1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.513 2007/06/18 13:57:49 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.
64 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
65 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
68 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
71 PH/02 In an ACL statement such as
73 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
75 if a client was not listed at all, or was listed with a value other than
76 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
77 the condition was not true (as it should be), so access was not denied.
78 The bug was that the ! inversion was incorrectly passed on to the second
79 item. This has been fixed.
81 PH/03 Added additional dnslists conditions == and =& which are different from
82 = and & when the dns lookup returns more than one IP address.
84 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
85 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
87 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
88 FSYNC, which compiles an option called disable_fsync that allows for
89 bypassing fsync(). The documentation is heavily laced with warnings.
91 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
93 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
94 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
95 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
96 including adding "make clean"; (3) Added -fPIC when compiling the test
97 dynamically loaded module, to get rid of a warning.
99 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
100 message fails, move_frozen_messages = true and ignore_bounce_errors_after
101 = 0s. The bug is otherwise harmless.
103 PH/07 There was a bug in the dovecot authenticator such that the value of
104 $auth1 could be overwritten, and so not correctly preserved, after a
105 successful authentication. This usually meant that the value preserved by
106 the server_setid option was incorrect.
108 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
110 PH/09 Installed PCRE release 7.0.
112 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
113 run for batched SMTP input. It is now run at the start of every message
114 in the batch. While fixing this I discovered that the process information
115 (output by running exiwhat) was not always getting set for -bs and -bS
116 input. This is fixed, and it now also says "batched" for BSMTP.
118 PH/11 Added control=no_pipelining.
120 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
121 patch, slightly modified), and move the expansion of helo_data till after
122 the connection is made in the smtp transport (so it can use these
125 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
127 PH/14 Added log_selector = +pid.
129 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
131 PH/16 Add ${if forany and ${if forall.
133 PH/17 Added dsn_from option to vary the From: line in DSNs.
135 PH/18 Flush SMTP output before performing a callout, unless control =
136 no_callout_flush is set.
138 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
139 was true (the default) a successful delivery failed to delete the retry
140 item, thus causing premature timeout of the address. The bug is now
143 PH/20 Added hosts_avoid_pipelining to the smtp transport.
145 PH/21 Long custom messages for fakedefer and fakereject are now split up
146 into multiline reponses in the same way that messages for "deny" and
147 other ACL rejections are.
149 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
150 with slight modification.
152 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
153 draft, changing the syntax and factoring some duplicate code".
155 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
156 for deliveries of the second and subsequent messages over the same SMTP
159 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
160 ${reduce, with only minor "tidies".
162 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
164 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
165 expansion side effects.
167 PH/27 When a message times out after an over-quota error from an Exim-imposed
168 quota, the bounce message says "mailbox is full". This message was not
169 being given when it was a system quota that was exceeded. It now should
172 MH/03 Made $recipients available in local_scan(). local_scan() already has
173 better access to the recipient list through recipients_list[], but
174 $recipients can be useful in postmaster-provided expansion strings.
176 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
177 in the case of a MAIL command with additional options following the
178 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
179 were accidentally chopped off.
181 PH/29 SMTP synchronization checks are implemented when a command is read -
182 there is a check that no more input is waiting when there shouldn't be
183 any. However, for some commands, a delay in an ACL can mean that it is
184 some time before the response is written. In this time, more input might
185 arrive, invalidly. So now there are extra checks after an ACL has run for
186 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
187 pipelining has not been advertised.
189 PH/30 MH's patch to allow iscntrl() characters to be list separators.
191 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
192 returned in the SMTP response when smtp_return_error_details was false.
195 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
196 instead of the C I/O that was originally supplied, because problems were
199 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
200 Exim which did not show up earlier: it was assuming that a call to
201 SSL_CTX_set_info_callback() might give an error value. In fact, there is
202 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
203 was a macro that became an assignment, so it seemed to work. This has
204 changed to a proper function call with a void return, hence the compile
205 error. Exim's code has been fixed.
207 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
210 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
212 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
214 PH/37 If a message is not accepted after it has had an id assigned (e.g.
215 because it turns out to be too big or there is a timeout) there is no
216 "Completed" line in the log. When some messages of this type were
217 selected by exigrep, they were listed as "not completed". Others were
218 picked up by some special patterns. I have improved the selection
219 criteria to be more general.
221 PH/38 The host_find_failed option in the manualroute router can now be set
222 to "ignore", to completely ignore a host whose IP address cannot be
223 found. If all hosts are ignored, the behaviour is controlled by the new
224 host_all_ignored option.
226 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
227 homing or because of multiple MX records with /mx) generated more than
228 one IP address, and the following item turned out to be the local host,
229 all the secondary addresses of the first item were incorrectly removed
230 from the list, along with the local host and any following hosts (which
231 is what is supposed to happen).
233 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
234 whoever called Exim into the -H file. In the case of the daemon it was
235 behaving confusingly. When first started, it used values for whoever
236 started the daemon, but after a SIGHUP it used the Exim user (because it
237 calls itself on a restart). I have changed the code so that it now always
240 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
241 message are rejected with the same error (e.g. no authentication or bad
242 sender address), and a DATA command is nevertheless sent (as can happen
243 with PIPELINING or a stupid MUA), the error message that was given to the
244 RCPT commands is included in the rejection of the DATA command. This is
245 intended to be helpful for MUAs that show only the final error to their
248 PH/42 Another patch from the Sieve maintainer.
250 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
251 Eximstats - Fixed some broken HTML links and added missing column headers
253 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
254 columns for Rejects, Temp Rejects, Ham, and Spam rows.
256 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
258 PH/43 Yet another patch from the Sieve maintainer.
260 PH/44 I found a way to check for a TCP/IP connection going away before sending
261 the response to the final '.' that terminates a message, but only in the
262 case where the client has not sent further data following the '.'
263 (unfortunately, this is allowed). However, in many cases there won't be
264 any further data because there won't be any more messages to send. A call
265 to select() can be used: if it shows that the input is "ready", there is
266 either input waiting, or the socket has been closed. An attempt to read
267 the next input character can distinguish the two cases. Previously, Exim
268 would have sent an OK response which the client would never have see.
269 This could lead to message repetition. This fix should cure that, at
270 least in a lot of common cases.
272 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
273 advertised in response to EHLO.
279 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
280 fixed by 4.65/MH/01 (is this a record?) are fixed:
282 (i) An empty string was always treated as zero by the numeric comparison
283 operators. This behaviour has been restored.
285 (ii) It is documented that the numeric comparison operators always treat
286 their arguments as decimal numbers. This was broken in that numbers
287 starting with 0 were being interpreted as octal.
289 While fixing these problems I realized that there was another issue that
290 hadn't been noticed. Values of message_size_limit (both the global option
291 and the transport option) were treated as octal if they started with 0.
292 The documentation was vague. These values are now always treated as
293 decimal, and I will make that clear in the documentation.
299 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
300 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
303 MH/01 Don't check that the operands of numeric comparison operators are
304 integers when their expansion is in "skipping" mode (fixes bug
305 introduced by 4.64-PH/07).
307 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
308 child addresses, Exim now panics and dies. Previously, because the count
309 is held in a short int, deliveries were likely to be lost. As such a
310 large number of recipients for a single message is ridiculous
311 (performance will be very, very poor), I have chosen to impose a limit
312 rather than extend the field.
318 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
319 leftover -K file (the existence of which was triggered by #402).
320 While we were at it, introduced process PID as part of the -K
321 filename. This should rule out race conditions when creating
324 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
325 processing considerably. Previous code took too long for large mails,
326 triggering a timeout which in turn triggers #401.
328 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
329 in the DK code in transports.c. sendfile() is not really portable,
330 hence the _LINUX specificness.
332 TF/01 In the add_headers option to the mail command in an Exim filter,
333 there was a bug that Exim would claim a syntax error in any
334 header after the first one which had an odd number of characters
337 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
338 callout verification, Exim cached a "reject" for the entire domain. This
339 is correct for most verifications, but it is not correct for a recipient
340 verification with use_sender or use_postmaster set, because in that case
341 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
342 case of MAIL FROM:<> rejection from other early rejections (e.g.
343 rejection of HELO). When verifying a recipient using a non-null MAIL
344 address, the cache is ignored if it shows MAIL FROM:<> rejection.
345 Whatever the result of the callout, the value of the domain cache is
346 left unchanged (for any other kind of callout, getting as far as trying
347 RCPT means that the domain itself is ok).
349 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
352 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
353 manifest itself as EPIPE rather than ECONNECT. When tidying away a
354 session, the daemon ignores ECONNECT errors and logs others; it now
355 ignores EPIPE as well.
357 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
358 (quoted-printable decoding).
360 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
361 later the small subsequent patch to fix an introduced bug.
363 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
365 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
367 PH/08 An error is now given if message_size_limit is specified negative.
369 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
370 to be given (somewhat) arbitrary names.
372 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
375 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
376 miscellaneous code fixes
378 PH/10 Added the log_reject_target ACL modifier to specify where to log
381 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
382 hostname. This is wrong, because it relates to the incoming message (and
383 probably the interface on which it is arriving) and not to the outgoing
384 callout (which could be using a different interface). This has been
385 changed to use the value of the helo_data option from the smtp transport
386 instead - this is what is used when a message is actually being sent. If
387 there is no remote transport (possible with a router that sets up host
388 addresses), $smtp_active_hostname is used.
390 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
391 tweaks were necessary in order to get it to work (see also 21 below):
392 (a) The code assumed that strncpy() returns a negative number on buffer
393 overflow, which isn't the case. Replaced with Exim's string_format()
395 (b) There were several signed/unsigned issues. I just did the minimum
396 hacking in of casts. There is scope for a larger refactoring.
397 (c) The code used strcasecmp() which is not a standard C function.
398 Replaced with Exim's strcmpic() function.
399 (d) The code set only $1; it now sets $auth1 as well.
400 (e) A simple test gave the error "authentication client didn't specify
401 service in request". It would seem that Dovecot has changed its
402 interface. Fortunately there's a specification; I followed it and
403 changed what the client sends and it appears to be working now.
405 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
408 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
409 address A is aliased to B and C, where B exists and C does not. Without
410 -v the output is "A verified" because verification stops after a
411 successful redirection if more than one address is generated. However,
412 with -v the child addresses are also verified. Exim was outputting "A
413 failed to verify" and then showing the successful verification for C,
414 with its parentage. It now outputs "B failed to verify", showing B's
415 parentage before showing the successful verification of C.
417 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
418 look up a TXT record in a specific list after matching in a combined
421 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
422 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
423 they consult the DNS. I had assumed they would set it the way they
424 wanted; and indeed my experiments on Linux seem to show that in some
425 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
426 To be on the safe side, however, I have now made the interface to
427 host_find_byname() similar to host_find_bydns(), with an argument
428 containing the DNS resolver options. The host_find_byname() function now
429 sets these options at its start, just as host_find_bydns() does. The smtp
430 transport options dns_qualify_single and dns_search_parents are passed to
431 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
432 of host_find_byname() use the default settings of RES_DEFNAMES
433 (qualify_single) but not RES_DNSRCH (search_parents).
435 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
436 spool_read_header() do less string testing, by means of a preliminary
437 switch on the second character of optional "-foo" lines. (This is
438 overdue, caused by the large number of possibilities that now exist.
439 Originally there were few.) While I was there, I also converted the
440 str(n)cmp tests so they don't re-test the leading "-" and the first
441 character, in the hope this might squeeze out yet more improvement.
443 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
444 flag allowing group syntax was set by the header_syntax check but not
445 turned off, possible causing trouble later; (2) The flag was not being
446 set at all for the header_verify test, causing "group"-style headers to
447 be rejected. I have now set it in this case, and also caused header_
448 verify to ignore an empty address taken from a group. While doing this, I
449 came across some other cases where the code for allowing group syntax
450 while scanning a header line wasn't quite right (mostly, not resetting
451 the flag correctly in the right place). These bugs could have caused
452 trouble for malformed header lines. I hope it is now all correct.
454 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
455 with the "reply" argument non-NULL. The code, however (which originally
456 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
457 but it didn't always do it. This confused somebody who was copying the
458 code for some other use. I have removed all the tests.
460 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
461 feature that was used to support insecure browsers during the U.S. crypto
462 embargo. It requires special client support, and Exim is probably the
463 only MTA that supported it -- and would never use it because real RSA is
464 always available. This code has been removed, because it had the bad
465 effect of slowing Exim down by computing (never used) parameters for the
466 RSA_EXPORT functionality.
468 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
469 authenticator to fail if there's a tab character in the incoming data
470 (there should never be unless someone is messing about, as it's supposed
471 to be base64-encoded). Also added, on Timo's advice, the "secured" option
472 if the connection is using TLS or if the remote IP is the same as the
473 local IP, and the "valid-client-cert option" if a client certificate has
476 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
477 authenticators. This can be used for authorization after authentication
478 succeeds. (In the case of plaintext, it servers for both authentication
481 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
482 if any retry times were supplied.
484 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
485 connection, where there is no client IP address to check. In this
486 situation, the verify now always succeeds.
488 PH/25 Applied John Jetmore's -Mset patch.
490 PH/26 Added -bem to be like -Mset, but loading a message from a file.
492 PH/27 In a string expansion for a processed (not raw) header when multiple
493 headers of the same name were present, leading whitespace was being
494 removed from all of them, but trailing whitespace was being removed only
495 from the last one. Now trailing whitespace is removed from each header
496 before concatenation. Completely empty headers in a concatenation (as
499 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
500 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
502 PH/29 [Removed. This was a change that I later backed out, and forgot to
503 correct the ChangeLog entry (that I had efficiently created) before
504 committing the later change.]
506 PH/30 Exim was sometimes attempting to deliver messages that had suffered
507 address errors (4xx response to RCPT) over the same connection as other
508 messages routed to the same hosts. Such deliveries are always "forced",
509 so retry times are not inspected. This resulted in far too many retries
510 for the affected addresses. The effect occurred only when there were more
511 hosts than the hosts_max_try setting in the smtp transport when it had
512 the 4xx errors. Those hosts that it had tried were not added to the list
513 of hosts for which the message was waiting, so if all were tried, there
514 was no problem. Two fixes have been applied:
516 (i) If there are any address or message errors in an SMTP delivery, none
517 of the hosts (tried or untried) are now added to the list of hosts
518 for which the message is waiting, so the message should not be a
519 candidate for sending over the same connection that was used for a
520 successful delivery of some other message. This seems entirely
521 reasonable: after all the message is NOT "waiting for some host".
522 This is so "obvious" that I'm not sure why it wasn't done
523 previously. Hope I haven't missed anything, but it can't do any
524 harm, as the worst effect is to miss an optimization.
526 (ii) If, despite (i), such a delivery is accidentally attempted, the
527 routing retry time is respected, so at least it doesn't keep
528 hammering the server.
530 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
531 in ${readsocket because some servers need this prod.
533 PH/32 Added some extra debug output when updating a wait-xxx database.
535 PH/33 The hint "could be header name not terminated by colon", which has been
536 given for certain expansion errors for a long time, was not being given
537 for the ${if def:h_colon_omitted{... case.
539 PH/34 The spec says: "With one important exception, whenever a domain list is
540 being scanned, $domain contains the subject domain." There was at least
541 one case where this was not true.
543 PH/35 The error "getsockname() failed: connection reset by peer" was being
544 written to the panic log as well as the main log, but it isn't really
545 panic-worthy as it just means the connection died rather early on. I have
546 removed the panic log writing for the ECONNRESET error when getsockname()
549 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
550 runs only) independently of the message's sender address. This meant
551 that, if the 4xx error was in fact related to the sender, a different
552 message to the same recipient with a different sender could confuse
553 things. In particualar, this can happen when sending to a greylisting
554 server, but other circumstances could also provoke similar problems.
555 I have changed the default so that the retry time for these errors is now
556 based a combination of the sender and recipient addresses. This change
557 can be overridden by setting address_retry_include_sender=false in the
560 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
561 remote server are returned as part of bounce messages. This was not
562 happening for LMTP over a pipe (the lmtp transport), but now it is the
563 same for both kinds of LMTP.
565 PH/38 Despite being documented as not happening, Exim was rewriting addresses
566 in header lines that were in fact CNAMEs. This is no longer the case.
568 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
569 and queue runs started by the daemon processed all messages. This has
570 been fixed so that -R and -S can now usefully be given with -q<time>.
572 PH/40 Import PCRE release 6.7 (fixes some bugs).
574 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
576 PH/42 Give an error if -q is specified more than once.
578 PH/43 Renamed the variables $interface_address and $interface_port as
579 $received_ip_address and $received_port, to make it clear that these
580 values apply to message reception, and not to the outgoing interface when
581 a message is delivered. (The old names remain recognized, of course.)
583 PH/44 There was no timeout on the connect() call when using a Unix domain
584 socket in the ${readsocket expansion. There now is.
586 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
587 be meaningful with "accept".
589 SC/01 Eximstats V1.43
590 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
592 SC/02 Eximstats V1.44
593 Use a glob alias rather than an array ref in the generated
594 parser. This improves both readability and performance.
596 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
597 Collect SpamAssassin and rejection statistics.
598 Don't display local sender or destination tables unless
599 there is data to show.
600 Added average volumes into the top table text output.
602 SC/04 Eximstats V1.46
603 Collect data on the number of addresses (recipients)
604 as well as the number of messages.
606 SC/05 Eximstats V1.47
607 Added 'Message too big' to the list of mail rejection
608 reasons (thanks to Marco Gaiarin).
610 SC/06 Eximstats V1.48
611 Mainlog lines which have GMT offsets and are too short to
612 have a flag are now skipped.
614 SC/07 Eximstats V1.49 (Alain Williams)
615 Added the -emptyok flag.
617 SC/08 Eximstats V1.50
618 Fixes for obtaining the IP address from reject messages.
620 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
621 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
622 whitesspace changes from 4.64-PH/27
624 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
627 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
628 are found, allow negative numbers in numeric criteria)
630 JJ/06 exipick.20061117.2, added new $message_body_missing variable
632 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
633 to match changes made in 4.64-PH/43
635 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
637 PH/47 Put in an explicit test for a DNS lookup of an address record where the
638 "domain" is actually an IP address, and force a failure. This locks out
639 those revolvers/nameservers that support "A-for-A" lookups, in
640 contravention of the specifications.
642 PH/48 When a host name was looked up from an IP address, and the subsequent
643 forward lookup of the name timed out, the host name was left in
644 $sender_host_name, contrary to the specification.
646 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
647 restricted to single-key lookups, Exim was not diagnosing an error if
648 * or *@ was used with a query-style lookup.
650 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
652 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
653 long ago, but noone interested enough thought of it. Let's just say that
654 the "1.1" means that there are some new functions that weren't there at
655 some point in the past.
657 PH/51 Error processing for expansion failure of helo_data from an smtp
658 transport during callout processing was broken.
660 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
661 tested/used via the -bh/-bhc/-bs options.
663 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
664 bug, fixed in subsequent PCRE releases).
666 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
667 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
669 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
675 SC/01 Use a glob alias rather than an array ref in eximstats generated
676 parser. This improves both readability and performance.
678 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
679 Don't display local sender or destination tables in eximstats unless
680 there is data to show.
681 Added average volumes into the eximstats top table text output.
683 SC/03 Collect data on the number of addresses (recipients) as well
684 as the number of messages in eximstats.
686 TF/01 Correct an error in the documentation for the redirect router. Exim
687 does (usually) call initgroups() when daemonizing.
689 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
690 with consistent privilege compared to when running as a daemon.
692 TF/03 Note in the spec that $authenticated_id is not set for local
693 submissions from trusted users.
695 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
696 Thanks to Dean Brooks <dean@iglou.com> for the patch.
698 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
699 by adding some example configuration directives to the default
700 configuration file. A little bit of work is required to uncomment the
701 directives and define how usernames and passwords are checked, but
702 there is now a framework to start from.
704 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
705 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
706 without this. I don't know how relevant this is to other LDAP libraries.
708 PH/02 Add the verb name to the "unknown ACL verb" error.
710 PH/03 Magnus Holmgren's patch for filter_prepend_home.
712 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
714 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
715 directory not expanded when it should be if an expanded home directory
716 was set for the address (which is overridden by the transport).
718 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
721 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
722 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
723 because it is too late at that time, and has no effect.
725 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
726 security issue with \' (bugzilla #107). I could not use the
727 PQescapeStringConn() function, because it needs a PGconn value as one of
730 PH/08 When testing addresses using -bt, indicate those final addresses that
731 are duplicates that would not cause an additional delivery. At least one
732 person was confused, thinking that -bt output corresponded to deliveries.
733 (Suppressing duplicates isn't a good idea as you lose the information
734 about possibly different redirections that led to the duplicates.)
736 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
737 systems where poll() doesn't work, in particular OS X.
739 PH/10 Added more information to debugging output for retry time not reached.
741 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
742 operations in malware.c.
744 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
747 PH/13 If write_rejectlog was set false when logging was sent to syslog with
748 syslog_duplication set false, log lines that would normally be written
749 both the the main log and to the reject log were not written to syslog at
752 PH/14 In the default configuration, change the use of "message" in ACL warn
753 statements to "add_header".
755 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
756 not followed by a command (e.g. "seen endif").
758 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
759 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
762 PH/17 Added extra conditions to the default value of delay_warning_condition
766 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
767 { match{$h_precedence:}{(?i)bulk|list|junk} } \
768 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
771 The Auto-Submitted: and various List- headers are standardised, whereas I
772 don't think Precedence: ever was.
774 PH/18 Refactored debugging code in route_finduser() to show more information,
775 in particular, the error code if getpwnam() issues one.
777 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
778 This is apparently needed in addition to the PH/07 change above to avoid
779 any possible encoding problems.
781 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
782 but not after initializing Perl.
784 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
785 output them only if debugging. By default they are written stderr,
786 apparently, which is not desirable.
788 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
791 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
794 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
796 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
797 authenticated or an ident call has been made. Suppress the default
798 values for $authenticated_id and $authenticated_sender (but permit -oMai
799 and -oMas) when testing with -bh.
801 PH/24 Re-jigged the order of the tests in the default configuration so that the
802 tests for valid domains and recipients precede the DNS black list and CSA
803 tests, on the grounds that those ones are more expensive.
805 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
806 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
807 This bug exists in every version of Exim that I still have, right back to
810 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
811 However, an attempt to turn on case-sensitivity in a regex key by
812 including (?-i) didn't work because the subject string was already
813 lowercased, and the effects were non-intuitive. It turns out that a
814 one-line patch can be used to allow (?-i) to work as expected.
820 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
821 other effects) broke the use of negated acl sub-conditions.
823 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
826 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
827 "Deny" causes Exim to reject the incoming connection with a 554 error.
828 Unfortunately, if there is a major crisis, such as a disk failure,
829 tcp-wrappers gives "deny", whereas what one would like would be some
830 kind of temporary error. A kludge has been added to help with this.
831 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
832 554 error is used if errno is still zero or contains ENOENT (which occurs
833 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
836 PH/03 Add -lutil to the default FreeBSD LIBS setting.
838 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
839 errors. Otherwise a message that provokes a temporary error (when other
840 messages do not) can cause a whole host to time out.
842 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
843 addresses were routed directly to files or pipes from a redirect router.
844 File deliveries just didn't batch; pipe deliveries might have suffered
847 PH/06 A failure to get a lock for a hints database would erroneously always say
848 "Failed to get write lock", even when it was really a read lock.
850 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
851 of 0600. This has been changed to use the value of the lockfile_mode
852 option (which defaults to 0600).
854 PH/08 Applied small patch from the Sieve maintainer.
856 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
857 folder from quota calculations, a direct delivery into this folder messed
858 up the contents of the maildirsize file. This was because the regex was
859 used only to exclude .Trash (or whatever) when the size of the mailbox
860 was calculated. There was no check that a delivery was happening into an
861 excluded directory. This bug has been fixed by ignoring all quota
862 processing for deliveries into excluded directories.
864 PH/10 Added the maildirfolder_create_regex option to appendfile.
870 PH/01 The code for finding all the local interface addresses on a FreeBSD
871 system running IPv6 was broken. This may well have applied to all BSD
872 systems, as well as to others that have similar system calls. The broken
873 code found IPv4 interfaces correctly, but gave incorrect values for the
874 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
875 that it would not match correctly against @[] and not recognize the IPv6
878 PH/02 The ipliteral router was not recognizing addresses of the form user@
879 [ipv6:....] because it didn't know about the "ipv6:" prefix.
881 PH/03 Added disable_ipv6.
883 PH/04 Changed $reply_address to use the raw form of the headers instead of the
884 decoded form, because it is most often used to construct To: headers
885 lines in autoreplies, and the decoded form may well be syntactically
886 invalid. However, $reply_address has leading white space removed, and all
887 newlines turned into spaces so that the autoreply transport does not
890 PH/05 If group was specified without a user on a router, and no group or user
891 was specified on a transport, the group from the router was ignored.
893 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
894 for visible compile-time settings that can be used to change these
895 numbers, for those that want even more. Backwards compatibility with old
896 spool files has been maintained. However, going back to a previous Exim
897 release will lost any variables that are in spool files.
899 PH/07 Two small changes when running in the test harness: increase delay when
900 passing a TCP/IP connection to a new process, in case the original
901 process has to generate a bounce, and remove special handling of
902 127.0.0.2 (sic), which is no longer necessary.
904 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
905 be the same on different OS.
907 PH/09 Moved a debug statement in filter processing to avoid a race problem when
910 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
911 whether --show-vars was specified or not
913 JJ/02 exipick: Added support for new ACL variable spool format introduced
916 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
917 syntactically invalid From: or Reply-to: line, and a filter used this to
918 generate an autoreply, and therefore failed to obtain an address for the
919 autoreply, Exim could try to deliver to a non-existent relative file
920 name, causing unrelated and misleading errors. What now happens is that
921 it logs this as a hard delivery error, but does not attempt to create a
924 PH/11 The exinext utility has a -C option for testing purposes, but although
925 the given file was scanned by exinext itself; it wasn't being passed on
926 when Exim was called.
928 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
929 an end-of-file indication when reading a command response.
931 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
932 compiled. In many other places in Exim, IPv6 addresses are always
933 recognized, so I have changed this. It also means that IPv4 domain
934 literals of the form [IPV4:n.n.n.n] are now always recognized.
936 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
937 used if the router is not running as root, for example, when verifying at
938 ACL time, or when using -bh. The debugging output from this situation was
939 non-existent - all you got was a failure to exec. I have made two
942 (a) Failures to set uid/gid, the current directory, or a process leader
943 in a subprocess such as that created by queryprogram now generate
944 suitable debugging ouput when -d is set.
946 (b) The queryprogram router detects when it is not running as root,
947 outputs suitable debugging information if -d is set, and then runs
948 the subprocess without attempting to change uid/gid.
950 PH/15 Minor change to Makefile for building test_host (undocumented testing
953 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
954 additional section of a DNS packet that returns MX or SRV records.
955 Instead, it always explicitly searches for A/AAAA records. This avoids
956 major problems that occur when a DNS server includes only records of one
957 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
958 fixed another bug: if SRV records were looked up and the corresponding
959 address records were *not* found in the additional section, the port
960 values from the SRV records were lost.
962 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
963 using the correct key (the original address) when searching the retry
964 rules in order to find which one to use for generating the retry hint.
966 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
967 adding the default one. Similarly, if it contains a Reply-To: header, the
968 errors_reply_to option, if set, is not used.
970 PH/19 When calculating a retry time, Exim used to measure the "time since
971 failure" by looking at the "first failed" field in the retry record. Now
972 it does not use this if it is later than than the arrival time of the
973 message. Instead it uses the arrival time. This makes for better
974 behaviour in cases where some deliveries succeed, thus re-setting the
975 "first failed" field. An example is a quota failure for a huge message
976 when small messages continue to be delivered. Without this change, the
977 "time since failure" will always be short, possible causing more frequent
978 delivery attempts for the huge message than are intended.
979 [Note: This change was subsequently modified - see PH/04 for 4.62.]
981 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
982 $1, $2, $3) because the numerical variables can be reset during some
983 expansion items (e.g. "match"), thereby losing the authentication data.
985 PH/21 Make -bV show the size of off_t variables so that the test suite can
986 decide whether to run tests for quotas > 2G.
988 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
989 mailbox_size, and mailbox_filecount in the appendfile transport. If a
990 filecount value is greater than 2G or if a quota value is greater than 2G
991 on a system where the size of off_t is not greater than 4, a panic error
994 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
995 never match. The debug and -bh output now contains an explicit error
996 message indicating a malformed IPv4 address or mask.
998 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
999 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1000 PH/23 above applies.
1002 PH/25 Do not write to syslog when running in the test harness. The only
1003 occasion when this arises is a failure to open the main or panic logs
1004 (for which there is an explicit test).
1006 PH/26 Added the /no_tell option to "control=freeze".
1008 PH/27 If a host name lookup failed very early in a connection, for example, if
1009 the IP address matched host_lookup and the reverse lookup yielded a name
1010 that did not have a forward lookup, an error message of the form "no IP
1011 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1012 could be logged. Now it outputs the IP address instead of "NULL".
1014 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1015 allows the sender and the authenticated sender to be set when
1016 submitting a message from within Exim. Since child_open_exim() is
1017 documented for local_scan(), the new function should be too.
1019 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1020 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1021 results in an empty string is now treated as unset.
1023 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1025 PH/31 Added sender_verify_fail logging option.
1027 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1028 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1029 tidied the source and removed it altogether.
1031 PH/33 When a queue run was abandoned because the load average was too high, a
1032 log line was always written; now it is written only if the queue_run log
1033 selector is set. In addition, the log line for abandonment now contains
1034 information about the queue run such as the pid. This is always present
1035 in "start" and "stop" lines but was omitted from the "abandon" line.
1037 PH/34 Omit spaces between a header name and the colon in the error message that
1038 is given when verify = headers_syntax fails (if there are lots of them,
1039 the message gets confusing).
1041 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1042 names, as there are now some PTR records that contain slashes. This check
1043 is only to protect against broken name servers that fall over on strange
1044 characters, so the fact that it applies to all lookups doesn't matter.
1046 PH/36 Now that the new test suite is complete, we can remove some of the
1047 special code in Exim that was needed for the old test suite. For example,
1048 sorting DNS records because real resolvers return them in an arbitrary
1049 order. The new test suite's fake resolver always returns records in the
1052 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1053 bounces) except when queue_only is set, to avoid logging races between
1054 the different processes.
1056 PH/38 Panic-die if .include specifies a non-absolute path.
1058 PH/39 A tweak to the "H" retry rule from its user.
1060 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1061 a label. They prevented compilation on older perls.
1063 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1064 a warning to be raised on newish perls.
1066 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1067 on queue. Changes to match documented behaviour of showing count of
1068 messages matching specified criteria.
1070 PH/40 Changed the default ident timeout from 30s to 5s.
1072 PH/41 Added support for the use of login_cap features, on those BSD systems
1073 that have them, for controlling the resources used by pipe deliveries.
1075 PH/42 The content-scanning code uses fopen() to create files in which to put
1076 message data. Previously it was not paying any attention to the mode of
1077 the files. Exim runs with umask(0) because the rest of the code creates
1078 files with open(), and sets the required mode explicitly. Thus, these
1079 files were ending up world-writeable. This was not a big issue, because,
1080 being within the spool directory, they were not world-accessible. I have
1081 created a function called modefopen, which takes an additional mode
1082 argument. It sets umask(777), creates the file, chmods it to the required
1083 mode, then resets the umask. All the relevant calls to fopen() in the
1084 content scanning code have been changed to use this function.
1086 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1087 to 24 hours. This avoids potential overflow problems when processing G
1088 and H retry rules. I suspect nobody ever tinkers with this value.
1090 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1092 PH/45 When the plaintext authenticator is running as a client, the server's
1093 challenges are checked to ensure they are valid base64 strings. By
1094 default, the authentication attempt is cancelled if an invalid string is
1095 received. Setting client_ignore_invalid_base64 true ignores these errors.
1096 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1097 they are received. Thus, the responses can be made to depend on the
1098 challenges. If an invalid string is ignored, an empty string is placed in
1101 PH/46 Messages that are created by the autoreply transport now contains a
1102 References: header, in accordance with RFCs 2822 and 3834.
1104 PH/47 Added authenticated_sender_force to the smtp transport.
1106 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1108 PH/49 Installed latest patch from the Sieve maintainer.
1110 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1111 was also set, the appendfile transport was unnecessarily scanning a
1112 directory of message files (e.g. for maildir delivery) to find the count
1113 of files (along with the size), even though it did not need this
1114 information. It now does the scan only if it needs to find either the
1115 size of the count of files.
1117 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1119 PH/52 Two bugs concerned with error handling when the smtp transport is
1122 (i) Exim was not creating retry information for temporary errors given
1123 for individual recipients after the DATA command when the smtp transport
1124 was used in LMTP mode. This meant that they could be retried too
1125 frequently, and not timed out correctly.
1127 (ii) Exim was setting the flag that allows error details to be returned
1128 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1129 recipients that were returned after the DATA command.
1131 PH/53 This is related to PH/52, but is more general: for any failing address,
1132 when detailed error information was permitted to be returned to the
1133 sender, but the error was temporary, then after the final timeout, only
1134 "retry timeout exceeded" was returned. Now it returns the full error as
1135 well as "retry timeout exceeded".
1137 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1138 do this, and (what is worse) MTAs that accept it.
1140 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1141 will now be deprecated.
1143 PH/56 New os.c-cygwin from the Cygwin maintainer.
1145 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1146 formats (previously only available in exim formats via -bpr, -bpru,
1147 and -bpra. Now also available in native and exiqgrep formats)
1149 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1150 with very large, slow to parse queues
1152 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1154 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1156 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1157 responses to authentication challenges, though it was showing the
1158 challenges; (ii) I've removed the CR characters from the debug output for
1161 PH/58 Allow for the insertion of a newline as well as a space when a string
1162 is turned into more than one encoded-word during RFC 2047 encoding. The
1163 Sieve code now uses this.
1165 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1166 data_4xx, lost_connection, tls_required.
1168 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1169 message was being sent as an SMTP response.
1171 PH/61 Add -l and -k options to exicyclog.
1173 PH/62 When verifying, if an address was redirected to one new address, so that
1174 verification continued, and the new address failed or deferred after
1175 having set something in $address_data, the value of $address_data was not
1176 passed back to the ACL. This was different to the case when no
1177 redirection occurred. The value is now passed back in both cases.
1179 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1180 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1181 use login_cap.h, so on its own it isn't the distinguishing feature. The
1182 new name refers directly to the setclassresources() function.
1184 PH/65 Added configuration files for NetBSD3.
1186 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1188 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1189 is preferred over IPv4.
1191 PH/68 The bounce_return_message and bounce_return_body options were not being
1192 honoured for bounces generated during the reception of non-SMTP messages.
1193 In particular, this applied to messages rejected by the ACL. This bug has
1194 been fixed. However, if bounce_return_message is true and bounce_return_
1195 body is false, the headers that are returned for a non-SMTP message
1196 include only those that have been read before the error was detected.
1197 (In the case of an ACL rejection, they have all been read.)
1199 PH/69 The HTML version of the specification is now built in a directory called
1200 spec_html instead of spec.html, because the latter looks like a path with
1201 a MIME-type, and this confuses some software.
1203 PH/70 Catch two compiler warnings in sieve.c.
1205 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1206 function verify_get_ident() calls ip_connect() to connect a socket, but
1207 if the "connect()" function timed out, ip_connect() used to close the
1208 socket. However, verify_get_ident() also closes the socket later, and in
1209 between Exim writes to the log, which may get opened at this point. When
1210 the socket was closed in ip_connect(), the log could get the same file
1211 descriptor number as the socket. This naturally causes chaos. The fix is
1212 not to close the socket in ip_connect(); the socket should be closed by
1213 the function that creates it. There was only one place in the code where
1214 this was missing, in the iplookup router, which I don't think anybody now
1215 uses, but I've fixed it anyway.
1217 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1218 well as to direct DNS lookups. Otherwise the handling of names in host
1219 lists is inconsistent and therefore confusing.
1225 PH/01 Two changes to the default runtime configuration:
1227 (1) Move the checks for relay_from_hosts and authenticated clients from
1228 after to before the (commented out) DNS black list checks.
1230 (2) Add control=submission to the relay_from_hosts and authenticated
1231 clients checks, on the grounds that messages accepted by these
1232 statements are most likely to be submissions.
1234 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1236 (1) Generate an error if the third argument for the ${prvs expansion is
1239 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1242 (3) Reset the variables that are obtained from the first argument of
1243 ${prvscheck and used in the second argument before leaving the code,
1244 because their memory is reclaimed, so using them afterwards may do
1247 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1248 one (it's much easier than Tom thought :-).
1250 (5) Because of (4), we can now allow for the use of $prvscheck_result
1251 inside the third argument.
1253 PH/03 For some reason, the default setting of PATH when running a command from
1254 a pipe transport was just "/usr/bin". I have changed it to
1257 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1258 anything to be listed in the output from -bV.
1260 PH/05 When a filter generated an autoreply, the entire To: header line was
1261 quoted in the delivery log line, like this:
1263 => >A.N.Other <ano@some.domain> <original@ddress> ...
1265 This has been changed so that it extracts the operative address. There
1266 may be more than one such address. If so, they are comma-separated, like
1269 => >ano@some.domain,ona@other.domain <original@ddress> ...
1271 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1272 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1273 not being looked up in the rDNS to get a host name, Exim was showing the
1274 IP address twice in Received: lines, even though the IP addresses were
1275 identical. For example:
1277 Received: from [1.2.3.4] (helo=[1.2.3.4])
1279 However, if the real host name was known, it was omitting the HELO data
1280 if it matched the actual IP address. This has been tidied up so that it
1281 doesn't show the same IP address twice.
1283 PH/07 When both +timestamp and +memory debugging was on, the value given by
1284 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1285 called by the debug printing, thereby overwriting the timestamp buffer.
1286 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1288 PH/08 When the original message was included in an autoreply transport, it
1289 always said "this is a copy of the message, including all the headers",
1290 even if body_only or headers_only was set. It now gives an appropriate
1293 PH/09 Applied a patch from the Sieve maintainer which:
1295 o fixes some comments
1296 o adds the (disabled) notify extension core
1297 o adds some debug output for the result of if/elsif tests
1298 o points to the current vacation draft in the documentation
1299 and documents the missing references header update
1303 o fixes a bug in processing the envelope test (when testing
1304 multiple envelope elements, the last element determinted the
1307 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1308 Electronic Mail") by including:
1310 Auto-submitted: auto-generated
1312 in the messages that it generates (bounce messages and others, such as
1313 warnings). In the case of bounce messages for non-SMTP mesages, there was
1314 also a typo: it was using "Auto_submitted" (underscore instead of
1315 hyphen). Since every message generated by Exim is necessarily in response
1316 to another message, thes have all been changed to:
1318 Auto-Submitted: auto-replied
1320 in accordance with these statements in the RFC:
1322 The auto-replied keyword:
1324 - SHOULD be used on messages sent in direct response to another
1325 message by an automatic process,
1327 - MUST NOT be used on manually-generated messages,
1329 - MAY be used on Delivery Status Notifications (DSNs) and Message
1330 Disposition Notifications (MDNs),
1332 - MUST NOT be used on messages generated by automatic or periodic
1333 processes, except for messages which are automatic responses to
1336 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1337 to the default Received: header definition.
1339 PH/12 Added log selector acl_warn_skipped (default on).
1341 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1342 variables because (a) they are in the wrong storage pool and (b) even if
1343 they were copied, it wouldn't work properly because of the caching.
1345 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1346 checking when decoding. Apparently there are clients that generate
1347 overlong encoded strings. Why am I not surprised?
1349 PH/15 If the first argument of "${if match_address" was not empty, but did not
1350 contain an "@" character, Exim crashed. Now it writes a panic log message
1351 and treats the condition as false.
1353 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1355 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1356 extension "envelope-auth". The code is finished and in agreement with
1357 other implementations, but there is no documentation so far and in fact,
1358 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1359 not changing the active code.
1361 Print executed "if" and "elsif" statements when debugging is used. This
1362 helps a great deal to understand what a filter does.
1364 Document more things not specified clearly in RFC3028. I had all this
1365 sorted out, when out of a sudden new issues came to my mind. Oops."
1367 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1370 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1371 canonical form (as documented). However, after a host name lookup from
1372 the IP address, check_host() was doing a simple string comparison with
1373 addresses acquired from the DNS when checking that the found name did
1374 have the original IP as one of its addresses. Since any found IPv6
1375 addresses are likely to be in abbreviated form, the comparison could
1376 fail. Luckily, there already exists a function for doing the comparison
1377 by converting both addresses to binary, so now that is used instead of
1378 the text comparison.
1380 PH/20 There was another similar case to PH/19, when a complete host name was
1381 given in a host list; looking up its IP address could give an abbreviated
1382 form, whereas the current host's name might or might not be abbreviated.
1383 The same fix has been applied.
1389 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1390 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1393 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1394 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1396 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1398 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1399 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1400 Also, for "Auto-Submitted", treat anything other than "no" as
1401 non-personal, in accordance with RFC 3834. (Previously it treated
1402 anything starting "auto-" as non-personal.)
1404 TF/01 The control=submission/name=... option had a problem with syntax
1405 errors if the name included a slash character. The /name= option
1406 now slurps the rest of the string, so it can include any characters
1407 but it must come last in the list of options (after /sender_retain
1410 PH/05 Some modifications to the interface to the fake nameserver for the new
1418 TK/01 Added the "success_on_redirect" address verification option. See
1419 NewStuff for rationale and an example.
1421 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1423 PH/02 Patch to exigrep to allow it to work on syslog lines.
1425 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1426 fread() to skip over the body file's header line, because in Cygwin the
1427 header line is locked and is inaccessible.
1429 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1430 co-exist for some time) to make it clear that it is the Exim ID that is
1431 referenced, not the Message-ID: header line.
1433 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1434 string_format() function, because snprintf() does not exist on all
1437 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1438 ${stat: expansion item.
1440 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1441 protocol synchronization error", to keep the pedants happy.
1443 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1444 well as for IRIX systems, when gcc is being used. See the host.c source
1447 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1449 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1452 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1453 transport and to the smtp transport in LMTP mode.
1455 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1457 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1458 run a filter in a subprocess. This could lead to confusion in subsequent
1459 lookups in the parent process. There should also be a search_tidyup() at
1460 the end of the subprocess.
1462 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1463 only if the host matched helo_try_verify_hosts, which caused the
1464 verification to occur when the EHLO/HELO command was issued. The ACL just
1465 tested the remembered result. Now, if a previous verification attempt has
1466 not happened, "verify = helo" does it there and then.
1468 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1470 TK/03 Fix log output including CR from clamd.
1472 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1473 exist provoked a memory error which could cause a segfault.
1475 PH/15 Installed PCRE 6.2
1477 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1479 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1480 of the problem. Specifically, suggested +O2 rather than +O1 for the
1483 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1485 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1486 record, the port was not correctly logged when the outgoing_port log
1487 selector was set (it logged the transort's default port).
1489 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1490 fallback_hosts, and "hosts" in the smtp transport.
1492 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1493 host errors such as "Connection refused".
1495 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1496 authentication with radiusclient 0.4.9:
1498 - Error returned from rc_read_config was caught wrongly
1499 - Username/password not passed on to radius server due to wrong length.
1501 The presumption is that some radiusclient API changes for 4.51/PH/17
1502 were not taken care of correctly. The code is still untested by me (my
1503 Linux distribution still has 0.3.2 of radiusclient), but it was
1504 contributed by a Radius user.
1506 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1507 expanding the "port" option of the smtp transport.
1509 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1510 while reading a MIME header. Thanks to Tom Hughes for a patch.
1512 PH/24 Include config.h inside local_scan.h so that configuration settings are
1515 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1516 that in an ACL for RCPT (for example), you can examine exactly what was
1519 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1520 commands, but it was not correctly comparing the address with the actual
1521 client host address. Thus, it would show the EHLO address in Received:
1522 header lines when this was not necessary.
1524 PH/27 Added the % operator to ${eval:}.
1526 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1527 it should be ignoring failures (because with -C, for example, it has lost
1528 privilege). It wasn't ignoring creation failures other than "already
1531 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1534 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1535 because an input line was too long, either on its own, or by virtue of
1536 too many continuations, the temporary file was not being removed, and the
1537 return code was incorrect.
1539 PH/31 Missing "BOOL" in function definition in filtertest.c.
1541 PH/32 Applied Sieve patches from the maintainer.
1543 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1545 PH/33 Added "verify = not_blind".
1547 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1548 Local/Makefile (with some defaults set). These are used in built scripts
1549 such as exicyclog, but they have never been used in the exim_install
1550 script (though there are many overriding facilities there). I have
1551 arranged that the exim_install script now takes note of these two
1554 PH/35 Installed configuration files for Dragonfly.
1556 PH/36 When a locally submitted message by a trusted user did not contain a
1557 From: header, and the sender address was obtained from -f or from an SMTP
1558 MAIL command, and the trusted user did not use -F to supply a sender
1559 name, $originator_name was incorrectly used when constructing a From:
1560 header. Furthermore, $originator_name was used for submission mode
1561 messages from external hosts without From: headers in a similar way,
1562 which is clearly wrong.
1564 PH/37 Added control=suppress_local_fixups.
1566 PH/38 When log_selector = +received_sender was set, and the addition of the
1567 sender made the log line's construction buffer exactly full, or one byte
1568 less than full, an overflow happened when the terminating "\n" was
1571 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1572 when the result of a list match is failure because a DNS lookup failed.
1574 PH/40 RM_COMMAND is now used in the building process.
1576 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1577 the "build-* directories that it finds.
1579 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1580 address in a domain literal was a prefix of an interface address.
1582 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1583 when verifying a sender address, unless rewrite_headers is false.
1585 PH/44 Wrote a long comment about why errors_to addresses are verified as
1586 recipients, not senders.
1588 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1589 the ratelimit ACL was added.
1591 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1593 PH/46 Added extra information about PostgreSQL errors to the error string.
1595 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1596 avoiding the need to install special zones in a real server. This is
1597 backwards compatible; if it can't find the fake resolver, it drops back.
1598 Thus, both old and new test suites can be run.
1600 TF/02 Added util/ratelimit.pl
1602 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1603 clock is set back in time.
1605 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1606 Candler <B.Candler@pobox.com>.
1608 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1609 OK for addresses that are the result of redirections.
1611 PH/48 A number of further additions for the benefit of the new test suite,
1612 including a fake gethostbyname() that interfaces to the fake DNS resolver
1615 TF/06 The fix for widen_domains has also been applied to qualify_single and
1616 search_parents which are the other dnslookup options that can cause
1619 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1622 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1624 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1625 service" if it encountered an MX record with an empty target hostname.
1626 The message is now "an MX or SRV record indicated no SMTP service".
1628 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1629 if the DNS of the sending site is misconfigured. This is quite a
1630 common situation. This change restores the behaviour of treating a
1631 helo verification defer as a failure.
1633 PH/51 If self=fail was set on a router, the bounce message did not include the
1634 actual error message.
1640 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1642 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1643 command itself ended in error, the underlying message about the transport
1644 filter timeout was being overwritten with the pipe command error. Now the
1645 underlying error message should be appended to the second error message.
1647 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1649 PH/02 Reduce the amount of output that "make" produces by default. Full output
1650 can still be requested.
1652 PH/03 The warning log line about a condition test deferring for a "warn" verb
1653 was being output only once per connection, rather than after each
1654 occurrence (because it was using the same function as for successful
1655 "warn" verbs). This seems wrong, so I have changed it.
1657 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1658 it should not have, which might have caused a crash in the right
1659 circumstances, but probably never did.
1661 PH/04 Installed a modified version of Tony Finch's patch to make submission
1662 mode fix the return path as well as the Sender: header line, and to
1663 add a /name= option so that you can make the user's friendly name appear
1666 TF/03 Added the control = fakedefer ACL modifier.
1668 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1669 Mark Lowes for thorough testing.
1671 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1673 TK/03 Merged latest SRS patch from Miles Wilton.
1675 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1676 with the definition in sysexits.h (which is #included earlier).
1677 Fortunately, Exim does not actually use EX_OK. The code used to try to
1678 preserve the sysexits.h value, by assumimg that macro definitions were
1679 scanned for macro replacements. I have been disabused of this notion,
1680 so now the code just undefines EX_OK before #including unistd.h.
1682 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1683 in the smtp transport. When a block could not be written in a single
1684 write() function, the timeout was being re-applied to each part-write.
1685 This seems wrong - if the receiver was accepting one byte at a time it
1686 would take for ever. The timeout is now adjusted when this happens. It
1687 doesn't have to be particularly precise.
1689 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1690 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1692 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1693 without a domain if the check to <postmaster@domain> fails.
1695 SC/01 Eximstats: added -xls and the ability to specify output files
1696 (patch written by Frank Heydlauf).
1698 SC/02 Eximstats: use FileHandles for outputing results.
1700 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1702 SC/04 Eximstats: fixed display of large numbers with -nvr option
1704 SC/05 Eximstats: fixed merging of reports with empty tables.
1706 SC/06 Eximstats: added the -include_original_destination flag
1708 SC/07 Eximstats: removed tabs and trailing whitespace.
1710 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1712 TK/06 MBOX spool code: Add real "From " MBOX separator line
1713 so the .eml file is really in mbox format (even though
1714 most programs do not really care). Patch from Alex Miller.
1716 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1717 The latter is generated from $received_to and is only set if the
1718 message has one envelope recipient. SA can use these headers,
1719 obviously out-of-the-box. Patch from Alex Miller.
1721 PH/08 The ${def test on a variable was returning false if the variable's
1722 value was "0", contrary to what the specification has always said!
1723 The result should be true unless the variable is empty.
1725 PH/09 The syntax error of a character other than { following "${if
1726 def:variable_name" (after optional whitespace) was not being diagnosed.
1727 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1728 accidental colon was present, for example, could give incorrect results.
1730 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1731 result is used (not including appendfile, where other changes are about
1734 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1735 This involved changing a lot of size variables from int to off_t. It
1736 should work with maildirs and everything.
1738 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1739 spamd dying while we are connected to it.
1741 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1744 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1745 function for BDB 4.3.
1747 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1749 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1750 the sources for PCRE have been re-arranged and more files are now
1753 PH/15 The code I had for printing potentially long long variables in PH/11
1754 above was not the best (it lost precision). The length of off_t variables
1755 is now inspected at build time, and an appropriate printing format (%ld
1756 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1757 to be "long long int" or "long int". This is needed for the internal
1758 formatting function string_vformat().
1760 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1761 the configuration file to be ":syslog", then the script "guesses" where
1762 the logs files are, rather than using the compiled in default. In our
1763 case the guess is not the same as the compiled default, so the script
1764 suddenly stopped working when I started to use syslog. The patch checks
1765 to see if log_file_path is "". If so, it attempts to read it from exim
1766 with no configuration file to get the compiled in version, before it
1767 falls back to the previous guessing code."
1769 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1770 implementing BATV in an Exim configuration. See NewStuff for the gory
1773 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1774 Makefile that are specific to HP-UX.
1776 PH/18 If the "use_postmaster" option was set for a recipient callout together
1777 with the "random" option, the postmaster address was used as the MAIL
1778 FROM address for the random test, but not for the subsequent recipient
1779 test. It is now used for both.
1781 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1782 patch removes a few documentation additions to RFC 3028, because the
1783 latest draft now contains them. It adds the new en;ascii-case comparator
1784 and a new error check for 8bit text in MIME parts. Comparator and
1785 require names are now matched exactly. I enabled the subaddress
1786 extension, but it is not well tested yet (read: it works for me)."
1788 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1789 rework some of the code of TK/09 above to avoid the hardwired use of
1790 "%lld" and "long long". Replaced the call to snprintf() with a call to
1793 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1794 records point to non-existent hosts", "retry timeout exceeded", and
1795 "retry time not reached for any host after a long failure period".
1797 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1798 experimental DomainKeys support:
1800 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1801 (2) On an error such as an illegally used "control", the wrong name for
1802 the control was given.
1804 These problems did NOT occur unless DomainKeys support was compiled.
1806 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1808 PH/24 Added ${if match_ip condition.
1810 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1811 databases so that it will be absolutely obvious if a crash occurs in the
1812 DB library. This is a regular occurrence (often caused by mis-matched
1815 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1816 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1817 picked up on a user's system that detects such things. There doesn't seem
1818 to be a gcc warning option for this - only an attribute that has to be
1819 put on the function's prototype. It seems that in Fedora Core 4 they have
1820 set this on a number of new functions. No doubt there will be more in due
1823 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1824 specify a transport. However, if an address that was verified by such a
1825 router was the subject of a callout, Exim crashed because it tried to
1826 read the rcpt_include_affixes from the non-existent transport. Now it
1827 just assumes that the setting of that option is false. This bug was
1828 introduced by 4.51/PH/31.
1830 PH/28 Changed -d+all to exclude +memory, because that information is very
1831 rarely of interest, but it makes the output a lot bigger. People tend to
1832 do -d+all out of habit.
1834 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1835 code in os-type was giving problems when libc.so lives in lib64, like on
1838 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1839 aren't the modern standard, and it seems that some systems' include files
1840 don't always have them. Exim was already checking for some of the newer
1841 ones like T_AAAA, and defining it itself. I've added checks for all the
1842 record types that Exim uses.
1844 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1845 not automatically generating a new one, as it is supposed to. This
1846 prevented TLS from working. If the file did exist, but contained invalid
1847 data, a new version was generated, as expected. It was only the case of a
1848 non-existent file that was broken.
1850 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1851 with a change in libdomainkeys > 0.64.
1853 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1854 from DNS. If the selector record carries the flag, it now has
1855 precedence over the domain-wide flag.
1857 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1859 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1860 the use of a port name that isn't defined in /etc/services) occurred, the
1861 message was deferred as in a normal delivery, and thus remained on the
1862 spool, instead of being failed because of the mua_wrapper setting. This
1863 is now fixed, and I tidied up some of the mua_wrapper messages at the
1866 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1867 the messages in a hash of arrays rather than using individual hashes.
1868 This is a bit cleaner and results in dramatic memory savings, albeit
1869 at a slight CPU cost.
1871 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1872 as requested by Marc Sherman.
1874 SC/10 Eximstats: added histograms for user specified patterns as requested
1877 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1879 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1880 fopen() in the content-scanning modules that did not already have it.
1886 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1887 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1889 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1891 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1893 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1894 to test Sieve filters that use "vacation".
1896 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1897 that changes the way the GnuTLS parameters are stored in the cache file.
1898 The new format can be generated externally. For backward compatibility,
1899 if the data in the cache doesn't make sense, Exim assumes it has read an
1900 old-format file, and it generates new data and writes a new file. This
1901 means that you can't go back to an older release without removing the
1904 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1905 work if there are any delivery delays because "one_time" forces the
1906 parent to be marked "delivered", so its unseen clone is never tried
1907 again. For this reason, Exim now forbids the simultaneous setting of
1910 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1911 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1912 entry if you want to know the details. The fix, however, neglected to
1913 consider the case when local delivery batching is involved. The test for
1914 "previously delivered" was not happening when checking to see if an
1915 address could be batched with a previous (undelivered) one; under
1916 certain circumstances this could lead to multiple deliveries to the same
1919 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1920 in its include files, and this causes problems building Exim.
1922 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1923 header_syntax) but Exim was just ignoring anything given after a slash.
1924 In particular, this caused confusion with an attempt to use "verify =
1925 reverse_host_lookup/defer_ok". An error is now given when options are
1926 supplied for verify items that do not have them. (Maybe reverse_host_
1927 lookup should have a defer_ok option, but that's a different point.)
1929 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1930 defined by RFC 821) to 2048, because there were problems with some AUTH
1931 commands, and RFC 1869 says the size should be increased for extended
1932 SMTP commands that take arguments.
1934 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1937 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1938 "unknown" error; now it says that the functionality isn't in the binary.
1940 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1941 an address' error message when a string expansion fails (syntax or
1942 whatever). Otherwise the password may appear in the log. Following change
1943 PH/42 below, there is no longer a chance of it appearing in a bounce
1946 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1948 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1949 of its addresses was ever tried. (Bugzilla bug #2.)
1951 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1952 the result incorrectly in the debug output. (It correctly added a newline
1953 to what was transported.)
1955 TF/01 Added $received_time.
1957 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1958 commented out examples of how to interface to a virus scanner and to
1959 SpamAssassin. Also added commented examples of av_scanner and
1960 spamd_address settings.
1962 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1963 and controls are allowed in which ACLs. There were a couple of minor
1964 errors. Some of the entries in the conditions table (which is a table of
1965 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1966 negation of where the condition IS allowed.
1968 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1970 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1971 header file does not have a version number, so I've had to invent a new
1972 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1973 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1974 radiusclient), but it was contributed by a Radius user.
1976 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1977 files or directories, for OpenSSL.
1979 PH/19 When an Exim process that is running as root has to create an Exim log
1980 file, it does so in a subprocess that runs as exim:exim so as to get the
1981 ownership right at creation (otherwise, other Exim processes might see
1982 the file with the wrong ownership). There was no test for failure of this
1983 fork() call, which would lead to the process getting stuck as it waited
1984 for a non-existent subprocess. Forks do occasionally fail when resources
1985 run out. I reviewed all the other calls to fork(); they all seem to check
1988 PH/20 When checking for unexpected SMTP input at connect time (before writing
1989 the banner), Exim was not dealing correctly with a non-positive return
1990 from the read() function. If the client had disconnected by this time,
1991 the result was a log entry for a synchronization error with an empty
1992 string after "input=" when read() returned zero. If read() returned -1
1993 (an event I could not check), uninitialized data bytes were printed.
1994 There were reports of junk text (parts of files, etc) appearing after
1997 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1999 PH/22 Added support for macro redefinition, and (re)definition in between
2000 driver and ACL definitions.
2002 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2003 forgetting to use the resulting value; it was using the unexpanded value.
2005 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2006 hadn't been configured. The fix is from Juergen Kreileder, who
2007 understands it better than I do:
2009 "Here's what I see happening with three configured cyrus_sasl
2010 authenticators configured (plain, login, cram-md5):
2012 On startup auth_cyrus_sasl_init() gets called for each of these.
2013 This means three calls to sasl_listmech() without a specified mech_list.
2014 => SASL tests which mechs of all available mechs actually work
2015 => three warnings about OTP not working
2016 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2018 With the patch, sasl_listmech() also gets called three times. But now
2019 SASL's mech_list option is set to the server_mech specified in the the
2020 authenticator. Or in other words, the answer from sasl_listmech()
2021 gets limited to just the mech you're testing for (which is different
2023 => the return list contains just 'plain' or 'login', 'cram-md5' or
2024 nothing depending on the value of ob->server_mech.
2026 I've just tested the patch: Authentication still works fine,
2027 unavailable mechs specified in the exim configuration are still
2028 caught, and the auth.log warnings about OTP are gone."
2030 PH/25 When debugging is enabled, the contents of the command line are added
2031 to the debugging output, even when log_selector=+arguments is not
2034 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2035 answer is "GNU", and only if the return is "GNU/something" is the answer
2038 PH/27 $acl_verify_message is now set immediately after the failure of a
2039 verification in an ACL, and so is available in subsequent modifiers. In
2040 particular, the message can be preserved by coding like this:
2042 warn !verify = sender
2043 set acl_m0 = $acl_verify_message
2045 Previously, $acl_verify_message was set only while expanding "message"
2046 and "log_message" when a very denied access.
2048 PH/28 Modified OS/os.c-Linux with
2050 -#ifndef OS_LOAD_AVERAGE
2051 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2053 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2054 nomenclature these days.)
2056 PH/29 Installed patch from the Sieve maintainer that adds the options
2057 sieve_useraddress and sieve_subaddress to the redirect router.
2059 PH/30 In these circumstances:
2060 . Two addresses routed to the same list of hosts;
2061 . First host does not offer TLS;
2062 . First host accepts first address;
2063 . First host gives temporary error to second address;
2064 . Second host offers TLS and a TLS session is established;
2065 . Second host accepts second address.
2066 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2067 and peerdn, if requested) that were in fact used only for the second
2070 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2071 attention to any local part prefix or suffix that was matched by the
2072 router that accepted the address. It now behaves in the same way as it
2073 does for delivery: the affixes are removed from the local part unless
2074 rcpt_include_affixes is set on the transport.
2076 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2077 timeout during the DATA phase of an incoming message.
2079 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2080 applied a patch sanctioned by the Sieve maintainer.
2082 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2083 the uid or gid is negative. A case of a negative gid caused this to be
2084 noticed. The fix allows for either to be negative.
2086 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2087 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2090 PH/36 The change PH/12 above was broken. Fixed it.
2092 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2093 the grounds that routing the same address twice would always produce the
2094 same answer. This might have been true once, but it is certainly no
2095 longer true now. Routing a child address may depend on the previous
2096 routing that produced that child. Some complicated redirection strategies
2097 went wrong when messages had multiple recipients, and made Exim's
2098 behaviour dependent on the order in which the addresses were given.
2100 I have moved the duplicate checking until after the routing is complete.
2101 Exim scans the addresses that are assigned to local and remote
2102 transports, and removes any duplicates. This means that more work will be
2103 done, as duplicates will always all be routed, but duplicates are
2104 presumably rare, so I don't expect this is of any significance.
2106 For deliveries to pipes, files, and autoreplies, the duplicate checking
2107 still happens during the routing process, since they are not going to be
2110 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2111 It corrects a timeout issue with spamd. This is Ian's comment: "The
2112 background is that sometimes spamd either never reads data from a
2113 connection it has accepted, or it never writes response data. The exiscan
2114 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2115 blindly assumes that writes won't block so it may never time out."
2117 PH/39 Allow G after quota size as well as K and M.
2119 PH/40 The value set for $authenticated_id in an authenticator may not contain
2120 binary zeroes or newlines because the value is written to log lines and
2121 to spool files. There was no check on this. Now the value is run through
2122 the string_printing() function so that such characters are converted to
2123 printable escape sequences.
2125 PH/41 $message_linecount is a new variable that contains the total number of
2126 lines in the message. Compare $body_linecount, which is the count for the
2129 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2130 bounce and delay warning messages, except in certain special cases, which
2133 (a) An SMTP error message from a remote host;
2134 (b) A message specified in a :fail: redirection;
2135 (c) A message specified in a "fail" command in a system filter;
2136 (d) A message specified in a FAIL return from the queryprogram router;
2137 (e) A message specified by the cannot_route_message router option.
2139 In these cases only, Exim does include the error details in bounce and
2140 warning messages. There are also a few cases where bland messages such
2141 as "unrouteable address" or "local delivery error" are given.
2143 PH/43 $value is now also set for the "else" part of a ${run expansion.
2145 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2146 being worked on, but at least Exim now implements the latest version to
2149 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2150 process to complete was treated as a delivery failure, a timeout while
2151 writing the message to the pipe was logged, but erroneously treated as a
2152 successful delivery. Such timeouts include transport filter timeouts. For
2153 consistency with the overall process timeout, these timeouts are now
2154 treated as errors, giving rise to delivery failures by default. However,
2155 there is now a new Boolean option for the pipe transport called
2156 timeout_defer, which, if set TRUE, converts the failures into defers for
2157 both kinds of timeout. A transport filter timeout is now identified in
2160 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2161 systems where "make" and "gmake" are different, calling "gmake" at top
2162 level broke things. I've arranged for the value of $(MAKE) to be passed
2163 from the Makefile to this script so that it can call the same version of
2167 A note about Exim versions 4.44 and 4.50
2168 ----------------------------------------
2170 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2171 changes of various kinds. As a consequence, a big documentation update was
2172 needed. This delayed the release for rather longer than seemed good, especially
2173 in the light of a couple of (minor) security issues. Therefore, the changes
2174 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2175 release. So 4.44 and 4.50 are in effect two different branches that both start
2178 I have left the 4.50 change log unchanged; it contains all the changes since
2179 4.43. The change log for 4.44 is below; many of its items are identical to
2180 those for 4.50. This seems to be the most sensible way to preserve the
2181 historical information.
2187 1. Minor wording change to the doc/README.SIEVE file.
2189 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2190 computation of the current number of files was incorrect.
2192 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2193 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2196 4. Give more explanation in the error message when the command for a transport
2197 filter fails to execute.
2199 5. There are several places where Exim runs a non-Exim command in a
2200 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2201 was being done only for the command run by the queryprogram router. It is
2202 now done for all such subprocesses. The other cases are: ${run, transport
2203 filters, and the commands run by the lmtp and pipe transports.
2205 6. Added CONFIGURE_GROUP build-time option.
2207 7. Some older OS have a limit of 256 on the maximum number of file
2208 descriptors. Exim was using setrlimit() to set 1000 as a large value
2209 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2210 systems. I've change it so that if it can't get 1000, it tries for 256.
2212 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2213 was an oversight, and furthermore, ever since the addition of extra
2214 controls (e.g. 4.43/32), the checks on when to allow different forms of
2215 "control" were broken. There should now be diagnostics for all cases when a
2216 control that does not make sense is encountered.
2218 9. Added the /retain_sender option to "control=submission".
2220 10. $recipients is now available in the predata ACL (oversight).
2222 11. Tidy the search cache before the fork to do a delivery from a message
2223 received from the command line. Otherwise the child will trigger a lookup
2224 failure and thereby defer the delivery if it tries to use (for example) a
2225 cached ldap connection that the parent has called unbind on.
2227 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2228 of $address_data from the recipient verification was clobbered by the
2229 sender verification.
2231 13. The value of address_data from a sender verification is now available in
2232 $sender_address_data in subsequent conditions in the ACL statement.
2234 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2236 15. Added a new option "connect=<time>" to callout options, to set a different
2239 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2240 was its contents. (It was OK if the option was not defined at all.)
2242 17. A "Completed" log line is now written for messages that are removed from
2243 the spool by the -Mrm option.
2245 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2246 information about exactly what failed.
2248 19. Added -dd to debug only the daemon process.
2250 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2251 handles timeouts, both on the server side and network timeouts. Renamed the
2252 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2254 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2255 It is now set to "smtps".
2257 22. $host_address is now set to the target address during the checking of
2258 ignore_target_hosts.
2260 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2261 being passed; this would have caused $sender_host_name to have been used if
2262 matching the list had actually called for a host name (not very likely,
2263 since this list is usually IP addresses). A host name is now passed as
2266 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2267 code that specifies a non-restarting handler (typically sigaction() in
2268 modern systems) in an attempt to fix a rare and obscure crash bug.
2270 25. Narrowed the window for a race in the daemon that could cause it to ignore
2271 SIGCHLD signals. This is not a major problem, because they are used only to
2272 wake it up if nothing else does.
2274 26. A malformed maildirsize file could cause Exim to calculate negative values
2275 for the mailbox size or file count. Odd effects could occur as a result.
2276 The maildirsize information is now recalculated if the size or filecount
2279 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2280 support for a long time. Removed HAVE_SYS_VFS_H.
2282 28. Installed the latest version of exipick from John Jetmore.
2284 29. In an address list, if the pattern was not a regular expression, an empty
2285 subject address (from a bounce message) matched only if the pattern was an
2286 empty string. Non-empty patterns were not even tested. This was the wrong
2287 because it is perfectly reasonable to use an empty address as part of a
2288 database query. An empty address is now tested by patterns that are
2289 lookups. However, all the other forms of pattern expect the subject to
2290 contain a local part and a domain, and therefore, for them, an empty
2291 address still always fails if the pattern is not itself empty.
2293 30. Exim went into a mad DNS loop when attempting to do a callout where the
2294 host was specified on an smtp transport, and looking it up yielded more
2295 than one IP address.
2297 31. Re-factored the code for checking spool and log partition space into a
2298 function that finds that data and another that does the check. The former
2299 is then used to implement four new variables: $spool_space, $log_space,
2300 $spool_inodes, and $log_inodes.
2302 32. The RFC2047 encoding function was originally intended for short strings
2303 such as real names; it was not keeping to the 75-character limit for
2304 encoded words that the RFC imposes. It now respects the limit, and
2305 generates multiple encoded words if necessary. To be on the safe side, I
2306 have increased the buffer size for the ${rfc2047: expansion operator from
2309 33. It is now permitted to omit both strings after an "if" condition; if the
2310 condition is true, the result is "true". As before, when the second string
2311 is omitted, a false condition yields an empty string. This makes it less
2312 cumbersome to write custom ACL and router conditions.
2314 34. Failure to deliver a bounce message always caused it to be frozen, even if
2315 there was an errors_to setting on the router. The errors_to setting is now
2318 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2319 canonical form (fully expanded) before being placed in
2320 $sender_host_address.
2322 36. The table in the code that translates DNS record types into text (T_A to
2323 "A" for instance) was missing entries for NS and CNAME. It is just possible
2324 that this could have caused confusion if both these types were looked up
2325 for the same domain, because the text type is used as part of Exim's
2326 per-process caching. But the chance of anyone hitting this buglet seems
2329 37. The dnsdb lookup has been extended in a number of ways.
2331 (1) There is a new type, "zns", which walks up the domain tree until it
2332 finds some nameserver records. It should be used with care.
2334 (2) There is a new type, "mxh", which is like "mx" except that it returns
2335 just the host names, not the priorities.
2337 (3) It is now possible to give a list of domains (or IP addresses) to be
2338 looked up. The behaviour when one of the lookups defers can be
2339 controlled by a keyword.
2341 (4) It is now possible to specify the separator character for use when
2342 multiple records are returned.
2344 38. The dnslists ACL condition has been extended: it is now possible to supply
2345 a list of IP addresses and/or domains to be looked up in a particular DNS
2348 39. Added log_selector=+queue_time_overall.
2350 40. When running the queue in the test harness, wait just a tad after forking a
2351 delivery process, to get repeatability of debugging output.
2353 41. Include certificate and key file names in error message when GnuTLS fails
2354 to set them up, because the GnuTLS error message doesn't include the name
2355 of the failing file when there is a problem reading it.
2357 42. Allow both -bf and -bF in the same test run.
2359 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2361 44. The "Exiscan patch" is now merged into the mainline Exim source.
2363 45. Sometimes the final signoff response after QUIT could fail to get
2364 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2365 < 0 before doing a fflush(). This bug looks as though it goes back to the
2366 introduction of TLS in release 3.20, but "sometimes" must have been rare
2367 because the tests only now provoked it.
2369 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2370 (this can affect the format of dates).
2372 47. exim_tidydb, when checking for the continued existence of a message for
2373 which it has found a message-specific retry record, was not finding
2374 messages that were in split spool directories. Consequently, it was
2375 deleting retry records that should have stayed in existence.
2377 48. Steve fixed some bugs in eximstats.
2379 49. The SPA authentication driver was not abandoning authentication and moving
2380 on to the next authenticator when an expansion was forced to fail,
2381 contradicting the general specification for all authenticators. Instead it
2382 was generating a temporary error. It now behaves as specified.
2384 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2385 (the order specifies the preference for clients). The order is now AES256,
2386 AES128, 3DES, ARCFOUR128.
2388 51. Small patch to Sieve code - explicitly set From: when generating an
2391 52. Exim crashed if a remote delivery caused a very long error message to be
2392 recorded - for instance if somebody sent an entire SpamAssassin report back
2393 as a large number of 550 error lines. This bug was coincidentally fixed by
2394 increasing the size of one of Exim's internal buffers (big_buffer) that
2395 happened as part of the Exiscan merge. However, to be on the safe side, I
2396 have made the code more robust (and fixed the comments that describe what
2399 53. Now that there can be additional text after "Completed" in log lines (if
2400 the queue_time_overall log selector is set), a one-byte patch to exigrep
2401 was needed to allow it to recognize "Completed" as not the last thing in
2404 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2405 patch that reportedly fixes this has been added. I am not expert enough to
2406 create a test for it. This is what the patch creator wrote:
2408 "I found a little strange behaviour of ldap code when working with
2409 Windows 2003 AD Domain, where users was placed in more than one
2410 Organization Units. When I tried to give exim partial DN, the exit code
2411 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2412 But simultaneously result of request was absolutely normal ldap result,
2413 so I produce this patch..."
2415 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2416 REFERENCE, so I have modified the code to exclude the patch when that macro
2419 55. Some experimental protocols are using DNS PTR records for new purposes. The
2420 keys for these records are domain names, not reversed IP addresses. The
2421 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2422 leaves it alone. Component reversal etc. now happens only for IP addresses.
2425 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2427 57. Double the size of the debug message buffer (to 2048) so that more of very
2428 long debug lines gets shown.
2430 58. The exicyclog utility now does better if the number of log files to keep
2431 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2433 59. Two changes related to the smtp_active_hostname option:
2435 (1) $smtp_active_hostname is now available as a variable.
2436 (2) The default for smtp_banner uses $smtp_active_hostname instead
2437 of $primary_hostname.
2439 60. The host_aton() function is supposed to be passed a string that is known
2440 to be a valid IP address. However, in the case of IPv6 addresses, it was
2441 not checking this. This is a hostage to fortune. Exim now panics and dies
2442 if the condition is not met. A case was found where this could be provoked
2443 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2444 components; fortuitously, this particular loophole had already been fixed
2445 by change 4.50/55 above.
2447 If there are any other similar loopholes, the new check in host_aton()
2448 itself should stop them being exploited. The report I received stated that
2449 data on the command line could provoke the exploit when Exim was running as
2450 exim, but did not say which command line option was involved. All I could
2451 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2452 running as the user.
2455 61. There was a buffer overflow vulnerability in the SPA authentication code
2456 (which came originally from the Samba project). I have added a test to the
2457 spa_base64_to_bits() function which I hope fixes it.
2460 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2461 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2463 63. The daemon start-up calls getloadavg() while still root for those OS that
2464 need the first call to be done as root, but it missed one case: when
2465 deliver_queue_load_max is set with deliver_drop_privilege. This is
2466 necessary for the benefit of the queue runner, because there is no re-exec
2467 when deliver_drop_privilege is set.
2469 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2470 This has been fixed.
2472 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2473 was in use, was not putting the data itself into the right store pool;
2474 consequently, it could be overwritten for a subsequent message in the same
2475 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2478 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2480 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2481 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2482 the function were treating the return as a boolean value, which happened to
2483 work because 0=false and not-0=true, but is not correct code.
2485 68. The host_aton() function was not handling scoped IPv6 addresses (those
2486 with, for example, "%eth0" on the end) correctly.
2488 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2489 negated items (that is, ~something) in unsigned ints. Some compilers
2490 apparently mutter when there is no cast.
2492 70. If an address verification called from an ACL failed, and did not produce a
2493 user-specific message (i.e. there was only a "system" message), nothing was
2494 put in $acl_verify_message. In this situation, it now puts the system
2497 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2498 session; change 4.31/43 added the unwanted input to the log line - except
2499 that it did not do this in the start of session case. It now does.
2501 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2502 This is wrong and can cause the other end to generate a synchronization
2503 error if it is another Exim or anything else that does the synchronization
2504 check. A QUIT command is no longer sent after a timeout.
2506 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2507 during host lookups.
2509 74. The defer_ok option of callout verification was not working if it was used
2510 when verifying addresses in header lines, that is, for this case:
2512 verify = header_sender/callout=defer_ok
2514 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2515 those file descriptors could be used for SMTP connections. If anything
2516 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2517 could be sent to the SMTP client, causing chaos. The daemon now opens
2518 stdin, stdout, and stderr to /dev/null when it puts itself into the
2521 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2522 log by default. The user can override this with suitable Perl magic.
2524 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2525 the log message when discard triggers, was not working for the DATA ACL or
2526 for the non-SMTP ACL.
2528 78. Error message wording change in sieve.c.
2530 79. If smtp_accept_max_per_host was set, the number of connections could be
2531 restricted to fewer than expected, because the daemon was trying to set up
2532 a new connection before checking whether the processes handling previous
2533 connections had finished. The check for completed processes is now done
2534 earlier. On busy systems, this bug wouldn't be noticed because something
2535 else would have woken the daemon, and it would have reaped the completed
2538 80. If a message was submitted locally by a user whose login name contained one
2539 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2540 It caused a spool format error. I have fixed the spool reading code. A
2541 related problem was that the "from" clause in the Received: line became
2542 illegal because of the space(s). It is now covered by ${quote_local_part.
2544 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2547 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2549 83. Patch from Sieve maintainer to fix unterminated string problem in
2550 "vacation" handling.
2552 84. Some minor changes to the Linux configuration files to help with other
2553 OS variants using glibc.
2555 85. One more patch for Sieve to update vacation handling to latest spec.
2558 ----------------------------------------------------
2559 See the note above about the 4.44 and 4.50 releases.
2560 ----------------------------------------------------
2566 1. Change 4.43/35 introduced a bug that caused file counts to be
2567 incorrectly computed when quota_filecount was set in an appendfile
2570 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2571 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2574 3. Give more explanation in the error message when the command for a transport
2575 filter fails to execute.
2577 4. There are several places where Exim runs a non-Exim command in a
2578 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2579 was being done only for the command run by the queryprogram router. It is
2580 now done for all such subprocesses. The other cases are: ${run, transport
2581 filters, and the commands run by the lmtp and pipe transports.
2583 5. Some older OS have a limit of 256 on the maximum number of file
2584 descriptors. Exim was using setrlimit() to set 1000 as a large value
2585 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2586 systems. I've change it so that if it can't get 1000, it tries for 256.
2588 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2589 was an oversight, and furthermore, ever since the addition of extra
2590 controls (e.g. 4.43/32), the checks on when to allow different forms of
2591 "control" were broken. There should now be diagnostics for all cases when a
2592 control that does not make sense is encountered.
2594 7. $recipients is now available in the predata ACL (oversight).
2596 8. Tidy the search cache before the fork to do a delivery from a message
2597 received from the command line. Otherwise the child will trigger a lookup
2598 failure and thereby defer the delivery if it tries to use (for example) a
2599 cached ldap connection that the parent has called unbind on.
2601 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2602 of $address_data from the recipient verification was clobbered by the
2603 sender verification.
2605 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2606 was its contents. (It was OK if the option was not defined at all.)
2608 11. A "Completed" log line is now written for messages that are removed from
2609 the spool by the -Mrm option.
2611 12. $host_address is now set to the target address during the checking of
2612 ignore_target_hosts.
2614 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2615 being passed; this would have caused $sender_host_name to have been used if
2616 matching the list had actually called for a host name (not very likely,
2617 since this list is usually IP addresses). A host name is now passed as
2620 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2621 code that specifies a non-restarting handler (typically sigaction() in
2622 modern systems) in an attempt to fix a rare and obscure crash bug.
2624 15. Narrowed the window for a race in the daemon that could cause it to ignore
2625 SIGCHLD signals. This is not a major problem, because they are used only to
2626 wake it up if nothing else does.
2628 16. A malformed maildirsize file could cause Exim to calculate negative values
2629 for the mailbox size or file count. Odd effects could occur as a result.
2630 The maildirsize information is now recalculated if the size or filecount
2633 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2634 support for a long time. Removed HAVE_SYS_VFS_H.
2636 18. Updated exipick to current release from John Jetmore.
2638 19. Allow an empty sender to be matched against a lookup in an address list.
2639 Previously the only cases considered were a regular expression, or an
2642 20. Exim went into a mad DNS lookup loop when doing a callout where the
2643 host was specified on the transport, if the DNS lookup yielded more than
2646 21. The RFC2047 encoding function was originally intended for short strings
2647 such as real names; it was not keeping to the 75-character limit for
2648 encoded words that the RFC imposes. It now respects the limit, and
2649 generates multiple encoded words if necessary. To be on the safe side, I
2650 have increased the buffer size for the ${rfc2047: expansion operator from
2653 22. Failure to deliver a bounce message always caused it to be frozen, even if
2654 there was an errors_to setting on the router. The errors_to setting is now
2657 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2658 canonical form (fully expanded) before being placed in
2659 $sender_host_address.
2661 24. Updated eximstats to version 1.33
2663 25. Include certificate and key file names in error message when GnuTLS fails
2664 to set them up, because the GnuTLS error message doesn't include the name
2665 of the failing file when there is a problem reading it.
2667 26. Expand error message when OpenSSL has problems setting up cert/key files.
2670 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2671 (this can affect the format of dates).
2673 28. exim_tidydb, when checking for the continued existence of a message for
2674 which it has found a message-specific retry record, was not finding
2675 messages that were in split spool directories. Consequently, it was
2676 deleting retry records that should have stayed in existence.
2678 29. eximstats updated to version 1.35
2679 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2680 1.35 - bugfix such that pie charts by volume are generated correctly
2682 30. The SPA authentication driver was not abandoning authentication and moving
2683 on to the next authenticator when an expansion was forced to fail,
2684 contradicting the general specification for all authenticators. Instead it
2685 was generating a temporary error. It now behaves as specified.
2687 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2688 (the order specifies the preference for clients). The order is now AES256,
2689 AES128, 3DES, ARCFOUR128.
2691 31. Small patch to Sieve code - explicitly set From: when generating an
2694 32. Exim crashed if a remote delivery caused a very long error message to be
2695 recorded - for instance if somebody sent an entire SpamAssassin report back
2696 as a large number of 550 error lines. This bug was coincidentally fixed by
2697 increasing the size of one of Exim's internal buffers (big_buffer) that
2698 happened as part of the Exiscan merge. However, to be on the safe side, I
2699 have made the code more robust (and fixed the comments that describe what
2702 33. Some experimental protocols are using DNS PTR records for new purposes. The
2703 keys for these records are domain names, not reversed IP addresses. The
2704 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2705 leaves it alone. Component reversal etc. now happens only for IP addresses.
2708 34. The host_aton() function is supposed to be passed a string that is known
2709 to be a valid IP address. However, in the case of IPv6 addresses, it was
2710 not checking this. This is a hostage to fortune. Exim now panics and dies
2711 if the condition is not met. A case was found where this could be provoked
2712 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2713 components; fortuitously, this particular loophole had already been fixed
2714 by change 4.50/55 or 4.44/33 above.
2716 If there are any other similar loopholes, the new check in host_aton()
2717 itself should stop them being exploited. The report I received stated that
2718 data on the command line could provoke the exploit when Exim was running as
2719 exim, but did not say which command line option was involved. All I could
2720 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2721 running as the user.
2724 35. There was a buffer overflow vulnerability in the SPA authentication code
2725 (which came originally from the Samba project). I have added a test to the
2726 spa_base64_to_bits() function which I hope fixes it.
2729 36. The daemon start-up calls getloadavg() while still root for those OS that
2730 need the first call to be done as root, but it missed one case: when
2731 deliver_queue_load_max is set with deliver_drop_privilege. This is
2732 necessary for the benefit of the queue runner, because there is no re-exec
2733 when deliver_drop_privilege is set.
2735 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2736 was in use, was not putting the data itself into the right store pool;
2737 consequently, it could be overwritten for a subsequent message in the same
2738 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2741 38. Sometimes the final signoff response after QUIT could fail to get
2742 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2743 < 0 before doing a fflush(). This bug looks as though it goes back to the
2744 introduction of TLS in release 3.20, but "sometimes" must have been rare
2745 because the tests only now provoked it.
2751 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2752 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2753 Now it returns an integer. A number of calls were still expecting a T/F
2754 return. Fortuitously, in all cases, the tests worked in OK situations,
2755 which is the norm. However, things would have gone wrong on any write
2756 failures on the smtp file descriptor. This function is used when sending
2757 messages over SMTP and also when doing verify callouts.
2759 2. When Exim is called to do synchronous delivery of a locally submitted
2760 message (the -odf or -odi options), it no longer closes stderr before doing
2763 3. Implemented the mua_wrapper option.
2765 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2767 5. Implemented the functions header_remove(), header_testname(),
2768 header_add_at_position(), and receive_remove_recipient(), and exported them
2771 6. If an ACL "warn" statement specified the addition of headers, Exim already
2772 inserted X-ACL-Warn: at the start if there was no header name. However, it
2773 was not making this test for the second and subsequent header lines if
2774 there were newlines in the string. This meant that an invalid header could
2775 be inserted if Exim was badly configured.
2777 7. Allow an ACL "warn" statement to add header lines at the start or after all
2778 the Received: headers, as well as at the end.
2780 8. Added the rcpt_4xx retry error code.
2782 9. Added postmaster_mailfrom=xxx to callout verification option.
2784 10. Added mailfrom=xxxx to the callout verification option, for verify=
2787 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2788 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2790 12. Inserted some casts to stop certain compilers warning when using pointer
2791 differences as field lengths or precisions in printf-type calls (mostly
2792 affecting debugging statements).
2794 13. Added optional readline() support for -be (dynamically loaded).
2796 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2797 same clock tick as a message's arrival, so that its received time was the
2798 same as the "first fail" time on the retry record, and that message
2799 remained on the queue past the ultimate address timeout, every queue runner
2800 would try a delivery (because it was past the ultimate address timeout) but
2801 after another failure, the ultimate address timeout, which should have then
2802 bounced the address, did not kick in. This was a "< instead of <=" error;
2803 in most cases the first failure would have been in the next clock tick
2804 after the received time, and all would be well.
2806 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2807 being recognized when the domain list was tested by the match_domain
2808 condition in an expansion string.
2810 16. Added the ${str2b64: operator.
2812 17. Exim was always calling setrlimit() to set a large limit for the number of
2813 processes, without checking whether the existing limit was already
2814 adequate. (It did check for the limit on file descriptors.) Furthermore,
2815 errors from getrlimit() and setrlimit() were being ignored. Now they are
2816 logged to the main and panic logs, but Exim does carry on, to try to do its
2817 job under whatever limits there are.
2819 18. Imported PCRE 5.0.
2821 19. Trivial typo in log message " temporarily refused connection" (the leading
2824 20. If the log selector return_path_on_delivery was set and an address was
2825 redirected to /dev/null, the delivery process crashed because it assumed
2826 that a return path would always be set for a "successful" delivery. In this
2827 case, the whole delivery is bypassed as an optimization, and therefore no
2830 21. Internal re-arrangement: the function for sending a challenge and reading
2831 a response while authentication was assuming a zero-terminated challenge
2832 string. It's now changed to take a pointer and a length, to allow for
2833 binary data in such strings.
2835 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2837 23. Exim was not respecting finduser_retries when seeking the login of the
2838 uid under which it was called; it was always trying 10 times. (The default
2839 setting of finduser_retries is zero.) Also, it was sleeping after the final
2840 failure, which is pointless.
2842 24. Implemented tls_on_connect_ports.
2844 25. Implemented acl_smtp_predata.
2846 26. If the domain in control=submission is set empty, Exim assumes that the
2847 authenticated id is a complete email address when it generates From: or
2848 Sender: header lines.
2850 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2851 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2852 chown and chgrp in /bin and hostname in /usr/bin.
2854 28. Exim was keeping the "process log" file open after each use, just as it
2855 does for the main log. This opens the possibility of it remaining open for
2856 long periods when the USR1 signal hits a daemon. Occasional processlog
2857 errors were reported, that could have been caused by this. Anyway, it seems
2858 much more sensible not to leave this file open at all, so that is what now
2861 29. The long-running daemon process does not normally write to the log once it
2862 has entered its main loop, and it closes the log before doing so. This is
2863 so that log files can straightforwardly be renamed and moved. However,
2864 there are a couple of unusual error situations where the daemon does write
2865 log entries, and I had neglected to close the log afterwards.
2867 30. The text of an SMTP error response that was received during a remote
2868 delivery was being truncated at 512 bytes. This is too short for some of
2869 the long messages that one sometimes sees. I've increased the limit to
2872 31. It is now possible to make retry rules that apply only when a message has a
2873 specific sender, in particular, an empty sender.
2875 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2876 it possible to be selective about when SMTP synchronization is enforced.
2878 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2880 32. Implemented hosts_connection_nolog.
2882 33. Added an ACL for QUIT.
2884 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2887 35. Added mailbox_size and mailbox_filecount to appendfile.
2889 36. Added control = no_multiline_responses to ACLs.
2891 37. There was a bug in the logic of the code that waits for the clock to tick
2892 in the case where the clock went backwards by a substantial amount such
2893 that the microsecond fraction of "now" was more than the microsecond
2894 fraction of "then" (but the whole seconds number was less).
2896 38. Added support for the libradius Radius client library this is found on
2897 FreeBSD (previously only the radiusclient library was supported).
2903 1. When certain lookups returned multiple values in the form name=value, the
2904 quoting of the values was not always being done properly. Specifically:
2905 (a) If the value started with a double quote, but contained no whitespace,
2907 (b) If the value contained whitespace other than a space character (i.e.
2908 tabs or newlines or carriage returns) it was not quoted.
2909 This fix has been applied to the mysql and pgsql lookups by writing a
2910 separate quoting function and calling it from the lookup code. The fix
2911 should probably also be applied to nisplus, ibase and oracle lookups, but
2912 since I cannot test any of those, I have not disturbed their existing code.
2914 2. A hit in the callout cache for a specific address caused a log line with no
2915 reason for rejecting RCPT. Now it says "Previous (cached) callout
2916 verification failure".
2918 3. There was an off-by-one bug in the queryprogram router. An over-long
2919 return line was truncated at 256 instead of 255 characters, thereby
2920 overflowing its buffer with the terminating zero. As well as fixing this, I
2921 have increased the buffer size to 1024 (and made a note to document this).
2923 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2924 when Exim is waiting for an SMTP response from a remote server, Exim
2925 restarts its select() call on the socket, thereby resetting its timeout.
2926 This is not a problem when such interrupts are rare. Somebody set up a cron
2927 job to run exiwhat every 2 minutes, which is less than the normal select()
2928 timeout (5 or 10 minutes). This meant that the select() timeout never
2929 kicked in because it was always reset. I have fixed this by comparing the
2930 time when an interrupt arrives with the time at the start of the first call
2931 to select(). If more time than the timeout has elapsed, the interrupt is
2932 treated as a timeout.
2934 5. Some internal re-factoring in preparation for the addition of Sieve
2935 extensions (by MH). In particular, the "personal" test is moved to a
2936 separate function, and given an option for scanning Cc: and Bcc: (which is
2937 not set for Exim filters).
2939 6. When Exim created an email address using the login of the caller as the
2940 local part (e.g. when creating a From: or Sender: header line), it was not
2941 quoting the local part when it contained special characters such as @.
2943 7. Installed new OpenBSD configuration files.
2945 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2946 try to make them clearer.
2948 9. Callout options, other than the timeout value, were being ignored when
2949 verifying sender addresses in header lines. For example, when using
2951 verify = header_sender/callout=no_cache
2953 the cache was (incorrectly) being used.
2955 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2956 only the Cygwin environment.
2958 11. When return_path_on_delivery was set as a log selector, if different remote
2959 addresses in the same message used different return paths and parallel
2960 remote delivery occurred, the wrong values would sometimes be logged.
2961 (Whenever a remote delivery process finished, the return path value from
2962 the most recently started remote delivery process was logged.)
2964 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2965 lines when AUTH and/or TLS are in use. This is the "received protocol"
2966 field. Exim used to use "asmtp" for authenticated SMTP, without any
2967 indication (in the protocol name) for TLS use. Now it follows the RFC and
2968 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2969 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2970 names appear in log lines as well as in Received: header lines.
2972 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2973 extensions, and comparison tests, and to fix some bugs.
2975 14. Changes to the "personal" filter test:
2977 (1) The test was buggy in that it was just doing the equivalent of
2978 "contains" tests on header lines. For example, if a user's address was
2979 anne@some.where, the "personal" test would incorrectly be true for
2981 To: susanne@some.where
2983 This test is now done by extracting each address from the header in turn,
2984 and checking the entire address. Other tests that are part of "personal"
2985 are now done using regular expressions (for example, to check local parts
2986 of addresses in From: header lines).
2988 (2) The list of non-personal local parts in From: addresses has been
2989 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2990 taken from the Sieve specification recommendations.
2992 (3) If the message contains any header line starting with "List-" it is
2993 treated as non-personal.
2995 (4) The test for "circular" in the Subject: header line has been removed
2996 because it now seems ill-conceived.
2998 15. Minor typos in src/EDITME comments corrected.
3000 16. Installed latest exipick from John Jetmore.
3002 17. If headers_add on a router specified a text string that was too long for
3003 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3004 of string_sprintf() is now avoided.
3006 18. $message_body_size was not set (it was always zero) when running the DATA
3007 ACL and the local_scan() function.
3009 19. For the "mail" command in an Exim filter, no default was being set for
3010 the once_repeat time, causing a random time value to be used if "once" was
3011 specified. (If the value happened to be <= 0, no repeat happened.) The
3012 default is now 0s, meaning "never repeat". The "vacation" command was OK
3013 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3014 (I found it when inspecting the code).
3016 20. There is now an overall timeout for performing a callout verification. It
3017 defaults to 4 times the callout timeout, which applies to individual SMTP
3018 commands during the callout. The overall timeout applies when there is more
3019 than one host that can be tried. The timeout is checked before trying the
3020 next host. This prevents very long delays if there are a large number of
3021 hosts and all are timing out (e.g. when the network connections are timing
3022 out). The value of the overall timeout can be changed by specifying an
3023 additional sub-option for "callout", called "maxwait". For example:
3025 verify = sender/callout=5s,maxwait=20s
3027 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3028 to the end before writing, but this should make it even safer).
3030 22. Exim was forgetting that it had advertised PIPELINING for the second and
3031 subsequent messages on an SMTP connection. It was also not resetting its
3032 memory on STARTTLS and an internal HELO.
3034 23. When Exim logs an SMTP synchronization error within a session, it now
3035 records whether PIPELINING has been advertised or not.
3037 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3038 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3039 rather than long int.
3041 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3043 26. Added the never_mail option to autoreply.
3049 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3050 crash if the getsockname() call failed; this can happen if a connection is
3051 closed very soon after it is established. The problem was simply in the
3052 order in which certain operations were done, causing Exim to try to write
3053 to the SMTP stream before it had set up the file descriptor. The bug has
3054 been fixed by making things happen in the correct order.
3060 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3061 before the connection was closed, thus losing the rejection response.
3063 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3064 some early Solaris releases, but causes trouble in current releases where
3065 socklen_t is defined.
3067 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3070 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3073 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3074 that releasing the top of it at the end releases what was used for sub-
3075 expansions (unless the block got too big). However, discard this block if
3076 the first thing is a variable or header, so that we can use its block when
3077 it is dynamic (useful for very large $message_headers, for example).
3079 6. Lookups now cache *every* query, not just the most recent. A new, separate
3080 store pool is used for this. It can be recovered when all lookup caches are
3081 flushed. Lookups now release memory at the end of their result strings.
3082 This has involved some general refactoring of the lookup sources.
3084 7. Some code has been added to the store_xxx() functions to reduce the amount
3085 of flapping under certain conditions.
3087 8. log_incoming_interface used to affect only the <= reception log lines. Now
3088 it causes the local interface and port to be added to several more SMTP log
3089 lines, for example "SMTP connection from", and rejection lines.
3091 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3093 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3095 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3096 could be overwritten at the end of the current message (or the start of a
3097 new message if it was set in a HELO ACL). The value is now preserved for
3098 the duration of the SMTP connection.
3100 12. If a transport had a headers_rewrite setting, and a matching header line
3101 contained an unqualified address, that address was qualified, even if it
3102 did not match any rewriting rules. The underlying bug was that the values
3103 of the flags that permit the existence of unqualified sender and recipient
3104 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3105 non-local messages, and by -bnq for local messages) were not being
3106 preserved with the message after it was received.
3108 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3109 "next input=" as part of the text comprising the host identity instead of
3110 the correct text. The code was using the same buffer for two different
3111 strings. However, depending on which order the printing function evaluated
3112 its arguments, the bug did not always show up. Under Linux, for example, my
3113 test suite worked just fine.
3115 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3116 doesn't work with some older versions of Perl. It has been changed to "my",
3117 which in any case is probably the better facility to use.
3119 15. A really picky compiler found some instances of statements for creating
3120 error messages that either had too many or two few arguments for the format
3123 16. The size of the buffer for calls to the DNS resolver has been increased
3124 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3125 for addresses that have a lot of PTR records. This alleviates a problem; it
3126 does not fully solve it.
3128 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3129 buffer now truncates the list and logs the incident, which is the same
3130 action as happens when Exim is looking up a host name and its aliases.
3131 Previously in this situation something unpredictable would happen;
3132 sometimes it was "internal error: store_reset failed".
3134 18. If a server dropped the connection unexpectedly when an Exim client was
3135 using GnuTLS and trying to read a response, the client delivery process
3136 crashed while trying to generate an error log message.
3138 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3139 string, for example:
3141 warn message = H1: something\nH2: something
3143 the text was added as a single header line from Exim's point of view
3144 though it ended up OK in the delivered message. However, searching for the
3145 second and subsequent header lines using $h_h2: did not work. This has been
3146 fixed. Similarly, if a system filter added multiple headers in this way,
3147 the routers could not see them.
3149 20. Expanded the error message when iplsearch is called with an invalid key to
3150 suggest using net-iplsearch in a host list.
3152 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3153 ACLs are no longer actually imposed (and a message to that effect is
3156 22. If a "gecos" field in a passwd entry contained escaped characters, in
3157 particular, if it contained a \" sequence, Exim got it wrong when building
3158 a From: or a Sender: header from that name. A second bug also caused
3159 incorrect handling when an unquoted " was present following a character
3160 that needed quoting.
3162 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3163 was not being matched caselessly.
3165 24. Arranged for all hyphens in the exim.8 source to be escaped with
3168 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3169 properly. Recipient callout cache records were still being keyed to include
3170 the sender, even when use_sender was set false. This led to far more
3171 callouts that were necessary. The sender is no longer included in the key
3172 when use_sender is false.
3174 26. Added "control = submission" modifier to ACLs.
3176 27. Added the ${base62d: operator to decode base 62 numbers.
3178 28. dnsdb lookups can now access SRV records.
3180 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3181 the configuration file.
3183 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3184 (-v) mode. This makes the output for a verbose queue run more intelligible.
3186 31. Added a use_postmaster feature to recipient callouts.
3188 32. Added the $body_zerocount variable, containing the number of binary zero
3189 bytes in the message body.
3191 33. The time of last modification of the "new" subdirectory is now used as the
3192 "mailbox time last read" when there is a quota error for a maildir
3195 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3197 35. Added +ignore_unknown as a special item in host lists.
3199 36. Code for decoding IPv6 addresses in host lists is now included, even if
3200 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3201 address was recognized as an IP address, but was then not correctly decoded
3202 into binary, causing unexpected and incorrect effects when compared with
3209 1. Very minor rewording of debugging text in manualroute to say "list of
3210 hosts" instead of "hostlist".
3212 2. If verify=header_syntax was set, and a header line with an unqualified
3213 address (no domain) and a large number of spaces between the end of the
3214 name and the colon was received, the reception process suffered a buffer
3215 overflow, and (when I tested it) crashed. This was caused by some obsolete
3216 code that should have been removed. The fix is to remove it!
3218 3. When running in the test harness, delay a bit after writing a bounce
3219 message to get a bit more predictability in the log output.
3221 4. Added a call to search_tidyup() just before forking a reception process. In
3222 theory, someone could use a lookup in the expansion of smtp_accept_max_
3223 per_host which, without the tidyup, could leave open a database connection.
3225 5. Added the variables $recipient_data and $sender_data which get set from a
3226 lookup success in an ACL "recipients" or "senders" condition, or a router
3227 "senders" option, similar to $domain_data and $local_part_data.
3229 6. Moved the writing of debug_print from before to after the "senders" test
3232 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3233 problems for message scanning, either using a data ACL, or using
3234 local_scan() because the Received: header was not generated till after they
3235 were called (in order to set the time as the time of reception completion).
3236 I have revised the way this works. The header is now generated after the
3237 body is received, but before the ACL or local_scan() are called. After they
3238 are run, the timestamp in the header is updated.
3244 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3245 before starting a queue runner without re-exec. This happened only when
3246 deliver_drop_privilege was set or when the Exim user was set to root. The
3247 effect of the bug was that timeouts during subsequent deliveries caused
3248 crashes instead of being properly handled. The handler is now left at its
3249 default (and expected) setting.
3251 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3252 message, again when deliver_drop_privilege is set or Exim is run as root.
3253 The bug described in (1) was not present in this case, but the tidying up
3254 of the other signals was missing. I have made the two cases consistent.
3256 3. The ignore_target_hosts setting on a manualroute router was being ignored
3257 for hosts that were looked up using the /MX notation.
3259 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3262 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3263 operated on the sender address. After changing the $sender_address to <>
3264 for the sender address verify, Exim was re-instated it as the original
3265 (before rewriting) address, but remembering that it had rewritten it, so it
3266 wasn't rewriting it again. This bug also had the effect of breaking the
3267 sender address verification caching when the sender address was rewritten.
3269 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3270 This has been changed so that if the ip literal address matches
3271 ignore_target_hosts, the router declines.
3273 7. Added expansion conditions match_domain, match_address, and match_local_
3274 part (NOT match_host).
3276 8. The placeholder for the Received: header didn't have a length field set.
3278 9. Added code to Exim itself and to exim_lock to test for a specific race
3279 condition that could lead to file corruption when using MBX delivery. The
3280 issue is with the lockfile that is created in /tmp. If this file is removed
3281 after a process has opened it but before that process has acquired a lock,
3282 there is the potential for a second process to recreate the file and also
3283 acquire a lock. This could lead to two Exim processes writing to the file
3284 at the same time. The added code performs the same test as UW imapd; it
3285 checks after acquiring the lock that its file descriptor still refers to
3286 the same named file.
3288 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3289 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3292 11. Added the smtp_active_hostname option. If used, this will typically be made
3293 to depend on the incoming interface address. Because $interface_address is
3294 not set up until the daemon has forked a reception process, error responses
3295 that can happen earlier (such as "too many connections") no longer contain
3298 12. If an expansion in a condition on a "warn" statement fails because a lookup
3299 defers, the "warn" statement is abandoned, and the next ACL statement is
3300 processed. Previously this caused the whole ACL to be aborted.
3302 13. Added the iplsearch lookup type.
3304 14. Added ident_timeout as a log selector.
3306 15. Added tls_certificate_verified as a log selector.
3308 16. Added a global option tls_require_ciphers (compare the smtp transport
3309 option of the same name). This controls incoming TLS connections.
3311 17. I finally figured out how to make tls_require_ciphers do a similar thing
3312 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3313 before starting the TLS session.
3315 18. Tabs are now shown as \t in -bP output.
3317 19. If the log selector return_path_on_delivery was set, Exim crashed when
3318 bouncing a message because it had too many Received: header lines.
3320 20. If two routers both had headers_remove settings, and the first one included
3321 a superfluous trailing colon, the final name in the first list and the
3322 first name in the second list were incorrectly joined into one item (with a
3323 colon in the middle).
3329 1. Added -C and -D options to the exinext utility, mainly to make it easier
3330 to include in the automated testing, but these could be helpful when
3331 multiple configurations are in use.
3333 2. The exinext utility was not formatting the output nicely when there was
3334 an alternate port involved in the retry record key, nor when there was a
3335 message id as well (for retries that were specific to a specific message
3336 and a specific host). It was also confused by IPv6 addresses, because of
3337 the additional colons they contain. I have fixed the IPv4 problem, and
3338 patched it up to do a reasonable job for IPv6.
3340 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3341 delivery, the log line now contains "pipelined" if PIPELINING was used.
3343 4. An SMTP transport process used to panic and die if the bind() call to set
3344 an explicit outgoing interface failed. This has been changed; it is now
3345 treated in the same way as a connect() failure.
3347 5. A reference to $sender_host_name in the part of a conditional expansion
3348 that was being skipped was still causing a DNS lookup. This no longer
3351 6. The def: expansion condition was not recognizing references to header lines
3352 that used bh_ and bheader_.
3354 7. Added the _cache feature to named lists.
3356 8. The code for checking quota_filecount in the appendfile transport was
3357 allowing one more file than it should have been.
3359 9. For compatibility with Sendmail, the command line option
3367 and sets the incoming protocol and host name (for trusted callers). The
3368 host name and its colon can be omitted when only the protocol is to be set.
3369 Note the Exim already has two private options, -pd and -ps, that refer to
3370 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3371 "s", but I don't think that's a major issue.
3373 10. A number of refactoring changes to the code, none of which should affect
3376 (a) The number of logging options was getting close to filling up the
3377 32-bit word that was used as a bit map. I have split them into two classes:
3378 those that are passed in the argument to log_write(), and those that are
3379 only ever tested independently outside of that function. These are now in
3380 separate 32-bit words, so there is plenty of room for expansion again.
3381 There is no change in the user interface or the logging behaviour.
3383 (b) When building, for example, log lines, the code previously used a
3384 macro that called string_cat() twice, in order to add two strings. This is
3385 not really sufficiently general. Furthermore, there was one instance where
3386 it was actually wrong because one of the argument was used twice, and in
3387 one call a function was used. (As it happened, calling the function twice
3388 did not affect the overall behaviour.) The macro has been replaced by a
3389 function that can join an arbitrary number of extra strings onto a growing
3392 (c) The code for expansion conditions now uses a table and a binary chop
3393 instead of a serial search (which was left over from when there were very
3394 few conditions). Also, it now recognizes conditions like "pam" even when
3395 the relevant support is not compiled in: a suitably worded error message is
3396 given if an attempt is made to use such a condition.
3398 11. Added ${time_interval:xxxxx}.
3400 12. A bug was causing one of the ddress fields not to be passed back correctly
3401 from remote delivery subprocesses. The field in question was not being
3402 subsequently used, so this caused to problems in practice.
3404 13. Added new log selectors queue_time and deliver_time.
3406 14. Might have fixed a bug in maildirsizefile handling that threw up
3407 "unexpected character" debug warnings, and recalculated the data
3408 unnecessarily. In any case, I expanded the warning message to give more
3411 15. Added the message "Restricted characters in address" to the statements in
3412 the default ACL that block characters like @ and % in local parts.
3414 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3415 Three changes have been made:
3417 (a) There was a serious bug; a negative response to MAIL caused the whole
3418 recipient domain to be cached as invalid, thereby blocking all messages
3419 to all local parts at the same domain, from all senders. This bug has
3420 been fixed. The domain is no longer cached after a negative response to
3421 MAIL if the sender used is not empty.
3423 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3426 (c) A new callout option, "use_sender" has been added for people who want
3427 the modified behaviour.
3433 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3436 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3437 indeed breaks things for older releases.
3439 3. Added additional logging to the case where there is a problem reading data
3440 from a filter that is running in a subprocess using a pipe, in order to
3441 try to track down a specific problem.
3443 4. Testing facility fudge: when running in the test harness and attempting
3444 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3445 getting "No route to host". Convert this to a timeout.
3447 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3450 6. Some OS don't have socklen_t but use size_t instead. This affects the
3451 fifth argument of getsockopt() amongst other things. This is now
3452 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3453 can be set for individual OS. I have set it for SunOS5, OSF1, and
3454 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3455 some earlier ones do not.
3457 7. Change 4.30/15 was not doing the test caselessly.
3459 8. The standard form for an IPv6 address literal was being rejected by address
3460 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3461 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3462 this, as well as the form without the "IPv6" on the front (but only when
3463 address literals are enabled, of course).
3465 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3467 10. Exim crashed if a message with an empty sender address specified by -f
3468 encountered a router with an errors_to setting. This could be provoked only
3469 by a command such as
3473 where an empty string was supplied; "<>" did not hit this bug.
3475 11. Installed PCRE release 4.5.
3477 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3478 remained set. It is now erased.
3480 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3481 times from message ids (which are base 36 rather than the normal 62).
3483 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3484 were being counted as actual protocol errors, and logged if the log
3485 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3486 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3487 and DATA following a set of rejected RCPTs do not count as protocol errors.
3488 In other words, Exim assumes they were pipelined, though this may not
3489 actually be the case. Of course, in all cases the client gets an
3490 appropriate error code.
3492 15. If a lookup fails in an ACL condition, a message about the failure may
3493 be available; it is used if testing the ACL cannot continue, because most
3494 such messages specify what the cause of the deferral is. However, some
3495 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3496 that caused an old message to be retained and used if a later statement
3497 caused a defer, replacing the real cause of the deferral.
3499 16. If an IP address had so many PTR records that the DNS lookup buffer
3500 was not large enough to hold them, Exim could crash while trying to process
3501 the truncated data. It now detects and logs this case.
3503 17. Further to 4.21/58, another change has been made: if (and only if) the
3504 first line of a message (the first header line) ends with CRLF, a bare LF
3505 in a subsequent header line has a space inserted after it, so as not to
3506 terminate the header.
3508 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3509 unnecessarily, used atoi() instead of strtol(), and didn't check the
3510 termination when getting file sizes from file names by regex.
3512 19. Completely re-implemented the support for maildirsize files, in the light
3513 of a number of problems with the previous contributed implementation
3514 (4.30/29). In particular:
3516 . If the quota is zero, the maildirsize file is maintained, but no quota is
3519 . If the maildir directory does not exist, it is created before any attempt
3520 to write a maildirsize file.
3522 . The quota value in the file is just a cache; if the quota is changed in
3523 the transport, the new value overrides.
3525 . A regular expression is available for excluding directories from the
3528 20. The autoreply transport checks the characters in options that define the
3529 message's headers; it allows continued headers, but it was checking with
3530 isspace() after an embedded newline instead of explicitly looking for a
3533 21. If all the "regular" hosts to which an address was routed had passed their
3534 expiry times, and had not reached their retry times, the address was
3535 bounced, even if fallback hosts were defined. Now Exim should go on to try
3538 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3539 equivalent code in the SMTP transport. Some hosts send humungous responses
3540 to HELO/EHLO, more than 1024 it seems.
3542 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3543 gives compiler warnings in some environments. I've now done it "properly",
3546 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3547 (because of problems with the built-in one) was declared to return uschar *
3548 instead of char *, causing compiler failure.
3550 25. Fixed a file descriptor leak when processing alias/forward files.
3552 26. Fixed a minor format string issue in dbfn.c.
3554 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3556 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3557 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3560 29. When a sender address is verified, it is cached, to save repeating the test
3561 when there is more than one recipient in a message. However, when the
3562 verification involves a callout, it is possible for different callout
3563 options to be set for different recipients. It is too complicated to keep
3564 track of this in the cache, so now Exim always runs a verification when a
3565 callout is required, relying on the callout cache for the optimization.
3566 The overhead is duplication of the address routing, but this should not be
3569 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3570 to be verified with callout=postmaster, and the main callout worked but the
3571 postmaster check failed, the verification correctly failed. However, if a
3572 subsequent RCPT command asked for sender verification *without* the
3573 postmaster check, incorrect caching caused this verification also to fail,
3576 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3577 it was not caching the DNS options (qualify_single, search_parents) that
3578 were used when the lookup failed. A subsequent lookup with different
3579 options therefore always gave the same answer, though there were cases
3580 where it should not have. (Example: a "domains = !$mx_any" option on a
3581 dnslookup router: the "domains" option is always processed without any
3582 widening, but the router might have qualify_single set.) Now Exim uses the
3583 cached value only when the same options are set.
3585 32. Added John Jetmore's "exipick" utility to the distribution.
3587 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3588 than a timeout (e.g. a certificate is required, and is not provided), an
3589 Exim server now closes the connection immediately. Previously it waited for
3590 the client to close - but if the client is SSL, it seems that they each
3591 wait for each other, leading to a delay before one of them times out.
3593 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3594 maintained 0.8.x compatibility because I don't think many are using it, and
3595 it is clearly obsolete.
3597 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3600 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3601 client certificate was expired. A simple patch fixes this, though I don't
3602 understand the full logic of why the verify callback is called multiple
3605 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3606 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3607 which causes problems with some clients (such as the Certicom SSL Plus
3608 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3609 disables the coutermeasure allowing Eudora to connect."
3611 38. Exim was not checking that a write() to a log file succeeded. This could
3612 lead to Bad Things if a log got too big, in particular if it hit a file
3613 size limit. Exim now panics and dies if it cannot write to a log file, just
3614 as it does if it cannot open a log file.
3616 39. Modified OS/Makefile-Linux so that it now contains
3618 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3620 The two -D definitions ensure that Exim is compiled with large file
3621 support, which makes it possible to handle log files that are bigger than
3624 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3625 instance) a domain was checked against a named list that involved a lookup,
3626 causing $domain_data to be set, then another domain was checked against the
3627 same list, then the first domain was re-checked, the value of $domain_data
3628 after the final check could be wrong. In particular, if the second check
3629 failed, it could be set empty. This bug probably also applied to
3632 41. The strip_trailing_dot option was not being applied to the address given
3633 with the -f command-line option.
3635 42. The code for reading a message's header from the spool was incrementing
3636 $received_count, but never initializing it. This meant that the value was
3637 incorrect (doubled) while delivering a message in the same process in which
3638 it was received. In the most common configuration of Exim, this never
3639 happens - a fresh exec is done - but it can happen when
3640 deliver_drop_privilege is set.
3642 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3643 it now includes up to 150 characters of the unexpected data in the log
3646 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3647 and building data strings. The size of both of these buffers was 10 000
3648 bytes - far larger than anybody would *ever* want, thought I. Needless to
3649 say, somebody hit the limit. I have increased the maximum line length to
3650 20 000 and the maximum data length of concatenated lines to 100 000. I have
3651 also fixed two bugs, because there was no checking on these buffers. Tsk,
3652 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3653 buffer is too small.
3655 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3656 lsearch lookups. Now it does.
3658 46. When parsing a route_list item in a manualroute router, a fixed-length
3659 buffer was used for the list of hosts. I made this 1024 bytes long,
3660 thinking that nobody would ever have a list of hosts that long. Wrong.
3661 Somebody had a whole pile of complicated expansion conditions, and the
3662 string was silently truncated, leading to an expansion error. It turns out
3663 that it is easier to change to an unlimited length (owing to other changes
3664 that have happened since this code was originally written) than to build
3665 structure for giving a limitation error. The length of the item that
3666 expands into the list of hosts is now unlimited.
3668 47. The lsearch lookup could not handle data where the length of text line was
3669 more than 4095 characters. Such lines were truncated, leading to shortened
3670 data being returned. It should now handle lines of any length.
3672 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3673 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3676 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3677 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3678 no longer happens in comment lines. A list of replacements is now placed
3679 at the head of all of the source files, except those whose only change is
3680 to replace PERL_COMMAND in the very first #! line.
3682 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3683 messages on the queue, with a bottom-up merge sort, using code contributed
3684 by Michael Haardt. This should make operations like -bp somewhat faster on
3685 large queues. It won't affect queue runners, except when queue_run_in_order
3688 51. Installed eximstats 1.31 in the distribution.
3690 52. Added support for SRV lookups to the dnslookup router.
3692 53. If an ACL referred to $message_body or $message_body_end, the value was not
3693 reset for any messages that followed in the same SMTP session.
3695 54. The store-handling optimization for building very long strings was not
3696 differentiating between the different store pools. I don't think this
3697 actually made any difference in practice, but I've tidied it.
3699 55. While running the routers to verify a sender address, $sender_address
3700 was still set to the sender address. This is wrong, because when routing to
3701 send a bounce to the sender, it would be empty. Therefore, I have changed
3702 it so that, while verifying a sender address, $sender_address is set to <>.
3703 (There is no change to what happens when verifying a recipient address.)
3705 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3706 target A or AAAA records (if not already returned) without resetting the
3707 qualify_single or search_parents options of the DNS resolver. These are
3708 inappropriate in this case because the targets of MX and SRV records must
3709 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3710 target that, when qualified, matched something in the local domain. These
3711 two options are now turned off when doing these lookups.
3713 57. It seems that at least some releases of Reiserfs (which does not have the
3714 concept of a fixed number of inodes) returns zero and not -1 for the
3715 number of available inodes. This interacted badly with check_spool_inodes,
3716 which assumed that -1 was the "no such thing" setting. What I have done is
3717 to check that the total number of inodes is greater than zero before doing
3718 the test of how many are available.
3720 58. When a "warn" ACL statement has a log_message modifier, the message is
3721 remembered, and not repeated. This is to avoid a lot of repetition when a
3722 message has many recipients that cause the same warning to be written.
3723 Howewer, Exim was preserving the list of already written lines for an
3724 entire SMTP session, which doesn't seem right. The memory is now reset if a
3725 new message is started.
3727 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3728 debugging output unless log_rewrite was also set.
3730 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3731 of (int)(handle) when we know that handle contains (void *)(-1).
3733 61. The Exim daemon panic-logs an error return when it closes the incoming
3734 connection. However "connection reset by peer" seems to be common, and
3735 isn't really an error worthy of noting specially, so that particular error
3738 62. When Exim is trying to find all the local interfaces, it used to panic and
3739 die if the ioctl to get the interface flags failed. However, it seems that
3740 on at least one OS (Solaris 9) it is possible to have an interface that is
3741 included in the list of interfaces, but for which you get a failure error
3742 for this call. This happens when the interface is not "plumbed" into a
3743 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3744 failure of the "get flags" call assumes that the interface is down.
3746 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3747 makes life easier for people who are doing arithmetic on fields extracted
3748 from dates, where you often get leading zeros that should not be
3749 interpreted as octal.
3751 64. Added qualify_domain to the redirect router, to override the global
3754 65. If a pathologically long header line contained very many addresses (the
3755 report of this problem mentioned 10 000) and each of them was rewritten,
3756 Exim could use up a very large amount of memory. (It kept on making new
3757 copies of the header line as it rewrote, and never released the old ones.)
3758 At the expense of a bit more processing, the header rewriting function has
3759 been changed so that it no longer eats memory in this way.
3761 66. The generation of the Received: header has been moved from the time that a
3762 message starts to be received, to the time that it finishes. The timestamp
3763 in the Received: header should now be very close to that of the <= log
3764 line. There are two side-effects of this change:
3766 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3767 logged header lines no longer include the local Received: line, because
3768 it has not yet been created. The same applies to a copy of the message
3769 that is returned to a non-SMTP sender when a message is rejected.
3771 (b) When a filter file is tested using -bf, no additional Received: header
3772 is added to the test message. After some thought, I decided that this
3775 This change does not affect the value of $received_for. It is still set
3776 after address rewriting, but before local_scan() is called.
3778 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3780 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3781 gave an unhelpful panic error message, and a defer error. I have managed to
3782 change this behaviour so that it now rejects any supplied certificate,
3783 which seems right, as the list of acceptable certificates is empty.
3785 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3786 gave an unhelpful defer error. I have not managed to make this reject any
3787 supplied certificates, but the error message it gives is "no certificate
3788 supplied", which is not helpful.
3790 70. exigrep's output now also includes lines that are not associated with any
3791 message, but which match the given pattern. Implemented by a patch from
3792 Martin Sluka, which also tidied up the Perl a bit.
3794 71. Recipient callout verification, like sender verification, was using <> in
3795 the MAIL FROM command. This isn't really the right thing, since the actual
3796 sender may affect whether the remote host accepts the recipient or not. I
3797 have changed it to use the actual sender in the callout; this means that
3798 the cache record is now keyed on a recipient/sender pair, not just the
3799 recipient address. There doesn't seem to be a real danger of callout loops,
3800 since a callout by the remote host to check the sender would use <>.
3801 [SEE ABOVE: changed after hitting problems.]
3803 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3804 temporary errors. However, in the case of such a code being given after
3805 the end of a data transmission (i.e. after ".") Exim was failing to write
3806 a retry record for the message. (Yes, there was some broken host that was
3807 actually sending 8xx at this point.)
3809 73. An unknown lookup type in a host list could cause Exim to panic-die when
3810 the list was checked. (An example that provoked this was putting <; in the
3811 middle of a list instead of at the start.) If this happened during a DATA
3812 ACL check, a -D file could be left lying around. This kind of configuration
3813 error no longer causes Exim to die; instead it causes a defer errror. The
3814 incident is still logged to the main and panic logs.
3816 74. Buglet left over from Exim 3 conversion. The message "too many messages
3817 in one connection" was written to the rejectlog but not the mainlog, except
3818 when address rewriting (yes!) was being logged.
3820 75. Added write_rejectlog option.
3822 76. When a system filter was run not as root (that is, when system_filter_user
3823 was set), the values of the $n variables were not being returned to the
3824 main process; thus, they were not subsequently available in the $sn
3827 77. Added +return_path_on_delivery log selector.
3829 78. A connection timeout was being treated differently from recipients deferred
3830 when testing hosts_max_try with a message that was older than the host's
3831 retry timeout. (The host should not be counted, thus allowing all hosts to
3832 be tried at least once before bouncing.) This may have been the cause of an
3833 occasionally reported bug whereby a message would remain on the queue
3834 longer than the retry timeout, but would be bounced if a delivery was
3835 forced. I say "may" because I never totally pinned down the problem;
3836 setting up timeout/retry tests is difficult. See also the next item.
3838 79. The ultimate address timeout was not being applied to errors that involved
3839 a combination of host plus message (for example, a timeout on a MAIL
3840 command). When an address resolved to a number of possible hosts, and they
3841 were not all tried for each delivery (e.g. because of hosts_max_try), a
3842 message could remain on the queue longer than the retry timeout.
3844 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3847 81. Fixed an obscure SMTP outgoing bug which required at least the following
3848 conditions: (a) there was another message waiting for the same server;
3849 (b) the server returned 5xx to all RCPT commands in the first message so
3850 that the message was not completed; (c) the server dropped the connection
3851 or gave a negative response to the RSET that Exim sends to abort the
3852 transaction. The observed case was a dropped connection after DATA that had
3853 been sent in pipelining mode. That is, the server had advertised PIPELINING
3854 but was not implementing it correctly. The effect of the bug was incorrect
3855 behaviour, such as trying another host, and this could lead to a crash.
3861 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3862 and daemon.c were passed as pointers to ints; they should have been
3863 pointers to socklen_t variables (which are typically unsigned ints).
3865 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3868 3. Fixed a really odd bug that affected only the testing scheme; patching a
3869 certain fixed string in the binary changed the value of another string that
3870 happened to be identical to the end of the original first string.
3872 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3873 name", it returns that address as the IP address. On some operating
3874 systems (e.g. Solaris), it also passes back the IP address string as the
3875 "host name". However, on others (e.g. Linux), it passes back an empty
3876 string. Exim wasn't checking for this, and was changing the host name to an
3877 empty string, assuming it had been canonicized.
3879 5. Although rare, it is permitted to have more than one PTR record for a given
3880 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3881 all the names associated with an address, because they do in Solaris.
3882 However, it seems that they do not in Linux for data that comes from the
3883 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3884 I found this out when I moved to a new Linux workstation and tried to run
3885 the Exim test suite.
3887 To get round this problem I have changed the code so that it now does its
3888 own call to the DNS to look up PTR records when searching for a host name.
3889 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3890 addresses that are only in /etc/hosts are still found.
3892 This behaviour is, however, controlled by an option called host_lookup_
3893 order, which defaults to "bydns:byaddr". If people want to use the other
3894 order, or indeed, just use one or the other means of lookup, they can
3895 specify it in this variable.
3897 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3898 some operating systems, this comes back from gethostbyaddr() as an empty
3899 string, and this is what Exim used to test for. However, it seems that in
3900 other systems, "." is yielded. Exim now tests for this case too.
3902 7. The values of check_spool_space and check_log_space are now held internally
3903 as a number of kilobytes instead of an absolute number of bytes. If a
3904 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3905 kilobyte. This means that much larger values can be stored.
3907 8. Exim monitor: an attempt to get the action menu when not actually pointing
3908 at a message produces an empty menu entitled "No message selected". This
3909 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3910 no entries in it ("Shell widget menu has zero width and/or height"). So I
3911 have added a single, blank menu entry in this case.
3913 9. Added ${quote_local_part.
3915 10. MIME decoding is now applied to the contents of Subject: header lines when
3918 11. Now that a reference to $sender_host_address automatically causes a reverse
3919 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3920 host lookup before query-style lookups in lists that might use this
3921 variable. This has therefore been abolished, and the "net-" prefix is no
3922 longer necessary for query-style lookups.
3924 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3925 have been a typo for LFLAGS, so it has been changed.
3927 13. The install script calls Exim with "-C /dev/null" in order to find the
3928 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3929 to be output. Howeve, since Exim outputs its version number before the
3930 error, it didn't break the script. It just looked ugly. I fixed this by
3931 always allowing "-C /dev/null" if the caller is root.
3933 14. Ignore overlarge ACL variable number when reading spool file - insurance
3934 against a later release with more variables having written the file.
3936 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3937 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3938 this, as well as the form without the "IPv6" on the front.
3940 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3941 OS/Makefile-Darwin file.
3943 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3944 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3946 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3947 variable. However, some OS use ints for inodes. I've added cast to long int
3948 to get rid of the compiler warning.
3950 19. I had forgotten to lock out "/../" in configuration file names when
3951 ALT_CONFIG_PREFIX was set.
3953 20. Routers used for verification do not need to specify transports. However,
3954 if such a router generated a host list, and callout was configured, Exim
3955 crashed, because it could not find a port number from the (non-existent)
3956 transport. It now assumes port 25 in this circumstance.
3958 21. Added the -t option to exigrep.
3960 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3961 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3962 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3963 them is set, LOOKUP_LSEARCH is forced.
3965 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3967 24. Added sender and host information to the "rejected by local_scan()" log
3968 line; previously there was no indication of these.
3970 25. Added .include_if_exists.
3972 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3973 Linux. It turns out that not all file systems support this. Apparently some
3974 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3975 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3976 not supported on the file descriptor, is now ignored when Exim is trying to
3977 sync a directory. This applies only to Linux.
3979 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3981 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3982 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3985 29. Added support for maildirsize files from supplied patch (modified a bit).
3987 30. The use of :fail: followed by an empty string could lead Exim to respond to
3988 sender verification failures with (e.g.):
3990 550 Verification failed for <xxx>
3991 550 Sender verify failed
3993 where the first response line was missing the '-' that indicates it is not
3994 the final line of the response.
3996 31. The loop for finding the name of the user that called Exim had a hardwired
3997 limit of 10; it now uses the value of finduser_retries, which is used for
3998 all other user lookups.
4000 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4003 33. Exim was neglecting to zero errno before one call of strtol() when
4004 expanding a string and expecting an integer value. On some systems this
4005 resulted in spurious "integer overflow" errors. Also, it was casting the
4006 result into an int without checking.
4008 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4009 did not work. The code looks as if it has *never* worked, though it appears
4010 to have been documented since at least releast 1.62. I have made it work.
4012 35. The "timeout_DNS" error in retry rules, also documented since at least
4013 1.62, also never worked. As it isn't clear exactly what this means, and
4014 clearly it isn't a major issue, I have abolished the feature by treating it
4015 as "timeout", and writing a warning to the main and panic logs.
4017 36. The display of retry rules for -brt wasn't always showing the error code
4020 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4021 timeout_connect_A, timeout_connect_MX.
4023 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4024 to the empty sender.
4026 39. The daemon was not analysing the content of -oX till after it had closed
4027 stderr and disconnected from the controlling terminal. This meant that any
4028 syntax errors were only noted on the panic log, and the return code from
4029 the command was 0. By re-arranging the code a little, I've made the
4030 decoding happen first, so such errors now appear on stderr, and the return
4031 code is 1. However, the actual setting up of the sockets still happens in
4032 the disconnected process, so errors there are still only recorded on the
4035 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4036 connections (as happens on some IP stacks) was logged at start up time as
4037 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4038 it from "IPv6 and IPv4", which means that two separate sockets are being
4041 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4042 says whether AF_INET or AF_INET6 was passed as an argument.
4044 42. Exiwhat output was messed up when time zones were included in log
4047 43. Exiwhat now gives more information about the daemon's listening ports,
4048 and whether -tls-on-connect was used.
4050 44. The "port" option of the smtp transport is now expanded.
4052 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4053 silently ignored. Now an error message is written to the main and panic
4056 46. There's a new ACL modifier called "logwrite" which writes to a log file
4057 as soon as it is encountered.
4059 47. Added $local_user_uid and $local_user_gid at routing time.
4061 48. Exim crashed when trying to verify a sender address that was being
4064 49. Exim was recognizing only a space character after ".include". It now also
4065 recognizes a tab character.
4067 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4068 extracting the relevant information from the specification. The man page no
4069 longer contains scrambled data for the -d option, and I've added a section
4070 at the front about calling Exim under different names.
4072 51. Added "extra_headers" argument to the "mail" command in filter files.
4074 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4077 53. Installed eximstats 1.29.
4079 54. Added transport_filter_timeout as a generic transport option.
4081 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4082 Cc: header lines. This was required by RFC 822, but it not required by RFC
4085 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4086 incoming messages that did not have them. Now it does so only if the
4087 message originates locally, that is, if there is no associated remote host
4088 address. When Resent- header lines are present, this applies to the Resent-
4089 lines rather than the non-Resent- lines.
4091 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4092 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4094 58. Messages for configuration errors now include the name of the main
4095 configuration file - useful now that there may be more than one file in a
4096 list (.included file names were always shown).
4098 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4099 for those rare installations that do not start the daemon as root or run it
4100 setuid root. I've cut out the call to initgroups() if the daemon is not
4103 60. The Exim user and group can now be bound into the binary as text strings
4104 that are looked up at the start of Exim's processing.
4106 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4108 62. Added $mailstore_basename variable.
4110 63. Installed patch to sieve.c from Michael Haardt.
4112 64. When Exim failed to open the panic log after failing to open the main log,
4113 the original message it was trying to log was written to stderr and debug
4114 output, but if they were not available (the usual case in production), it
4115 was lost. Now it is written to syslog before the two lines that record the
4116 failures to open the logs.
4118 65. Users' Exim filters run in subprocesses under the user's uid. It is
4119 possible for a "deliver" command or an alias in a "personal" command to
4120 provoke an address rewrite. If logging of address rewriting is configured,
4121 this fails because the process is not running as root or exim. There may be
4122 a better way of dealing with this, but for the moment (because 4.30 needs
4123 to be released), I have disabled address rewrite logging when running a
4124 filter in a non-root, non-exim process.
4130 1. The buildconfig auxiliary program wasn't quoting the value set for
4131 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4132 not defined. This bug was masked in 4.22 by the effect that was fixed in
4135 2. Some messages that were rejected after a message id was allocated were
4136 shown as "incomplete" by exigrep. It no longer does this for messages that
4137 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4139 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4140 have allow_domain_literals set, the ID did not get logged in the <= line.
4141 Domain literals are now always recognized in Message-ID: header lines.
4143 4. The first argument for a ${extract expansion item is the key name or field
4144 number. Leading and trailing spaces in this item were not being ignored,
4145 causing some misleading effects.
4147 5. When deliver_drop_privilege was set, single queue runner processes started
4148 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4149 same command in the process it spins off) were not dropping privilege.
4151 6. When the daemon running as "exim" started a queue runner, it always
4152 re-executed Exim in the spun-off process. This is a waste of effort when
4153 deliver_drop_privilege is set. The new process now just calls the
4154 queue-runner function directly.
4160 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4163 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4164 ignored. Though the use of -oP was forcing the writing of a pid file, it
4165 was always written to the default place.
4167 3. If the message "no IP address found for host xxxx" is generated during
4168 incoming verification, it is now followed by identification of the incoming
4169 connection (so you can more easily find what provoked it).
4171 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4173 5. Added some features to "harden" Exim a bit more against certain attacks:
4175 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4176 be put in Local/Makefile. This is like the never_users runtime option,
4177 but it cannot be overridden. The default setting is "root".
4179 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4180 prefix string with which any file named in a -C command line option
4183 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4184 is retained for -C and -D only if the caller of Exim is root. Without
4185 it, the exim user may also use -C and -D and retain privilege.
4187 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4188 command line option is disabled.
4190 6. Macro names set by the -D option must start with an upper case letter, just
4191 like macro names defined in the configuration file.
4193 7. Added "dereference=" facility to LDAP.
4195 8. Two instances of the typo "uknown" in the source files are fixed.
4197 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4198 the Configure-Makefile script screwed up while processing it.
4200 10. Incorporated PCRE 4.4.
4202 11. The SMTP synchronization check was not operating right at the start of an
4203 SMTP session. For example, it could not catch a HELO sent before the client
4204 waited for the greeting. There is now a check for outstanding input at the
4205 point when the greeting is written. Because of the duplex, asynchronous
4206 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4207 way, but not yet received, when the check is performed.
4209 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4210 on TCP/IP sockets, because this apparently causes some broken clients to
4213 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4214 unchanged) from the Cygwin maintainer.
4216 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4217 was supported instead of testing for mbx. Effectively a typo.
4219 15. The spa authenticator server code was not checking that the input it
4220 received was valid base64.
4222 16. The debug output line for the "set" modifier in ACLs was not showing the
4223 name of the variable that was being set.
4225 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4227 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4228 lookup on its own. Something else was needed to trigger the lookup. For
4229 example, a match in host_lookup or the need for a host name in a host list.
4230 Now, if $sender_host_name is referenced and the host name has not yet been
4231 looked up, a lookup is performed. If the lookup fails, the variable remains
4232 empty, and $host_lookup_failed is set to "1".
4234 19. Added "eqi" as a case-independent comparison operator.
4236 20. The saslauthd authentication condition could segfault if neither service
4237 nor realm was specified.
4239 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4240 error message that was logged was misleading, and incoming SMTP
4241 connections were dropped. The message is now more accurate, and temporary
4242 errors are given to SMTP connections.
4244 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4245 (except RSET) with a 421 error, until QUIT is received. However, it was
4246 failing to send a response to QUIT.
4248 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4249 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4250 ACL is now run after the helo_try_verify_hosts code.
4252 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4253 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4254 case-independent, so other case variants are also recognized). Apparently
4255 some systems use these upper case variants.
4257 25. If more than two messages were waiting for the same host, and a transport
4258 filter was specified for the transport, Exim sent two messages over the
4259 same TCP/IP connection, and then failed with "socket operation on non-
4260 socket" when it tried to send the third.
4262 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4264 27. The extern definition of crypt16() in expand.c was not being excluded when
4265 the OS had its own crypt16() function.
4267 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4268 as a preferred synonym for return_size_limit, both as an option and as an
4271 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4273 30. Changed the default configuration ACL to relax the local part checking rule
4274 for addresses that are not in any local domains. For these addresses,
4275 slashes and pipe symbols are allowed within local parts, but the sequence
4276 /../ is explicitly forbidden.
4278 31. SPA server authentication was not clearing the challenge buffer before
4281 32. log_message in a "warn" ACL statement was writing to the reject log as
4282 well as to the main log, which contradicts the documentation and doesn't
4283 seem right (because no rejection is happening). So I have stopped it.
4285 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4286 However, I am unable to do any testing of this.
4288 34. Fixed an infelicity in the appendfile transport. When checking directories
4289 for a mailbox, to see if any needed to be created, it was accidentally
4290 using path names with one or more superfluous leading slashes; tracing
4291 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4293 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4294 contents are added to the log line that is written for every discarded
4295 recipient. (Previously a log_message setting was ignored.)
4297 36. The ${quote: operator now quotes the string if it is empty.
4299 37. The install script runs exim in order to find its version number. If for
4300 some reason other than non-existence or emptiness, which it checks, it
4301 could not run './exim', it was installing it with an empty version number,
4302 i.e. as "exim-". This error state is now caught, and the installation is
4305 38. An argument was missing from the function that creates an error message
4306 when Exim fails to connect to the socket for saslauthd authentication.
4307 This could cause Exim to crash, or give a corrupted message.
4309 39. Added isip, isip4, and isip6 to ${if conditions.
4311 40. The ACL variables $acl_xx are now saved with the message, and can be
4312 accessed later in routers, transports, and filters.
4314 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4315 strings in the file are not string-expanded.
4317 42. If a MAIL command specified a SIZE value that was too large to fit into an
4318 int variable, the check against message_size_limit failed. Such values are
4319 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4320 day this will have to be increased, but I don't think I want to be around
4321 when emails are that large.
4328 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4329 iconv() is not standard in FreeBSD.
4331 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4332 IPv6 enabled. The observed symptom was a segmentation fault on return from
4333 the function os_common_find_running_interfaces() in src/os.c.
4335 3. In the check_special_case() function in daemon.c I had used "errno" as an
4336 argument name, which causes warnings on some systems. This was basically a
4337 typo, since it was named "eno" in the comments!
4339 4. The code that waits for the clock to tick (at a resolution of some fraction
4340 of a second) so as to ensure message-id uniqueness was always waiting for
4341 at least one whole tick, when it could have waited for less. [This is
4342 almost certainly not relevant at current processor speeds, where it is
4343 unlikely to ever wait at all. But we try to future-proof.]
4345 5. The function that sleeps for a time interval that includes fractions of a
4346 second contained a race. It did not block SIGALRM between setting the
4347 timer, and suspending (a couple of lines later). If the interval was short
4348 and the sigsuspend() was delayed until after it had expired, the suspension
4349 never ended. On busy systems this could lead to processes getting stuck for
4352 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4353 process, before it forks any delivery processes. The open lookup caching
4354 mechanism meant that the open file or database connection was passed into
4355 the delivery process. The problem was that delivery processes always tidy
4356 up cached lookup data. This could cause a problem for the next delivery
4357 process started by the queue runner, because the external queue runner
4358 process does not know about the closure. So the next delivery process
4359 still has data in the lookup cache. In the case of a file lookup, there was
4360 no problem because closing a file descriptor in a subprocess doesn't affect
4361 the parent. However, if the lookup was caching a connection to a database,
4362 the connection was closed, and the second delivery process was likely to
4363 see errors such as "PGSQL: query failed: server closed the connection
4364 unexpectedly". The problem has been fixed by closing all cached lookups
4365 in a queue runner before running a delivery process.
4367 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4368 seem to have the "const" qualifier which it has on other OS. I've
4371 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4372 *of the same type* (client or server) with the same public name that an
4373 error should be diagnosed.
4375 9. When Exim looked up a host name for an IP address, but failed to find the
4376 original IP address when looking up the host name (a safety check), it
4377 output the message "<ip address> does not match any IP for NULL", which was
4378 confusing, to say the least. The bug was that the host name should have
4379 appeared instead of "NULL".
4381 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4382 user that is built into the binary, and the -C or -D options is used, root
4383 privilege is dropped before the configuration file is read. In addition,
4384 logging is switched to stderr instead of the normal log files. If the
4385 configuration then re-defines the Exim user, the unprivileged environment
4386 is probably not what is expected, so Exim logs a panic warning message (but
4389 However, if deliver_drop_privilege is set, the unprivileged state may well
4390 be exactly what is intended, so the warning has been cut out in that case,
4391 and Exim is allowed to try to write to its normal log files.
4397 1. smtp_return_error_details was not giving details for temporary sender
4398 or receiver verification errors.
4400 2. Diagnose a configuration error if two authenticators have the same public
4403 3. Exim used not to create the message log file for a message until the first
4404 delivery attempt. This could be confusing when incoming messages were held
4405 for policy or load reasons. The message log file is now created at the time
4406 the message is received, and an initial "Received" line is written to it.
4408 4. The automatically generated man page for command line options had a minor
4409 bug that caused no ill effects; however, a more serious problem was that
4410 the procedure for building the man page automatically didn't always
4411 operate. Consequently, release 4.20 contains an out-of-date version. This
4412 shouldn't happen again.
4414 5. When building Exim with embedded Perl support, the script that builds the
4415 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4416 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4418 6. The freeze_tell option was not being used for messages that were frozen on
4419 arrival, either by an ACL or by local_scan().
4421 7. Added the smtp_incomplete_transaction log selector.
4423 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4424 was accepting AUTH without a new EHLO.
4426 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4427 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4430 10. Logging of TCP/IP connections (when configured) now happens in the main
4431 daemon process instead of the child process, so that the TCP/IP connection
4432 count is more accurate (but it can never be perfect).
4434 11. The use of "drop" in a nested ACL was not being handled correctly in the
4435 outer ACL. Now, if condition failure induced by the nested "drop" causes
4436 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4437 or "require"), the connection is dropped.
4439 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4440 that yield "discard" can now be used with an "accept" or a "discard" verb,
4441 but an error is generated for any others (because I can't see a useful way
4442 to define what should happen).
4444 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4445 are now processed in the same way as lines in the Exim configuration file.
4446 In particular, continuation lines are supported.
4448 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4450 15. Added -ti meaning -t -i.
4452 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4453 domains, and warn by logging if others are found.
4455 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4456 returned from GIFCONF when Exim is trying to find the list of interfaces on
4457 a host. The code in os.c has been modified to copy each ifreq to an aligned
4458 structure in all cases.
4460 Also, in some cases, the returned ifreq's were being copied to a 'struct
4461 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4462 means the last couple of bytes of an IPv6 address could be chopped if the
4463 ifreq contained only a normal sockaddr (14 bytes storage).
4465 18. Named domain lists were not supported in the hosts_treat_as_local option.
4466 An entry such as +xxxx was not recognized, and was treated as a literal
4469 19. Ensure that header lines added by a DATA ACL are included in the reject log
4470 if the ACL subsequently rejects the message.
4472 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4473 MD5 (which is deprecated).
4475 21. When testing a filter file using -bf, Exim was writing a message when it
4476 took the sender from a "From " line in the message, but it was not doing so
4477 when it took $return_path from a Return-Path: header line. It now does.
4479 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4480 with a valid header line field name (a series of printing characters
4481 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4483 23. Changed "disc" in the source to "disk" to conform to the documentation and
4484 the book and for uniformity.
4486 24. Ignore Sendmail's -Ooption=value command line item.
4488 25. When execve() failed while trying to run a command in a pipe transport,
4489 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4490 could be confused with a return value of 69 from the command itself. This
4491 has been changed to 127, the value the shell returns if it is asked to run
4492 a non-existent command. The wording for the related log line suggests a
4493 non-existent command as the problem.
4495 26. If received_header_text expands to an empty string, do not add a Received:
4496 header line to the message. (Well, it adds a token one on the spool, but
4497 marks it "old" so that it doesn't get used or transmitted.)
4499 27. Installed eximstats 1.28 (addition of -nt option).
4501 28. There was no check for failure on the call to getsockname() in the daemon
4502 code. This can fail if there is a shortage of resources on the system, with
4503 ENOMEM, for example. A temporary error is now given on failure.
4505 29. Contrary to the C standard, it seems that in some environments, the
4506 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4507 program. Exim now does this explicitly; it affects the formatting of
4508 timestamps using strftime().
4510 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4511 complaints. I've now initialized all the variables, to avoid this.
4513 32. Header lines added by a system filter were not being "seen" during
4514 transport-time rewrites.
4516 33. The info_callback() function passed to OpenSSL is set up with type void
4517 (*)(SSL *, int, int), as described somewhere. However, when calling the
4518 function (actually a macro) that sets it up, the type void(*)() is
4519 expected. I've put in a cast to prevent warnings from picky compilers.
4521 34. If a DNS black list lookup found a CNAME record, but there were no A
4522 records associated with the domain it pointed at, Exim crashed.
4524 35. If a DNS black list lookup returned more than one A record, Exim ignored
4525 all but the first. It now scans all returned addresses if a particular IP
4526 value is being sought. In this situation, the contents of the
4527 $dnslist_value variable are a list of all the addresses, separated by a
4530 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4531 to accept a host name and all its aliases if the forward lookup for any of
4532 them yielded the IP address of the incoming connection. Now it accepts only
4533 those names whose forward lookup yields the correct IP address. Any other
4534 names are discarded. This closes a loophole whereby a rogue DNS
4535 administrator could create reverse DNS records to break through a
4536 wildcarded host restriction in an ACL.
4538 37. If a user filter or a system filter that ran in a subprocess used any of
4539 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4540 the wrong values were passed to the pipe command ($thisaddress had the
4541 value of $0, $0 had the value of $1, etc). This bug was introduced by
4542 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4544 38. Improved the line breaking for long SMTP error messages from ACLs.
4545 Previously, if there was no break point between 40 and 75 characters, Exim
4546 left the rest of the message alone. Two changes have been made: (a) I've
4547 reduced the minimum length to 35 characters; (b) if it can't find a break
4548 point between 35 and 75 characters, it looks ahead and uses the first one
4549 that it finds. This may give the occasional overlong line, but at least the
4550 remaining text gets split now.
4552 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4553 file descriptors might be low, and that setting 1000 would always raise it.
4554 It turns out that in some environments, the limit is already over 1000 and
4555 that lowering it causes trouble. So now Exim takes care not to decrease it.
4557 40. When delivering a message, the value of $return_path is set to $sender_
4558 address at the start of routing (routers may change the value). By an
4559 oversight, this default was not being set up when an address was tested by
4560 -bt or -bv, which affected the outcome if any router or filter referred to
4563 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4564 log only when "message" or "log_message" are set. However, if one of the
4565 conditions was an address verification, or a call to a nested ACL, the
4566 messages generated by the underlying test were being passed through. This
4567 no longer happens. The underlying message is available in $acl_verify_
4568 message for both "message" and "log_message" expansions, so it can be
4569 passed through if needed.
4571 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4572 new expansion $bh_ to give the encoded byte string without charset
4573 translation. Translation happens only if iconv() is available; HAVE_ICONV
4574 indicates this at build time. HEADERS_CHARSET gives the charset to
4575 translate to; headers_charset can change it in the configuration, and
4576 "headers charset" can change it in an individual filter file.
4578 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4579 that creates RFC 2047 encoded "words" labels them as that charset instead
4580 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4581 expansion operator; (ii) when Exim creates a From: line for a local
4582 message; (iii) when a header line is rewritten to include a "phrase" part.
4584 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4585 buggy, causing it to skip the first lines of messages whose message ID
4586 ended in 'D'. This would not have bitten before Exim release 4.14, because
4587 message IDs were unlikely to end in 'D' before then. The effect was to have
4588 incorrect size information for certain domains.
4590 45. #include "config.h" was missing at the start of the crypt16.c module. This
4591 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4594 46. If there was a timeout during a "random" callout check, Exim treated it as
4595 a failure of the random address, and carried on sending RSET and the real
4596 address. If the delay was just some slowness somewhere, the response to the
4597 original RCPT would be taken as a response to RSET and so on, causing
4598 mayhem of various kinds.
4600 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4601 when I implemented it. It didn't allow for the fact that some option values
4602 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4603 the right test for positive values.
4605 48. Domain names in DNS records are case-independent. Exim always looks them up
4606 in lower case. Some resolvers return domain names in exactly the case they
4607 appear in the zone file, that is, they may contain uppercase letters. Not
4608 all resolvers do this - some return always lower case. Exim was treating a
4609 change of case by a resolver as a change of domain, similar to a widening
4610 of a domain abbreviation. This triggered its re-routing code and so it was
4611 trying to route what was effectively the same domain again. This normally
4612 caused routing to fail (because the router wouldn't handle the domain
4613 twice). Now Exim checks for this case specially, and just changes the
4614 casing of the domain that it ultimately uses when it transmits the message
4617 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4620 50. If a filter generated a file delivery with a non-absolute name (possible if
4621 no home directory exists for the router), the forbid_file option was not
4624 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4625 the existing equality matching.
4627 52. Exim was using ints instead of ino_t variables in some places where it was
4628 dealing with inode numbers.
4630 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4631 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4632 called TMPDIR, and if it finds it is different, it changes its value.
4634 54. The smtp_printf() function is now made available to local_scan() so
4635 additional output lines can be written before returning. There is also an
4636 smtp_fflush() function to enable the detection of a dropped connection.
4637 The variables smtp_input and smtp_batched_input are exported to
4640 55. Changed the default runtime configuration: the message "Unknown user"
4641 has been removed from the ACL, and instead placed on the localuser router,
4642 using the cannot_route_message feature. This means that any verification
4643 failures that generate their own messages won't get overridden. Similarly,
4644 the "Unrouteable address" message that was in the ACL for unverifiable
4645 relay addresses has also been removed.
4647 56. Added hosts_avoid_esmtp to the smtp transport.
4649 57. The exicyclog script was not checking for the esoteric option
4650 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4651 will work only if exicyclog is run under the appropriate euid.
4653 58. Following a discussion on the list, the rules by which Exim recognises line
4654 endings on incoming messages have been changed. The -dropcr and drop_cr
4655 options are now no-ops, retained only for backwards compatibility. The
4656 following line terminators are recognized: LF CRLF CR. However, special
4657 processing applies to CR:
4659 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4660 nor a local message in the state where . is a terminator.
4662 (ii) If a bare CR is encountered in a header line, an extra space is added
4663 after the line terminator so as not to end the header. The reasoning
4664 behind this is that bare CRs in header lines are most likely either
4665 to be mistakes, or people trying to play silly games.
4667 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4668 was being incorrectly given as 18 bytes larger than it should have been.
4669 This is a VOB (very old bug).
4671 60. This may never have affected anything current, but just in case it has:
4672 When the local host is found other than at the start of a list of hosts,
4673 the local host, those with the same MX, and any that follow, are discarded.
4674 When the list in question was part of a longer list of hosts, the following
4675 hosts (not currently being processed) were also being discarded. This no
4676 longer happens. I'm not sure if this situation could ever has previously
4679 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4682 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4683 header. This is something that is recommended in a new Internet Draft, and
4684 is something that is documented as being done by Sendmail. There are two
4685 possible values. For messages generated by the autoreply transport, Exim
4688 Auto-Submitted: auto-replied
4690 whereas for all other generated messages (e.g. bounces) it adds
4692 Auto-Submitted: auto-generated
4694 63. The "personal" condition in filters now includes a test for the
4695 Auto-Submitted: header. If it contains the string "auto-" the message it
4696 not considered personal.
4698 64. Added rcpt_include_affixes as a generic transport option.
4700 65. Added queue_only_override (default true).
4702 66. Added the syslog_duplication option.
4704 67. If what should have been the first header line of a message consisted of
4705 a space followed by a colon, Exim was mis-interpreting it as a header line.
4706 It isn't of course - it is syntactically invalid and should therefore be
4707 treated as the start of the message body. The misbehaviour could have
4708 caused a number of strange effects, including loss of data in subsequent
4709 header lines, and spool format errors.
4711 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4712 client host had authenticated. This control can now be exercised by an ACL
4713 for more flexibility.
4715 69. By default, callouts do not happen when testing with -bh. There is now a
4716 variant, -bhc, which does actually run the callout code, including
4717 consulting and updating the callout cache.
4719 70. Added support for saslauthd authentication, courtesy of Alexander
4722 71. If statvfs() failed on the spool or log directories while checking their
4723 size for availability, Exim confusingly gave the error "space shortage".
4724 Furthermore, in debugging mode it crashed with a floating point exception.
4725 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4726 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4727 serious problem, Exim now writes to the main and panic logs when this
4728 happens, with details of the failure. It then refuses to accept the
4729 incoming message, giving the message "spool directory problem" or "log
4730 directory problem" with a 421 code for SMTP messages.
4732 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4733 0, 1, and 2 exist, because some OS complain for execs without them (see
4734 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4735 descriptors. However, the code omitted to check that the open succeeded,
4736 causing mysterious errors if for some reason the permissions on /dev/null
4737 got screwed. Now Exim writes a message to the main and panic logs, and
4738 bombs out if it can't open /dev/null.
4740 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4741 interact so that it is all more flexible. It is supposed to remain
4742 backwards compatible. Also added extra_local_interfaces.
4744 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4745 to bomb out with an assertion failure - to the client this appears as a
4746 connection drop. This problem occurs in the part of the code that was taken
4747 from the Samba project. Fortunately, the assertion is in a very simple
4748 function, so I have fixed this by reproducing the function inline in the
4749 one place where it is called, and arranging for authentication to fail
4750 instead of killing the process with assert().
4752 75. The SPA client code was not working when the server requested OEM rather
4753 than Unicode encoding.
4755 76. Added code to make require_files with a specific uid setting more usable in
4756 the case where statting the file as root fails - usually a non-root-mounted
4757 NFS file system. When this happens and the failure is EACCES, Exim now
4758 forks a subprocess and does the per-uid checking as the relevant uid.
4760 77. Added process_log_path.
4762 78. If log_file_path was not explicitly set, a setting of check_log_space or
4763 check_log_inodes was ignored.
4765 79. If a space check for the spool or log partitions fails, the incident is now
4766 logged. Of course, in the latter case the data may get lost...
4768 80. Added the %p formatting code to string_format() so that it can be used to
4769 print addresses in debug_print(). Adjusted all the address printing in the
4770 debugging in store.c to use %p rather than %d.
4772 81. There was a concern that a line of code in smtp_in.c could overflow a
4773 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4774 initially expressed, the concern was not well-founded, because trailing
4775 spaces are removed early. However, if the trailing spaces were followed by
4776 a NULL, they did not get removed, so the overflow was possible. Two fixes
4779 (a) I re-wrote the offending code in a cleaner fashion.
4780 (b) If an incoming SMTP command contains a NULL character, it is rejected
4783 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4784 runs initgroups(), so that if the Exim user is in any additional groups,
4785 they will be used during message reception.
4791 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob