> does not match any IP for NULL", which was
5429 confusing, to say the least. The bug was that the host name should have
5430 appeared instead of "NULL".
5432 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
5433 user that is built into the binary, and the -C or -D options is used, root
5434 privilege is dropped before the configuration file is read. In addition,
5435 logging is switched to stderr instead of the normal log files. If the
5436 configuration then re-defines the Exim user, the unprivileged environment
5437 is probably not what is expected, so Exim logs a panic warning message (but
5440 However, if deliver_drop_privilege is set, the unprivileged state may well
5441 be exactly what is intended, so the warning has been cut out in that case,
5442 and Exim is allowed to try to write to its normal log files.
5448 1. smtp_return_error_details was not giving details for temporary sender
5449 or receiver verification errors.
5451 2. Diagnose a configuration error if two authenticators have the same public
5454 3. Exim used not to create the message log file for a message until the first
5455 delivery attempt. This could be confusing when incoming messages were held
5456 for policy or load reasons. The message log file is now created at the time
5457 the message is received, and an initial "Received" line is written to it.
5459 4. The automatically generated man page for command line options had a minor
5460 bug that caused no ill effects; however, a more serious problem was that
5461 the procedure for building the man page automatically didn't always
5462 operate. Consequently, release 4.20 contains an out-of-date version. This
5463 shouldn't happen again.
5465 5. When building Exim with embedded Perl support, the script that builds the
5466 Makefile was calling 'perl' to find its compile-time parameters, ignoring
5467 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
5469 6. The freeze_tell option was not being used for messages that were frozen on
5470 arrival, either by an ACL or by local_scan().
5472 7. Added the smtp_incomplete_transaction log selector.
5474 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
5475 was accepting AUTH without a new EHLO.
5477 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
5478 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
5481 10. Logging of TCP/IP connections (when configured) now happens in the main
5482 daemon process instead of the child process, so that the TCP/IP connection
5483 count is more accurate (but it can never be perfect).
5485 11. The use of "drop" in a nested ACL was not being handled correctly in the
5486 outer ACL. Now, if condition failure induced by the nested "drop" causes
5487 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
5488 or "require"), the connection is dropped.
5490 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
5491 that yield "discard" can now be used with an "accept" or a "discard" verb,
5492 but an error is generated for any others (because I can't see a useful way
5493 to define what should happen).
5495 13. When an ACL is read dynamically from a file (or anywhere else), the lines
5496 are now processed in the same way as lines in the Exim configuration file.
5497 In particular, continuation lines are supported.
5499 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
5501 15. Added -ti meaning -t -i.
5503 16. Check for letters, digits, hyphens, and dots in the names of dnslist
5504 domains, and warn by logging if others are found.
5506 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
5507 returned from GIFCONF when Exim is trying to find the list of interfaces on
5508 a host. The code in os.c has been modified to copy each ifreq to an aligned
5509 structure in all cases.
5511 Also, in some cases, the returned ifreq's were being copied to a 'struct
5512 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
5513 means the last couple of bytes of an IPv6 address could be chopped if the
5514 ifreq contained only a normal sockaddr (14 bytes storage).
5516 18. Named domain lists were not supported in the hosts_treat_as_local option.
5517 An entry such as +xxxx was not recognized, and was treated as a literal
5520 19. Ensure that header lines added by a DATA ACL are included in the reject log
5521 if the ACL subsequently rejects the message.
5523 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
5524 MD5 (which is deprecated).
5526 21. When testing a filter file using -bf, Exim was writing a message when it
5527 took the sender from a "From " line in the message, but it was not doing so
5528 when it took $return_path from a Return-Path: header line. It now does.
5530 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
5531 with a valid header line field name (a series of printing characters
5532 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
5534 23. Changed "disc" in the source to "disk" to conform to the documentation and
5535 the book and for uniformity.
5537 24. Ignore Sendmail's -Ooption=value command line item.
5539 25. When execve() failed while trying to run a command in a pipe transport,
5540 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
5541 could be confused with a return value of 69 from the command itself. This
5542 has been changed to 127, the value the shell returns if it is asked to run
5543 a non-existent command. The wording for the related log line suggests a
5544 non-existent command as the problem.
5546 26. If received_header_text expands to an empty string, do not add a Received:
5547 header line to the message. (Well, it adds a token one on the spool, but
5548 marks it "old" so that it doesn't get used or transmitted.)
5550 27. Installed eximstats 1.28 (addition of -nt option).
5552 28. There was no check for failure on the call to getsockname() in the daemon
5553 code. This can fail if there is a shortage of resources on the system, with
5554 ENOMEM, for example. A temporary error is now given on failure.
5556 29. Contrary to the C standard, it seems that in some environments, the
5557 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
5558 program. Exim now does this explicitly; it affects the formatting of
5559 timestamps using strftime().
5561 30. If exiqsumm was given junk data, it threw up some uninitialized variable
5562 complaints. I've now initialized all the variables, to avoid this.
5564 32. Header lines added by a system filter were not being "seen" during
5565 transport-time rewrites.
5567 33. The info_callback() function passed to OpenSSL is set up with type void
5568 (*)(SSL *, int, int), as described somewhere. However, when calling the
5569 function (actually a macro) that sets it up, the type void(*)() is
5570 expected. I've put in a cast to prevent warnings from picky compilers.
5572 34. If a DNS black list lookup found a CNAME record, but there were no A
5573 records associated with the domain it pointed at, Exim crashed.
5575 35. If a DNS black list lookup returned more than one A record, Exim ignored
5576 all but the first. It now scans all returned addresses if a particular IP
5577 value is being sought. In this situation, the contents of the
5578 $dnslist_value variable are a list of all the addresses, separated by a
5581 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
5582 to accept a host name and all its aliases if the forward lookup for any of
5583 them yielded the IP address of the incoming connection. Now it accepts only
5584 those names whose forward lookup yields the correct IP address. Any other
5585 names are discarded. This closes a loophole whereby a rogue DNS
5586 administrator could create reverse DNS records to break through a
5587 wildcarded host restriction in an ACL.
5589 37. If a user filter or a system filter that ran in a subprocess used any of
5590 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
5591 the wrong values were passed to the pipe command ($thisaddress had the
5592 value of $0, $0 had the value of $1, etc). This bug was introduced by
5593 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
5595 38. Improved the line breaking for long SMTP error messages from ACLs.
5596 Previously, if there was no break point between 40 and 75 characters, Exim
5597 left the rest of the message alone. Two changes have been made: (a) I've
5598 reduced the minimum length to 35 characters; (b) if it can't find a break
5599 point between 35 and 75 characters, it looks ahead and uses the first one
5600 that it finds. This may give the occasional overlong line, but at least the
5601 remaining text gets split now.
5603 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
5604 file descriptors might be low, and that setting 1000 would always raise it.
5605 It turns out that in some environments, the limit is already over 1000 and
5606 that lowering it causes trouble. So now Exim takes care not to decrease it.
5608 40. When delivering a message, the value of $return_path is set to $sender_
5609 address at the start of routing (routers may change the value). By an
5610 oversight, this default was not being set up when an address was tested by
5611 -bt or -bv, which affected the outcome if any router or filter referred to
5614 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
5615 log only when "message" or "log_message" are set. However, if one of the
5616 conditions was an address verification, or a call to a nested ACL, the
5617 messages generated by the underlying test were being passed through. This
5618 no longer happens. The underlying message is available in $acl_verify_
5619 message for both "message" and "log_message" expansions, so it can be
5620 passed through if needed.
5622 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
5623 new expansion $bh_ to give the encoded byte string without charset
5624 translation. Translation happens only if iconv() is available; HAVE_ICONV
5625 indicates this at build time. HEADERS_CHARSET gives the charset to
5626 translate to; headers_charset can change it in the configuration, and
5627 "headers charset" can change it in an individual filter file.
5629 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
5630 that creates RFC 2047 encoded "words" labels them as that charset instead
5631 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
5632 expansion operator; (ii) when Exim creates a From: line for a local
5633 message; (iii) when a header line is rewritten to include a "phrase" part.
5635 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
5636 buggy, causing it to skip the first lines of messages whose message ID
5637 ended in 'D'. This would not have bitten before Exim release 4.14, because
5638 message IDs were unlikely to end in 'D' before then. The effect was to have
5639 incorrect size information for certain domains.
5641 45. #include "config.h" was missing at the start of the crypt16.c module. This
5642 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
5645 46. If there was a timeout during a "random" callout check, Exim treated it as
5646 a failure of the random address, and carried on sending RSET and the real
5647 address. If the delay was just some slowness somewhere, the response to the
5648 original RCPT would be taken as a response to RSET and so on, causing
5649 mayhem of various kinds.
5651 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
5652 when I implemented it. It didn't allow for the fact that some option values
5653 may legitimatetly be negative (e.g. size_addition), and it didn't even do
5654 the right test for positive values.
5656 48. Domain names in DNS records are case-independent. Exim always looks them up
5657 in lower case. Some resolvers return domain names in exactly the case they
5658 appear in the zone file, that is, they may contain uppercase letters. Not
5659 all resolvers do this - some return always lower case. Exim was treating a
5660 change of case by a resolver as a change of domain, similar to a widening
5661 of a domain abbreviation. This triggered its re-routing code and so it was
5662 trying to route what was effectively the same domain again. This normally
5663 caused routing to fail (because the router wouldn't handle the domain
5664 twice). Now Exim checks for this case specially, and just changes the
5665 casing of the domain that it ultimately uses when it transmits the message
5668 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
5671 50. If a filter generated a file delivery with a non-absolute name (possible if
5672 no home directory exists for the router), the forbid_file option was not
5675 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
5676 the existing equality matching.
5678 52. Exim was using ints instead of ino_t variables in some places where it was
5679 dealing with inode numbers.
5681 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
5682 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
5683 called TMPDIR, and if it finds it is different, it changes its value.
5685 54. The smtp_printf() function is now made available to local_scan() so
5686 additional output lines can be written before returning. There is also an
5687 smtp_fflush() function to enable the detection of a dropped connection.
5688 The variables smtp_input and smtp_batched_input are exported to
5691 55. Changed the default runtime configuration: the message "Unknown user"
5692 has been removed from the ACL, and instead placed on the localuser router,
5693 using the cannot_route_message feature. This means that any verification
5694 failures that generate their own messages won't get overridden. Similarly,
5695 the "Unrouteable address" message that was in the ACL for unverifiable
5696 relay addresses has also been removed.
5698 56. Added hosts_avoid_esmtp to the smtp transport.
5700 57. The exicyclog script was not checking for the esoteric option
5701 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
5702 will work only if exicyclog is run under the appropriate euid.
5704 58. Following a discussion on the list, the rules by which Exim recognises line
5705 endings on incoming messages have been changed. The -dropcr and drop_cr
5706 options are now no-ops, retained only for backwards compatibility. The
5707 following line terminators are recognized: LF CRLF CR. However, special
5708 processing applies to CR:
5710 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
5711 nor a local message in the state where . is a terminator.
5713 (ii) If a bare CR is encountered in a header line, an extra space is added
5714 after the line terminator so as not to end the header. The reasoning
5715 behind this is that bare CRs in header lines are most likely either
5716 to be mistakes, or people trying to play silly games.
5718 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
5719 was being incorrectly given as 18 bytes larger than it should have been.
5720 This is a VOB (very old bug).
5722 60. This may never have affected anything current, but just in case it has:
5723 When the local host is found other than at the start of a list of hosts,
5724 the local host, those with the same MX, and any that follow, are discarded.
5725 When the list in question was part of a longer list of hosts, the following
5726 hosts (not currently being processed) were also being discarded. This no
5727 longer happens. I'm not sure if this situation could ever has previously
5730 61. Added the "/MX" feature to lists of hosts in the manualroute and query
5733 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
5734 header. This is something that is recommended in a new Internet Draft, and
5735 is something that is documented as being done by Sendmail. There are two
5736 possible values. For messages generated by the autoreply transport, Exim
5739 Auto-Submitted: auto-replied
5741 whereas for all other generated messages (e.g. bounces) it adds
5743 Auto-Submitted: auto-generated
5745 63. The "personal" condition in filters now includes a test for the
5746 Auto-Submitted: header. If it contains the string "auto-" the message it
5747 not considered personal.
5749 64. Added rcpt_include_affixes as a generic transport option.
5751 65. Added queue_only_override (default true).
5753 66. Added the syslog_duplication option.
5755 67. If what should have been the first header line of a message consisted of
5756 a space followed by a colon, Exim was mis-interpreting it as a header line.
5757 It isn't of course - it is syntactically invalid and should therefore be
5758 treated as the start of the message body. The misbehaviour could have
5759 caused a number of strange effects, including loss of data in subsequent
5760 header lines, and spool format errors.
5762 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
5763 client host had authenticated. This control can now be exercised by an ACL
5764 for more flexibility.
5766 69. By default, callouts do not happen when testing with -bh. There is now a
5767 variant, -bhc, which does actually run the callout code, including
5768 consulting and updating the callout cache.
5770 70. Added support for saslauthd authentication, courtesy of Alexander
5773 71. If statvfs() failed on the spool or log directories while checking their
5774 size for availability, Exim confusingly gave the error "space shortage".
5775 Furthermore, in debugging mode it crashed with a floating point exception.
5776 These checks are done if check_{spool,log}_{space,inodes} are set, and when
5777 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
5778 serious problem, Exim now writes to the main and panic logs when this
5779 happens, with details of the failure. It then refuses to accept the
5780 incoming message, giving the message "spool directory problem" or "log
5781 directory problem" with a 421 code for SMTP messages.
5783 72. When Exim is about to re-exec itself, it ensures that the file descriptors
5784 0, 1, and 2 exist, because some OS complain for execs without them (see
5785 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
5786 descriptors. However, the code omitted to check that the open succeeded,
5787 causing mysterious errors if for some reason the permissions on /dev/null
5788 got screwed. Now Exim writes a message to the main and panic logs, and
5789 bombs out if it can't open /dev/null.
5791 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
5792 interact so that it is all more flexible. It is supposed to remain
5793 backwards compatible. Also added extra_local_interfaces.
5795 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
5796 to bomb out with an assertion failure - to the client this appears as a
5797 connection drop. This problem occurs in the part of the code that was taken
5798 from the Samba project. Fortunately, the assertion is in a very simple
5799 function, so I have fixed this by reproducing the function inline in the
5800 one place where it is called, and arranging for authentication to fail
5801 instead of killing the process with assert().
5803 75. The SPA client code was not working when the server requested OEM rather
5804 than Unicode encoding.
5806 76. Added code to make require_files with a specific uid setting more usable in
5807 the case where statting the file as root fails - usually a non-root-mounted
5808 NFS file system. When this happens and the failure is EACCES, Exim now
5809 forks a subprocess and does the per-uid checking as the relevant uid.
5811 77. Added process_log_path.
5813 78. If log_file_path was not explicitly set, a setting of check_log_space or
5814 check_log_inodes was ignored.
5816 79. If a space check for the spool or log partitions fails, the incident is now
5817 logged. Of course, in the latter case the data may get lost...
5819 80. Added the %p formatting code to string_format() so that it can be used to
5820 print addresses in debug_print(). Adjusted all the address printing in the
5821 debugging in store.c to use %p rather than %d.
5823 81. There was a concern that a line of code in smtp_in.c could overflow a
5824 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
5825 initially expressed, the concern was not well-founded, because trailing
5826 spaces are removed early. However, if the trailing spaces were followed by
5827 a NULL, they did not get removed, so the overflow was possible. Two fixes
5830 (a) I re-wrote the offending code in a cleaner fashion.
5831 (b) If an incoming SMTP command contains a NULL character, it is rejected
5834 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
5835 runs initgroups(), so that if the Exim user is in any additional groups,
5836 they will be used during message reception.
5842 The change log for 4.20 and earlier releases has been archived.