1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.477 2007/02/08 15:16:19 ph10 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
108 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
109 fixed by 4.65/MH/01 (is this a record?) are fixed:
111 (i) An empty string was always treated as zero by the numeric comparison
112 operators. This behaviour has been restored.
114 (ii) It is documented that the numeric comparison operators always treat
115 their arguments as decimal numbers. This was broken in that numbers
116 starting with 0 were being interpreted as octal.
118 While fixing these problems I realized that there was another issue that
119 hadn't been noticed. Values of message_size_limit (both the global option
120 and the transport option) were treated as octal if they started with 0.
121 The documentation was vague. These values are now always treated as
122 decimal, and I will make that clear in the documentation.
128 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
129 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
132 MH/01 Don't check that the operands of numeric comparison operators are
133 integers when their expansion is in "skipping" mode (fixes bug
134 introduced by 4.64-PH/07).
136 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
137 child addresses, Exim now panics and dies. Previously, because the count
138 is held in a short int, deliveries were likely to be lost. As such a
139 large number of recipients for a single message is ridiculous
140 (performance will be very, very poor), I have chosen to impose a limit
141 rather than extend the field.
147 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
148 leftover -K file (the existence of which was triggered by #402).
149 While we were at it, introduced process PID as part of the -K
150 filename. This should rule out race conditions when creating
153 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
154 processing considerably. Previous code took too long for large mails,
155 triggering a timeout which in turn triggers #401.
157 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
158 in the DK code in transports.c. sendfile() is not really portable,
159 hence the _LINUX specificness.
161 TF/01 In the add_headers option to the mail command in an Exim filter,
162 there was a bug that Exim would claim a syntax error in any
163 header after the first one which had an odd number of characters
166 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
167 callout verification, Exim cached a "reject" for the entire domain. This
168 is correct for most verifications, but it is not correct for a recipient
169 verification with use_sender or use_postmaster set, because in that case
170 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
171 case of MAIL FROM:<> rejection from other early rejections (e.g.
172 rejection of HELO). When verifying a recipient using a non-null MAIL
173 address, the cache is ignored if it shows MAIL FROM:<> rejection.
174 Whatever the result of the callout, the value of the domain cache is
175 left unchanged (for any other kind of callout, getting as far as trying
176 RCPT means that the domain itself is ok).
178 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
181 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
182 manifest itself as EPIPE rather than ECONNECT. When tidying away a
183 session, the daemon ignores ECONNECT errors and logs others; it now
184 ignores EPIPE as well.
186 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
187 (quoted-printable decoding).
189 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
190 later the small subsequent patch to fix an introduced bug.
192 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
194 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
196 PH/08 An error is now given if message_size_limit is specified negative.
198 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
199 to be given (somewhat) arbitrary names.
201 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
204 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
205 miscellaneous code fixes
207 PH/10 Added the log_reject_target ACL modifier to specify where to log
210 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
211 hostname. This is wrong, because it relates to the incoming message (and
212 probably the interface on which it is arriving) and not to the outgoing
213 callout (which could be using a different interface). This has been
214 changed to use the value of the helo_data option from the smtp transport
215 instead - this is what is used when a message is actually being sent. If
216 there is no remote transport (possible with a router that sets up host
217 addresses), $smtp_active_hostname is used.
219 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
220 tweaks were necessary in order to get it to work (see also 21 below):
221 (a) The code assumed that strncpy() returns a negative number on buffer
222 overflow, which isn't the case. Replaced with Exim's string_format()
224 (b) There were several signed/unsigned issues. I just did the minimum
225 hacking in of casts. There is scope for a larger refactoring.
226 (c) The code used strcasecmp() which is not a standard C function.
227 Replaced with Exim's strcmpic() function.
228 (d) The code set only $1; it now sets $auth1 as well.
229 (e) A simple test gave the error "authentication client didn't specify
230 service in request". It would seem that Dovecot has changed its
231 interface. Fortunately there's a specification; I followed it and
232 changed what the client sends and it appears to be working now.
234 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
237 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
238 address A is aliased to B and C, where B exists and C does not. Without
239 -v the output is "A verified" because verification stops after a
240 successful redirection if more than one address is generated. However,
241 with -v the child addresses are also verified. Exim was outputting "A
242 failed to verify" and then showing the successful verification for C,
243 with its parentage. It now outputs "B failed to verify", showing B's
244 parentage before showing the successful verification of C.
246 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
247 look up a TXT record in a specific list after matching in a combined
250 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
251 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
252 they consult the DNS. I had assumed they would set it the way they
253 wanted; and indeed my experiments on Linux seem to show that in some
254 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
255 To be on the safe side, however, I have now made the interface to
256 host_find_byname() similar to host_find_bydns(), with an argument
257 containing the DNS resolver options. The host_find_byname() function now
258 sets these options at its start, just as host_find_bydns() does. The smtp
259 transport options dns_qualify_single and dns_search_parents are passed to
260 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
261 of host_find_byname() use the default settings of RES_DEFNAMES
262 (qualify_single) but not RES_DNSRCH (search_parents).
264 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
265 spool_read_header() do less string testing, by means of a preliminary
266 switch on the second character of optional "-foo" lines. (This is
267 overdue, caused by the large number of possibilities that now exist.
268 Originally there were few.) While I was there, I also converted the
269 str(n)cmp tests so they don't re-test the leading "-" and the first
270 character, in the hope this might squeeze out yet more improvement.
272 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
273 flag allowing group syntax was set by the header_syntax check but not
274 turned off, possible causing trouble later; (2) The flag was not being
275 set at all for the header_verify test, causing "group"-style headers to
276 be rejected. I have now set it in this case, and also caused header_
277 verify to ignore an empty address taken from a group. While doing this, I
278 came across some other cases where the code for allowing group syntax
279 while scanning a header line wasn't quite right (mostly, not resetting
280 the flag correctly in the right place). These bugs could have caused
281 trouble for malformed header lines. I hope it is now all correct.
283 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
284 with the "reply" argument non-NULL. The code, however (which originally
285 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
286 but it didn't always do it. This confused somebody who was copying the
287 code for some other use. I have removed all the tests.
289 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
290 feature that was used to support insecure browsers during the U.S. crypto
291 embargo. It requires special client support, and Exim is probably the
292 only MTA that supported it -- and would never use it because real RSA is
293 always available. This code has been removed, because it had the bad
294 effect of slowing Exim down by computing (never used) parameters for the
295 RSA_EXPORT functionality.
297 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
298 authenticator to fail if there's a tab character in the incoming data
299 (there should never be unless someone is messing about, as it's supposed
300 to be base64-encoded). Also added, on Timo's advice, the "secured" option
301 if the connection is using TLS or if the remote IP is the same as the
302 local IP, and the "valid-client-cert option" if a client certificate has
305 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
306 authenticators. This can be used for authorization after authentication
307 succeeds. (In the case of plaintext, it servers for both authentication
310 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
311 if any retry times were supplied.
313 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
314 connection, where there is no client IP address to check. In this
315 situation, the verify now always succeeds.
317 PH/25 Applied John Jetmore's -Mset patch.
319 PH/26 Added -bem to be like -Mset, but loading a message from a file.
321 PH/27 In a string expansion for a processed (not raw) header when multiple
322 headers of the same name were present, leading whitespace was being
323 removed from all of them, but trailing whitespace was being removed only
324 from the last one. Now trailing whitespace is removed from each header
325 before concatenation. Completely empty headers in a concatenation (as
328 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
329 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
331 PH/29 [Removed. This was a change that I later backed out, and forgot to
332 correct the ChangeLog entry (that I had efficiently created) before
333 committing the later change.]
335 PH/30 Exim was sometimes attempting to deliver messages that had suffered
336 address errors (4xx response to RCPT) over the same connection as other
337 messages routed to the same hosts. Such deliveries are always "forced",
338 so retry times are not inspected. This resulted in far too many retries
339 for the affected addresses. The effect occurred only when there were more
340 hosts than the hosts_max_try setting in the smtp transport when it had
341 the 4xx errors. Those hosts that it had tried were not added to the list
342 of hosts for which the message was waiting, so if all were tried, there
343 was no problem. Two fixes have been applied:
345 (i) If there are any address or message errors in an SMTP delivery, none
346 of the hosts (tried or untried) are now added to the list of hosts
347 for which the message is waiting, so the message should not be a
348 candidate for sending over the same connection that was used for a
349 successful delivery of some other message. This seems entirely
350 reasonable: after all the message is NOT "waiting for some host".
351 This is so "obvious" that I'm not sure why it wasn't done
352 previously. Hope I haven't missed anything, but it can't do any
353 harm, as the worst effect is to miss an optimization.
355 (ii) If, despite (i), such a delivery is accidentally attempted, the
356 routing retry time is respected, so at least it doesn't keep
357 hammering the server.
359 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
360 in ${readsocket because some servers need this prod.
362 PH/32 Added some extra debug output when updating a wait-xxx database.
364 PH/33 The hint "could be header name not terminated by colon", which has been
365 given for certain expansion errors for a long time, was not being given
366 for the ${if def:h_colon_omitted{... case.
368 PH/34 The spec says: "With one important exception, whenever a domain list is
369 being scanned, $domain contains the subject domain." There was at least
370 one case where this was not true.
372 PH/35 The error "getsockname() failed: connection reset by peer" was being
373 written to the panic log as well as the main log, but it isn't really
374 panic-worthy as it just means the connection died rather early on. I have
375 removed the panic log writing for the ECONNRESET error when getsockname()
378 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
379 runs only) independently of the message's sender address. This meant
380 that, if the 4xx error was in fact related to the sender, a different
381 message to the same recipient with a different sender could confuse
382 things. In particualar, this can happen when sending to a greylisting
383 server, but other circumstances could also provoke similar problems.
384 I have changed the default so that the retry time for these errors is now
385 based a combination of the sender and recipient addresses. This change
386 can be overridden by setting address_retry_include_sender=false in the
389 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
390 remote server are returned as part of bounce messages. This was not
391 happening for LMTP over a pipe (the lmtp transport), but now it is the
392 same for both kinds of LMTP.
394 PH/38 Despite being documented as not happening, Exim was rewriting addresses
395 in header lines that were in fact CNAMEs. This is no longer the case.
397 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
398 and queue runs started by the daemon processed all messages. This has
399 been fixed so that -R and -S can now usefully be given with -q<time>.
401 PH/40 Import PCRE release 6.7 (fixes some bugs).
403 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
405 PH/42 Give an error if -q is specified more than once.
407 PH/43 Renamed the variables $interface_address and $interface_port as
408 $received_ip_address and $received_port, to make it clear that these
409 values apply to message reception, and not to the outgoing interface when
410 a message is delivered. (The old names remain recognized, of course.)
412 PH/44 There was no timeout on the connect() call when using a Unix domain
413 socket in the ${readsocket expansion. There now is.
415 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
416 be meaningful with "accept".
418 SC/01 Eximstats V1.43
419 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
421 SC/02 Eximstats V1.44
422 Use a glob alias rather than an array ref in the generated
423 parser. This improves both readability and performance.
425 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
426 Collect SpamAssassin and rejection statistics.
427 Don't display local sender or destination tables unless
428 there is data to show.
429 Added average volumes into the top table text output.
431 SC/04 Eximstats V1.46
432 Collect data on the number of addresses (recipients)
433 as well as the number of messages.
435 SC/05 Eximstats V1.47
436 Added 'Message too big' to the list of mail rejection
437 reasons (thanks to Marco Gaiarin).
439 SC/06 Eximstats V1.48
440 Mainlog lines which have GMT offsets and are too short to
441 have a flag are now skipped.
443 SC/07 Eximstats V1.49 (Alain Williams)
444 Added the -emptyok flag.
446 SC/08 Eximstats V1.50
447 Fixes for obtaining the IP address from reject messages.
449 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
450 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
451 whitesspace changes from 4.64-PH/27
453 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
456 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
457 are found, allow negative numbers in numeric criteria)
459 JJ/06 exipick.20061117.2, added new $message_body_missing variable
461 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
462 to match changes made in 4.64-PH/43
464 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
466 PH/47 Put in an explicit test for a DNS lookup of an address record where the
467 "domain" is actually an IP address, and force a failure. This locks out
468 those revolvers/nameservers that support "A-for-A" lookups, in
469 contravention of the specifications.
471 PH/48 When a host name was looked up from an IP address, and the subsequent
472 forward lookup of the name timed out, the host name was left in
473 $sender_host_name, contrary to the specification.
475 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
476 restricted to single-key lookups, Exim was not diagnosing an error if
477 * or *@ was used with a query-style lookup.
479 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
481 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
482 long ago, but noone interested enough thought of it. Let's just say that
483 the "1.1" means that there are some new functions that weren't there at
484 some point in the past.
486 PH/51 Error processing for expansion failure of helo_data from an smtp
487 transport during callout processing was broken.
489 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
490 tested/used via the -bh/-bhc/-bs options.
492 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
493 bug, fixed in subsequent PCRE releases).
495 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
496 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
498 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
504 SC/01 Use a glob alias rather than an array ref in eximstats generated
505 parser. This improves both readability and performance.
507 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
508 Don't display local sender or destination tables in eximstats unless
509 there is data to show.
510 Added average volumes into the eximstats top table text output.
512 SC/03 Collect data on the number of addresses (recipients) as well
513 as the number of messages in eximstats.
515 TF/01 Correct an error in the documentation for the redirect router. Exim
516 does (usually) call initgroups() when daemonizing.
518 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
519 with consistent privilege compared to when running as a daemon.
521 TF/03 Note in the spec that $authenticated_id is not set for local
522 submissions from trusted users.
524 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
525 Thanks to Dean Brooks <dean@iglou.com> for the patch.
527 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
528 by adding some example configuration directives to the default
529 configuration file. A little bit of work is required to uncomment the
530 directives and define how usernames and passwords are checked, but
531 there is now a framework to start from.
533 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
534 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
535 without this. I don't know how relevant this is to other LDAP libraries.
537 PH/02 Add the verb name to the "unknown ACL verb" error.
539 PH/03 Magnus Holmgren's patch for filter_prepend_home.
541 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
543 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
544 directory not expanded when it should be if an expanded home directory
545 was set for the address (which is overridden by the transport).
547 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
550 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
551 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
552 because it is too late at that time, and has no effect.
554 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
555 security issue with \' (bugzilla #107). I could not use the
556 PQescapeStringConn() function, because it needs a PGconn value as one of
559 PH/08 When testing addresses using -bt, indicate those final addresses that
560 are duplicates that would not cause an additional delivery. At least one
561 person was confused, thinking that -bt output corresponded to deliveries.
562 (Suppressing duplicates isn't a good idea as you lose the information
563 about possibly different redirections that led to the duplicates.)
565 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
566 systems where poll() doesn't work, in particular OS X.
568 PH/10 Added more information to debugging output for retry time not reached.
570 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
571 operations in malware.c.
573 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
576 PH/13 If write_rejectlog was set false when logging was sent to syslog with
577 syslog_duplication set false, log lines that would normally be written
578 both the the main log and to the reject log were not written to syslog at
581 PH/14 In the default configuration, change the use of "message" in ACL warn
582 statements to "add_header".
584 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
585 not followed by a command (e.g. "seen endif").
587 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
588 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
591 PH/17 Added extra conditions to the default value of delay_warning_condition
595 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
596 { match{$h_precedence:}{(?i)bulk|list|junk} } \
597 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
600 The Auto-Submitted: and various List- headers are standardised, whereas I
601 don't think Precedence: ever was.
603 PH/18 Refactored debugging code in route_finduser() to show more information,
604 in particular, the error code if getpwnam() issues one.
606 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
607 This is apparently needed in addition to the PH/07 change above to avoid
608 any possible encoding problems.
610 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
611 but not after initializing Perl.
613 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
614 output them only if debugging. By default they are written stderr,
615 apparently, which is not desirable.
617 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
620 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
623 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
625 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
626 authenticated or an ident call has been made. Suppress the default
627 values for $authenticated_id and $authenticated_sender (but permit -oMai
628 and -oMas) when testing with -bh.
630 PH/24 Re-jigged the order of the tests in the default configuration so that the
631 tests for valid domains and recipients precede the DNS black list and CSA
632 tests, on the grounds that those ones are more expensive.
634 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
635 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
636 This bug exists in every version of Exim that I still have, right back to
639 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
640 However, an attempt to turn on case-sensitivity in a regex key by
641 including (?-i) didn't work because the subject string was already
642 lowercased, and the effects were non-intuitive. It turns out that a
643 one-line patch can be used to allow (?-i) to work as expected.
649 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
650 other effects) broke the use of negated acl sub-conditions.
652 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
655 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
656 "Deny" causes Exim to reject the incoming connection with a 554 error.
657 Unfortunately, if there is a major crisis, such as a disk failure,
658 tcp-wrappers gives "deny", whereas what one would like would be some
659 kind of temporary error. A kludge has been added to help with this.
660 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
661 554 error is used if errno is still zero or contains ENOENT (which occurs
662 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
665 PH/03 Add -lutil to the default FreeBSD LIBS setting.
667 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
668 errors. Otherwise a message that provokes a temporary error (when other
669 messages do not) can cause a whole host to time out.
671 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
672 addresses were routed directly to files or pipes from a redirect router.
673 File deliveries just didn't batch; pipe deliveries might have suffered
676 PH/06 A failure to get a lock for a hints database would erroneously always say
677 "Failed to get write lock", even when it was really a read lock.
679 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
680 of 0600. This has been changed to use the value of the lockfile_mode
681 option (which defaults to 0600).
683 PH/08 Applied small patch from the Sieve maintainer.
685 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
686 folder from quota calculations, a direct delivery into this folder messed
687 up the contents of the maildirsize file. This was because the regex was
688 used only to exclude .Trash (or whatever) when the size of the mailbox
689 was calculated. There was no check that a delivery was happening into an
690 excluded directory. This bug has been fixed by ignoring all quota
691 processing for deliveries into excluded directories.
693 PH/10 Added the maildirfolder_create_regex option to appendfile.
699 PH/01 The code for finding all the local interface addresses on a FreeBSD
700 system running IPv6 was broken. This may well have applied to all BSD
701 systems, as well as to others that have similar system calls. The broken
702 code found IPv4 interfaces correctly, but gave incorrect values for the
703 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
704 that it would not match correctly against @[] and not recognize the IPv6
707 PH/02 The ipliteral router was not recognizing addresses of the form user@
708 [ipv6:....] because it didn't know about the "ipv6:" prefix.
710 PH/03 Added disable_ipv6.
712 PH/04 Changed $reply_address to use the raw form of the headers instead of the
713 decoded form, because it is most often used to construct To: headers
714 lines in autoreplies, and the decoded form may well be syntactically
715 invalid. However, $reply_address has leading white space removed, and all
716 newlines turned into spaces so that the autoreply transport does not
719 PH/05 If group was specified without a user on a router, and no group or user
720 was specified on a transport, the group from the router was ignored.
