1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.514 2007/06/19 13:32:05 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Another patch from the Sieve maintainer.
11 PH/02 When an IPv6 address is converted to a string for single-key lookup
12 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
13 dots are used instead of colons so that keys in lsearch files need not
14 contain colons. This was done some time before quoting was made available
15 in lsearch files. However, iplsearch files do require colons in IPv6 keys
16 (notated using the quote facility) so as to distinguish them from IPv4
17 keys. This meant that lookups for IP addresses in host lists did not work
18 for iplsearch lookups.
20 This has been fixed by arranging for IPv6 addresses to be expressed with
21 colons if the lookup type is iplsearch. This is not incompatible, because
22 previously such lookups could never work.
24 The situation is now rather anomolous, since one *can* have colons in
25 ordinary lsearch keys. However, making the change in all cases is
26 incompatible and would probably break a number of configurations.
28 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
31 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
32 conversion specification without a maximum field width, thereby enabling
33 a rogue spamd server to cause a buffer overflow. While nobody in their
34 right mind would setup Exim to query an untrusted spamd server, an
35 attacker that gains access to a server running spamd could potentially
36 exploit this vulnerability to run arbitrary code as the Exim user.
38 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
39 $primary_hostname instead of what libspf2 thinks the hosts name is.
41 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
42 a directory entry by the name of the lookup key. Previously, if a
43 symlink pointed to a non-existing file or a file in a directory that
44 Exim lacked permissions to read, a lookup for a key matching that
45 symlink would fail. Now it is enough that a matching directory entry
46 exists, symlink or not. (Bugzilla 503.)
48 PH/03 The body_linecount and body_zerocount variables are now exported in the
51 PH/04 Added the $dnslist_matched variable.
53 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
54 This means they are set thereafter only if the connection becomes
57 PH/06 Added the client_condition to authenticators so that some can be skipped
58 by clients under certain conditions.
60 PH/07 The error message for a badly-placed control=no_multiline_responses left
61 "_responses" off the end of the name.
67 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
68 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
71 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
74 PH/02 In an ACL statement such as
76 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
78 if a client was not listed at all, or was listed with a value other than
79 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
80 the condition was not true (as it should be), so access was not denied.
81 The bug was that the ! inversion was incorrectly passed on to the second
82 item. This has been fixed.
84 PH/03 Added additional dnslists conditions == and =& which are different from
85 = and & when the dns lookup returns more than one IP address.
87 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
88 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
90 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
91 FSYNC, which compiles an option called disable_fsync that allows for
92 bypassing fsync(). The documentation is heavily laced with warnings.
94 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
96 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
97 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
98 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
99 including adding "make clean"; (3) Added -fPIC when compiling the test
100 dynamically loaded module, to get rid of a warning.
102 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
103 message fails, move_frozen_messages = true and ignore_bounce_errors_after
104 = 0s. The bug is otherwise harmless.
106 PH/07 There was a bug in the dovecot authenticator such that the value of
107 $auth1 could be overwritten, and so not correctly preserved, after a
108 successful authentication. This usually meant that the value preserved by
109 the server_setid option was incorrect.
111 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
113 PH/09 Installed PCRE release 7.0.
115 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
116 run for batched SMTP input. It is now run at the start of every message
117 in the batch. While fixing this I discovered that the process information
118 (output by running exiwhat) was not always getting set for -bs and -bS
119 input. This is fixed, and it now also says "batched" for BSMTP.
121 PH/11 Added control=no_pipelining.
123 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
124 patch, slightly modified), and move the expansion of helo_data till after
125 the connection is made in the smtp transport (so it can use these
128 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
130 PH/14 Added log_selector = +pid.
132 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
134 PH/16 Add ${if forany and ${if forall.
136 PH/17 Added dsn_from option to vary the From: line in DSNs.
138 PH/18 Flush SMTP output before performing a callout, unless control =
139 no_callout_flush is set.
141 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
142 was true (the default) a successful delivery failed to delete the retry
143 item, thus causing premature timeout of the address. The bug is now
146 PH/20 Added hosts_avoid_pipelining to the smtp transport.
148 PH/21 Long custom messages for fakedefer and fakereject are now split up
149 into multiline reponses in the same way that messages for "deny" and
150 other ACL rejections are.
152 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
153 with slight modification.
155 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
156 draft, changing the syntax and factoring some duplicate code".
158 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
159 for deliveries of the second and subsequent messages over the same SMTP
162 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
163 ${reduce, with only minor "tidies".
165 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
167 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
168 expansion side effects.
170 PH/27 When a message times out after an over-quota error from an Exim-imposed
171 quota, the bounce message says "mailbox is full". This message was not
172 being given when it was a system quota that was exceeded. It now should
175 MH/03 Made $recipients available in local_scan(). local_scan() already has
176 better access to the recipient list through recipients_list[], but
177 $recipients can be useful in postmaster-provided expansion strings.
179 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
180 in the case of a MAIL command with additional options following the
181 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
182 were accidentally chopped off.
184 PH/29 SMTP synchronization checks are implemented when a command is read -
185 there is a check that no more input is waiting when there shouldn't be
186 any. However, for some commands, a delay in an ACL can mean that it is
187 some time before the response is written. In this time, more input might
188 arrive, invalidly. So now there are extra checks after an ACL has run for
189 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
190 pipelining has not been advertised.
192 PH/30 MH's patch to allow iscntrl() characters to be list separators.
194 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
195 returned in the SMTP response when smtp_return_error_details was false.
198 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
199 instead of the C I/O that was originally supplied, because problems were
202 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
203 Exim which did not show up earlier: it was assuming that a call to
204 SSL_CTX_set_info_callback() might give an error value. In fact, there is
205 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
206 was a macro that became an assignment, so it seemed to work. This has
207 changed to a proper function call with a void return, hence the compile
208 error. Exim's code has been fixed.
210 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
213 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
215 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
217 PH/37 If a message is not accepted after it has had an id assigned (e.g.
218 because it turns out to be too big or there is a timeout) there is no
219 "Completed" line in the log. When some messages of this type were
220 selected by exigrep, they were listed as "not completed". Others were
221 picked up by some special patterns. I have improved the selection
222 criteria to be more general.
224 PH/38 The host_find_failed option in the manualroute router can now be set
225 to "ignore", to completely ignore a host whose IP address cannot be
226 found. If all hosts are ignored, the behaviour is controlled by the new
227 host_all_ignored option.
229 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
230 homing or because of multiple MX records with /mx) generated more than
231 one IP address, and the following item turned out to be the local host,
232 all the secondary addresses of the first item were incorrectly removed
233 from the list, along with the local host and any following hosts (which
234 is what is supposed to happen).
236 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
237 whoever called Exim into the -H file. In the case of the daemon it was
238 behaving confusingly. When first started, it used values for whoever
239 started the daemon, but after a SIGHUP it used the Exim user (because it
240 calls itself on a restart). I have changed the code so that it now always
243 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
244 message are rejected with the same error (e.g. no authentication or bad
245 sender address), and a DATA command is nevertheless sent (as can happen
246 with PIPELINING or a stupid MUA), the error message that was given to the
247 RCPT commands is included in the rejection of the DATA command. This is
248 intended to be helpful for MUAs that show only the final error to their
251 PH/42 Another patch from the Sieve maintainer.
253 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
254 Eximstats - Fixed some broken HTML links and added missing column headers
256 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
257 columns for Rejects, Temp Rejects, Ham, and Spam rows.
259 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
261 PH/43 Yet another patch from the Sieve maintainer.
263 PH/44 I found a way to check for a TCP/IP connection going away before sending
264 the response to the final '.' that terminates a message, but only in the
265 case where the client has not sent further data following the '.'
266 (unfortunately, this is allowed). However, in many cases there won't be
267 any further data because there won't be any more messages to send. A call
268 to select() can be used: if it shows that the input is "ready", there is
269 either input waiting, or the socket has been closed. An attempt to read
270 the next input character can distinguish the two cases. Previously, Exim
271 would have sent an OK response which the client would never have see.
272 This could lead to message repetition. This fix should cure that, at
273 least in a lot of common cases.
275 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
276 advertised in response to EHLO.
282 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
283 fixed by 4.65/MH/01 (is this a record?) are fixed:
285 (i) An empty string was always treated as zero by the numeric comparison
286 operators. This behaviour has been restored.
288 (ii) It is documented that the numeric comparison operators always treat
289 their arguments as decimal numbers. This was broken in that numbers
290 starting with 0 were being interpreted as octal.
292 While fixing these problems I realized that there was another issue that
293 hadn't been noticed. Values of message_size_limit (both the global option
294 and the transport option) were treated as octal if they started with 0.
295 The documentation was vague. These values are now always treated as
296 decimal, and I will make that clear in the documentation.
302 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
303 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
306 MH/01 Don't check that the operands of numeric comparison operators are
307 integers when their expansion is in "skipping" mode (fixes bug
308 introduced by 4.64-PH/07).
310 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
311 child addresses, Exim now panics and dies. Previously, because the count
312 is held in a short int, deliveries were likely to be lost. As such a
313 large number of recipients for a single message is ridiculous
314 (performance will be very, very poor), I have chosen to impose a limit
315 rather than extend the field.
321 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
322 leftover -K file (the existence of which was triggered by #402).
323 While we were at it, introduced process PID as part of the -K
324 filename. This should rule out race conditions when creating
327 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
328 processing considerably. Previous code took too long for large mails,
329 triggering a timeout which in turn triggers #401.
331 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
332 in the DK code in transports.c. sendfile() is not really portable,
333 hence the _LINUX specificness.
335 TF/01 In the add_headers option to the mail command in an Exim filter,
336 there was a bug that Exim would claim a syntax error in any
337 header after the first one which had an odd number of characters
340 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
341 callout verification, Exim cached a "reject" for the entire domain. This
342 is correct for most verifications, but it is not correct for a recipient
343 verification with use_sender or use_postmaster set, because in that case
344 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
345 case of MAIL FROM:<> rejection from other early rejections (e.g.
346 rejection of HELO). When verifying a recipient using a non-null MAIL
347 address, the cache is ignored if it shows MAIL FROM:<> rejection.
348 Whatever the result of the callout, the value of the domain cache is
349 left unchanged (for any other kind of callout, getting as far as trying
350 RCPT means that the domain itself is ok).
352 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
355 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
356 manifest itself as EPIPE rather than ECONNECT. When tidying away a
357 session, the daemon ignores ECONNECT errors and logs others; it now
358 ignores EPIPE as well.
360 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
361 (quoted-printable decoding).
363 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
364 later the small subsequent patch to fix an introduced bug.
366 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
368 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
370 PH/08 An error is now given if message_size_limit is specified negative.
372 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
373 to be given (somewhat) arbitrary names.
375 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
378 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
379 miscellaneous code fixes
381 PH/10 Added the log_reject_target ACL modifier to specify where to log
384 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
385 hostname. This is wrong, because it relates to the incoming message (and
386 probably the interface on which it is arriving) and not to the outgoing
387 callout (which could be using a different interface). This has been
388 changed to use the value of the helo_data option from the smtp transport
389 instead - this is what is used when a message is actually being sent. If
390 there is no remote transport (possible with a router that sets up host
391 addresses), $smtp_active_hostname is used.
393 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
394 tweaks were necessary in order to get it to work (see also 21 below):
395 (a) The code assumed that strncpy() returns a negative number on buffer
396 overflow, which isn't the case. Replaced with Exim's string_format()
398 (b) There were several signed/unsigned issues. I just did the minimum
399 hacking in of casts. There is scope for a larger refactoring.
400 (c) The code used strcasecmp() which is not a standard C function.
401 Replaced with Exim's strcmpic() function.
402 (d) The code set only $1; it now sets $auth1 as well.
403 (e) A simple test gave the error "authentication client didn't specify
404 service in request". It would seem that Dovecot has changed its
405 interface. Fortunately there's a specification; I followed it and
406 changed what the client sends and it appears to be working now.
408 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
411 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
412 address A is aliased to B and C, where B exists and C does not. Without
413 -v the output is "A verified" because verification stops after a
414 successful redirection if more than one address is generated. However,
415 with -v the child addresses are also verified. Exim was outputting "A
416 failed to verify" and then showing the successful verification for C,
417 with its parentage. It now outputs "B failed to verify", showing B's
418 parentage before showing the successful verification of C.
420 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
421 look up a TXT record in a specific list after matching in a combined
424 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
425 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
426 they consult the DNS. I had assumed they would set it the way they
427 wanted; and indeed my experiments on Linux seem to show that in some
428 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
429 To be on the safe side, however, I have now made the interface to
430 host_find_byname() similar to host_find_bydns(), with an argument
431 containing the DNS resolver options. The host_find_byname() function now
432 sets these options at its start, just as host_find_bydns() does. The smtp
433 transport options dns_qualify_single and dns_search_parents are passed to
434 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
435 of host_find_byname() use the default settings of RES_DEFNAMES
436 (qualify_single) but not RES_DNSRCH (search_parents).
438 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
439 spool_read_header() do less string testing, by means of a preliminary
440 switch on the second character of optional "-foo" lines. (This is
441 overdue, caused by the large number of possibilities that now exist.
442 Originally there were few.) While I was there, I also converted the
443 str(n)cmp tests so they don't re-test the leading "-" and the first
444 character, in the hope this might squeeze out yet more improvement.
446 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
447 flag allowing group syntax was set by the header_syntax check but not
448 turned off, possible causing trouble later; (2) The flag was not being
449 set at all for the header_verify test, causing "group"-style headers to
450 be rejected. I have now set it in this case, and also caused header_
451 verify to ignore an empty address taken from a group. While doing this, I
452 came across some other cases where the code for allowing group syntax
453 while scanning a header line wasn't quite right (mostly, not resetting
454 the flag correctly in the right place). These bugs could have caused
455 trouble for malformed header lines. I hope it is now all correct.
457 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
458 with the "reply" argument non-NULL. The code, however (which originally
459 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
460 but it didn't always do it. This confused somebody who was copying the
461 code for some other use. I have removed all the tests.
463 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
464 feature that was used to support insecure browsers during the U.S. crypto
465 embargo. It requires special client support, and Exim is probably the
466 only MTA that supported it -- and would never use it because real RSA is
467 always available. This code has been removed, because it had the bad
468 effect of slowing Exim down by computing (never used) parameters for the
469 RSA_EXPORT functionality.
471 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
472 authenticator to fail if there's a tab character in the incoming data
473 (there should never be unless someone is messing about, as it's supposed
474 to be base64-encoded). Also added, on Timo's advice, the "secured" option
475 if the connection is using TLS or if the remote IP is the same as the
476 local IP, and the "valid-client-cert option" if a client certificate has
479 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
480 authenticators. This can be used for authorization after authentication
481 succeeds. (In the case of plaintext, it servers for both authentication
484 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
485 if any retry times were supplied.
487 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
488 connection, where there is no client IP address to check. In this
489 situation, the verify now always succeeds.
491 PH/25 Applied John Jetmore's -Mset patch.
493 PH/26 Added -bem to be like -Mset, but loading a message from a file.
495 PH/27 In a string expansion for a processed (not raw) header when multiple
496 headers of the same name were present, leading whitespace was being
497 removed from all of them, but trailing whitespace was being removed only
498 from the last one. Now trailing whitespace is removed from each header
499 before concatenation. Completely empty headers in a concatenation (as
502 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
503 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
505 PH/29 [Removed. This was a change that I later backed out, and forgot to
506 correct the ChangeLog entry (that I had efficiently created) before
507 committing the later change.]
509 PH/30 Exim was sometimes attempting to deliver messages that had suffered
510 address errors (4xx response to RCPT) over the same connection as other
511 messages routed to the same hosts. Such deliveries are always "forced",
512 so retry times are not inspected. This resulted in far too many retries
513 for the affected addresses. The effect occurred only when there were more
514 hosts than the hosts_max_try setting in the smtp transport when it had
515 the 4xx errors. Those hosts that it had tried were not added to the list
516 of hosts for which the message was waiting, so if all were tried, there
517 was no problem. Two fixes have been applied:
519 (i) If there are any address or message errors in an SMTP delivery, none
520 of the hosts (tried or untried) are now added to the list of hosts
521 for which the message is waiting, so the message should not be a
522 candidate for sending over the same connection that was used for a
523 successful delivery of some other message. This seems entirely
524 reasonable: after all the message is NOT "waiting for some host".
525 This is so "obvious" that I'm not sure why it wasn't done
526 previously. Hope I haven't missed anything, but it can't do any
527 harm, as the worst effect is to miss an optimization.
529 (ii) If, despite (i), such a delivery is accidentally attempted, the
530 routing retry time is respected, so at least it doesn't keep
531 hammering the server.
533 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
534 in ${readsocket because some servers need this prod.
536 PH/32 Added some extra debug output when updating a wait-xxx database.
538 PH/33 The hint "could be header name not terminated by colon", which has been
539 given for certain expansion errors for a long time, was not being given
540 for the ${if def:h_colon_omitted{... case.
542 PH/34 The spec says: "With one important exception, whenever a domain list is
543 being scanned, $domain contains the subject domain." There was at least
544 one case where this was not true.
546 PH/35 The error "getsockname() failed: connection reset by peer" was being
547 written to the panic log as well as the main log, but it isn't really
548 panic-worthy as it just means the connection died rather early on. I have
549 removed the panic log writing for the ECONNRESET error when getsockname()
552 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
553 runs only) independently of the message's sender address. This meant
554 that, if the 4xx error was in fact related to the sender, a different
555 message to the same recipient with a different sender could confuse
556 things. In particualar, this can happen when sending to a greylisting
557 server, but other circumstances could also provoke similar problems.
558 I have changed the default so that the retry time for these errors is now
559 based a combination of the sender and recipient addresses. This change
560 can be overridden by setting address_retry_include_sender=false in the
563 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
564 remote server are returned as part of bounce messages. This was not
565 happening for LMTP over a pipe (the lmtp transport), but now it is the
566 same for both kinds of LMTP.
568 PH/38 Despite being documented as not happening, Exim was rewriting addresses
569 in header lines that were in fact CNAMEs. This is no longer the case.
571 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
572 and queue runs started by the daemon processed all messages. This has
573 been fixed so that -R and -S can now usefully be given with -q<time>.
575 PH/40 Import PCRE release 6.7 (fixes some bugs).
577 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
579 PH/42 Give an error if -q is specified more than once.
581 PH/43 Renamed the variables $interface_address and $interface_port as
582 $received_ip_address and $received_port, to make it clear that these
583 values apply to message reception, and not to the outgoing interface when
584 a message is delivered. (The old names remain recognized, of course.)
586 PH/44 There was no timeout on the connect() call when using a Unix domain
587 socket in the ${readsocket expansion. There now is.
589 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
590 be meaningful with "accept".
592 SC/01 Eximstats V1.43
593 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
595 SC/02 Eximstats V1.44
596 Use a glob alias rather than an array ref in the generated
597 parser. This improves both readability and performance.
599 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
600 Collect SpamAssassin and rejection statistics.
601 Don't display local sender or destination tables unless
602 there is data to show.
603 Added average volumes into the top table text output.
605 SC/04 Eximstats V1.46
606 Collect data on the number of addresses (recipients)
607 as well as the number of messages.
609 SC/05 Eximstats V1.47
610 Added 'Message too big' to the list of mail rejection
611 reasons (thanks to Marco Gaiarin).
613 SC/06 Eximstats V1.48
614 Mainlog lines which have GMT offsets and are too short to
615 have a flag are now skipped.
617 SC/07 Eximstats V1.49 (Alain Williams)
618 Added the -emptyok flag.
620 SC/08 Eximstats V1.50
621 Fixes for obtaining the IP address from reject messages.
623 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
624 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
625 whitesspace changes from 4.64-PH/27
627 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
630 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
631 are found, allow negative numbers in numeric criteria)
633 JJ/06 exipick.20061117.2, added new $message_body_missing variable
635 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
636 to match changes made in 4.64-PH/43
638 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
640 PH/47 Put in an explicit test for a DNS lookup of an address record where the
641 "domain" is actually an IP address, and force a failure. This locks out
642 those revolvers/nameservers that support "A-for-A" lookups, in
643 contravention of the specifications.
645 PH/48 When a host name was looked up from an IP address, and the subsequent
646 forward lookup of the name timed out, the host name was left in
647 $sender_host_name, contrary to the specification.
649 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
650 restricted to single-key lookups, Exim was not diagnosing an error if
651 * or *@ was used with a query-style lookup.
653 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
655 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
656 long ago, but noone interested enough thought of it. Let's just say that
657 the "1.1" means that there are some new functions that weren't there at
658 some point in the past.
660 PH/51 Error processing for expansion failure of helo_data from an smtp
661 transport during callout processing was broken.
663 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
664 tested/used via the -bh/-bhc/-bs options.
666 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
667 bug, fixed in subsequent PCRE releases).
669 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
670 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
672 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
678 SC/01 Use a glob alias rather than an array ref in eximstats generated
679 parser. This improves both readability and performance.
681 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
682 Don't display local sender or destination tables in eximstats unless
683 there is data to show.
684 Added average volumes into the eximstats top table text output.
686 SC/03 Collect data on the number of addresses (recipients) as well
687 as the number of messages in eximstats.
689 TF/01 Correct an error in the documentation for the redirect router. Exim
690 does (usually) call initgroups() when daemonizing.
692 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
693 with consistent privilege compared to when running as a daemon.
695 TF/03 Note in the spec that $authenticated_id is not set for local
696 submissions from trusted users.
698 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
699 Thanks to Dean Brooks <dean@iglou.com> for the patch.
701 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
702 by adding some example configuration directives to the default
703 configuration file. A little bit of work is required to uncomment the
704 directives and define how usernames and passwords are checked, but
705 there is now a framework to start from.
707 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
708 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
709 without this. I don't know how relevant this is to other LDAP libraries.
711 PH/02 Add the verb name to the "unknown ACL verb" error.
713 PH/03 Magnus Holmgren's patch for filter_prepend_home.
715 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
717 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
718 directory not expanded when it should be if an expanded home directory
719 was set for the address (which is overridden by the transport).
721 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
724 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
725 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
726 because it is too late at that time, and has no effect.
728 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
729 security issue with \' (bugzilla #107). I could not use the
730 PQescapeStringConn() function, because it needs a PGconn value as one of
733 PH/08 When testing addresses using -bt, indicate those final addresses that
734 are duplicates that would not cause an additional delivery. At least one
735 person was confused, thinking that -bt output corresponded to deliveries.
736 (Suppressing duplicates isn't a good idea as you lose the information
737 about possibly different redirections that led to the duplicates.)
739 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
740 systems where poll() doesn't work, in particular OS X.
742 PH/10 Added more information to debugging output for retry time not reached.
744 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
745 operations in malware.c.
747 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
750 PH/13 If write_rejectlog was set false when logging was sent to syslog with
751 syslog_duplication set false, log lines that would normally be written
752 both the the main log and to the reject log were not written to syslog at
755 PH/14 In the default configuration, change the use of "message" in ACL warn
756 statements to "add_header".
758 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
759 not followed by a command (e.g. "seen endif").
761 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
762 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
765 PH/17 Added extra conditions to the default value of delay_warning_condition
769 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
770 { match{$h_precedence:}{(?i)bulk|list|junk} } \
771 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
774 The Auto-Submitted: and various List- headers are standardised, whereas I
775 don't think Precedence: ever was.
777 PH/18 Refactored debugging code in route_finduser() to show more information,
778 in particular, the error code if getpwnam() issues one.
780 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
781 This is apparently needed in addition to the PH/07 change above to avoid
782 any possible encoding problems.
784 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
785 but not after initializing Perl.
787 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
788 output them only if debugging. By default they are written stderr,
789 apparently, which is not desirable.
791 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
794 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
797 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
799 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
800 authenticated or an ident call has been made. Suppress the default
801 values for $authenticated_id and $authenticated_sender (but permit -oMai
802 and -oMas) when testing with -bh.
804 PH/24 Re-jigged the order of the tests in the default configuration so that the
805 tests for valid domains and recipients precede the DNS black list and CSA
806 tests, on the grounds that those ones are more expensive.
808 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
809 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
810 This bug exists in every version of Exim that I still have, right back to
813 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
814 However, an attempt to turn on case-sensitivity in a regex key by
815 including (?-i) didn't work because the subject string was already
816 lowercased, and the effects were non-intuitive. It turns out that a
817 one-line patch can be used to allow (?-i) to work as expected.
823 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
824 other effects) broke the use of negated acl sub-conditions.
826 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
829 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
830 "Deny" causes Exim to reject the incoming connection with a 554 error.
831 Unfortunately, if there is a major crisis, such as a disk failure,
832 tcp-wrappers gives "deny", whereas what one would like would be some
833 kind of temporary error. A kludge has been added to help with this.
834 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
835 554 error is used if errno is still zero or contains ENOENT (which occurs
836 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
839 PH/03 Add -lutil to the default FreeBSD LIBS setting.
841 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
842 errors. Otherwise a message that provokes a temporary error (when other
843 messages do not) can cause a whole host to time out.
845 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
846 addresses were routed directly to files or pipes from a redirect router.
847 File deliveries just didn't batch; pipe deliveries might have suffered
850 PH/06 A failure to get a lock for a hints database would erroneously always say
851 "Failed to get write lock", even when it was really a read lock.
853 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
854 of 0600. This has been changed to use the value of the lockfile_mode
855 option (which defaults to 0600).
857 PH/08 Applied small patch from the Sieve maintainer.
859 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
860 folder from quota calculations, a direct delivery into this folder messed
861 up the contents of the maildirsize file. This was because the regex was
862 used only to exclude .Trash (or whatever) when the size of the mailbox
863 was calculated. There was no check that a delivery was happening into an
864 excluded directory. This bug has been fixed by ignoring all quota
865 processing for deliveries into excluded directories.
867 PH/10 Added the maildirfolder_create_regex option to appendfile.
873 PH/01 The code for finding all the local interface addresses on a FreeBSD
874 system running IPv6 was broken. This may well have applied to all BSD
875 systems, as well as to others that have similar system calls. The broken
876 code found IPv4 interfaces correctly, but gave incorrect values for the
877 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
878 that it would not match correctly against @[] and not recognize the IPv6
881 PH/02 The ipliteral router was not recognizing addresses of the form user@
882 [ipv6:....] because it didn't know about the "ipv6:" prefix.
884 PH/03 Added disable_ipv6.
886 PH/04 Changed $reply_address to use the raw form of the headers instead of the
887 decoded form, because it is most often used to construct To: headers
888 lines in autoreplies, and the decoded form may well be syntactically
889 invalid. However, $reply_address has leading white space removed, and all
890 newlines turned into spaces so that the autoreply transport does not
893 PH/05 If group was specified without a user on a router, and no group or user
894 was specified on a transport, the group from the router was ignored.
896 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
897 for visible compile-time settings that can be used to change these
898 numbers, for those that want even more. Backwards compatibility with old
899 spool files has been maintained. However, going back to a previous Exim
900 release will lost any variables that are in spool files.
902 PH/07 Two small changes when running in the test harness: increase delay when
903 passing a TCP/IP connection to a new process, in case the original
904 process has to generate a bounce, and remove special handling of
905 127.0.0.2 (sic), which is no longer necessary.
907 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
908 be the same on different OS.
910 PH/09 Moved a debug statement in filter processing to avoid a race problem when
913 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
914 whether --show-vars was specified or not
916 JJ/02 exipick: Added support for new ACL variable spool format introduced
919 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
920 syntactically invalid From: or Reply-to: line, and a filter used this to
921 generate an autoreply, and therefore failed to obtain an address for the
922 autoreply, Exim could try to deliver to a non-existent relative file
923 name, causing unrelated and misleading errors. What now happens is that
924 it logs this as a hard delivery error, but does not attempt to create a
927 PH/11 The exinext utility has a -C option for testing purposes, but although
928 the given file was scanned by exinext itself; it wasn't being passed on
929 when Exim was called.
931 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
932 an end-of-file indication when reading a command response.
934 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
935 compiled. In many other places in Exim, IPv6 addresses are always
936 recognized, so I have changed this. It also means that IPv4 domain
937 literals of the form [IPV4:n.n.n.n] are now always recognized.
939 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
940 used if the router is not running as root, for example, when verifying at
941 ACL time, or when using -bh. The debugging output from this situation was
942 non-existent - all you got was a failure to exec. I have made two
945 (a) Failures to set uid/gid, the current directory, or a process leader
946 in a subprocess such as that created by queryprogram now generate
947 suitable debugging ouput when -d is set.
949 (b) The queryprogram router detects when it is not running as root,
950 outputs suitable debugging information if -d is set, and then runs
951 the subprocess without attempting to change uid/gid.
953 PH/15 Minor change to Makefile for building test_host (undocumented testing
956 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
957 additional section of a DNS packet that returns MX or SRV records.
958 Instead, it always explicitly searches for A/AAAA records. This avoids
959 major problems that occur when a DNS server includes only records of one
960 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
961 fixed another bug: if SRV records were looked up and the corresponding
962 address records were *not* found in the additional section, the port
963 values from the SRV records were lost.
965 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
966 using the correct key (the original address) when searching the retry
967 rules in order to find which one to use for generating the retry hint.
969 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
970 adding the default one. Similarly, if it contains a Reply-To: header, the
971 errors_reply_to option, if set, is not used.
973 PH/19 When calculating a retry time, Exim used to measure the "time since
974 failure" by looking at the "first failed" field in the retry record. Now
975 it does not use this if it is later than than the arrival time of the
976 message. Instead it uses the arrival time. This makes for better
977 behaviour in cases where some deliveries succeed, thus re-setting the
978 "first failed" field. An example is a quota failure for a huge message
979 when small messages continue to be delivered. Without this change, the
980 "time since failure" will always be short, possible causing more frequent
981 delivery attempts for the huge message than are intended.
982 [Note: This change was subsequently modified - see PH/04 for 4.62.]
984 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
985 $1, $2, $3) because the numerical variables can be reset during some
986 expansion items (e.g. "match"), thereby losing the authentication data.
988 PH/21 Make -bV show the size of off_t variables so that the test suite can
989 decide whether to run tests for quotas > 2G.
991 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
992 mailbox_size, and mailbox_filecount in the appendfile transport. If a
993 filecount value is greater than 2G or if a quota value is greater than 2G
994 on a system where the size of off_t is not greater than 4, a panic error
997 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
998 never match. The debug and -bh output now contains an explicit error
999 message indicating a malformed IPv4 address or mask.
1001 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1002 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1003 PH/23 above applies.
1005 PH/25 Do not write to syslog when running in the test harness. The only
1006 occasion when this arises is a failure to open the main or panic logs
1007 (for which there is an explicit test).
1009 PH/26 Added the /no_tell option to "control=freeze".
1011 PH/27 If a host name lookup failed very early in a connection, for example, if
1012 the IP address matched host_lookup and the reverse lookup yielded a name
1013 that did not have a forward lookup, an error message of the form "no IP
1014 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1015 could be logged. Now it outputs the IP address instead of "NULL".
1017 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1018 allows the sender and the authenticated sender to be set when
1019 submitting a message from within Exim. Since child_open_exim() is
1020 documented for local_scan(), the new function should be too.
1022 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1023 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1024 results in an empty string is now treated as unset.
1026 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1028 PH/31 Added sender_verify_fail logging option.
1030 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1031 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1032 tidied the source and removed it altogether.
1034 PH/33 When a queue run was abandoned because the load average was too high, a
1035 log line was always written; now it is written only if the queue_run log
1036 selector is set. In addition, the log line for abandonment now contains
1037 information about the queue run such as the pid. This is always present
1038 in "start" and "stop" lines but was omitted from the "abandon" line.
1040 PH/34 Omit spaces between a header name and the colon in the error message that
1041 is given when verify = headers_syntax fails (if there are lots of them,
1042 the message gets confusing).
1044 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1045 names, as there are now some PTR records that contain slashes. This check
1046 is only to protect against broken name servers that fall over on strange
1047 characters, so the fact that it applies to all lookups doesn't matter.
1049 PH/36 Now that the new test suite is complete, we can remove some of the
1050 special code in Exim that was needed for the old test suite. For example,
1051 sorting DNS records because real resolvers return them in an arbitrary
1052 order. The new test suite's fake resolver always returns records in the
1055 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1056 bounces) except when queue_only is set, to avoid logging races between
1057 the different processes.
1059 PH/38 Panic-die if .include specifies a non-absolute path.
1061 PH/39 A tweak to the "H" retry rule from its user.
1063 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1064 a label. They prevented compilation on older perls.
1066 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1067 a warning to be raised on newish perls.
1069 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1070 on queue. Changes to match documented behaviour of showing count of
1071 messages matching specified criteria.
1073 PH/40 Changed the default ident timeout from 30s to 5s.
1075 PH/41 Added support for the use of login_cap features, on those BSD systems
1076 that have them, for controlling the resources used by pipe deliveries.
1078 PH/42 The content-scanning code uses fopen() to create files in which to put
1079 message data. Previously it was not paying any attention to the mode of
1080 the files. Exim runs with umask(0) because the rest of the code creates
1081 files with open(), and sets the required mode explicitly. Thus, these
1082 files were ending up world-writeable. This was not a big issue, because,
1083 being within the spool directory, they were not world-accessible. I have
1084 created a function called modefopen, which takes an additional mode
1085 argument. It sets umask(777), creates the file, chmods it to the required
1086 mode, then resets the umask. All the relevant calls to fopen() in the
1087 content scanning code have been changed to use this function.
1089 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1090 to 24 hours. This avoids potential overflow problems when processing G
1091 and H retry rules. I suspect nobody ever tinkers with this value.
1093 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1095 PH/45 When the plaintext authenticator is running as a client, the server's
1096 challenges are checked to ensure they are valid base64 strings. By
1097 default, the authentication attempt is cancelled if an invalid string is
1098 received. Setting client_ignore_invalid_base64 true ignores these errors.
1099 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1100 they are received. Thus, the responses can be made to depend on the
1101 challenges. If an invalid string is ignored, an empty string is placed in
1104 PH/46 Messages that are created by the autoreply transport now contains a
1105 References: header, in accordance with RFCs 2822 and 3834.
1107 PH/47 Added authenticated_sender_force to the smtp transport.
1109 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1111 PH/49 Installed latest patch from the Sieve maintainer.
1113 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1114 was also set, the appendfile transport was unnecessarily scanning a
1115 directory of message files (e.g. for maildir delivery) to find the count
1116 of files (along with the size), even though it did not need this
1117 information. It now does the scan only if it needs to find either the
1118 size of the count of files.
1120 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1122 PH/52 Two bugs concerned with error handling when the smtp transport is
1125 (i) Exim was not creating retry information for temporary errors given
1126 for individual recipients after the DATA command when the smtp transport
1127 was used in LMTP mode. This meant that they could be retried too
1128 frequently, and not timed out correctly.
1130 (ii) Exim was setting the flag that allows error details to be returned
1131 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1132 recipients that were returned after the DATA command.
1134 PH/53 This is related to PH/52, but is more general: for any failing address,
1135 when detailed error information was permitted to be returned to the
1136 sender, but the error was temporary, then after the final timeout, only
1137 "retry timeout exceeded" was returned. Now it returns the full error as
1138 well as "retry timeout exceeded".
1140 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1141 do this, and (what is worse) MTAs that accept it.
1143 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1144 will now be deprecated.
1146 PH/56 New os.c-cygwin from the Cygwin maintainer.
1148 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1149 formats (previously only available in exim formats via -bpr, -bpru,
1150 and -bpra. Now also available in native and exiqgrep formats)
1152 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1153 with very large, slow to parse queues
1155 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1157 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1159 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1160 responses to authentication challenges, though it was showing the
1161 challenges; (ii) I've removed the CR characters from the debug output for
1164 PH/58 Allow for the insertion of a newline as well as a space when a string
1165 is turned into more than one encoded-word during RFC 2047 encoding. The
1166 Sieve code now uses this.
1168 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1169 data_4xx, lost_connection, tls_required.
1171 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1172 message was being sent as an SMTP response.
1174 PH/61 Add -l and -k options to exicyclog.
1176 PH/62 When verifying, if an address was redirected to one new address, so that
1177 verification continued, and the new address failed or deferred after
1178 having set something in $address_data, the value of $address_data was not
1179 passed back to the ACL. This was different to the case when no
1180 redirection occurred. The value is now passed back in both cases.
1182 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1183 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1184 use login_cap.h, so on its own it isn't the distinguishing feature. The
1185 new name refers directly to the setclassresources() function.
1187 PH/65 Added configuration files for NetBSD3.
1189 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1191 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1192 is preferred over IPv4.
1194 PH/68 The bounce_return_message and bounce_return_body options were not being
1195 honoured for bounces generated during the reception of non-SMTP messages.
1196 In particular, this applied to messages rejected by the ACL. This bug has
1197 been fixed. However, if bounce_return_message is true and bounce_return_
1198 body is false, the headers that are returned for a non-SMTP message
1199 include only those that have been read before the error was detected.
1200 (In the case of an ACL rejection, they have all been read.)
1202 PH/69 The HTML version of the specification is now built in a directory called
1203 spec_html instead of spec.html, because the latter looks like a path with
1204 a MIME-type, and this confuses some software.
1206 PH/70 Catch two compiler warnings in sieve.c.
1208 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1209 function verify_get_ident() calls ip_connect() to connect a socket, but
1210 if the "connect()" function timed out, ip_connect() used to close the
1211 socket. However, verify_get_ident() also closes the socket later, and in
1212 between Exim writes to the log, which may get opened at this point. When
1213 the socket was closed in ip_connect(), the log could get the same file
1214 descriptor number as the socket. This naturally causes chaos. The fix is
1215 not to close the socket in ip_connect(); the socket should be closed by
1216 the function that creates it. There was only one place in the code where
1217 this was missing, in the iplookup router, which I don't think anybody now
1218 uses, but I've fixed it anyway.
1220 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1221 well as to direct DNS lookups. Otherwise the handling of names in host
1222 lists is inconsistent and therefore confusing.
1228 PH/01 Two changes to the default runtime configuration:
1230 (1) Move the checks for relay_from_hosts and authenticated clients from
1231 after to before the (commented out) DNS black list checks.
1233 (2) Add control=submission to the relay_from_hosts and authenticated
1234 clients checks, on the grounds that messages accepted by these
1235 statements are most likely to be submissions.
1237 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1239 (1) Generate an error if the third argument for the ${prvs expansion is
1242 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1245 (3) Reset the variables that are obtained from the first argument of
1246 ${prvscheck and used in the second argument before leaving the code,
1247 because their memory is reclaimed, so using them afterwards may do
1250 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1251 one (it's much easier than Tom thought :-).
1253 (5) Because of (4), we can now allow for the use of $prvscheck_result
1254 inside the third argument.
1256 PH/03 For some reason, the default setting of PATH when running a command from
1257 a pipe transport was just "/usr/bin". I have changed it to
1260 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1261 anything to be listed in the output from -bV.
1263 PH/05 When a filter generated an autoreply, the entire To: header line was
1264 quoted in the delivery log line, like this:
1266 => >A.N.Other <ano@some.domain> <original@ddress> ...
1268 This has been changed so that it extracts the operative address. There
1269 may be more than one such address. If so, they are comma-separated, like
1272 => >ano@some.domain,ona@other.domain <original@ddress> ...
1274 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1275 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1276 not being looked up in the rDNS to get a host name, Exim was showing the
1277 IP address twice in Received: lines, even though the IP addresses were
1278 identical. For example:
1280 Received: from [1.2.3.4] (helo=[1.2.3.4])
1282 However, if the real host name was known, it was omitting the HELO data
1283 if it matched the actual IP address. This has been tidied up so that it
1284 doesn't show the same IP address twice.
1286 PH/07 When both +timestamp and +memory debugging was on, the value given by
1287 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1288 called by the debug printing, thereby overwriting the timestamp buffer.
1289 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1291 PH/08 When the original message was included in an autoreply transport, it
1292 always said "this is a copy of the message, including all the headers",
1293 even if body_only or headers_only was set. It now gives an appropriate
1296 PH/09 Applied a patch from the Sieve maintainer which:
1298 o fixes some comments
1299 o adds the (disabled) notify extension core
1300 o adds some debug output for the result of if/elsif tests
1301 o points to the current vacation draft in the documentation
1302 and documents the missing references header update
1306 o fixes a bug in processing the envelope test (when testing
1307 multiple envelope elements, the last element determinted the
1310 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1311 Electronic Mail") by including:
1313 Auto-submitted: auto-generated
1315 in the messages that it generates (bounce messages and others, such as
1316 warnings). In the case of bounce messages for non-SMTP mesages, there was
1317 also a typo: it was using "Auto_submitted" (underscore instead of
1318 hyphen). Since every message generated by Exim is necessarily in response
1319 to another message, thes have all been changed to:
1321 Auto-Submitted: auto-replied
1323 in accordance with these statements in the RFC:
1325 The auto-replied keyword:
1327 - SHOULD be used on messages sent in direct response to another
1328 message by an automatic process,
1330 - MUST NOT be used on manually-generated messages,
1332 - MAY be used on Delivery Status Notifications (DSNs) and Message
1333 Disposition Notifications (MDNs),
1335 - MUST NOT be used on messages generated by automatic or periodic
1336 processes, except for messages which are automatic responses to
1339 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1340 to the default Received: header definition.
1342 PH/12 Added log selector acl_warn_skipped (default on).
1344 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1345 variables because (a) they are in the wrong storage pool and (b) even if
1346 they were copied, it wouldn't work properly because of the caching.
1348 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1349 checking when decoding. Apparently there are clients that generate
1350 overlong encoded strings. Why am I not surprised?
1352 PH/15 If the first argument of "${if match_address" was not empty, but did not
1353 contain an "@" character, Exim crashed. Now it writes a panic log message
1354 and treats the condition as false.
1356 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1358 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1359 extension "envelope-auth". The code is finished and in agreement with
1360 other implementations, but there is no documentation so far and in fact,
1361 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1362 not changing the active code.
1364 Print executed "if" and "elsif" statements when debugging is used. This
1365 helps a great deal to understand what a filter does.
1367 Document more things not specified clearly in RFC3028. I had all this
1368 sorted out, when out of a sudden new issues came to my mind. Oops."
1370 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1373 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1374 canonical form (as documented). However, after a host name lookup from
1375 the IP address, check_host() was doing a simple string comparison with
1376 addresses acquired from the DNS when checking that the found name did
1377 have the original IP as one of its addresses. Since any found IPv6
1378 addresses are likely to be in abbreviated form, the comparison could
1379 fail. Luckily, there already exists a function for doing the comparison
1380 by converting both addresses to binary, so now that is used instead of
1381 the text comparison.
1383 PH/20 There was another similar case to PH/19, when a complete host name was
1384 given in a host list; looking up its IP address could give an abbreviated
1385 form, whereas the current host's name might or might not be abbreviated.
1386 The same fix has been applied.
1392 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1393 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1396 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1397 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1399 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1401 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1402 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1403 Also, for "Auto-Submitted", treat anything other than "no" as
1404 non-personal, in accordance with RFC 3834. (Previously it treated
1405 anything starting "auto-" as non-personal.)
1407 TF/01 The control=submission/name=... option had a problem with syntax
1408 errors if the name included a slash character. The /name= option
1409 now slurps the rest of the string, so it can include any characters
1410 but it must come last in the list of options (after /sender_retain
1413 PH/05 Some modifications to the interface to the fake nameserver for the new
1421 TK/01 Added the "success_on_redirect" address verification option. See
1422 NewStuff for rationale and an example.
1424 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1426 PH/02 Patch to exigrep to allow it to work on syslog lines.
1428 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1429 fread() to skip over the body file's header line, because in Cygwin the
1430 header line is locked and is inaccessible.
1432 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1433 co-exist for some time) to make it clear that it is the Exim ID that is
1434 referenced, not the Message-ID: header line.
1436 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1437 string_format() function, because snprintf() does not exist on all
1440 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1441 ${stat: expansion item.
1443 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1444 protocol synchronization error", to keep the pedants happy.
1446 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1447 well as for IRIX systems, when gcc is being used. See the host.c source
1450 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1452 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1455 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1456 transport and to the smtp transport in LMTP mode.
1458 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1460 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1461 run a filter in a subprocess. This could lead to confusion in subsequent
1462 lookups in the parent process. There should also be a search_tidyup() at
1463 the end of the subprocess.
1465 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1466 only if the host matched helo_try_verify_hosts, which caused the
1467 verification to occur when the EHLO/HELO command was issued. The ACL just
1468 tested the remembered result. Now, if a previous verification attempt has
1469 not happened, "verify = helo" does it there and then.
1471 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1473 TK/03 Fix log output including CR from clamd.
1475 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1476 exist provoked a memory error which could cause a segfault.
1478 PH/15 Installed PCRE 6.2
1480 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1482 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1483 of the problem. Specifically, suggested +O2 rather than +O1 for the
1486 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1488 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1489 record, the port was not correctly logged when the outgoing_port log
1490 selector was set (it logged the transort's default port).
1492 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1493 fallback_hosts, and "hosts" in the smtp transport.
1495 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1496 host errors such as "Connection refused".
1498 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1499 authentication with radiusclient 0.4.9:
1501 - Error returned from rc_read_config was caught wrongly
1502 - Username/password not passed on to radius server due to wrong length.
1504 The presumption is that some radiusclient API changes for 4.51/PH/17
1505 were not taken care of correctly. The code is still untested by me (my
1506 Linux distribution still has 0.3.2 of radiusclient), but it was
1507 contributed by a Radius user.
1509 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1510 expanding the "port" option of the smtp transport.
1512 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1513 while reading a MIME header. Thanks to Tom Hughes for a patch.
1515 PH/24 Include config.h inside local_scan.h so that configuration settings are
1518 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1519 that in an ACL for RCPT (for example), you can examine exactly what was
1522 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1523 commands, but it was not correctly comparing the address with the actual
1524 client host address. Thus, it would show the EHLO address in Received:
1525 header lines when this was not necessary.
1527 PH/27 Added the % operator to ${eval:}.
1529 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1530 it should be ignoring failures (because with -C, for example, it has lost
1531 privilege). It wasn't ignoring creation failures other than "already
1534 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1537 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1538 because an input line was too long, either on its own, or by virtue of
1539 too many continuations, the temporary file was not being removed, and the
1540 return code was incorrect.
1542 PH/31 Missing "BOOL" in function definition in filtertest.c.
1544 PH/32 Applied Sieve patches from the maintainer.
1546 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1548 PH/33 Added "verify = not_blind".
1550 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1551 Local/Makefile (with some defaults set). These are used in built scripts
1552 such as exicyclog, but they have never been used in the exim_install
1553 script (though there are many overriding facilities there). I have
1554 arranged that the exim_install script now takes note of these two
1557 PH/35 Installed configuration files for Dragonfly.
1559 PH/36 When a locally submitted message by a trusted user did not contain a
1560 From: header, and the sender address was obtained from -f or from an SMTP
1561 MAIL command, and the trusted user did not use -F to supply a sender
1562 name, $originator_name was incorrectly used when constructing a From:
1563 header. Furthermore, $originator_name was used for submission mode
1564 messages from external hosts without From: headers in a similar way,
1565 which is clearly wrong.
1567 PH/37 Added control=suppress_local_fixups.
1569 PH/38 When log_selector = +received_sender was set, and the addition of the
1570 sender made the log line's construction buffer exactly full, or one byte
1571 less than full, an overflow happened when the terminating "\n" was
1574 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1575 when the result of a list match is failure because a DNS lookup failed.
1577 PH/40 RM_COMMAND is now used in the building process.
1579 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1580 the "build-* directories that it finds.
1582 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1583 address in a domain literal was a prefix of an interface address.
1585 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1586 when verifying a sender address, unless rewrite_headers is false.
1588 PH/44 Wrote a long comment about why errors_to addresses are verified as
1589 recipients, not senders.
1591 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1592 the ratelimit ACL was added.
1594 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1596 PH/46 Added extra information about PostgreSQL errors to the error string.
1598 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1599 avoiding the need to install special zones in a real server. This is
1600 backwards compatible; if it can't find the fake resolver, it drops back.
1601 Thus, both old and new test suites can be run.
1603 TF/02 Added util/ratelimit.pl
1605 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1606 clock is set back in time.
1608 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1609 Candler <B.Candler@pobox.com>.
1611 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1612 OK for addresses that are the result of redirections.
1614 PH/48 A number of further additions for the benefit of the new test suite,
1615 including a fake gethostbyname() that interfaces to the fake DNS resolver
1618 TF/06 The fix for widen_domains has also been applied to qualify_single and
1619 search_parents which are the other dnslookup options that can cause
1622 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1625 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1627 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1628 service" if it encountered an MX record with an empty target hostname.
1629 The message is now "an MX or SRV record indicated no SMTP service".
1631 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1632 if the DNS of the sending site is misconfigured. This is quite a
1633 common situation. This change restores the behaviour of treating a
1634 helo verification defer as a failure.
1636 PH/51 If self=fail was set on a router, the bounce message did not include the
1637 actual error message.
1643 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1645 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1646 command itself ended in error, the underlying message about the transport
1647 filter timeout was being overwritten with the pipe command error. Now the
1648 underlying error message should be appended to the second error message.
1650 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1652 PH/02 Reduce the amount of output that "make" produces by default. Full output
1653 can still be requested.
1655 PH/03 The warning log line about a condition test deferring for a "warn" verb
1656 was being output only once per connection, rather than after each
1657 occurrence (because it was using the same function as for successful
1658 "warn" verbs). This seems wrong, so I have changed it.
1660 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1661 it should not have, which might have caused a crash in the right
1662 circumstances, but probably never did.
1664 PH/04 Installed a modified version of Tony Finch's patch to make submission
1665 mode fix the return path as well as the Sender: header line, and to
1666 add a /name= option so that you can make the user's friendly name appear
1669 TF/03 Added the control = fakedefer ACL modifier.
1671 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1672 Mark Lowes for thorough testing.
1674 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1676 TK/03 Merged latest SRS patch from Miles Wilton.
1678 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1679 with the definition in sysexits.h (which is #included earlier).
1680 Fortunately, Exim does not actually use EX_OK. The code used to try to
1681 preserve the sysexits.h value, by assumimg that macro definitions were
1682 scanned for macro replacements. I have been disabused of this notion,
1683 so now the code just undefines EX_OK before #including unistd.h.
1685 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1686 in the smtp transport. When a block could not be written in a single
1687 write() function, the timeout was being re-applied to each part-write.
1688 This seems wrong - if the receiver was accepting one byte at a time it
1689 would take for ever. The timeout is now adjusted when this happens. It
1690 doesn't have to be particularly precise.
1692 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1693 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1695 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1696 without a domain if the check to <postmaster@domain> fails.
1698 SC/01 Eximstats: added -xls and the ability to specify output files
1699 (patch written by Frank Heydlauf).
1701 SC/02 Eximstats: use FileHandles for outputing results.
1703 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1705 SC/04 Eximstats: fixed display of large numbers with -nvr option
1707 SC/05 Eximstats: fixed merging of reports with empty tables.
1709 SC/06 Eximstats: added the -include_original_destination flag
1711 SC/07 Eximstats: removed tabs and trailing whitespace.
1713 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1715 TK/06 MBOX spool code: Add real "From " MBOX separator line
1716 so the .eml file is really in mbox format (even though
1717 most programs do not really care). Patch from Alex Miller.
1719 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1720 The latter is generated from $received_to and is only set if the
1721 message has one envelope recipient. SA can use these headers,
1722 obviously out-of-the-box. Patch from Alex Miller.
1724 PH/08 The ${def test on a variable was returning false if the variable's
1725 value was "0", contrary to what the specification has always said!
1726 The result should be true unless the variable is empty.
1728 PH/09 The syntax error of a character other than { following "${if
1729 def:variable_name" (after optional whitespace) was not being diagnosed.
1730 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1731 accidental colon was present, for example, could give incorrect results.
1733 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1734 result is used (not including appendfile, where other changes are about
1737 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1738 This involved changing a lot of size variables from int to off_t. It
1739 should work with maildirs and everything.
1741 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1742 spamd dying while we are connected to it.
1744 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1747 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1748 function for BDB 4.3.
1750 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1752 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1753 the sources for PCRE have been re-arranged and more files are now
1756 PH/15 The code I had for printing potentially long long variables in PH/11
1757 above was not the best (it lost precision). The length of off_t variables
1758 is now inspected at build time, and an appropriate printing format (%ld
1759 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1760 to be "long long int" or "long int". This is needed for the internal
1761 formatting function string_vformat().
1763 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1764 the configuration file to be ":syslog", then the script "guesses" where
1765 the logs files are, rather than using the compiled in default. In our
1766 case the guess is not the same as the compiled default, so the script
1767 suddenly stopped working when I started to use syslog. The patch checks
1768 to see if log_file_path is "". If so, it attempts to read it from exim
1769 with no configuration file to get the compiled in version, before it
1770 falls back to the previous guessing code."
1772 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1773 implementing BATV in an Exim configuration. See NewStuff for the gory
1776 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1777 Makefile that are specific to HP-UX.
1779 PH/18 If the "use_postmaster" option was set for a recipient callout together
1780 with the "random" option, the postmaster address was used as the MAIL
1781 FROM address for the random test, but not for the subsequent recipient
1782 test. It is now used for both.
1784 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1785 patch removes a few documentation additions to RFC 3028, because the
1786 latest draft now contains them. It adds the new en;ascii-case comparator
1787 and a new error check for 8bit text in MIME parts. Comparator and
1788 require names are now matched exactly. I enabled the subaddress
1789 extension, but it is not well tested yet (read: it works for me)."
1791 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1792 rework some of the code of TK/09 above to avoid the hardwired use of
1793 "%lld" and "long long". Replaced the call to snprintf() with a call to
1796 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1797 records point to non-existent hosts", "retry timeout exceeded", and
1798 "retry time not reached for any host after a long failure period".
1800 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1801 experimental DomainKeys support:
1803 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1804 (2) On an error such as an illegally used "control", the wrong name for
1805 the control was given.
1807 These problems did NOT occur unless DomainKeys support was compiled.
1809 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1811 PH/24 Added ${if match_ip condition.
1813 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1814 databases so that it will be absolutely obvious if a crash occurs in the
1815 DB library. This is a regular occurrence (often caused by mis-matched
1818 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1819 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1820 picked up on a user's system that detects such things. There doesn't seem
1821 to be a gcc warning option for this - only an attribute that has to be
1822 put on the function's prototype. It seems that in Fedora Core 4 they have
1823 set this on a number of new functions. No doubt there will be more in due
1826 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1827 specify a transport. However, if an address that was verified by such a
1828 router was the subject of a callout, Exim crashed because it tried to
1829 read the rcpt_include_affixes from the non-existent transport. Now it
1830 just assumes that the setting of that option is false. This bug was
1831 introduced by 4.51/PH/31.
1833 PH/28 Changed -d+all to exclude +memory, because that information is very
1834 rarely of interest, but it makes the output a lot bigger. People tend to
1835 do -d+all out of habit.
1837 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1838 code in os-type was giving problems when libc.so lives in lib64, like on
1841 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1842 aren't the modern standard, and it seems that some systems' include files
1843 don't always have them. Exim was already checking for some of the newer
1844 ones like T_AAAA, and defining it itself. I've added checks for all the
1845 record types that Exim uses.
1847 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1848 not automatically generating a new one, as it is supposed to. This
1849 prevented TLS from working. If the file did exist, but contained invalid
1850 data, a new version was generated, as expected. It was only the case of a
1851 non-existent file that was broken.
1853 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1854 with a change in libdomainkeys > 0.64.
1856 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1857 from DNS. If the selector record carries the flag, it now has
1858 precedence over the domain-wide flag.
1860 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1862 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1863 the use of a port name that isn't defined in /etc/services) occurred, the
1864 message was deferred as in a normal delivery, and thus remained on the
1865 spool, instead of being failed because of the mua_wrapper setting. This
1866 is now fixed, and I tidied up some of the mua_wrapper messages at the
1869 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1870 the messages in a hash of arrays rather than using individual hashes.
1871 This is a bit cleaner and results in dramatic memory savings, albeit
1872 at a slight CPU cost.
1874 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1875 as requested by Marc Sherman.
1877 SC/10 Eximstats: added histograms for user specified patterns as requested
1880 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1882 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1883 fopen() in the content-scanning modules that did not already have it.
1889 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1890 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1892 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1894 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1896 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1897 to test Sieve filters that use "vacation".
1899 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1900 that changes the way the GnuTLS parameters are stored in the cache file.
1901 The new format can be generated externally. For backward compatibility,
1902 if the data in the cache doesn't make sense, Exim assumes it has read an
1903 old-format file, and it generates new data and writes a new file. This
1904 means that you can't go back to an older release without removing the
1907 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1908 work if there are any delivery delays because "one_time" forces the
1909 parent to be marked "delivered", so its unseen clone is never tried
1910 again. For this reason, Exim now forbids the simultaneous setting of
1913 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1914 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1915 entry if you want to know the details. The fix, however, neglected to
1916 consider the case when local delivery batching is involved. The test for
1917 "previously delivered" was not happening when checking to see if an
1918 address could be batched with a previous (undelivered) one; under
1919 certain circumstances this could lead to multiple deliveries to the same
1922 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1923 in its include files, and this causes problems building Exim.
1925 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1926 header_syntax) but Exim was just ignoring anything given after a slash.
1927 In particular, this caused confusion with an attempt to use "verify =
1928 reverse_host_lookup/defer_ok". An error is now given when options are
1929 supplied for verify items that do not have them. (Maybe reverse_host_
1930 lookup should have a defer_ok option, but that's a different point.)
1932 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1933 defined by RFC 821) to 2048, because there were problems with some AUTH
1934 commands, and RFC 1869 says the size should be increased for extended
1935 SMTP commands that take arguments.
1937 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1940 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1941 "unknown" error; now it says that the functionality isn't in the binary.
1943 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1944 an address' error message when a string expansion fails (syntax or
1945 whatever). Otherwise the password may appear in the log. Following change
1946 PH/42 below, there is no longer a chance of it appearing in a bounce
1949 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1951 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1952 of its addresses was ever tried. (Bugzilla bug #2.)
1954 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1955 the result incorrectly in the debug output. (It correctly added a newline
1956 to what was transported.)
1958 TF/01 Added $received_time.
1960 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1961 commented out examples of how to interface to a virus scanner and to
1962 SpamAssassin. Also added commented examples of av_scanner and
1963 spamd_address settings.
1965 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1966 and controls are allowed in which ACLs. There were a couple of minor
1967 errors. Some of the entries in the conditions table (which is a table of
1968 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1969 negation of where the condition IS allowed.
1971 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1973 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1974 header file does not have a version number, so I've had to invent a new
1975 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1976 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1977 radiusclient), but it was contributed by a Radius user.
1979 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1980 files or directories, for OpenSSL.
1982 PH/19 When an Exim process that is running as root has to create an Exim log
1983 file, it does so in a subprocess that runs as exim:exim so as to get the
1984 ownership right at creation (otherwise, other Exim processes might see
1985 the file with the wrong ownership). There was no test for failure of this
1986 fork() call, which would lead to the process getting stuck as it waited
1987 for a non-existent subprocess. Forks do occasionally fail when resources
1988 run out. I reviewed all the other calls to fork(); they all seem to check
1991 PH/20 When checking for unexpected SMTP input at connect time (before writing
1992 the banner), Exim was not dealing correctly with a non-positive return
1993 from the read() function. If the client had disconnected by this time,
1994 the result was a log entry for a synchronization error with an empty
1995 string after "input=" when read() returned zero. If read() returned -1
1996 (an event I could not check), uninitialized data bytes were printed.
1997 There were reports of junk text (parts of files, etc) appearing after
2000 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2002 PH/22 Added support for macro redefinition, and (re)definition in between
2003 driver and ACL definitions.
2005 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2006 forgetting to use the resulting value; it was using the unexpanded value.
2008 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2009 hadn't been configured. The fix is from Juergen Kreileder, who
2010 understands it better than I do:
2012 "Here's what I see happening with three configured cyrus_sasl
2013 authenticators configured (plain, login, cram-md5):
2015 On startup auth_cyrus_sasl_init() gets called for each of these.
2016 This means three calls to sasl_listmech() without a specified mech_list.
2017 => SASL tests which mechs of all available mechs actually work
2018 => three warnings about OTP not working
2019 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2021 With the patch, sasl_listmech() also gets called three times. But now
2022 SASL's mech_list option is set to the server_mech specified in the the
2023 authenticator. Or in other words, the answer from sasl_listmech()
2024 gets limited to just the mech you're testing for (which is different
2026 => the return list contains just 'plain' or 'login', 'cram-md5' or
2027 nothing depending on the value of ob->server_mech.
2029 I've just tested the patch: Authentication still works fine,
2030 unavailable mechs specified in the exim configuration are still
2031 caught, and the auth.log warnings about OTP are gone."
2033 PH/25 When debugging is enabled, the contents of the command line are added
2034 to the debugging output, even when log_selector=+arguments is not
2037 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2038 answer is "GNU", and only if the return is "GNU/something" is the answer
2041 PH/27 $acl_verify_message is now set immediately after the failure of a
2042 verification in an ACL, and so is available in subsequent modifiers. In
2043 particular, the message can be preserved by coding like this:
2045 warn !verify = sender
2046 set acl_m0 = $acl_verify_message
2048 Previously, $acl_verify_message was set only while expanding "message"
2049 and "log_message" when a very denied access.
2051 PH/28 Modified OS/os.c-Linux with
2053 -#ifndef OS_LOAD_AVERAGE
2054 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2056 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2057 nomenclature these days.)
2059 PH/29 Installed patch from the Sieve maintainer that adds the options
2060 sieve_useraddress and sieve_subaddress to the redirect router.
2062 PH/30 In these circumstances:
2063 . Two addresses routed to the same list of hosts;
2064 . First host does not offer TLS;
2065 . First host accepts first address;
2066 . First host gives temporary error to second address;
2067 . Second host offers TLS and a TLS session is established;
2068 . Second host accepts second address.
2069 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2070 and peerdn, if requested) that were in fact used only for the second
2073 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2074 attention to any local part prefix or suffix that was matched by the
2075 router that accepted the address. It now behaves in the same way as it
2076 does for delivery: the affixes are removed from the local part unless
2077 rcpt_include_affixes is set on the transport.
2079 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2080 timeout during the DATA phase of an incoming message.
2082 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2083 applied a patch sanctioned by the Sieve maintainer.
2085 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2086 the uid or gid is negative. A case of a negative gid caused this to be
2087 noticed. The fix allows for either to be negative.
2089 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2090 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2093 PH/36 The change PH/12 above was broken. Fixed it.
2095 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2096 the grounds that routing the same address twice would always produce the
2097 same answer. This might have been true once, but it is certainly no
2098 longer true now. Routing a child address may depend on the previous
2099 routing that produced that child. Some complicated redirection strategies
2100 went wrong when messages had multiple recipients, and made Exim's
2101 behaviour dependent on the order in which the addresses were given.
2103 I have moved the duplicate checking until after the routing is complete.
2104 Exim scans the addresses that are assigned to local and remote
2105 transports, and removes any duplicates. This means that more work will be
2106 done, as duplicates will always all be routed, but duplicates are
2107 presumably rare, so I don't expect this is of any significance.
2109 For deliveries to pipes, files, and autoreplies, the duplicate checking
2110 still happens during the routing process, since they are not going to be
2113 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2114 It corrects a timeout issue with spamd. This is Ian's comment: "The
2115 background is that sometimes spamd either never reads data from a
2116 connection it has accepted, or it never writes response data. The exiscan
2117 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2118 blindly assumes that writes won't block so it may never time out."
2120 PH/39 Allow G after quota size as well as K and M.
2122 PH/40 The value set for $authenticated_id in an authenticator may not contain
2123 binary zeroes or newlines because the value is written to log lines and
2124 to spool files. There was no check on this. Now the value is run through
2125 the string_printing() function so that such characters are converted to
2126 printable escape sequences.
2128 PH/41 $message_linecount is a new variable that contains the total number of
2129 lines in the message. Compare $body_linecount, which is the count for the
2132 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2133 bounce and delay warning messages, except in certain special cases, which
2136 (a) An SMTP error message from a remote host;
2137 (b) A message specified in a :fail: redirection;
2138 (c) A message specified in a "fail" command in a system filter;
2139 (d) A message specified in a FAIL return from the queryprogram router;
2140 (e) A message specified by the cannot_route_message router option.
2142 In these cases only, Exim does include the error details in bounce and
2143 warning messages. There are also a few cases where bland messages such
2144 as "unrouteable address" or "local delivery error" are given.
2146 PH/43 $value is now also set for the "else" part of a ${run expansion.
2148 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2149 being worked on, but at least Exim now implements the latest version to
2152 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2153 process to complete was treated as a delivery failure, a timeout while
2154 writing the message to the pipe was logged, but erroneously treated as a
2155 successful delivery. Such timeouts include transport filter timeouts. For
2156 consistency with the overall process timeout, these timeouts are now
2157 treated as errors, giving rise to delivery failures by default. However,
2158 there is now a new Boolean option for the pipe transport called
2159 timeout_defer, which, if set TRUE, converts the failures into defers for
2160 both kinds of timeout. A transport filter timeout is now identified in
2163 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2164 systems where "make" and "gmake" are different, calling "gmake" at top
2165 level broke things. I've arranged for the value of $(MAKE) to be passed
2166 from the Makefile to this script so that it can call the same version of
2170 A note about Exim versions 4.44 and 4.50
2171 ----------------------------------------
2173 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2174 changes of various kinds. As a consequence, a big documentation update was
2175 needed. This delayed the release for rather longer than seemed good, especially
2176 in the light of a couple of (minor) security issues. Therefore, the changes
2177 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2178 release. So 4.44 and 4.50 are in effect two different branches that both start
2181 I have left the 4.50 change log unchanged; it contains all the changes since
2182 4.43. The change log for 4.44 is below; many of its items are identical to
2183 those for 4.50. This seems to be the most sensible way to preserve the
2184 historical information.
2190 1. Minor wording change to the doc/README.SIEVE file.
2192 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2193 computation of the current number of files was incorrect.
2195 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2196 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2199 4. Give more explanation in the error message when the command for a transport
2200 filter fails to execute.
2202 5. There are several places where Exim runs a non-Exim command in a
2203 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2204 was being done only for the command run by the queryprogram router. It is
2205 now done for all such subprocesses. The other cases are: ${run, transport
2206 filters, and the commands run by the lmtp and pipe transports.
2208 6. Added CONFIGURE_GROUP build-time option.
2210 7. Some older OS have a limit of 256 on the maximum number of file
2211 descriptors. Exim was using setrlimit() to set 1000 as a large value
2212 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2213 systems. I've change it so that if it can't get 1000, it tries for 256.
2215 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2216 was an oversight, and furthermore, ever since the addition of extra
2217 controls (e.g. 4.43/32), the checks on when to allow different forms of
2218 "control" were broken. There should now be diagnostics for all cases when a
2219 control that does not make sense is encountered.
2221 9. Added the /retain_sender option to "control=submission".
2223 10. $recipients is now available in the predata ACL (oversight).
2225 11. Tidy the search cache before the fork to do a delivery from a message
2226 received from the command line. Otherwise the child will trigger a lookup
2227 failure and thereby defer the delivery if it tries to use (for example) a
2228 cached ldap connection that the parent has called unbind on.
2230 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2231 of $address_data from the recipient verification was clobbered by the
2232 sender verification.
2234 13. The value of address_data from a sender verification is now available in
2235 $sender_address_data in subsequent conditions in the ACL statement.
2237 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2239 15. Added a new option "connect=<time>" to callout options, to set a different
2242 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2243 was its contents. (It was OK if the option was not defined at all.)
2245 17. A "Completed" log line is now written for messages that are removed from
2246 the spool by the -Mrm option.
2248 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2249 information about exactly what failed.
2251 19. Added -dd to debug only the daemon process.
2253 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2254 handles timeouts, both on the server side and network timeouts. Renamed the
2255 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2257 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2258 It is now set to "smtps".
2260 22. $host_address is now set to the target address during the checking of
2261 ignore_target_hosts.
2263 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2264 being passed; this would have caused $sender_host_name to have been used if
2265 matching the list had actually called for a host name (not very likely,
2266 since this list is usually IP addresses). A host name is now passed as
2269 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2270 code that specifies a non-restarting handler (typically sigaction() in
2271 modern systems) in an attempt to fix a rare and obscure crash bug.
2273 25. Narrowed the window for a race in the daemon that could cause it to ignore
2274 SIGCHLD signals. This is not a major problem, because they are used only to
2275 wake it up if nothing else does.
2277 26. A malformed maildirsize file could cause Exim to calculate negative values
2278 for the mailbox size or file count. Odd effects could occur as a result.
2279 The maildirsize information is now recalculated if the size or filecount
2282 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2283 support for a long time. Removed HAVE_SYS_VFS_H.
2285 28. Installed the latest version of exipick from John Jetmore.
2287 29. In an address list, if the pattern was not a regular expression, an empty
2288 subject address (from a bounce message) matched only if the pattern was an
2289 empty string. Non-empty patterns were not even tested. This was the wrong
2290 because it is perfectly reasonable to use an empty address as part of a
2291 database query. An empty address is now tested by patterns that are
2292 lookups. However, all the other forms of pattern expect the subject to
2293 contain a local part and a domain, and therefore, for them, an empty
2294 address still always fails if the pattern is not itself empty.
2296 30. Exim went into a mad DNS loop when attempting to do a callout where the
2297 host was specified on an smtp transport, and looking it up yielded more
2298 than one IP address.
2300 31. Re-factored the code for checking spool and log partition space into a
2301 function that finds that data and another that does the check. The former
2302 is then used to implement four new variables: $spool_space, $log_space,
2303 $spool_inodes, and $log_inodes.
2305 32. The RFC2047 encoding function was originally intended for short strings
2306 such as real names; it was not keeping to the 75-character limit for
2307 encoded words that the RFC imposes. It now respects the limit, and
2308 generates multiple encoded words if necessary. To be on the safe side, I
2309 have increased the buffer size for the ${rfc2047: expansion operator from
2312 33. It is now permitted to omit both strings after an "if" condition; if the
2313 condition is true, the result is "true". As before, when the second string
2314 is omitted, a false condition yields an empty string. This makes it less
2315 cumbersome to write custom ACL and router conditions.
2317 34. Failure to deliver a bounce message always caused it to be frozen, even if
2318 there was an errors_to setting on the router. The errors_to setting is now
2321 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2322 canonical form (fully expanded) before being placed in
2323 $sender_host_address.
2325 36. The table in the code that translates DNS record types into text (T_A to
2326 "A" for instance) was missing entries for NS and CNAME. It is just possible
2327 that this could have caused confusion if both these types were looked up
2328 for the same domain, because the text type is used as part of Exim's
2329 per-process caching. But the chance of anyone hitting this buglet seems
2332 37. The dnsdb lookup has been extended in a number of ways.
2334 (1) There is a new type, "zns", which walks up the domain tree until it
2335 finds some nameserver records. It should be used with care.
2337 (2) There is a new type, "mxh", which is like "mx" except that it returns
2338 just the host names, not the priorities.
2340 (3) It is now possible to give a list of domains (or IP addresses) to be
2341 looked up. The behaviour when one of the lookups defers can be
2342 controlled by a keyword.
2344 (4) It is now possible to specify the separator character for use when
2345 multiple records are returned.
2347 38. The dnslists ACL condition has been extended: it is now possible to supply
2348 a list of IP addresses and/or domains to be looked up in a particular DNS
2351 39. Added log_selector=+queue_time_overall.
2353 40. When running the queue in the test harness, wait just a tad after forking a
2354 delivery process, to get repeatability of debugging output.
2356 41. Include certificate and key file names in error message when GnuTLS fails
2357 to set them up, because the GnuTLS error message doesn't include the name
2358 of the failing file when there is a problem reading it.
2360 42. Allow both -bf and -bF in the same test run.
2362 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2364 44. The "Exiscan patch" is now merged into the mainline Exim source.
2366 45. Sometimes the final signoff response after QUIT could fail to get
2367 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2368 < 0 before doing a fflush(). This bug looks as though it goes back to the
2369 introduction of TLS in release 3.20, but "sometimes" must have been rare
2370 because the tests only now provoked it.
2372 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2373 (this can affect the format of dates).
2375 47. exim_tidydb, when checking for the continued existence of a message for
2376 which it has found a message-specific retry record, was not finding
2377 messages that were in split spool directories. Consequently, it was
2378 deleting retry records that should have stayed in existence.
2380 48. Steve fixed some bugs in eximstats.
2382 49. The SPA authentication driver was not abandoning authentication and moving
2383 on to the next authenticator when an expansion was forced to fail,
2384 contradicting the general specification for all authenticators. Instead it
2385 was generating a temporary error. It now behaves as specified.
2387 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2388 (the order specifies the preference for clients). The order is now AES256,
2389 AES128, 3DES, ARCFOUR128.
2391 51. Small patch to Sieve code - explicitly set From: when generating an
2394 52. Exim crashed if a remote delivery caused a very long error message to be
2395 recorded - for instance if somebody sent an entire SpamAssassin report back
2396 as a large number of 550 error lines. This bug was coincidentally fixed by
2397 increasing the size of one of Exim's internal buffers (big_buffer) that
2398 happened as part of the Exiscan merge. However, to be on the safe side, I
2399 have made the code more robust (and fixed the comments that describe what
2402 53. Now that there can be additional text after "Completed" in log lines (if
2403 the queue_time_overall log selector is set), a one-byte patch to exigrep
2404 was needed to allow it to recognize "Completed" as not the last thing in
2407 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2408 patch that reportedly fixes this has been added. I am not expert enough to
2409 create a test for it. This is what the patch creator wrote:
2411 "I found a little strange behaviour of ldap code when working with
2412 Windows 2003 AD Domain, where users was placed in more than one
2413 Organization Units. When I tried to give exim partial DN, the exit code
2414 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2415 But simultaneously result of request was absolutely normal ldap result,
2416 so I produce this patch..."
2418 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2419 REFERENCE, so I have modified the code to exclude the patch when that macro
2422 55. Some experimental protocols are using DNS PTR records for new purposes. The
2423 keys for these records are domain names, not reversed IP addresses. The
2424 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2425 leaves it alone. Component reversal etc. now happens only for IP addresses.
2428 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2430 57. Double the size of the debug message buffer (to 2048) so that more of very
2431 long debug lines gets shown.
2433 58. The exicyclog utility now does better if the number of log files to keep
2434 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2436 59. Two changes related to the smtp_active_hostname option:
2438 (1) $smtp_active_hostname is now available as a variable.
2439 (2) The default for smtp_banner uses $smtp_active_hostname instead
2440 of $primary_hostname.
2442 60. The host_aton() function is supposed to be passed a string that is known
2443 to be a valid IP address. However, in the case of IPv6 addresses, it was
2444 not checking this. This is a hostage to fortune. Exim now panics and dies
2445 if the condition is not met. A case was found where this could be provoked
2446 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2447 components; fortuitously, this particular loophole had already been fixed
2448 by change 4.50/55 above.
2450 If there are any other similar loopholes, the new check in host_aton()
2451 itself should stop them being exploited. The report I received stated that
2452 data on the command line could provoke the exploit when Exim was running as
2453 exim, but did not say which command line option was involved. All I could
2454 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2455 running as the user.
2458 61. There was a buffer overflow vulnerability in the SPA authentication code
2459 (which came originally from the Samba project). I have added a test to the
2460 spa_base64_to_bits() function which I hope fixes it.
2463 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2464 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2466 63. The daemon start-up calls getloadavg() while still root for those OS that
2467 need the first call to be done as root, but it missed one case: when
2468 deliver_queue_load_max is set with deliver_drop_privilege. This is
2469 necessary for the benefit of the queue runner, because there is no re-exec
2470 when deliver_drop_privilege is set.
2472 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2473 This has been fixed.
2475 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2476 was in use, was not putting the data itself into the right store pool;
2477 consequently, it could be overwritten for a subsequent message in the same
2478 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2481 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2483 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2484 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2485 the function were treating the return as a boolean value, which happened to
2486 work because 0=false and not-0=true, but is not correct code.
2488 68. The host_aton() function was not handling scoped IPv6 addresses (those
2489 with, for example, "%eth0" on the end) correctly.
2491 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2492 negated items (that is, ~something) in unsigned ints. Some compilers
2493 apparently mutter when there is no cast.
2495 70. If an address verification called from an ACL failed, and did not produce a
2496 user-specific message (i.e. there was only a "system" message), nothing was
2497 put in $acl_verify_message. In this situation, it now puts the system
2500 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2501 session; change 4.31/43 added the unwanted input to the log line - except
2502 that it did not do this in the start of session case. It now does.
2504 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2505 This is wrong and can cause the other end to generate a synchronization
2506 error if it is another Exim or anything else that does the synchronization
2507 check. A QUIT command is no longer sent after a timeout.
2509 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2510 during host lookups.
2512 74. The defer_ok option of callout verification was not working if it was used
2513 when verifying addresses in header lines, that is, for this case:
2515 verify = header_sender/callout=defer_ok
2517 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2518 those file descriptors could be used for SMTP connections. If anything
2519 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2520 could be sent to the SMTP client, causing chaos. The daemon now opens
2521 stdin, stdout, and stderr to /dev/null when it puts itself into the
2524 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2525 log by default. The user can override this with suitable Perl magic.
2527 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2528 the log message when discard triggers, was not working for the DATA ACL or
2529 for the non-SMTP ACL.
2531 78. Error message wording change in sieve.c.
2533 79. If smtp_accept_max_per_host was set, the number of connections could be
2534 restricted to fewer than expected, because the daemon was trying to set up
2535 a new connection before checking whether the processes handling previous
2536 connections had finished. The check for completed processes is now done
2537 earlier. On busy systems, this bug wouldn't be noticed because something
2538 else would have woken the daemon, and it would have reaped the completed
2541 80. If a message was submitted locally by a user whose login name contained one
2542 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2543 It caused a spool format error. I have fixed the spool reading code. A
2544 related problem was that the "from" clause in the Received: line became
2545 illegal because of the space(s). It is now covered by ${quote_local_part.
2547 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2550 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2552 83. Patch from Sieve maintainer to fix unterminated string problem in
2553 "vacation" handling.
2555 84. Some minor changes to the Linux configuration files to help with other
2556 OS variants using glibc.
2558 85. One more patch for Sieve to update vacation handling to latest spec.
2561 ----------------------------------------------------
2562 See the note above about the 4.44 and 4.50 releases.
2563 ----------------------------------------------------
2569 1. Change 4.43/35 introduced a bug that caused file counts to be
2570 incorrectly computed when quota_filecount was set in an appendfile
2573 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2574 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2577 3. Give more explanation in the error message when the command for a transport
2578 filter fails to execute.
2580 4. There are several places where Exim runs a non-Exim command in a
2581 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2582 was being done only for the command run by the queryprogram router. It is
2583 now done for all such subprocesses. The other cases are: ${run, transport
2584 filters, and the commands run by the lmtp and pipe transports.
2586 5. Some older OS have a limit of 256 on the maximum number of file
2587 descriptors. Exim was using setrlimit() to set 1000 as a large value
2588 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2589 systems. I've change it so that if it can't get 1000, it tries for 256.
2591 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2592 was an oversight, and furthermore, ever since the addition of extra
2593 controls (e.g. 4.43/32), the checks on when to allow different forms of
2594 "control" were broken. There should now be diagnostics for all cases when a
2595 control that does not make sense is encountered.
2597 7. $recipients is now available in the predata ACL (oversight).
2599 8. Tidy the search cache before the fork to do a delivery from a message
2600 received from the command line. Otherwise the child will trigger a lookup
2601 failure and thereby defer the delivery if it tries to use (for example) a
2602 cached ldap connection that the parent has called unbind on.
2604 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2605 of $address_data from the recipient verification was clobbered by the
2606 sender verification.
2608 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2609 was its contents. (It was OK if the option was not defined at all.)
2611 11. A "Completed" log line is now written for messages that are removed from
2612 the spool by the -Mrm option.
2614 12. $host_address is now set to the target address during the checking of
2615 ignore_target_hosts.
2617 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2618 being passed; this would have caused $sender_host_name to have been used if
2619 matching the list had actually called for a host name (not very likely,
2620 since this list is usually IP addresses). A host name is now passed as
2623 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2624 code that specifies a non-restarting handler (typically sigaction() in
2625 modern systems) in an attempt to fix a rare and obscure crash bug.
2627 15. Narrowed the window for a race in the daemon that could cause it to ignore
2628 SIGCHLD signals. This is not a major problem, because they are used only to
2629 wake it up if nothing else does.
2631 16. A malformed maildirsize file could cause Exim to calculate negative values
2632 for the mailbox size or file count. Odd effects could occur as a result.
2633 The maildirsize information is now recalculated if the size or filecount
2636 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2637 support for a long time. Removed HAVE_SYS_VFS_H.
2639 18. Updated exipick to current release from John Jetmore.
2641 19. Allow an empty sender to be matched against a lookup in an address list.
2642 Previously the only cases considered were a regular expression, or an
2645 20. Exim went into a mad DNS lookup loop when doing a callout where the
2646 host was specified on the transport, if the DNS lookup yielded more than
2649 21. The RFC2047 encoding function was originally intended for short strings
2650 such as real names; it was not keeping to the 75-character limit for
2651 encoded words that the RFC imposes. It now respects the limit, and
2652 generates multiple encoded words if necessary. To be on the safe side, I
2653 have increased the buffer size for the ${rfc2047: expansion operator from
2656 22. Failure to deliver a bounce message always caused it to be frozen, even if
2657 there was an errors_to setting on the router. The errors_to setting is now
2660 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2661 canonical form (fully expanded) before being placed in
2662 $sender_host_address.
2664 24. Updated eximstats to version 1.33
2666 25. Include certificate and key file names in error message when GnuTLS fails
2667 to set them up, because the GnuTLS error message doesn't include the name
2668 of the failing file when there is a problem reading it.
2670 26. Expand error message when OpenSSL has problems setting up cert/key files.
2673 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2674 (this can affect the format of dates).
2676 28. exim_tidydb, when checking for the continued existence of a message for
2677 which it has found a message-specific retry record, was not finding
2678 messages that were in split spool directories. Consequently, it was
2679 deleting retry records that should have stayed in existence.
2681 29. eximstats updated to version 1.35
2682 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2683 1.35 - bugfix such that pie charts by volume are generated correctly
2685 30. The SPA authentication driver was not abandoning authentication and moving
2686 on to the next authenticator when an expansion was forced to fail,
2687 contradicting the general specification for all authenticators. Instead it
2688 was generating a temporary error. It now behaves as specified.
2690 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2691 (the order specifies the preference for clients). The order is now AES256,
2692 AES128, 3DES, ARCFOUR128.
2694 31. Small patch to Sieve code - explicitly set From: when generating an
2697 32. Exim crashed if a remote delivery caused a very long error message to be
2698 recorded - for instance if somebody sent an entire SpamAssassin report back
2699 as a large number of 550 error lines. This bug was coincidentally fixed by
2700 increasing the size of one of Exim's internal buffers (big_buffer) that
2701 happened as part of the Exiscan merge. However, to be on the safe side, I
2702 have made the code more robust (and fixed the comments that describe what
2705 33. Some experimental protocols are using DNS PTR records for new purposes. The
2706 keys for these records are domain names, not reversed IP addresses. The
2707 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2708 leaves it alone. Component reversal etc. now happens only for IP addresses.
2711 34. The host_aton() function is supposed to be passed a string that is known
2712 to be a valid IP address. However, in the case of IPv6 addresses, it was
2713 not checking this. This is a hostage to fortune. Exim now panics and dies
2714 if the condition is not met. A case was found where this could be provoked
2715 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2716 components; fortuitously, this particular loophole had already been fixed
2717 by change 4.50/55 or 4.44/33 above.
2719 If there are any other similar loopholes, the new check in host_aton()
2720 itself should stop them being exploited. The report I received stated that
2721 data on the command line could provoke the exploit when Exim was running as
2722 exim, but did not say which command line option was involved. All I could
2723 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2724 running as the user.
2727 35. There was a buffer overflow vulnerability in the SPA authentication code
2728 (which came originally from the Samba project). I have added a test to the
2729 spa_base64_to_bits() function which I hope fixes it.
2732 36. The daemon start-up calls getloadavg() while still root for those OS that
2733 need the first call to be done as root, but it missed one case: when
2734 deliver_queue_load_max is set with deliver_drop_privilege. This is
2735 necessary for the benefit of the queue runner, because there is no re-exec
2736 when deliver_drop_privilege is set.
2738 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2739 was in use, was not putting the data itself into the right store pool;
2740 consequently, it could be overwritten for a subsequent message in the same
2741 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2744 38. Sometimes the final signoff response after QUIT could fail to get
2745 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2746 < 0 before doing a fflush(). This bug looks as though it goes back to the
2747 introduction of TLS in release 3.20, but "sometimes" must have been rare
2748 because the tests only now provoked it.
2754 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2755 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2756 Now it returns an integer. A number of calls were still expecting a T/F
2757 return. Fortuitously, in all cases, the tests worked in OK situations,
2758 which is the norm. However, things would have gone wrong on any write
2759 failures on the smtp file descriptor. This function is used when sending
2760 messages over SMTP and also when doing verify callouts.
2762 2. When Exim is called to do synchronous delivery of a locally submitted
2763 message (the -odf or -odi options), it no longer closes stderr before doing
2766 3. Implemented the mua_wrapper option.
2768 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2770 5. Implemented the functions header_remove(), header_testname(),
2771 header_add_at_position(), and receive_remove_recipient(), and exported them
2774 6. If an ACL "warn" statement specified the addition of headers, Exim already
2775 inserted X-ACL-Warn: at the start if there was no header name. However, it
2776 was not making this test for the second and subsequent header lines if
2777 there were newlines in the string. This meant that an invalid header could
2778 be inserted if Exim was badly configured.
2780 7. Allow an ACL "warn" statement to add header lines at the start or after all
2781 the Received: headers, as well as at the end.
2783 8. Added the rcpt_4xx retry error code.
2785 9. Added postmaster_mailfrom=xxx to callout verification option.
2787 10. Added mailfrom=xxxx to the callout verification option, for verify=
2790 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2791 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2793 12. Inserted some casts to stop certain compilers warning when using pointer
2794 differences as field lengths or precisions in printf-type calls (mostly
2795 affecting debugging statements).
2797 13. Added optional readline() support for -be (dynamically loaded).
2799 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2800 same clock tick as a message's arrival, so that its received time was the
2801 same as the "first fail" time on the retry record, and that message
2802 remained on the queue past the ultimate address timeout, every queue runner
2803 would try a delivery (because it was past the ultimate address timeout) but
2804 after another failure, the ultimate address timeout, which should have then
2805 bounced the address, did not kick in. This was a "< instead of <=" error;
2806 in most cases the first failure would have been in the next clock tick
2807 after the received time, and all would be well.
2809 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2810 being recognized when the domain list was tested by the match_domain
2811 condition in an expansion string.
2813 16. Added the ${str2b64: operator.
2815 17. Exim was always calling setrlimit() to set a large limit for the number of
2816 processes, without checking whether the existing limit was already
2817 adequate. (It did check for the limit on file descriptors.) Furthermore,
2818 errors from getrlimit() and setrlimit() were being ignored. Now they are
2819 logged to the main and panic logs, but Exim does carry on, to try to do its
2820 job under whatever limits there are.
2822 18. Imported PCRE 5.0.
2824 19. Trivial typo in log message " temporarily refused connection" (the leading
2827 20. If the log selector return_path_on_delivery was set and an address was
2828 redirected to /dev/null, the delivery process crashed because it assumed
2829 that a return path would always be set for a "successful" delivery. In this
2830 case, the whole delivery is bypassed as an optimization, and therefore no
2833 21. Internal re-arrangement: the function for sending a challenge and reading
2834 a response while authentication was assuming a zero-terminated challenge
2835 string. It's now changed to take a pointer and a length, to allow for
2836 binary data in such strings.
2838 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2840 23. Exim was not respecting finduser_retries when seeking the login of the
2841 uid under which it was called; it was always trying 10 times. (The default
2842 setting of finduser_retries is zero.) Also, it was sleeping after the final
2843 failure, which is pointless.
2845 24. Implemented tls_on_connect_ports.
2847 25. Implemented acl_smtp_predata.
2849 26. If the domain in control=submission is set empty, Exim assumes that the
2850 authenticated id is a complete email address when it generates From: or
2851 Sender: header lines.
2853 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2854 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2855 chown and chgrp in /bin and hostname in /usr/bin.
2857 28. Exim was keeping the "process log" file open after each use, just as it
2858 does for the main log. This opens the possibility of it remaining open for
2859 long periods when the USR1 signal hits a daemon. Occasional processlog
2860 errors were reported, that could have been caused by this. Anyway, it seems
2861 much more sensible not to leave this file open at all, so that is what now
2864 29. The long-running daemon process does not normally write to the log once it
2865 has entered its main loop, and it closes the log before doing so. This is
2866 so that log files can straightforwardly be renamed and moved. However,
2867 there are a couple of unusual error situations where the daemon does write
2868 log entries, and I had neglected to close the log afterwards.
2870 30. The text of an SMTP error response that was received during a remote
2871 delivery was being truncated at 512 bytes. This is too short for some of
2872 the long messages that one sometimes sees. I've increased the limit to
2875 31. It is now possible to make retry rules that apply only when a message has a
2876 specific sender, in particular, an empty sender.
2878 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2879 it possible to be selective about when SMTP synchronization is enforced.
2881 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2883 32. Implemented hosts_connection_nolog.
2885 33. Added an ACL for QUIT.
2887 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2890 35. Added mailbox_size and mailbox_filecount to appendfile.
2892 36. Added control = no_multiline_responses to ACLs.
2894 37. There was a bug in the logic of the code that waits for the clock to tick
2895 in the case where the clock went backwards by a substantial amount such
2896 that the microsecond fraction of "now" was more than the microsecond
2897 fraction of "then" (but the whole seconds number was less).
2899 38. Added support for the libradius Radius client library this is found on
2900 FreeBSD (previously only the radiusclient library was supported).
2906 1. When certain lookups returned multiple values in the form name=value, the
2907 quoting of the values was not always being done properly. Specifically:
2908 (a) If the value started with a double quote, but contained no whitespace,
2910 (b) If the value contained whitespace other than a space character (i.e.
2911 tabs or newlines or carriage returns) it was not quoted.
2912 This fix has been applied to the mysql and pgsql lookups by writing a
2913 separate quoting function and calling it from the lookup code. The fix
2914 should probably also be applied to nisplus, ibase and oracle lookups, but
2915 since I cannot test any of those, I have not disturbed their existing code.
2917 2. A hit in the callout cache for a specific address caused a log line with no
2918 reason for rejecting RCPT. Now it says "Previous (cached) callout
2919 verification failure".
2921 3. There was an off-by-one bug in the queryprogram router. An over-long
2922 return line was truncated at 256 instead of 255 characters, thereby
2923 overflowing its buffer with the terminating zero. As well as fixing this, I
2924 have increased the buffer size to 1024 (and made a note to document this).
2926 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2927 when Exim is waiting for an SMTP response from a remote server, Exim
2928 restarts its select() call on the socket, thereby resetting its timeout.
2929 This is not a problem when such interrupts are rare. Somebody set up a cron
2930 job to run exiwhat every 2 minutes, which is less than the normal select()
2931 timeout (5 or 10 minutes). This meant that the select() timeout never
2932 kicked in because it was always reset. I have fixed this by comparing the
2933 time when an interrupt arrives with the time at the start of the first call
2934 to select(). If more time than the timeout has elapsed, the interrupt is
2935 treated as a timeout.
2937 5. Some internal re-factoring in preparation for the addition of Sieve
2938 extensions (by MH). In particular, the "personal" test is moved to a
2939 separate function, and given an option for scanning Cc: and Bcc: (which is
2940 not set for Exim filters).
2942 6. When Exim created an email address using the login of the caller as the
2943 local part (e.g. when creating a From: or Sender: header line), it was not
2944 quoting the local part when it contained special characters such as @.
2946 7. Installed new OpenBSD configuration files.
2948 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2949 try to make them clearer.
2951 9. Callout options, other than the timeout value, were being ignored when
2952 verifying sender addresses in header lines. For example, when using
2954 verify = header_sender/callout=no_cache
2956 the cache was (incorrectly) being used.
2958 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2959 only the Cygwin environment.
2961 11. When return_path_on_delivery was set as a log selector, if different remote
2962 addresses in the same message used different return paths and parallel
2963 remote delivery occurred, the wrong values would sometimes be logged.
2964 (Whenever a remote delivery process finished, the return path value from
2965 the most recently started remote delivery process was logged.)
2967 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2968 lines when AUTH and/or TLS are in use. This is the "received protocol"
2969 field. Exim used to use "asmtp" for authenticated SMTP, without any
2970 indication (in the protocol name) for TLS use. Now it follows the RFC and
2971 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2972 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2973 names appear in log lines as well as in Received: header lines.
2975 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2976 extensions, and comparison tests, and to fix some bugs.
2978 14. Changes to the "personal" filter test:
2980 (1) The test was buggy in that it was just doing the equivalent of
2981 "contains" tests on header lines. For example, if a user's address was
2982 anne@some.where, the "personal" test would incorrectly be true for
2984 To: susanne@some.where
2986 This test is now done by extracting each address from the header in turn,
2987 and checking the entire address. Other tests that are part of "personal"
2988 are now done using regular expressions (for example, to check local parts
2989 of addresses in From: header lines).
2991 (2) The list of non-personal local parts in From: addresses has been
2992 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2993 taken from the Sieve specification recommendations.
2995 (3) If the message contains any header line starting with "List-" it is
2996 treated as non-personal.
2998 (4) The test for "circular" in the Subject: header line has been removed
2999 because it now seems ill-conceived.
3001 15. Minor typos in src/EDITME comments corrected.
3003 16. Installed latest exipick from John Jetmore.
3005 17. If headers_add on a router specified a text string that was too long for
3006 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3007 of string_sprintf() is now avoided.
3009 18. $message_body_size was not set (it was always zero) when running the DATA
3010 ACL and the local_scan() function.
3012 19. For the "mail" command in an Exim filter, no default was being set for
3013 the once_repeat time, causing a random time value to be used if "once" was
3014 specified. (If the value happened to be <= 0, no repeat happened.) The
3015 default is now 0s, meaning "never repeat". The "vacation" command was OK
3016 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3017 (I found it when inspecting the code).
3019 20. There is now an overall timeout for performing a callout verification. It
3020 defaults to 4 times the callout timeout, which applies to individual SMTP
3021 commands during the callout. The overall timeout applies when there is more
3022 than one host that can be tried. The timeout is checked before trying the
3023 next host. This prevents very long delays if there are a large number of
3024 hosts and all are timing out (e.g. when the network connections are timing
3025 out). The value of the overall timeout can be changed by specifying an
3026 additional sub-option for "callout", called "maxwait". For example:
3028 verify = sender/callout=5s,maxwait=20s
3030 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3031 to the end before writing, but this should make it even safer).
3033 22. Exim was forgetting that it had advertised PIPELINING for the second and
3034 subsequent messages on an SMTP connection. It was also not resetting its
3035 memory on STARTTLS and an internal HELO.
3037 23. When Exim logs an SMTP synchronization error within a session, it now
3038 records whether PIPELINING has been advertised or not.
3040 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3041 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3042 rather than long int.
3044 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3046 26. Added the never_mail option to autoreply.
3052 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3053 crash if the getsockname() call failed; this can happen if a connection is
3054 closed very soon after it is established. The problem was simply in the
3055 order in which certain operations were done, causing Exim to try to write
3056 to the SMTP stream before it had set up the file descriptor. The bug has
3057 been fixed by making things happen in the correct order.
3063 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3064 before the connection was closed, thus losing the rejection response.
3066 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3067 some early Solaris releases, but causes trouble in current releases where
3068 socklen_t is defined.
3070 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3073 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3076 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3077 that releasing the top of it at the end releases what was used for sub-
3078 expansions (unless the block got too big). However, discard this block if
3079 the first thing is a variable or header, so that we can use its block when
3080 it is dynamic (useful for very large $message_headers, for example).
3082 6. Lookups now cache *every* query, not just the most recent. A new, separate
3083 store pool is used for this. It can be recovered when all lookup caches are
3084 flushed. Lookups now release memory at the end of their result strings.
3085 This has involved some general refactoring of the lookup sources.
3087 7. Some code has been added to the store_xxx() functions to reduce the amount
3088 of flapping under certain conditions.
3090 8. log_incoming_interface used to affect only the <= reception log lines. Now
3091 it causes the local interface and port to be added to several more SMTP log
3092 lines, for example "SMTP connection from", and rejection lines.
3094 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3096 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3098 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3099 could be overwritten at the end of the current message (or the start of a
3100 new message if it was set in a HELO ACL). The value is now preserved for
3101 the duration of the SMTP connection.
3103 12. If a transport had a headers_rewrite setting, and a matching header line
3104 contained an unqualified address, that address was qualified, even if it
3105 did not match any rewriting rules. The underlying bug was that the values
3106 of the flags that permit the existence of unqualified sender and recipient
3107 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3108 non-local messages, and by -bnq for local messages) were not being
3109 preserved with the message after it was received.
3111 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3112 "next input=" as part of the text comprising the host identity instead of
3113 the correct text. The code was using the same buffer for two different
3114 strings. However, depending on which order the printing function evaluated
3115 its arguments, the bug did not always show up. Under Linux, for example, my
3116 test suite worked just fine.
3118 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3119 doesn't work with some older versions of Perl. It has been changed to "my",
3120 which in any case is probably the better facility to use.
3122 15. A really picky compiler found some instances of statements for creating
3123 error messages that either had too many or two few arguments for the format
3126 16. The size of the buffer for calls to the DNS resolver has been increased
3127 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3128 for addresses that have a lot of PTR records. This alleviates a problem; it
3129 does not fully solve it.
3131 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3132 buffer now truncates the list and logs the incident, which is the same
3133 action as happens when Exim is looking up a host name and its aliases.
3134 Previously in this situation something unpredictable would happen;
3135 sometimes it was "internal error: store_reset failed".
3137 18. If a server dropped the connection unexpectedly when an Exim client was
3138 using GnuTLS and trying to read a response, the client delivery process
3139 crashed while trying to generate an error log message.
3141 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3142 string, for example:
3144 warn message = H1: something\nH2: something
3146 the text was added as a single header line from Exim's point of view
3147 though it ended up OK in the delivered message. However, searching for the
3148 second and subsequent header lines using $h_h2: did not work. This has been
3149 fixed. Similarly, if a system filter added multiple headers in this way,
3150 the routers could not see them.
3152 20. Expanded the error message when iplsearch is called with an invalid key to
3153 suggest using net-iplsearch in a host list.
3155 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3156 ACLs are no longer actually imposed (and a message to that effect is
3159 22. If a "gecos" field in a passwd entry contained escaped characters, in
3160 particular, if it contained a \" sequence, Exim got it wrong when building
3161 a From: or a Sender: header from that name. A second bug also caused
3162 incorrect handling when an unquoted " was present following a character
3163 that needed quoting.
3165 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3166 was not being matched caselessly.
3168 24. Arranged for all hyphens in the exim.8 source to be escaped with
3171 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3172 properly. Recipient callout cache records were still being keyed to include
3173 the sender, even when use_sender was set false. This led to far more
3174 callouts that were necessary. The sender is no longer included in the key
3175 when use_sender is false.
3177 26. Added "control = submission" modifier to ACLs.
3179 27. Added the ${base62d: operator to decode base 62 numbers.
3181 28. dnsdb lookups can now access SRV records.
3183 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3184 the configuration file.
3186 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3187 (-v) mode. This makes the output for a verbose queue run more intelligible.
3189 31. Added a use_postmaster feature to recipient callouts.
3191 32. Added the $body_zerocount variable, containing the number of binary zero
3192 bytes in the message body.
3194 33. The time of last modification of the "new" subdirectory is now used as the
3195 "mailbox time last read" when there is a quota error for a maildir
3198 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3200 35. Added +ignore_unknown as a special item in host lists.
3202 36. Code for decoding IPv6 addresses in host lists is now included, even if
3203 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3204 address was recognized as an IP address, but was then not correctly decoded
3205 into binary, causing unexpected and incorrect effects when compared with
3212 1. Very minor rewording of debugging text in manualroute to say "list of
3213 hosts" instead of "hostlist".
3215 2. If verify=header_syntax was set, and a header line with an unqualified
3216 address (no domain) and a large number of spaces between the end of the
3217 name and the colon was received, the reception process suffered a buffer
3218 overflow, and (when I tested it) crashed. This was caused by some obsolete
3219 code that should have been removed. The fix is to remove it!
3221 3. When running in the test harness, delay a bit after writing a bounce
3222 message to get a bit more predictability in the log output.
3224 4. Added a call to search_tidyup() just before forking a reception process. In
3225 theory, someone could use a lookup in the expansion of smtp_accept_max_
3226 per_host which, without the tidyup, could leave open a database connection.
3228 5. Added the variables $recipient_data and $sender_data which get set from a
3229 lookup success in an ACL "recipients" or "senders" condition, or a router
3230 "senders" option, similar to $domain_data and $local_part_data.
3232 6. Moved the writing of debug_print from before to after the "senders" test
3235 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3236 problems for message scanning, either using a data ACL, or using
3237 local_scan() because the Received: header was not generated till after they
3238 were called (in order to set the time as the time of reception completion).
3239 I have revised the way this works. The header is now generated after the
3240 body is received, but before the ACL or local_scan() are called. After they
3241 are run, the timestamp in the header is updated.
3247 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3248 before starting a queue runner without re-exec. This happened only when
3249 deliver_drop_privilege was set or when the Exim user was set to root. The
3250 effect of the bug was that timeouts during subsequent deliveries caused
3251 crashes instead of being properly handled. The handler is now left at its
3252 default (and expected) setting.
3254 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3255 message, again when deliver_drop_privilege is set or Exim is run as root.
3256 The bug described in (1) was not present in this case, but the tidying up
3257 of the other signals was missing. I have made the two cases consistent.
3259 3. The ignore_target_hosts setting on a manualroute router was being ignored
3260 for hosts that were looked up using the /MX notation.
3262 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3265 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3266 operated on the sender address. After changing the $sender_address to <>
3267 for the sender address verify, Exim was re-instated it as the original
3268 (before rewriting) address, but remembering that it had rewritten it, so it
3269 wasn't rewriting it again. This bug also had the effect of breaking the
3270 sender address verification caching when the sender address was rewritten.
3272 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3273 This has been changed so that if the ip literal address matches
3274 ignore_target_hosts, the router declines.
3276 7. Added expansion conditions match_domain, match_address, and match_local_
3277 part (NOT match_host).
3279 8. The placeholder for the Received: header didn't have a length field set.
3281 9. Added code to Exim itself and to exim_lock to test for a specific race
3282 condition that could lead to file corruption when using MBX delivery. The
3283 issue is with the lockfile that is created in /tmp. If this file is removed
3284 after a process has opened it but before that process has acquired a lock,
3285 there is the potential for a second process to recreate the file and also
3286 acquire a lock. This could lead to two Exim processes writing to the file
3287 at the same time. The added code performs the same test as UW imapd; it
3288 checks after acquiring the lock that its file descriptor still refers to
3289 the same named file.
3291 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3292 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3295 11. Added the smtp_active_hostname option. If used, this will typically be made
3296 to depend on the incoming interface address. Because $interface_address is
3297 not set up until the daemon has forked a reception process, error responses
3298 that can happen earlier (such as "too many connections") no longer contain
3301 12. If an expansion in a condition on a "warn" statement fails because a lookup
3302 defers, the "warn" statement is abandoned, and the next ACL statement is
3303 processed. Previously this caused the whole ACL to be aborted.
3305 13. Added the iplsearch lookup type.
3307 14. Added ident_timeout as a log selector.
3309 15. Added tls_certificate_verified as a log selector.
3311 16. Added a global option tls_require_ciphers (compare the smtp transport
3312 option of the same name). This controls incoming TLS connections.
3314 17. I finally figured out how to make tls_require_ciphers do a similar thing
3315 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3316 before starting the TLS session.
3318 18. Tabs are now shown as \t in -bP output.
3320 19. If the log selector return_path_on_delivery was set, Exim crashed when
3321 bouncing a message because it had too many Received: header lines.
3323 20. If two routers both had headers_remove settings, and the first one included
3324 a superfluous trailing colon, the final name in the first list and the
3325 first name in the second list were incorrectly joined into one item (with a
3326 colon in the middle).
3332 1. Added -C and -D options to the exinext utility, mainly to make it easier
3333 to include in the automated testing, but these could be helpful when
3334 multiple configurations are in use.
3336 2. The exinext utility was not formatting the output nicely when there was
3337 an alternate port involved in the retry record key, nor when there was a
3338 message id as well (for retries that were specific to a specific message
3339 and a specific host). It was also confused by IPv6 addresses, because of
3340 the additional colons they contain. I have fixed the IPv4 problem, and
3341 patched it up to do a reasonable job for IPv6.
3343 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3344 delivery, the log line now contains "pipelined" if PIPELINING was used.
3346 4. An SMTP transport process used to panic and die if the bind() call to set
3347 an explicit outgoing interface failed. This has been changed; it is now
3348 treated in the same way as a connect() failure.
3350 5. A reference to $sender_host_name in the part of a conditional expansion
3351 that was being skipped was still causing a DNS lookup. This no longer
3354 6. The def: expansion condition was not recognizing references to header lines
3355 that used bh_ and bheader_.
3357 7. Added the _cache feature to named lists.
3359 8. The code for checking quota_filecount in the appendfile transport was
3360 allowing one more file than it should have been.
3362 9. For compatibility with Sendmail, the command line option
3370 and sets the incoming protocol and host name (for trusted callers). The
3371 host name and its colon can be omitted when only the protocol is to be set.
3372 Note the Exim already has two private options, -pd and -ps, that refer to
3373 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3374 "s", but I don't think that's a major issue.
3376 10. A number of refactoring changes to the code, none of which should affect
3379 (a) The number of logging options was getting close to filling up the
3380 32-bit word that was used as a bit map. I have split them into two classes:
3381 those that are passed in the argument to log_write(), and those that are
3382 only ever tested independently outside of that function. These are now in
3383 separate 32-bit words, so there is plenty of room for expansion again.
3384 There is no change in the user interface or the logging behaviour.
3386 (b) When building, for example, log lines, the code previously used a
3387 macro that called string_cat() twice, in order to add two strings. This is
3388 not really sufficiently general. Furthermore, there was one instance where
3389 it was actually wrong because one of the argument was used twice, and in
3390 one call a function was used. (As it happened, calling the function twice
3391 did not affect the overall behaviour.) The macro has been replaced by a
3392 function that can join an arbitrary number of extra strings onto a growing
3395 (c) The code for expansion conditions now uses a table and a binary chop
3396 instead of a serial search (which was left over from when there were very
3397 few conditions). Also, it now recognizes conditions like "pam" even when
3398 the relevant support is not compiled in: a suitably worded error message is
3399 given if an attempt is made to use such a condition.
3401 11. Added ${time_interval:xxxxx}.
3403 12. A bug was causing one of the ddress fields not to be passed back correctly
3404 from remote delivery subprocesses. The field in question was not being
3405 subsequently used, so this caused to problems in practice.
3407 13. Added new log selectors queue_time and deliver_time.
3409 14. Might have fixed a bug in maildirsizefile handling that threw up
3410 "unexpected character" debug warnings, and recalculated the data
3411 unnecessarily. In any case, I expanded the warning message to give more
3414 15. Added the message "Restricted characters in address" to the statements in
3415 the default ACL that block characters like @ and % in local parts.
3417 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3418 Three changes have been made:
3420 (a) There was a serious bug; a negative response to MAIL caused the whole
3421 recipient domain to be cached as invalid, thereby blocking all messages
3422 to all local parts at the same domain, from all senders. This bug has
3423 been fixed. The domain is no longer cached after a negative response to
3424 MAIL if the sender used is not empty.
3426 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3429 (c) A new callout option, "use_sender" has been added for people who want
3430 the modified behaviour.
3436 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3439 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3440 indeed breaks things for older releases.
3442 3. Added additional logging to the case where there is a problem reading data
3443 from a filter that is running in a subprocess using a pipe, in order to
3444 try to track down a specific problem.
3446 4. Testing facility fudge: when running in the test harness and attempting
3447 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3448 getting "No route to host". Convert this to a timeout.
3450 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3453 6. Some OS don't have socklen_t but use size_t instead. This affects the
3454 fifth argument of getsockopt() amongst other things. This is now
3455 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3456 can be set for individual OS. I have set it for SunOS5, OSF1, and
3457 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3458 some earlier ones do not.
3460 7. Change 4.30/15 was not doing the test caselessly.
3462 8. The standard form for an IPv6 address literal was being rejected by address
3463 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3464 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3465 this, as well as the form without the "IPv6" on the front (but only when
3466 address literals are enabled, of course).
3468 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3470 10. Exim crashed if a message with an empty sender address specified by -f
3471 encountered a router with an errors_to setting. This could be provoked only
3472 by a command such as
3476 where an empty string was supplied; "<>" did not hit this bug.
3478 11. Installed PCRE release 4.5.
3480 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3481 remained set. It is now erased.
3483 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3484 times from message ids (which are base 36 rather than the normal 62).
3486 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3487 were being counted as actual protocol errors, and logged if the log
3488 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3489 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3490 and DATA following a set of rejected RCPTs do not count as protocol errors.
3491 In other words, Exim assumes they were pipelined, though this may not
3492 actually be the case. Of course, in all cases the client gets an
3493 appropriate error code.
3495 15. If a lookup fails in an ACL condition, a message about the failure may
3496 be available; it is used if testing the ACL cannot continue, because most
3497 such messages specify what the cause of the deferral is. However, some
3498 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3499 that caused an old message to be retained and used if a later statement
3500 caused a defer, replacing the real cause of the deferral.
3502 16. If an IP address had so many PTR records that the DNS lookup buffer
3503 was not large enough to hold them, Exim could crash while trying to process
3504 the truncated data. It now detects and logs this case.
3506 17. Further to 4.21/58, another change has been made: if (and only if) the
3507 first line of a message (the first header line) ends with CRLF, a bare LF
3508 in a subsequent header line has a space inserted after it, so as not to
3509 terminate the header.
3511 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3512 unnecessarily, used atoi() instead of strtol(), and didn't check the
3513 termination when getting file sizes from file names by regex.
3515 19. Completely re-implemented the support for maildirsize files, in the light
3516 of a number of problems with the previous contributed implementation
3517 (4.30/29). In particular:
3519 . If the quota is zero, the maildirsize file is maintained, but no quota is
3522 . If the maildir directory does not exist, it is created before any attempt
3523 to write a maildirsize file.
3525 . The quota value in the file is just a cache; if the quota is changed in
3526 the transport, the new value overrides.
3528 . A regular expression is available for excluding directories from the
3531 20. The autoreply transport checks the characters in options that define the
3532 message's headers; it allows continued headers, but it was checking with
3533 isspace() after an embedded newline instead of explicitly looking for a
3536 21. If all the "regular" hosts to which an address was routed had passed their
3537 expiry times, and had not reached their retry times, the address was
3538 bounced, even if fallback hosts were defined. Now Exim should go on to try
3541 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3542 equivalent code in the SMTP transport. Some hosts send humungous responses
3543 to HELO/EHLO, more than 1024 it seems.
3545 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3546 gives compiler warnings in some environments. I've now done it "properly",
3549 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3550 (because of problems with the built-in one) was declared to return uschar *
3551 instead of char *, causing compiler failure.
3553 25. Fixed a file descriptor leak when processing alias/forward files.
3555 26. Fixed a minor format string issue in dbfn.c.
3557 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3559 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3560 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3563 29. When a sender address is verified, it is cached, to save repeating the test
3564 when there is more than one recipient in a message. However, when the
3565 verification involves a callout, it is possible for different callout
3566 options to be set for different recipients. It is too complicated to keep
3567 track of this in the cache, so now Exim always runs a verification when a
3568 callout is required, relying on the callout cache for the optimization.
3569 The overhead is duplication of the address routing, but this should not be
3572 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3573 to be verified with callout=postmaster, and the main callout worked but the
3574 postmaster check failed, the verification correctly failed. However, if a
3575 subsequent RCPT command asked for sender verification *without* the
3576 postmaster check, incorrect caching caused this verification also to fail,
3579 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3580 it was not caching the DNS options (qualify_single, search_parents) that
3581 were used when the lookup failed. A subsequent lookup with different
3582 options therefore always gave the same answer, though there were cases
3583 where it should not have. (Example: a "domains = !$mx_any" option on a
3584 dnslookup router: the "domains" option is always processed without any
3585 widening, but the router might have qualify_single set.) Now Exim uses the
3586 cached value only when the same options are set.
3588 32. Added John Jetmore's "exipick" utility to the distribution.
3590 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3591 than a timeout (e.g. a certificate is required, and is not provided), an
3592 Exim server now closes the connection immediately. Previously it waited for
3593 the client to close - but if the client is SSL, it seems that they each
3594 wait for each other, leading to a delay before one of them times out.
3596 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3597 maintained 0.8.x compatibility because I don't think many are using it, and
3598 it is clearly obsolete.
3600 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3603 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3604 client certificate was expired. A simple patch fixes this, though I don't
3605 understand the full logic of why the verify callback is called multiple
3608 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3609 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3610 which causes problems with some clients (such as the Certicom SSL Plus
3611 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3612 disables the coutermeasure allowing Eudora to connect."
3614 38. Exim was not checking that a write() to a log file succeeded. This could
3615 lead to Bad Things if a log got too big, in particular if it hit a file
3616 size limit. Exim now panics and dies if it cannot write to a log file, just
3617 as it does if it cannot open a log file.
3619 39. Modified OS/Makefile-Linux so that it now contains
3621 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3623 The two -D definitions ensure that Exim is compiled with large file
3624 support, which makes it possible to handle log files that are bigger than
3627 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3628 instance) a domain was checked against a named list that involved a lookup,
3629 causing $domain_data to be set, then another domain was checked against the
3630 same list, then the first domain was re-checked, the value of $domain_data
3631 after the final check could be wrong. In particular, if the second check
3632 failed, it could be set empty. This bug probably also applied to
3635 41. The strip_trailing_dot option was not being applied to the address given
3636 with the -f command-line option.
3638 42. The code for reading a message's header from the spool was incrementing
3639 $received_count, but never initializing it. This meant that the value was
3640 incorrect (doubled) while delivering a message in the same process in which
3641 it was received. In the most common configuration of Exim, this never
3642 happens - a fresh exec is done - but it can happen when
3643 deliver_drop_privilege is set.
3645 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3646 it now includes up to 150 characters of the unexpected data in the log
3649 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3650 and building data strings. The size of both of these buffers was 10 000
3651 bytes - far larger than anybody would *ever* want, thought I. Needless to
3652 say, somebody hit the limit. I have increased the maximum line length to
3653 20 000 and the maximum data length of concatenated lines to 100 000. I have
3654 also fixed two bugs, because there was no checking on these buffers. Tsk,
3655 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3656 buffer is too small.
3658 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3659 lsearch lookups. Now it does.
3661 46. When parsing a route_list item in a manualroute router, a fixed-length
3662 buffer was used for the list of hosts. I made this 1024 bytes long,
3663 thinking that nobody would ever have a list of hosts that long. Wrong.
3664 Somebody had a whole pile of complicated expansion conditions, and the
3665 string was silently truncated, leading to an expansion error. It turns out
3666 that it is easier to change to an unlimited length (owing to other changes
3667 that have happened since this code was originally written) than to build
3668 structure for giving a limitation error. The length of the item that
3669 expands into the list of hosts is now unlimited.
3671 47. The lsearch lookup could not handle data where the length of text line was
3672 more than 4095 characters. Such lines were truncated, leading to shortened
3673 data being returned. It should now handle lines of any length.
3675 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3676 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3679 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3680 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3681 no longer happens in comment lines. A list of replacements is now placed
3682 at the head of all of the source files, except those whose only change is
3683 to replace PERL_COMMAND in the very first #! line.
3685 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3686 messages on the queue, with a bottom-up merge sort, using code contributed
3687 by Michael Haardt. This should make operations like -bp somewhat faster on
3688 large queues. It won't affect queue runners, except when queue_run_in_order
3691 51. Installed eximstats 1.31 in the distribution.
3693 52. Added support for SRV lookups to the dnslookup router.
3695 53. If an ACL referred to $message_body or $message_body_end, the value was not
3696 reset for any messages that followed in the same SMTP session.
3698 54. The store-handling optimization for building very long strings was not
3699 differentiating between the different store pools. I don't think this
3700 actually made any difference in practice, but I've tidied it.
3702 55. While running the routers to verify a sender address, $sender_address
3703 was still set to the sender address. This is wrong, because when routing to
3704 send a bounce to the sender, it would be empty. Therefore, I have changed
3705 it so that, while verifying a sender address, $sender_address is set to <>.
3706 (There is no change to what happens when verifying a recipient address.)
3708 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3709 target A or AAAA records (if not already returned) without resetting the
3710 qualify_single or search_parents options of the DNS resolver. These are
3711 inappropriate in this case because the targets of MX and SRV records must
3712 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3713 target that, when qualified, matched something in the local domain. These
3714 two options are now turned off when doing these lookups.
3716 57. It seems that at least some releases of Reiserfs (which does not have the
3717 concept of a fixed number of inodes) returns zero and not -1 for the
3718 number of available inodes. This interacted badly with check_spool_inodes,
3719 which assumed that -1 was the "no such thing" setting. What I have done is
3720 to check that the total number of inodes is greater than zero before doing
3721 the test of how many are available.
3723 58. When a "warn" ACL statement has a log_message modifier, the message is
3724 remembered, and not repeated. This is to avoid a lot of repetition when a
3725 message has many recipients that cause the same warning to be written.
3726 Howewer, Exim was preserving the list of already written lines for an
3727 entire SMTP session, which doesn't seem right. The memory is now reset if a
3728 new message is started.
3730 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3731 debugging output unless log_rewrite was also set.
3733 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3734 of (int)(handle) when we know that handle contains (void *)(-1).
3736 61. The Exim daemon panic-logs an error return when it closes the incoming
3737 connection. However "connection reset by peer" seems to be common, and
3738 isn't really an error worthy of noting specially, so that particular error
3741 62. When Exim is trying to find all the local interfaces, it used to panic and
3742 die if the ioctl to get the interface flags failed. However, it seems that
3743 on at least one OS (Solaris 9) it is possible to have an interface that is
3744 included in the list of interfaces, but for which you get a failure error
3745 for this call. This happens when the interface is not "plumbed" into a
3746 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3747 failure of the "get flags" call assumes that the interface is down.
3749 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3750 makes life easier for people who are doing arithmetic on fields extracted
3751 from dates, where you often get leading zeros that should not be
3752 interpreted as octal.
3754 64. Added qualify_domain to the redirect router, to override the global
3757 65. If a pathologically long header line contained very many addresses (the
3758 report of this problem mentioned 10 000) and each of them was rewritten,
3759 Exim could use up a very large amount of memory. (It kept on making new
3760 copies of the header line as it rewrote, and never released the old ones.)
3761 At the expense of a bit more processing, the header rewriting function has
3762 been changed so that it no longer eats memory in this way.
3764 66. The generation of the Received: header has been moved from the time that a
3765 message starts to be received, to the time that it finishes. The timestamp
3766 in the Received: header should now be very close to that of the <= log
3767 line. There are two side-effects of this change:
3769 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3770 logged header lines no longer include the local Received: line, because
3771 it has not yet been created. The same applies to a copy of the message
3772 that is returned to a non-SMTP sender when a message is rejected.
3774 (b) When a filter file is tested using -bf, no additional Received: header
3775 is added to the test message. After some thought, I decided that this
3778 This change does not affect the value of $received_for. It is still set
3779 after address rewriting, but before local_scan() is called.
3781 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3783 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3784 gave an unhelpful panic error message, and a defer error. I have managed to
3785 change this behaviour so that it now rejects any supplied certificate,
3786 which seems right, as the list of acceptable certificates is empty.
3788 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3789 gave an unhelpful defer error. I have not managed to make this reject any
3790 supplied certificates, but the error message it gives is "no certificate
3791 supplied", which is not helpful.
3793 70. exigrep's output now also includes lines that are not associated with any
3794 message, but which match the given pattern. Implemented by a patch from
3795 Martin Sluka, which also tidied up the Perl a bit.
3797 71. Recipient callout verification, like sender verification, was using <> in
3798 the MAIL FROM command. This isn't really the right thing, since the actual
3799 sender may affect whether the remote host accepts the recipient or not. I
3800 have changed it to use the actual sender in the callout; this means that
3801 the cache record is now keyed on a recipient/sender pair, not just the
3802 recipient address. There doesn't seem to be a real danger of callout loops,
3803 since a callout by the remote host to check the sender would use <>.
3804 [SEE ABOVE: changed after hitting problems.]
3806 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3807 temporary errors. However, in the case of such a code being given after
3808 the end of a data transmission (i.e. after ".") Exim was failing to write
3809 a retry record for the message. (Yes, there was some broken host that was
3810 actually sending 8xx at this point.)
3812 73. An unknown lookup type in a host list could cause Exim to panic-die when
3813 the list was checked. (An example that provoked this was putting <; in the
3814 middle of a list instead of at the start.) If this happened during a DATA
3815 ACL check, a -D file could be left lying around. This kind of configuration
3816 error no longer causes Exim to die; instead it causes a defer errror. The
3817 incident is still logged to the main and panic logs.
3819 74. Buglet left over from Exim 3 conversion. The message "too many messages
3820 in one connection" was written to the rejectlog but not the mainlog, except
3821 when address rewriting (yes!) was being logged.
3823 75. Added write_rejectlog option.
3825 76. When a system filter was run not as root (that is, when system_filter_user
3826 was set), the values of the $n variables were not being returned to the
3827 main process; thus, they were not subsequently available in the $sn
3830 77. Added +return_path_on_delivery log selector.
3832 78. A connection timeout was being treated differently from recipients deferred
3833 when testing hosts_max_try with a message that was older than the host's
3834 retry timeout. (The host should not be counted, thus allowing all hosts to
3835 be tried at least once before bouncing.) This may have been the cause of an
3836 occasionally reported bug whereby a message would remain on the queue
3837 longer than the retry timeout, but would be bounced if a delivery was
3838 forced. I say "may" because I never totally pinned down the problem;
3839 setting up timeout/retry tests is difficult. See also the next item.
3841 79. The ultimate address timeout was not being applied to errors that involved
3842 a combination of host plus message (for example, a timeout on a MAIL
3843 command). When an address resolved to a number of possible hosts, and they
3844 were not all tried for each delivery (e.g. because of hosts_max_try), a
3845 message could remain on the queue longer than the retry timeout.
3847 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3850 81. Fixed an obscure SMTP outgoing bug which required at least the following
3851 conditions: (a) there was another message waiting for the same server;
3852 (b) the server returned 5xx to all RCPT commands in the first message so
3853 that the message was not completed; (c) the server dropped the connection
3854 or gave a negative response to the RSET that Exim sends to abort the
3855 transaction. The observed case was a dropped connection after DATA that had
3856 been sent in pipelining mode. That is, the server had advertised PIPELINING
3857 but was not implementing it correctly. The effect of the bug was incorrect
3858 behaviour, such as trying another host, and this could lead to a crash.
3864 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3865 and daemon.c were passed as pointers to ints; they should have been
3866 pointers to socklen_t variables (which are typically unsigned ints).
3868 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3871 3. Fixed a really odd bug that affected only the testing scheme; patching a
3872 certain fixed string in the binary changed the value of another string that
3873 happened to be identical to the end of the original first string.
3875 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3876 name", it returns that address as the IP address. On some operating
3877 systems (e.g. Solaris), it also passes back the IP address string as the
3878 "host name". However, on others (e.g. Linux), it passes back an empty
3879 string. Exim wasn't checking for this, and was changing the host name to an
3880 empty string, assuming it had been canonicized.
3882 5. Although rare, it is permitted to have more than one PTR record for a given
3883 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3884 all the names associated with an address, because they do in Solaris.
3885 However, it seems that they do not in Linux for data that comes from the
3886 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3887 I found this out when I moved to a new Linux workstation and tried to run
3888 the Exim test suite.
3890 To get round this problem I have changed the code so that it now does its
3891 own call to the DNS to look up PTR records when searching for a host name.
3892 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3893 addresses that are only in /etc/hosts are still found.
3895 This behaviour is, however, controlled by an option called host_lookup_
3896 order, which defaults to "bydns:byaddr". If people want to use the other
3897 order, or indeed, just use one or the other means of lookup, they can
3898 specify it in this variable.
3900 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3901 some operating systems, this comes back from gethostbyaddr() as an empty
3902 string, and this is what Exim used to test for. However, it seems that in
3903 other systems, "." is yielded. Exim now tests for this case too.
3905 7. The values of check_spool_space and check_log_space are now held internally
3906 as a number of kilobytes instead of an absolute number of bytes. If a
3907 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3908 kilobyte. This means that much larger values can be stored.
3910 8. Exim monitor: an attempt to get the action menu when not actually pointing
3911 at a message produces an empty menu entitled "No message selected". This
3912 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3913 no entries in it ("Shell widget menu has zero width and/or height"). So I
3914 have added a single, blank menu entry in this case.
3916 9. Added ${quote_local_part.
3918 10. MIME decoding is now applied to the contents of Subject: header lines when
3921 11. Now that a reference to $sender_host_address automatically causes a reverse
3922 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3923 host lookup before query-style lookups in lists that might use this
3924 variable. This has therefore been abolished, and the "net-" prefix is no
3925 longer necessary for query-style lookups.
3927 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3928 have been a typo for LFLAGS, so it has been changed.
3930 13. The install script calls Exim with "-C /dev/null" in order to find the
3931 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3932 to be output. Howeve, since Exim outputs its version number before the
3933 error, it didn't break the script. It just looked ugly. I fixed this by
3934 always allowing "-C /dev/null" if the caller is root.
3936 14. Ignore overlarge ACL variable number when reading spool file - insurance
3937 against a later release with more variables having written the file.
3939 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3940 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3941 this, as well as the form without the "IPv6" on the front.
3943 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3944 OS/Makefile-Darwin file.
3946 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3947 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3949 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3950 variable. However, some OS use ints for inodes. I've added cast to long int
3951 to get rid of the compiler warning.
3953 19. I had forgotten to lock out "/../" in configuration file names when
3954 ALT_CONFIG_PREFIX was set.
3956 20. Routers used for verification do not need to specify transports. However,
3957 if such a router generated a host list, and callout was configured, Exim
3958 crashed, because it could not find a port number from the (non-existent)
3959 transport. It now assumes port 25 in this circumstance.
3961 21. Added the -t option to exigrep.
3963 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3964 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3965 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3966 them is set, LOOKUP_LSEARCH is forced.
3968 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3970 24. Added sender and host information to the "rejected by local_scan()" log
3971 line; previously there was no indication of these.
3973 25. Added .include_if_exists.
3975 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3976 Linux. It turns out that not all file systems support this. Apparently some
3977 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3978 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3979 not supported on the file descriptor, is now ignored when Exim is trying to
3980 sync a directory. This applies only to Linux.
3982 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3984 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3985 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3988 29. Added support for maildirsize files from supplied patch (modified a bit).
3990 30. The use of :fail: followed by an empty string could lead Exim to respond to
3991 sender verification failures with (e.g.):
3993 550 Verification failed for <xxx>
3994 550 Sender verify failed
3996 where the first response line was missing the '-' that indicates it is not
3997 the final line of the response.
3999 31. The loop for finding the name of the user that called Exim had a hardwired
4000 limit of 10; it now uses the value of finduser_retries, which is used for
4001 all other user lookups.
4003 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4006 33. Exim was neglecting to zero errno before one call of strtol() when
4007 expanding a string and expecting an integer value. On some systems this
4008 resulted in spurious "integer overflow" errors. Also, it was casting the
4009 result into an int without checking.
4011 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4012 did not work. The code looks as if it has *never* worked, though it appears
4013 to have been documented since at least releast 1.62. I have made it work.
4015 35. The "timeout_DNS" error in retry rules, also documented since at least
4016 1.62, also never worked. As it isn't clear exactly what this means, and
4017 clearly it isn't a major issue, I have abolished the feature by treating it
4018 as "timeout", and writing a warning to the main and panic logs.
4020 36. The display of retry rules for -brt wasn't always showing the error code
4023 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4024 timeout_connect_A, timeout_connect_MX.
4026 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4027 to the empty sender.
4029 39. The daemon was not analysing the content of -oX till after it had closed
4030 stderr and disconnected from the controlling terminal. This meant that any
4031 syntax errors were only noted on the panic log, and the return code from
4032 the command was 0. By re-arranging the code a little, I've made the
4033 decoding happen first, so such errors now appear on stderr, and the return
4034 code is 1. However, the actual setting up of the sockets still happens in
4035 the disconnected process, so errors there are still only recorded on the
4038 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4039 connections (as happens on some IP stacks) was logged at start up time as
4040 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4041 it from "IPv6 and IPv4", which means that two separate sockets are being
4044 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4045 says whether AF_INET or AF_INET6 was passed as an argument.
4047 42. Exiwhat output was messed up when time zones were included in log
4050 43. Exiwhat now gives more information about the daemon's listening ports,
4051 and whether -tls-on-connect was used.
4053 44. The "port" option of the smtp transport is now expanded.
4055 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4056 silently ignored. Now an error message is written to the main and panic
4059 46. There's a new ACL modifier called "logwrite" which writes to a log file
4060 as soon as it is encountered.
4062 47. Added $local_user_uid and $local_user_gid at routing time.
4064 48. Exim crashed when trying to verify a sender address that was being
4067 49. Exim was recognizing only a space character after ".include". It now also
4068 recognizes a tab character.
4070 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4071 extracting the relevant information from the specification. The man page no
4072 longer contains scrambled data for the -d option, and I've added a section
4073 at the front about calling Exim under different names.
4075 51. Added "extra_headers" argument to the "mail" command in filter files.
4077 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4080 53. Installed eximstats 1.29.
4082 54. Added transport_filter_timeout as a generic transport option.
4084 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4085 Cc: header lines. This was required by RFC 822, but it not required by RFC
4088 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4089 incoming messages that did not have them. Now it does so only if the
4090 message originates locally, that is, if there is no associated remote host
4091 address. When Resent- header lines are present, this applies to the Resent-
4092 lines rather than the non-Resent- lines.
4094 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4095 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4097 58. Messages for configuration errors now include the name of the main
4098 configuration file - useful now that there may be more than one file in a
4099 list (.included file names were always shown).
4101 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4102 for those rare installations that do not start the daemon as root or run it
4103 setuid root. I've cut out the call to initgroups() if the daemon is not
4106 60. The Exim user and group can now be bound into the binary as text strings
4107 that are looked up at the start of Exim's processing.
4109 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4111 62. Added $mailstore_basename variable.
4113 63. Installed patch to sieve.c from Michael Haardt.
4115 64. When Exim failed to open the panic log after failing to open the main log,
4116 the original message it was trying to log was written to stderr and debug
4117 output, but if they were not available (the usual case in production), it
4118 was lost. Now it is written to syslog before the two lines that record the
4119 failures to open the logs.
4121 65. Users' Exim filters run in subprocesses under the user's uid. It is
4122 possible for a "deliver" command or an alias in a "personal" command to
4123 provoke an address rewrite. If logging of address rewriting is configured,
4124 this fails because the process is not running as root or exim. There may be
4125 a better way of dealing with this, but for the moment (because 4.30 needs
4126 to be released), I have disabled address rewrite logging when running a
4127 filter in a non-root, non-exim process.
4133 1. The buildconfig auxiliary program wasn't quoting the value set for
4134 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4135 not defined. This bug was masked in 4.22 by the effect that was fixed in
4138 2. Some messages that were rejected after a message id was allocated were
4139 shown as "incomplete" by exigrep. It no longer does this for messages that
4140 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4142 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4143 have allow_domain_literals set, the ID did not get logged in the <= line.
4144 Domain literals are now always recognized in Message-ID: header lines.
4146 4. The first argument for a ${extract expansion item is the key name or field
4147 number. Leading and trailing spaces in this item were not being ignored,
4148 causing some misleading effects.
4150 5. When deliver_drop_privilege was set, single queue runner processes started
4151 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4152 same command in the process it spins off) were not dropping privilege.
4154 6. When the daemon running as "exim" started a queue runner, it always
4155 re-executed Exim in the spun-off process. This is a waste of effort when
4156 deliver_drop_privilege is set. The new process now just calls the
4157 queue-runner function directly.
4163 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4166 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4167 ignored. Though the use of -oP was forcing the writing of a pid file, it
4168 was always written to the default place.
4170 3. If the message "no IP address found for host xxxx" is generated during
4171 incoming verification, it is now followed by identification of the incoming
4172 connection (so you can more easily find what provoked it).
4174 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4176 5. Added some features to "harden" Exim a bit more against certain attacks:
4178 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4179 be put in Local/Makefile. This is like the never_users runtime option,
4180 but it cannot be overridden. The default setting is "root".
4182 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4183 prefix string with which any file named in a -C command line option
4186 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4187 is retained for -C and -D only if the caller of Exim is root. Without
4188 it, the exim user may also use -C and -D and retain privilege.
4190 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4191 command line option is disabled.
4193 6. Macro names set by the -D option must start with an upper case letter, just
4194 like macro names defined in the configuration file.
4196 7. Added "dereference=" facility to LDAP.
4198 8. Two instances of the typo "uknown" in the source files are fixed.
4200 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4201 the Configure-Makefile script screwed up while processing it.
4203 10. Incorporated PCRE 4.4.
4205 11. The SMTP synchronization check was not operating right at the start of an
4206 SMTP session. For example, it could not catch a HELO sent before the client
4207 waited for the greeting. There is now a check for outstanding input at the
4208 point when the greeting is written. Because of the duplex, asynchronous
4209 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4210 way, but not yet received, when the check is performed.
4212 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4213 on TCP/IP sockets, because this apparently causes some broken clients to
4216 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4217 unchanged) from the Cygwin maintainer.
4219 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4220 was supported instead of testing for mbx. Effectively a typo.
4222 15. The spa authenticator server code was not checking that the input it
4223 received was valid base64.
4225 16. The debug output line for the "set" modifier in ACLs was not showing the
4226 name of the variable that was being set.
4228 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4230 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4231 lookup on its own. Something else was needed to trigger the lookup. For
4232 example, a match in host_lookup or the need for a host name in a host list.
4233 Now, if $sender_host_name is referenced and the host name has not yet been
4234 looked up, a lookup is performed. If the lookup fails, the variable remains
4235 empty, and $host_lookup_failed is set to "1".
4237 19. Added "eqi" as a case-independent comparison operator.
4239 20. The saslauthd authentication condition could segfault if neither service
4240 nor realm was specified.
4242 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4243 error message that was logged was misleading, and incoming SMTP
4244 connections were dropped. The message is now more accurate, and temporary
4245 errors are given to SMTP connections.
4247 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4248 (except RSET) with a 421 error, until QUIT is received. However, it was
4249 failing to send a response to QUIT.
4251 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4252 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4253 ACL is now run after the helo_try_verify_hosts code.
4255 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4256 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4257 case-independent, so other case variants are also recognized). Apparently
4258 some systems use these upper case variants.
4260 25. If more than two messages were waiting for the same host, and a transport
4261 filter was specified for the transport, Exim sent two messages over the
4262 same TCP/IP connection, and then failed with "socket operation on non-
4263 socket" when it tried to send the third.
4265 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4267 27. The extern definition of crypt16() in expand.c was not being excluded when
4268 the OS had its own crypt16() function.
4270 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4271 as a preferred synonym for return_size_limit, both as an option and as an
4274 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4276 30. Changed the default configuration ACL to relax the local part checking rule
4277 for addresses that are not in any local domains. For these addresses,
4278 slashes and pipe symbols are allowed within local parts, but the sequence
4279 /../ is explicitly forbidden.
4281 31. SPA server authentication was not clearing the challenge buffer before
4284 32. log_message in a "warn" ACL statement was writing to the reject log as
4285 well as to the main log, which contradicts the documentation and doesn't
4286 seem right (because no rejection is happening). So I have stopped it.
4288 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4289 However, I am unable to do any testing of this.
4291 34. Fixed an infelicity in the appendfile transport. When checking directories
4292 for a mailbox, to see if any needed to be created, it was accidentally
4293 using path names with one or more superfluous leading slashes; tracing
4294 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4296 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4297 contents are added to the log line that is written for every discarded
4298 recipient. (Previously a log_message setting was ignored.)
4300 36. The ${quote: operator now quotes the string if it is empty.
4302 37. The install script runs exim in order to find its version number. If for
4303 some reason other than non-existence or emptiness, which it checks, it
4304 could not run './exim', it was installing it with an empty version number,
4305 i.e. as "exim-". This error state is now caught, and the installation is
4308 38. An argument was missing from the function that creates an error message
4309 when Exim fails to connect to the socket for saslauthd authentication.
4310 This could cause Exim to crash, or give a corrupted message.
4312 39. Added isip, isip4, and isip6 to ${if conditions.
4314 40. The ACL variables $acl_xx are now saved with the message, and can be
4315 accessed later in routers, transports, and filters.
4317 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4318 strings in the file are not string-expanded.
4320 42. If a MAIL command specified a SIZE value that was too large to fit into an
4321 int variable, the check against message_size_limit failed. Such values are
4322 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4323 day this will have to be increased, but I don't think I want to be around
4324 when emails are that large.
4331 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4332 iconv() is not standard in FreeBSD.
4334 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4335 IPv6 enabled. The observed symptom was a segmentation fault on return from
4336 the function os_common_find_running_interfaces() in src/os.c.
4338 3. In the check_special_case() function in daemon.c I had used "errno" as an
4339 argument name, which causes warnings on some systems. This was basically a
4340 typo, since it was named "eno" in the comments!
4342 4. The code that waits for the clock to tick (at a resolution of some fraction
4343 of a second) so as to ensure message-id uniqueness was always waiting for
4344 at least one whole tick, when it could have waited for less. [This is
4345 almost certainly not relevant at current processor speeds, where it is
4346 unlikely to ever wait at all. But we try to future-proof.]
4348 5. The function that sleeps for a time interval that includes fractions of a
4349 second contained a race. It did not block SIGALRM between setting the
4350 timer, and suspending (a couple of lines later). If the interval was short
4351 and the sigsuspend() was delayed until after it had expired, the suspension
4352 never ended. On busy systems this could lead to processes getting stuck for
4355 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4356 process, before it forks any delivery processes. The open lookup caching
4357 mechanism meant that the open file or database connection was passed into
4358 the delivery process. The problem was that delivery processes always tidy
4359 up cached lookup data. This could cause a problem for the next delivery
4360 process started by the queue runner, because the external queue runner
4361 process does not know about the closure. So the next delivery process
4362 still has data in the lookup cache. In the case of a file lookup, there was
4363 no problem because closing a file descriptor in a subprocess doesn't affect
4364 the parent. However, if the lookup was caching a connection to a database,
4365 the connection was closed, and the second delivery process was likely to
4366 see errors such as "PGSQL: query failed: server closed the connection
4367 unexpectedly". The problem has been fixed by closing all cached lookups
4368 in a queue runner before running a delivery process.
4370 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4371 seem to have the "const" qualifier which it has on other OS. I've
4374 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4375 *of the same type* (client or server) with the same public name that an
4376 error should be diagnosed.
4378 9. When Exim looked up a host name for an IP address, but failed to find the
4379 original IP address when looking up the host name (a safety check), it
4380 output the message "<ip address> does not match any IP for NULL", which was
4381 confusing, to say the least. The bug was that the host name should have
4382 appeared instead of "NULL".
4384 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4385 user that is built into the binary, and the -C or -D options is used, root
4386 privilege is dropped before the configuration file is read. In addition,
4387 logging is switched to stderr instead of the normal log files. If the
4388 configuration then re-defines the Exim user, the unprivileged environment
4389 is probably not what is expected, so Exim logs a panic warning message (but
4392 However, if deliver_drop_privilege is set, the unprivileged state may well
4393 be exactly what is intended, so the warning has been cut out in that case,
4394 and Exim is allowed to try to write to its normal log files.
4400 1. smtp_return_error_details was not giving details for temporary sender
4401 or receiver verification errors.
4403 2. Diagnose a configuration error if two authenticators have the same public
4406 3. Exim used not to create the message log file for a message until the first
4407 delivery attempt. This could be confusing when incoming messages were held
4408 for policy or load reasons. The message log file is now created at the time
4409 the message is received, and an initial "Received" line is written to it.
4411 4. The automatically generated man page for command line options had a minor
4412 bug that caused no ill effects; however, a more serious problem was that
4413 the procedure for building the man page automatically didn't always
4414 operate. Consequently, release 4.20 contains an out-of-date version. This
4415 shouldn't happen again.
4417 5. When building Exim with embedded Perl support, the script that builds the
4418 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4419 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4421 6. The freeze_tell option was not being used for messages that were frozen on
4422 arrival, either by an ACL or by local_scan().
4424 7. Added the smtp_incomplete_transaction log selector.
4426 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4427 was accepting AUTH without a new EHLO.
4429 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4430 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4433 10. Logging of TCP/IP connections (when configured) now happens in the main
4434 daemon process instead of the child process, so that the TCP/IP connection
4435 count is more accurate (but it can never be perfect).
4437 11. The use of "drop" in a nested ACL was not being handled correctly in the
4438 outer ACL. Now, if condition failure induced by the nested "drop" causes
4439 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4440 or "require"), the connection is dropped.
4442 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4443 that yield "discard" can now be used with an "accept" or a "discard" verb,
4444 but an error is generated for any others (because I can't see a useful way
4445 to define what should happen).
4447 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4448 are now processed in the same way as lines in the Exim configuration file.
4449 In particular, continuation lines are supported.
4451 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4453 15. Added -ti meaning -t -i.
4455 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4456 domains, and warn by logging if others are found.
4458 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4459 returned from GIFCONF when Exim is trying to find the list of interfaces on
4460 a host. The code in os.c has been modified to copy each ifreq to an aligned
4461 structure in all cases.
4463 Also, in some cases, the returned ifreq's were being copied to a 'struct
4464 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4465 means the last couple of bytes of an IPv6 address could be chopped if the
4466 ifreq contained only a normal sockaddr (14 bytes storage).
4468 18. Named domain lists were not supported in the hosts_treat_as_local option.
4469 An entry such as +xxxx was not recognized, and was treated as a literal
4472 19. Ensure that header lines added by a DATA ACL are included in the reject log
4473 if the ACL subsequently rejects the message.
4475 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4476 MD5 (which is deprecated).
4478 21. When testing a filter file using -bf, Exim was writing a message when it
4479 took the sender from a "From " line in the message, but it was not doing so
4480 when it took $return_path from a Return-Path: header line. It now does.
4482 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4483 with a valid header line field name (a series of printing characters
4484 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4486 23. Changed "disc" in the source to "disk" to conform to the documentation and
4487 the book and for uniformity.
4489 24. Ignore Sendmail's -Ooption=value command line item.
4491 25. When execve() failed while trying to run a command in a pipe transport,
4492 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4493 could be confused with a return value of 69 from the command itself. This
4494 has been changed to 127, the value the shell returns if it is asked to run
4495 a non-existent command. The wording for the related log line suggests a
4496 non-existent command as the problem.
4498 26. If received_header_text expands to an empty string, do not add a Received:
4499 header line to the message. (Well, it adds a token one on the spool, but
4500 marks it "old" so that it doesn't get used or transmitted.)
4502 27. Installed eximstats 1.28 (addition of -nt option).
4504 28. There was no check for failure on the call to getsockname() in the daemon
4505 code. This can fail if there is a shortage of resources on the system, with
4506 ENOMEM, for example. A temporary error is now given on failure.
4508 29. Contrary to the C standard, it seems that in some environments, the
4509 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4510 program. Exim now does this explicitly; it affects the formatting of
4511 timestamps using strftime().
4513 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4514 complaints. I've now initialized all the variables, to avoid this.
4516 32. Header lines added by a system filter were not being "seen" during
4517 transport-time rewrites.
4519 33. The info_callback() function passed to OpenSSL is set up with type void
4520 (*)(SSL *, int, int), as described somewhere. However, when calling the
4521 function (actually a macro) that sets it up, the type void(*)() is
4522 expected. I've put in a cast to prevent warnings from picky compilers.
4524 34. If a DNS black list lookup found a CNAME record, but there were no A
4525 records associated with the domain it pointed at, Exim crashed.
4527 35. If a DNS black list lookup returned more than one A record, Exim ignored
4528 all but the first. It now scans all returned addresses if a particular IP
4529 value is being sought. In this situation, the contents of the
4530 $dnslist_value variable are a list of all the addresses, separated by a
4533 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4534 to accept a host name and all its aliases if the forward lookup for any of
4535 them yielded the IP address of the incoming connection. Now it accepts only
4536 those names whose forward lookup yields the correct IP address. Any other
4537 names are discarded. This closes a loophole whereby a rogue DNS
4538 administrator could create reverse DNS records to break through a
4539 wildcarded host restriction in an ACL.
4541 37. If a user filter or a system filter that ran in a subprocess used any of
4542 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4543 the wrong values were passed to the pipe command ($thisaddress had the
4544 value of $0, $0 had the value of $1, etc). This bug was introduced by
4545 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4547 38. Improved the line breaking for long SMTP error messages from ACLs.
4548 Previously, if there was no break point between 40 and 75 characters, Exim
4549 left the rest of the message alone. Two changes have been made: (a) I've
4550 reduced the minimum length to 35 characters; (b) if it can't find a break
4551 point between 35 and 75 characters, it looks ahead and uses the first one
4552 that it finds. This may give the occasional overlong line, but at least the
4553 remaining text gets split now.
4555 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4556 file descriptors might be low, and that setting 1000 would always raise it.
4557 It turns out that in some environments, the limit is already over 1000 and
4558 that lowering it causes trouble. So now Exim takes care not to decrease it.
4560 40. When delivering a message, the value of $return_path is set to $sender_
4561 address at the start of routing (routers may change the value). By an
4562 oversight, this default was not being set up when an address was tested by
4563 -bt or -bv, which affected the outcome if any router or filter referred to
4566 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4567 log only when "message" or "log_message" are set. However, if one of the
4568 conditions was an address verification, or a call to a nested ACL, the
4569 messages generated by the underlying test were being passed through. This
4570 no longer happens. The underlying message is available in $acl_verify_
4571 message for both "message" and "log_message" expansions, so it can be
4572 passed through if needed.
4574 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4575 new expansion $bh_ to give the encoded byte string without charset
4576 translation. Translation happens only if iconv() is available; HAVE_ICONV
4577 indicates this at build time. HEADERS_CHARSET gives the charset to
4578 translate to; headers_charset can change it in the configuration, and
4579 "headers charset" can change it in an individual filter file.
4581 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4582 that creates RFC 2047 encoded "words" labels them as that charset instead
4583 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4584 expansion operator; (ii) when Exim creates a From: line for a local
4585 message; (iii) when a header line is rewritten to include a "phrase" part.
4587 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4588 buggy, causing it to skip the first lines of messages whose message ID
4589 ended in 'D'. This would not have bitten before Exim release 4.14, because
4590 message IDs were unlikely to end in 'D' before then. The effect was to have
4591 incorrect size information for certain domains.
4593 45. #include "config.h" was missing at the start of the crypt16.c module. This
4594 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4597 46. If there was a timeout during a "random" callout check, Exim treated it as
4598 a failure of the random address, and carried on sending RSET and the real
4599 address. If the delay was just some slowness somewhere, the response to the
4600 original RCPT would be taken as a response to RSET and so on, causing
4601 mayhem of various kinds.
4603 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4604 when I implemented it. It didn't allow for the fact that some option values
4605 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4606 the right test for positive values.
4608 48. Domain names in DNS records are case-independent. Exim always looks them up
4609 in lower case. Some resolvers return domain names in exactly the case they
4610 appear in the zone file, that is, they may contain uppercase letters. Not
4611 all resolvers do this - some return always lower case. Exim was treating a
4612 change of case by a resolver as a change of domain, similar to a widening
4613 of a domain abbreviation. This triggered its re-routing code and so it was
4614 trying to route what was effectively the same domain again. This normally
4615 caused routing to fail (because the router wouldn't handle the domain
4616 twice). Now Exim checks for this case specially, and just changes the
4617 casing of the domain that it ultimately uses when it transmits the message
4620 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4623 50. If a filter generated a file delivery with a non-absolute name (possible if
4624 no home directory exists for the router), the forbid_file option was not
4627 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4628 the existing equality matching.
4630 52. Exim was using ints instead of ino_t variables in some places where it was
4631 dealing with inode numbers.
4633 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4634 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4635 called TMPDIR, and if it finds it is different, it changes its value.
4637 54. The smtp_printf() function is now made available to local_scan() so
4638 additional output lines can be written before returning. There is also an
4639 smtp_fflush() function to enable the detection of a dropped connection.
4640 The variables smtp_input and smtp_batched_input are exported to
4643 55. Changed the default runtime configuration: the message "Unknown user"
4644 has been removed from the ACL, and instead placed on the localuser router,
4645 using the cannot_route_message feature. This means that any verification
4646 failures that generate their own messages won't get overridden. Similarly,
4647 the "Unrouteable address" message that was in the ACL for unverifiable
4648 relay addresses has also been removed.
4650 56. Added hosts_avoid_esmtp to the smtp transport.
4652 57. The exicyclog script was not checking for the esoteric option
4653 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4654 will work only if exicyclog is run under the appropriate euid.
4656 58. Following a discussion on the list, the rules by which Exim recognises line
4657 endings on incoming messages have been changed. The -dropcr and drop_cr
4658 options are now no-ops, retained only for backwards compatibility. The
4659 following line terminators are recognized: LF CRLF CR. However, special
4660 processing applies to CR:
4662 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4663 nor a local message in the state where . is a terminator.
4665 (ii) If a bare CR is encountered in a header line, an extra space is added
4666 after the line terminator so as not to end the header. The reasoning
4667 behind this is that bare CRs in header lines are most likely either
4668 to be mistakes, or people trying to play silly games.
4670 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4671 was being incorrectly given as 18 bytes larger than it should have been.
4672 This is a VOB (very old bug).
4674 60. This may never have affected anything current, but just in case it has:
4675 When the local host is found other than at the start of a list of hosts,
4676 the local host, those with the same MX, and any that follow, are discarded.
4677 When the list in question was part of a longer list of hosts, the following
4678 hosts (not currently being processed) were also being discarded. This no
4679 longer happens. I'm not sure if this situation could ever has previously
4682 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4685 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4686 header. This is something that is recommended in a new Internet Draft, and
4687 is something that is documented as being done by Sendmail. There are two
4688 possible values. For messages generated by the autoreply transport, Exim
4691 Auto-Submitted: auto-replied
4693 whereas for all other generated messages (e.g. bounces) it adds
4695 Auto-Submitted: auto-generated
4697 63. The "personal" condition in filters now includes a test for the
4698 Auto-Submitted: header. If it contains the string "auto-" the message it
4699 not considered personal.
4701 64. Added rcpt_include_affixes as a generic transport option.
4703 65. Added queue_only_override (default true).
4705 66. Added the syslog_duplication option.
4707 67. If what should have been the first header line of a message consisted of
4708 a space followed by a colon, Exim was mis-interpreting it as a header line.
4709 It isn't of course - it is syntactically invalid and should therefore be
4710 treated as the start of the message body. The misbehaviour could have
4711 caused a number of strange effects, including loss of data in subsequent
4712 header lines, and spool format errors.
4714 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4715 client host had authenticated. This control can now be exercised by an ACL
4716 for more flexibility.
4718 69. By default, callouts do not happen when testing with -bh. There is now a
4719 variant, -bhc, which does actually run the callout code, including
4720 consulting and updating the callout cache.
4722 70. Added support for saslauthd authentication, courtesy of Alexander
4725 71. If statvfs() failed on the spool or log directories while checking their
4726 size for availability, Exim confusingly gave the error "space shortage".
4727 Furthermore, in debugging mode it crashed with a floating point exception.
4728 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4729 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4730 serious problem, Exim now writes to the main and panic logs when this
4731 happens, with details of the failure. It then refuses to accept the
4732 incoming message, giving the message "spool directory problem" or "log
4733 directory problem" with a 421 code for SMTP messages.
4735 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4736 0, 1, and 2 exist, because some OS complain for execs without them (see
4737 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4738 descriptors. However, the code omitted to check that the open succeeded,
4739 causing mysterious errors if for some reason the permissions on /dev/null
4740 got screwed. Now Exim writes a message to the main and panic logs, and
4741 bombs out if it can't open /dev/null.
4743 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4744 interact so that it is all more flexible. It is supposed to remain
4745 backwards compatible. Also added extra_local_interfaces.
4747 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4748 to bomb out with an assertion failure - to the client this appears as a
4749 connection drop. This problem occurs in the part of the code that was taken
4750 from the Samba project. Fortunately, the assertion is in a very simple
4751 function, so I have fixed this by reproducing the function inline in the
4752 one place where it is called, and arranging for authentication to fail
4753 instead of killing the process with assert().
4755 75. The SPA client code was not working when the server requested OEM rather
4756 than Unicode encoding.
4758 76. Added code to make require_files with a specific uid setting more usable in
4759 the case where statting the file as root fails - usually a non-root-mounted
4760 NFS file system. When this happens and the failure is EACCES, Exim now
4761 forks a subprocess and does the per-uid checking as the relevant uid.
4763 77. Added process_log_path.
4765 78. If log_file_path was not explicitly set, a setting of check_log_space or
4766 check_log_inodes was ignored.
4768 79. If a space check for the spool or log partitions fails, the incident is now
4769 logged. Of course, in the latter case the data may get lost...
4771 80. Added the %p formatting code to string_format() so that it can be used to
4772 print addresses in debug_print(). Adjusted all the address printing in the
4773 debugging in store.c to use %p rather than %d.
4775 81. There was a concern that a line of code in smtp_in.c could overflow a
4776 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4777 initially expressed, the concern was not well-founded, because trailing
4778 spaces are removed early. However, if the trailing spaces were followed by
4779 a NULL, they did not get removed, so the overflow was possible. Two fixes
4782 (a) I re-wrote the offending code in a cleaner fashion.
4783 (b) If an incoming SMTP command contains a NULL character, it is rejected
4786 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4787 runs initgroups(), so that if the Exim user is in any additional groups,
4788 they will be used during message reception.
4794 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob