1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.481 2007/02/16 22:23:35 magnus Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
10 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
13 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
16 PH/02 In an ACL statement such as
18 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
20 if a client was not listed at all, or was listed with a value other than
21 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
22 the condition was not true (as it should be), so access was not denied.
23 The bug was that the ! inversion was incorrectly passed on to the second
24 item. This has been fixed.
26 PH/03 Added additional dnslists conditions == and =& which are different from
27 = and & when the dns lookup returns more than one IP address.
29 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
30 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
32 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
33 FSYNC, which compiles an option called disable_fsync that allows for
34 bypassing fsync(). The documentation is heavily laced with warnings.
36 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
38 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
39 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
40 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
41 including adding "make clean"; (3) Added -fPIC when compiling the test
42 dynamically loaded module, to get rid of a warning.
44 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
45 message fails, move_frozen_messages = true and ignore_bounce_errors_after
46 = 0s. The bug is otherwise harmless.
48 PH/07 There was a bug in the dovecot authenticator such that the value of
49 $auth1 could be overwritten, and so not correctly preserved, after a
50 successful authentication. This usually meant that the value preserved by
51 the server_setid option was incorrect.
53 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
55 PH/09 Installed PCRE release 7.0.
57 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
58 run for batched SMTP input. It is now run at the start of every message
59 in the batch. While fixing this I discovered that the process information
60 (output by running exiwhat) was not always getting set for -bs and -bS
61 input. This is fixed, and it now also says "batched" for BSMTP.
63 PH/11 Added control=no_pipelining.
65 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
66 patch, slightly modified), and move the expansion of helo_data till after
67 the connection is made in the smtp transport (so it can use these
70 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
72 PH/14 Added log_selector = +pid.
74 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
76 PH/16 Add ${if forany and ${if forall.
78 PH/17 Added dsn_from option to vary the From: line in DSNs.
80 PH/18 Flush SMTP output before performing a callout, unless control =
81 no_callout_flush is set.
83 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
84 was true (the default) a successful delivery failed to delete the retry
85 item, thus causing premature timeout of the address. The bug is now
88 PH/20 Added hosts_avoid_pipelining to the smtp transport.
90 PH/21 Long custom messages for fakedefer and fakereject are now split up
91 into multiline reponses in the same way that messages for "deny" and
92 other ACL rejections are.
94 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
95 with slight modification.
97 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
98 draft, changing the syntax and factoring some duplicate code".
100 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
101 for deliveries of the second and subsequent messages over the same SMTP
104 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
105 ${reduce, with only minor "tidies".
107 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
109 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
110 expansion side effects.
112 MH/03 Made $recipients available in local_scan(). local_scan() already has
113 better access to the recipient list through recipients_list[], but
114 $recipients can be useful in postmaster-provided expansion strings.
120 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
121 fixed by 4.65/MH/01 (is this a record?) are fixed:
123 (i) An empty string was always treated as zero by the numeric comparison
124 operators. This behaviour has been restored.
126 (ii) It is documented that the numeric comparison operators always treat
127 their arguments as decimal numbers. This was broken in that numbers
128 starting with 0 were being interpreted as octal.
130 While fixing these problems I realized that there was another issue that
131 hadn't been noticed. Values of message_size_limit (both the global option
132 and the transport option) were treated as octal if they started with 0.
133 The documentation was vague. These values are now always treated as
134 decimal, and I will make that clear in the documentation.
140 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
141 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
144 MH/01 Don't check that the operands of numeric comparison operators are
145 integers when their expansion is in "skipping" mode (fixes bug
146 introduced by 4.64-PH/07).
148 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
149 child addresses, Exim now panics and dies. Previously, because the count
150 is held in a short int, deliveries were likely to be lost. As such a
151 large number of recipients for a single message is ridiculous
152 (performance will be very, very poor), I have chosen to impose a limit
153 rather than extend the field.
159 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
160 leftover -K file (the existence of which was triggered by #402).
161 While we were at it, introduced process PID as part of the -K
162 filename. This should rule out race conditions when creating
165 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
166 processing considerably. Previous code took too long for large mails,
167 triggering a timeout which in turn triggers #401.
169 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
170 in the DK code in transports.c. sendfile() is not really portable,
171 hence the _LINUX specificness.
173 TF/01 In the add_headers option to the mail command in an Exim filter,
174 there was a bug that Exim would claim a syntax error in any
175 header after the first one which had an odd number of characters
178 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
179 callout verification, Exim cached a "reject" for the entire domain. This
180 is correct for most verifications, but it is not correct for a recipient
181 verification with use_sender or use_postmaster set, because in that case
182 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
183 case of MAIL FROM:<> rejection from other early rejections (e.g.
184 rejection of HELO). When verifying a recipient using a non-null MAIL
185 address, the cache is ignored if it shows MAIL FROM:<> rejection.
186 Whatever the result of the callout, the value of the domain cache is
187 left unchanged (for any other kind of callout, getting as far as trying
188 RCPT means that the domain itself is ok).
190 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
193 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
194 manifest itself as EPIPE rather than ECONNECT. When tidying away a
195 session, the daemon ignores ECONNECT errors and logs others; it now
196 ignores EPIPE as well.
198 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
199 (quoted-printable decoding).
201 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
202 later the small subsequent patch to fix an introduced bug.
204 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
206 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
208 PH/08 An error is now given if message_size_limit is specified negative.
210 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
211 to be given (somewhat) arbitrary names.
213 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
216 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
217 miscellaneous code fixes
219 PH/10 Added the log_reject_target ACL modifier to specify where to log
222 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
223 hostname. This is wrong, because it relates to the incoming message (and
224 probably the interface on which it is arriving) and not to the outgoing
225 callout (which could be using a different interface). This has been
226 changed to use the value of the helo_data option from the smtp transport
227 instead - this is what is used when a message is actually being sent. If
228 there is no remote transport (possible with a router that sets up host
229 addresses), $smtp_active_hostname is used.
231 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
232 tweaks were necessary in order to get it to work (see also 21 below):
233 (a) The code assumed that strncpy() returns a negative number on buffer
234 overflow, which isn't the case. Replaced with Exim's string_format()
236 (b) There were several signed/unsigned issues. I just did the minimum
237 hacking in of casts. There is scope for a larger refactoring.
238 (c) The code used strcasecmp() which is not a standard C function.
239 Replaced with Exim's strcmpic() function.
240 (d) The code set only $1; it now sets $auth1 as well.
241 (e) A simple test gave the error "authentication client didn't specify
242 service in request". It would seem that Dovecot has changed its
243 interface. Fortunately there's a specification; I followed it and
244 changed what the client sends and it appears to be working now.
246 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
249 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
250 address A is aliased to B and C, where B exists and C does not. Without
251 -v the output is "A verified" because verification stops after a
252 successful redirection if more than one address is generated. However,
253 with -v the child addresses are also verified. Exim was outputting "A
254 failed to verify" and then showing the successful verification for C,
255 with its parentage. It now outputs "B failed to verify", showing B's
256 parentage before showing the successful verification of C.
258 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
259 look up a TXT record in a specific list after matching in a combined
262 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
263 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
264 they consult the DNS. I had assumed they would set it the way they
265 wanted; and indeed my experiments on Linux seem to show that in some
266 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
267 To be on the safe side, however, I have now made the interface to
268 host_find_byname() similar to host_find_bydns(), with an argument
269 containing the DNS resolver options. The host_find_byname() function now
270 sets these options at its start, just as host_find_bydns() does. The smtp
271 transport options dns_qualify_single and dns_search_parents are passed to
272 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
273 of host_find_byname() use the default settings of RES_DEFNAMES
274 (qualify_single) but not RES_DNSRCH (search_parents).
276 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
277 spool_read_header() do less string testing, by means of a preliminary
278 switch on the second character of optional "-foo" lines. (This is
279 overdue, caused by the large number of possibilities that now exist.
280 Originally there were few.) While I was there, I also converted the
281 str(n)cmp tests so they don't re-test the leading "-" and the first
282 character, in the hope this might squeeze out yet more improvement.
284 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
285 flag allowing group syntax was set by the header_syntax check but not
286 turned off, possible causing trouble later; (2) The flag was not being
287 set at all for the header_verify test, causing "group"-style headers to
288 be rejected. I have now set it in this case, and also caused header_
289 verify to ignore an empty address taken from a group. While doing this, I
290 came across some other cases where the code for allowing group syntax
291 while scanning a header line wasn't quite right (mostly, not resetting
292 the flag correctly in the right place). These bugs could have caused
293 trouble for malformed header lines. I hope it is now all correct.
295 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
296 with the "reply" argument non-NULL. The code, however (which originally
297 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
298 but it didn't always do it. This confused somebody who was copying the
299 code for some other use. I have removed all the tests.
301 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
302 feature that was used to support insecure browsers during the U.S. crypto
303 embargo. It requires special client support, and Exim is probably the
304 only MTA that supported it -- and would never use it because real RSA is
305 always available. This code has been removed, because it had the bad
306 effect of slowing Exim down by computing (never used) parameters for the
307 RSA_EXPORT functionality.
309 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
310 authenticator to fail if there's a tab character in the incoming data
311 (there should never be unless someone is messing about, as it's supposed
312 to be base64-encoded). Also added, on Timo's advice, the "secured" option
313 if the connection is using TLS or if the remote IP is the same as the
314 local IP, and the "valid-client-cert option" if a client certificate has
317 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
318 authenticators. This can be used for authorization after authentication
319 succeeds. (In the case of plaintext, it servers for both authentication
322 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
323 if any retry times were supplied.
325 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
326 connection, where there is no client IP address to check. In this
327 situation, the verify now always succeeds.
329 PH/25 Applied John Jetmore's -Mset patch.
331 PH/26 Added -bem to be like -Mset, but loading a message from a file.
333 PH/27 In a string expansion for a processed (not raw) header when multiple
334 headers of the same name were present, leading whitespace was being
335 removed from all of them, but trailing whitespace was being removed only
336 from the last one. Now trailing whitespace is removed from each header
337 before concatenation. Completely empty headers in a concatenation (as
340 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
341 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
343 PH/29 [Removed. This was a change that I later backed out, and forgot to
344 correct the ChangeLog entry (that I had efficiently created) before
345 committing the later change.]
347 PH/30 Exim was sometimes attempting to deliver messages that had suffered
348 address errors (4xx response to RCPT) over the same connection as other
349 messages routed to the same hosts. Such deliveries are always "forced",
350 so retry times are not inspected. This resulted in far too many retries
351 for the affected addresses. The effect occurred only when there were more
352 hosts than the hosts_max_try setting in the smtp transport when it had
353 the 4xx errors. Those hosts that it had tried were not added to the list
354 of hosts for which the message was waiting, so if all were tried, there
355 was no problem. Two fixes have been applied:
357 (i) If there are any address or message errors in an SMTP delivery, none
358 of the hosts (tried or untried) are now added to the list of hosts
359 for which the message is waiting, so the message should not be a
360 candidate for sending over the same connection that was used for a
361 successful delivery of some other message. This seems entirely
362 reasonable: after all the message is NOT "waiting for some host".
363 This is so "obvious" that I'm not sure why it wasn't done
364 previously. Hope I haven't missed anything, but it can't do any
365 harm, as the worst effect is to miss an optimization.
367 (ii) If, despite (i), such a delivery is accidentally attempted, the
368 routing retry time is respected, so at least it doesn't keep
369 hammering the server.
371 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
372 in ${readsocket because some servers need this prod.
374 PH/32 Added some extra debug output when updating a wait-xxx database.
376 PH/33 The hint "could be header name not terminated by colon", which has been
377 given for certain expansion errors for a long time, was not being given
378 for the ${if def:h_colon_omitted{... case.
380 PH/34 The spec says: "With one important exception, whenever a domain list is
381 being scanned, $domain contains the subject domain." There was at least
382 one case where this was not true.
384 PH/35 The error "getsockname() failed: connection reset by peer" was being
385 written to the panic log as well as the main log, but it isn't really
386 panic-worthy as it just means the connection died rather early on. I have
387 removed the panic log writing for the ECONNRESET error when getsockname()
390 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
391 runs only) independently of the message's sender address. This meant
392 that, if the 4xx error was in fact related to the sender, a different
393 message to the same recipient with a different sender could confuse
394 things. In particualar, this can happen when sending to a greylisting
395 server, but other circumstances could also provoke similar problems.
396 I have changed the default so that the retry time for these errors is now
397 based a combination of the sender and recipient addresses. This change
398 can be overridden by setting address_retry_include_sender=false in the
401 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
402 remote server are returned as part of bounce messages. This was not
403 happening for LMTP over a pipe (the lmtp transport), but now it is the
404 same for both kinds of LMTP.
406 PH/38 Despite being documented as not happening, Exim was rewriting addresses
407 in header lines that were in fact CNAMEs. This is no longer the case.
409 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
410 and queue runs started by the daemon processed all messages. This has
411 been fixed so that -R and -S can now usefully be given with -q<time>.
413 PH/40 Import PCRE release 6.7 (fixes some bugs).
415 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
417 PH/42 Give an error if -q is specified more than once.
419 PH/43 Renamed the variables $interface_address and $interface_port as
420 $received_ip_address and $received_port, to make it clear that these
421 values apply to message reception, and not to the outgoing interface when
422 a message is delivered. (The old names remain recognized, of course.)
424 PH/44 There was no timeout on the connect() call when using a Unix domain
425 socket in the ${readsocket expansion. There now is.
427 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
428 be meaningful with "accept".
430 SC/01 Eximstats V1.43
431 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
433 SC/02 Eximstats V1.44
434 Use a glob alias rather than an array ref in the generated
435 parser. This improves both readability and performance.
437 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
438 Collect SpamAssassin and rejection statistics.
439 Don't display local sender or destination tables unless
440 there is data to show.
441 Added average volumes into the top table text output.
443 SC/04 Eximstats V1.46
444 Collect data on the number of addresses (recipients)
445 as well as the number of messages.
447 SC/05 Eximstats V1.47
448 Added 'Message too big' to the list of mail rejection
449 reasons (thanks to Marco Gaiarin).
451 SC/06 Eximstats V1.48
452 Mainlog lines which have GMT offsets and are too short to
453 have a flag are now skipped.
455 SC/07 Eximstats V1.49 (Alain Williams)
456 Added the -emptyok flag.
458 SC/08 Eximstats V1.50
459 Fixes for obtaining the IP address from reject messages.
461 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
462 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
463 whitesspace changes from 4.64-PH/27
465 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
468 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
469 are found, allow negative numbers in numeric criteria)
471 JJ/06 exipick.20061117.2, added new $message_body_missing variable
473 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
474 to match changes made in 4.64-PH/43
476 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
478 PH/47 Put in an explicit test for a DNS lookup of an address record where the
479 "domain" is actually an IP address, and force a failure. This locks out
480 those revolvers/nameservers that support "A-for-A" lookups, in
481 contravention of the specifications.
483 PH/48 When a host name was looked up from an IP address, and the subsequent
484 forward lookup of the name timed out, the host name was left in
485 $sender_host_name, contrary to the specification.
487 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
488 restricted to single-key lookups, Exim was not diagnosing an error if
489 * or *@ was used with a query-style lookup.
491 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
493 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
494 long ago, but noone interested enough thought of it. Let's just say that
495 the "1.1" means that there are some new functions that weren't there at
496 some point in the past.
498 PH/51 Error processing for expansion failure of helo_data from an smtp
499 transport during callout processing was broken.
501 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
502 tested/used via the -bh/-bhc/-bs options.
504 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
505 bug, fixed in subsequent PCRE releases).
507 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
508 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
510 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
516 SC/01 Use a glob alias rather than an array ref in eximstats generated
517 parser. This improves both readability and performance.
519 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
520 Don't display local sender or destination tables in eximstats unless
521 there is data to show.
522 Added average volumes into the eximstats top table text output.
524 SC/03 Collect data on the number of addresses (recipients) as well
525 as the number of messages in eximstats.
527 TF/01 Correct an error in the documentation for the redirect router. Exim
528 does (usually) call initgroups() when daemonizing.
530 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
531 with consistent privilege compared to when running as a daemon.
533 TF/03 Note in the spec that $authenticated_id is not set for local
534 submissions from trusted users.
536 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
537 Thanks to Dean Brooks <dean@iglou.com> for the patch.
539 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
540 by adding some example configuration directives to the default
541 configuration file. A little bit of work is required to uncomment the
542 directives and define how usernames and passwords are checked, but
543 there is now a framework to start from.
545 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
546 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
547 without this. I don't know how relevant this is to other LDAP libraries.
549 PH/02 Add the verb name to the "unknown ACL verb" error.
551 PH/03 Magnus Holmgren's patch for filter_prepend_home.
553 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
555 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
556 directory not expanded when it should be if an expanded home directory
557 was set for the address (which is overridden by the transport).
559 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
562 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
563 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
564 because it is too late at that time, and has no effect.
566 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
567 security issue with \' (bugzilla #107). I could not use the
568 PQescapeStringConn() function, because it needs a PGconn value as one of
571 PH/08 When testing addresses using -bt, indicate those final addresses that
572 are duplicates that would not cause an additional delivery. At least one
573 person was confused, thinking that -bt output corresponded to deliveries.
574 (Suppressing duplicates isn't a good idea as you lose the information
575 about possibly different redirections that led to the duplicates.)
577 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
578 systems where poll() doesn't work, in particular OS X.
580 PH/10 Added more information to debugging output for retry time not reached.
582 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
583 operations in malware.c.
585 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
588 PH/13 If write_rejectlog was set false when logging was sent to syslog with
589 syslog_duplication set false, log lines that would normally be written
590 both the the main log and to the reject log were not written to syslog at
593 PH/14 In the default configuration, change the use of "message" in ACL warn
594 statements to "add_header".
596 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
597 not followed by a command (e.g. "seen endif").
599 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
600 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
603 PH/17 Added extra conditions to the default value of delay_warning_condition
607 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
608 { match{$h_precedence:}{(?i)bulk|list|junk} } \
609 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
612 The Auto-Submitted: and various List- headers are standardised, whereas I
613 don't think Precedence: ever was.
615 PH/18 Refactored debugging code in route_finduser() to show more information,
616 in particular, the error code if getpwnam() issues one.
618 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
619 This is apparently needed in addition to the PH/07 change above to avoid
620 any possible encoding problems.
622 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
623 but not after initializing Perl.
625 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
626 output them only if debugging. By default they are written stderr,
627 apparently, which is not desirable.
629 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
632 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
635 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
637 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
638 authenticated or an ident call has been made. Suppress the default
639 values for $authenticated_id and $authenticated_sender (but permit -oMai
640 and -oMas) when testing with -bh.
642 PH/24 Re-jigged the order of the tests in the default configuration so that the
643 tests for valid domains and recipients precede the DNS black list and CSA
644 tests, on the grounds that those ones are more expensive.
646 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
647 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
648 This bug exists in every version of Exim that I still have, right back to
651 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
652 However, an attempt to turn on case-sensitivity in a regex key by
653 including (?-i) didn't work because the subject string was already
654 lowercased, and the effects were non-intuitive. It turns out that a
655 one-line patch can be used to allow (?-i) to work as expected.
661 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
662 other effects) broke the use of negated acl sub-conditions.
664 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
667 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
668 "Deny" causes Exim to reject the incoming connection with a 554 error.
669 Unfortunately, if there is a major crisis, such as a disk failure,
670 tcp-wrappers gives "deny", whereas what one would like would be some
671 kind of temporary error. A kludge has been added to help with this.
672 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
673 554 error is used if errno is still zero or contains ENOENT (which occurs
674 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
677 PH/03 Add -lutil to the default FreeBSD LIBS setting.
679 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
680 errors. Otherwise a message that provokes a temporary error (when other
681 messages do not) can cause a whole host to time out.
683 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
684 addresses were routed directly to files or pipes from a redirect router.
685 File deliveries just didn't batch; pipe deliveries might have suffered
688 PH/06 A failure to get a lock for a hints database would erroneously always say
689 "Failed to get write lock", even when it was really a read lock.
691 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
692 of 0600. This has been changed to use the value of the lockfile_mode
693 option (which defaults to 0600).
695 PH/08 Applied small patch from the Sieve maintainer.
697 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
698 folder from quota calculations, a direct delivery into this folder messed
699 up the contents of the maildirsize file. This was because the regex was
700 used only to exclude .Trash (or whatever) when the size of the mailbox
701 was calculated. There was no check that a delivery was happening into an
702 excluded directory. This bug has been fixed by ignoring all quota
703 processing for deliveries into excluded directories.
705 PH/10 Added the maildirfolder_create_regex option to appendfile.
711 PH/01 The code for finding all the local interface addresses on a FreeBSD
712 system running IPv6 was broken. This may well have applied to all BSD
713 systems, as well as to others that have similar system calls. The broken
714 code found IPv4 interfaces correctly, but gave incorrect values for the
715 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
716 that it would not match correctly against @[] and not recognize the IPv6
719 PH/02 The ipliteral router was not recognizing addresses of the form user@
720 [ipv6:....] because it didn't know about the "ipv6:" prefix.
722 PH/03 Added disable_ipv6.
724 PH/04 Changed $reply_address to use the raw form of the headers instead of the
725 decoded form, because it is most often used to construct To: headers
726 lines in autoreplies, and the decoded form may well be syntactically
727 invalid. However, $reply_address has leading white space removed, and all
728 newlines turned into spaces so that the autoreply transport does not
731 PH/05 If group was specified without a user on a router, and no group or user
732 was specified on a transport, the group from the router was ignored.
734 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
735 for visible compile-time settings that can be used to change these
736 numbers, for those that want even more. Backwards compatibility with old
737 spool files has been maintained. However, going back to a previous Exim
738 release will lost any variables that are in spool files.
740 PH/07 Two small changes when running in the test harness: increase delay when
741 passing a TCP/IP connection to a new process, in case the original
742 process has to generate a bounce, and remove special handling of
743 127.0.0.2 (sic), which is no longer necessary.
745 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
746 be the same on different OS.
748 PH/09 Moved a debug statement in filter processing to avoid a race problem when
751 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
752 whether --show-vars was specified or not
754 JJ/02 exipick: Added support for new ACL variable spool format introduced
757 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
758 syntactically invalid From: or Reply-to: line, and a filter used this to
759 generate an autoreply, and therefore failed to obtain an address for the
760 autoreply, Exim could try to deliver to a non-existent relative file
761 name, causing unrelated and misleading errors. What now happens is that
762 it logs this as a hard delivery error, but does not attempt to create a
765 PH/11 The exinext utility has a -C option for testing purposes, but although
766 the given file was scanned by exinext itself; it wasn't being passed on
767 when Exim was called.
769 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
770 an end-of-file indication when reading a command response.
772 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
773 compiled. In many other places in Exim, IPv6 addresses are always
774 recognized, so I have changed this. It also means that IPv4 domain
775 literals of the form [IPV4:n.n.n.n] are now always recognized.
777 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
778 used if the router is not running as root, for example, when verifying at
779 ACL time, or when using -bh. The debugging output from this situation was
780 non-existent - all you got was a failure to exec. I have made two
783 (a) Failures to set uid/gid, the current directory, or a process leader
784 in a subprocess such as that created by queryprogram now generate
785 suitable debugging ouput when -d is set.
787 (b) The queryprogram router detects when it is not running as root,
788 outputs suitable debugging information if -d is set, and then runs
789 the subprocess without attempting to change uid/gid.
791 PH/15 Minor change to Makefile for building test_host (undocumented testing
794 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
795 additional section of a DNS packet that returns MX or SRV records.
796 Instead, it always explicitly searches for A/AAAA records. This avoids
797 major problems that occur when a DNS server includes only records of one
798 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
799 fixed another bug: if SRV records were looked up and the corresponding
800 address records were *not* found in the additional section, the port
801 values from the SRV records were lost.
803 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
804 using the correct key (the original address) when searching the retry
805 rules in order to find which one to use for generating the retry hint.
807 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
808 adding the default one. Similarly, if it contains a Reply-To: header, the
809 errors_reply_to option, if set, is not used.
811 PH/19 When calculating a retry time, Exim used to measure the "time since
812 failure" by looking at the "first failed" field in the retry record. Now
813 it does not use this if it is later than than the arrival time of the
814 message. Instead it uses the arrival time. This makes for better
815 behaviour in cases where some deliveries succeed, thus re-setting the
816 "first failed" field. An example is a quota failure for a huge message
817 when small messages continue to be delivered. Without this change, the
818 "time since failure" will always be short, possible causing more frequent
819 delivery attempts for the huge message than are intended.
820 [Note: This change was subsequently modified - see PH/04 for 4.62.]
822 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
823 $1, $2, $3) because the numerical variables can be reset during some
824 expansion items (e.g. "match"), thereby losing the authentication data.
826 PH/21 Make -bV show the size of off_t variables so that the test suite can
827 decide whether to run tests for quotas > 2G.
829 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
830 mailbox_size, and mailbox_filecount in the appendfile transport. If a
831 filecount value is greater than 2G or if a quota value is greater than 2G
832 on a system where the size of off_t is not greater than 4, a panic error
835 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
836 never match. The debug and -bh output now contains an explicit error
837 message indicating a malformed IPv4 address or mask.
839 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
840 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
843 PH/25 Do not write to syslog when running in the test harness. The only
844 occasion when this arises is a failure to open the main or panic logs
845 (for which there is an explicit test).
847 PH/26 Added the /no_tell option to "control=freeze".
849 PH/27 If a host name lookup failed very early in a connection, for example, if
850 the IP address matched host_lookup and the reverse lookup yielded a name
851 that did not have a forward lookup, an error message of the form "no IP
852 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
853 could be logged. Now it outputs the IP address instead of "NULL".
855 PH/28 An enabling patch from MH: add new function child_open_exim2() which
856 allows the sender and the authenticated sender to be set when
857 submitting a message from within Exim. Since child_open_exim() is
858 documented for local_scan(), the new function should be too.
860 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
861 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
862 results in an empty string is now treated as unset.
864 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
866 PH/31 Added sender_verify_fail logging option.
868 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
869 needed by RFC 822 but not by RFC 2822 was commented out. I have now
870 tidied the source and removed it altogether.
872 PH/33 When a queue run was abandoned because the load average was too high, a
873 log line was always written; now it is written only if the queue_run log
874 selector is set. In addition, the log line for abandonment now contains
875 information about the queue run such as the pid. This is always present
876 in "start" and "stop" lines but was omitted from the "abandon" line.
878 PH/34 Omit spaces between a header name and the colon in the error message that
879 is given when verify = headers_syntax fails (if there are lots of them,
880 the message gets confusing).
882 PH/35 Change the default for dns_check_names_pattern to allow slashes within
883 names, as there are now some PTR records that contain slashes. This check
884 is only to protect against broken name servers that fall over on strange
885 characters, so the fact that it applies to all lookups doesn't matter.
887 PH/36 Now that the new test suite is complete, we can remove some of the
888 special code in Exim that was needed for the old test suite. For example,
889 sorting DNS records because real resolvers return them in an arbitrary
890 order. The new test suite's fake resolver always returns records in the
893 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
894 bounces) except when queue_only is set, to avoid logging races between
895 the different processes.
897 PH/38 Panic-die if .include specifies a non-absolute path.
899 PH/39 A tweak to the "H" retry rule from its user.
901 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
902 a label. They prevented compilation on older perls.
904 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
905 a warning to be raised on newish perls.
907 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
908 on queue. Changes to match documented behaviour of showing count of
909 messages matching specified criteria.
911 PH/40 Changed the default ident timeout from 30s to 5s.
913 PH/41 Added support for the use of login_cap features, on those BSD systems
914 that have them, for controlling the resources used by pipe deliveries.
916 PH/42 The content-scanning code uses fopen() to create files in which to put
917 message data. Previously it was not paying any attention to the mode of
918 the files. Exim runs with umask(0) because the rest of the code creates
919 files with open(), and sets the required mode explicitly. Thus, these
920 files were ending up world-writeable. This was not a big issue, because,
921 being within the spool directory, they were not world-accessible. I have
922 created a function called modefopen, which takes an additional mode
923 argument. It sets umask(777), creates the file, chmods it to the required
924 mode, then resets the umask. All the relevant calls to fopen() in the
925 content scanning code have been changed to use this function.
927 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
928 to 24 hours. This avoids potential overflow problems when processing G
929 and H retry rules. I suspect nobody ever tinkers with this value.
931 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
933 PH/45 When the plaintext authenticator is running as a client, the server's
934 challenges are checked to ensure they are valid base64 strings. By
935 default, the authentication attempt is cancelled if an invalid string is
936 received. Setting client_ignore_invalid_base64 true ignores these errors.
937 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
938 they are received. Thus, the responses can be made to depend on the
939 challenges. If an invalid string is ignored, an empty string is placed in
942 PH/46 Messages that are created by the autoreply transport now contains a
943 References: header, in accordance with RFCs 2822 and 3834.
945 PH/47 Added authenticated_sender_force to the smtp transport.
947 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
949 PH/49 Installed latest patch from the Sieve maintainer.
951 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
952 was also set, the appendfile transport was unnecessarily scanning a
953 directory of message files (e.g. for maildir delivery) to find the count
954 of files (along with the size), even though it did not need this
955 information. It now does the scan only if it needs to find either the
956 size of the count of files.
958 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
960 PH/52 Two bugs concerned with error handling when the smtp transport is
963 (i) Exim was not creating retry information for temporary errors given
964 for individual recipients after the DATA command when the smtp transport
965 was used in LMTP mode. This meant that they could be retried too
966 frequently, and not timed out correctly.
968 (ii) Exim was setting the flag that allows error details to be returned
969 for LMTP errors on RCPT commands, but not for LMTP errors for individual
970 recipients that were returned after the DATA command.
972 PH/53 This is related to PH/52, but is more general: for any failing address,
973 when detailed error information was permitted to be returned to the
974 sender, but the error was temporary, then after the final timeout, only
975 "retry timeout exceeded" was returned. Now it returns the full error as
976 well as "retry timeout exceeded".
978 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
979 do this, and (what is worse) MTAs that accept it.
981 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
982 will now be deprecated.
984 PH/56 New os.c-cygwin from the Cygwin maintainer.
986 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
987 formats (previously only available in exim formats via -bpr, -bpru,
988 and -bpra. Now also available in native and exiqgrep formats)
990 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
991 with very large, slow to parse queues
993 JJ/08 exipick: added ! as generic prefix to negate any criteria format
995 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
997 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
998 responses to authentication challenges, though it was showing the
999 challenges; (ii) I've removed the CR characters from the debug output for
1002 PH/58 Allow for the insertion of a newline as well as a space when a string
1003 is turned into more than one encoded-word during RFC 2047 encoding. The
1004 Sieve code now uses this.
1006 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1007 data_4xx, lost_connection, tls_required.
1009 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1010 message was being sent as an SMTP response.
1012 PH/61 Add -l and -k options to exicyclog.
1014 PH/62 When verifying, if an address was redirected to one new address, so that
1015 verification continued, and the new address failed or deferred after
1016 having set something in $address_data, the value of $address_data was not
1017 passed back to the ACL. This was different to the case when no
1018 redirection occurred. The value is now passed back in both cases.
1020 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1021 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1022 use login_cap.h, so on its own it isn't the distinguishing feature. The
1023 new name refers directly to the setclassresources() function.
1025 PH/65 Added configuration files for NetBSD3.
1027 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1029 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1030 is preferred over IPv4.
1032 PH/68 The bounce_return_message and bounce_return_body options were not being
1033 honoured for bounces generated during the reception of non-SMTP messages.
1034 In particular, this applied to messages rejected by the ACL. This bug has
1035 been fixed. However, if bounce_return_message is true and bounce_return_
1036 body is false, the headers that are returned for a non-SMTP message
1037 include only those that have been read before the error was detected.
1038 (In the case of an ACL rejection, they have all been read.)
1040 PH/69 The HTML version of the specification is now built in a directory called
1041 spec_html instead of spec.html, because the latter looks like a path with
1042 a MIME-type, and this confuses some software.
1044 PH/70 Catch two compiler warnings in sieve.c.
1046 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1047 function verify_get_ident() calls ip_connect() to connect a socket, but
1048 if the "connect()" function timed out, ip_connect() used to close the
1049 socket. However, verify_get_ident() also closes the socket later, and in
1050 between Exim writes to the log, which may get opened at this point. When
1051 the socket was closed in ip_connect(), the log could get the same file
1052 descriptor number as the socket. This naturally causes chaos. The fix is
1053 not to close the socket in ip_connect(); the socket should be closed by
1054 the function that creates it. There was only one place in the code where
1055 this was missing, in the iplookup router, which I don't think anybody now
1056 uses, but I've fixed it anyway.
1058 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1059 well as to direct DNS lookups. Otherwise the handling of names in host
1060 lists is inconsistent and therefore confusing.
1066 PH/01 Two changes to the default runtime configuration:
1068 (1) Move the checks for relay_from_hosts and authenticated clients from
1069 after to before the (commented out) DNS black list checks.
1071 (2) Add control=submission to the relay_from_hosts and authenticated
1072 clients checks, on the grounds that messages accepted by these
1073 statements are most likely to be submissions.
1075 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1077 (1) Generate an error if the third argument for the ${prvs expansion is
1080 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1083 (3) Reset the variables that are obtained from the first argument of
1084 ${prvscheck and used in the second argument before leaving the code,
1085 because their memory is reclaimed, so using them afterwards may do
1088 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1089 one (it's much easier than Tom thought :-).
1091 (5) Because of (4), we can now allow for the use of $prvscheck_result
1092 inside the third argument.
1094 PH/03 For some reason, the default setting of PATH when running a command from
1095 a pipe transport was just "/usr/bin". I have changed it to
1098 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1099 anything to be listed in the output from -bV.
1101 PH/05 When a filter generated an autoreply, the entire To: header line was
1102 quoted in the delivery log line, like this:
1104 => >A.N.Other <ano@some.domain> <original@ddress> ...
1106 This has been changed so that it extracts the operative address. There
1107 may be more than one such address. If so, they are comma-separated, like
1110 => >ano@some.domain,ona@other.domain <original@ddress> ...
1112 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1113 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1114 not being looked up in the rDNS to get a host name, Exim was showing the
1115 IP address twice in Received: lines, even though the IP addresses were
1116 identical. For example:
1118 Received: from [1.2.3.4] (helo=[1.2.3.4])
1120 However, if the real host name was known, it was omitting the HELO data
1121 if it matched the actual IP address. This has been tidied up so that it
1122 doesn't show the same IP address twice.
1124 PH/07 When both +timestamp and +memory debugging was on, the value given by
1125 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1126 called by the debug printing, thereby overwriting the timestamp buffer.
1127 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1129 PH/08 When the original message was included in an autoreply transport, it
1130 always said "this is a copy of the message, including all the headers",
1131 even if body_only or headers_only was set. It now gives an appropriate
1134 PH/09 Applied a patch from the Sieve maintainer which:
1136 o fixes some comments
1137 o adds the (disabled) notify extension core
1138 o adds some debug output for the result of if/elsif tests
1139 o points to the current vacation draft in the documentation
1140 and documents the missing references header update
1144 o fixes a bug in processing the envelope test (when testing
1145 multiple envelope elements, the last element determinted the
1148 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1149 Electronic Mail") by including:
1151 Auto-submitted: auto-generated
1153 in the messages that it generates (bounce messages and others, such as
1154 warnings). In the case of bounce messages for non-SMTP mesages, there was
1155 also a typo: it was using "Auto_submitted" (underscore instead of
1156 hyphen). Since every message generated by Exim is necessarily in response
1157 to another message, thes have all been changed to:
1159 Auto-Submitted: auto-replied
1161 in accordance with these statements in the RFC:
1163 The auto-replied keyword:
1165 - SHOULD be used on messages sent in direct response to another
1166 message by an automatic process,
1168 - MUST NOT be used on manually-generated messages,
1170 - MAY be used on Delivery Status Notifications (DSNs) and Message
1171 Disposition Notifications (MDNs),
1173 - MUST NOT be used on messages generated by automatic or periodic
1174 processes, except for messages which are automatic responses to
1177 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1178 to the default Received: header definition.
1180 PH/12 Added log selector acl_warn_skipped (default on).
1182 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1183 variables because (a) they are in the wrong storage pool and (b) even if
1184 they were copied, it wouldn't work properly because of the caching.
1186 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1187 checking when decoding. Apparently there are clients that generate
1188 overlong encoded strings. Why am I not surprised?
1190 PH/15 If the first argument of "${if match_address" was not empty, but did not
1191 contain an "@" character, Exim crashed. Now it writes a panic log message
1192 and treats the condition as false.
1194 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1196 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1197 extension "envelope-auth". The code is finished and in agreement with
1198 other implementations, but there is no documentation so far and in fact,
1199 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1200 not changing the active code.
1202 Print executed "if" and "elsif" statements when debugging is used. This
1203 helps a great deal to understand what a filter does.
1205 Document more things not specified clearly in RFC3028. I had all this
1206 sorted out, when out of a sudden new issues came to my mind. Oops."
1208 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1211 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1212 canonical form (as documented). However, after a host name lookup from
1213 the IP address, check_host() was doing a simple string comparison with
1214 addresses acquired from the DNS when checking that the found name did
1215 have the original IP as one of its addresses. Since any found IPv6
1216 addresses are likely to be in abbreviated form, the comparison could
1217 fail. Luckily, there already exists a function for doing the comparison
1218 by converting both addresses to binary, so now that is used instead of
1219 the text comparison.
1221 PH/20 There was another similar case to PH/19, when a complete host name was
1222 given in a host list; looking up its IP address could give an abbreviated
1223 form, whereas the current host's name might or might not be abbreviated.
1224 The same fix has been applied.
1230 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1231 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1234 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1235 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1237 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1239 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1240 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1241 Also, for "Auto-Submitted", treat anything other than "no" as
1242 non-personal, in accordance with RFC 3834. (Previously it treated
1243 anything starting "auto-" as non-personal.)
1245 TF/01 The control=submission/name=... option had a problem with syntax
1246 errors if the name included a slash character. The /name= option
1247 now slurps the rest of the string, so it can include any characters
1248 but it must come last in the list of options (after /sender_retain
1251 PH/05 Some modifications to the interface to the fake nameserver for the new
1259 TK/01 Added the "success_on_redirect" address verification option. See
1260 NewStuff for rationale and an example.
1262 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1264 PH/02 Patch to exigrep to allow it to work on syslog lines.
1266 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1267 fread() to skip over the body file's header line, because in Cygwin the
1268 header line is locked and is inaccessible.
1270 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1271 co-exist for some time) to make it clear that it is the Exim ID that is
1272 referenced, not the Message-ID: header line.
1274 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1275 string_format() function, because snprintf() does not exist on all
1278 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1279 ${stat: expansion item.
1281 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1282 protocol synchronization error", to keep the pedants happy.
1284 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1285 well as for IRIX systems, when gcc is being used. See the host.c source
1288 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1290 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1293 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1294 transport and to the smtp transport in LMTP mode.
1296 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1298 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1299 run a filter in a subprocess. This could lead to confusion in subsequent
1300 lookups in the parent process. There should also be a search_tidyup() at
1301 the end of the subprocess.
1303 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1304 only if the host matched helo_try_verify_hosts, which caused the
1305 verification to occur when the EHLO/HELO command was issued. The ACL just
1306 tested the remembered result. Now, if a previous verification attempt has
1307 not happened, "verify = helo" does it there and then.
1309 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1311 TK/03 Fix log output including CR from clamd.
1313 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1314 exist provoked a memory error which could cause a segfault.
1316 PH/15 Installed PCRE 6.2
1318 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1320 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1321 of the problem. Specifically, suggested +O2 rather than +O1 for the
1324 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1326 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1327 record, the port was not correctly logged when the outgoing_port log
1328 selector was set (it logged the transort's default port).
1330 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1331 fallback_hosts, and "hosts" in the smtp transport.
1333 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1334 host errors such as "Connection refused".
1336 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1337 authentication with radiusclient 0.4.9:
1339 - Error returned from rc_read_config was caught wrongly
1340 - Username/password not passed on to radius server due to wrong length.
1342 The presumption is that some radiusclient API changes for 4.51/PH/17
1343 were not taken care of correctly. The code is still untested by me (my
1344 Linux distribution still has 0.3.2 of radiusclient), but it was
1345 contributed by a Radius user.
1347 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1348 expanding the "port" option of the smtp transport.
1350 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1351 while reading a MIME header. Thanks to Tom Hughes for a patch.
1353 PH/24 Include config.h inside local_scan.h so that configuration settings are
1356 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1357 that in an ACL for RCPT (for example), you can examine exactly what was
1360 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1361 commands, but it was not correctly comparing the address with the actual
1362 client host address. Thus, it would show the EHLO address in Received:
1363 header lines when this was not necessary.
1365 PH/27 Added the % operator to ${eval:}.
1367 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1368 it should be ignoring failures (because with -C, for example, it has lost
1369 privilege). It wasn't ignoring creation failures other than "already
1372 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1375 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1376 because an input line was too long, either on its own, or by virtue of
1377 too many continuations, the temporary file was not being removed, and the
1378 return code was incorrect.
1380 PH/31 Missing "BOOL" in function definition in filtertest.c.
1382 PH/32 Applied Sieve patches from the maintainer.
1384 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1386 PH/33 Added "verify = not_blind".
1388 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1389 Local/Makefile (with some defaults set). These are used in built scripts
1390 such as exicyclog, but they have never been used in the exim_install
1391 script (though there are many overriding facilities there). I have
1392 arranged that the exim_install script now takes note of these two
1395 PH/35 Installed configuration files for Dragonfly.
1397 PH/36 When a locally submitted message by a trusted user did not contain a
1398 From: header, and the sender address was obtained from -f or from an SMTP
1399 MAIL command, and the trusted user did not use -F to supply a sender
1400 name, $originator_name was incorrectly used when constructing a From:
1401 header. Furthermore, $originator_name was used for submission mode
1402 messages from external hosts without From: headers in a similar way,
1403 which is clearly wrong.
1405 PH/37 Added control=suppress_local_fixups.
1407 PH/38 When log_selector = +received_sender was set, and the addition of the
1408 sender made the log line's construction buffer exactly full, or one byte
1409 less than full, an overflow happened when the terminating "\n" was
1412 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1413 when the result of a list match is failure because a DNS lookup failed.
1415 PH/40 RM_COMMAND is now used in the building process.
1417 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1418 the "build-* directories that it finds.
1420 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1421 address in a domain literal was a prefix of an interface address.
1423 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1424 when verifying a sender address, unless rewrite_headers is false.
1426 PH/44 Wrote a long comment about why errors_to addresses are verified as
1427 recipients, not senders.
1429 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1430 the ratelimit ACL was added.
1432 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1434 PH/46 Added extra information about PostgreSQL errors to the error string.
1436 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1437 avoiding the need to install special zones in a real server. This is
1438 backwards compatible; if it can't find the fake resolver, it drops back.
1439 Thus, both old and new test suites can be run.
1441 TF/02 Added util/ratelimit.pl
1443 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1444 clock is set back in time.
1446 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1447 Candler <B.Candler@pobox.com>.
1449 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1450 OK for addresses that are the result of redirections.
1452 PH/48 A number of further additions for the benefit of the new test suite,
1453 including a fake gethostbyname() that interfaces to the fake DNS resolver
1456 TF/06 The fix for widen_domains has also been applied to qualify_single and
1457 search_parents which are the other dnslookup options that can cause
1460 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1463 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1465 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1466 service" if it encountered an MX record with an empty target hostname.
1467 The message is now "an MX or SRV record indicated no SMTP service".
1469 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1470 if the DNS of the sending site is misconfigured. This is quite a
1471 common situation. This change restores the behaviour of treating a
1472 helo verification defer as a failure.
1474 PH/51 If self=fail was set on a router, the bounce message did not include the
1475 actual error message.
1481 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1483 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1484 command itself ended in error, the underlying message about the transport
1485 filter timeout was being overwritten with the pipe command error. Now the
1486 underlying error message should be appended to the second error message.
1488 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1490 PH/02 Reduce the amount of output that "make" produces by default. Full output
1491 can still be requested.
1493 PH/03 The warning log line about a condition test deferring for a "warn" verb
1494 was being output only once per connection, rather than after each
1495 occurrence (because it was using the same function as for successful
1496 "warn" verbs). This seems wrong, so I have changed it.
1498 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1499 it should not have, which might have caused a crash in the right
1500 circumstances, but probably never did.
1502 PH/04 Installed a modified version of Tony Finch's patch to make submission
1503 mode fix the return path as well as the Sender: header line, and to
1504 add a /name= option so that you can make the user's friendly name appear
1507 TF/03 Added the control = fakedefer ACL modifier.
1509 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1510 Mark Lowes for thorough testing.
1512 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1514 TK/03 Merged latest SRS patch from Miles Wilton.
1516 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1517 with the definition in sysexits.h (which is #included earlier).
1518 Fortunately, Exim does not actually use EX_OK. The code used to try to
1519 preserve the sysexits.h value, by assumimg that macro definitions were
1520 scanned for macro replacements. I have been disabused of this notion,
1521 so now the code just undefines EX_OK before #including unistd.h.
1523 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1524 in the smtp transport. When a block could not be written in a single
1525 write() function, the timeout was being re-applied to each part-write.
1526 This seems wrong - if the receiver was accepting one byte at a time it
1527 would take for ever. The timeout is now adjusted when this happens. It
1528 doesn't have to be particularly precise.
1530 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1531 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1533 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1534 without a domain if the check to <postmaster@domain> fails.
1536 SC/01 Eximstats: added -xls and the ability to specify output files
1537 (patch written by Frank Heydlauf).
1539 SC/02 Eximstats: use FileHandles for outputing results.
1541 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1543 SC/04 Eximstats: fixed display of large numbers with -nvr option
1545 SC/05 Eximstats: fixed merging of reports with empty tables.
1547 SC/06 Eximstats: added the -include_original_destination flag
1549 SC/07 Eximstats: removed tabs and trailing whitespace.
1551 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1553 TK/06 MBOX spool code: Add real "From " MBOX separator line
1554 so the .eml file is really in mbox format (even though
1555 most programs do not really care). Patch from Alex Miller.
1557 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1558 The latter is generated from $received_to and is only set if the
1559 message has one envelope recipient. SA can use these headers,
1560 obviously out-of-the-box. Patch from Alex Miller.
1562 PH/08 The ${def test on a variable was returning false if the variable's
1563 value was "0", contrary to what the specification has always said!
1564 The result should be true unless the variable is empty.
1566 PH/09 The syntax error of a character other than { following "${if
1567 def:variable_name" (after optional whitespace) was not being diagnosed.
1568 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1569 accidental colon was present, for example, could give incorrect results.
1571 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1572 result is used (not including appendfile, where other changes are about
1575 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1576 This involved changing a lot of size variables from int to off_t. It
1577 should work with maildirs and everything.
1579 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1580 spamd dying while we are connected to it.
1582 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1585 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1586 function for BDB 4.3.
1588 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1590 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1591 the sources for PCRE have been re-arranged and more files are now
1594 PH/15 The code I had for printing potentially long long variables in PH/11
1595 above was not the best (it lost precision). The length of off_t variables
1596 is now inspected at build time, and an appropriate printing format (%ld
1597 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1598 to be "long long int" or "long int". This is needed for the internal
1599 formatting function string_vformat().
1601 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1602 the configuration file to be ":syslog", then the script "guesses" where
1603 the logs files are, rather than using the compiled in default. In our
1604 case the guess is not the same as the compiled default, so the script
1605 suddenly stopped working when I started to use syslog. The patch checks
1606 to see if log_file_path is "". If so, it attempts to read it from exim
1607 with no configuration file to get the compiled in version, before it
1608 falls back to the previous guessing code."
1610 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1611 implementing BATV in an Exim configuration. See NewStuff for the gory
1614 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1615 Makefile that are specific to HP-UX.
1617 PH/18 If the "use_postmaster" option was set for a recipient callout together
1618 with the "random" option, the postmaster address was used as the MAIL
1619 FROM address for the random test, but not for the subsequent recipient
1620 test. It is now used for both.
1622 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1623 patch removes a few documentation additions to RFC 3028, because the
1624 latest draft now contains them. It adds the new en;ascii-case comparator
1625 and a new error check for 8bit text in MIME parts. Comparator and
1626 require names are now matched exactly. I enabled the subaddress
1627 extension, but it is not well tested yet (read: it works for me)."
1629 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1630 rework some of the code of TK/09 above to avoid the hardwired use of
1631 "%lld" and "long long". Replaced the call to snprintf() with a call to
1634 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1635 records point to non-existent hosts", "retry timeout exceeded", and
1636 "retry time not reached for any host after a long failure period".
1638 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1639 experimental DomainKeys support:
1641 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1642 (2) On an error such as an illegally used "control", the wrong name for
1643 the control was given.
1645 These problems did NOT occur unless DomainKeys support was compiled.
1647 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1649 PH/24 Added ${if match_ip condition.
1651 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1652 databases so that it will be absolutely obvious if a crash occurs in the
1653 DB library. This is a regular occurrence (often caused by mis-matched
1656 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1657 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1658 picked up on a user's system that detects such things. There doesn't seem
1659 to be a gcc warning option for this - only an attribute that has to be
1660 put on the function's prototype. It seems that in Fedora Core 4 they have
1661 set this on a number of new functions. No doubt there will be more in due
1664 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1665 specify a transport. However, if an address that was verified by such a
1666 router was the subject of a callout, Exim crashed because it tried to
1667 read the rcpt_include_affixes from the non-existent transport. Now it
1668 just assumes that the setting of that option is false. This bug was
1669 introduced by 4.51/PH/31.
1671 PH/28 Changed -d+all to exclude +memory, because that information is very
1672 rarely of interest, but it makes the output a lot bigger. People tend to
1673 do -d+all out of habit.
1675 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1676 code in os-type was giving problems when libc.so lives in lib64, like on
1679 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1680 aren't the modern standard, and it seems that some systems' include files
1681 don't always have them. Exim was already checking for some of the newer
1682 ones like T_AAAA, and defining it itself. I've added checks for all the
1683 record types that Exim uses.
1685 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1686 not automatically generating a new one, as it is supposed to. This
1687 prevented TLS from working. If the file did exist, but contained invalid
1688 data, a new version was generated, as expected. It was only the case of a
1689 non-existent file that was broken.
1691 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1692 with a change in libdomainkeys > 0.64.
1694 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1695 from DNS. If the selector record carries the flag, it now has
1696 precedence over the domain-wide flag.
1698 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1700 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1701 the use of a port name that isn't defined in /etc/services) occurred, the
1702 message was deferred as in a normal delivery, and thus remained on the
1703 spool, instead of being failed because of the mua_wrapper setting. This
1704 is now fixed, and I tidied up some of the mua_wrapper messages at the
1707 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1708 the messages in a hash of arrays rather than using individual hashes.
1709 This is a bit cleaner and results in dramatic memory savings, albeit
1710 at a slight CPU cost.
1712 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1713 as requested by Marc Sherman.
1715 SC/10 Eximstats: added histograms for user specified patterns as requested
1718 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1720 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1721 fopen() in the content-scanning modules that did not already have it.
1727 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1728 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1730 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1732 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1734 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1735 to test Sieve filters that use "vacation".
1737 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1738 that changes the way the GnuTLS parameters are stored in the cache file.
1739 The new format can be generated externally. For backward compatibility,
1740 if the data in the cache doesn't make sense, Exim assumes it has read an
1741 old-format file, and it generates new data and writes a new file. This
1742 means that you can't go back to an older release without removing the
1745 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1746 work if there are any delivery delays because "one_time" forces the
1747 parent to be marked "delivered", so its unseen clone is never tried
1748 again. For this reason, Exim now forbids the simultaneous setting of
1751 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1752 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1753 entry if you want to know the details. The fix, however, neglected to
1754 consider the case when local delivery batching is involved. The test for
1755 "previously delivered" was not happening when checking to see if an
1756 address could be batched with a previous (undelivered) one; under
1757 certain circumstances this could lead to multiple deliveries to the same
1760 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1761 in its include files, and this causes problems building Exim.
1763 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1764 header_syntax) but Exim was just ignoring anything given after a slash.
1765 In particular, this caused confusion with an attempt to use "verify =
1766 reverse_host_lookup/defer_ok". An error is now given when options are
1767 supplied for verify items that do not have them. (Maybe reverse_host_
1768 lookup should have a defer_ok option, but that's a different point.)
1770 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1771 defined by RFC 821) to 2048, because there were problems with some AUTH
1772 commands, and RFC 1869 says the size should be increased for extended
1773 SMTP commands that take arguments.
1775 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1778 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1779 "unknown" error; now it says that the functionality isn't in the binary.
1781 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1782 an address' error message when a string expansion fails (syntax or
1783 whatever). Otherwise the password may appear in the log. Following change
1784 PH/42 below, there is no longer a chance of it appearing in a bounce
1787 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1789 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1790 of its addresses was ever tried. (Bugzilla bug #2.)
1792 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1793 the result incorrectly in the debug output. (It correctly added a newline
1794 to what was transported.)
1796 TF/01 Added $received_time.
1798 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1799 commented out examples of how to interface to a virus scanner and to
1800 SpamAssassin. Also added commented examples of av_scanner and
1801 spamd_address settings.
1803 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1804 and controls are allowed in which ACLs. There were a couple of minor
1805 errors. Some of the entries in the conditions table (which is a table of
1806 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1807 negation of where the condition IS allowed.
1809 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1811 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1812 header file does not have a version number, so I've had to invent a new
1813 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1814 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1815 radiusclient), but it was contributed by a Radius user.
1817 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1818 files or directories, for OpenSSL.
1820 PH/19 When an Exim process that is running as root has to create an Exim log
1821 file, it does so in a subprocess that runs as exim:exim so as to get the
1822 ownership right at creation (otherwise, other Exim processes might see
1823 the file with the wrong ownership). There was no test for failure of this
1824 fork() call, which would lead to the process getting stuck as it waited
1825 for a non-existent subprocess. Forks do occasionally fail when resources
1826 run out. I reviewed all the other calls to fork(); they all seem to check
1829 PH/20 When checking for unexpected SMTP input at connect time (before writing
1830 the banner), Exim was not dealing correctly with a non-positive return
1831 from the read() function. If the client had disconnected by this time,
1832 the result was a log entry for a synchronization error with an empty
1833 string after "input=" when read() returned zero. If read() returned -1
1834 (an event I could not check), uninitialized data bytes were printed.
1835 There were reports of junk text (parts of files, etc) appearing after
1838 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1840 PH/22 Added support for macro redefinition, and (re)definition in between
1841 driver and ACL definitions.
1843 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1844 forgetting to use the resulting value; it was using the unexpanded value.
1846 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1847 hadn't been configured. The fix is from Juergen Kreileder, who
1848 understands it better than I do:
1850 "Here's what I see happening with three configured cyrus_sasl
1851 authenticators configured (plain, login, cram-md5):
1853 On startup auth_cyrus_sasl_init() gets called for each of these.
1854 This means three calls to sasl_listmech() without a specified mech_list.
1855 => SASL tests which mechs of all available mechs actually work
1856 => three warnings about OTP not working
1857 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1859 With the patch, sasl_listmech() also gets called three times. But now
1860 SASL's mech_list option is set to the server_mech specified in the the
1861 authenticator. Or in other words, the answer from sasl_listmech()
1862 gets limited to just the mech you're testing for (which is different
1864 => the return list contains just 'plain' or 'login', 'cram-md5' or
1865 nothing depending on the value of ob->server_mech.
1867 I've just tested the patch: Authentication still works fine,
1868 unavailable mechs specified in the exim configuration are still
1869 caught, and the auth.log warnings about OTP are gone."
1871 PH/25 When debugging is enabled, the contents of the command line are added
1872 to the debugging output, even when log_selector=+arguments is not
1875 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1876 answer is "GNU", and only if the return is "GNU/something" is the answer
1879 PH/27 $acl_verify_message is now set immediately after the failure of a
1880 verification in an ACL, and so is available in subsequent modifiers. In
1881 particular, the message can be preserved by coding like this:
1883 warn !verify = sender
1884 set acl_m0 = $acl_verify_message
1886 Previously, $acl_verify_message was set only while expanding "message"
1887 and "log_message" when a very denied access.
1889 PH/28 Modified OS/os.c-Linux with
1891 -#ifndef OS_LOAD_AVERAGE
1892 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1894 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1895 nomenclature these days.)
1897 PH/29 Installed patch from the Sieve maintainer that adds the options
1898 sieve_useraddress and sieve_subaddress to the redirect router.
1900 PH/30 In these circumstances:
1901 . Two addresses routed to the same list of hosts;
1902 . First host does not offer TLS;
1903 . First host accepts first address;
1904 . First host gives temporary error to second address;
1905 . Second host offers TLS and a TLS session is established;
1906 . Second host accepts second address.
1907 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1908 and peerdn, if requested) that were in fact used only for the second
1911 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1912 attention to any local part prefix or suffix that was matched by the
1913 router that accepted the address. It now behaves in the same way as it
1914 does for delivery: the affixes are removed from the local part unless
1915 rcpt_include_affixes is set on the transport.
1917 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1918 timeout during the DATA phase of an incoming message.
1920 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1921 applied a patch sanctioned by the Sieve maintainer.
1923 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1924 the uid or gid is negative. A case of a negative gid caused this to be
1925 noticed. The fix allows for either to be negative.
1927 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1928 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1931 PH/36 The change PH/12 above was broken. Fixed it.
1933 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1934 the grounds that routing the same address twice would always produce the
1935 same answer. This might have been true once, but it is certainly no
1936 longer true now. Routing a child address may depend on the previous
1937 routing that produced that child. Some complicated redirection strategies
1938 went wrong when messages had multiple recipients, and made Exim's
1939 behaviour dependent on the order in which the addresses were given.
1941 I have moved the duplicate checking until after the routing is complete.
1942 Exim scans the addresses that are assigned to local and remote
1943 transports, and removes any duplicates. This means that more work will be
1944 done, as duplicates will always all be routed, but duplicates are
1945 presumably rare, so I don't expect this is of any significance.
1947 For deliveries to pipes, files, and autoreplies, the duplicate checking
1948 still happens during the routing process, since they are not going to be
1951 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1952 It corrects a timeout issue with spamd. This is Ian's comment: "The
1953 background is that sometimes spamd either never reads data from a
1954 connection it has accepted, or it never writes response data. The exiscan
1955 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1956 blindly assumes that writes won't block so it may never time out."
1958 PH/39 Allow G after quota size as well as K and M.
1960 PH/40 The value set for $authenticated_id in an authenticator may not contain
1961 binary zeroes or newlines because the value is written to log lines and
1962 to spool files. There was no check on this. Now the value is run through
1963 the string_printing() function so that such characters are converted to
1964 printable escape sequences.
1966 PH/41 $message_linecount is a new variable that contains the total number of
1967 lines in the message. Compare $body_linecount, which is the count for the
1970 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1971 bounce and delay warning messages, except in certain special cases, which
1974 (a) An SMTP error message from a remote host;
1975 (b) A message specified in a :fail: redirection;
1976 (c) A message specified in a "fail" command in a system filter;
1977 (d) A message specified in a FAIL return from the queryprogram router;
1978 (e) A message specified by the cannot_route_message router option.
1980 In these cases only, Exim does include the error details in bounce and
1981 warning messages. There are also a few cases where bland messages such
1982 as "unrouteable address" or "local delivery error" are given.
1984 PH/43 $value is now also set for the "else" part of a ${run expansion.
1986 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1987 being worked on, but at least Exim now implements the latest version to
1990 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1991 process to complete was treated as a delivery failure, a timeout while
1992 writing the message to the pipe was logged, but erroneously treated as a
1993 successful delivery. Such timeouts include transport filter timeouts. For
1994 consistency with the overall process timeout, these timeouts are now
1995 treated as errors, giving rise to delivery failures by default. However,
1996 there is now a new Boolean option for the pipe transport called
1997 timeout_defer, which, if set TRUE, converts the failures into defers for
1998 both kinds of timeout. A transport filter timeout is now identified in
2001 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2002 systems where "make" and "gmake" are different, calling "gmake" at top
2003 level broke things. I've arranged for the value of $(MAKE) to be passed
2004 from the Makefile to this script so that it can call the same version of
2008 A note about Exim versions 4.44 and 4.50
2009 ----------------------------------------
2011 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2012 changes of various kinds. As a consequence, a big documentation update was
2013 needed. This delayed the release for rather longer than seemed good, especially
2014 in the light of a couple of (minor) security issues. Therefore, the changes
2015 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2016 release. So 4.44 and 4.50 are in effect two different branches that both start
2019 I have left the 4.50 change log unchanged; it contains all the changes since
2020 4.43. The change log for 4.44 is below; many of its items are identical to
2021 those for 4.50. This seems to be the most sensible way to preserve the
2022 historical information.
2028 1. Minor wording change to the doc/README.SIEVE file.
2030 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2031 computation of the current number of files was incorrect.
2033 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2034 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2037 4. Give more explanation in the error message when the command for a transport
2038 filter fails to execute.
2040 5. There are several places where Exim runs a non-Exim command in a
2041 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2042 was being done only for the command run by the queryprogram router. It is
2043 now done for all such subprocesses. The other cases are: ${run, transport
2044 filters, and the commands run by the lmtp and pipe transports.
2046 6. Added CONFIGURE_GROUP build-time option.
2048 7. Some older OS have a limit of 256 on the maximum number of file
2049 descriptors. Exim was using setrlimit() to set 1000 as a large value
2050 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2051 systems. I've change it so that if it can't get 1000, it tries for 256.
2053 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2054 was an oversight, and furthermore, ever since the addition of extra
2055 controls (e.g. 4.43/32), the checks on when to allow different forms of
2056 "control" were broken. There should now be diagnostics for all cases when a
2057 control that does not make sense is encountered.
2059 9. Added the /retain_sender option to "control=submission".
2061 10. $recipients is now available in the predata ACL (oversight).
2063 11. Tidy the search cache before the fork to do a delivery from a message
2064 received from the command line. Otherwise the child will trigger a lookup
2065 failure and thereby defer the delivery if it tries to use (for example) a
2066 cached ldap connection that the parent has called unbind on.
2068 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2069 of $address_data from the recipient verification was clobbered by the
2070 sender verification.
2072 13. The value of address_data from a sender verification is now available in
2073 $sender_address_data in subsequent conditions in the ACL statement.
2075 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2077 15. Added a new option "connect=<time>" to callout options, to set a different
2080 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2081 was its contents. (It was OK if the option was not defined at all.)
2083 17. A "Completed" log line is now written for messages that are removed from
2084 the spool by the -Mrm option.
2086 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2087 information about exactly what failed.
2089 19. Added -dd to debug only the daemon process.
2091 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2092 handles timeouts, both on the server side and network timeouts. Renamed the
2093 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2095 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2096 It is now set to "smtps".
2098 22. $host_address is now set to the target address during the checking of
2099 ignore_target_hosts.
2101 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2102 being passed; this would have caused $sender_host_name to have been used if
2103 matching the list had actually called for a host name (not very likely,
2104 since this list is usually IP addresses). A host name is now passed as
2107 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2108 code that specifies a non-restarting handler (typically sigaction() in
2109 modern systems) in an attempt to fix a rare and obscure crash bug.
2111 25. Narrowed the window for a race in the daemon that could cause it to ignore
2112 SIGCHLD signals. This is not a major problem, because they are used only to
2113 wake it up if nothing else does.
2115 26. A malformed maildirsize file could cause Exim to calculate negative values
2116 for the mailbox size or file count. Odd effects could occur as a result.
2117 The maildirsize information is now recalculated if the size or filecount
2120 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2121 support for a long time. Removed HAVE_SYS_VFS_H.
2123 28. Installed the latest version of exipick from John Jetmore.
2125 29. In an address list, if the pattern was not a regular expression, an empty
2126 subject address (from a bounce message) matched only if the pattern was an
2127 empty string. Non-empty patterns were not even tested. This was the wrong
2128 because it is perfectly reasonable to use an empty address as part of a
2129 database query. An empty address is now tested by patterns that are
2130 lookups. However, all the other forms of pattern expect the subject to
2131 contain a local part and a domain, and therefore, for them, an empty
2132 address still always fails if the pattern is not itself empty.
2134 30. Exim went into a mad DNS loop when attempting to do a callout where the
2135 host was specified on an smtp transport, and looking it up yielded more
2136 than one IP address.
2138 31. Re-factored the code for checking spool and log partition space into a
2139 function that finds that data and another that does the check. The former
2140 is then used to implement four new variables: $spool_space, $log_space,
2141 $spool_inodes, and $log_inodes.
2143 32. The RFC2047 encoding function was originally intended for short strings
2144 such as real names; it was not keeping to the 75-character limit for
2145 encoded words that the RFC imposes. It now respects the limit, and
2146 generates multiple encoded words if necessary. To be on the safe side, I
2147 have increased the buffer size for the ${rfc2047: expansion operator from
2150 33. It is now permitted to omit both strings after an "if" condition; if the
2151 condition is true, the result is "true". As before, when the second string
2152 is omitted, a false condition yields an empty string. This makes it less
2153 cumbersome to write custom ACL and router conditions.
2155 34. Failure to deliver a bounce message always caused it to be frozen, even if
2156 there was an errors_to setting on the router. The errors_to setting is now
2159 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2160 canonical form (fully expanded) before being placed in
2161 $sender_host_address.
2163 36. The table in the code that translates DNS record types into text (T_A to
2164 "A" for instance) was missing entries for NS and CNAME. It is just possible
2165 that this could have caused confusion if both these types were looked up
2166 for the same domain, because the text type is used as part of Exim's
2167 per-process caching. But the chance of anyone hitting this buglet seems
2170 37. The dnsdb lookup has been extended in a number of ways.
2172 (1) There is a new type, "zns", which walks up the domain tree until it
2173 finds some nameserver records. It should be used with care.
2175 (2) There is a new type, "mxh", which is like "mx" except that it returns
2176 just the host names, not the priorities.
2178 (3) It is now possible to give a list of domains (or IP addresses) to be
2179 looked up. The behaviour when one of the lookups defers can be
2180 controlled by a keyword.
2182 (4) It is now possible to specify the separator character for use when
2183 multiple records are returned.
2185 38. The dnslists ACL condition has been extended: it is now possible to supply
2186 a list of IP addresses and/or domains to be looked up in a particular DNS
2189 39. Added log_selector=+queue_time_overall.
2191 40. When running the queue in the test harness, wait just a tad after forking a
2192 delivery process, to get repeatability of debugging output.
2194 41. Include certificate and key file names in error message when GnuTLS fails
2195 to set them up, because the GnuTLS error message doesn't include the name
2196 of the failing file when there is a problem reading it.
2198 42. Allow both -bf and -bF in the same test run.
2200 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2202 44. The "Exiscan patch" is now merged into the mainline Exim source.
2204 45. Sometimes the final signoff response after QUIT could fail to get
2205 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2206 < 0 before doing a fflush(). This bug looks as though it goes back to the
2207 introduction of TLS in release 3.20, but "sometimes" must have been rare
2208 because the tests only now provoked it.
2210 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2211 (this can affect the format of dates).
2213 47. exim_tidydb, when checking for the continued existence of a message for
2214 which it has found a message-specific retry record, was not finding
2215 messages that were in split spool directories. Consequently, it was
2216 deleting retry records that should have stayed in existence.
2218 48. Steve fixed some bugs in eximstats.
2220 49. The SPA authentication driver was not abandoning authentication and moving
2221 on to the next authenticator when an expansion was forced to fail,
2222 contradicting the general specification for all authenticators. Instead it
2223 was generating a temporary error. It now behaves as specified.
2225 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2226 (the order specifies the preference for clients). The order is now AES256,
2227 AES128, 3DES, ARCFOUR128.
2229 51. Small patch to Sieve code - explicitly set From: when generating an
2232 52. Exim crashed if a remote delivery caused a very long error message to be
2233 recorded - for instance if somebody sent an entire SpamAssassin report back
2234 as a large number of 550 error lines. This bug was coincidentally fixed by
2235 increasing the size of one of Exim's internal buffers (big_buffer) that
2236 happened as part of the Exiscan merge. However, to be on the safe side, I
2237 have made the code more robust (and fixed the comments that describe what
2240 53. Now that there can be additional text after "Completed" in log lines (if
2241 the queue_time_overall log selector is set), a one-byte patch to exigrep
2242 was needed to allow it to recognize "Completed" as not the last thing in
2245 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2246 patch that reportedly fixes this has been added. I am not expert enough to
2247 create a test for it. This is what the patch creator wrote:
2249 "I found a little strange behaviour of ldap code when working with
2250 Windows 2003 AD Domain, where users was placed in more than one
2251 Organization Units. When I tried to give exim partial DN, the exit code
2252 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2253 But simultaneously result of request was absolutely normal ldap result,
2254 so I produce this patch..."
2256 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2257 REFERENCE, so I have modified the code to exclude the patch when that macro
2260 55. Some experimental protocols are using DNS PTR records for new purposes. The
2261 keys for these records are domain names, not reversed IP addresses. The
2262 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2263 leaves it alone. Component reversal etc. now happens only for IP addresses.
2266 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2268 57. Double the size of the debug message buffer (to 2048) so that more of very
2269 long debug lines gets shown.
2271 58. The exicyclog utility now does better if the number of log files to keep
2272 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2274 59. Two changes related to the smtp_active_hostname option:
2276 (1) $smtp_active_hostname is now available as a variable.
2277 (2) The default for smtp_banner uses $smtp_active_hostname instead
2278 of $primary_hostname.
2280 60. The host_aton() function is supposed to be passed a string that is known
2281 to be a valid IP address. However, in the case of IPv6 addresses, it was
2282 not checking this. This is a hostage to fortune. Exim now panics and dies
2283 if the condition is not met. A case was found where this could be provoked
2284 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2285 components; fortuitously, this particular loophole had already been fixed
2286 by change 4.50/55 above.
2288 If there are any other similar loopholes, the new check in host_aton()
2289 itself should stop them being exploited. The report I received stated that
2290 data on the command line could provoke the exploit when Exim was running as
2291 exim, but did not say which command line option was involved. All I could
2292 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2293 running as the user.
2296 61. There was a buffer overflow vulnerability in the SPA authentication code
2297 (which came originally from the Samba project). I have added a test to the
2298 spa_base64_to_bits() function which I hope fixes it.
2301 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2302 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2304 63. The daemon start-up calls getloadavg() while still root for those OS that
2305 need the first call to be done as root, but it missed one case: when
2306 deliver_queue_load_max is set with deliver_drop_privilege. This is
2307 necessary for the benefit of the queue runner, because there is no re-exec
2308 when deliver_drop_privilege is set.
2310 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2311 This has been fixed.
2313 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2314 was in use, was not putting the data itself into the right store pool;
2315 consequently, it could be overwritten for a subsequent message in the same
2316 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2319 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2321 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2322 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2323 the function were treating the return as a boolean value, which happened to
2324 work because 0=false and not-0=true, but is not correct code.
2326 68. The host_aton() function was not handling scoped IPv6 addresses (those
2327 with, for example, "%eth0" on the end) correctly.
2329 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2330 negated items (that is, ~something) in unsigned ints. Some compilers
2331 apparently mutter when there is no cast.
2333 70. If an address verification called from an ACL failed, and did not produce a
2334 user-specific message (i.e. there was only a "system" message), nothing was
2335 put in $acl_verify_message. In this situation, it now puts the system
2338 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2339 session; change 4.31/43 added the unwanted input to the log line - except
2340 that it did not do this in the start of session case. It now does.
2342 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2343 This is wrong and can cause the other end to generate a synchronization
2344 error if it is another Exim or anything else that does the synchronization
2345 check. A QUIT command is no longer sent after a timeout.
2347 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2348 during host lookups.
2350 74. The defer_ok option of callout verification was not working if it was used
2351 when verifying addresses in header lines, that is, for this case:
2353 verify = header_sender/callout=defer_ok
2355 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2356 those file descriptors could be used for SMTP connections. If anything
2357 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2358 could be sent to the SMTP client, causing chaos. The daemon now opens
2359 stdin, stdout, and stderr to /dev/null when it puts itself into the
2362 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2363 log by default. The user can override this with suitable Perl magic.
2365 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2366 the log message when discard triggers, was not working for the DATA ACL or
2367 for the non-SMTP ACL.
2369 78. Error message wording change in sieve.c.
2371 79. If smtp_accept_max_per_host was set, the number of connections could be
2372 restricted to fewer than expected, because the daemon was trying to set up
2373 a new connection before checking whether the processes handling previous
2374 connections had finished. The check for completed processes is now done
2375 earlier. On busy systems, this bug wouldn't be noticed because something
2376 else would have woken the daemon, and it would have reaped the completed
2379 80. If a message was submitted locally by a user whose login name contained one
2380 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2381 It caused a spool format error. I have fixed the spool reading code. A
2382 related problem was that the "from" clause in the Received: line became
2383 illegal because of the space(s). It is now covered by ${quote_local_part.
2385 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2388 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2390 83. Patch from Sieve maintainer to fix unterminated string problem in
2391 "vacation" handling.
2393 84. Some minor changes to the Linux configuration files to help with other
2394 OS variants using glibc.
2396 85. One more patch for Sieve to update vacation handling to latest spec.
2399 ----------------------------------------------------
2400 See the note above about the 4.44 and 4.50 releases.
2401 ----------------------------------------------------
2407 1. Change 4.43/35 introduced a bug that caused file counts to be
2408 incorrectly computed when quota_filecount was set in an appendfile
2411 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2412 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2415 3. Give more explanation in the error message when the command for a transport
2416 filter fails to execute.
2418 4. There are several places where Exim runs a non-Exim command in a
2419 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2420 was being done only for the command run by the queryprogram router. It is
2421 now done for all such subprocesses. The other cases are: ${run, transport
2422 filters, and the commands run by the lmtp and pipe transports.
2424 5. Some older OS have a limit of 256 on the maximum number of file
2425 descriptors. Exim was using setrlimit() to set 1000 as a large value
2426 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2427 systems. I've change it so that if it can't get 1000, it tries for 256.
2429 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2430 was an oversight, and furthermore, ever since the addition of extra
2431 controls (e.g. 4.43/32), the checks on when to allow different forms of
2432 "control" were broken. There should now be diagnostics for all cases when a
2433 control that does not make sense is encountered.
2435 7. $recipients is now available in the predata ACL (oversight).
2437 8. Tidy the search cache before the fork to do a delivery from a message
2438 received from the command line. Otherwise the child will trigger a lookup
2439 failure and thereby defer the delivery if it tries to use (for example) a
2440 cached ldap connection that the parent has called unbind on.
2442 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2443 of $address_data from the recipient verification was clobbered by the
2444 sender verification.
2446 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2447 was its contents. (It was OK if the option was not defined at all.)
2449 11. A "Completed" log line is now written for messages that are removed from
2450 the spool by the -Mrm option.
2452 12. $host_address is now set to the target address during the checking of
2453 ignore_target_hosts.
2455 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2456 being passed; this would have caused $sender_host_name to have been used if
2457 matching the list had actually called for a host name (not very likely,
2458 since this list is usually IP addresses). A host name is now passed as
2461 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2462 code that specifies a non-restarting handler (typically sigaction() in
2463 modern systems) in an attempt to fix a rare and obscure crash bug.
2465 15. Narrowed the window for a race in the daemon that could cause it to ignore
2466 SIGCHLD signals. This is not a major problem, because they are used only to
2467 wake it up if nothing else does.
2469 16. A malformed maildirsize file could cause Exim to calculate negative values
2470 for the mailbox size or file count. Odd effects could occur as a result.
2471 The maildirsize information is now recalculated if the size or filecount
2474 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2475 support for a long time. Removed HAVE_SYS_VFS_H.
2477 18. Updated exipick to current release from John Jetmore.
2479 19. Allow an empty sender to be matched against a lookup in an address list.
2480 Previously the only cases considered were a regular expression, or an
2483 20. Exim went into a mad DNS lookup loop when doing a callout where the
2484 host was specified on the transport, if the DNS lookup yielded more than
2487 21. The RFC2047 encoding function was originally intended for short strings
2488 such as real names; it was not keeping to the 75-character limit for
2489 encoded words that the RFC imposes. It now respects the limit, and
2490 generates multiple encoded words if necessary. To be on the safe side, I
2491 have increased the buffer size for the ${rfc2047: expansion operator from
2494 22. Failure to deliver a bounce message always caused it to be frozen, even if
2495 there was an errors_to setting on the router. The errors_to setting is now
2498 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2499 canonical form (fully expanded) before being placed in
2500 $sender_host_address.
2502 24. Updated eximstats to version 1.33
2504 25. Include certificate and key file names in error message when GnuTLS fails
2505 to set them up, because the GnuTLS error message doesn't include the name
2506 of the failing file when there is a problem reading it.
2508 26. Expand error message when OpenSSL has problems setting up cert/key files.
2511 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2512 (this can affect the format of dates).
2514 28. exim_tidydb, when checking for the continued existence of a message for
2515 which it has found a message-specific retry record, was not finding
2516 messages that were in split spool directories. Consequently, it was
2517 deleting retry records that should have stayed in existence.
2519 29. eximstats updated to version 1.35
2520 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2521 1.35 - bugfix such that pie charts by volume are generated correctly
2523 30. The SPA authentication driver was not abandoning authentication and moving
2524 on to the next authenticator when an expansion was forced to fail,
2525 contradicting the general specification for all authenticators. Instead it
2526 was generating a temporary error. It now behaves as specified.
2528 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2529 (the order specifies the preference for clients). The order is now AES256,
2530 AES128, 3DES, ARCFOUR128.
2532 31. Small patch to Sieve code - explicitly set From: when generating an
2535 32. Exim crashed if a remote delivery caused a very long error message to be
2536 recorded - for instance if somebody sent an entire SpamAssassin report back
2537 as a large number of 550 error lines. This bug was coincidentally fixed by
2538 increasing the size of one of Exim's internal buffers (big_buffer) that
2539 happened as part of the Exiscan merge. However, to be on the safe side, I
2540 have made the code more robust (and fixed the comments that describe what
2543 33. Some experimental protocols are using DNS PTR records for new purposes. The
2544 keys for these records are domain names, not reversed IP addresses. The
2545 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2546 leaves it alone. Component reversal etc. now happens only for IP addresses.
2549 34. The host_aton() function is supposed to be passed a string that is known
2550 to be a valid IP address. However, in the case of IPv6 addresses, it was
2551 not checking this. This is a hostage to fortune. Exim now panics and dies
2552 if the condition is not met. A case was found where this could be provoked
2553 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2554 components; fortuitously, this particular loophole had already been fixed
2555 by change 4.50/55 or 4.44/33 above.
2557 If there are any other similar loopholes, the new check in host_aton()
2558 itself should stop them being exploited. The report I received stated that
2559 data on the command line could provoke the exploit when Exim was running as
2560 exim, but did not say which command line option was involved. All I could
2561 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2562 running as the user.
2565 35. There was a buffer overflow vulnerability in the SPA authentication code
2566 (which came originally from the Samba project). I have added a test to the
2567 spa_base64_to_bits() function which I hope fixes it.
2570 36. The daemon start-up calls getloadavg() while still root for those OS that
2571 need the first call to be done as root, but it missed one case: when
2572 deliver_queue_load_max is set with deliver_drop_privilege. This is
2573 necessary for the benefit of the queue runner, because there is no re-exec
2574 when deliver_drop_privilege is set.
2576 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2577 was in use, was not putting the data itself into the right store pool;
2578 consequently, it could be overwritten for a subsequent message in the same
2579 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2582 38. Sometimes the final signoff response after QUIT could fail to get
2583 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2584 < 0 before doing a fflush(). This bug looks as though it goes back to the
2585 introduction of TLS in release 3.20, but "sometimes" must have been rare
2586 because the tests only now provoked it.
2592 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2593 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2594 Now it returns an integer. A number of calls were still expecting a T/F
2595 return. Fortuitously, in all cases, the tests worked in OK situations,
2596 which is the norm. However, things would have gone wrong on any write
2597 failures on the smtp file descriptor. This function is used when sending
2598 messages over SMTP and also when doing verify callouts.
2600 2. When Exim is called to do synchronous delivery of a locally submitted
2601 message (the -odf or -odi options), it no longer closes stderr before doing
2604 3. Implemented the mua_wrapper option.
2606 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2608 5. Implemented the functions header_remove(), header_testname(),
2609 header_add_at_position(), and receive_remove_recipient(), and exported them
2612 6. If an ACL "warn" statement specified the addition of headers, Exim already
2613 inserted X-ACL-Warn: at the start if there was no header name. However, it
2614 was not making this test for the second and subsequent header lines if
2615 there were newlines in the string. This meant that an invalid header could
2616 be inserted if Exim was badly configured.
2618 7. Allow an ACL "warn" statement to add header lines at the start or after all
2619 the Received: headers, as well as at the end.
2621 8. Added the rcpt_4xx retry error code.
2623 9. Added postmaster_mailfrom=xxx to callout verification option.
2625 10. Added mailfrom=xxxx to the callout verification option, for verify=
2628 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2629 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2631 12. Inserted some casts to stop certain compilers warning when using pointer
2632 differences as field lengths or precisions in printf-type calls (mostly
2633 affecting debugging statements).
2635 13. Added optional readline() support for -be (dynamically loaded).
2637 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2638 same clock tick as a message's arrival, so that its received time was the
2639 same as the "first fail" time on the retry record, and that message
2640 remained on the queue past the ultimate address timeout, every queue runner
2641 would try a delivery (because it was past the ultimate address timeout) but
2642 after another failure, the ultimate address timeout, which should have then
2643 bounced the address, did not kick in. This was a "< instead of <=" error;
2644 in most cases the first failure would have been in the next clock tick
2645 after the received time, and all would be well.
2647 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2648 being recognized when the domain list was tested by the match_domain
2649 condition in an expansion string.
2651 16. Added the ${str2b64: operator.
2653 17. Exim was always calling setrlimit() to set a large limit for the number of
2654 processes, without checking whether the existing limit was already
2655 adequate. (It did check for the limit on file descriptors.) Furthermore,
2656 errors from getrlimit() and setrlimit() were being ignored. Now they are
2657 logged to the main and panic logs, but Exim does carry on, to try to do its
2658 job under whatever limits there are.
2660 18. Imported PCRE 5.0.
2662 19. Trivial typo in log message " temporarily refused connection" (the leading
2665 20. If the log selector return_path_on_delivery was set and an address was
2666 redirected to /dev/null, the delivery process crashed because it assumed
2667 that a return path would always be set for a "successful" delivery. In this
2668 case, the whole delivery is bypassed as an optimization, and therefore no
2671 21. Internal re-arrangement: the function for sending a challenge and reading
2672 a response while authentication was assuming a zero-terminated challenge
2673 string. It's now changed to take a pointer and a length, to allow for
2674 binary data in such strings.
2676 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2678 23. Exim was not respecting finduser_retries when seeking the login of the
2679 uid under which it was called; it was always trying 10 times. (The default
2680 setting of finduser_retries is zero.) Also, it was sleeping after the final
2681 failure, which is pointless.
2683 24. Implemented tls_on_connect_ports.
2685 25. Implemented acl_smtp_predata.
2687 26. If the domain in control=submission is set empty, Exim assumes that the
2688 authenticated id is a complete email address when it generates From: or
2689 Sender: header lines.
2691 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2692 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2693 chown and chgrp in /bin and hostname in /usr/bin.
2695 28. Exim was keeping the "process log" file open after each use, just as it
2696 does for the main log. This opens the possibility of it remaining open for
2697 long periods when the USR1 signal hits a daemon. Occasional processlog
2698 errors were reported, that could have been caused by this. Anyway, it seems
2699 much more sensible not to leave this file open at all, so that is what now
2702 29. The long-running daemon process does not normally write to the log once it
2703 has entered its main loop, and it closes the log before doing so. This is
2704 so that log files can straightforwardly be renamed and moved. However,
2705 there are a couple of unusual error situations where the daemon does write
2706 log entries, and I had neglected to close the log afterwards.
2708 30. The text of an SMTP error response that was received during a remote
2709 delivery was being truncated at 512 bytes. This is too short for some of
2710 the long messages that one sometimes sees. I've increased the limit to
2713 31. It is now possible to make retry rules that apply only when a message has a
2714 specific sender, in particular, an empty sender.
2716 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2717 it possible to be selective about when SMTP synchronization is enforced.
2719 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2721 32. Implemented hosts_connection_nolog.
2723 33. Added an ACL for QUIT.
2725 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2728 35. Added mailbox_size and mailbox_filecount to appendfile.
2730 36. Added control = no_multiline_responses to ACLs.
2732 37. There was a bug in the logic of the code that waits for the clock to tick
2733 in the case where the clock went backwards by a substantial amount such
2734 that the microsecond fraction of "now" was more than the microsecond
2735 fraction of "then" (but the whole seconds number was less).
2737 38. Added support for the libradius Radius client library this is found on
2738 FreeBSD (previously only the radiusclient library was supported).
2744 1. When certain lookups returned multiple values in the form name=value, the
2745 quoting of the values was not always being done properly. Specifically:
2746 (a) If the value started with a double quote, but contained no whitespace,
2748 (b) If the value contained whitespace other than a space character (i.e.
2749 tabs or newlines or carriage returns) it was not quoted.
2750 This fix has been applied to the mysql and pgsql lookups by writing a
2751 separate quoting function and calling it from the lookup code. The fix
2752 should probably also be applied to nisplus, ibase and oracle lookups, but
2753 since I cannot test any of those, I have not disturbed their existing code.
2755 2. A hit in the callout cache for a specific address caused a log line with no
2756 reason for rejecting RCPT. Now it says "Previous (cached) callout
2757 verification failure".
2759 3. There was an off-by-one bug in the queryprogram router. An over-long
2760 return line was truncated at 256 instead of 255 characters, thereby
2761 overflowing its buffer with the terminating zero. As well as fixing this, I
2762 have increased the buffer size to 1024 (and made a note to document this).
2764 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2765 when Exim is waiting for an SMTP response from a remote server, Exim
2766 restarts its select() call on the socket, thereby resetting its timeout.
2767 This is not a problem when such interrupts are rare. Somebody set up a cron
2768 job to run exiwhat every 2 minutes, which is less than the normal select()
2769 timeout (5 or 10 minutes). This meant that the select() timeout never
2770 kicked in because it was always reset. I have fixed this by comparing the
2771 time when an interrupt arrives with the time at the start of the first call
2772 to select(). If more time than the timeout has elapsed, the interrupt is
2773 treated as a timeout.
2775 5. Some internal re-factoring in preparation for the addition of Sieve
2776 extensions (by MH). In particular, the "personal" test is moved to a
2777 separate function, and given an option for scanning Cc: and Bcc: (which is
2778 not set for Exim filters).
2780 6. When Exim created an email address using the login of the caller as the
2781 local part (e.g. when creating a From: or Sender: header line), it was not
2782 quoting the local part when it contained special characters such as @.
2784 7. Installed new OpenBSD configuration files.
2786 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2787 try to make them clearer.
2789 9. Callout options, other than the timeout value, were being ignored when
2790 verifying sender addresses in header lines. For example, when using
2792 verify = header_sender/callout=no_cache
2794 the cache was (incorrectly) being used.
2796 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2797 only the Cygwin environment.
2799 11. When return_path_on_delivery was set as a log selector, if different remote
2800 addresses in the same message used different return paths and parallel
2801 remote delivery occurred, the wrong values would sometimes be logged.
2802 (Whenever a remote delivery process finished, the return path value from
2803 the most recently started remote delivery process was logged.)
2805 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2806 lines when AUTH and/or TLS are in use. This is the "received protocol"
2807 field. Exim used to use "asmtp" for authenticated SMTP, without any
2808 indication (in the protocol name) for TLS use. Now it follows the RFC and
2809 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2810 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2811 names appear in log lines as well as in Received: header lines.
2813 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2814 extensions, and comparison tests, and to fix some bugs.
2816 14. Changes to the "personal" filter test:
2818 (1) The test was buggy in that it was just doing the equivalent of
2819 "contains" tests on header lines. For example, if a user's address was
2820 anne@some.where, the "personal" test would incorrectly be true for
2822 To: susanne@some.where
2824 This test is now done by extracting each address from the header in turn,
2825 and checking the entire address. Other tests that are part of "personal"
2826 are now done using regular expressions (for example, to check local parts
2827 of addresses in From: header lines).
2829 (2) The list of non-personal local parts in From: addresses has been
2830 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2831 taken from the Sieve specification recommendations.
2833 (3) If the message contains any header line starting with "List-" it is
2834 treated as non-personal.
2836 (4) The test for "circular" in the Subject: header line has been removed
2837 because it now seems ill-conceived.
2839 15. Minor typos in src/EDITME comments corrected.
2841 16. Installed latest exipick from John Jetmore.
2843 17. If headers_add on a router specified a text string that was too long for
2844 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2845 of string_sprintf() is now avoided.
2847 18. $message_body_size was not set (it was always zero) when running the DATA
2848 ACL and the local_scan() function.
2850 19. For the "mail" command in an Exim filter, no default was being set for
2851 the once_repeat time, causing a random time value to be used if "once" was
2852 specified. (If the value happened to be <= 0, no repeat happened.) The
2853 default is now 0s, meaning "never repeat". The "vacation" command was OK
2854 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2855 (I found it when inspecting the code).
2857 20. There is now an overall timeout for performing a callout verification. It
2858 defaults to 4 times the callout timeout, which applies to individual SMTP
2859 commands during the callout. The overall timeout applies when there is more
2860 than one host that can be tried. The timeout is checked before trying the
2861 next host. This prevents very long delays if there are a large number of
2862 hosts and all are timing out (e.g. when the network connections are timing
2863 out). The value of the overall timeout can be changed by specifying an
2864 additional sub-option for "callout", called "maxwait". For example:
2866 verify = sender/callout=5s,maxwait=20s
2868 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2869 to the end before writing, but this should make it even safer).
2871 22. Exim was forgetting that it had advertised PIPELINING for the second and
2872 subsequent messages on an SMTP connection. It was also not resetting its
2873 memory on STARTTLS and an internal HELO.
2875 23. When Exim logs an SMTP synchronization error within a session, it now
2876 records whether PIPELINING has been advertised or not.
2878 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2879 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2880 rather than long int.
2882 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2884 26. Added the never_mail option to autoreply.
2890 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2891 crash if the getsockname() call failed; this can happen if a connection is
2892 closed very soon after it is established. The problem was simply in the
2893 order in which certain operations were done, causing Exim to try to write
2894 to the SMTP stream before it had set up the file descriptor. The bug has
2895 been fixed by making things happen in the correct order.
2901 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2902 before the connection was closed, thus losing the rejection response.
2904 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2905 some early Solaris releases, but causes trouble in current releases where
2906 socklen_t is defined.
2908 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2911 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2914 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2915 that releasing the top of it at the end releases what was used for sub-
2916 expansions (unless the block got too big). However, discard this block if
2917 the first thing is a variable or header, so that we can use its block when
2918 it is dynamic (useful for very large $message_headers, for example).
2920 6. Lookups now cache *every* query, not just the most recent. A new, separate
2921 store pool is used for this. It can be recovered when all lookup caches are
2922 flushed. Lookups now release memory at the end of their result strings.
2923 This has involved some general refactoring of the lookup sources.
2925 7. Some code has been added to the store_xxx() functions to reduce the amount
2926 of flapping under certain conditions.
2928 8. log_incoming_interface used to affect only the <= reception log lines. Now
2929 it causes the local interface and port to be added to several more SMTP log
2930 lines, for example "SMTP connection from", and rejection lines.
2932 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2934 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2936 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2937 could be overwritten at the end of the current message (or the start of a
2938 new message if it was set in a HELO ACL). The value is now preserved for
2939 the duration of the SMTP connection.
2941 12. If a transport had a headers_rewrite setting, and a matching header line
2942 contained an unqualified address, that address was qualified, even if it
2943 did not match any rewriting rules. The underlying bug was that the values
2944 of the flags that permit the existence of unqualified sender and recipient
2945 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2946 non-local messages, and by -bnq for local messages) were not being
2947 preserved with the message after it was received.
2949 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2950 "next input=" as part of the text comprising the host identity instead of
2951 the correct text. The code was using the same buffer for two different
2952 strings. However, depending on which order the printing function evaluated
2953 its arguments, the bug did not always show up. Under Linux, for example, my
2954 test suite worked just fine.
2956 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2957 doesn't work with some older versions of Perl. It has been changed to "my",
2958 which in any case is probably the better facility to use.
2960 15. A really picky compiler found some instances of statements for creating
2961 error messages that either had too many or two few arguments for the format
2964 16. The size of the buffer for calls to the DNS resolver has been increased
2965 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2966 for addresses that have a lot of PTR records. This alleviates a problem; it
2967 does not fully solve it.
2969 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2970 buffer now truncates the list and logs the incident, which is the same
2971 action as happens when Exim is looking up a host name and its aliases.
2972 Previously in this situation something unpredictable would happen;
2973 sometimes it was "internal error: store_reset failed".
2975 18. If a server dropped the connection unexpectedly when an Exim client was
2976 using GnuTLS and trying to read a response, the client delivery process
2977 crashed while trying to generate an error log message.
2979 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2980 string, for example:
2982 warn message = H1: something\nH2: something
2984 the text was added as a single header line from Exim's point of view
2985 though it ended up OK in the delivered message. However, searching for the
2986 second and subsequent header lines using $h_h2: did not work. This has been
2987 fixed. Similarly, if a system filter added multiple headers in this way,
2988 the routers could not see them.
2990 20. Expanded the error message when iplsearch is called with an invalid key to
2991 suggest using net-iplsearch in a host list.
2993 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2994 ACLs are no longer actually imposed (and a message to that effect is
2997 22. If a "gecos" field in a passwd entry contained escaped characters, in
2998 particular, if it contained a \" sequence, Exim got it wrong when building
2999 a From: or a Sender: header from that name. A second bug also caused
3000 incorrect handling when an unquoted " was present following a character
3001 that needed quoting.
3003 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3004 was not being matched caselessly.
3006 24. Arranged for all hyphens in the exim.8 source to be escaped with
3009 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3010 properly. Recipient callout cache records were still being keyed to include
3011 the sender, even when use_sender was set false. This led to far more
3012 callouts that were necessary. The sender is no longer included in the key
3013 when use_sender is false.
3015 26. Added "control = submission" modifier to ACLs.
3017 27. Added the ${base62d: operator to decode base 62 numbers.
3019 28. dnsdb lookups can now access SRV records.
3021 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3022 the configuration file.
3024 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3025 (-v) mode. This makes the output for a verbose queue run more intelligible.
3027 31. Added a use_postmaster feature to recipient callouts.
3029 32. Added the $body_zerocount variable, containing the number of binary zero
3030 bytes in the message body.
3032 33. The time of last modification of the "new" subdirectory is now used as the
3033 "mailbox time last read" when there is a quota error for a maildir
3036 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3038 35. Added +ignore_unknown as a special item in host lists.
3040 36. Code for decoding IPv6 addresses in host lists is now included, even if
3041 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3042 address was recognized as an IP address, but was then not correctly decoded
3043 into binary, causing unexpected and incorrect effects when compared with
3050 1. Very minor rewording of debugging text in manualroute to say "list of
3051 hosts" instead of "hostlist".
3053 2. If verify=header_syntax was set, and a header line with an unqualified
3054 address (no domain) and a large number of spaces between the end of the
3055 name and the colon was received, the reception process suffered a buffer
3056 overflow, and (when I tested it) crashed. This was caused by some obsolete
3057 code that should have been removed. The fix is to remove it!
3059 3. When running in the test harness, delay a bit after writing a bounce
3060 message to get a bit more predictability in the log output.
3062 4. Added a call to search_tidyup() just before forking a reception process. In
3063 theory, someone could use a lookup in the expansion of smtp_accept_max_
3064 per_host which, without the tidyup, could leave open a database connection.
3066 5. Added the variables $recipient_data and $sender_data which get set from a
3067 lookup success in an ACL "recipients" or "senders" condition, or a router
3068 "senders" option, similar to $domain_data and $local_part_data.
3070 6. Moved the writing of debug_print from before to after the "senders" test
3073 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3074 problems for message scanning, either using a data ACL, or using
3075 local_scan() because the Received: header was not generated till after they
3076 were called (in order to set the time as the time of reception completion).
3077 I have revised the way this works. The header is now generated after the
3078 body is received, but before the ACL or local_scan() are called. After they
3079 are run, the timestamp in the header is updated.
3085 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3086 before starting a queue runner without re-exec. This happened only when
3087 deliver_drop_privilege was set or when the Exim user was set to root. The
3088 effect of the bug was that timeouts during subsequent deliveries caused
3089 crashes instead of being properly handled. The handler is now left at its
3090 default (and expected) setting.
3092 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3093 message, again when deliver_drop_privilege is set or Exim is run as root.
3094 The bug described in (1) was not present in this case, but the tidying up
3095 of the other signals was missing. I have made the two cases consistent.
3097 3. The ignore_target_hosts setting on a manualroute router was being ignored
3098 for hosts that were looked up using the /MX notation.
3100 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3103 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3104 operated on the sender address. After changing the $sender_address to <>
3105 for the sender address verify, Exim was re-instated it as the original
3106 (before rewriting) address, but remembering that it had rewritten it, so it
3107 wasn't rewriting it again. This bug also had the effect of breaking the
3108 sender address verification caching when the sender address was rewritten.
3110 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3111 This has been changed so that if the ip literal address matches
3112 ignore_target_hosts, the router declines.
3114 7. Added expansion conditions match_domain, match_address, and match_local_
3115 part (NOT match_host).
3117 8. The placeholder for the Received: header didn't have a length field set.
3119 9. Added code to Exim itself and to exim_lock to test for a specific race
3120 condition that could lead to file corruption when using MBX delivery. The
3121 issue is with the lockfile that is created in /tmp. If this file is removed
3122 after a process has opened it but before that process has acquired a lock,
3123 there is the potential for a second process to recreate the file and also
3124 acquire a lock. This could lead to two Exim processes writing to the file
3125 at the same time. The added code performs the same test as UW imapd; it
3126 checks after acquiring the lock that its file descriptor still refers to
3127 the same named file.
3129 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3130 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3133 11. Added the smtp_active_hostname option. If used, this will typically be made
3134 to depend on the incoming interface address. Because $interface_address is
3135 not set up until the daemon has forked a reception process, error responses
3136 that can happen earlier (such as "too many connections") no longer contain
3139 12. If an expansion in a condition on a "warn" statement fails because a lookup
3140 defers, the "warn" statement is abandoned, and the next ACL statement is
3141 processed. Previously this caused the whole ACL to be aborted.
3143 13. Added the iplsearch lookup type.
3145 14. Added ident_timeout as a log selector.
3147 15. Added tls_certificate_verified as a log selector.
3149 16. Added a global option tls_require_ciphers (compare the smtp transport
3150 option of the same name). This controls incoming TLS connections.
3152 17. I finally figured out how to make tls_require_ciphers do a similar thing
3153 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3154 before starting the TLS session.
3156 18. Tabs are now shown as \t in -bP output.
3158 19. If the log selector return_path_on_delivery was set, Exim crashed when
3159 bouncing a message because it had too many Received: header lines.
3161 20. If two routers both had headers_remove settings, and the first one included
3162 a superfluous trailing colon, the final name in the first list and the
3163 first name in the second list were incorrectly joined into one item (with a
3164 colon in the middle).
3170 1. Added -C and -D options to the exinext utility, mainly to make it easier
3171 to include in the automated testing, but these could be helpful when
3172 multiple configurations are in use.
3174 2. The exinext utility was not formatting the output nicely when there was
3175 an alternate port involved in the retry record key, nor when there was a
3176 message id as well (for retries that were specific to a specific message
3177 and a specific host). It was also confused by IPv6 addresses, because of
3178 the additional colons they contain. I have fixed the IPv4 problem, and
3179 patched it up to do a reasonable job for IPv6.
3181 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3182 delivery, the log line now contains "pipelined" if PIPELINING was used.
3184 4. An SMTP transport process used to panic and die if the bind() call to set
3185 an explicit outgoing interface failed. This has been changed; it is now
3186 treated in the same way as a connect() failure.
3188 5. A reference to $sender_host_name in the part of a conditional expansion
3189 that was being skipped was still causing a DNS lookup. This no longer
3192 6. The def: expansion condition was not recognizing references to header lines
3193 that used bh_ and bheader_.
3195 7. Added the _cache feature to named lists.
3197 8. The code for checking quota_filecount in the appendfile transport was
3198 allowing one more file than it should have been.
3200 9. For compatibility with Sendmail, the command line option
3208 and sets the incoming protocol and host name (for trusted callers). The
3209 host name and its colon can be omitted when only the protocol is to be set.
3210 Note the Exim already has two private options, -pd and -ps, that refer to
3211 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3212 "s", but I don't think that's a major issue.
3214 10. A number of refactoring changes to the code, none of which should affect
3217 (a) The number of logging options was getting close to filling up the
3218 32-bit word that was used as a bit map. I have split them into two classes:
3219 those that are passed in the argument to log_write(), and those that are
3220 only ever tested independently outside of that function. These are now in
3221 separate 32-bit words, so there is plenty of room for expansion again.
3222 There is no change in the user interface or the logging behaviour.
3224 (b) When building, for example, log lines, the code previously used a
3225 macro that called string_cat() twice, in order to add two strings. This is
3226 not really sufficiently general. Furthermore, there was one instance where
3227 it was actually wrong because one of the argument was used twice, and in
3228 one call a function was used. (As it happened, calling the function twice
3229 did not affect the overall behaviour.) The macro has been replaced by a
3230 function that can join an arbitrary number of extra strings onto a growing
3233 (c) The code for expansion conditions now uses a table and a binary chop
3234 instead of a serial search (which was left over from when there were very
3235 few conditions). Also, it now recognizes conditions like "pam" even when
3236 the relevant support is not compiled in: a suitably worded error message is
3237 given if an attempt is made to use such a condition.
3239 11. Added ${time_interval:xxxxx}.
3241 12. A bug was causing one of the ddress fields not to be passed back correctly
3242 from remote delivery subprocesses. The field in question was not being
3243 subsequently used, so this caused to problems in practice.
3245 13. Added new log selectors queue_time and deliver_time.
3247 14. Might have fixed a bug in maildirsizefile handling that threw up
3248 "unexpected character" debug warnings, and recalculated the data
3249 unnecessarily. In any case, I expanded the warning message to give more
3252 15. Added the message "Restricted characters in address" to the statements in
3253 the default ACL that block characters like @ and % in local parts.
3255 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3256 Three changes have been made:
3258 (a) There was a serious bug; a negative response to MAIL caused the whole
3259 recipient domain to be cached as invalid, thereby blocking all messages
3260 to all local parts at the same domain, from all senders. This bug has
3261 been fixed. The domain is no longer cached after a negative response to
3262 MAIL if the sender used is not empty.
3264 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3267 (c) A new callout option, "use_sender" has been added for people who want
3268 the modified behaviour.
3274 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3277 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3278 indeed breaks things for older releases.
3280 3. Added additional logging to the case where there is a problem reading data
3281 from a filter that is running in a subprocess using a pipe, in order to
3282 try to track down a specific problem.
3284 4. Testing facility fudge: when running in the test harness and attempting
3285 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3286 getting "No route to host". Convert this to a timeout.
3288 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3291 6. Some OS don't have socklen_t but use size_t instead. This affects the
3292 fifth argument of getsockopt() amongst other things. This is now
3293 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3294 can be set for individual OS. I have set it for SunOS5, OSF1, and
3295 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3296 some earlier ones do not.
3298 7. Change 4.30/15 was not doing the test caselessly.
3300 8. The standard form for an IPv6 address literal was being rejected by address
3301 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3302 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3303 this, as well as the form without the "IPv6" on the front (but only when
3304 address literals are enabled, of course).
3306 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3308 10. Exim crashed if a message with an empty sender address specified by -f
3309 encountered a router with an errors_to setting. This could be provoked only
3310 by a command such as
3314 where an empty string was supplied; "<>" did not hit this bug.
3316 11. Installed PCRE release 4.5.
3318 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3319 remained set. It is now erased.
3321 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3322 times from message ids (which are base 36 rather than the normal 62).
3324 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3325 were being counted as actual protocol errors, and logged if the log
3326 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3327 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3328 and DATA following a set of rejected RCPTs do not count as protocol errors.
3329 In other words, Exim assumes they were pipelined, though this may not
3330 actually be the case. Of course, in all cases the client gets an
3331 appropriate error code.
3333 15. If a lookup fails in an ACL condition, a message about the failure may
3334 be available; it is used if testing the ACL cannot continue, because most
3335 such messages specify what the cause of the deferral is. However, some
3336 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3337 that caused an old message to be retained and used if a later statement
3338 caused a defer, replacing the real cause of the deferral.
3340 16. If an IP address had so many PTR records that the DNS lookup buffer
3341 was not large enough to hold them, Exim could crash while trying to process
3342 the truncated data. It now detects and logs this case.
3344 17. Further to 4.21/58, another change has been made: if (and only if) the
3345 first line of a message (the first header line) ends with CRLF, a bare LF
3346 in a subsequent header line has a space inserted after it, so as not to
3347 terminate the header.
3349 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3350 unnecessarily, used atoi() instead of strtol(), and didn't check the
3351 termination when getting file sizes from file names by regex.
3353 19. Completely re-implemented the support for maildirsize files, in the light
3354 of a number of problems with the previous contributed implementation
3355 (4.30/29). In particular:
3357 . If the quota is zero, the maildirsize file is maintained, but no quota is
3360 . If the maildir directory does not exist, it is created before any attempt
3361 to write a maildirsize file.
3363 . The quota value in the file is just a cache; if the quota is changed in
3364 the transport, the new value overrides.
3366 . A regular expression is available for excluding directories from the
3369 20. The autoreply transport checks the characters in options that define the
3370 message's headers; it allows continued headers, but it was checking with
3371 isspace() after an embedded newline instead of explicitly looking for a
3374 21. If all the "regular" hosts to which an address was routed had passed their
3375 expiry times, and had not reached their retry times, the address was
3376 bounced, even if fallback hosts were defined. Now Exim should go on to try
3379 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3380 equivalent code in the SMTP transport. Some hosts send humungous responses
3381 to HELO/EHLO, more than 1024 it seems.
3383 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3384 gives compiler warnings in some environments. I've now done it "properly",
3387 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3388 (because of problems with the built-in one) was declared to return uschar *
3389 instead of char *, causing compiler failure.
3391 25. Fixed a file descriptor leak when processing alias/forward files.
3393 26. Fixed a minor format string issue in dbfn.c.
3395 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3397 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3398 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3401 29. When a sender address is verified, it is cached, to save repeating the test
3402 when there is more than one recipient in a message. However, when the
3403 verification involves a callout, it is possible for different callout
3404 options to be set for different recipients. It is too complicated to keep
3405 track of this in the cache, so now Exim always runs a verification when a
3406 callout is required, relying on the callout cache for the optimization.
3407 The overhead is duplication of the address routing, but this should not be
3410 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3411 to be verified with callout=postmaster, and the main callout worked but the
3412 postmaster check failed, the verification correctly failed. However, if a
3413 subsequent RCPT command asked for sender verification *without* the
3414 postmaster check, incorrect caching caused this verification also to fail,
3417 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3418 it was not caching the DNS options (qualify_single, search_parents) that
3419 were used when the lookup failed. A subsequent lookup with different
3420 options therefore always gave the same answer, though there were cases
3421 where it should not have. (Example: a "domains = !$mx_any" option on a
3422 dnslookup router: the "domains" option is always processed without any
3423 widening, but the router might have qualify_single set.) Now Exim uses the
3424 cached value only when the same options are set.
3426 32. Added John Jetmore's "exipick" utility to the distribution.
3428 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3429 than a timeout (e.g. a certificate is required, and is not provided), an
3430 Exim server now closes the connection immediately. Previously it waited for
3431 the client to close - but if the client is SSL, it seems that they each
3432 wait for each other, leading to a delay before one of them times out.
3434 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3435 maintained 0.8.x compatibility because I don't think many are using it, and
3436 it is clearly obsolete.
3438 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3441 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3442 client certificate was expired. A simple patch fixes this, though I don't
3443 understand the full logic of why the verify callback is called multiple
3446 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3447 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3448 which causes problems with some clients (such as the Certicom SSL Plus
3449 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3450 disables the coutermeasure allowing Eudora to connect."
3452 38. Exim was not checking that a write() to a log file succeeded. This could
3453 lead to Bad Things if a log got too big, in particular if it hit a file
3454 size limit. Exim now panics and dies if it cannot write to a log file, just
3455 as it does if it cannot open a log file.
3457 39. Modified OS/Makefile-Linux so that it now contains
3459 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3461 The two -D definitions ensure that Exim is compiled with large file
3462 support, which makes it possible to handle log files that are bigger than
3465 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3466 instance) a domain was checked against a named list that involved a lookup,
3467 causing $domain_data to be set, then another domain was checked against the
3468 same list, then the first domain was re-checked, the value of $domain_data
3469 after the final check could be wrong. In particular, if the second check
3470 failed, it could be set empty. This bug probably also applied to
3473 41. The strip_trailing_dot option was not being applied to the address given
3474 with the -f command-line option.
3476 42. The code for reading a message's header from the spool was incrementing
3477 $received_count, but never initializing it. This meant that the value was
3478 incorrect (doubled) while delivering a message in the same process in which
3479 it was received. In the most common configuration of Exim, this never
3480 happens - a fresh exec is done - but it can happen when
3481 deliver_drop_privilege is set.
3483 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3484 it now includes up to 150 characters of the unexpected data in the log
3487 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3488 and building data strings. The size of both of these buffers was 10 000
3489 bytes - far larger than anybody would *ever* want, thought I. Needless to
3490 say, somebody hit the limit. I have increased the maximum line length to
3491 20 000 and the maximum data length of concatenated lines to 100 000. I have
3492 also fixed two bugs, because there was no checking on these buffers. Tsk,
3493 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3494 buffer is too small.
3496 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3497 lsearch lookups. Now it does.
3499 46. When parsing a route_list item in a manualroute router, a fixed-length
3500 buffer was used for the list of hosts. I made this 1024 bytes long,
3501 thinking that nobody would ever have a list of hosts that long. Wrong.
3502 Somebody had a whole pile of complicated expansion conditions, and the
3503 string was silently truncated, leading to an expansion error. It turns out
3504 that it is easier to change to an unlimited length (owing to other changes
3505 that have happened since this code was originally written) than to build
3506 structure for giving a limitation error. The length of the item that
3507 expands into the list of hosts is now unlimited.
3509 47. The lsearch lookup could not handle data where the length of text line was
3510 more than 4095 characters. Such lines were truncated, leading to shortened
3511 data being returned. It should now handle lines of any length.
3513 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3514 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3517 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3518 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3519 no longer happens in comment lines. A list of replacements is now placed
3520 at the head of all of the source files, except those whose only change is
3521 to replace PERL_COMMAND in the very first #! line.
3523 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3524 messages on the queue, with a bottom-up merge sort, using code contributed
3525 by Michael Haardt. This should make operations like -bp somewhat faster on
3526 large queues. It won't affect queue runners, except when queue_run_in_order
3529 51. Installed eximstats 1.31 in the distribution.
3531 52. Added support for SRV lookups to the dnslookup router.
3533 53. If an ACL referred to $message_body or $message_body_end, the value was not
3534 reset for any messages that followed in the same SMTP session.
3536 54. The store-handling optimization for building very long strings was not
3537 differentiating between the different store pools. I don't think this
3538 actually made any difference in practice, but I've tidied it.
3540 55. While running the routers to verify a sender address, $sender_address
3541 was still set to the sender address. This is wrong, because when routing to
3542 send a bounce to the sender, it would be empty. Therefore, I have changed
3543 it so that, while verifying a sender address, $sender_address is set to <>.
3544 (There is no change to what happens when verifying a recipient address.)
3546 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3547 target A or AAAA records (if not already returned) without resetting the
3548 qualify_single or search_parents options of the DNS resolver. These are
3549 inappropriate in this case because the targets of MX and SRV records must
3550 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3551 target that, when qualified, matched something in the local domain. These
3552 two options are now turned off when doing these lookups.
3554 57. It seems that at least some releases of Reiserfs (which does not have the
3555 concept of a fixed number of inodes) returns zero and not -1 for the
3556 number of available inodes. This interacted badly with check_spool_inodes,
3557 which assumed that -1 was the "no such thing" setting. What I have done is
3558 to check that the total number of inodes is greater than zero before doing
3559 the test of how many are available.
3561 58. When a "warn" ACL statement has a log_message modifier, the message is
3562 remembered, and not repeated. This is to avoid a lot of repetition when a
3563 message has many recipients that cause the same warning to be written.
3564 Howewer, Exim was preserving the list of already written lines for an
3565 entire SMTP session, which doesn't seem right. The memory is now reset if a
3566 new message is started.
3568 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3569 debugging output unless log_rewrite was also set.
3571 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3572 of (int)(handle) when we know that handle contains (void *)(-1).
3574 61. The Exim daemon panic-logs an error return when it closes the incoming
3575 connection. However "connection reset by peer" seems to be common, and
3576 isn't really an error worthy of noting specially, so that particular error
3579 62. When Exim is trying to find all the local interfaces, it used to panic and
3580 die if the ioctl to get the interface flags failed. However, it seems that
3581 on at least one OS (Solaris 9) it is possible to have an interface that is
3582 included in the list of interfaces, but for which you get a failure error
3583 for this call. This happens when the interface is not "plumbed" into a
3584 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3585 failure of the "get flags" call assumes that the interface is down.
3587 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3588 makes life easier for people who are doing arithmetic on fields extracted
3589 from dates, where you often get leading zeros that should not be
3590 interpreted as octal.
3592 64. Added qualify_domain to the redirect router, to override the global
3595 65. If a pathologically long header line contained very many addresses (the
3596 report of this problem mentioned 10 000) and each of them was rewritten,
3597 Exim could use up a very large amount of memory. (It kept on making new
3598 copies of the header line as it rewrote, and never released the old ones.)
3599 At the expense of a bit more processing, the header rewriting function has
3600 been changed so that it no longer eats memory in this way.
3602 66. The generation of the Received: header has been moved from the time that a
3603 message starts to be received, to the time that it finishes. The timestamp
3604 in the Received: header should now be very close to that of the <= log
3605 line. There are two side-effects of this change:
3607 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3608 logged header lines no longer include the local Received: line, because
3609 it has not yet been created. The same applies to a copy of the message
3610 that is returned to a non-SMTP sender when a message is rejected.
3612 (b) When a filter file is tested using -bf, no additional Received: header
3613 is added to the test message. After some thought, I decided that this
3616 This change does not affect the value of $received_for. It is still set
3617 after address rewriting, but before local_scan() is called.
3619 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3621 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3622 gave an unhelpful panic error message, and a defer error. I have managed to
3623 change this behaviour so that it now rejects any supplied certificate,
3624 which seems right, as the list of acceptable certificates is empty.
3626 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3627 gave an unhelpful defer error. I have not managed to make this reject any
3628 supplied certificates, but the error message it gives is "no certificate
3629 supplied", which is not helpful.
3631 70. exigrep's output now also includes lines that are not associated with any
3632 message, but which match the given pattern. Implemented by a patch from
3633 Martin Sluka, which also tidied up the Perl a bit.
3635 71. Recipient callout verification, like sender verification, was using <> in
3636 the MAIL FROM command. This isn't really the right thing, since the actual
3637 sender may affect whether the remote host accepts the recipient or not. I
3638 have changed it to use the actual sender in the callout; this means that
3639 the cache record is now keyed on a recipient/sender pair, not just the
3640 recipient address. There doesn't seem to be a real danger of callout loops,
3641 since a callout by the remote host to check the sender would use <>.
3642 [SEE ABOVE: changed after hitting problems.]
3644 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3645 temporary errors. However, in the case of such a code being given after
3646 the end of a data transmission (i.e. after ".") Exim was failing to write
3647 a retry record for the message. (Yes, there was some broken host that was
3648 actually sending 8xx at this point.)
3650 73. An unknown lookup type in a host list could cause Exim to panic-die when
3651 the list was checked. (An example that provoked this was putting <; in the
3652 middle of a list instead of at the start.) If this happened during a DATA
3653 ACL check, a -D file could be left lying around. This kind of configuration
3654 error no longer causes Exim to die; instead it causes a defer errror. The
3655 incident is still logged to the main and panic logs.
3657 74. Buglet left over from Exim 3 conversion. The message "too many messages
3658 in one connection" was written to the rejectlog but not the mainlog, except
3659 when address rewriting (yes!) was being logged.
3661 75. Added write_rejectlog option.
3663 76. When a system filter was run not as root (that is, when system_filter_user
3664 was set), the values of the $n variables were not being returned to the
3665 main process; thus, they were not subsequently available in the $sn
3668 77. Added +return_path_on_delivery log selector.
3670 78. A connection timeout was being treated differently from recipients deferred
3671 when testing hosts_max_try with a message that was older than the host's
3672 retry timeout. (The host should not be counted, thus allowing all hosts to
3673 be tried at least once before bouncing.) This may have been the cause of an
3674 occasionally reported bug whereby a message would remain on the queue
3675 longer than the retry timeout, but would be bounced if a delivery was
3676 forced. I say "may" because I never totally pinned down the problem;
3677 setting up timeout/retry tests is difficult. See also the next item.
3679 79. The ultimate address timeout was not being applied to errors that involved
3680 a combination of host plus message (for example, a timeout on a MAIL
3681 command). When an address resolved to a number of possible hosts, and they
3682 were not all tried for each delivery (e.g. because of hosts_max_try), a
3683 message could remain on the queue longer than the retry timeout.
3685 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3688 81. Fixed an obscure SMTP outgoing bug which required at least the following
3689 conditions: (a) there was another message waiting for the same server;
3690 (b) the server returned 5xx to all RCPT commands in the first message so
3691 that the message was not completed; (c) the server dropped the connection
3692 or gave a negative response to the RSET that Exim sends to abort the
3693 transaction. The observed case was a dropped connection after DATA that had
3694 been sent in pipelining mode. That is, the server had advertised PIPELINING
3695 but was not implementing it correctly. The effect of the bug was incorrect
3696 behaviour, such as trying another host, and this could lead to a crash.
3702 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3703 and daemon.c were passed as pointers to ints; they should have been
3704 pointers to socklen_t variables (which are typically unsigned ints).
3706 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3709 3. Fixed a really odd bug that affected only the testing scheme; patching a
3710 certain fixed string in the binary changed the value of another string that
3711 happened to be identical to the end of the original first string.
3713 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3714 name", it returns that address as the IP address. On some operating
3715 systems (e.g. Solaris), it also passes back the IP address string as the
3716 "host name". However, on others (e.g. Linux), it passes back an empty
3717 string. Exim wasn't checking for this, and was changing the host name to an
3718 empty string, assuming it had been canonicized.
3720 5. Although rare, it is permitted to have more than one PTR record for a given
3721 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3722 all the names associated with an address, because they do in Solaris.
3723 However, it seems that they do not in Linux for data that comes from the
3724 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3725 I found this out when I moved to a new Linux workstation and tried to run
3726 the Exim test suite.
3728 To get round this problem I have changed the code so that it now does its
3729 own call to the DNS to look up PTR records when searching for a host name.
3730 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3731 addresses that are only in /etc/hosts are still found.
3733 This behaviour is, however, controlled by an option called host_lookup_
3734 order, which defaults to "bydns:byaddr". If people want to use the other
3735 order, or indeed, just use one or the other means of lookup, they can
3736 specify it in this variable.
3738 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3739 some operating systems, this comes back from gethostbyaddr() as an empty
3740 string, and this is what Exim used to test for. However, it seems that in
3741 other systems, "." is yielded. Exim now tests for this case too.
3743 7. The values of check_spool_space and check_log_space are now held internally
3744 as a number of kilobytes instead of an absolute number of bytes. If a
3745 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3746 kilobyte. This means that much larger values can be stored.
3748 8. Exim monitor: an attempt to get the action menu when not actually pointing
3749 at a message produces an empty menu entitled "No message selected". This
3750 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3751 no entries in it ("Shell widget menu has zero width and/or height"). So I
3752 have added a single, blank menu entry in this case.
3754 9. Added ${quote_local_part.
3756 10. MIME decoding is now applied to the contents of Subject: header lines when
3759 11. Now that a reference to $sender_host_address automatically causes a reverse
3760 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3761 host lookup before query-style lookups in lists that might use this
3762 variable. This has therefore been abolished, and the "net-" prefix is no
3763 longer necessary for query-style lookups.
3765 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3766 have been a typo for LFLAGS, so it has been changed.
3768 13. The install script calls Exim with "-C /dev/null" in order to find the
3769 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3770 to be output. Howeve, since Exim outputs its version number before the
3771 error, it didn't break the script. It just looked ugly. I fixed this by
3772 always allowing "-C /dev/null" if the caller is root.
3774 14. Ignore overlarge ACL variable number when reading spool file - insurance
3775 against a later release with more variables having written the file.
3777 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3778 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3779 this, as well as the form without the "IPv6" on the front.
3781 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3782 OS/Makefile-Darwin file.
3784 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3785 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3787 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3788 variable. However, some OS use ints for inodes. I've added cast to long int
3789 to get rid of the compiler warning.
3791 19. I had forgotten to lock out "/../" in configuration file names when
3792 ALT_CONFIG_PREFIX was set.
3794 20. Routers used for verification do not need to specify transports. However,
3795 if such a router generated a host list, and callout was configured, Exim
3796 crashed, because it could not find a port number from the (non-existent)
3797 transport. It now assumes port 25 in this circumstance.
3799 21. Added the -t option to exigrep.
3801 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3802 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3803 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3804 them is set, LOOKUP_LSEARCH is forced.
3806 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3808 24. Added sender and host information to the "rejected by local_scan()" log
3809 line; previously there was no indication of these.
3811 25. Added .include_if_exists.
3813 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3814 Linux. It turns out that not all file systems support this. Apparently some
3815 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3816 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3817 not supported on the file descriptor, is now ignored when Exim is trying to
3818 sync a directory. This applies only to Linux.
3820 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3822 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3823 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3826 29. Added support for maildirsize files from supplied patch (modified a bit).
3828 30. The use of :fail: followed by an empty string could lead Exim to respond to
3829 sender verification failures with (e.g.):
3831 550 Verification failed for <xxx>
3832 550 Sender verify failed
3834 where the first response line was missing the '-' that indicates it is not
3835 the final line of the response.
3837 31. The loop for finding the name of the user that called Exim had a hardwired
3838 limit of 10; it now uses the value of finduser_retries, which is used for
3839 all other user lookups.
3841 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3844 33. Exim was neglecting to zero errno before one call of strtol() when
3845 expanding a string and expecting an integer value. On some systems this
3846 resulted in spurious "integer overflow" errors. Also, it was casting the
3847 result into an int without checking.
3849 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3850 did not work. The code looks as if it has *never* worked, though it appears
3851 to have been documented since at least releast 1.62. I have made it work.
3853 35. The "timeout_DNS" error in retry rules, also documented since at least
3854 1.62, also never worked. As it isn't clear exactly what this means, and
3855 clearly it isn't a major issue, I have abolished the feature by treating it
3856 as "timeout", and writing a warning to the main and panic logs.
3858 36. The display of retry rules for -brt wasn't always showing the error code
3861 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3862 timeout_connect_A, timeout_connect_MX.
3864 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3865 to the empty sender.
3867 39. The daemon was not analysing the content of -oX till after it had closed
3868 stderr and disconnected from the controlling terminal. This meant that any
3869 syntax errors were only noted on the panic log, and the return code from
3870 the command was 0. By re-arranging the code a little, I've made the
3871 decoding happen first, so such errors now appear on stderr, and the return
3872 code is 1. However, the actual setting up of the sockets still happens in
3873 the disconnected process, so errors there are still only recorded on the
3876 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3877 connections (as happens on some IP stacks) was logged at start up time as
3878 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3879 it from "IPv6 and IPv4", which means that two separate sockets are being
3882 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3883 says whether AF_INET or AF_INET6 was passed as an argument.
3885 42. Exiwhat output was messed up when time zones were included in log
3888 43. Exiwhat now gives more information about the daemon's listening ports,
3889 and whether -tls-on-connect was used.
3891 44. The "port" option of the smtp transport is now expanded.
3893 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3894 silently ignored. Now an error message is written to the main and panic
3897 46. There's a new ACL modifier called "logwrite" which writes to a log file
3898 as soon as it is encountered.
3900 47. Added $local_user_uid and $local_user_gid at routing time.
3902 48. Exim crashed when trying to verify a sender address that was being
3905 49. Exim was recognizing only a space character after ".include". It now also
3906 recognizes a tab character.
3908 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3909 extracting the relevant information from the specification. The man page no
3910 longer contains scrambled data for the -d option, and I've added a section
3911 at the front about calling Exim under different names.
3913 51. Added "extra_headers" argument to the "mail" command in filter files.
3915 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3918 53. Installed eximstats 1.29.
3920 54. Added transport_filter_timeout as a generic transport option.
3922 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3923 Cc: header lines. This was required by RFC 822, but it not required by RFC
3926 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3927 incoming messages that did not have them. Now it does so only if the
3928 message originates locally, that is, if there is no associated remote host
3929 address. When Resent- header lines are present, this applies to the Resent-
3930 lines rather than the non-Resent- lines.
3932 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3933 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3935 58. Messages for configuration errors now include the name of the main
3936 configuration file - useful now that there may be more than one file in a
3937 list (.included file names were always shown).
3939 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3940 for those rare installations that do not start the daemon as root or run it
3941 setuid root. I've cut out the call to initgroups() if the daemon is not
3944 60. The Exim user and group can now be bound into the binary as text strings
3945 that are looked up at the start of Exim's processing.
3947 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3949 62. Added $mailstore_basename variable.
3951 63. Installed patch to sieve.c from Michael Haardt.
3953 64. When Exim failed to open the panic log after failing to open the main log,
3954 the original message it was trying to log was written to stderr and debug
3955 output, but if they were not available (the usual case in production), it
3956 was lost. Now it is written to syslog before the two lines that record the
3957 failures to open the logs.
3959 65. Users' Exim filters run in subprocesses under the user's uid. It is
3960 possible for a "deliver" command or an alias in a "personal" command to
3961 provoke an address rewrite. If logging of address rewriting is configured,
3962 this fails because the process is not running as root or exim. There may be
3963 a better way of dealing with this, but for the moment (because 4.30 needs
3964 to be released), I have disabled address rewrite logging when running a
3965 filter in a non-root, non-exim process.
3971 1. The buildconfig auxiliary program wasn't quoting the value set for
3972 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3973 not defined. This bug was masked in 4.22 by the effect that was fixed in
3976 2. Some messages that were rejected after a message id was allocated were
3977 shown as "incomplete" by exigrep. It no longer does this for messages that
3978 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3980 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3981 have allow_domain_literals set, the ID did not get logged in the <= line.
3982 Domain literals are now always recognized in Message-ID: header lines.
3984 4. The first argument for a ${extract expansion item is the key name or field
3985 number. Leading and trailing spaces in this item were not being ignored,
3986 causing some misleading effects.
3988 5. When deliver_drop_privilege was set, single queue runner processes started
3989 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3990 same command in the process it spins off) were not dropping privilege.
3992 6. When the daemon running as "exim" started a queue runner, it always
3993 re-executed Exim in the spun-off process. This is a waste of effort when
3994 deliver_drop_privilege is set. The new process now just calls the
3995 queue-runner function directly.
4001 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4004 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4005 ignored. Though the use of -oP was forcing the writing of a pid file, it
4006 was always written to the default place.
4008 3. If the message "no IP address found for host xxxx" is generated during
4009 incoming verification, it is now followed by identification of the incoming
4010 connection (so you can more easily find what provoked it).
4012 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4014 5. Added some features to "harden" Exim a bit more against certain attacks:
4016 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4017 be put in Local/Makefile. This is like the never_users runtime option,
4018 but it cannot be overridden. The default setting is "root".
4020 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4021 prefix string with which any file named in a -C command line option
4024 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4025 is retained for -C and -D only if the caller of Exim is root. Without
4026 it, the exim user may also use -C and -D and retain privilege.
4028 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4029 command line option is disabled.
4031 6. Macro names set by the -D option must start with an upper case letter, just
4032 like macro names defined in the configuration file.
4034 7. Added "dereference=" facility to LDAP.
4036 8. Two instances of the typo "uknown" in the source files are fixed.
4038 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4039 the Configure-Makefile script screwed up while processing it.
4041 10. Incorporated PCRE 4.4.
4043 11. The SMTP synchronization check was not operating right at the start of an
4044 SMTP session. For example, it could not catch a HELO sent before the client
4045 waited for the greeting. There is now a check for outstanding input at the
4046 point when the greeting is written. Because of the duplex, asynchronous
4047 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4048 way, but not yet received, when the check is performed.
4050 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4051 on TCP/IP sockets, because this apparently causes some broken clients to
4054 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4055 unchanged) from the Cygwin maintainer.
4057 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4058 was supported instead of testing for mbx. Effectively a typo.
4060 15. The spa authenticator server code was not checking that the input it
4061 received was valid base64.
4063 16. The debug output line for the "set" modifier in ACLs was not showing the
4064 name of the variable that was being set.
4066 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4068 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4069 lookup on its own. Something else was needed to trigger the lookup. For
4070 example, a match in host_lookup or the need for a host name in a host list.
4071 Now, if $sender_host_name is referenced and the host name has not yet been
4072 looked up, a lookup is performed. If the lookup fails, the variable remains
4073 empty, and $host_lookup_failed is set to "1".
4075 19. Added "eqi" as a case-independent comparison operator.
4077 20. The saslauthd authentication condition could segfault if neither service
4078 nor realm was specified.
4080 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4081 error message that was logged was misleading, and incoming SMTP
4082 connections were dropped. The message is now more accurate, and temporary
4083 errors are given to SMTP connections.
4085 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4086 (except RSET) with a 421 error, until QUIT is received. However, it was
4087 failing to send a response to QUIT.
4089 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4090 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4091 ACL is now run after the helo_try_verify_hosts code.
4093 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4094 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4095 case-independent, so other case variants are also recognized). Apparently
4096 some systems use these upper case variants.
4098 25. If more than two messages were waiting for the same host, and a transport
4099 filter was specified for the transport, Exim sent two messages over the
4100 same TCP/IP connection, and then failed with "socket operation on non-
4101 socket" when it tried to send the third.
4103 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4105 27. The extern definition of crypt16() in expand.c was not being excluded when
4106 the OS had its own crypt16() function.
4108 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4109 as a preferred synonym for return_size_limit, both as an option and as an
4112 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4114 30. Changed the default configuration ACL to relax the local part checking rule
4115 for addresses that are not in any local domains. For these addresses,
4116 slashes and pipe symbols are allowed within local parts, but the sequence
4117 /../ is explicitly forbidden.
4119 31. SPA server authentication was not clearing the challenge buffer before
4122 32. log_message in a "warn" ACL statement was writing to the reject log as
4123 well as to the main log, which contradicts the documentation and doesn't
4124 seem right (because no rejection is happening). So I have stopped it.
4126 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4127 However, I am unable to do any testing of this.
4129 34. Fixed an infelicity in the appendfile transport. When checking directories
4130 for a mailbox, to see if any needed to be created, it was accidentally
4131 using path names with one or more superfluous leading slashes; tracing
4132 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4134 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4135 contents are added to the log line that is written for every discarded
4136 recipient. (Previously a log_message setting was ignored.)
4138 36. The ${quote: operator now quotes the string if it is empty.
4140 37. The install script runs exim in order to find its version number. If for
4141 some reason other than non-existence or emptiness, which it checks, it
4142 could not run './exim', it was installing it with an empty version number,
4143 i.e. as "exim-". This error state is now caught, and the installation is
4146 38. An argument was missing from the function that creates an error message
4147 when Exim fails to connect to the socket for saslauthd authentication.
4148 This could cause Exim to crash, or give a corrupted message.
4150 39. Added isip, isip4, and isip6 to ${if conditions.
4152 40. The ACL variables $acl_xx are now saved with the message, and can be
4153 accessed later in routers, transports, and filters.
4155 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4156 strings in the file are not string-expanded.
4158 42. If a MAIL command specified a SIZE value that was too large to fit into an
4159 int variable, the check against message_size_limit failed. Such values are
4160 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4161 day this will have to be increased, but I don't think I want to be around
4162 when emails are that large.
4169 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4170 iconv() is not standard in FreeBSD.
4172 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4173 IPv6 enabled. The observed symptom was a segmentation fault on return from
4174 the function os_common_find_running_interfaces() in src/os.c.
4176 3. In the check_special_case() function in daemon.c I had used "errno" as an
4177 argument name, which causes warnings on some systems. This was basically a
4178 typo, since it was named "eno" in the comments!
4180 4. The code that waits for the clock to tick (at a resolution of some fraction
4181 of a second) so as to ensure message-id uniqueness was always waiting for
4182 at least one whole tick, when it could have waited for less. [This is
4183 almost certainly not relevant at current processor speeds, where it is
4184 unlikely to ever wait at all. But we try to future-proof.]
4186 5. The function that sleeps for a time interval that includes fractions of a
4187 second contained a race. It did not block SIGALRM between setting the
4188 timer, and suspending (a couple of lines later). If the interval was short
4189 and the sigsuspend() was delayed until after it had expired, the suspension
4190 never ended. On busy systems this could lead to processes getting stuck for
4193 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4194 process, before it forks any delivery processes. The open lookup caching
4195 mechanism meant that the open file or database connection was passed into
4196 the delivery process. The problem was that delivery processes always tidy
4197 up cached lookup data. This could cause a problem for the next delivery
4198 process started by the queue runner, because the external queue runner
4199 process does not know about the closure. So the next delivery process
4200 still has data in the lookup cache. In the case of a file lookup, there was
4201 no problem because closing a file descriptor in a subprocess doesn't affect
4202 the parent. However, if the lookup was caching a connection to a database,
4203 the connection was closed, and the second delivery process was likely to
4204 see errors such as "PGSQL: query failed: server closed the connection
4205 unexpectedly". The problem has been fixed by closing all cached lookups
4206 in a queue runner before running a delivery process.
4208 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4209 seem to have the "const" qualifier which it has on other OS. I've
4212 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4213 *of the same type* (client or server) with the same public name that an
4214 error should be diagnosed.
4216 9. When Exim looked up a host name for an IP address, but failed to find the
4217 original IP address when looking up the host name (a safety check), it
4218 output the message "<ip address> does not match any IP for NULL", which was
4219 confusing, to say the least. The bug was that the host name should have
4220 appeared instead of "NULL".
4222 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4223 user that is built into the binary, and the -C or -D options is used, root
4224 privilege is dropped before the configuration file is read. In addition,
4225 logging is switched to stderr instead of the normal log files. If the
4226 configuration then re-defines the Exim user, the unprivileged environment
4227 is probably not what is expected, so Exim logs a panic warning message (but
4230 However, if deliver_drop_privilege is set, the unprivileged state may well
4231 be exactly what is intended, so the warning has been cut out in that case,
4232 and Exim is allowed to try to write to its normal log files.
4238 1. smtp_return_error_details was not giving details for temporary sender
4239 or receiver verification errors.
4241 2. Diagnose a configuration error if two authenticators have the same public
4244 3. Exim used not to create the message log file for a message until the first
4245 delivery attempt. This could be confusing when incoming messages were held
4246 for policy or load reasons. The message log file is now created at the time
4247 the message is received, and an initial "Received" line is written to it.
4249 4. The automatically generated man page for command line options had a minor
4250 bug that caused no ill effects; however, a more serious problem was that
4251 the procedure for building the man page automatically didn't always
4252 operate. Consequently, release 4.20 contains an out-of-date version. This
4253 shouldn't happen again.
4255 5. When building Exim with embedded Perl support, the script that builds the
4256 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4257 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4259 6. The freeze_tell option was not being used for messages that were frozen on
4260 arrival, either by an ACL or by local_scan().
4262 7. Added the smtp_incomplete_transaction log selector.
4264 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4265 was accepting AUTH without a new EHLO.
4267 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4268 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4271 10. Logging of TCP/IP connections (when configured) now happens in the main
4272 daemon process instead of the child process, so that the TCP/IP connection
4273 count is more accurate (but it can never be perfect).
4275 11. The use of "drop" in a nested ACL was not being handled correctly in the
4276 outer ACL. Now, if condition failure induced by the nested "drop" causes
4277 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4278 or "require"), the connection is dropped.
4280 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4281 that yield "discard" can now be used with an "accept" or a "discard" verb,
4282 but an error is generated for any others (because I can't see a useful way
4283 to define what should happen).
4285 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4286 are now processed in the same way as lines in the Exim configuration file.
4287 In particular, continuation lines are supported.
4289 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4291 15. Added -ti meaning -t -i.
4293 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4294 domains, and warn by logging if others are found.
4296 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4297 returned from GIFCONF when Exim is trying to find the list of interfaces on
4298 a host. The code in os.c has been modified to copy each ifreq to an aligned
4299 structure in all cases.
4301 Also, in some cases, the returned ifreq's were being copied to a 'struct
4302 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4303 means the last couple of bytes of an IPv6 address could be chopped if the
4304 ifreq contained only a normal sockaddr (14 bytes storage).
4306 18. Named domain lists were not supported in the hosts_treat_as_local option.
4307 An entry such as +xxxx was not recognized, and was treated as a literal
4310 19. Ensure that header lines added by a DATA ACL are included in the reject log
4311 if the ACL subsequently rejects the message.
4313 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4314 MD5 (which is deprecated).
4316 21. When testing a filter file using -bf, Exim was writing a message when it
4317 took the sender from a "From " line in the message, but it was not doing so
4318 when it took $return_path from a Return-Path: header line. It now does.
4320 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4321 with a valid header line field name (a series of printing characters
4322 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4324 23. Changed "disc" in the source to "disk" to conform to the documentation and
4325 the book and for uniformity.
4327 24. Ignore Sendmail's -Ooption=value command line item.
4329 25. When execve() failed while trying to run a command in a pipe transport,
4330 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4331 could be confused with a return value of 69 from the command itself. This
4332 has been changed to 127, the value the shell returns if it is asked to run
4333 a non-existent command. The wording for the related log line suggests a
4334 non-existent command as the problem.
4336 26. If received_header_text expands to an empty string, do not add a Received:
4337 header line to the message. (Well, it adds a token one on the spool, but
4338 marks it "old" so that it doesn't get used or transmitted.)
4340 27. Installed eximstats 1.28 (addition of -nt option).
4342 28. There was no check for failure on the call to getsockname() in the daemon
4343 code. This can fail if there is a shortage of resources on the system, with
4344 ENOMEM, for example. A temporary error is now given on failure.
4346 29. Contrary to the C standard, it seems that in some environments, the
4347 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4348 program. Exim now does this explicitly; it affects the formatting of
4349 timestamps using strftime().
4351 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4352 complaints. I've now initialized all the variables, to avoid this.
4354 32. Header lines added by a system filter were not being "seen" during
4355 transport-time rewrites.
4357 33. The info_callback() function passed to OpenSSL is set up with type void
4358 (*)(SSL *, int, int), as described somewhere. However, when calling the
4359 function (actually a macro) that sets it up, the type void(*)() is
4360 expected. I've put in a cast to prevent warnings from picky compilers.
4362 34. If a DNS black list lookup found a CNAME record, but there were no A
4363 records associated with the domain it pointed at, Exim crashed.
4365 35. If a DNS black list lookup returned more than one A record, Exim ignored
4366 all but the first. It now scans all returned addresses if a particular IP
4367 value is being sought. In this situation, the contents of the
4368 $dnslist_value variable are a list of all the addresses, separated by a
4371 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4372 to accept a host name and all its aliases if the forward lookup for any of
4373 them yielded the IP address of the incoming connection. Now it accepts only
4374 those names whose forward lookup yields the correct IP address. Any other
4375 names are discarded. This closes a loophole whereby a rogue DNS
4376 administrator could create reverse DNS records to break through a
4377 wildcarded host restriction in an ACL.
4379 37. If a user filter or a system filter that ran in a subprocess used any of
4380 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4381 the wrong values were passed to the pipe command ($thisaddress had the
4382 value of $0, $0 had the value of $1, etc). This bug was introduced by
4383 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4385 38. Improved the line breaking for long SMTP error messages from ACLs.
4386 Previously, if there was no break point between 40 and 75 characters, Exim
4387 left the rest of the message alone. Two changes have been made: (a) I've
4388 reduced the minimum length to 35 characters; (b) if it can't find a break
4389 point between 35 and 75 characters, it looks ahead and uses the first one
4390 that it finds. This may give the occasional overlong line, but at least the
4391 remaining text gets split now.
4393 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4394 file descriptors might be low, and that setting 1000 would always raise it.
4395 It turns out that in some environments, the limit is already over 1000 and
4396 that lowering it causes trouble. So now Exim takes care not to decrease it.
4398 40. When delivering a message, the value of $return_path is set to $sender_
4399 address at the start of routing (routers may change the value). By an
4400 oversight, this default was not being set up when an address was tested by
4401 -bt or -bv, which affected the outcome if any router or filter referred to
4404 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4405 log only when "message" or "log_message" are set. However, if one of the
4406 conditions was an address verification, or a call to a nested ACL, the
4407 messages generated by the underlying test were being passed through. This
4408 no longer happens. The underlying message is available in $acl_verify_
4409 message for both "message" and "log_message" expansions, so it can be
4410 passed through if needed.
4412 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4413 new expansion $bh_ to give the encoded byte string without charset
4414 translation. Translation happens only if iconv() is available; HAVE_ICONV
4415 indicates this at build time. HEADERS_CHARSET gives the charset to
4416 translate to; headers_charset can change it in the configuration, and
4417 "headers charset" can change it in an individual filter file.
4419 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4420 that creates RFC 2047 encoded "words" labels them as that charset instead
4421 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4422 expansion operator; (ii) when Exim creates a From: line for a local
4423 message; (iii) when a header line is rewritten to include a "phrase" part.
4425 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4426 buggy, causing it to skip the first lines of messages whose message ID
4427 ended in 'D'. This would not have bitten before Exim release 4.14, because
4428 message IDs were unlikely to end in 'D' before then. The effect was to have
4429 incorrect size information for certain domains.
4431 45. #include "config.h" was missing at the start of the crypt16.c module. This
4432 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4435 46. If there was a timeout during a "random" callout check, Exim treated it as
4436 a failure of the random address, and carried on sending RSET and the real
4437 address. If the delay was just some slowness somewhere, the response to the
4438 original RCPT would be taken as a response to RSET and so on, causing
4439 mayhem of various kinds.
4441 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4442 when I implemented it. It didn't allow for the fact that some option values
4443 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4444 the right test for positive values.
4446 48. Domain names in DNS records are case-independent. Exim always looks them up
4447 in lower case. Some resolvers return domain names in exactly the case they
4448 appear in the zone file, that is, they may contain uppercase letters. Not
4449 all resolvers do this - some return always lower case. Exim was treating a
4450 change of case by a resolver as a change of domain, similar to a widening
4451 of a domain abbreviation. This triggered its re-routing code and so it was
4452 trying to route what was effectively the same domain again. This normally
4453 caused routing to fail (because the router wouldn't handle the domain
4454 twice). Now Exim checks for this case specially, and just changes the
4455 casing of the domain that it ultimately uses when it transmits the message
4458 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4461 50. If a filter generated a file delivery with a non-absolute name (possible if
4462 no home directory exists for the router), the forbid_file option was not
4465 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4466 the existing equality matching.
4468 52. Exim was using ints instead of ino_t variables in some places where it was
4469 dealing with inode numbers.
4471 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4472 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4473 called TMPDIR, and if it finds it is different, it changes its value.
4475 54. The smtp_printf() function is now made available to local_scan() so
4476 additional output lines can be written before returning. There is also an
4477 smtp_fflush() function to enable the detection of a dropped connection.
4478 The variables smtp_input and smtp_batched_input are exported to
4481 55. Changed the default runtime configuration: the message "Unknown user"
4482 has been removed from the ACL, and instead placed on the localuser router,
4483 using the cannot_route_message feature. This means that any verification
4484 failures that generate their own messages won't get overridden. Similarly,
4485 the "Unrouteable address" message that was in the ACL for unverifiable
4486 relay addresses has also been removed.
4488 56. Added hosts_avoid_esmtp to the smtp transport.
4490 57. The exicyclog script was not checking for the esoteric option
4491 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4492 will work only if exicyclog is run under the appropriate euid.
4494 58. Following a discussion on the list, the rules by which Exim recognises line
4495 endings on incoming messages have been changed. The -dropcr and drop_cr
4496 options are now no-ops, retained only for backwards compatibility. The
4497 following line terminators are recognized: LF CRLF CR. However, special
4498 processing applies to CR:
4500 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4501 nor a local message in the state where . is a terminator.
4503 (ii) If a bare CR is encountered in a header line, an extra space is added
4504 after the line terminator so as not to end the header. The reasoning
4505 behind this is that bare CRs in header lines are most likely either
4506 to be mistakes, or people trying to play silly games.
4508 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4509 was being incorrectly given as 18 bytes larger than it should have been.
4510 This is a VOB (very old bug).
4512 60. This may never have affected anything current, but just in case it has:
4513 When the local host is found other than at the start of a list of hosts,
4514 the local host, those with the same MX, and any that follow, are discarded.
4515 When the list in question was part of a longer list of hosts, the following
4516 hosts (not currently being processed) were also being discarded. This no
4517 longer happens. I'm not sure if this situation could ever has previously
4520 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4523 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4524 header. This is something that is recommended in a new Internet Draft, and
4525 is something that is documented as being done by Sendmail. There are two
4526 possible values. For messages generated by the autoreply transport, Exim
4529 Auto-Submitted: auto-replied
4531 whereas for all other generated messages (e.g. bounces) it adds
4533 Auto-Submitted: auto-generated
4535 63. The "personal" condition in filters now includes a test for the
4536 Auto-Submitted: header. If it contains the string "auto-" the message it
4537 not considered personal.
4539 64. Added rcpt_include_affixes as a generic transport option.
4541 65. Added queue_only_override (default true).
4543 66. Added the syslog_duplication option.
4545 67. If what should have been the first header line of a message consisted of
4546 a space followed by a colon, Exim was mis-interpreting it as a header line.
4547 It isn't of course - it is syntactically invalid and should therefore be
4548 treated as the start of the message body. The misbehaviour could have
4549 caused a number of strange effects, including loss of data in subsequent
4550 header lines, and spool format errors.
4552 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4553 client host had authenticated. This control can now be exercised by an ACL
4554 for more flexibility.
4556 69. By default, callouts do not happen when testing with -bh. There is now a
4557 variant, -bhc, which does actually run the callout code, including
4558 consulting and updating the callout cache.
4560 70. Added support for saslauthd authentication, courtesy of Alexander
4563 71. If statvfs() failed on the spool or log directories while checking their
4564 size for availability, Exim confusingly gave the error "space shortage".
4565 Furthermore, in debugging mode it crashed with a floating point exception.
4566 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4567 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4568 serious problem, Exim now writes to the main and panic logs when this
4569 happens, with details of the failure. It then refuses to accept the
4570 incoming message, giving the message "spool directory problem" or "log
4571 directory problem" with a 421 code for SMTP messages.
4573 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4574 0, 1, and 2 exist, because some OS complain for execs without them (see
4575 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4576 descriptors. However, the code omitted to check that the open succeeded,
4577 causing mysterious errors if for some reason the permissions on /dev/null
4578 got screwed. Now Exim writes a message to the main and panic logs, and
4579 bombs out if it can't open /dev/null.
4581 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4582 interact so that it is all more flexible. It is supposed to remain
4583 backwards compatible. Also added extra_local_interfaces.
4585 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4586 to bomb out with an assertion failure - to the client this appears as a
4587 connection drop. This problem occurs in the part of the code that was taken
4588 from the Samba project. Fortunately, the assertion is in a very simple
4589 function, so I have fixed this by reproducing the function inline in the
4590 one place where it is called, and arranging for authentication to fail
4591 instead of killing the process with assert().
4593 75. The SPA client code was not working when the server requested OEM rather
4594 than Unicode encoding.
4596 76. Added code to make require_files with a specific uid setting more usable in
4597 the case where statting the file as root fails - usually a non-root-mounted
4598 NFS file system. When this happens and the failure is EACCES, Exim now
4599 forks a subprocess and does the per-uid checking as the relevant uid.
4601 77. Added process_log_path.
4603 78. If log_file_path was not explicitly set, a setting of check_log_space or
4604 check_log_inodes was ignored.
4606 79. If a space check for the spool or log partitions fails, the incident is now
4607 logged. Of course, in the latter case the data may get lost...
4609 80. Added the %p formatting code to string_format() so that it can be used to
4610 print addresses in debug_print(). Adjusted all the address printing in the
4611 debugging in store.c to use %p rather than %d.
4613 81. There was a concern that a line of code in smtp_in.c could overflow a
4614 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4615 initially expressed, the concern was not well-founded, because trailing
4616 spaces are removed early. However, if the trailing spaces were followed by
4617 a NULL, they did not get removed, so the overflow was possible. Two fixes
4620 (a) I re-wrote the offending code in a cleaner fashion.
4621 (b) If an incoming SMTP command contains a NULL character, it is rejected
4624 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4625 runs initgroups(), so that if the Exim user is in any additional groups,
4626 they will be used during message reception.
4632 The change log for 4.20 and earlier releases has been archived.