722 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
723 for visible compile-time settings that can be used to change these
724 numbers, for those that want even more. Backwards compatibility with old
725 spool files has been maintained. However, going back to a previous Exim
726 release will lost any variables that are in spool files.
728 PH/07 Two small changes when running in the test harness: increase delay when
729 passing a TCP/IP connection to a new process, in case the original
730 process has to generate a bounce, and remove special handling of
731 127.0.0.2 (sic), which is no longer necessary.
733 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
734 be the same on different OS.
736 PH/09 Moved a debug statement in filter processing to avoid a race problem when
739 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
740 whether --show-vars was specified or not
742 JJ/02 exipick: Added support for new ACL variable spool format introduced
745 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
746 syntactically invalid From: or Reply-to: line, and a filter used this to
747 generate an autoreply, and therefore failed to obtain an address for the
748 autoreply, Exim could try to deliver to a non-existent relative file
749 name, causing unrelated and misleading errors. What now happens is that
750 it logs this as a hard delivery error, but does not attempt to create a
753 PH/11 The exinext utility has a -C option for testing purposes, but although
754 the given file was scanned by exinext itself; it wasn't being passed on
755 when Exim was called.
757 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
758 an end-of-file indication when reading a command response.
760 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
761 compiled. In many other places in Exim, IPv6 addresses are always
762 recognized, so I have changed this. It also means that IPv4 domain
763 literals of the form [IPV4:n.n.n.n] are now always recognized.
765 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
766 used if the router is not running as root, for example, when verifying at
767 ACL time, or when using -bh. The debugging output from this situation was
768 non-existent - all you got was a failure to exec. I have made two
771 (a) Failures to set uid/gid, the current directory, or a process leader
772 in a subprocess such as that created by queryprogram now generate
773 suitable debugging ouput when -d is set.
775 (b) The queryprogram router detects when it is not running as root,
776 outputs suitable debugging information if -d is set, and then runs
777 the subprocess without attempting to change uid/gid.
779 PH/15 Minor change to Makefile for building test_host (undocumented testing
782 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
783 additional section of a DNS packet that returns MX or SRV records.
784 Instead, it always explicitly searches for A/AAAA records. This avoids
785 major problems that occur when a DNS server includes only records of one
786 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
787 fixed another bug: if SRV records were looked up and the corresponding
788 address records were *not* found in the additional section, the port
789 values from the SRV records were lost.
791 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
792 using the correct key (the original address) when searching the retry
793 rules in order to find which one to use for generating the retry hint.
795 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
796 adding the default one. Similarly, if it contains a Reply-To: header, the
797 errors_reply_to option, if set, is not used.
799 PH/19 When calculating a retry time, Exim used to measure the "time since
800 failure" by looking at the "first failed" field in the retry record. Now
801 it does not use this if it is later than than the arrival time of the
802 message. Instead it uses the arrival time. This makes for better
803 behaviour in cases where some deliveries succeed, thus re-setting the
804 "first failed" field. An example is a quota failure for a huge message
805 when small messages continue to be delivered. Without this change, the
806 "time since failure" will always be short, possible causing more frequent
807 delivery attempts for the huge message than are intended.
808 [Note: This change was subsequently modified - see PH/04 for 4.62.]
810 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
811 $1, $2, $3) because the numerical variables can be reset during some
812 expansion items (e.g. "match"), thereby losing the authentication data.
814 PH/21 Make -bV show the size of off_t variables so that the test suite can
815 decide whether to run tests for quotas > 2G.
817 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
818 mailbox_size, and mailbox_filecount in the appendfile transport. If a
819 filecount value is greater than 2G or if a quota value is greater than 2G
820 on a system where the size of off_t is not greater than 4, a panic error
823 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
824 never match. The debug and -bh output now contains an explicit error
825 message indicating a malformed IPv4 address or mask.
827 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
828 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
831 PH/25 Do not write to syslog when running in the test harness. The only
832 occasion when this arises is a failure to open the main or panic logs
833 (for which there is an explicit test).
835 PH/26 Added the /no_tell option to "control=freeze".
837 PH/27 If a host name lookup failed very early in a connection, for example, if
838 the IP address matched host_lookup and the reverse lookup yielded a name
839 that did not have a forward lookup, an error message of the form "no IP
840 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
841 could be logged. Now it outputs the IP address instead of "NULL".
843 PH/28 An enabling patch from MH: add new function child_open_exim2() which
844 allows the sender and the authenticated sender to be set when
845 submitting a message from within Exim. Since child_open_exim() is
846 documented for local_scan(), the new function should be too.
848 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
849 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
850 results in an empty string is now treated as unset.
852 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
854 PH/31 Added sender_verify_fail logging option.
856 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
857 needed by RFC 822 but not by RFC 2822 was commented out. I have now
858 tidied the source and removed it altogether.
860 PH/33 When a queue run was abandoned because the load average was too high, a
861 log line was always written; now it is written only if the queue_run log
862 selector is set. In addition, the log line for abandonment now contains
863 information about the queue run such as the pid. This is always present
864 in "start" and "stop" lines but was omitted from the "abandon" line.
866 PH/34 Omit spaces between a header name and the colon in the error message that
867 is given when verify = headers_syntax fails (if there are lots of them,
868 the message gets confusing).
870 PH/35 Change the default for dns_check_names_pattern to allow slashes within
871 names, as there are now some PTR records that contain slashes. This check
872 is only to protect against broken name servers that fall over on strange
873 characters, so the fact that it applies to all lookups doesn't matter.
875 PH/36 Now that the new test suite is complete, we can remove some of the
876 special code in Exim that was needed for the old test suite. For example,
877 sorting DNS records because real resolvers return them in an arbitrary
878 order. The new test suite's fake resolver always returns records in the
881 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
882 bounces) except when queue_only is set, to avoid logging races between
883 the different processes.
885 PH/38 Panic-die if .include specifies a non-absolute path.
887 PH/39 A tweak to the "H" retry rule from its user.
889 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
890 a label. They prevented compilation on older perls.
892 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
893 a warning to be raised on newish perls.
895 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
896 on queue. Changes to match documented behaviour of showing count of
897 messages matching specified criteria.
899 PH/40 Changed the default ident timeout from 30s to 5s.
901 PH/41 Added support for the use of login_cap features, on those BSD systems
902 that have them, for controlling the resources used by pipe deliveries.
904 PH/42 The content-scanning code uses fopen() to create files in which to put
905 message data. Previously it was not paying any attention to the mode of
906 the files. Exim runs with umask(0) because the rest of the code creates
907 files with open(), and sets the required mode explicitly. Thus, these
908 files were ending up world-writeable. This was not a big issue, because,
909 being within the spool directory, they were not world-accessible. I have
910 created a function called modefopen, which takes an additional mode
911 argument. It sets umask(777), creates the file, chmods it to the required
912 mode, then resets the umask. All the relevant calls to fopen() in the
913 content scanning code have been changed to use this function.
915 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
916 to 24 hours. This avoids potential overflow problems when processing G
917 and H retry rules. I suspect nobody ever tinkers with this value.
919 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
921 PH/45 When the plaintext authenticator is running as a client, the server's
922 challenges are checked to ensure they are valid base64 strings. By
923 default, the authentication attempt is cancelled if an invalid string is
924 received. Setting client_ignore_invalid_base64 true ignores these errors.
925 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
926 they are received. Thus, the responses can be made to depend on the
927 challenges. If an invalid string is ignored, an empty string is placed in
930 PH/46 Messages that are created by the autoreply transport now contains a
931 References: header, in accordance with RFCs 2822 and 3834.
933 PH/47 Added authenticated_sender_force to the smtp transport.
935 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
937 PH/49 Installed latest patch from the Sieve maintainer.
939 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
940 was also set, the appendfile transport was unnecessarily scanning a
941 directory of message files (e.g. for maildir delivery) to find the count
942 of files (along with the size), even though it did not need this
943 information. It now does the scan only if it needs to find either the
944 size of the count of files.
946 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
948 PH/52 Two bugs concerned with error handling when the smtp transport is
951 (i) Exim was not creating retry information for temporary errors given
952 for individual recipients after the DATA command when the smtp transport
953 was used in LMTP mode. This meant that they could be retried too
954 frequently, and not timed out correctly.
956 (ii) Exim was setting the flag that allows error details to be returned
957 for LMTP errors on RCPT commands, but not for LMTP errors for individual
958 recipients that were returned after the DATA command.
960 PH/53 This is related to PH/52, but is more general: for any failing address,
961 when detailed error information was permitted to be returned to the
962 sender, but the error was temporary, then after the final timeout, only
963 "retry timeout exceeded" was returned. Now it returns the full error as
964 well as "retry timeout exceeded".
966 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
967 do this, and (what is worse) MTAs that accept it.
969 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
970 will now be deprecated.
972 PH/56 New os.c-cygwin from the Cygwin maintainer.
974 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
975 formats (previously only available in exim formats via -bpr, -bpru,
976 and -bpra. Now also available in native and exiqgrep formats)
978 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
979 with very large, slow to parse queues
981 JJ/08 exipick: added ! as generic prefix to negate any criteria format
983 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
985 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
986 responses to authentication challenges, though it was showing the
987 challenges; (ii) I've removed the CR characters from the debug output for
990 PH/58 Allow for the insertion of a newline as well as a space when a string
991 is turned into more than one encoded-word during RFC 2047 encoding. The
992 Sieve code now uses this.
994 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
995 data_4xx, lost_connection, tls_required.
997 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
998 message was being sent as an SMTP response.
1000 PH/61 Add -l and -k options to exicyclog.
1002 PH/62 When verifying, if an address was redirected to one new address, so that
1003 verification continued, and the new address failed or deferred after
1004 having set something in $address_data, the value of $address_data was not
1005 passed back to the ACL. This was different to the case when no
1006 redirection occurred. The value is now passed back in both cases.
1008 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1009 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1010 use login_cap.h, so on its own it isn't the distinguishing feature. The
1011 new name refers directly to the setclassresources() function.
1013 PH/65 Added configuration files for NetBSD3.
1015 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1017 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1018 is preferred over IPv4.
1020 PH/68 The bounce_return_message and bounce_return_body options were not being
1021 honoured for bounces generated during the reception of non-SMTP messages.
1022 In particular, this applied to messages rejected by the ACL. This bug has
1023 been fixed. However, if bounce_return_message is true and bounce_return_
1024 body is false, the headers that are returned for a non-SMTP message
1025 include only those that have been read before the error was detected.
1026 (In the case of an ACL rejection, they have all been read.)
1028 PH/69 The HTML version of the specification is now built in a directory called
1029 spec_html instead of spec.html, because the latter looks like a path with
1030 a MIME-type, and this confuses some software.
1032 PH/70 Catch two compiler warnings in sieve.c.
1034 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1035 function verify_get_ident() calls ip_connect() to connect a socket, but
1036 if the "connect()" function timed out, ip_connect() used to close the
1037 socket. However, verify_get_ident() also closes the socket later, and in
1038 between Exim writes to the log, which may get opened at this point. When
1039 the socket was closed in ip_connect(), the log could get the same file
1040 descriptor number as the socket. This naturally causes chaos. The fix is
1041 not to close the socket in ip_connect(); the socket should be closed by
1042 the function that creates it. There was only one place in the code where
1043 this was missing, in the iplookup router, which I don't think anybody now
1044 uses, but I've fixed it anyway.
1046 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1047 well as to direct DNS lookups. Otherwise the handling of names in host
1048 lists is inconsistent and therefore confusing.
1054 PH/01 Two changes to the default runtime configuration:
1056 (1) Move the checks for relay_from_hosts and authenticated clients from
1057 after to before the (commented out) DNS black list checks.
1059 (2) Add control=submission to the relay_from_hosts and authenticated
1060 clients checks, on the grounds that messages accepted by these
1061 statements are most likely to be submissions.
1063 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1065 (1) Generate an error if the third argument for the ${prvs expansion is
1068 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1071 (3) Reset the variables that are obtained from the first argument of
1072 ${prvscheck and used in the second argument before leaving the code,
1073 because their memory is reclaimed, so using them afterwards may do
1076 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1077 one (it's much easier than Tom thought :-).
1079 (5) Because of (4), we can now allow for the use of $prvscheck_result
1080 inside the third argument.
1082 PH/03 For some reason, the default setting of PATH when running a command from
1083 a pipe transport was just "/usr/bin". I have changed it to
1086 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1087 anything to be listed in the output from -bV.
1089 PH/05 When a filter generated an autoreply, the entire To: header line was
1090 quoted in the delivery log line, like this:
1092 => >A.N.Other <ano@some.domain> <original@ddress> ...
1094 This has been changed so that it extracts the operative address. There
1095 may be more than one such address. If so, they are comma-separated, like
1098 => >ano@some.domain,ona@other.domain <original@ddress> ...
1100 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1101 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1102 not being looked up in the rDNS to get a host name, Exim was showing the
1103 IP address twice in Received: lines, even though the IP addresses were
1104 identical. For example:
1106 Received: from [1.2.3.4] (helo=[1.2.3.4])
1108 However, if the real host name was known, it was omitting the HELO data
1109 if it matched the actual IP address. This has been tidied up so that it
1110 doesn't show the same IP address twice.
1112 PH/07 When both +timestamp and +memory debugging was on, the value given by
1113 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1114 called by the debug printing, thereby overwriting the timestamp buffer.
1115 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1117 PH/08 When the original message was included in an autoreply transport, it
1118 always said "this is a copy of the message, including all the headers",
1119 even if body_only or headers_only was set. It now gives an appropriate
1122 PH/09 Applied a patch from the Sieve maintainer which:
1124 o fixes some comments
1125 o adds the (disabled) notify extension core
1126 o adds some debug output for the result of if/elsif tests
1127 o points to the current vacation draft in the documentation
1128 and documents the missing references header update
1132 o fixes a bug in processing the envelope test (when testing
1133 multiple envelope elements, the last element determinted the
1136 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1137 Electronic Mail") by including:
1139 Auto-submitted: auto-generated
1141 in the messages that it generates (bounce messages and others, such as
1142 warnings). In the case of bounce messages for non-SMTP mesages, there was
1143 also a typo: it was using "Auto_submitted" (underscore instead of
1144 hyphen). Since every message generated by Exim is necessarily in response
1145 to another message, thes have all been changed to:
1147 Auto-Submitted: auto-replied
1149 in accordance with these statements in the RFC:
1151 The auto-replied keyword:
1153 - SHOULD be used on messages sent in direct response to another
1154 message by an automatic process,
1156 - MUST NOT be used on manually-generated messages,
1158 - MAY be used on Delivery Status Notifications (DSNs) and Message
1159 Disposition Notifications (MDNs),
1161 - MUST NOT be used on messages generated by automatic or periodic
1162 processes, except for messages which are automatic responses to
1165 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1166 to the default Received: header definition.
1168 PH/12 Added log selector acl_warn_skipped (default on).
1170 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1171 variables because (a) they are in the wrong storage pool and (b) even if
1172 they were copied, it wouldn't work properly because of the caching.
1174 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1175 checking when decoding. Apparently there are clients that generate
1176 overlong encoded strings. Why am I not surprised?
1178 PH/15 If the first argument of "${if match_address" was not empty, but did not
1179 contain an "@" character, Exim crashed. Now it writes a panic log message
1180 and treats the condition as false.
1182 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1184 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1185 extension "envelope-auth". The code is finished and in agreement with
1186 other implementations, but there is no documentation so far and in fact,
1187 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1188 not changing the active code.
1190 Print executed "if" and "elsif" statements when debugging is used. This
1191 helps a great deal to understand what a filter does.
1193 Document more things not specified clearly in RFC3028. I had all this
1194 sorted out, when out of a sudden new issues came to my mind. Oops."
1196 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1199 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1200 canonical form (as documented). However, after a host name lookup from
1201 the IP address, check_host() was doing a simple string comparison with
1202 addresses acquired from the DNS when checking that the found name did
1203 have the original IP as one of its addresses. Since any found IPv6
1204 addresses are likely to be in abbreviated form, the comparison could
1205 fail. Luckily, there already exists a function for doing the comparison
1206 by converting both addresses to binary, so now that is used instead of
1207 the text comparison.
1209 PH/20 There was another similar case to PH/19, when a complete host name was
1210 given in a host list; looking up its IP address could give an abbreviated
1211 form, whereas the current host's name might or might not be abbreviated.
1212 The same fix has been applied.
1218 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1219 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1222 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1223 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1225 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1227 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1228 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1229 Also, for "Auto-Submitted", treat anything other than "no" as
1230 non-personal, in accordance with RFC 3834. (Previously it treated
1231 anything starting "auto-" as non-personal.)
1233 TF/01 The control=submission/name=... option had a problem with syntax
1234 errors if the name included a slash character. The /name= option
1235 now slurps the rest of the string, so it can include any characters
1236 but it must come last in the list of options (after /sender_retain
1239 PH/05 Some modifications to the interface to the fake nameserver for the new
1247 TK/01 Added the "success_on_redirect" address verification option. See
1248 NewStuff for rationale and an example.
1250 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1252 PH/02 Patch to exigrep to allow it to work on syslog lines.
1254 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1255 fread() to skip over the body file's header line, because in Cygwin the
1256 header line is locked and is inaccessible.
1258 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1259 co-exist for some time) to make it clear that it is the Exim ID that is
1260 referenced, not the Message-ID: header line.
1262 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1263 string_format() function, because snprintf() does not exist on all
1266 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1267 ${stat: expansion item.
1269 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1270 protocol synchronization error", to keep the pedants happy.
1272 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1273 well as for IRIX systems, when gcc is being used. See the host.c source
1276 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1278 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1281 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1282 transport and to the smtp transport in LMTP mode.
1284 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1286 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1287 run a filter in a subprocess. This could lead to confusion in subsequent
1288 lookups in the parent process. There should also be a search_tidyup() at
1289 the end of the subprocess.
1291 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1292 only if the host matched helo_try_verify_hosts, which caused the
1293 verification to occur when the EHLO/HELO command was issued. The ACL just
1294 tested the remembered result. Now, if a previous verification attempt has
1295 not happened, "verify = helo" does it there and then.
1297 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1299 TK/03 Fix log output including CR from clamd.
1301 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1302 exist provoked a memory error which could cause a segfault.
1304 PH/15 Installed PCRE 6.2
1306 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1308 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1309 of the problem. Specifically, suggested +O2 rather than +O1 for the
1312 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1314 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1315 record, the port was not correctly logged when the outgoing_port log
1316 selector was set (it logged the transort's default port).
1318 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1319 fallback_hosts, and "hosts" in the smtp transport.
1321 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1322 host errors such as "Connection refused".
1324 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1325 authentication with radiusclient 0.4.9:
1327 - Error returned from rc_read_config was caught wrongly
1328 - Username/password not passed on to radius server due to wrong length.
1330 The presumption is that some radiusclient API changes for 4.51/PH/17
1331 were not taken care of correctly. The code is still untested by me (my
1332 Linux distribution still has 0.3.2 of radiusclient), but it was
1333 contributed by a Radius user.
1335 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1336 expanding the "port" option of the smtp transport.
1338 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1339 while reading a MIME header. Thanks to Tom Hughes for a patch.
1341 PH/24 Include config.h inside local_scan.h so that configuration settings are
1344 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1345 that in an ACL for RCPT (for example), you can examine exactly what was
1348 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1349 commands, but it was not correctly comparing the address with the actual
1350 client host address. Thus, it would show the EHLO address in Received:
1351 header lines when this was not necessary.
1353 PH/27 Added the % operator to ${eval:}.
1355 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1356 it should be ignoring failures (because with -C, for example, it has lost
1357 privilege). It wasn't ignoring creation failures other than "already
1360 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1363 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1364 because an input line was too long, either on its own, or by virtue of
1365 too many continuations, the temporary file was not being removed, and the
1366 return code was incorrect.
1368 PH/31 Missing "BOOL" in function definition in filtertest.c.
1370 PH/32 Applied Sieve patches from the maintainer.
1372 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1374 PH/33 Added "verify = not_blind".
1376 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1377 Local/Makefile (with some defaults set). These are used in built scripts
1378 such as exicyclog, but they have never been used in the exim_install
1379 script (though there are many overriding facilities there). I have
1380 arranged that the exim_install script now takes note of these two
1383 PH/35 Installed configuration files for Dragonfly.
1385 PH/36 When a locally submitted message by a trusted user did not contain a
1386 From: header, and the sender address was obtained from -f or from an SMTP
1387 MAIL command, and the trusted user did not use -F to supply a sender
1388 name, $originator_name was incorrectly used when constructing a From:
1389 header. Furthermore, $originator_name was used for submission mode
1390 messages from external hosts without From: headers in a similar way,
1391 which is clearly wrong.
1393 PH/37 Added control=suppress_local_fixups.
1395 PH/38 When log_selector = +received_sender was set, and the addition of the
1396 sender made the log line's construction buffer exactly full, or one byte
1397 less than full, an overflow happened when the terminating "\n" was
1400 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1401 when the result of a list match is failure because a DNS lookup failed.
1403 PH/40 RM_COMMAND is now used in the building process.
1405 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1406 the "build-* directories that it finds.
1408 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1409 address in a domain literal was a prefix of an interface address.
1411 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1412 when verifying a sender address, unless rewrite_headers is false.
1414 PH/44 Wrote a long comment about why errors_to addresses are verified as
1415 recipients, not senders.
1417 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1418 the ratelimit ACL was added.
1420 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1422 PH/46 Added extra information about PostgreSQL errors to the error string.
1424 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1425 avoiding the need to install special zones in a real server. This is
1426 backwards compatible; if it can't find the fake resolver, it drops back.
1427 Thus, both old and new test suites can be run.
1429 TF/02 Added util/ratelimit.pl
1431 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1432 clock is set back in time.
1434 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1435 Candler <B.Candler@pobox.com>.
1437 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1438 OK for addresses that are the result of redirections.
1440 PH/48 A number of further additions for the benefit of the new test suite,
1441 including a fake gethostbyname() that interfaces to the fake DNS resolver
1444 TF/06 The fix for widen_domains has also been applied to qualify_single and
1445 search_parents which are the other dnslookup options that can cause
1448 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1451 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1453 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1454 service" if it encountered an MX record with an empty target hostname.
1455 The message is now "an MX or SRV record indicated no SMTP service".
1457 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1458 if the DNS of the sending site is misconfigured. This is quite a
1459 common situation. This change restores the behaviour of treating a
1460 helo verification defer as a failure.
1462 PH/51 If self=fail was set on a router, the bounce message did not include the
1463 actual error message.
1469 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1471 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1472 command itself ended in error, the underlying message about the transport
1473 filter timeout was being overwritten with the pipe command error. Now the
1474 underlying error message should be appended to the second error message.
1476 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1478 PH/02 Reduce the amount of output that "make" produces by default. Full output
1479 can still be requested.
1481 PH/03 The warning log line about a condition test deferring for a "warn" verb
1482 was being output only once per connection, rather than after each
1483 occurrence (because it was using the same function as for successful
1484 "warn" verbs). This seems wrong, so I have changed it.
1486 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1487 it should not have, which might have caused a crash in the right
1488 circumstances, but probably never did.
1490 PH/04 Installed a modified version of Tony Finch's patch to make submission
1491 mode fix the return path as well as the Sender: header line, and to
1492 add a /name= option so that you can make the user's friendly name appear
1495 TF/03 Added the control = fakedefer ACL modifier.
1497 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1498 Mark Lowes for thorough testing.
1500 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1502 TK/03 Merged latest SRS patch from Miles Wilton.
1504 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1505 with the definition in sysexits.h (which is #included earlier).
1506 Fortunately, Exim does not actually use EX_OK. The code used to try to
1507 preserve the sysexits.h value, by assumimg that macro definitions were
1508 scanned for macro replacements. I have been disabused of this notion,
1509 so now the code just undefines EX_OK before #including unistd.h.
1511 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1512 in the smtp transport. When a block could not be written in a single
1513 write() function, the timeout was being re-applied to each part-write.
1514 This seems wrong - if the receiver was accepting one byte at a time it
1515 would take for ever. The timeout is now adjusted when this happens. It
1516 doesn't have to be particularly precise.
1518 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1519 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1521 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1522 without a domain if the check to <postmaster@domain> fails.
1524 SC/01 Eximstats: added -xls and the ability to specify output files
1525 (patch written by Frank Heydlauf).
1527 SC/02 Eximstats: use FileHandles for outputing results.
1529 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1531 SC/04 Eximstats: fixed display of large numbers with -nvr option
1533 SC/05 Eximstats: fixed merging of reports with empty tables.
1535 SC/06 Eximstats: added the -include_original_destination flag
1537 SC/07 Eximstats: removed tabs and trailing whitespace.
1539 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1541 TK/06 MBOX spool code: Add real "From " MBOX separator line
1542 so the .eml file is really in mbox format (even though
1543 most programs do not really care). Patch from Alex Miller.
1545 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1546 The latter is generated from $received_to and is only set if the
1547 message has one envelope recipient. SA can use these headers,
1548 obviously out-of-the-box. Patch from Alex Miller.
1550 PH/08 The ${def test on a variable was returning false if the variable's
1551 value was "0", contrary to what the specification has always said!
1552 The result should be true unless the variable is empty.
1554 PH/09 The syntax error of a character other than { following "${if
1555 def:variable_name" (after optional whitespace) was not being diagnosed.
1556 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1557 accidental colon was present, for example, could give incorrect results.
1559 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1560 result is used (not including appendfile, where other changes are about
1563 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1564 This involved changing a lot of size variables from int to off_t. It
1565 should work with maildirs and everything.
1567 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1568 spamd dying while we are connected to it.
1570 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1573 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1574 function for BDB 4.3.
1576 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1578 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1579 the sources for PCRE have been re-arranged and more files are now
1582 PH/15 The code I had for printing potentially long long variables in PH/11
1583 above was not the best (it lost precision). The length of off_t variables
1584 is now inspected at build time, and an appropriate printing format (%ld
1585 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1586 to be "long long int" or "long int". This is needed for the internal
1587 formatting function string_vformat().
1589 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1590 the configuration file to be ":syslog", then the script "guesses" where
1591 the logs files are, rather than using the compiled in default. In our
1592 case the guess is not the same as the compiled default, so the script
1593 suddenly stopped working when I started to use syslog. The patch checks
1594 to see if log_file_path is "". If so, it attempts to read it from exim
1595 with no configuration file to get the compiled in version, before it
1596 falls back to the previous guessing code."
1598 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1599 implementing BATV in an Exim configuration. See NewStuff for the gory
1602 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1603 Makefile that are specific to HP-UX.
1605 PH/18 If the "use_postmaster" option was set for a recipient callout together
1606 with the "random" option, the postmaster address was used as the MAIL
1607 FROM address for the random test, but not for the subsequent recipient
1608 test. It is now used for both.
1610 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1611 patch removes a few documentation additions to RFC 3028, because the
1612 latest draft now contains them. It adds the new en;ascii-case comparator
1613 and a new error check for 8bit text in MIME parts. Comparator and
1614 require names are now matched exactly. I enabled the subaddress
1615 extension, but it is not well tested yet (read: it works for me)."
1617 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1618 rework some of the code of TK/09 above to avoid the hardwired use of
1619 "%lld" and "long long". Replaced the call to snprintf() with a call to
1622 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1623 records point to non-existent hosts", "retry timeout exceeded", and
1624 "retry time not reached for any host after a long failure period".
1626 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1627 experimental DomainKeys support:
1629 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1630 (2) On an error such as an illegally used "control", the wrong name for
1631 the control was given.
1633 These problems did NOT occur unless DomainKeys support was compiled.
1635 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1637 PH/24 Added ${if match_ip condition.
1639 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1640 databases so that it will be absolutely obvious if a crash occurs in the
1641 DB library. This is a regular occurrence (often caused by mis-matched
1644 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1645 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1646 picked up on a user's system that detects such things. There doesn't seem
1647 to be a gcc warning option for this - only an attribute that has to be
1648 put on the function's prototype. It seems that in Fedora Core 4 they have
1649 set this on a number of new functions. No doubt there will be more in due
1652 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1653 specify a transport. However, if an address that was verified by such a
1654 router was the subject of a callout, Exim crashed because it tried to
1655 read the rcpt_include_affixes from the non-existent transport. Now it
1656 just assumes that the setting of that option is false. This bug was
1657 introduced by 4.51/PH/31.
1659 PH/28 Changed -d+all to exclude +memory, because that information is very
1660 rarely of interest, but it makes the output a lot bigger. People tend to
1661 do -d+all out of habit.
1663 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1664 code in os-type was giving problems when libc.so lives in lib64, like on
1667 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1668 aren't the modern standard, and it seems that some systems' include files
1669 don't always have them. Exim was already checking for some of the newer
1670 ones like T_AAAA, and defining it itself. I've added checks for all the
1671 record types that Exim uses.
1673 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1674 not automatically generating a new one, as it is supposed to. This
1675 prevented TLS from working. If the file did exist, but contained invalid
1676 data, a new version was generated, as expected. It was only the case of a
1677 non-existent file that was broken.
1679 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1680 with a change in libdomainkeys > 0.64.
1682 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1683 from DNS. If the selector record carries the flag, it now has
1684 precedence over the domain-wide flag.
1686 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1688 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1689 the use of a port name that isn't defined in /etc/services) occurred, the
1690 message was deferred as in a normal delivery, and thus remained on the
1691 spool, instead of being failed because of the mua_wrapper setting. This
1692 is now fixed, and I tidied up some of the mua_wrapper messages at the
1695 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1696 the messages in a hash of arrays rather than using individual hashes.
1697 This is a bit cleaner and results in dramatic memory savings, albeit
1698 at a slight CPU cost.
1700 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1701 as requested by Marc Sherman.
1703 SC/10 Eximstats: added histograms for user specified patterns as requested
1706 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1708 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1709 fopen() in the content-scanning modules that did not already have it.
1715 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1716 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1718 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1720 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1722 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1723 to test Sieve filters that use "vacation".
1725 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1726 that changes the way the GnuTLS parameters are stored in the cache file.
1727 The new format can be generated externally. For backward compatibility,
1728 if the data in the cache doesn't make sense, Exim assumes it has read an
1729 old-format file, and it generates new data and writes a new file. This
1730 means that you can't go back to an older release without removing the
1733 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1734 work if there are any delivery delays because "one_time" forces the
1735 parent to be marked "delivered", so its unseen clone is never tried
1736 again. For this reason, Exim now forbids the simultaneous setting of
1739 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1740 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1741 entry if you want to know the details. The fix, however, neglected to
1742 consider the case when local delivery batching is involved. The test for
1743 "previously delivered" was not happening when checking to see if an
1744 address could be batched with a previous (undelivered) one; under
1745 certain circumstances this could lead to multiple deliveries to the same
1748 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1749 in its include files, and this causes problems building Exim.
1751 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1752 header_syntax) but Exim was just ignoring anything given after a slash.
1753 In particular, this caused confusion with an attempt to use "verify =
1754 reverse_host_lookup/defer_ok". An error is now given when options are
1755 supplied for verify items that do not have them. (Maybe reverse_host_
1756 lookup should have a defer_ok option, but that's a different point.)
1758 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1759 defined by RFC 821) to 2048, because there were problems with some AUTH
1760 commands, and RFC 1869 says the size should be increased for extended
1761 SMTP commands that take arguments.
1763 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1766 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1767 "unknown" error; now it says that the functionality isn't in the binary.
1769 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1770 an address' error message when a string expansion fails (syntax or
1771 whatever). Otherwise the password may appear in the log. Following change
1772 PH/42 below, there is no longer a chance of it appearing in a bounce
1775 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1777 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1778 of its addresses was ever tried. (Bugzilla bug #2.)
1780 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1781 the result incorrectly in the debug output. (It correctly added a newline
1782 to what was transported.)
1784 TF/01 Added $received_time.
1786 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1787 commented out examples of how to interface to a virus scanner and to
1788 SpamAssassin. Also added commented examples of av_scanner and
1789 spamd_address settings.
1791 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1792 and controls are allowed in which ACLs. There were a couple of minor
1793 errors. Some of the entries in the conditions table (which is a table of
1794 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1795 negation of where the condition IS allowed.
1797 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1799 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1800 header file does not have a version number, so I've had to invent a new
1801 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1802 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1803 radiusclient), but it was contributed by a Radius user.
1805 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1806 files or directories, for OpenSSL.
1808 PH/19 When an Exim process that is running as root has to create an Exim log
1809 file, it does so in a subprocess that runs as exim:exim so as to get the
1810 ownership right at creation (otherwise, other Exim processes might see
1811 the file with the wrong ownership). There was no test for failure of this
1812 fork() call, which would lead to the process getting stuck as it waited
1813 for a non-existent subprocess. Forks do occasionally fail when resources
1814 run out. I reviewed all the other calls to fork(); they all seem to check
1817 PH/20 When checking for unexpected SMTP input at connect time (before writing
1818 the banner), Exim was not dealing correctly with a non-positive return
1819 from the read() function. If the client had disconnected by this time,
1820 the result was a log entry for a synchronization error with an empty
1821 string after "input=" when read() returned zero. If read() returned -1
1822 (an event I could not check), uninitialized data bytes were printed.
1823 There were reports of junk text (parts of files, etc) appearing after
1826 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1828 PH/22 Added support for macro redefinition, and (re)definition in between
1829 driver and ACL definitions.
1831 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1832 forgetting to use the resulting value; it was using the unexpanded value.
1834 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1835 hadn't been configured. The fix is from Juergen Kreileder, who
1836 understands it better than I do:
1838 "Here's what I see happening with three configured cyrus_sasl
1839 authenticators configured (plain, login, cram-md5):
1841 On startup auth_cyrus_sasl_init() gets called for each of these.
1842 This means three calls to sasl_listmech() without a specified mech_list.
1843 => SASL tests which mechs of all available mechs actually work
1844 => three warnings about OTP not working
1845 => the returned list contains: plain, login, cram-md5, digest-md5, ...
1847 With the patch, sasl_listmech() also gets called three times. But now
1848 SASL's mech_list option is set to the server_mech specified in the the
1849 authenticator. Or in other words, the answer from sasl_listmech()
1850 gets limited to just the mech you're testing for (which is different
1852 => the return list contains just 'plain' or 'login', 'cram-md5' or
1853 nothing depending on the value of ob->server_mech.
1855 I've just tested the patch: Authentication still works fine,
1856 unavailable mechs specified in the exim configuration are still
1857 caught, and the auth.log warnings about OTP are gone."
1859 PH/25 When debugging is enabled, the contents of the command line are added
1860 to the debugging output, even when log_selector=+arguments is not
1863 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
1864 answer is "GNU", and only if the return is "GNU/something" is the answer
1867 PH/27 $acl_verify_message is now set immediately after the failure of a
1868 verification in an ACL, and so is available in subsequent modifiers. In
1869 particular, the message can be preserved by coding like this:
1871 warn !verify = sender
1872 set acl_m0 = $acl_verify_message
1874 Previously, $acl_verify_message was set only while expanding "message"
1875 and "log_message" when a very denied access.
1877 PH/28 Modified OS/os.c-Linux with
1879 -#ifndef OS_LOAD_AVERAGE
1880 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
1882 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
1883 nomenclature these days.)
1885 PH/29 Installed patch from the Sieve maintainer that adds the options
1886 sieve_useraddress and sieve_subaddress to the redirect router.
1888 PH/30 In these circumstances:
1889 . Two addresses routed to the same list of hosts;
1890 . First host does not offer TLS;
1891 . First host accepts first address;
1892 . First host gives temporary error to second address;
1893 . Second host offers TLS and a TLS session is established;
1894 . Second host accepts second address.
1895 Exim incorrectly logged both deliveries with the TLS parameters (cipher
1896 and peerdn, if requested) that were in fact used only for the second
1899 PH/31 When doing a callout as part of verifying an address, Exim was not paying
1900 attention to any local part prefix or suffix that was matched by the
1901 router that accepted the address. It now behaves in the same way as it
1902 does for delivery: the affixes are removed from the local part unless
1903 rcpt_include_affixes is set on the transport.
1905 PH/32 Add the sender address, as F=<...>, to the log line when logging a
1906 timeout during the DATA phase of an incoming message.
1908 PH/33 Sieve envelope tests were broken for match types other than :is. I have
1909 applied a patch sanctioned by the Sieve maintainer.
1911 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
1912 the uid or gid is negative. A case of a negative gid caused this to be
1913 noticed. The fix allows for either to be negative.
1915 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
1916 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
1919 PH/36 The change PH/12 above was broken. Fixed it.
1921 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
1922 the grounds that routing the same address twice would always produce the
1923 same answer. This might have been true once, but it is certainly no
1924 longer true now. Routing a child address may depend on the previous
1925 routing that produced that child. Some complicated redirection strategies
1926 went wrong when messages had multiple recipients, and made Exim's
1927 behaviour dependent on the order in which the addresses were given.
1929 I have moved the duplicate checking until after the routing is complete.
1930 Exim scans the addresses that are assigned to local and remote
1931 transports, and removes any duplicates. This means that more work will be
1932 done, as duplicates will always all be routed, but duplicates are
1933 presumably rare, so I don't expect this is of any significance.
1935 For deliveries to pipes, files, and autoreplies, the duplicate checking
1936 still happens during the routing process, since they are not going to be
1939 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
1940 It corrects a timeout issue with spamd. This is Ian's comment: "The
1941 background is that sometimes spamd either never reads data from a
1942 connection it has accepted, or it never writes response data. The exiscan
1943 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
1944 blindly assumes that writes won't block so it may never time out."
1946 PH/39 Allow G after quota size as well as K and M.
1948 PH/40 The value set for $authenticated_id in an authenticator may not contain
1949 binary zeroes or newlines because the value is written to log lines and
1950 to spool files. There was no check on this. Now the value is run through
1951 the string_printing() function so that such characters are converted to
1952 printable escape sequences.
1954 PH/41 $message_linecount is a new variable that contains the total number of
1955 lines in the message. Compare $body_linecount, which is the count for the
1958 PH/42 Exim no longer gives details of delivery errors for specific addresses in
1959 bounce and delay warning messages, except in certain special cases, which
1962 (a) An SMTP error message from a remote host;
1963 (b) A message specified in a :fail: redirection;
1964 (c) A message specified in a "fail" command in a system filter;
1965 (d) A message specified in a FAIL return from the queryprogram router;
1966 (e) A message specified by the cannot_route_message router option.
1968 In these cases only, Exim does include the error details in bounce and
1969 warning messages. There are also a few cases where bland messages such
1970 as "unrouteable address" or "local delivery error" are given.
1972 PH/43 $value is now also set for the "else" part of a ${run expansion.
1974 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
1975 being worked on, but at least Exim now implements the latest version to
1978 PH/45 In a pipe transport, although a timeout while waiting for the pipe
1979 process to complete was treated as a delivery failure, a timeout while
1980 writing the message to the pipe was logged, but erroneously treated as a
1981 successful delivery. Such timeouts include transport filter timeouts. For
1982 consistency with the overall process timeout, these timeouts are now
1983 treated as errors, giving rise to delivery failures by default. However,
1984 there is now a new Boolean option for the pipe transport called
1985 timeout_defer, which, if set TRUE, converts the failures into defers for
1986 both kinds of timeout. A transport filter timeout is now identified in
1989 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
1990 systems where "make" and "gmake" are different, calling "gmake" at top
1991 level broke things. I've arranged for the value of $(MAKE) to be passed
1992 from the Makefile to this script so that it can call the same version of
1996 A note about Exim versions 4.44 and 4.50
1997 ----------------------------------------
1999 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2000 changes of various kinds. As a consequence, a big documentation update was
2001 needed. This delayed the release for rather longer than seemed good, especially
2002 in the light of a couple of (minor) security issues. Therefore, the changes
2003 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2004 release. So 4.44 and 4.50 are in effect two different branches that both start
2007 I have left the 4.50 change log unchanged; it contains all the changes since
2008 4.43. The change log for 4.44 is below; many of its items are identical to
2009 those for 4.50. This seems to be the most sensible way to preserve the
2010 historical information.
2016 1. Minor wording change to the doc/README.SIEVE file.
2018 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2019 computation of the current number of files was incorrect.
2021 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2022 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2025 4. Give more explanation in the error message when the command for a transport
2026 filter fails to execute.
2028 5. There are several places where Exim runs a non-Exim command in a
2029 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2030 was being done only for the command run by the queryprogram router. It is
2031 now done for all such subprocesses. The other cases are: ${run, transport
2032 filters, and the commands run by the lmtp and pipe transports.
2034 6. Added CONFIGURE_GROUP build-time option.
2036 7. Some older OS have a limit of 256 on the maximum number of file
2037 descriptors. Exim was using setrlimit() to set 1000 as a large value
2038 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2039 systems. I've change it so that if it can't get 1000, it tries for 256.
2041 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2042 was an oversight, and furthermore, ever since the addition of extra
2043 controls (e.g. 4.43/32), the checks on when to allow different forms of
2044 "control" were broken. There should now be diagnostics for all cases when a
2045 control that does not make sense is encountered.
2047 9. Added the /retain_sender option to "control=submission".
2049 10. $recipients is now available in the predata ACL (oversight).
2051 11. Tidy the search cache before the fork to do a delivery from a message
2052 received from the command line. Otherwise the child will trigger a lookup
2053 failure and thereby defer the delivery if it tries to use (for example) a
2054 cached ldap connection that the parent has called unbind on.
2056 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2057 of $address_data from the recipient verification was clobbered by the
2058 sender verification.
2060 13. The value of address_data from a sender verification is now available in
2061 $sender_address_data in subsequent conditions in the ACL statement.
2063 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2065 15. Added a new option "connect=<time>" to callout options, to set a different
2068 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2069 was its contents. (It was OK if the option was not defined at all.)
2071 17. A "Completed" log line is now written for messages that are removed from
2072 the spool by the -Mrm option.
2074 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2075 information about exactly what failed.
2077 19. Added -dd to debug only the daemon process.
2079 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2080 handles timeouts, both on the server side and network timeouts. Renamed the
2081 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2083 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2084 It is now set to "smtps".
2086 22. $host_address is now set to the target address during the checking of
2087 ignore_target_hosts.
2089 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2090 being passed; this would have caused $sender_host_name to have been used if
2091 matching the list had actually called for a host name (not very likely,
2092 since this list is usually IP addresses). A host name is now passed as
2095 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2096 code that specifies a non-restarting handler (typically sigaction() in
2097 modern systems) in an attempt to fix a rare and obscure crash bug.
2099 25. Narrowed the window for a race in the daemon that could cause it to ignore
2100 SIGCHLD signals. This is not a major problem, because they are used only to
2101 wake it up if nothing else does.
2103 26. A malformed maildirsize file could cause Exim to calculate negative values
2104 for the mailbox size or file count. Odd effects could occur as a result.
2105 The maildirsize information is now recalculated if the size or filecount
2108 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2109 support for a long time. Removed HAVE_SYS_VFS_H.
2111 28. Installed the latest version of exipick from John Jetmore.
2113 29. In an address list, if the pattern was not a regular expression, an empty
2114 subject address (from a bounce message) matched only if the pattern was an
2115 empty string. Non-empty patterns were not even tested. This was the wrong
2116 because it is perfectly reasonable to use an empty address as part of a
2117 database query. An empty address is now tested by patterns that are
2118 lookups. However, all the other forms of pattern expect the subject to
2119 contain a local part and a domain, and therefore, for them, an empty
2120 address still always fails if the pattern is not itself empty.
2122 30. Exim went into a mad DNS loop when attempting to do a callout where the
2123 host was specified on an smtp transport, and looking it up yielded more
2124 than one IP address.
2126 31. Re-factored the code for checking spool and log partition space into a
2127 function that finds that data and another that does the check. The former
2128 is then used to implement four new variables: $spool_space, $log_space,
2129 $spool_inodes, and $log_inodes.
2131 32. The RFC2047 encoding function was originally intended for short strings
2132 such as real names; it was not keeping to the 75-character limit for
2133 encoded words that the RFC imposes. It now respects the limit, and
2134 generates multiple encoded words if necessary. To be on the safe side, I
2135 have increased the buffer size for the ${rfc2047: expansion operator from
2138 33. It is now permitted to omit both strings after an "if" condition; if the
2139 condition is true, the result is "true". As before, when the second string
2140 is omitted, a false condition yields an empty string. This makes it less
2141 cumbersome to write custom ACL and router conditions.
2143 34. Failure to deliver a bounce message always caused it to be frozen, even if
2144 there was an errors_to setting on the router. The errors_to setting is now
2147 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2148 canonical form (fully expanded) before being placed in
2149 $sender_host_address.
2151 36. The table in the code that translates DNS record types into text (T_A to
2152 "A" for instance) was missing entries for NS and CNAME. It is just possible
2153 that this could have caused confusion if both these types were looked up
2154 for the same domain, because the text type is used as part of Exim's
2155 per-process caching. But the chance of anyone hitting this buglet seems
2158 37. The dnsdb lookup has been extended in a number of ways.
2160 (1) There is a new type, "zns", which walks up the domain tree until it
2161 finds some nameserver records. It should be used with care.
2163 (2) There is a new type, "mxh", which is like "mx" except that it returns
2164 just the host names, not the priorities.
2166 (3) It is now possible to give a list of domains (or IP addresses) to be
2167 looked up. The behaviour when one of the lookups defers can be
2168 controlled by a keyword.
2170 (4) It is now possible to specify the separator character for use when
2171 multiple records are returned.
2173 38. The dnslists ACL condition has been extended: it is now possible to supply
2174 a list of IP addresses and/or domains to be looked up in a particular DNS
2177 39. Added log_selector=+queue_time_overall.
2179 40. When running the queue in the test harness, wait just a tad after forking a
2180 delivery process, to get repeatability of debugging output.
2182 41. Include certificate and key file names in error message when GnuTLS fails
2183 to set them up, because the GnuTLS error message doesn't include the name
2184 of the failing file when there is a problem reading it.
2186 42. Allow both -bf and -bF in the same test run.
2188 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2190 44. The "Exiscan patch" is now merged into the mainline Exim source.
2192 45. Sometimes the final signoff response after QUIT could fail to get
2193 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2194 < 0 before doing a fflush(). This bug looks as though it goes back to the
2195 introduction of TLS in release 3.20, but "sometimes" must have been rare
2196 because the tests only now provoked it.
2198 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2199 (this can affect the format of dates).
2201 47. exim_tidydb, when checking for the continued existence of a message for
2202 which it has found a message-specific retry record, was not finding
2203 messages that were in split spool directories. Consequently, it was
2204 deleting retry records that should have stayed in existence.
2206 48. Steve fixed some bugs in eximstats.
2208 49. The SPA authentication driver was not abandoning authentication and moving
2209 on to the next authenticator when an expansion was forced to fail,
2210 contradicting the general specification for all authenticators. Instead it
2211 was generating a temporary error. It now behaves as specified.
2213 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2214 (the order specifies the preference for clients). The order is now AES256,
2215 AES128, 3DES, ARCFOUR128.
2217 51. Small patch to Sieve code - explicitly set From: when generating an
2220 52. Exim crashed if a remote delivery caused a very long error message to be
2221 recorded - for instance if somebody sent an entire SpamAssassin report back
2222 as a large number of 550 error lines. This bug was coincidentally fixed by
2223 increasing the size of one of Exim's internal buffers (big_buffer) that
2224 happened as part of the Exiscan merge. However, to be on the safe side, I
2225 have made the code more robust (and fixed the comments that describe what
2228 53. Now that there can be additional text after "Completed" in log lines (if
2229 the queue_time_overall log selector is set), a one-byte patch to exigrep
2230 was needed to allow it to recognize "Completed" as not the last thing in
2233 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2234 patch that reportedly fixes this has been added. I am not expert enough to
2235 create a test for it. This is what the patch creator wrote:
2237 "I found a little strange behaviour of ldap code when working with
2238 Windows 2003 AD Domain, where users was placed in more than one
2239 Organization Units. When I tried to give exim partial DN, the exit code
2240 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2241 But simultaneously result of request was absolutely normal ldap result,
2242 so I produce this patch..."
2244 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2245 REFERENCE, so I have modified the code to exclude the patch when that macro
2248 55. Some experimental protocols are using DNS PTR records for new purposes. The
2249 keys for these records are domain names, not reversed IP addresses. The
2250 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2251 leaves it alone. Component reversal etc. now happens only for IP addresses.
2254 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2256 57. Double the size of the debug message buffer (to 2048) so that more of very
2257 long debug lines gets shown.
2259 58. The exicyclog utility now does better if the number of log files to keep
2260 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2262 59. Two changes related to the smtp_active_hostname option:
2264 (1) $smtp_active_hostname is now available as a variable.
2265 (2) The default for smtp_banner uses $smtp_active_hostname instead
2266 of $primary_hostname.
2268 60. The host_aton() function is supposed to be passed a string that is known
2269 to be a valid IP address. However, in the case of IPv6 addresses, it was
2270 not checking this. This is a hostage to fortune. Exim now panics and dies
2271 if the condition is not met. A case was found where this could be provoked
2272 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2273 components; fortuitously, this particular loophole had already been fixed
2274 by change 4.50/55 above.
2276 If there are any other similar loopholes, the new check in host_aton()
2277 itself should stop them being exploited. The report I received stated that
2278 data on the command line could provoke the exploit when Exim was running as
2279 exim, but did not say which command line option was involved. All I could
2280 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2281 running as the user.
2284 61. There was a buffer overflow vulnerability in the SPA authentication code
2285 (which came originally from the Samba project). I have added a test to the
2286 spa_base64_to_bits() function which I hope fixes it.
2289 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2290 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2292 63. The daemon start-up calls getloadavg() while still root for those OS that
2293 need the first call to be done as root, but it missed one case: when
2294 deliver_queue_load_max is set with deliver_drop_privilege. This is
2295 necessary for the benefit of the queue runner, because there is no re-exec
2296 when deliver_drop_privilege is set.
2298 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2299 This has been fixed.
2301 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2302 was in use, was not putting the data itself into the right store pool;
2303 consequently, it could be overwritten for a subsequent message in the same
2304 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2307 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2309 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2310 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2311 the function were treating the return as a boolean value, which happened to
2312 work because 0=false and not-0=true, but is not correct code.
2314 68. The host_aton() function was not handling scoped IPv6 addresses (those
2315 with, for example, "%eth0" on the end) correctly.
2317 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2318 negated items (that is, ~something) in unsigned ints. Some compilers
2319 apparently mutter when there is no cast.
2321 70. If an address verification called from an ACL failed, and did not produce a
2322 user-specific message (i.e. there was only a "system" message), nothing was
2323 put in $acl_verify_message. In this situation, it now puts the system
2326 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2327 session; change 4.31/43 added the unwanted input to the log line - except
2328 that it did not do this in the start of session case. It now does.
2330 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2331 This is wrong and can cause the other end to generate a synchronization
2332 error if it is another Exim or anything else that does the synchronization
2333 check. A QUIT command is no longer sent after a timeout.
2335 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2336 during host lookups.
2338 74. The defer_ok option of callout verification was not working if it was used
2339 when verifying addresses in header lines, that is, for this case:
2341 verify = header_sender/callout=defer_ok
2343 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2344 those file descriptors could be used for SMTP connections. If anything
2345 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2346 could be sent to the SMTP client, causing chaos. The daemon now opens
2347 stdin, stdout, and stderr to /dev/null when it puts itself into the
2350 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2351 log by default. The user can override this with suitable Perl magic.
2353 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2354 the log message when discard triggers, was not working for the DATA ACL or
2355 for the non-SMTP ACL.
2357 78. Error message wording change in sieve.c.
2359 79. If smtp_accept_max_per_host was set, the number of connections could be
2360 restricted to fewer than expected, because the daemon was trying to set up
2361 a new connection before checking whether the processes handling previous
2362 connections had finished. The check for completed processes is now done
2363 earlier. On busy systems, this bug wouldn't be noticed because something
2364 else would have woken the daemon, and it would have reaped the completed
2367 80. If a message was submitted locally by a user whose login name contained one
2368 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2369 It caused a spool format error. I have fixed the spool reading code. A
2370 related problem was that the "from" clause in the Received: line became
2371 illegal because of the space(s). It is now covered by ${quote_local_part.
2373 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2376 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2378 83. Patch from Sieve maintainer to fix unterminated string problem in
2379 "vacation" handling.
2381 84. Some minor changes to the Linux configuration files to help with other
2382 OS variants using glibc.
2384 85. One more patch for Sieve to update vacation handling to latest spec.
2387 ----------------------------------------------------
2388 See the note above about the 4.44 and 4.50 releases.
2389 ----------------------------------------------------
2395 1. Change 4.43/35 introduced a bug that caused file counts to be
2396 incorrectly computed when quota_filecount was set in an appendfile
2399 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2400 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2403 3. Give more explanation in the error message when the command for a transport
2404 filter fails to execute.
2406 4. There are several places where Exim runs a non-Exim command in a
2407 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2408 was being done only for the command run by the queryprogram router. It is
2409 now done for all such subprocesses. The other cases are: ${run, transport
2410 filters, and the commands run by the lmtp and pipe transports.
2412 5. Some older OS have a limit of 256 on the maximum number of file
2413 descriptors. Exim was using setrlimit() to set 1000 as a large value
2414 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2415 systems. I've change it so that if it can't get 1000, it tries for 256.
2417 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2418 was an oversight, and furthermore, ever since the addition of extra
2419 controls (e.g. 4.43/32), the checks on when to allow different forms of
2420 "control" were broken. There should now be diagnostics for all cases when a
2421 control that does not make sense is encountered.
2423 7. $recipients is now available in the predata ACL (oversight).
2425 8. Tidy the search cache before the fork to do a delivery from a message
2426 received from the command line. Otherwise the child will trigger a lookup
2427 failure and thereby defer the delivery if it tries to use (for example) a
2428 cached ldap connection that the parent has called unbind on.
2430 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2431 of $address_data from the recipient verification was clobbered by the
2432 sender verification.
2434 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2435 was its contents. (It was OK if the option was not defined at all.)
2437 11. A "Completed" log line is now written for messages that are removed from
2438 the spool by the -Mrm option.
2440 12. $host_address is now set to the target address during the checking of
2441 ignore_target_hosts.
2443 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2444 being passed; this would have caused $sender_host_name to have been used if
2445 matching the list had actually called for a host name (not very likely,
2446 since this list is usually IP addresses). A host name is now passed as
2449 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2450 code that specifies a non-restarting handler (typically sigaction() in
2451 modern systems) in an attempt to fix a rare and obscure crash bug.
2453 15. Narrowed the window for a race in the daemon that could cause it to ignore
2454 SIGCHLD signals. This is not a major problem, because they are used only to
2455 wake it up if nothing else does.
2457 16. A malformed maildirsize file could cause Exim to calculate negative values
2458 for the mailbox size or file count. Odd effects could occur as a result.
2459 The maildirsize information is now recalculated if the size or filecount
2462 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2463 support for a long time. Removed HAVE_SYS_VFS_H.
2465 18. Updated exipick to current release from John Jetmore.
2467 19. Allow an empty sender to be matched against a lookup in an address list.
2468 Previously the only cases considered were a regular expression, or an
2471 20. Exim went into a mad DNS lookup loop when doing a callout where the
2472 host was specified on the transport, if the DNS lookup yielded more than
2475 21. The RFC2047 encoding function was originally intended for short strings
2476 such as real names; it was not keeping to the 75-character limit for
2477 encoded words that the RFC imposes. It now respects the limit, and
2478 generates multiple encoded words if necessary. To be on the safe side, I
2479 have increased the buffer size for the ${rfc2047: expansion operator from
2482 22. Failure to deliver a bounce message always caused it to be frozen, even if
2483 there was an errors_to setting on the router. The errors_to setting is now
2486 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2487 canonical form (fully expanded) before being placed in
2488 $sender_host_address.
2490 24. Updated eximstats to version 1.33
2492 25. Include certificate and key file names in error message when GnuTLS fails
2493 to set them up, because the GnuTLS error message doesn't include the name
2494 of the failing file when there is a problem reading it.
2496 26. Expand error message when OpenSSL has problems setting up cert/key files.
2499 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2500 (this can affect the format of dates).
2502 28. exim_tidydb, when checking for the continued existence of a message for
2503 which it has found a message-specific retry record, was not finding
2504 messages that were in split spool directories. Consequently, it was
2505 deleting retry records that should have stayed in existence.
2507 29. eximstats updated to version 1.35
2508 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2509 1.35 - bugfix such that pie charts by volume are generated correctly
2511 30. The SPA authentication driver was not abandoning authentication and moving
2512 on to the next authenticator when an expansion was forced to fail,
2513 contradicting the general specification for all authenticators. Instead it
2514 was generating a temporary error. It now behaves as specified.
2516 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2517 (the order specifies the preference for clients). The order is now AES256,
2518 AES128, 3DES, ARCFOUR128.
2520 31. Small patch to Sieve code - explicitly set From: when generating an
2523 32. Exim crashed if a remote delivery caused a very long error message to be
2524 recorded - for instance if somebody sent an entire SpamAssassin report back
2525 as a large number of 550 error lines. This bug was coincidentally fixed by
2526 increasing the size of one of Exim's internal buffers (big_buffer) that
2527 happened as part of the Exiscan merge. However, to be on the safe side, I
2528 have made the code more robust (and fixed the comments that describe what
2531 33. Some experimental protocols are using DNS PTR records for new purposes. The
2532 keys for these records are domain names, not reversed IP addresses. The
2533 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2534 leaves it alone. Component reversal etc. now happens only for IP addresses.
2537 34. The host_aton() function is supposed to be passed a string that is known
2538 to be a valid IP address. However, in the case of IPv6 addresses, it was
2539 not checking this. This is a hostage to fortune. Exim now panics and dies
2540 if the condition is not met. A case was found where this could be provoked
2541 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2542 components; fortuitously, this particular loophole had already been fixed
2543 by change 4.50/55 or 4.44/33 above.
2545 If there are any other similar loopholes, the new check in host_aton()
2546 itself should stop them being exploited. The report I received stated that
2547 data on the command line could provoke the exploit when Exim was running as
2548 exim, but did not say which command line option was involved. All I could
2549 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2550 running as the user.
2553 35. There was a buffer overflow vulnerability in the SPA authentication code
2554 (which came originally from the Samba project). I have added a test to the
2555 spa_base64_to_bits() function which I hope fixes it.
2558 36. The daemon start-up calls getloadavg() while still root for those OS that
2559 need the first call to be done as root, but it missed one case: when
2560 deliver_queue_load_max is set with deliver_drop_privilege. This is
2561 necessary for the benefit of the queue runner, because there is no re-exec
2562 when deliver_drop_privilege is set.
2564 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2565 was in use, was not putting the data itself into the right store pool;
2566 consequently, it could be overwritten for a subsequent message in the same
2567 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2570 38. Sometimes the final signoff response after QUIT could fail to get
2571 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2572 < 0 before doing a fflush(). This bug looks as though it goes back to the
2573 introduction of TLS in release 3.20, but "sometimes" must have been rare
2574 because the tests only now provoked it.
2580 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2581 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2582 Now it returns an integer. A number of calls were still expecting a T/F
2583 return. Fortuitously, in all cases, the tests worked in OK situations,
2584 which is the norm. However, things would have gone wrong on any write
2585 failures on the smtp file descriptor. This function is used when sending
2586 messages over SMTP and also when doing verify callouts.
2588 2. When Exim is called to do synchronous delivery of a locally submitted
2589 message (the -odf or -odi options), it no longer closes stderr before doing
2592 3. Implemented the mua_wrapper option.
2594 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2596 5. Implemented the functions header_remove(), header_testname(),
2597 header_add_at_position(), and receive_remove_recipient(), and exported them
2600 6. If an ACL "warn" statement specified the addition of headers, Exim already
2601 inserted X-ACL-Warn: at the start if there was no header name. However, it
2602 was not making this test for the second and subsequent header lines if
2603 there were newlines in the string. This meant that an invalid header could
2604 be inserted if Exim was badly configured.
2606 7. Allow an ACL "warn" statement to add header lines at the start or after all
2607 the Received: headers, as well as at the end.
2609 8. Added the rcpt_4xx retry error code.
2611 9. Added postmaster_mailfrom=xxx to callout verification option.
2613 10. Added mailfrom=xxxx to the callout verification option, for verify=
2616 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2617 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2619 12. Inserted some casts to stop certain compilers warning when using pointer
2620 differences as field lengths or precisions in printf-type calls (mostly
2621 affecting debugging statements).
2623 13. Added optional readline() support for -be (dynamically loaded).
2625 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2626 same clock tick as a message's arrival, so that its received time was the
2627 same as the "first fail" time on the retry record, and that message
2628 remained on the queue past the ultimate address timeout, every queue runner
2629 would try a delivery (because it was past the ultimate address timeout) but
2630 after another failure, the ultimate address timeout, which should have then
2631 bounced the address, did not kick in. This was a "< instead of <=" error;
2632 in most cases the first failure would have been in the next clock tick
2633 after the received time, and all would be well.
2635 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2636 being recognized when the domain list was tested by the match_domain
2637 condition in an expansion string.
2639 16. Added the ${str2b64: operator.
2641 17. Exim was always calling setrlimit() to set a large limit for the number of
2642 processes, without checking whether the existing limit was already
2643 adequate. (It did check for the limit on file descriptors.) Furthermore,
2644 errors from getrlimit() and setrlimit() were being ignored. Now they are
2645 logged to the main and panic logs, but Exim does carry on, to try to do its
2646 job under whatever limits there are.
2648 18. Imported PCRE 5.0.
2650 19. Trivial typo in log message " temporarily refused connection" (the leading
2653 20. If the log selector return_path_on_delivery was set and an address was
2654 redirected to /dev/null, the delivery process crashed because it assumed
2655 that a return path would always be set for a "successful" delivery. In this
2656 case, the whole delivery is bypassed as an optimization, and therefore no
2659 21. Internal re-arrangement: the function for sending a challenge and reading
2660 a response while authentication was assuming a zero-terminated challenge
2661 string. It's now changed to take a pointer and a length, to allow for
2662 binary data in such strings.
2664 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2666 23. Exim was not respecting finduser_retries when seeking the login of the
2667 uid under which it was called; it was always trying 10 times. (The default
2668 setting of finduser_retries is zero.) Also, it was sleeping after the final
2669 failure, which is pointless.
2671 24. Implemented tls_on_connect_ports.
2673 25. Implemented acl_smtp_predata.
2675 26. If the domain in control=submission is set empty, Exim assumes that the
2676 authenticated id is a complete email address when it generates From: or
2677 Sender: header lines.
2679 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2680 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2681 chown and chgrp in /bin and hostname in /usr/bin.
2683 28. Exim was keeping the "process log" file open after each use, just as it
2684 does for the main log. This opens the possibility of it remaining open for
2685 long periods when the USR1 signal hits a daemon. Occasional processlog
2686 errors were reported, that could have been caused by this. Anyway, it seems
2687 much more sensible not to leave this file open at all, so that is what now
2690 29. The long-running daemon process does not normally write to the log once it
2691 has entered its main loop, and it closes the log before doing so. This is
2692 so that log files can straightforwardly be renamed and moved. However,
2693 there are a couple of unusual error situations where the daemon does write
2694 log entries, and I had neglected to close the log afterwards.
2696 30. The text of an SMTP error response that was received during a remote
2697 delivery was being truncated at 512 bytes. This is too short for some of
2698 the long messages that one sometimes sees. I've increased the limit to
2701 31. It is now possible to make retry rules that apply only when a message has a
2702 specific sender, in particular, an empty sender.
2704 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2705 it possible to be selective about when SMTP synchronization is enforced.
2707 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2709 32. Implemented hosts_connection_nolog.
2711 33. Added an ACL for QUIT.
2713 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2716 35. Added mailbox_size and mailbox_filecount to appendfile.
2718 36. Added control = no_multiline_responses to ACLs.
2720 37. There was a bug in the logic of the code that waits for the clock to tick
2721 in the case where the clock went backwards by a substantial amount such
2722 that the microsecond fraction of "now" was more than the microsecond
2723 fraction of "then" (but the whole seconds number was less).
2725 38. Added support for the libradius Radius client library this is found on
2726 FreeBSD (previously only the radiusclient library was supported).
2732 1. When certain lookups returned multiple values in the form name=value, the
2733 quoting of the values was not always being done properly. Specifically:
2734 (a) If the value started with a double quote, but contained no whitespace,
2736 (b) If the value contained whitespace other than a space character (i.e.
2737 tabs or newlines or carriage returns) it was not quoted.
2738 This fix has been applied to the mysql and pgsql lookups by writing a
2739 separate quoting function and calling it from the lookup code. The fix
2740 should probably also be applied to nisplus, ibase and oracle lookups, but
2741 since I cannot test any of those, I have not disturbed their existing code.
2743 2. A hit in the callout cache for a specific address caused a log line with no
2744 reason for rejecting RCPT. Now it says "Previous (cached) callout
2745 verification failure".
2747 3. There was an off-by-one bug in the queryprogram router. An over-long
2748 return line was truncated at 256 instead of 255 characters, thereby
2749 overflowing its buffer with the terminating zero. As well as fixing this, I
2750 have increased the buffer size to 1024 (and made a note to document this).
2752 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2753 when Exim is waiting for an SMTP response from a remote server, Exim
2754 restarts its select() call on the socket, thereby resetting its timeout.
2755 This is not a problem when such interrupts are rare. Somebody set up a cron
2756 job to run exiwhat every 2 minutes, which is less than the normal select()
2757 timeout (5 or 10 minutes). This meant that the select() timeout never
2758 kicked in because it was always reset. I have fixed this by comparing the
2759 time when an interrupt arrives with the time at the start of the first call
2760 to select(). If more time than the timeout has elapsed, the interrupt is
2761 treated as a timeout.
2763 5. Some internal re-factoring in preparation for the addition of Sieve
2764 extensions (by MH). In particular, the "personal" test is moved to a
2765 separate function, and given an option for scanning Cc: and Bcc: (which is
2766 not set for Exim filters).
2768 6. When Exim created an email address using the login of the caller as the
2769 local part (e.g. when creating a From: or Sender: header line), it was not
2770 quoting the local part when it contained special characters such as @.
2772 7. Installed new OpenBSD configuration files.
2774 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2775 try to make them clearer.
2777 9. Callout options, other than the timeout value, were being ignored when
2778 verifying sender addresses in header lines. For example, when using
2780 verify = header_sender/callout=no_cache
2782 the cache was (incorrectly) being used.
2784 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2785 only the Cygwin environment.
2787 11. When return_path_on_delivery was set as a log selector, if different remote
2788 addresses in the same message used different return paths and parallel
2789 remote delivery occurred, the wrong values would sometimes be logged.
2790 (Whenever a remote delivery process finished, the return path value from
2791 the most recently started remote delivery process was logged.)
2793 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2794 lines when AUTH and/or TLS are in use. This is the "received protocol"
2795 field. Exim used to use "asmtp" for authenticated SMTP, without any
2796 indication (in the protocol name) for TLS use. Now it follows the RFC and
2797 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2798 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2799 names appear in log lines as well as in Received: header lines.
2801 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2802 extensions, and comparison tests, and to fix some bugs.
2804 14. Changes to the "personal" filter test:
2806 (1) The test was buggy in that it was just doing the equivalent of
2807 "contains" tests on header lines. For example, if a user's address was
2808 anne@some.where, the "personal" test would incorrectly be true for
2810 To: susanne@some.where
2812 This test is now done by extracting each address from the header in turn,
2813 and checking the entire address. Other tests that are part of "personal"
2814 are now done using regular expressions (for example, to check local parts
2815 of addresses in From: header lines).
2817 (2) The list of non-personal local parts in From: addresses has been
2818 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2819 taken from the Sieve specification recommendations.
2821 (3) If the message contains any header line starting with "List-" it is
2822 treated as non-personal.
2824 (4) The test for "circular" in the Subject: header line has been removed
2825 because it now seems ill-conceived.
2827 15. Minor typos in src/EDITME comments corrected.
2829 16. Installed latest exipick from John Jetmore.
2831 17. If headers_add on a router specified a text string that was too long for
2832 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2833 of string_sprintf() is now avoided.
2835 18. $message_body_size was not set (it was always zero) when running the DATA
2836 ACL and the local_scan() function.
2838 19. For the "mail" command in an Exim filter, no default was being set for
2839 the once_repeat time, causing a random time value to be used if "once" was
2840 specified. (If the value happened to be <= 0, no repeat happened.) The
2841 default is now 0s, meaning "never repeat". The "vacation" command was OK
2842 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
2843 (I found it when inspecting the code).
2845 20. There is now an overall timeout for performing a callout verification. It
2846 defaults to 4 times the callout timeout, which applies to individual SMTP
2847 commands during the callout. The overall timeout applies when there is more
2848 than one host that can be tried. The timeout is checked before trying the
2849 next host. This prevents very long delays if there are a large number of
2850 hosts and all are timing out (e.g. when the network connections are timing
2851 out). The value of the overall timeout can be changed by specifying an
2852 additional sub-option for "callout", called "maxwait". For example:
2854 verify = sender/callout=5s,maxwait=20s
2856 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
2857 to the end before writing, but this should make it even safer).
2859 22. Exim was forgetting that it had advertised PIPELINING for the second and
2860 subsequent messages on an SMTP connection. It was also not resetting its
2861 memory on STARTTLS and an internal HELO.
2863 23. When Exim logs an SMTP synchronization error within a session, it now
2864 records whether PIPELINING has been advertised or not.
2866 24. Added 3 instances of "(long int)" casts to time_t variables that were being
2867 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
2868 rather than long int.
2870 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
2872 26. Added the never_mail option to autoreply.
2878 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
2879 crash if the getsockname() call failed; this can happen if a connection is
2880 closed very soon after it is established. The problem was simply in the
2881 order in which certain operations were done, causing Exim to try to write
2882 to the SMTP stream before it had set up the file descriptor. The bug has
2883 been fixed by making things happen in the correct order.
2889 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
2890 before the connection was closed, thus losing the rejection response.
2892 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
2893 some early Solaris releases, but causes trouble in current releases where
2894 socklen_t is defined.
2896 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
2899 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
2902 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
2903 that releasing the top of it at the end releases what was used for sub-
2904 expansions (unless the block got too big). However, discard this block if
2905 the first thing is a variable or header, so that we can use its block when
2906 it is dynamic (useful for very large $message_headers, for example).
2908 6. Lookups now cache *every* query, not just the most recent. A new, separate
2909 store pool is used for this. It can be recovered when all lookup caches are
2910 flushed. Lookups now release memory at the end of their result strings.
2911 This has involved some general refactoring of the lookup sources.
2913 7. Some code has been added to the store_xxx() functions to reduce the amount
2914 of flapping under certain conditions.
2916 8. log_incoming_interface used to affect only the <= reception log lines. Now
2917 it causes the local interface and port to be added to several more SMTP log
2918 lines, for example "SMTP connection from", and rejection lines.
2920 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
2922 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
2924 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
2925 could be overwritten at the end of the current message (or the start of a
2926 new message if it was set in a HELO ACL). The value is now preserved for
2927 the duration of the SMTP connection.
2929 12. If a transport had a headers_rewrite setting, and a matching header line
2930 contained an unqualified address, that address was qualified, even if it
2931 did not match any rewriting rules. The underlying bug was that the values
2932 of the flags that permit the existence of unqualified sender and recipient
2933 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
2934 non-local messages, and by -bnq for local messages) were not being
2935 preserved with the message after it was received.
2937 13. When Exim was logging an SMTP synchronization error, it could sometimes log
2938 "next input=" as part of the text comprising the host identity instead of
2939 the correct text. The code was using the same buffer for two different
2940 strings. However, depending on which order the printing function evaluated
2941 its arguments, the bug did not always show up. Under Linux, for example, my
2942 test suite worked just fine.
2944 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
2945 doesn't work with some older versions of Perl. It has been changed to "my",
2946 which in any case is probably the better facility to use.
2948 15. A really picky compiler found some instances of statements for creating
2949 error messages that either had too many or two few arguments for the format
2952 16. The size of the buffer for calls to the DNS resolver has been increased
2953 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
2954 for addresses that have a lot of PTR records. This alleviates a problem; it
2955 does not fully solve it.
2957 17. A dnsdb lookup for PTR records that receives more data than will fit in the
2958 buffer now truncates the list and logs the incident, which is the same
2959 action as happens when Exim is looking up a host name and its aliases.
2960 Previously in this situation something unpredictable would happen;
2961 sometimes it was "internal error: store_reset failed".
2963 18. If a server dropped the connection unexpectedly when an Exim client was
2964 using GnuTLS and trying to read a response, the client delivery process
2965 crashed while trying to generate an error log message.
2967 19. If a "warn" verb in an ACL added multiple headers to a message in a single
2968 string, for example:
2970 warn message = H1: something\nH2: something
2972 the text was added as a single header line from Exim's point of view
2973 though it ended up OK in the delivered message. However, searching for the
2974 second and subsequent header lines using $h_h2: did not work. This has been
2975 fixed. Similarly, if a system filter added multiple headers in this way,
2976 the routers could not see them.
2978 20. Expanded the error message when iplsearch is called with an invalid key to
2979 suggest using net-iplsearch in a host list.
2981 21. When running tests using -bh, any delays imposed by "delay" modifiers in
2982 ACLs are no longer actually imposed (and a message to that effect is
2985 22. If a "gecos" field in a passwd entry contained escaped characters, in
2986 particular, if it contained a \" sequence, Exim got it wrong when building
2987 a From: or a Sender: header from that name. A second bug also caused
2988 incorrect handling when an unquoted " was present following a character
2989 that needed quoting.
2991 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
2992 was not being matched caselessly.
2994 24. Arranged for all hyphens in the exim.8 source to be escaped with
2997 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
2998 properly. Recipient callout cache records were still being keyed to include
2999 the sender, even when use_sender was set false. This led to far more
3000 callouts that were necessary. The sender is no longer included in the key
3001 when use_sender is false.
3003 26. Added "control = submission" modifier to ACLs.
3005 27. Added the ${base62d: operator to decode base 62 numbers.
3007 28. dnsdb lookups can now access SRV records.
3009 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3010 the configuration file.
3012 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3013 (-v) mode. This makes the output for a verbose queue run more intelligible.
3015 31. Added a use_postmaster feature to recipient callouts.
3017 32. Added the $body_zerocount variable, containing the number of binary zero
3018 bytes in the message body.
3020 33. The time of last modification of the "new" subdirectory is now used as the
3021 "mailbox time last read" when there is a quota error for a maildir
3024 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3026 35. Added +ignore_unknown as a special item in host lists.
3028 36. Code for decoding IPv6 addresses in host lists is now included, even if
3029 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3030 address was recognized as an IP address, but was then not correctly decoded
3031 into binary, causing unexpected and incorrect effects when compared with
3038 1. Very minor rewording of debugging text in manualroute to say "list of
3039 hosts" instead of "hostlist".
3041 2. If verify=header_syntax was set, and a header line with an unqualified
3042 address (no domain) and a large number of spaces between the end of the
3043 name and the colon was received, the reception process suffered a buffer
3044 overflow, and (when I tested it) crashed. This was caused by some obsolete
3045 code that should have been removed. The fix is to remove it!
3047 3. When running in the test harness, delay a bit after writing a bounce
3048 message to get a bit more predictability in the log output.
3050 4. Added a call to search_tidyup() just before forking a reception process. In
3051 theory, someone could use a lookup in the expansion of smtp_accept_max_
3052 per_host which, without the tidyup, could leave open a database connection.
3054 5. Added the variables $recipient_data and $sender_data which get set from a
3055 lookup success in an ACL "recipients" or "senders" condition, or a router
3056 "senders" option, similar to $domain_data and $local_part_data.
3058 6. Moved the writing of debug_print from before to after the "senders" test
3061 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3062 problems for message scanning, either using a data ACL, or using
3063 local_scan() because the Received: header was not generated till after they
3064 were called (in order to set the time as the time of reception completion).
3065 I have revised the way this works. The header is now generated after the
3066 body is received, but before the ACL or local_scan() are called. After they
3067 are run, the timestamp in the header is updated.
3073 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3074 before starting a queue runner without re-exec. This happened only when
3075 deliver_drop_privilege was set or when the Exim user was set to root. The
3076 effect of the bug was that timeouts during subsequent deliveries caused
3077 crashes instead of being properly handled. The handler is now left at its
3078 default (and expected) setting.
3080 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3081 message, again when deliver_drop_privilege is set or Exim is run as root.
3082 The bug described in (1) was not present in this case, but the tidying up
3083 of the other signals was missing. I have made the two cases consistent.
3085 3. The ignore_target_hosts setting on a manualroute router was being ignored
3086 for hosts that were looked up using the /MX notation.
3088 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3091 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3092 operated on the sender address. After changing the $sender_address to <>
3093 for the sender address verify, Exim was re-instated it as the original
3094 (before rewriting) address, but remembering that it had rewritten it, so it
3095 wasn't rewriting it again. This bug also had the effect of breaking the
3096 sender address verification caching when the sender address was rewritten.
3098 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3099 This has been changed so that if the ip literal address matches
3100 ignore_target_hosts, the router declines.
3102 7. Added expansion conditions match_domain, match_address, and match_local_
3103 part (NOT match_host).
3105 8. The placeholder for the Received: header didn't have a length field set.
3107 9. Added code to Exim itself and to exim_lock to test for a specific race
3108 condition that could lead to file corruption when using MBX delivery. The
3109 issue is with the lockfile that is created in /tmp. If this file is removed
3110 after a process has opened it but before that process has acquired a lock,
3111 there is the potential for a second process to recreate the file and also
3112 acquire a lock. This could lead to two Exim processes writing to the file
3113 at the same time. The added code performs the same test as UW imapd; it
3114 checks after acquiring the lock that its file descriptor still refers to
3115 the same named file.
3117 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3118 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3121 11. Added the smtp_active_hostname option. If used, this will typically be made
3122 to depend on the incoming interface address. Because $interface_address is
3123 not set up until the daemon has forked a reception process, error responses
3124 that can happen earlier (such as "too many connections") no longer contain
3127 12. If an expansion in a condition on a "warn" statement fails because a lookup
3128 defers, the "warn" statement is abandoned, and the next ACL statement is
3129 processed. Previously this caused the whole ACL to be aborted.
3131 13. Added the iplsearch lookup type.
3133 14. Added ident_timeout as a log selector.
3135 15. Added tls_certificate_verified as a log selector.
3137 16. Added a global option tls_require_ciphers (compare the smtp transport
3138 option of the same name). This controls incoming TLS connections.
3140 17. I finally figured out how to make tls_require_ciphers do a similar thing
3141 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3142 before starting the TLS session.
3144 18. Tabs are now shown as \t in -bP output.
3146 19. If the log selector return_path_on_delivery was set, Exim crashed when
3147 bouncing a message because it had too many Received: header lines.
3149 20. If two routers both had headers_remove settings, and the first one included
3150 a superfluous trailing colon, the final name in the first list and the
3151 first name in the second list were incorrectly joined into one item (with a
3152 colon in the middle).
3158 1. Added -C and -D options to the exinext utility, mainly to make it easier
3159 to include in the automated testing, but these could be helpful when
3160 multiple configurations are in use.
3162 2. The exinext utility was not formatting the output nicely when there was
3163 an alternate port involved in the retry record key, nor when there was a
3164 message id as well (for retries that were specific to a specific message
3165 and a specific host). It was also confused by IPv6 addresses, because of
3166 the additional colons they contain. I have fixed the IPv4 problem, and
3167 patched it up to do a reasonable job for IPv6.
3169 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3170 delivery, the log line now contains "pipelined" if PIPELINING was used.
3172 4. An SMTP transport process used to panic and die if the bind() call to set
3173 an explicit outgoing interface failed. This has been changed; it is now
3174 treated in the same way as a connect() failure.
3176 5. A reference to $sender_host_name in the part of a conditional expansion
3177 that was being skipped was still causing a DNS lookup. This no longer
3180 6. The def: expansion condition was not recognizing references to header lines
3181 that used bh_ and bheader_.
3183 7. Added the _cache feature to named lists.
3185 8. The code for checking quota_filecount in the appendfile transport was
3186 allowing one more file than it should have been.
3188 9. For compatibility with Sendmail, the command line option
3196 and sets the incoming protocol and host name (for trusted callers). The
3197 host name and its colon can be omitted when only the protocol is to be set.
3198 Note the Exim already has two private options, -pd and -ps, that refer to
3199 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3200 "s", but I don't think that's a major issue.
3202 10. A number of refactoring changes to the code, none of which should affect
3205 (a) The number of logging options was getting close to filling up the
3206 32-bit word that was used as a bit map. I have split them into two classes:
3207 those that are passed in the argument to log_write(), and those that are
3208 only ever tested independently outside of that function. These are now in
3209 separate 32-bit words, so there is plenty of room for expansion again.
3210 There is no change in the user interface or the logging behaviour.
3212 (b) When building, for example, log lines, the code previously used a
3213 macro that called string_cat() twice, in order to add two strings. This is
3214 not really sufficiently general. Furthermore, there was one instance where
3215 it was actually wrong because one of the argument was used twice, and in
3216 one call a function was used. (As it happened, calling the function twice
3217 did not affect the overall behaviour.) The macro has been replaced by a
3218 function that can join an arbitrary number of extra strings onto a growing
3221 (c) The code for expansion conditions now uses a table and a binary chop
3222 instead of a serial search (which was left over from when there were very
3223 few conditions). Also, it now recognizes conditions like "pam" even when
3224 the relevant support is not compiled in: a suitably worded error message is
3225 given if an attempt is made to use such a condition.
3227 11. Added ${time_interval:xxxxx}.
3229 12. A bug was causing one of the ddress fields not to be passed back correctly
3230 from remote delivery subprocesses. The field in question was not being
3231 subsequently used, so this caused to problems in practice.
3233 13. Added new log selectors queue_time and deliver_time.
3235 14. Might have fixed a bug in maildirsizefile handling that threw up
3236 "unexpected character" debug warnings, and recalculated the data
3237 unnecessarily. In any case, I expanded the warning message to give more
3240 15. Added the message "Restricted characters in address" to the statements in
3241 the default ACL that block characters like @ and % in local parts.
3243 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3244 Three changes have been made:
3246 (a) There was a serious bug; a negative response to MAIL caused the whole
3247 recipient domain to be cached as invalid, thereby blocking all messages
3248 to all local parts at the same domain, from all senders. This bug has
3249 been fixed. The domain is no longer cached after a negative response to
3250 MAIL if the sender used is not empty.
3252 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3255 (c) A new callout option, "use_sender" has been added for people who want
3256 the modified behaviour.
3262 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3265 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3266 indeed breaks things for older releases.
3268 3. Added additional logging to the case where there is a problem reading data
3269 from a filter that is running in a subprocess using a pipe, in order to
3270 try to track down a specific problem.
3272 4. Testing facility fudge: when running in the test harness and attempting
3273 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3274 getting "No route to host". Convert this to a timeout.
3276 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3279 6. Some OS don't have socklen_t but use size_t instead. This affects the
3280 fifth argument of getsockopt() amongst other things. This is now
3281 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3282 can be set for individual OS. I have set it for SunOS5, OSF1, and
3283 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3284 some earlier ones do not.
3286 7. Change 4.30/15 was not doing the test caselessly.
3288 8. The standard form for an IPv6 address literal was being rejected by address
3289 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3290 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3291 this, as well as the form without the "IPv6" on the front (but only when
3292 address literals are enabled, of course).
3294 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3296 10. Exim crashed if a message with an empty sender address specified by -f
3297 encountered a router with an errors_to setting. This could be provoked only
3298 by a command such as
3302 where an empty string was supplied; "<>" did not hit this bug.
3304 11. Installed PCRE release 4.5.
3306 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3307 remained set. It is now erased.
3309 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3310 times from message ids (which are base 36 rather than the normal 62).
3312 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3313 were being counted as actual protocol errors, and logged if the log
3314 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3315 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3316 and DATA following a set of rejected RCPTs do not count as protocol errors.
3317 In other words, Exim assumes they were pipelined, though this may not
3318 actually be the case. Of course, in all cases the client gets an
3319 appropriate error code.
3321 15. If a lookup fails in an ACL condition, a message about the failure may
3322 be available; it is used if testing the ACL cannot continue, because most
3323 such messages specify what the cause of the deferral is. However, some
3324 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3325 that caused an old message to be retained and used if a later statement
3326 caused a defer, replacing the real cause of the deferral.
3328 16. If an IP address had so many PTR records that the DNS lookup buffer
3329 was not large enough to hold them, Exim could crash while trying to process
3330 the truncated data. It now detects and logs this case.
3332 17. Further to 4.21/58, another change has been made: if (and only if) the
3333 first line of a message (the first header line) ends with CRLF, a bare LF
3334 in a subsequent header line has a space inserted after it, so as not to
3335 terminate the header.
3337 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3338 unnecessarily, used atoi() instead of strtol(), and didn't check the
3339 termination when getting file sizes from file names by regex.
3341 19. Completely re-implemented the support for maildirsize files, in the light
3342 of a number of problems with the previous contributed implementation
3343 (4.30/29). In particular:
3345 . If the quota is zero, the maildirsize file is maintained, but no quota is
3348 . If the maildir directory does not exist, it is created before any attempt
3349 to write a maildirsize file.
3351 . The quota value in the file is just a cache; if the quota is changed in
3352 the transport, the new value overrides.
3354 . A regular expression is available for excluding directories from the
3357 20. The autoreply transport checks the characters in options that define the
3358 message's headers; it allows continued headers, but it was checking with
3359 isspace() after an embedded newline instead of explicitly looking for a
3362 21. If all the "regular" hosts to which an address was routed had passed their
3363 expiry times, and had not reached their retry times, the address was
3364 bounced, even if fallback hosts were defined. Now Exim should go on to try
3367 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3368 equivalent code in the SMTP transport. Some hosts send humungous responses
3369 to HELO/EHLO, more than 1024 it seems.
3371 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3372 gives compiler warnings in some environments. I've now done it "properly",
3375 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3376 (because of problems with the built-in one) was declared to return uschar *
3377 instead of char *, causing compiler failure.
3379 25. Fixed a file descriptor leak when processing alias/forward files.
3381 26. Fixed a minor format string issue in dbfn.c.
3383 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3385 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3386 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3389 29. When a sender address is verified, it is cached, to save repeating the test
3390 when there is more than one recipient in a message. However, when the
3391 verification involves a callout, it is possible for different callout
3392 options to be set for different recipients. It is too complicated to keep
3393 track of this in the cache, so now Exim always runs a verification when a
3394 callout is required, relying on the callout cache for the optimization.
3395 The overhead is duplication of the address routing, but this should not be
3398 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3399 to be verified with callout=postmaster, and the main callout worked but the
3400 postmaster check failed, the verification correctly failed. However, if a
3401 subsequent RCPT command asked for sender verification *without* the
3402 postmaster check, incorrect caching caused this verification also to fail,
3405 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3406 it was not caching the DNS options (qualify_single, search_parents) that
3407 were used when the lookup failed. A subsequent lookup with different
3408 options therefore always gave the same answer, though there were cases
3409 where it should not have. (Example: a "domains = !$mx_any" option on a
3410 dnslookup router: the "domains" option is always processed without any
3411 widening, but the router might have qualify_single set.) Now Exim uses the
3412 cached value only when the same options are set.
3414 32. Added John Jetmore's "exipick" utility to the distribution.
3416 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3417 than a timeout (e.g. a certificate is required, and is not provided), an
3418 Exim server now closes the connection immediately. Previously it waited for
3419 the client to close - but if the client is SSL, it seems that they each
3420 wait for each other, leading to a delay before one of them times out.
3422 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3423 maintained 0.8.x compatibility because I don't think many are using it, and
3424 it is clearly obsolete.
3426 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3429 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3430 client certificate was expired. A simple patch fixes this, though I don't
3431 understand the full logic of why the verify callback is called multiple
3434 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3435 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3436 which causes problems with some clients (such as the Certicom SSL Plus
3437 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3438 disables the coutermeasure allowing Eudora to connect."
3440 38. Exim was not checking that a write() to a log file succeeded. This could
3441 lead to Bad Things if a log got too big, in particular if it hit a file
3442 size limit. Exim now panics and dies if it cannot write to a log file, just
3443 as it does if it cannot open a log file.
3445 39. Modified OS/Makefile-Linux so that it now contains
3447 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3449 The two -D definitions ensure that Exim is compiled with large file
3450 support, which makes it possible to handle log files that are bigger than
3453 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3454 instance) a domain was checked against a named list that involved a lookup,
3455 causing $domain_data to be set, then another domain was checked against the
3456 same list, then the first domain was re-checked, the value of $domain_data
3457 after the final check could be wrong. In particular, if the second check
3458 failed, it could be set empty. This bug probably also applied to
3461 41. The strip_trailing_dot option was not being applied to the address given
3462 with the -f command-line option.
3464 42. The code for reading a message's header from the spool was incrementing
3465 $received_count, but never initializing it. This meant that the value was
3466 incorrect (doubled) while delivering a message in the same process in which
3467 it was received. In the most common configuration of Exim, this never
3468 happens - a fresh exec is done - but it can happen when
3469 deliver_drop_privilege is set.
3471 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3472 it now includes up to 150 characters of the unexpected data in the log
3475 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3476 and building data strings. The size of both of these buffers was 10 000
3477 bytes - far larger than anybody would *ever* want, thought I. Needless to
3478 say, somebody hit the limit. I have increased the maximum line length to
3479 20 000 and the maximum data length of concatenated lines to 100 000. I have
3480 also fixed two bugs, because there was no checking on these buffers. Tsk,
3481 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3482 buffer is too small.
3484 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3485 lsearch lookups. Now it does.
3487 46. When parsing a route_list item in a manualroute router, a fixed-length
3488 buffer was used for the list of hosts. I made this 1024 bytes long,
3489 thinking that nobody would ever have a list of hosts that long. Wrong.
3490 Somebody had a whole pile of complicated expansion conditions, and the
3491 string was silently truncated, leading to an expansion error. It turns out
3492 that it is easier to change to an unlimited length (owing to other changes
3493 that have happened since this code was originally written) than to build
3494 structure for giving a limitation error. The length of the item that
3495 expands into the list of hosts is now unlimited.
3497 47. The lsearch lookup could not handle data where the length of text line was
3498 more than 4095 characters. Such lines were truncated, leading to shortened
3499 data being returned. It should now handle lines of any length.
3501 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3502 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3505 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3506 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3507 no longer happens in comment lines. A list of replacements is now placed
3508 at the head of all of the source files, except those whose only change is
3509 to replace PERL_COMMAND in the very first #! line.
3511 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3512 messages on the queue, with a bottom-up merge sort, using code contributed
3513 by Michael Haardt. This should make operations like -bp somewhat faster on
3514 large queues. It won't affect queue runners, except when queue_run_in_order
3517 51. Installed eximstats 1.31 in the distribution.
3519 52. Added support for SRV lookups to the dnslookup router.
3521 53. If an ACL referred to $message_body or $message_body_end, the value was not
3522 reset for any messages that followed in the same SMTP session.
3524 54. The store-handling optimization for building very long strings was not
3525 differentiating between the different store pools. I don't think this
3526 actually made any difference in practice, but I've tidied it.
3528 55. While running the routers to verify a sender address, $sender_address
3529 was still set to the sender address. This is wrong, because when routing to
3530 send a bounce to the sender, it would be empty. Therefore, I have changed
3531 it so that, while verifying a sender address, $sender_address is set to <>.
3532 (There is no change to what happens when verifying a recipient address.)
3534 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3535 target A or AAAA records (if not already returned) without resetting the
3536 qualify_single or search_parents options of the DNS resolver. These are
3537 inappropriate in this case because the targets of MX and SRV records must
3538 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3539 target that, when qualified, matched something in the local domain. These
3540 two options are now turned off when doing these lookups.
3542 57. It seems that at least some releases of Reiserfs (which does not have the
3543 concept of a fixed number of inodes) returns zero and not -1 for the
3544 number of available inodes. This interacted badly with check_spool_inodes,
3545 which assumed that -1 was the "no such thing" setting. What I have done is
3546 to check that the total number of inodes is greater than zero before doing
3547 the test of how many are available.
3549 58. When a "warn" ACL statement has a log_message modifier, the message is
3550 remembered, and not repeated. This is to avoid a lot of repetition when a
3551 message has many recipients that cause the same warning to be written.
3552 Howewer, Exim was preserving the list of already written lines for an
3553 entire SMTP session, which doesn't seem right. The memory is now reset if a
3554 new message is started.
3556 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3557 debugging output unless log_rewrite was also set.
3559 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3560 of (int)(handle) when we know that handle contains (void *)(-1).
3562 61. The Exim daemon panic-logs an error return when it closes the incoming
3563 connection. However "connection reset by peer" seems to be common, and
3564 isn't really an error worthy of noting specially, so that particular error
3567 62. When Exim is trying to find all the local interfaces, it used to panic and
3568 die if the ioctl to get the interface flags failed. However, it seems that
3569 on at least one OS (Solaris 9) it is possible to have an interface that is
3570 included in the list of interfaces, but for which you get a failure error
3571 for this call. This happens when the interface is not "plumbed" into a
3572 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3573 failure of the "get flags" call assumes that the interface is down.
3575 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3576 makes life easier for people who are doing arithmetic on fields extracted
3577 from dates, where you often get leading zeros that should not be
3578 interpreted as octal.
3580 64. Added qualify_domain to the redirect router, to override the global
3583 65. If a pathologically long header line contained very many addresses (the
3584 report of this problem mentioned 10 000) and each of them was rewritten,
3585 Exim could use up a very large amount of memory. (It kept on making new
3586 copies of the header line as it rewrote, and never released the old ones.)
3587 At the expense of a bit more processing, the header rewriting function has
3588 been changed so that it no longer eats memory in this way.
3590 66. The generation of the Received: header has been moved from the time that a
3591 message starts to be received, to the time that it finishes. The timestamp
3592 in the Received: header should now be very close to that of the <= log
3593 line. There are two side-effects of this change:
3595 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3596 logged header lines no longer include the local Received: line, because
3597 it has not yet been created. The same applies to a copy of the message
3598 that is returned to a non-SMTP sender when a message is rejected.
3600 (b) When a filter file is tested using -bf, no additional Received: header
3601 is added to the test message. After some thought, I decided that this
3604 This change does not affect the value of $received_for. It is still set
3605 after address rewriting, but before local_scan() is called.
3607 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3609 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3610 gave an unhelpful panic error message, and a defer error. I have managed to
3611 change this behaviour so that it now rejects any supplied certificate,
3612 which seems right, as the list of acceptable certificates is empty.
3614 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3615 gave an unhelpful defer error. I have not managed to make this reject any
3616 supplied certificates, but the error message it gives is "no certificate
3617 supplied", which is not helpful.
3619 70. exigrep's output now also includes lines that are not associated with any
3620 message, but which match the given pattern. Implemented by a patch from
3621 Martin Sluka, which also tidied up the Perl a bit.
3623 71. Recipient callout verification, like sender verification, was using <> in
3624 the MAIL FROM command. This isn't really the right thing, since the actual
3625 sender may affect whether the remote host accepts the recipient or not. I
3626 have changed it to use the actual sender in the callout; this means that
3627 the cache record is now keyed on a recipient/sender pair, not just the
3628 recipient address. There doesn't seem to be a real danger of callout loops,
3629 since a callout by the remote host to check the sender would use <>.
3630 [SEE ABOVE: changed after hitting problems.]
3632 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3633 temporary errors. However, in the case of such a code being given after
3634 the end of a data transmission (i.e. after ".") Exim was failing to write
3635 a retry record for the message. (Yes, there was some broken host that was
3636 actually sending 8xx at this point.)
3638 73. An unknown lookup type in a host list could cause Exim to panic-die when
3639 the list was checked. (An example that provoked this was putting <; in the
3640 middle of a list instead of at the start.) If this happened during a DATA
3641 ACL check, a -D file could be left lying around. This kind of configuration
3642 error no longer causes Exim to die; instead it causes a defer errror. The
3643 incident is still logged to the main and panic logs.
3645 74. Buglet left over from Exim 3 conversion. The message "too many messages
3646 in one connection" was written to the rejectlog but not the mainlog, except
3647 when address rewriting (yes!) was being logged.
3649 75. Added write_rejectlog option.
3651 76. When a system filter was run not as root (that is, when system_filter_user
3652 was set), the values of the $n variables were not being returned to the
3653 main process; thus, they were not subsequently available in the $sn
3656 77. Added +return_path_on_delivery log selector.
3658 78. A connection timeout was being treated differently from recipients deferred
3659 when testing hosts_max_try with a message that was older than the host's
3660 retry timeout. (The host should not be counted, thus allowing all hosts to
3661 be tried at least once before bouncing.) This may have been the cause of an
3662 occasionally reported bug whereby a message would remain on the queue
3663 longer than the retry timeout, but would be bounced if a delivery was
3664 forced. I say "may" because I never totally pinned down the problem;
3665 setting up timeout/retry tests is difficult. See also the next item.
3667 79. The ultimate address timeout was not being applied to errors that involved
3668 a combination of host plus message (for example, a timeout on a MAIL
3669 command). When an address resolved to a number of possible hosts, and they
3670 were not all tried for each delivery (e.g. because of hosts_max_try), a
3671 message could remain on the queue longer than the retry timeout.
3673 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3676 81. Fixed an obscure SMTP outgoing bug which required at least the following
3677 conditions: (a) there was another message waiting for the same server;
3678 (b) the server returned 5xx to all RCPT commands in the first message so
3679 that the message was not completed; (c) the server dropped the connection
3680 or gave a negative response to the RSET that Exim sends to abort the
3681 transaction. The observed case was a dropped connection after DATA that had
3682 been sent in pipelining mode. That is, the server had advertised PIPELINING
3683 but was not implementing it correctly. The effect of the bug was incorrect
3684 behaviour, such as trying another host, and this could lead to a crash.
3690 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3691 and daemon.c were passed as pointers to ints; they should have been
3692 pointers to socklen_t variables (which are typically unsigned ints).
3694 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3697 3. Fixed a really odd bug that affected only the testing scheme; patching a
3698 certain fixed string in the binary changed the value of another string that
3699 happened to be identical to the end of the original first string.
3701 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3702 name", it returns that address as the IP address. On some operating
3703 systems (e.g. Solaris), it also passes back the IP address string as the
3704 "host name". However, on others (e.g. Linux), it passes back an empty
3705 string. Exim wasn't checking for this, and was changing the host name to an
3706 empty string, assuming it had been canonicized.
3708 5. Although rare, it is permitted to have more than one PTR record for a given
3709 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3710 all the names associated with an address, because they do in Solaris.
3711 However, it seems that they do not in Linux for data that comes from the
3712 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3713 I found this out when I moved to a new Linux workstation and tried to run
3714 the Exim test suite.
3716 To get round this problem I have changed the code so that it now does its
3717 own call to the DNS to look up PTR records when searching for a host name.
3718 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3719 addresses that are only in /etc/hosts are still found.
3721 This behaviour is, however, controlled by an option called host_lookup_
3722 order, which defaults to "bydns:byaddr". If people want to use the other
3723 order, or indeed, just use one or the other means of lookup, they can
3724 specify it in this variable.
3726 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3727 some operating systems, this comes back from gethostbyaddr() as an empty
3728 string, and this is what Exim used to test for. However, it seems that in
3729 other systems, "." is yielded. Exim now tests for this case too.
3731 7. The values of check_spool_space and check_log_space are now held internally
3732 as a number of kilobytes instead of an absolute number of bytes. If a
3733 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3734 kilobyte. This means that much larger values can be stored.
3736 8. Exim monitor: an attempt to get the action menu when not actually pointing
3737 at a message produces an empty menu entitled "No message selected". This
3738 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3739 no entries in it ("Shell widget menu has zero width and/or height"). So I
3740 have added a single, blank menu entry in this case.
3742 9. Added ${quote_local_part.
3744 10. MIME decoding is now applied to the contents of Subject: header lines when
3747 11. Now that a reference to $sender_host_address automatically causes a reverse
3748 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3749 host lookup before query-style lookups in lists that might use this
3750 variable. This has therefore been abolished, and the "net-" prefix is no
3751 longer necessary for query-style lookups.
3753 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3754 have been a typo for LFLAGS, so it has been changed.
3756 13. The install script calls Exim with "-C /dev/null" in order to find the
3757 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3758 to be output. Howeve, since Exim outputs its version number before the
3759 error, it didn't break the script. It just looked ugly. I fixed this by
3760 always allowing "-C /dev/null" if the caller is root.
3762 14. Ignore overlarge ACL variable number when reading spool file - insurance
3763 against a later release with more variables having written the file.
3765 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3766 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3767 this, as well as the form without the "IPv6" on the front.
3769 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3770 OS/Makefile-Darwin file.
3772 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3773 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3775 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3776 variable. However, some OS use ints for inodes. I've added cast to long int
3777 to get rid of the compiler warning.
3779 19. I had forgotten to lock out "/../" in configuration file names when
3780 ALT_CONFIG_PREFIX was set.
3782 20. Routers used for verification do not need to specify transports. However,
3783 if such a router generated a host list, and callout was configured, Exim
3784 crashed, because it could not find a port number from the (non-existent)
3785 transport. It now assumes port 25 in this circumstance.
3787 21. Added the -t option to exigrep.
3789 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3790 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3791 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3792 them is set, LOOKUP_LSEARCH is forced.
3794 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3796 24. Added sender and host information to the "rejected by local_scan()" log
3797 line; previously there was no indication of these.
3799 25. Added .include_if_exists.
3801 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3802 Linux. It turns out that not all file systems support this. Apparently some
3803 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3804 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3805 not supported on the file descriptor, is now ignored when Exim is trying to
3806 sync a directory. This applies only to Linux.
3808 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3810 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3811 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3814 29. Added support for maildirsize files from supplied patch (modified a bit).
3816 30. The use of :fail: followed by an empty string could lead Exim to respond to
3817 sender verification failures with (e.g.):
3819 550 Verification failed for <xxx>
3820 550 Sender verify failed
3822 where the first response line was missing the '-' that indicates it is not
3823 the final line of the response.
3825 31. The loop for finding the name of the user that called Exim had a hardwired
3826 limit of 10; it now uses the value of finduser_retries, which is used for
3827 all other user lookups.
3829 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3832 33. Exim was neglecting to zero errno before one call of strtol() when
3833 expanding a string and expecting an integer value. On some systems this
3834 resulted in spurious "integer overflow" errors. Also, it was casting the
3835 result into an int without checking.
3837 34. Testing for a connection timeout using "timeout_connect" in the retry rules
3838 did not work. The code looks as if it has *never* worked, though it appears
3839 to have been documented since at least releast 1.62. I have made it work.
3841 35. The "timeout_DNS" error in retry rules, also documented since at least
3842 1.62, also never worked. As it isn't clear exactly what this means, and
3843 clearly it isn't a major issue, I have abolished the feature by treating it
3844 as "timeout", and writing a warning to the main and panic logs.
3846 36. The display of retry rules for -brt wasn't always showing the error code
3849 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
3850 timeout_connect_A, timeout_connect_MX.
3852 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
3853 to the empty sender.
3855 39. The daemon was not analysing the content of -oX till after it had closed
3856 stderr and disconnected from the controlling terminal. This meant that any
3857 syntax errors were only noted on the panic log, and the return code from
3858 the command was 0. By re-arranging the code a little, I've made the
3859 decoding happen first, so such errors now appear on stderr, and the return
3860 code is 1. However, the actual setting up of the sockets still happens in
3861 the disconnected process, so errors there are still only recorded on the
3864 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
3865 connections (as happens on some IP stacks) was logged at start up time as
3866 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
3867 it from "IPv6 and IPv4", which means that two separate sockets are being
3870 41. The debug output for gethostbyname2() or getipnodebyname() failures now
3871 says whether AF_INET or AF_INET6 was passed as an argument.
3873 42. Exiwhat output was messed up when time zones were included in log
3876 43. Exiwhat now gives more information about the daemon's listening ports,
3877 and whether -tls-on-connect was used.
3879 44. The "port" option of the smtp transport is now expanded.
3881 45. A "message" modifier in a "warn" statement in a non-message ACL was being
3882 silently ignored. Now an error message is written to the main and panic
3885 46. There's a new ACL modifier called "logwrite" which writes to a log file
3886 as soon as it is encountered.
3888 47. Added $local_user_uid and $local_user_gid at routing time.
3890 48. Exim crashed when trying to verify a sender address that was being
3893 49. Exim was recognizing only a space character after ".include". It now also
3894 recognizes a tab character.
3896 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
3897 extracting the relevant information from the specification. The man page no
3898 longer contains scrambled data for the -d option, and I've added a section
3899 at the front about calling Exim under different names.
3901 51. Added "extra_headers" argument to the "mail" command in filter files.
3903 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
3906 53. Installed eximstats 1.29.
3908 54. Added transport_filter_timeout as a generic transport option.
3910 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
3911 Cc: header lines. This was required by RFC 822, but it not required by RFC
3914 56. Exim used to add From:, Date:, and Message-Id: header lines to any
3915 incoming messages that did not have them. Now it does so only if the
3916 message originates locally, that is, if there is no associated remote host
3917 address. When Resent- header lines are present, this applies to the Resent-
3918 lines rather than the non-Resent- lines.
3920 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
3921 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
3923 58. Messages for configuration errors now include the name of the main
3924 configuration file - useful now that there may be more than one file in a
3925 list (.included file names were always shown).
3927 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
3928 for those rare installations that do not start the daemon as root or run it
3929 setuid root. I've cut out the call to initgroups() if the daemon is not
3932 60. The Exim user and group can now be bound into the binary as text strings
3933 that are looked up at the start of Exim's processing.
3935 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
3937 62. Added $mailstore_basename variable.
3939 63. Installed patch to sieve.c from Michael Haardt.
3941 64. When Exim failed to open the panic log after failing to open the main log,
3942 the original message it was trying to log was written to stderr and debug
3943 output, but if they were not available (the usual case in production), it
3944 was lost. Now it is written to syslog before the two lines that record the
3945 failures to open the logs.
3947 65. Users' Exim filters run in subprocesses under the user's uid. It is
3948 possible for a "deliver" command or an alias in a "personal" command to
3949 provoke an address rewrite. If logging of address rewriting is configured,
3950 this fails because the process is not running as root or exim. There may be
3951 a better way of dealing with this, but for the moment (because 4.30 needs
3952 to be released), I have disabled address rewrite logging when running a
3953 filter in a non-root, non-exim process.
3959 1. The buildconfig auxiliary program wasn't quoting the value set for
3960 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
3961 not defined. This bug was masked in 4.22 by the effect that was fixed in
3964 2. Some messages that were rejected after a message id was allocated were
3965 shown as "incomplete" by exigrep. It no longer does this for messages that
3966 are rejected by local_scan() or the DATA or non-SMTP ACLs.
3968 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
3969 have allow_domain_literals set, the ID did not get logged in the <= line.
3970 Domain literals are now always recognized in Message-ID: header lines.
3972 4. The first argument for a ${extract expansion item is the key name or field
3973 number. Leading and trailing spaces in this item were not being ignored,
3974 causing some misleading effects.
3976 5. When deliver_drop_privilege was set, single queue runner processes started
3977 manually (i.e. by the command "exim -q") or by the daemon (which uses the
3978 same command in the process it spins off) were not dropping privilege.
3980 6. When the daemon running as "exim" started a queue runner, it always
3981 re-executed Exim in the spun-off process. This is a waste of effort when
3982 deliver_drop_privilege is set. The new process now just calls the
3983 queue-runner function directly.
3989 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
3992 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
3993 ignored. Though the use of -oP was forcing the writing of a pid file, it
3994 was always written to the default place.
3996 3. If the message "no IP address found for host xxxx" is generated during
3997 incoming verification, it is now followed by identification of the incoming
3998 connection (so you can more easily find what provoked it).
4000 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4002 5. Added some features to "harden" Exim a bit more against certain attacks:
4004 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4005 be put in Local/Makefile. This is like the never_users runtime option,
4006 but it cannot be overridden. The default setting is "root".
4008 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4009 prefix string with which any file named in a -C command line option
4012 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4013 is retained for -C and -D only if the caller of Exim is root. Without
4014 it, the exim user may also use -C and -D and retain privilege.
4016 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4017 command line option is disabled.
4019 6. Macro names set by the -D option must start with an upper case letter, just
4020 like macro names defined in the configuration file.
4022 7. Added "dereference=" facility to LDAP.
4024 8. Two instances of the typo "uknown" in the source files are fixed.
4026 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4027 the Configure-Makefile script screwed up while processing it.
4029 10. Incorporated PCRE 4.4.
4031 11. The SMTP synchronization check was not operating right at the start of an
4032 SMTP session. For example, it could not catch a HELO sent before the client
4033 waited for the greeting. There is now a check for outstanding input at the
4034 point when the greeting is written. Because of the duplex, asynchronous
4035 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4036 way, but not yet received, when the check is performed.
4038 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4039 on TCP/IP sockets, because this apparently causes some broken clients to
4042 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4043 unchanged) from the Cygwin maintainer.
4045 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4046 was supported instead of testing for mbx. Effectively a typo.
4048 15. The spa authenticator server code was not checking that the input it
4049 received was valid base64.
4051 16. The debug output line for the "set" modifier in ACLs was not showing the
4052 name of the variable that was being set.
4054 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4056 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4057 lookup on its own. Something else was needed to trigger the lookup. For
4058 example, a match in host_lookup or the need for a host name in a host list.
4059 Now, if $sender_host_name is referenced and the host name has not yet been
4060 looked up, a lookup is performed. If the lookup fails, the variable remains
4061 empty, and $host_lookup_failed is set to "1".
4063 19. Added "eqi" as a case-independent comparison operator.
4065 20. The saslauthd authentication condition could segfault if neither service
4066 nor realm was specified.
4068 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4069 error message that was logged was misleading, and incoming SMTP
4070 connections were dropped. The message is now more accurate, and temporary
4071 errors are given to SMTP connections.
4073 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4074 (except RSET) with a 421 error, until QUIT is received. However, it was
4075 failing to send a response to QUIT.
4077 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4078 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4079 ACL is now run after the helo_try_verify_hosts code.
4081 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4082 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4083 case-independent, so other case variants are also recognized). Apparently
4084 some systems use these upper case variants.
4086 25. If more than two messages were waiting for the same host, and a transport
4087 filter was specified for the transport, Exim sent two messages over the
4088 same TCP/IP connection, and then failed with "socket operation on non-
4089 socket" when it tried to send the third.
4091 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4093 27. The extern definition of crypt16() in expand.c was not being excluded when
4094 the OS had its own crypt16() function.
4096 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4097 as a preferred synonym for return_size_limit, both as an option and as an
4100 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4102 30. Changed the default configuration ACL to relax the local part checking rule
4103 for addresses that are not in any local domains. For these addresses,
4104 slashes and pipe symbols are allowed within local parts, but the sequence
4105 /../ is explicitly forbidden.
4107 31. SPA server authentication was not clearing the challenge buffer before
4110 32. log_message in a "warn" ACL statement was writing to the reject log as
4111 well as to the main log, which contradicts the documentation and doesn't
4112 seem right (because no rejection is happening). So I have stopped it.
4114 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4115 However, I am unable to do any testing of this.
4117 34. Fixed an infelicity in the appendfile transport. When checking directories
4118 for a mailbox, to see if any needed to be created, it was accidentally
4119 using path names with one or more superfluous leading slashes; tracing
4120 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4122 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4123 contents are added to the log line that is written for every discarded
4124 recipient. (Previously a log_message setting was ignored.)
4126 36. The ${quote: operator now quotes the string if it is empty.
4128 37. The install script runs exim in order to find its version number. If for
4129 some reason other than non-existence or emptiness, which it checks, it
4130 could not run './exim', it was installing it with an empty version number,
4131 i.e. as "exim-". This error state is now caught, and the installation is
4134 38. An argument was missing from the function that creates an error message
4135 when Exim fails to connect to the socket for saslauthd authentication.
4136 This could cause Exim to crash, or give a corrupted message.
4138 39. Added isip, isip4, and isip6 to ${if conditions.
4140 40. The ACL variables $acl_xx are now saved with the message, and can be
4141 accessed later in routers, transports, and filters.
4143 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4144 strings in the file are not string-expanded.
4146 42. If a MAIL command specified a SIZE value that was too large to fit into an
4147 int variable, the check against message_size_limit failed. Such values are
4148 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4149 day this will have to be increased, but I don't think I want to be around
4150 when emails are that large.
4157 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4158 iconv() is not standard in FreeBSD.
4160 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4161 IPv6 enabled. The observed symptom was a segmentation fault on return from
4162 the function os_common_find_running_interfaces() in src/os.c.
4164 3. In the check_special_case() function in daemon.c I had used "errno" as an
4165 argument name, which causes warnings on some systems. This was basically a
4166 typo, since it was named "eno" in the comments!
4168 4. The code that waits for the clock to tick (at a resolution of some fraction
4169 of a second) so as to ensure message-id uniqueness was always waiting for
4170 at least one whole tick, when it could have waited for less. [This is
4171 almost certainly not relevant at current processor speeds, where it is
4172 unlikely to ever wait at all. But we try to future-proof.]
4174 5. The function that sleeps for a time interval that includes fractions of a
4175 second contained a race. It did not block SIGALRM between setting the
4176 timer, and suspending (a couple of lines later). If the interval was short
4177 and the sigsuspend() was delayed until after it had expired, the suspension
4178 never ended. On busy systems this could lead to processes getting stuck for
4181 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4182 process, before it forks any delivery processes. The open lookup caching
4183 mechanism meant that the open file or database connection was passed into
4184 the delivery process. The problem was that delivery processes always tidy
4185 up cached lookup data. This could cause a problem for the next delivery
4186 process started by the queue runner, because the external queue runner
4187 process does not know about the closure. So the next delivery process
4188 still has data in the lookup cache. In the case of a file lookup, there was
4189 no problem because closing a file descriptor in a subprocess doesn't affect
4190 the parent. However, if the lookup was caching a connection to a database,
4191 the connection was closed, and the second delivery process was likely to
4192 see errors such as "PGSQL: query failed: server closed the connection
4193 unexpectedly". The problem has been fixed by closing all cached lookups
4194 in a queue runner before running a delivery process.
4196 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4197 seem to have the "const" qualifier which it has on other OS. I've
4200 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4201 *of the same type* (client or server) with the same public name that an
4202 error should be diagnosed.
4204 9. When Exim looked up a host name for an IP address, but failed to find the
4205 original IP address when looking up the host name (a safety check), it
4206 output the message "<ip address> does not match any IP for NULL", which was
4207 confusing, to say the least. The bug was that the host name should have
4208 appeared instead of "NULL".
4210 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4211 user that is built into the binary, and the -C or -D options is used, root
4212 privilege is dropped before the configuration file is read. In addition,
4213 logging is switched to stderr instead of the normal log files. If the
4214 configuration then re-defines the Exim user, the unprivileged environment
4215 is probably not what is expected, so Exim logs a panic warning message (but
4218 However, if deliver_drop_privilege is set, the unprivileged state may well
4219 be exactly what is intended, so the warning has been cut out in that case,
4220 and Exim is allowed to try to write to its normal log files.
4226 1. smtp_return_error_details was not giving details for temporary sender
4227 or receiver verification errors.
4229 2. Diagnose a configuration error if two authenticators have the same public
4232 3. Exim used not to create the message log file for a message until the first
4233 delivery attempt. This could be confusing when incoming messages were held
4234 for policy or load reasons. The message log file is now created at the time
4235 the message is received, and an initial "Received" line is written to it.
4237 4. The automatically generated man page for command line options had a minor
4238 bug that caused no ill effects; however, a more serious problem was that
4239 the procedure for building the man page automatically didn't always
4240 operate. Consequently, release 4.20 contains an out-of-date version. This
4241 shouldn't happen again.
4243 5. When building Exim with embedded Perl support, the script that builds the
4244 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4245 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4247 6. The freeze_tell option was not being used for messages that were frozen on
4248 arrival, either by an ACL or by local_scan().
4250 7. Added the smtp_incomplete_transaction log selector.
4252 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4253 was accepting AUTH without a new EHLO.
4255 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4256 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4259 10. Logging of TCP/IP connections (when configured) now happens in the main
4260 daemon process instead of the child process, so that the TCP/IP connection
4261 count is more accurate (but it can never be perfect).
4263 11. The use of "drop" in a nested ACL was not being handled correctly in the
4264 outer ACL. Now, if condition failure induced by the nested "drop" causes
4265 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4266 or "require"), the connection is dropped.
4268 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4269 that yield "discard" can now be used with an "accept" or a "discard" verb,
4270 but an error is generated for any others (because I can't see a useful way
4271 to define what should happen).
4273 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4274 are now processed in the same way as lines in the Exim configuration file.
4275 In particular, continuation lines are supported.
4277 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4279 15. Added -ti meaning -t -i.
4281 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4282 domains, and warn by logging if others are found.
4284 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4285 returned from GIFCONF when Exim is trying to find the list of interfaces on
4286 a host. The code in os.c has been modified to copy each ifreq to an aligned
4287 structure in all cases.
4289 Also, in some cases, the returned ifreq's were being copied to a 'struct
4290 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4291 means the last couple of bytes of an IPv6 address could be chopped if the
4292 ifreq contained only a normal sockaddr (14 bytes storage).
4294 18. Named domain lists were not supported in the hosts_treat_as_local option.
4295 An entry such as +xxxx was not recognized, and was treated as a literal
4298 19. Ensure that header lines added by a DATA ACL are included in the reject log
4299 if the ACL subsequently rejects the message.
4301 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4302 MD5 (which is deprecated).
4304 21. When testing a filter file using -bf, Exim was writing a message when it
4305 took the sender from a "From " line in the message, but it was not doing so
4306 when it took $return_path from a Return-Path: header line. It now does.
4308 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4309 with a valid header line field name (a series of printing characters
4310 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4312 23. Changed "disc" in the source to "disk" to conform to the documentation and
4313 the book and for uniformity.
4315 24. Ignore Sendmail's -Ooption=value command line item.
4317 25. When execve() failed while trying to run a command in a pipe transport,
4318 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4319 could be confused with a return value of 69 from the command itself. This
4320 has been changed to 127, the value the shell returns if it is asked to run
4321 a non-existent command. The wording for the related log line suggests a
4322 non-existent command as the problem.
4324 26. If received_header_text expands to an empty string, do not add a Received:
4325 header line to the message. (Well, it adds a token one on the spool, but
4326 marks it "old" so that it doesn't get used or transmitted.)
4328 27. Installed eximstats 1.28 (addition of -nt option).
4330 28. There was no check for failure on the call to getsockname() in the daemon
4331 code. This can fail if there is a shortage of resources on the system, with
4332 ENOMEM, for example. A temporary error is now given on failure.
4334 29. Contrary to the C standard, it seems that in some environments, the
4335 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4336 program. Exim now does this explicitly; it affects the formatting of
4337 timestamps using strftime().
4339 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4340 complaints. I've now initialized all the variables, to avoid this.
4342 32. Header lines added by a system filter were not being "seen" during
4343 transport-time rewrites.
4345 33. The info_callback() function passed to OpenSSL is set up with type void
4346 (*)(SSL *, int, int), as described somewhere. However, when calling the
4347 function (actually a macro) that sets it up, the type void(*)() is
4348 expected. I've put in a cast to prevent warnings from picky compilers.
4350 34. If a DNS black list lookup found a CNAME record, but there were no A
4351 records associated with the domain it pointed at, Exim crashed.
4353 35. If a DNS black list lookup returned more than one A record, Exim ignored
4354 all but the first. It now scans all returned addresses if a particular IP
4355 value is being sought. In this situation, the contents of the
4356 $dnslist_value variable are a list of all the addresses, separated by a
4359 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4360 to accept a host name and all its aliases if the forward lookup for any of
4361 them yielded the IP address of the incoming connection. Now it accepts only
4362 those names whose forward lookup yields the correct IP address. Any other
4363 names are discarded. This closes a loophole whereby a rogue DNS
4364 administrator could create reverse DNS records to break through a
4365 wildcarded host restriction in an ACL.
4367 37. If a user filter or a system filter that ran in a subprocess used any of
4368 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4369 the wrong values were passed to the pipe command ($thisaddress had the
4370 value of $0, $0 had the value of $1, etc). This bug was introduced by
4371 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4373 38. Improved the line breaking for long SMTP error messages from ACLs.
4374 Previously, if there was no break point between 40 and 75 characters, Exim
4375 left the rest of the message alone. Two changes have been made: (a) I've
4376 reduced the minimum length to 35 characters; (b) if it can't find a break
4377 point between 35 and 75 characters, it looks ahead and uses the first one
4378 that it finds. This may give the occasional overlong line, but at least the
4379 remaining text gets split now.
4381 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4382 file descriptors might be low, and that setting 1000 would always raise it.
4383 It turns out that in some environments, the limit is already over 1000 and
4384 that lowering it causes trouble. So now Exim takes care not to decrease it.
4386 40. When delivering a message, the value of $return_path is set to $sender_
4387 address at the start of routing (routers may change the value). By an
4388 oversight, this default was not being set up when an address was tested by
4389 -bt or -bv, which affected the outcome if any router or filter referred to
4392 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4393 log only when "message" or "log_message" are set. However, if one of the
4394 conditions was an address verification, or a call to a nested ACL, the
4395 messages generated by the underlying test were being passed through. This
4396 no longer happens. The underlying message is available in $acl_verify_
4397 message for both "message" and "log_message" expansions, so it can be
4398 passed through if needed.
4400 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4401 new expansion $bh_ to give the encoded byte string without charset
4402 translation. Translation happens only if iconv() is available; HAVE_ICONV
4403 indicates this at build time. HEADERS_CHARSET gives the charset to
4404 translate to; headers_charset can change it in the configuration, and
4405 "headers charset" can change it in an individual filter file.
4407 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4408 that creates RFC 2047 encoded "words" labels them as that charset instead
4409 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4410 expansion operator; (ii) when Exim creates a From: line for a local
4411 message; (iii) when a header line is rewritten to include a "phrase" part.
4413 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4414 buggy, causing it to skip the first lines of messages whose message ID
4415 ended in 'D'. This would not have bitten before Exim release 4.14, because
4416 message IDs were unlikely to end in 'D' before then. The effect was to have
4417 incorrect size information for certain domains.
4419 45. #include "config.h" was missing at the start of the crypt16.c module. This
4420 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4423 46. If there was a timeout during a "random" callout check, Exim treated it as
4424 a failure of the random address, and carried on sending RSET and the real
4425 address. If the delay was just some slowness somewhere, the response to the
4426 original RCPT would be taken as a response to RSET and so on, causing
4427 mayhem of various kinds.
4429 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4430 when I implemented it. It didn't allow for the fact that some option values
4431 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4432 the right test for positive values.
4434 48. Domain names in DNS records are case-independent. Exim always looks them up
4435 in lower case. Some resolvers return domain names in exactly the case they
4436 appear in the zone file, that is, they may contain uppercase letters. Not
4437 all resolvers do this - some return always lower case. Exim was treating a
4438 change of case by a resolver as a change of domain, similar to a widening
4439 of a domain abbreviation. This triggered its re-routing code and so it was
4440 trying to route what was effectively the same domain again. This normally
4441 caused routing to fail (because the router wouldn't handle the domain
4442 twice). Now Exim checks for this case specially, and just changes the
4443 casing of the domain that it ultimately uses when it transmits the message
4446 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4449 50. If a filter generated a file delivery with a non-absolute name (possible if
4450 no home directory exists for the router), the forbid_file option was not
4453 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4454 the existing equality matching.
4456 52. Exim was using ints instead of ino_t variables in some places where it was
4457 dealing with inode numbers.
4459 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4460 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4461 called TMPDIR, and if it finds it is different, it changes its value.
4463 54. The smtp_printf() function is now made available to local_scan() so
4464 additional output lines can be written before returning. There is also an
4465 smtp_fflush() function to enable the detection of a dropped connection.
4466 The variables smtp_input and smtp_batched_input are exported to
4469 55. Changed the default runtime configuration: the message "Unknown user"
4470 has been removed from the ACL, and instead placed on the localuser router,
4471 using the cannot_route_message feature. This means that any verification
4472 failures that generate their own messages won't get overridden. Similarly,
4473 the "Unrouteable address" message that was in the ACL for unverifiable
4474 relay addresses has also been removed.
4476 56. Added hosts_avoid_esmtp to the smtp transport.
4478 57. The exicyclog script was not checking for the esoteric option
4479 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4480 will work only if exicyclog is run under the appropriate euid.
4482 58. Following a discussion on the list, the rules by which Exim recognises line
4483 endings on incoming messages have been changed. The -dropcr and drop_cr
4484 options are now no-ops, retained only for backwards compatibility. The
4485 following line terminators are recognized: LF CRLF CR. However, special
4486 processing applies to CR:
4488 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4489 nor a local message in the state where . is a terminator.
4491 (ii) If a bare CR is encountered in a header line, an extra space is added
4492 after the line terminator so as not to end the header. The reasoning
4493 behind this is that bare CRs in header lines are most likely either
4494 to be mistakes, or people trying to play silly games.
4496 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4497 was being incorrectly given as 18 bytes larger than it should have been.
4498 This is a VOB (very old bug).
4500 60. This may never have affected anything current, but just in case it has:
4501 When the local host is found other than at the start of a list of hosts,
4502 the local host, those with the same MX, and any that follow, are discarded.
4503 When the list in question was part of a longer list of hosts, the following
4504 hosts (not currently being processed) were also being discarded. This no
4505 longer happens. I'm not sure if this situation could ever has previously
4508 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4511 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4512 header. This is something that is recommended in a new Internet Draft, and
4513 is something that is documented as being done by Sendmail. There are two
4514 possible values. For messages generated by the autoreply transport, Exim
4517 Auto-Submitted: auto-replied
4519 whereas for all other generated messages (e.g. bounces) it adds
4521 Auto-Submitted: auto-generated
4523 63. The "personal" condition in filters now includes a test for the
4524 Auto-Submitted: header. If it contains the string "auto-" the message it
4525 not considered personal.
4527 64. Added rcpt_include_affixes as a generic transport option.
4529 65. Added queue_only_override (default true).
4531 66. Added the syslog_duplication option.
4533 67. If what should have been the first header line of a message consisted of
4534 a space followed by a colon, Exim was mis-interpreting it as a header line.
4535 It isn't of course - it is syntactically invalid and should therefore be
4536 treated as the start of the message body. The misbehaviour could have
4537 caused a number of strange effects, including loss of data in subsequent
4538 header lines, and spool format errors.
4540 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4541 client host had authenticated. This control can now be exercised by an ACL
4542 for more flexibility.
4544 69. By default, callouts do not happen when testing with -bh. There is now a
4545 variant, -bhc, which does actually run the callout code, including
4546 consulting and updating the callout cache.
4548 70. Added support for saslauthd authentication, courtesy of Alexander
4551 71. If statvfs() failed on the spool or log directories while checking their
4552 size for availability, Exim confusingly gave the error "space shortage".
4553 Furthermore, in debugging mode it crashed with a floating point exception.
4554 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4555 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4556 serious problem, Exim now writes to the main and panic logs when this
4557 happens, with details of the failure. It then refuses to accept the
4558 incoming message, giving the message "spool directory problem" or "log
4559 directory problem" with a 421 code for SMTP messages.
4561 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4562 0, 1, and 2 exist, because some OS complain for execs without them (see
4563 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4564 descriptors. However, the code omitted to check that the open succeeded,
4565 causing mysterious errors if for some reason the permissions on /dev/null
4566 got screwed. Now Exim writes a message to the main and panic logs, and
4567 bombs out if it can't open /dev/null.
4569 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4570 interact so that it is all more flexible. It is supposed to remain
4571 backwards compatible. Also added extra_local_interfaces.
4573 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4574 to bomb out with an assertion failure - to the client this appears as a
4575 connection drop. This problem occurs in the part of the code that was taken
4576 from the Samba project. Fortunately, the assertion is in a very simple
4577 function, so I have fixed this by reproducing the function inline in the
4578 one place where it is called, and arranging for authentication to fail
4579 instead of killing the process with assert().
4581 75. The SPA client code was not working when the server requested OEM rather
4582 than Unicode encoding.
4584 76. Added code to make require_files with a specific uid setting more usable in
4585 the case where statting the file as root fails - usually a non-root-mounted
4586 NFS file system. When this happens and the failure is EACCES, Exim now
4587 forks a subprocess and does the per-uid checking as the relevant uid.
4589 77. Added process_log_path.
4591 78. If log_file_path was not explicitly set, a setting of check_log_space or
4592 check_log_inodes was ignored.
4594 79. If a space check for the spool or log partitions fails, the incident is now
4595 logged. Of course, in the latter case the data may get lost...
4597 80. Added the %p formatting code to string_format() so that it can be used to
4598 print addresses in debug_print(). Adjusted all the address printing in the
4599 debugging in store.c to use %p rather than %d.
4601 81. There was a concern that a line of code in smtp_in.c could overflow a
4602 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4603 initially expressed, the concern was not well-founded, because trailing
4604 spaces are removed early. However, if the trailing spaces were followed by
4605 a NULL, they did not get removed, so the overflow was possible. Two fixes
4608 (a) I re-wrote the offending code in a cleaner fashion.
4609 (b) If an incoming SMTP command contains a NULL character, it is rejected
4612 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4613 runs initgroups(), so that if the Exim user is in any additional groups,
4614 they will be used during message reception.
4620 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob