1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.521 2007/07/04 10:37:03 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Another patch from the Sieve maintainer.
11 PH/02 When an IPv6 address is converted to a string for single-key lookup
12 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
13 dots are used instead of colons so that keys in lsearch files need not
14 contain colons. This was done some time before quoting was made available
15 in lsearch files. However, iplsearch files do require colons in IPv6 keys
16 (notated using the quote facility) so as to distinguish them from IPv4
17 keys. This meant that lookups for IP addresses in host lists did not work
18 for iplsearch lookups.
20 This has been fixed by arranging for IPv6 addresses to be expressed with
21 colons if the lookup type is iplsearch. This is not incompatible, because
22 previously such lookups could never work.
24 The situation is now rather anomolous, since one *can* have colons in
25 ordinary lsearch keys. However, making the change in all cases is
26 incompatible and would probably break a number of configurations.
28 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
31 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
32 conversion specification without a maximum field width, thereby enabling
33 a rogue spamd server to cause a buffer overflow. While nobody in their
34 right mind would setup Exim to query an untrusted spamd server, an
35 attacker that gains access to a server running spamd could potentially
36 exploit this vulnerability to run arbitrary code as the Exim user.
38 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
39 $primary_hostname instead of what libspf2 thinks the hosts name is.
41 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
42 a directory entry by the name of the lookup key. Previously, if a
43 symlink pointed to a non-existing file or a file in a directory that
44 Exim lacked permissions to read, a lookup for a key matching that
45 symlink would fail. Now it is enough that a matching directory entry
46 exists, symlink or not. (Bugzilla 503.)
48 PH/03 The body_linecount and body_zerocount variables are now exported in the
51 PH/04 Added the $dnslist_matched variable.
53 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
54 This means they are set thereafter only if the connection becomes
57 PH/06 Added the client_condition to authenticators so that some can be skipped
58 by clients under certain conditions.
60 PH/07 The error message for a badly-placed control=no_multiline_responses left
61 "_responses" off the end of the name.
63 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
65 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
66 (without spaces) instead of just copying the configuration text.
68 PH/10 Added the /noupdate option to the ratelimit ACL condition.
70 PH/11 Added $max_received_linelength.
72 PH/12 Added +ignore_defer and +include_defer to host lists.
74 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
75 way in which PCRE > 7.0 is built.
77 PH/14 Implemented queue_only_load_latch.
79 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
80 MAIL command. The effect was to mangle the value on 64-bit systems.
86 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
87 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
90 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
93 PH/02 In an ACL statement such as
95 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
97 if a client was not listed at all, or was listed with a value other than
98 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
99 the condition was not true (as it should be), so access was not denied.
100 The bug was that the ! inversion was incorrectly passed on to the second
101 item. This has been fixed.
103 PH/03 Added additional dnslists conditions == and =& which are different from
104 = and & when the dns lookup returns more than one IP address.
106 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
107 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
109 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
110 FSYNC, which compiles an option called disable_fsync that allows for
111 bypassing fsync(). The documentation is heavily laced with warnings.
113 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
115 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
116 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
117 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
118 including adding "make clean"; (3) Added -fPIC when compiling the test
119 dynamically loaded module, to get rid of a warning.
121 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
122 message fails, move_frozen_messages = true and ignore_bounce_errors_after
123 = 0s. The bug is otherwise harmless.
125 PH/07 There was a bug in the dovecot authenticator such that the value of
126 $auth1 could be overwritten, and so not correctly preserved, after a
127 successful authentication. This usually meant that the value preserved by
128 the server_setid option was incorrect.
130 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
132 PH/09 Installed PCRE release 7.0.
134 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
135 run for batched SMTP input. It is now run at the start of every message
136 in the batch. While fixing this I discovered that the process information
137 (output by running exiwhat) was not always getting set for -bs and -bS
138 input. This is fixed, and it now also says "batched" for BSMTP.
140 PH/11 Added control=no_pipelining.
142 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
143 patch, slightly modified), and move the expansion of helo_data till after
144 the connection is made in the smtp transport (so it can use these
147 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
149 PH/14 Added log_selector = +pid.
151 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
153 PH/16 Add ${if forany and ${if forall.
155 PH/17 Added dsn_from option to vary the From: line in DSNs.
157 PH/18 Flush SMTP output before performing a callout, unless control =
158 no_callout_flush is set.
160 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
161 was true (the default) a successful delivery failed to delete the retry
162 item, thus causing premature timeout of the address. The bug is now
165 PH/20 Added hosts_avoid_pipelining to the smtp transport.
167 PH/21 Long custom messages for fakedefer and fakereject are now split up
168 into multiline reponses in the same way that messages for "deny" and
169 other ACL rejections are.
171 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
172 with slight modification.
174 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
175 draft, changing the syntax and factoring some duplicate code".
177 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
178 for deliveries of the second and subsequent messages over the same SMTP
181 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
182 ${reduce, with only minor "tidies".
184 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
186 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
187 expansion side effects.
189 PH/27 When a message times out after an over-quota error from an Exim-imposed
190 quota, the bounce message says "mailbox is full". This message was not
191 being given when it was a system quota that was exceeded. It now should
194 MH/03 Made $recipients available in local_scan(). local_scan() already has
195 better access to the recipient list through recipients_list[], but
196 $recipients can be useful in postmaster-provided expansion strings.
198 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
199 in the case of a MAIL command with additional options following the
200 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
201 were accidentally chopped off.
203 PH/29 SMTP synchronization checks are implemented when a command is read -
204 there is a check that no more input is waiting when there shouldn't be
205 any. However, for some commands, a delay in an ACL can mean that it is
206 some time before the response is written. In this time, more input might
207 arrive, invalidly. So now there are extra checks after an ACL has run for
208 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
209 pipelining has not been advertised.
211 PH/30 MH's patch to allow iscntrl() characters to be list separators.
213 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
214 returned in the SMTP response when smtp_return_error_details was false.
217 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
218 instead of the C I/O that was originally supplied, because problems were
221 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
222 Exim which did not show up earlier: it was assuming that a call to
223 SSL_CTX_set_info_callback() might give an error value. In fact, there is
224 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
225 was a macro that became an assignment, so it seemed to work. This has
226 changed to a proper function call with a void return, hence the compile
227 error. Exim's code has been fixed.
229 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
232 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
234 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
236 PH/37 If a message is not accepted after it has had an id assigned (e.g.
237 because it turns out to be too big or there is a timeout) there is no
238 "Completed" line in the log. When some messages of this type were
239 selected by exigrep, they were listed as "not completed". Others were
240 picked up by some special patterns. I have improved the selection
241 criteria to be more general.
243 PH/38 The host_find_failed option in the manualroute router can now be set
244 to "ignore", to completely ignore a host whose IP address cannot be
245 found. If all hosts are ignored, the behaviour is controlled by the new
246 host_all_ignored option.
248 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
249 homing or because of multiple MX records with /mx) generated more than
250 one IP address, and the following item turned out to be the local host,
251 all the secondary addresses of the first item were incorrectly removed
252 from the list, along with the local host and any following hosts (which
253 is what is supposed to happen).
255 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
256 whoever called Exim into the -H file. In the case of the daemon it was
257 behaving confusingly. When first started, it used values for whoever
258 started the daemon, but after a SIGHUP it used the Exim user (because it
259 calls itself on a restart). I have changed the code so that it now always
262 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
263 message are rejected with the same error (e.g. no authentication or bad
264 sender address), and a DATA command is nevertheless sent (as can happen
265 with PIPELINING or a stupid MUA), the error message that was given to the
266 RCPT commands is included in the rejection of the DATA command. This is
267 intended to be helpful for MUAs that show only the final error to their
270 PH/42 Another patch from the Sieve maintainer.
272 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
273 Eximstats - Fixed some broken HTML links and added missing column headers
275 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
276 columns for Rejects, Temp Rejects, Ham, and Spam rows.
278 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
280 PH/43 Yet another patch from the Sieve maintainer.
282 PH/44 I found a way to check for a TCP/IP connection going away before sending
283 the response to the final '.' that terminates a message, but only in the
284 case where the client has not sent further data following the '.'
285 (unfortunately, this is allowed). However, in many cases there won't be
286 any further data because there won't be any more messages to send. A call
287 to select() can be used: if it shows that the input is "ready", there is
288 either input waiting, or the socket has been closed. An attempt to read
289 the next input character can distinguish the two cases. Previously, Exim
290 would have sent an OK response which the client would never have see.
291 This could lead to message repetition. This fix should cure that, at
292 least in a lot of common cases.
294 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
295 advertised in response to EHLO.
301 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
302 fixed by 4.65/MH/01 (is this a record?) are fixed:
304 (i) An empty string was always treated as zero by the numeric comparison
305 operators. This behaviour has been restored.
307 (ii) It is documented that the numeric comparison operators always treat
308 their arguments as decimal numbers. This was broken in that numbers
309 starting with 0 were being interpreted as octal.
311 While fixing these problems I realized that there was another issue that
312 hadn't been noticed. Values of message_size_limit (both the global option
313 and the transport option) were treated as octal if they started with 0.
314 The documentation was vague. These values are now always treated as
315 decimal, and I will make that clear in the documentation.
321 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
322 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
325 MH/01 Don't check that the operands of numeric comparison operators are
326 integers when their expansion is in "skipping" mode (fixes bug
327 introduced by 4.64-PH/07).
329 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
330 child addresses, Exim now panics and dies. Previously, because the count
331 is held in a short int, deliveries were likely to be lost. As such a
332 large number of recipients for a single message is ridiculous
333 (performance will be very, very poor), I have chosen to impose a limit
334 rather than extend the field.
340 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
341 leftover -K file (the existence of which was triggered by #402).
342 While we were at it, introduced process PID as part of the -K
343 filename. This should rule out race conditions when creating
346 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
347 processing considerably. Previous code took too long for large mails,
348 triggering a timeout which in turn triggers #401.
350 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
351 in the DK code in transports.c. sendfile() is not really portable,
352 hence the _LINUX specificness.
354 TF/01 In the add_headers option to the mail command in an Exim filter,
355 there was a bug that Exim would claim a syntax error in any
356 header after the first one which had an odd number of characters
359 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
360 callout verification, Exim cached a "reject" for the entire domain. This
361 is correct for most verifications, but it is not correct for a recipient
362 verification with use_sender or use_postmaster set, because in that case
363 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
364 case of MAIL FROM:<> rejection from other early rejections (e.g.
365 rejection of HELO). When verifying a recipient using a non-null MAIL
366 address, the cache is ignored if it shows MAIL FROM:<> rejection.
367 Whatever the result of the callout, the value of the domain cache is
368 left unchanged (for any other kind of callout, getting as far as trying
369 RCPT means that the domain itself is ok).
371 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
374 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
375 manifest itself as EPIPE rather than ECONNECT. When tidying away a
376 session, the daemon ignores ECONNECT errors and logs others; it now
377 ignores EPIPE as well.
379 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
380 (quoted-printable decoding).
382 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
383 later the small subsequent patch to fix an introduced bug.
385 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
387 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
389 PH/08 An error is now given if message_size_limit is specified negative.
391 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
392 to be given (somewhat) arbitrary names.
394 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
397 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
398 miscellaneous code fixes
400 PH/10 Added the log_reject_target ACL modifier to specify where to log
403 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
404 hostname. This is wrong, because it relates to the incoming message (and
405 probably the interface on which it is arriving) and not to the outgoing
406 callout (which could be using a different interface). This has been
407 changed to use the value of the helo_data option from the smtp transport
408 instead - this is what is used when a message is actually being sent. If
409 there is no remote transport (possible with a router that sets up host
410 addresses), $smtp_active_hostname is used.
412 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
413 tweaks were necessary in order to get it to work (see also 21 below):
414 (a) The code assumed that strncpy() returns a negative number on buffer
415 overflow, which isn't the case. Replaced with Exim's string_format()
417 (b) There were several signed/unsigned issues. I just did the minimum
418 hacking in of casts. There is scope for a larger refactoring.
419 (c) The code used strcasecmp() which is not a standard C function.
420 Replaced with Exim's strcmpic() function.
421 (d) The code set only $1; it now sets $auth1 as well.
422 (e) A simple test gave the error "authentication client didn't specify
423 service in request". It would seem that Dovecot has changed its
424 interface. Fortunately there's a specification; I followed it and
425 changed what the client sends and it appears to be working now.
427 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
430 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
431 address A is aliased to B and C, where B exists and C does not. Without
432 -v the output is "A verified" because verification stops after a
433 successful redirection if more than one address is generated. However,
434 with -v the child addresses are also verified. Exim was outputting "A
435 failed to verify" and then showing the successful verification for C,
436 with its parentage. It now outputs "B failed to verify", showing B's
437 parentage before showing the successful verification of C.
439 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
440 look up a TXT record in a specific list after matching in a combined
443 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
444 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
445 they consult the DNS. I had assumed they would set it the way they
446 wanted; and indeed my experiments on Linux seem to show that in some
447 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
448 To be on the safe side, however, I have now made the interface to
449 host_find_byname() similar to host_find_bydns(), with an argument
450 containing the DNS resolver options. The host_find_byname() function now
451 sets these options at its start, just as host_find_bydns() does. The smtp
452 transport options dns_qualify_single and dns_search_parents are passed to
453 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
454 of host_find_byname() use the default settings of RES_DEFNAMES
455 (qualify_single) but not RES_DNSRCH (search_parents).
457 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
458 spool_read_header() do less string testing, by means of a preliminary
459 switch on the second character of optional "-foo" lines. (This is
460 overdue, caused by the large number of possibilities that now exist.
461 Originally there were few.) While I was there, I also converted the
462 str(n)cmp tests so they don't re-test the leading "-" and the first
463 character, in the hope this might squeeze out yet more improvement.
465 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
466 flag allowing group syntax was set by the header_syntax check but not
467 turned off, possible causing trouble later; (2) The flag was not being
468 set at all for the header_verify test, causing "group"-style headers to
469 be rejected. I have now set it in this case, and also caused header_
470 verify to ignore an empty address taken from a group. While doing this, I
471 came across some other cases where the code for allowing group syntax
472 while scanning a header line wasn't quite right (mostly, not resetting
473 the flag correctly in the right place). These bugs could have caused
474 trouble for malformed header lines. I hope it is now all correct.
476 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
477 with the "reply" argument non-NULL. The code, however (which originally
478 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
479 but it didn't always do it. This confused somebody who was copying the
480 code for some other use. I have removed all the tests.
482 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
483 feature that was used to support insecure browsers during the U.S. crypto
484 embargo. It requires special client support, and Exim is probably the
485 only MTA that supported it -- and would never use it because real RSA is
486 always available. This code has been removed, because it had the bad
487 effect of slowing Exim down by computing (never used) parameters for the
488 RSA_EXPORT functionality.
490 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
491 authenticator to fail if there's a tab character in the incoming data
492 (there should never be unless someone is messing about, as it's supposed
493 to be base64-encoded). Also added, on Timo's advice, the "secured" option
494 if the connection is using TLS or if the remote IP is the same as the
495 local IP, and the "valid-client-cert option" if a client certificate has
498 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
499 authenticators. This can be used for authorization after authentication
500 succeeds. (In the case of plaintext, it servers for both authentication
503 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
504 if any retry times were supplied.
506 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
507 connection, where there is no client IP address to check. In this
508 situation, the verify now always succeeds.
510 PH/25 Applied John Jetmore's -Mset patch.
512 PH/26 Added -bem to be like -Mset, but loading a message from a file.
514 PH/27 In a string expansion for a processed (not raw) header when multiple
515 headers of the same name were present, leading whitespace was being
516 removed from all of them, but trailing whitespace was being removed only
517 from the last one. Now trailing whitespace is removed from each header
518 before concatenation. Completely empty headers in a concatenation (as
521 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
522 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
524 PH/29 [Removed. This was a change that I later backed out, and forgot to
525 correct the ChangeLog entry (that I had efficiently created) before
526 committing the later change.]
528 PH/30 Exim was sometimes attempting to deliver messages that had suffered
529 address errors (4xx response to RCPT) over the same connection as other
530 messages routed to the same hosts. Such deliveries are always "forced",
531 so retry times are not inspected. This resulted in far too many retries
532 for the affected addresses. The effect occurred only when there were more
533 hosts than the hosts_max_try setting in the smtp transport when it had
534 the 4xx errors. Those hosts that it had tried were not added to the list
535 of hosts for which the message was waiting, so if all were tried, there
536 was no problem. Two fixes have been applied:
538 (i) If there are any address or message errors in an SMTP delivery, none
539 of the hosts (tried or untried) are now added to the list of hosts
540 for which the message is waiting, so the message should not be a
541 candidate for sending over the same connection that was used for a
542 successful delivery of some other message. This seems entirely
543 reasonable: after all the message is NOT "waiting for some host".
544 This is so "obvious" that I'm not sure why it wasn't done
545 previously. Hope I haven't missed anything, but it can't do any
546 harm, as the worst effect is to miss an optimization.
548 (ii) If, despite (i), such a delivery is accidentally attempted, the
549 routing retry time is respected, so at least it doesn't keep
550 hammering the server.
552 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
553 in ${readsocket because some servers need this prod.
555 PH/32 Added some extra debug output when updating a wait-xxx database.
557 PH/33 The hint "could be header name not terminated by colon", which has been
558 given for certain expansion errors for a long time, was not being given
559 for the ${if def:h_colon_omitted{... case.
561 PH/34 The spec says: "With one important exception, whenever a domain list is
562 being scanned, $domain contains the subject domain." There was at least
563 one case where this was not true.
565 PH/35 The error "getsockname() failed: connection reset by peer" was being
566 written to the panic log as well as the main log, but it isn't really
567 panic-worthy as it just means the connection died rather early on. I have
568 removed the panic log writing for the ECONNRESET error when getsockname()
571 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
572 runs only) independently of the message's sender address. This meant
573 that, if the 4xx error was in fact related to the sender, a different
574 message to the same recipient with a different sender could confuse
575 things. In particualar, this can happen when sending to a greylisting
576 server, but other circumstances could also provoke similar problems.
577 I have changed the default so that the retry time for these errors is now
578 based a combination of the sender and recipient addresses. This change
579 can be overridden by setting address_retry_include_sender=false in the
582 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
583 remote server are returned as part of bounce messages. This was not
584 happening for LMTP over a pipe (the lmtp transport), but now it is the
585 same for both kinds of LMTP.
587 PH/38 Despite being documented as not happening, Exim was rewriting addresses
588 in header lines that were in fact CNAMEs. This is no longer the case.
590 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
591 and queue runs started by the daemon processed all messages. This has
592 been fixed so that -R and -S can now usefully be given with -q<time>.
594 PH/40 Import PCRE release 6.7 (fixes some bugs).
596 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
598 PH/42 Give an error if -q is specified more than once.
600 PH/43 Renamed the variables $interface_address and $interface_port as
601 $received_ip_address and $received_port, to make it clear that these
602 values apply to message reception, and not to the outgoing interface when
603 a message is delivered. (The old names remain recognized, of course.)
605 PH/44 There was no timeout on the connect() call when using a Unix domain
606 socket in the ${readsocket expansion. There now is.
608 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
609 be meaningful with "accept".
611 SC/01 Eximstats V1.43
612 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
614 SC/02 Eximstats V1.44
615 Use a glob alias rather than an array ref in the generated
616 parser. This improves both readability and performance.
618 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
619 Collect SpamAssassin and rejection statistics.
620 Don't display local sender or destination tables unless
621 there is data to show.
622 Added average volumes into the top table text output.
624 SC/04 Eximstats V1.46
625 Collect data on the number of addresses (recipients)
626 as well as the number of messages.
628 SC/05 Eximstats V1.47
629 Added 'Message too big' to the list of mail rejection
630 reasons (thanks to Marco Gaiarin).
632 SC/06 Eximstats V1.48
633 Mainlog lines which have GMT offsets and are too short to
634 have a flag are now skipped.
636 SC/07 Eximstats V1.49 (Alain Williams)
637 Added the -emptyok flag.
639 SC/08 Eximstats V1.50
640 Fixes for obtaining the IP address from reject messages.
642 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
643 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
644 whitesspace changes from 4.64-PH/27
646 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
649 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
650 are found, allow negative numbers in numeric criteria)
652 JJ/06 exipick.20061117.2, added new $message_body_missing variable
654 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
655 to match changes made in 4.64-PH/43
657 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
659 PH/47 Put in an explicit test for a DNS lookup of an address record where the
660 "domain" is actually an IP address, and force a failure. This locks out
661 those revolvers/nameservers that support "A-for-A" lookups, in
662 contravention of the specifications.
664 PH/48 When a host name was looked up from an IP address, and the subsequent
665 forward lookup of the name timed out, the host name was left in
666 $sender_host_name, contrary to the specification.
668 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
669 restricted to single-key lookups, Exim was not diagnosing an error if
670 * or *@ was used with a query-style lookup.
672 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
674 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
675 long ago, but noone interested enough thought of it. Let's just say that
676 the "1.1" means that there are some new functions that weren't there at
677 some point in the past.
679 PH/51 Error processing for expansion failure of helo_data from an smtp
680 transport during callout processing was broken.
682 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
683 tested/used via the -bh/-bhc/-bs options.
685 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
686 bug, fixed in subsequent PCRE releases).
688 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
689 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
691 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
697 SC/01 Use a glob alias rather than an array ref in eximstats generated
698 parser. This improves both readability and performance.
700 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
701 Don't display local sender or destination tables in eximstats unless
702 there is data to show.
703 Added average volumes into the eximstats top table text output.
705 SC/03 Collect data on the number of addresses (recipients) as well
706 as the number of messages in eximstats.
708 TF/01 Correct an error in the documentation for the redirect router. Exim
709 does (usually) call initgroups() when daemonizing.
711 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
712 with consistent privilege compared to when running as a daemon.
714 TF/03 Note in the spec that $authenticated_id is not set for local
715 submissions from trusted users.
717 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
718 Thanks to Dean Brooks <dean@iglou.com> for the patch.
720 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
721 by adding some example configuration directives to the default
722 configuration file. A little bit of work is required to uncomment the
723 directives and define how usernames and passwords are checked, but
724 there is now a framework to start from.
726 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
727 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
728 without this. I don't know how relevant this is to other LDAP libraries.
730 PH/02 Add the verb name to the "unknown ACL verb" error.
732 PH/03 Magnus Holmgren's patch for filter_prepend_home.
734 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
736 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
737 directory not expanded when it should be if an expanded home directory
738 was set for the address (which is overridden by the transport).
740 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
743 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
744 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
745 because it is too late at that time, and has no effect.
747 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
748 security issue with \' (bugzilla #107). I could not use the
749 PQescapeStringConn() function, because it needs a PGconn value as one of
752 PH/08 When testing addresses using -bt, indicate those final addresses that
753 are duplicates that would not cause an additional delivery. At least one
754 person was confused, thinking that -bt output corresponded to deliveries.
755 (Suppressing duplicates isn't a good idea as you lose the information
756 about possibly different redirections that led to the duplicates.)
758 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
759 systems where poll() doesn't work, in particular OS X.
761 PH/10 Added more information to debugging output for retry time not reached.
763 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
764 operations in malware.c.
766 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
769 PH/13 If write_rejectlog was set false when logging was sent to syslog with
770 syslog_duplication set false, log lines that would normally be written
771 both the the main log and to the reject log were not written to syslog at
774 PH/14 In the default configuration, change the use of "message" in ACL warn
775 statements to "add_header".
777 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
778 not followed by a command (e.g. "seen endif").
780 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
781 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
784 PH/17 Added extra conditions to the default value of delay_warning_condition
788 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
789 { match{$h_precedence:}{(?i)bulk|list|junk} } \
790 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
793 The Auto-Submitted: and various List- headers are standardised, whereas I
794 don't think Precedence: ever was.
796 PH/18 Refactored debugging code in route_finduser() to show more information,
797 in particular, the error code if getpwnam() issues one.
799 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
800 This is apparently needed in addition to the PH/07 change above to avoid
801 any possible encoding problems.
803 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
804 but not after initializing Perl.
806 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
807 output them only if debugging. By default they are written stderr,
808 apparently, which is not desirable.
810 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
813 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
816 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
818 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
819 authenticated or an ident call has been made. Suppress the default
820 values for $authenticated_id and $authenticated_sender (but permit -oMai
821 and -oMas) when testing with -bh.
823 PH/24 Re-jigged the order of the tests in the default configuration so that the
824 tests for valid domains and recipients precede the DNS black list and CSA
825 tests, on the grounds that those ones are more expensive.
827 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
828 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
829 This bug exists in every version of Exim that I still have, right back to
832 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
833 However, an attempt to turn on case-sensitivity in a regex key by
834 including (?-i) didn't work because the subject string was already
835 lowercased, and the effects were non-intuitive. It turns out that a
836 one-line patch can be used to allow (?-i) to work as expected.
842 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
843 other effects) broke the use of negated acl sub-conditions.
845 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
848 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
849 "Deny" causes Exim to reject the incoming connection with a 554 error.
850 Unfortunately, if there is a major crisis, such as a disk failure,
851 tcp-wrappers gives "deny", whereas what one would like would be some
852 kind of temporary error. A kludge has been added to help with this.
853 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
854 554 error is used if errno is still zero or contains ENOENT (which occurs
855 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
858 PH/03 Add -lutil to the default FreeBSD LIBS setting.
860 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
861 errors. Otherwise a message that provokes a temporary error (when other
862 messages do not) can cause a whole host to time out.
864 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
865 addresses were routed directly to files or pipes from a redirect router.
866 File deliveries just didn't batch; pipe deliveries might have suffered
869 PH/06 A failure to get a lock for a hints database would erroneously always say
870 "Failed to get write lock", even when it was really a read lock.
872 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
873 of 0600. This has been changed to use the value of the lockfile_mode
874 option (which defaults to 0600).
876 PH/08 Applied small patch from the Sieve maintainer.
878 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
879 folder from quota calculations, a direct delivery into this folder messed
880 up the contents of the maildirsize file. This was because the regex was
881 used only to exclude .Trash (or whatever) when the size of the mailbox
882 was calculated. There was no check that a delivery was happening into an
883 excluded directory. This bug has been fixed by ignoring all quota
884 processing for deliveries into excluded directories.
886 PH/10 Added the maildirfolder_create_regex option to appendfile.
892 PH/01 The code for finding all the local interface addresses on a FreeBSD
893 system running IPv6 was broken. This may well have applied to all BSD
894 systems, as well as to others that have similar system calls. The broken
895 code found IPv4 interfaces correctly, but gave incorrect values for the
896 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
897 that it would not match correctly against @[] and not recognize the IPv6
900 PH/02 The ipliteral router was not recognizing addresses of the form user@
901 [ipv6:....] because it didn't know about the "ipv6:" prefix.
903 PH/03 Added disable_ipv6.
905 PH/04 Changed $reply_address to use the raw form of the headers instead of the
906 decoded form, because it is most often used to construct To: headers
907 lines in autoreplies, and the decoded form may well be syntactically
908 invalid. However, $reply_address has leading white space removed, and all
909 newlines turned into spaces so that the autoreply transport does not
912 PH/05 If group was specified without a user on a router, and no group or user
913 was specified on a transport, the group from the router was ignored.
915 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
916 for visible compile-time settings that can be used to change these
917 numbers, for those that want even more. Backwards compatibility with old
918 spool files has been maintained. However, going back to a previous Exim
919 release will lost any variables that are in spool files.
921 PH/07 Two small changes when running in the test harness: increase delay when
922 passing a TCP/IP connection to a new process, in case the original
923 process has to generate a bounce, and remove special handling of
924 127.0.0.2 (sic), which is no longer necessary.
926 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
927 be the same on different OS.
929 PH/09 Moved a debug statement in filter processing to avoid a race problem when
932 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
933 whether --show-vars was specified or not
935 JJ/02 exipick: Added support for new ACL variable spool format introduced
938 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
939 syntactically invalid From: or Reply-to: line, and a filter used this to
940 generate an autoreply, and therefore failed to obtain an address for the
941 autoreply, Exim could try to deliver to a non-existent relative file
942 name, causing unrelated and misleading errors. What now happens is that
943 it logs this as a hard delivery error, but does not attempt to create a
946 PH/11 The exinext utility has a -C option for testing purposes, but although
947 the given file was scanned by exinext itself; it wasn't being passed on
948 when Exim was called.
950 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
951 an end-of-file indication when reading a command response.
953 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
954 compiled. In many other places in Exim, IPv6 addresses are always
955 recognized, so I have changed this. It also means that IPv4 domain
956 literals of the form [IPV4:n.n.n.n] are now always recognized.
958 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
959 used if the router is not running as root, for example, when verifying at
960 ACL time, or when using -bh. The debugging output from this situation was
961 non-existent - all you got was a failure to exec. I have made two
964 (a) Failures to set uid/gid, the current directory, or a process leader
965 in a subprocess such as that created by queryprogram now generate
966 suitable debugging ouput when -d is set.
968 (b) The queryprogram router detects when it is not running as root,
969 outputs suitable debugging information if -d is set, and then runs
970 the subprocess without attempting to change uid/gid.
972 PH/15 Minor change to Makefile for building test_host (undocumented testing
975 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
976 additional section of a DNS packet that returns MX or SRV records.
977 Instead, it always explicitly searches for A/AAAA records. This avoids
978 major problems that occur when a DNS server includes only records of one
979 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
980 fixed another bug: if SRV records were looked up and the corresponding
981 address records were *not* found in the additional section, the port
982 values from the SRV records were lost.
984 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
985 using the correct key (the original address) when searching the retry
986 rules in order to find which one to use for generating the retry hint.
988 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
989 adding the default one. Similarly, if it contains a Reply-To: header, the
990 errors_reply_to option, if set, is not used.
992 PH/19 When calculating a retry time, Exim used to measure the "time since
993 failure" by looking at the "first failed" field in the retry record. Now
994 it does not use this if it is later than than the arrival time of the
995 message. Instead it uses the arrival time. This makes for better
996 behaviour in cases where some deliveries succeed, thus re-setting the
997 "first failed" field. An example is a quota failure for a huge message
998 when small messages continue to be delivered. Without this change, the
999 "time since failure" will always be short, possible causing more frequent
1000 delivery attempts for the huge message than are intended.
1001 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1003 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1004 $1, $2, $3) because the numerical variables can be reset during some
1005 expansion items (e.g. "match"), thereby losing the authentication data.
1007 PH/21 Make -bV show the size of off_t variables so that the test suite can
1008 decide whether to run tests for quotas > 2G.
1010 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1011 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1012 filecount value is greater than 2G or if a quota value is greater than 2G
1013 on a system where the size of off_t is not greater than 4, a panic error
1016 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1017 never match. The debug and -bh output now contains an explicit error
1018 message indicating a malformed IPv4 address or mask.
1020 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1021 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1022 PH/23 above applies.
1024 PH/25 Do not write to syslog when running in the test harness. The only
1025 occasion when this arises is a failure to open the main or panic logs
1026 (for which there is an explicit test).
1028 PH/26 Added the /no_tell option to "control=freeze".
1030 PH/27 If a host name lookup failed very early in a connection, for example, if
1031 the IP address matched host_lookup and the reverse lookup yielded a name
1032 that did not have a forward lookup, an error message of the form "no IP
1033 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1034 could be logged. Now it outputs the IP address instead of "NULL".
1036 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1037 allows the sender and the authenticated sender to be set when
1038 submitting a message from within Exim. Since child_open_exim() is
1039 documented for local_scan(), the new function should be too.
1041 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1042 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1043 results in an empty string is now treated as unset.
1045 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1047 PH/31 Added sender_verify_fail logging option.
1049 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1050 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1051 tidied the source and removed it altogether.
1053 PH/33 When a queue run was abandoned because the load average was too high, a
1054 log line was always written; now it is written only if the queue_run log
1055 selector is set. In addition, the log line for abandonment now contains
1056 information about the queue run such as the pid. This is always present
1057 in "start" and "stop" lines but was omitted from the "abandon" line.
1059 PH/34 Omit spaces between a header name and the colon in the error message that
1060 is given when verify = headers_syntax fails (if there are lots of them,
1061 the message gets confusing).
1063 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1064 names, as there are now some PTR records that contain slashes. This check
1065 is only to protect against broken name servers that fall over on strange
1066 characters, so the fact that it applies to all lookups doesn't matter.
1068 PH/36 Now that the new test suite is complete, we can remove some of the
1069 special code in Exim that was needed for the old test suite. For example,
1070 sorting DNS records because real resolvers return them in an arbitrary
1071 order. The new test suite's fake resolver always returns records in the
1074 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1075 bounces) except when queue_only is set, to avoid logging races between
1076 the different processes.
1078 PH/38 Panic-die if .include specifies a non-absolute path.
1080 PH/39 A tweak to the "H" retry rule from its user.
1082 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1083 a label. They prevented compilation on older perls.
1085 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1086 a warning to be raised on newish perls.
1088 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1089 on queue. Changes to match documented behaviour of showing count of
1090 messages matching specified criteria.
1092 PH/40 Changed the default ident timeout from 30s to 5s.
1094 PH/41 Added support for the use of login_cap features, on those BSD systems
1095 that have them, for controlling the resources used by pipe deliveries.
1097 PH/42 The content-scanning code uses fopen() to create files in which to put
1098 message data. Previously it was not paying any attention to the mode of
1099 the files. Exim runs with umask(0) because the rest of the code creates
1100 files with open(), and sets the required mode explicitly. Thus, these
1101 files were ending up world-writeable. This was not a big issue, because,
1102 being within the spool directory, they were not world-accessible. I have
1103 created a function called modefopen, which takes an additional mode
1104 argument. It sets umask(777), creates the file, chmods it to the required
1105 mode, then resets the umask. All the relevant calls to fopen() in the
1106 content scanning code have been changed to use this function.
1108 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1109 to 24 hours. This avoids potential overflow problems when processing G
1110 and H retry rules. I suspect nobody ever tinkers with this value.
1112 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1114 PH/45 When the plaintext authenticator is running as a client, the server's
1115 challenges are checked to ensure they are valid base64 strings. By
1116 default, the authentication attempt is cancelled if an invalid string is
1117 received. Setting client_ignore_invalid_base64 true ignores these errors.
1118 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1119 they are received. Thus, the responses can be made to depend on the
1120 challenges. If an invalid string is ignored, an empty string is placed in
1123 PH/46 Messages that are created by the autoreply transport now contains a
1124 References: header, in accordance with RFCs 2822 and 3834.
1126 PH/47 Added authenticated_sender_force to the smtp transport.
1128 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1130 PH/49 Installed latest patch from the Sieve maintainer.
1132 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1133 was also set, the appendfile transport was unnecessarily scanning a
1134 directory of message files (e.g. for maildir delivery) to find the count
1135 of files (along with the size), even though it did not need this
1136 information. It now does the scan only if it needs to find either the
1137 size of the count of files.
1139 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1141 PH/52 Two bugs concerned with error handling when the smtp transport is
1144 (i) Exim was not creating retry information for temporary errors given
1145 for individual recipients after the DATA command when the smtp transport
1146 was used in LMTP mode. This meant that they could be retried too
1147 frequently, and not timed out correctly.
1149 (ii) Exim was setting the flag that allows error details to be returned
1150 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1151 recipients that were returned after the DATA command.
1153 PH/53 This is related to PH/52, but is more general: for any failing address,
1154 when detailed error information was permitted to be returned to the
1155 sender, but the error was temporary, then after the final timeout, only
1156 "retry timeout exceeded" was returned. Now it returns the full error as
1157 well as "retry timeout exceeded".
1159 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1160 do this, and (what is worse) MTAs that accept it.
1162 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1163 will now be deprecated.
1165 PH/56 New os.c-cygwin from the Cygwin maintainer.
1167 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1168 formats (previously only available in exim formats via -bpr, -bpru,
1169 and -bpra. Now also available in native and exiqgrep formats)
1171 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1172 with very large, slow to parse queues
1174 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1176 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1178 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1179 responses to authentication challenges, though it was showing the
1180 challenges; (ii) I've removed the CR characters from the debug output for
1183 PH/58 Allow for the insertion of a newline as well as a space when a string
1184 is turned into more than one encoded-word during RFC 2047 encoding. The
1185 Sieve code now uses this.
1187 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1188 data_4xx, lost_connection, tls_required.
1190 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1191 message was being sent as an SMTP response.
1193 PH/61 Add -l and -k options to exicyclog.
1195 PH/62 When verifying, if an address was redirected to one new address, so that
1196 verification continued, and the new address failed or deferred after
1197 having set something in $address_data, the value of $address_data was not
1198 passed back to the ACL. This was different to the case when no
1199 redirection occurred. The value is now passed back in both cases.
1201 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1202 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1203 use login_cap.h, so on its own it isn't the distinguishing feature. The
1204 new name refers directly to the setclassresources() function.
1206 PH/65 Added configuration files for NetBSD3.
1208 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1210 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1211 is preferred over IPv4.
1213 PH/68 The bounce_return_message and bounce_return_body options were not being
1214 honoured for bounces generated during the reception of non-SMTP messages.
1215 In particular, this applied to messages rejected by the ACL. This bug has
1216 been fixed. However, if bounce_return_message is true and bounce_return_
1217 body is false, the headers that are returned for a non-SMTP message
1218 include only those that have been read before the error was detected.
1219 (In the case of an ACL rejection, they have all been read.)
1221 PH/69 The HTML version of the specification is now built in a directory called
1222 spec_html instead of spec.html, because the latter looks like a path with
1223 a MIME-type, and this confuses some software.
1225 PH/70 Catch two compiler warnings in sieve.c.
1227 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1228 function verify_get_ident() calls ip_connect() to connect a socket, but
1229 if the "connect()" function timed out, ip_connect() used to close the
1230 socket. However, verify_get_ident() also closes the socket later, and in
1231 between Exim writes to the log, which may get opened at this point. When
1232 the socket was closed in ip_connect(), the log could get the same file
1233 descriptor number as the socket. This naturally causes chaos. The fix is
1234 not to close the socket in ip_connect(); the socket should be closed by
1235 the function that creates it. There was only one place in the code where
1236 this was missing, in the iplookup router, which I don't think anybody now
1237 uses, but I've fixed it anyway.
1239 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1240 well as to direct DNS lookups. Otherwise the handling of names in host
1241 lists is inconsistent and therefore confusing.
1247 PH/01 Two changes to the default runtime configuration:
1249 (1) Move the checks for relay_from_hosts and authenticated clients from
1250 after to before the (commented out) DNS black list checks.
1252 (2) Add control=submission to the relay_from_hosts and authenticated
1253 clients checks, on the grounds that messages accepted by these
1254 statements are most likely to be submissions.
1256 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1258 (1) Generate an error if the third argument for the ${prvs expansion is
1261 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1264 (3) Reset the variables that are obtained from the first argument of
1265 ${prvscheck and used in the second argument before leaving the code,
1266 because their memory is reclaimed, so using them afterwards may do
1269 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1270 one (it's much easier than Tom thought :-).
1272 (5) Because of (4), we can now allow for the use of $prvscheck_result
1273 inside the third argument.
1275 PH/03 For some reason, the default setting of PATH when running a command from
1276 a pipe transport was just "/usr/bin". I have changed it to
1279 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1280 anything to be listed in the output from -bV.
1282 PH/05 When a filter generated an autoreply, the entire To: header line was
1283 quoted in the delivery log line, like this:
1285 => >A.N.Other <ano@some.domain> <original@ddress> ...
1287 This has been changed so that it extracts the operative address. There
1288 may be more than one such address. If so, they are comma-separated, like
1291 => >ano@some.domain,ona@other.domain <original@ddress> ...
1293 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1294 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1295 not being looked up in the rDNS to get a host name, Exim was showing the
1296 IP address twice in Received: lines, even though the IP addresses were
1297 identical. For example:
1299 Received: from [1.2.3.4] (helo=[1.2.3.4])
1301 However, if the real host name was known, it was omitting the HELO data
1302 if it matched the actual IP address. This has been tidied up so that it
1303 doesn't show the same IP address twice.
1305 PH/07 When both +timestamp and +memory debugging was on, the value given by
1306 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1307 called by the debug printing, thereby overwriting the timestamp buffer.
1308 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1310 PH/08 When the original message was included in an autoreply transport, it
1311 always said "this is a copy of the message, including all the headers",
1312 even if body_only or headers_only was set. It now gives an appropriate
1315 PH/09 Applied a patch from the Sieve maintainer which:
1317 o fixes some comments
1318 o adds the (disabled) notify extension core
1319 o adds some debug output for the result of if/elsif tests
1320 o points to the current vacation draft in the documentation
1321 and documents the missing references header update
1325 o fixes a bug in processing the envelope test (when testing
1326 multiple envelope elements, the last element determinted the
1329 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1330 Electronic Mail") by including:
1332 Auto-submitted: auto-generated
1334 in the messages that it generates (bounce messages and others, such as
1335 warnings). In the case of bounce messages for non-SMTP mesages, there was
1336 also a typo: it was using "Auto_submitted" (underscore instead of
1337 hyphen). Since every message generated by Exim is necessarily in response
1338 to another message, thes have all been changed to:
1340 Auto-Submitted: auto-replied
1342 in accordance with these statements in the RFC:
1344 The auto-replied keyword:
1346 - SHOULD be used on messages sent in direct response to another
1347 message by an automatic process,
1349 - MUST NOT be used on manually-generated messages,
1351 - MAY be used on Delivery Status Notifications (DSNs) and Message
1352 Disposition Notifications (MDNs),
1354 - MUST NOT be used on messages generated by automatic or periodic
1355 processes, except for messages which are automatic responses to
1358 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1359 to the default Received: header definition.
1361 PH/12 Added log selector acl_warn_skipped (default on).
1363 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1364 variables because (a) they are in the wrong storage pool and (b) even if
1365 they were copied, it wouldn't work properly because of the caching.
1367 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1368 checking when decoding. Apparently there are clients that generate
1369 overlong encoded strings. Why am I not surprised?
1371 PH/15 If the first argument of "${if match_address" was not empty, but did not
1372 contain an "@" character, Exim crashed. Now it writes a panic log message
1373 and treats the condition as false.
1375 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1377 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1378 extension "envelope-auth". The code is finished and in agreement with
1379 other implementations, but there is no documentation so far and in fact,
1380 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1381 not changing the active code.
1383 Print executed "if" and "elsif" statements when debugging is used. This
1384 helps a great deal to understand what a filter does.
1386 Document more things not specified clearly in RFC3028. I had all this
1387 sorted out, when out of a sudden new issues came to my mind. Oops."
1389 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1392 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1393 canonical form (as documented). However, after a host name lookup from
1394 the IP address, check_host() was doing a simple string comparison with
1395 addresses acquired from the DNS when checking that the found name did
1396 have the original IP as one of its addresses. Since any found IPv6
1397 addresses are likely to be in abbreviated form, the comparison could
1398 fail. Luckily, there already exists a function for doing the comparison
1399 by converting both addresses to binary, so now that is used instead of
1400 the text comparison.
1402 PH/20 There was another similar case to PH/19, when a complete host name was
1403 given in a host list; looking up its IP address could give an abbreviated
1404 form, whereas the current host's name might or might not be abbreviated.
1405 The same fix has been applied.
1411 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1412 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1415 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1416 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1418 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1420 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1421 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1422 Also, for "Auto-Submitted", treat anything other than "no" as
1423 non-personal, in accordance with RFC 3834. (Previously it treated
1424 anything starting "auto-" as non-personal.)
1426 TF/01 The control=submission/name=... option had a problem with syntax
1427 errors if the name included a slash character. The /name= option
1428 now slurps the rest of the string, so it can include any characters
1429 but it must come last in the list of options (after /sender_retain
1432 PH/05 Some modifications to the interface to the fake nameserver for the new
1440 TK/01 Added the "success_on_redirect" address verification option. See
1441 NewStuff for rationale and an example.
1443 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1445 PH/02 Patch to exigrep to allow it to work on syslog lines.
1447 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1448 fread() to skip over the body file's header line, because in Cygwin the
1449 header line is locked and is inaccessible.
1451 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1452 co-exist for some time) to make it clear that it is the Exim ID that is
1453 referenced, not the Message-ID: header line.
1455 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1456 string_format() function, because snprintf() does not exist on all
1459 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1460 ${stat: expansion item.
1462 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1463 protocol synchronization error", to keep the pedants happy.
1465 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1466 well as for IRIX systems, when gcc is being used. See the host.c source
1469 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1471 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1474 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1475 transport and to the smtp transport in LMTP mode.
1477 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1479 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1480 run a filter in a subprocess. This could lead to confusion in subsequent
1481 lookups in the parent process. There should also be a search_tidyup() at
1482 the end of the subprocess.
1484 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1485 only if the host matched helo_try_verify_hosts, which caused the
1486 verification to occur when the EHLO/HELO command was issued. The ACL just
1487 tested the remembered result. Now, if a previous verification attempt has
1488 not happened, "verify = helo" does it there and then.
1490 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1492 TK/03 Fix log output including CR from clamd.
1494 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1495 exist provoked a memory error which could cause a segfault.
1497 PH/15 Installed PCRE 6.2
1499 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1501 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1502 of the problem. Specifically, suggested +O2 rather than +O1 for the
1505 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1507 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1508 record, the port was not correctly logged when the outgoing_port log
1509 selector was set (it logged the transort's default port).
1511 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1512 fallback_hosts, and "hosts" in the smtp transport.
1514 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1515 host errors such as "Connection refused".
1517 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1518 authentication with radiusclient 0.4.9:
1520 - Error returned from rc_read_config was caught wrongly
1521 - Username/password not passed on to radius server due to wrong length.
1523 The presumption is that some radiusclient API changes for 4.51/PH/17
1524 were not taken care of correctly. The code is still untested by me (my
1525 Linux distribution still has 0.3.2 of radiusclient), but it was
1526 contributed by a Radius user.
1528 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1529 expanding the "port" option of the smtp transport.
1531 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1532 while reading a MIME header. Thanks to Tom Hughes for a patch.
1534 PH/24 Include config.h inside local_scan.h so that configuration settings are
1537 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1538 that in an ACL for RCPT (for example), you can examine exactly what was
1541 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1542 commands, but it was not correctly comparing the address with the actual
1543 client host address. Thus, it would show the EHLO address in Received:
1544 header lines when this was not necessary.
1546 PH/27 Added the % operator to ${eval:}.
1548 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1549 it should be ignoring failures (because with -C, for example, it has lost
1550 privilege). It wasn't ignoring creation failures other than "already
1553 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1556 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1557 because an input line was too long, either on its own, or by virtue of
1558 too many continuations, the temporary file was not being removed, and the
1559 return code was incorrect.
1561 PH/31 Missing "BOOL" in function definition in filtertest.c.
1563 PH/32 Applied Sieve patches from the maintainer.
1565 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1567 PH/33 Added "verify = not_blind".
1569 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1570 Local/Makefile (with some defaults set). These are used in built scripts
1571 such as exicyclog, but they have never been used in the exim_install
1572 script (though there are many overriding facilities there). I have
1573 arranged that the exim_install script now takes note of these two
1576 PH/35 Installed configuration files for Dragonfly.
1578 PH/36 When a locally submitted message by a trusted user did not contain a
1579 From: header, and the sender address was obtained from -f or from an SMTP
1580 MAIL command, and the trusted user did not use -F to supply a sender
1581 name, $originator_name was incorrectly used when constructing a From:
1582 header. Furthermore, $originator_name was used for submission mode
1583 messages from external hosts without From: headers in a similar way,
1584 which is clearly wrong.
1586 PH/37 Added control=suppress_local_fixups.
1588 PH/38 When log_selector = +received_sender was set, and the addition of the
1589 sender made the log line's construction buffer exactly full, or one byte
1590 less than full, an overflow happened when the terminating "\n" was
1593 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1594 when the result of a list match is failure because a DNS lookup failed.
1596 PH/40 RM_COMMAND is now used in the building process.
1598 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1599 the "build-* directories that it finds.
1601 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1602 address in a domain literal was a prefix of an interface address.
1604 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1605 when verifying a sender address, unless rewrite_headers is false.
1607 PH/44 Wrote a long comment about why errors_to addresses are verified as
1608 recipients, not senders.
1610 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1611 the ratelimit ACL was added.
1613 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1615 PH/46 Added extra information about PostgreSQL errors to the error string.
1617 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1618 avoiding the need to install special zones in a real server. This is
1619 backwards compatible; if it can't find the fake resolver, it drops back.
1620 Thus, both old and new test suites can be run.
1622 TF/02 Added util/ratelimit.pl
1624 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1625 clock is set back in time.
1627 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1628 Candler <B.Candler@pobox.com>.
1630 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1631 OK for addresses that are the result of redirections.
1633 PH/48 A number of further additions for the benefit of the new test suite,
1634 including a fake gethostbyname() that interfaces to the fake DNS resolver
1637 TF/06 The fix for widen_domains has also been applied to qualify_single and
1638 search_parents which are the other dnslookup options that can cause
1641 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1644 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1646 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1647 service" if it encountered an MX record with an empty target hostname.
1648 The message is now "an MX or SRV record indicated no SMTP service".
1650 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1651 if the DNS of the sending site is misconfigured. This is quite a
1652 common situation. This change restores the behaviour of treating a
1653 helo verification defer as a failure.
1655 PH/51 If self=fail was set on a router, the bounce message did not include the
1656 actual error message.
1662 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1664 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1665 command itself ended in error, the underlying message about the transport
1666 filter timeout was being overwritten with the pipe command error. Now the
1667 underlying error message should be appended to the second error message.
1669 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1671 PH/02 Reduce the amount of output that "make" produces by default. Full output
1672 can still be requested.
1674 PH/03 The warning log line about a condition test deferring for a "warn" verb
1675 was being output only once per connection, rather than after each
1676 occurrence (because it was using the same function as for successful
1677 "warn" verbs). This seems wrong, so I have changed it.
1679 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1680 it should not have, which might have caused a crash in the right
1681 circumstances, but probably never did.
1683 PH/04 Installed a modified version of Tony Finch's patch to make submission
1684 mode fix the return path as well as the Sender: header line, and to
1685 add a /name= option so that you can make the user's friendly name appear
1688 TF/03 Added the control = fakedefer ACL modifier.
1690 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1691 Mark Lowes for thorough testing.
1693 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1695 TK/03 Merged latest SRS patch from Miles Wilton.
1697 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1698 with the definition in sysexits.h (which is #included earlier).
1699 Fortunately, Exim does not actually use EX_OK. The code used to try to
1700 preserve the sysexits.h value, by assumimg that macro definitions were
1701 scanned for macro replacements. I have been disabused of this notion,
1702 so now the code just undefines EX_OK before #including unistd.h.
1704 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1705 in the smtp transport. When a block could not be written in a single
1706 write() function, the timeout was being re-applied to each part-write.
1707 This seems wrong - if the receiver was accepting one byte at a time it
1708 would take for ever. The timeout is now adjusted when this happens. It
1709 doesn't have to be particularly precise.
1711 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1712 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1714 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1715 without a domain if the check to <postmaster@domain> fails.
1717 SC/01 Eximstats: added -xls and the ability to specify output files
1718 (patch written by Frank Heydlauf).
1720 SC/02 Eximstats: use FileHandles for outputing results.
1722 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1724 SC/04 Eximstats: fixed display of large numbers with -nvr option
1726 SC/05 Eximstats: fixed merging of reports with empty tables.
1728 SC/06 Eximstats: added the -include_original_destination flag
1730 SC/07 Eximstats: removed tabs and trailing whitespace.
1732 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1734 TK/06 MBOX spool code: Add real "From " MBOX separator line
1735 so the .eml file is really in mbox format (even though
1736 most programs do not really care). Patch from Alex Miller.
1738 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1739 The latter is generated from $received_to and is only set if the
1740 message has one envelope recipient. SA can use these headers,
1741 obviously out-of-the-box. Patch from Alex Miller.
1743 PH/08 The ${def test on a variable was returning false if the variable's
1744 value was "0", contrary to what the specification has always said!
1745 The result should be true unless the variable is empty.
1747 PH/09 The syntax error of a character other than { following "${if
1748 def:variable_name" (after optional whitespace) was not being diagnosed.
1749 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1750 accidental colon was present, for example, could give incorrect results.
1752 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1753 result is used (not including appendfile, where other changes are about
1756 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1757 This involved changing a lot of size variables from int to off_t. It
1758 should work with maildirs and everything.
1760 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1761 spamd dying while we are connected to it.
1763 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1766 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1767 function for BDB 4.3.
1769 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1771 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1772 the sources for PCRE have been re-arranged and more files are now
1775 PH/15 The code I had for printing potentially long long variables in PH/11
1776 above was not the best (it lost precision). The length of off_t variables
1777 is now inspected at build time, and an appropriate printing format (%ld
1778 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1779 to be "long long int" or "long int". This is needed for the internal
1780 formatting function string_vformat().
1782 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1783 the configuration file to be ":syslog", then the script "guesses" where
1784 the logs files are, rather than using the compiled in default. In our
1785 case the guess is not the same as the compiled default, so the script
1786 suddenly stopped working when I started to use syslog. The patch checks
1787 to see if log_file_path is "". If so, it attempts to read it from exim
1788 with no configuration file to get the compiled in version, before it
1789 falls back to the previous guessing code."
1791 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1792 implementing BATV in an Exim configuration. See NewStuff for the gory
1795 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1796 Makefile that are specific to HP-UX.
1798 PH/18 If the "use_postmaster" option was set for a recipient callout together
1799 with the "random" option, the postmaster address was used as the MAIL
1800 FROM address for the random test, but not for the subsequent recipient
1801 test. It is now used for both.
1803 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1804 patch removes a few documentation additions to RFC 3028, because the
1805 latest draft now contains them. It adds the new en;ascii-case comparator
1806 and a new error check for 8bit text in MIME parts. Comparator and
1807 require names are now matched exactly. I enabled the subaddress
1808 extension, but it is not well tested yet (read: it works for me)."
1810 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1811 rework some of the code of TK/09 above to avoid the hardwired use of
1812 "%lld" and "long long". Replaced the call to snprintf() with a call to
1815 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1816 records point to non-existent hosts", "retry timeout exceeded", and
1817 "retry time not reached for any host after a long failure period".
1819 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1820 experimental DomainKeys support:
1822 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1823 (2) On an error such as an illegally used "control", the wrong name for
1824 the control was given.
1826 These problems did NOT occur unless DomainKeys support was compiled.
1828 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1830 PH/24 Added ${if match_ip condition.
1832 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1833 databases so that it will be absolutely obvious if a crash occurs in the
1834 DB library. This is a regular occurrence (often caused by mis-matched
1837 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1838 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1839 picked up on a user's system that detects such things. There doesn't seem
1840 to be a gcc warning option for this - only an attribute that has to be
1841 put on the function's prototype. It seems that in Fedora Core 4 they have
1842 set this on a number of new functions. No doubt there will be more in due
1845 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1846 specify a transport. However, if an address that was verified by such a
1847 router was the subject of a callout, Exim crashed because it tried to
1848 read the rcpt_include_affixes from the non-existent transport. Now it
1849 just assumes that the setting of that option is false. This bug was
1850 introduced by 4.51/PH/31.
1852 PH/28 Changed -d+all to exclude +memory, because that information is very
1853 rarely of interest, but it makes the output a lot bigger. People tend to
1854 do -d+all out of habit.
1856 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1857 code in os-type was giving problems when libc.so lives in lib64, like on
1860 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1861 aren't the modern standard, and it seems that some systems' include files
1862 don't always have them. Exim was already checking for some of the newer
1863 ones like T_AAAA, and defining it itself. I've added checks for all the
1864 record types that Exim uses.
1866 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1867 not automatically generating a new one, as it is supposed to. This
1868 prevented TLS from working. If the file did exist, but contained invalid
1869 data, a new version was generated, as expected. It was only the case of a
1870 non-existent file that was broken.
1872 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1873 with a change in libdomainkeys > 0.64.
1875 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1876 from DNS. If the selector record carries the flag, it now has
1877 precedence over the domain-wide flag.
1879 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1881 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1882 the use of a port name that isn't defined in /etc/services) occurred, the
1883 message was deferred as in a normal delivery, and thus remained on the
1884 spool, instead of being failed because of the mua_wrapper setting. This
1885 is now fixed, and I tidied up some of the mua_wrapper messages at the
1888 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1889 the messages in a hash of arrays rather than using individual hashes.
1890 This is a bit cleaner and results in dramatic memory savings, albeit
1891 at a slight CPU cost.
1893 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1894 as requested by Marc Sherman.
1896 SC/10 Eximstats: added histograms for user specified patterns as requested
1899 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1901 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1902 fopen() in the content-scanning modules that did not already have it.
1908 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1909 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1911 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1913 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1915 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1916 to test Sieve filters that use "vacation".
1918 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1919 that changes the way the GnuTLS parameters are stored in the cache file.
1920 The new format can be generated externally. For backward compatibility,
1921 if the data in the cache doesn't make sense, Exim assumes it has read an
1922 old-format file, and it generates new data and writes a new file. This
1923 means that you can't go back to an older release without removing the
1926 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1927 work if there are any delivery delays because "one_time" forces the
1928 parent to be marked "delivered", so its unseen clone is never tried
1929 again. For this reason, Exim now forbids the simultaneous setting of
1932 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1933 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1934 entry if you want to know the details. The fix, however, neglected to
1935 consider the case when local delivery batching is involved. The test for
1936 "previously delivered" was not happening when checking to see if an
1937 address could be batched with a previous (undelivered) one; under
1938 certain circumstances this could lead to multiple deliveries to the same
1941 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1942 in its include files, and this causes problems building Exim.
1944 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1945 header_syntax) but Exim was just ignoring anything given after a slash.
1946 In particular, this caused confusion with an attempt to use "verify =
1947 reverse_host_lookup/defer_ok". An error is now given when options are
1948 supplied for verify items that do not have them. (Maybe reverse_host_
1949 lookup should have a defer_ok option, but that's a different point.)
1951 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1952 defined by RFC 821) to 2048, because there were problems with some AUTH
1953 commands, and RFC 1869 says the size should be increased for extended
1954 SMTP commands that take arguments.
1956 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1959 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1960 "unknown" error; now it says that the functionality isn't in the binary.
1962 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1963 an address' error message when a string expansion fails (syntax or
1964 whatever). Otherwise the password may appear in the log. Following change
1965 PH/42 below, there is no longer a chance of it appearing in a bounce
1968 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1970 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1971 of its addresses was ever tried. (Bugzilla bug #2.)
1973 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1974 the result incorrectly in the debug output. (It correctly added a newline
1975 to what was transported.)
1977 TF/01 Added $received_time.
1979 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1980 commented out examples of how to interface to a virus scanner and to
1981 SpamAssassin. Also added commented examples of av_scanner and
1982 spamd_address settings.
1984 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1985 and controls are allowed in which ACLs. There were a couple of minor
1986 errors. Some of the entries in the conditions table (which is a table of
1987 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1988 negation of where the condition IS allowed.
1990 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1992 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1993 header file does not have a version number, so I've had to invent a new
1994 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1995 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1996 radiusclient), but it was contributed by a Radius user.
1998 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1999 files or directories, for OpenSSL.
2001 PH/19 When an Exim process that is running as root has to create an Exim log
2002 file, it does so in a subprocess that runs as exim:exim so as to get the
2003 ownership right at creation (otherwise, other Exim processes might see
2004 the file with the wrong ownership). There was no test for failure of this
2005 fork() call, which would lead to the process getting stuck as it waited
2006 for a non-existent subprocess. Forks do occasionally fail when resources
2007 run out. I reviewed all the other calls to fork(); they all seem to check
2010 PH/20 When checking for unexpected SMTP input at connect time (before writing
2011 the banner), Exim was not dealing correctly with a non-positive return
2012 from the read() function. If the client had disconnected by this time,
2013 the result was a log entry for a synchronization error with an empty
2014 string after "input=" when read() returned zero. If read() returned -1
2015 (an event I could not check), uninitialized data bytes were printed.
2016 There were reports of junk text (parts of files, etc) appearing after
2019 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2021 PH/22 Added support for macro redefinition, and (re)definition in between
2022 driver and ACL definitions.
2024 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2025 forgetting to use the resulting value; it was using the unexpanded value.
2027 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2028 hadn't been configured. The fix is from Juergen Kreileder, who
2029 understands it better than I do:
2031 "Here's what I see happening with three configured cyrus_sasl
2032 authenticators configured (plain, login, cram-md5):
2034 On startup auth_cyrus_sasl_init() gets called for each of these.
2035 This means three calls to sasl_listmech() without a specified mech_list.
2036 => SASL tests which mechs of all available mechs actually work
2037 => three warnings about OTP not working
2038 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2040 With the patch, sasl_listmech() also gets called three times. But now
2041 SASL's mech_list option is set to the server_mech specified in the the
2042 authenticator. Or in other words, the answer from sasl_listmech()
2043 gets limited to just the mech you're testing for (which is different
2045 => the return list contains just 'plain' or 'login', 'cram-md5' or
2046 nothing depending on the value of ob->server_mech.
2048 I've just tested the patch: Authentication still works fine,
2049 unavailable mechs specified in the exim configuration are still
2050 caught, and the auth.log warnings about OTP are gone."
2052 PH/25 When debugging is enabled, the contents of the command line are added
2053 to the debugging output, even when log_selector=+arguments is not
2056 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2057 answer is "GNU", and only if the return is "GNU/something" is the answer
2060 PH/27 $acl_verify_message is now set immediately after the failure of a
2061 verification in an ACL, and so is available in subsequent modifiers. In
2062 particular, the message can be preserved by coding like this:
2064 warn !verify = sender
2065 set acl_m0 = $acl_verify_message
2067 Previously, $acl_verify_message was set only while expanding "message"
2068 and "log_message" when a very denied access.
2070 PH/28 Modified OS/os.c-Linux with
2072 -#ifndef OS_LOAD_AVERAGE
2073 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2075 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2076 nomenclature these days.)
2078 PH/29 Installed patch from the Sieve maintainer that adds the options
2079 sieve_useraddress and sieve_subaddress to the redirect router.
2081 PH/30 In these circumstances:
2082 . Two addresses routed to the same list of hosts;
2083 . First host does not offer TLS;
2084 . First host accepts first address;
2085 . First host gives temporary error to second address;
2086 . Second host offers TLS and a TLS session is established;
2087 . Second host accepts second address.
2088 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2089 and peerdn, if requested) that were in fact used only for the second
2092 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2093 attention to any local part prefix or suffix that was matched by the
2094 router that accepted the address. It now behaves in the same way as it
2095 does for delivery: the affixes are removed from the local part unless
2096 rcpt_include_affixes is set on the transport.
2098 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2099 timeout during the DATA phase of an incoming message.
2101 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2102 applied a patch sanctioned by the Sieve maintainer.
2104 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2105 the uid or gid is negative. A case of a negative gid caused this to be
2106 noticed. The fix allows for either to be negative.
2108 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2109 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2112 PH/36 The change PH/12 above was broken. Fixed it.
2114 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2115 the grounds that routing the same address twice would always produce the
2116 same answer. This might have been true once, but it is certainly no
2117 longer true now. Routing a child address may depend on the previous
2118 routing that produced that child. Some complicated redirection strategies
2119 went wrong when messages had multiple recipients, and made Exim's
2120 behaviour dependent on the order in which the addresses were given.
2122 I have moved the duplicate checking until after the routing is complete.
2123 Exim scans the addresses that are assigned to local and remote
2124 transports, and removes any duplicates. This means that more work will be
2125 done, as duplicates will always all be routed, but duplicates are
2126 presumably rare, so I don't expect this is of any significance.
2128 For deliveries to pipes, files, and autoreplies, the duplicate checking
2129 still happens during the routing process, since they are not going to be
2132 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2133 It corrects a timeout issue with spamd. This is Ian's comment: "The
2134 background is that sometimes spamd either never reads data from a
2135 connection it has accepted, or it never writes response data. The exiscan
2136 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2137 blindly assumes that writes won't block so it may never time out."
2139 PH/39 Allow G after quota size as well as K and M.
2141 PH/40 The value set for $authenticated_id in an authenticator may not contain
2142 binary zeroes or newlines because the value is written to log lines and
2143 to spool files. There was no check on this. Now the value is run through
2144 the string_printing() function so that such characters are converted to
2145 printable escape sequences.
2147 PH/41 $message_linecount is a new variable that contains the total number of
2148 lines in the message. Compare $body_linecount, which is the count for the
2151 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2152 bounce and delay warning messages, except in certain special cases, which
2155 (a) An SMTP error message from a remote host;
2156 (b) A message specified in a :fail: redirection;
2157 (c) A message specified in a "fail" command in a system filter;
2158 (d) A message specified in a FAIL return from the queryprogram router;
2159 (e) A message specified by the cannot_route_message router option.
2161 In these cases only, Exim does include the error details in bounce and
2162 warning messages. There are also a few cases where bland messages such
2163 as "unrouteable address" or "local delivery error" are given.
2165 PH/43 $value is now also set for the "else" part of a ${run expansion.
2167 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2168 being worked on, but at least Exim now implements the latest version to
2171 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2172 process to complete was treated as a delivery failure, a timeout while
2173 writing the message to the pipe was logged, but erroneously treated as a
2174 successful delivery. Such timeouts include transport filter timeouts. For
2175 consistency with the overall process timeout, these timeouts are now
2176 treated as errors, giving rise to delivery failures by default. However,
2177 there is now a new Boolean option for the pipe transport called
2178 timeout_defer, which, if set TRUE, converts the failures into defers for
2179 both kinds of timeout. A transport filter timeout is now identified in
2182 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2183 systems where "make" and "gmake" are different, calling "gmake" at top
2184 level broke things. I've arranged for the value of $(MAKE) to be passed
2185 from the Makefile to this script so that it can call the same version of
2189 A note about Exim versions 4.44 and 4.50
2190 ----------------------------------------
2192 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2193 changes of various kinds. As a consequence, a big documentation update was
2194 needed. This delayed the release for rather longer than seemed good, especially
2195 in the light of a couple of (minor) security issues. Therefore, the changes
2196 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2197 release. So 4.44 and 4.50 are in effect two different branches that both start
2200 I have left the 4.50 change log unchanged; it contains all the changes since
2201 4.43. The change log for 4.44 is below; many of its items are identical to
2202 those for 4.50. This seems to be the most sensible way to preserve the
2203 historical information.
2209 1. Minor wording change to the doc/README.SIEVE file.
2211 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2212 computation of the current number of files was incorrect.
2214 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2215 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2218 4. Give more explanation in the error message when the command for a transport
2219 filter fails to execute.
2221 5. There are several places where Exim runs a non-Exim command in a
2222 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2223 was being done only for the command run by the queryprogram router. It is
2224 now done for all such subprocesses. The other cases are: ${run, transport
2225 filters, and the commands run by the lmtp and pipe transports.
2227 6. Added CONFIGURE_GROUP build-time option.
2229 7. Some older OS have a limit of 256 on the maximum number of file
2230 descriptors. Exim was using setrlimit() to set 1000 as a large value
2231 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2232 systems. I've change it so that if it can't get 1000, it tries for 256.
2234 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2235 was an oversight, and furthermore, ever since the addition of extra
2236 controls (e.g. 4.43/32), the checks on when to allow different forms of
2237 "control" were broken. There should now be diagnostics for all cases when a
2238 control that does not make sense is encountered.
2240 9. Added the /retain_sender option to "control=submission".
2242 10. $recipients is now available in the predata ACL (oversight).
2244 11. Tidy the search cache before the fork to do a delivery from a message
2245 received from the command line. Otherwise the child will trigger a lookup
2246 failure and thereby defer the delivery if it tries to use (for example) a
2247 cached ldap connection that the parent has called unbind on.
2249 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2250 of $address_data from the recipient verification was clobbered by the
2251 sender verification.
2253 13. The value of address_data from a sender verification is now available in
2254 $sender_address_data in subsequent conditions in the ACL statement.
2256 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2258 15. Added a new option "connect=<time>" to callout options, to set a different
2261 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2262 was its contents. (It was OK if the option was not defined at all.)
2264 17. A "Completed" log line is now written for messages that are removed from
2265 the spool by the -Mrm option.
2267 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2268 information about exactly what failed.
2270 19. Added -dd to debug only the daemon process.
2272 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2273 handles timeouts, both on the server side and network timeouts. Renamed the
2274 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2276 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2277 It is now set to "smtps".
2279 22. $host_address is now set to the target address during the checking of
2280 ignore_target_hosts.
2282 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2283 being passed; this would have caused $sender_host_name to have been used if
2284 matching the list had actually called for a host name (not very likely,
2285 since this list is usually IP addresses). A host name is now passed as
2288 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2289 code that specifies a non-restarting handler (typically sigaction() in
2290 modern systems) in an attempt to fix a rare and obscure crash bug.
2292 25. Narrowed the window for a race in the daemon that could cause it to ignore
2293 SIGCHLD signals. This is not a major problem, because they are used only to
2294 wake it up if nothing else does.
2296 26. A malformed maildirsize file could cause Exim to calculate negative values
2297 for the mailbox size or file count. Odd effects could occur as a result.
2298 The maildirsize information is now recalculated if the size or filecount
2301 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2302 support for a long time. Removed HAVE_SYS_VFS_H.
2304 28. Installed the latest version of exipick from John Jetmore.
2306 29. In an address list, if the pattern was not a regular expression, an empty
2307 subject address (from a bounce message) matched only if the pattern was an
2308 empty string. Non-empty patterns were not even tested. This was the wrong
2309 because it is perfectly reasonable to use an empty address as part of a
2310 database query. An empty address is now tested by patterns that are
2311 lookups. However, all the other forms of pattern expect the subject to
2312 contain a local part and a domain, and therefore, for them, an empty
2313 address still always fails if the pattern is not itself empty.
2315 30. Exim went into a mad DNS loop when attempting to do a callout where the
2316 host was specified on an smtp transport, and looking it up yielded more
2317 than one IP address.
2319 31. Re-factored the code for checking spool and log partition space into a
2320 function that finds that data and another that does the check. The former
2321 is then used to implement four new variables: $spool_space, $log_space,
2322 $spool_inodes, and $log_inodes.
2324 32. The RFC2047 encoding function was originally intended for short strings
2325 such as real names; it was not keeping to the 75-character limit for
2326 encoded words that the RFC imposes. It now respects the limit, and
2327 generates multiple encoded words if necessary. To be on the safe side, I
2328 have increased the buffer size for the ${rfc2047: expansion operator from
2331 33. It is now permitted to omit both strings after an "if" condition; if the
2332 condition is true, the result is "true". As before, when the second string
2333 is omitted, a false condition yields an empty string. This makes it less
2334 cumbersome to write custom ACL and router conditions.
2336 34. Failure to deliver a bounce message always caused it to be frozen, even if
2337 there was an errors_to setting on the router. The errors_to setting is now
2340 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2341 canonical form (fully expanded) before being placed in
2342 $sender_host_address.
2344 36. The table in the code that translates DNS record types into text (T_A to
2345 "A" for instance) was missing entries for NS and CNAME. It is just possible
2346 that this could have caused confusion if both these types were looked up
2347 for the same domain, because the text type is used as part of Exim's
2348 per-process caching. But the chance of anyone hitting this buglet seems
2351 37. The dnsdb lookup has been extended in a number of ways.
2353 (1) There is a new type, "zns", which walks up the domain tree until it
2354 finds some nameserver records. It should be used with care.
2356 (2) There is a new type, "mxh", which is like "mx" except that it returns
2357 just the host names, not the priorities.
2359 (3) It is now possible to give a list of domains (or IP addresses) to be
2360 looked up. The behaviour when one of the lookups defers can be
2361 controlled by a keyword.
2363 (4) It is now possible to specify the separator character for use when
2364 multiple records are returned.
2366 38. The dnslists ACL condition has been extended: it is now possible to supply
2367 a list of IP addresses and/or domains to be looked up in a particular DNS
2370 39. Added log_selector=+queue_time_overall.
2372 40. When running the queue in the test harness, wait just a tad after forking a
2373 delivery process, to get repeatability of debugging output.
2375 41. Include certificate and key file names in error message when GnuTLS fails
2376 to set them up, because the GnuTLS error message doesn't include the name
2377 of the failing file when there is a problem reading it.
2379 42. Allow both -bf and -bF in the same test run.
2381 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2383 44. The "Exiscan patch" is now merged into the mainline Exim source.
2385 45. Sometimes the final signoff response after QUIT could fail to get
2386 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2387 < 0 before doing a fflush(). This bug looks as though it goes back to the
2388 introduction of TLS in release 3.20, but "sometimes" must have been rare
2389 because the tests only now provoked it.
2391 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2392 (this can affect the format of dates).
2394 47. exim_tidydb, when checking for the continued existence of a message for
2395 which it has found a message-specific retry record, was not finding
2396 messages that were in split spool directories. Consequently, it was
2397 deleting retry records that should have stayed in existence.
2399 48. Steve fixed some bugs in eximstats.
2401 49. The SPA authentication driver was not abandoning authentication and moving
2402 on to the next authenticator when an expansion was forced to fail,
2403 contradicting the general specification for all authenticators. Instead it
2404 was generating a temporary error. It now behaves as specified.
2406 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2407 (the order specifies the preference for clients). The order is now AES256,
2408 AES128, 3DES, ARCFOUR128.
2410 51. Small patch to Sieve code - explicitly set From: when generating an
2413 52. Exim crashed if a remote delivery caused a very long error message to be
2414 recorded - for instance if somebody sent an entire SpamAssassin report back
2415 as a large number of 550 error lines. This bug was coincidentally fixed by
2416 increasing the size of one of Exim's internal buffers (big_buffer) that
2417 happened as part of the Exiscan merge. However, to be on the safe side, I
2418 have made the code more robust (and fixed the comments that describe what
2421 53. Now that there can be additional text after "Completed" in log lines (if
2422 the queue_time_overall log selector is set), a one-byte patch to exigrep
2423 was needed to allow it to recognize "Completed" as not the last thing in
2426 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2427 patch that reportedly fixes this has been added. I am not expert enough to
2428 create a test for it. This is what the patch creator wrote:
2430 "I found a little strange behaviour of ldap code when working with
2431 Windows 2003 AD Domain, where users was placed in more than one
2432 Organization Units. When I tried to give exim partial DN, the exit code
2433 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2434 But simultaneously result of request was absolutely normal ldap result,
2435 so I produce this patch..."
2437 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2438 REFERENCE, so I have modified the code to exclude the patch when that macro
2441 55. Some experimental protocols are using DNS PTR records for new purposes. The
2442 keys for these records are domain names, not reversed IP addresses. The
2443 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2444 leaves it alone. Component reversal etc. now happens only for IP addresses.
2447 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2449 57. Double the size of the debug message buffer (to 2048) so that more of very
2450 long debug lines gets shown.
2452 58. The exicyclog utility now does better if the number of log files to keep
2453 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2455 59. Two changes related to the smtp_active_hostname option:
2457 (1) $smtp_active_hostname is now available as a variable.
2458 (2) The default for smtp_banner uses $smtp_active_hostname instead
2459 of $primary_hostname.
2461 60. The host_aton() function is supposed to be passed a string that is known
2462 to be a valid IP address. However, in the case of IPv6 addresses, it was
2463 not checking this. This is a hostage to fortune. Exim now panics and dies
2464 if the condition is not met. A case was found where this could be provoked
2465 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2466 components; fortuitously, this particular loophole had already been fixed
2467 by change 4.50/55 above.
2469 If there are any other similar loopholes, the new check in host_aton()
2470 itself should stop them being exploited. The report I received stated that
2471 data on the command line could provoke the exploit when Exim was running as
2472 exim, but did not say which command line option was involved. All I could
2473 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2474 running as the user.
2477 61. There was a buffer overflow vulnerability in the SPA authentication code
2478 (which came originally from the Samba project). I have added a test to the
2479 spa_base64_to_bits() function which I hope fixes it.
2482 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2483 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2485 63. The daemon start-up calls getloadavg() while still root for those OS that
2486 need the first call to be done as root, but it missed one case: when
2487 deliver_queue_load_max is set with deliver_drop_privilege. This is
2488 necessary for the benefit of the queue runner, because there is no re-exec
2489 when deliver_drop_privilege is set.
2491 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2492 This has been fixed.
2494 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2495 was in use, was not putting the data itself into the right store pool;
2496 consequently, it could be overwritten for a subsequent message in the same
2497 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2500 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2502 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2503 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2504 the function were treating the return as a boolean value, which happened to
2505 work because 0=false and not-0=true, but is not correct code.
2507 68. The host_aton() function was not handling scoped IPv6 addresses (those
2508 with, for example, "%eth0" on the end) correctly.
2510 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2511 negated items (that is, ~something) in unsigned ints. Some compilers
2512 apparently mutter when there is no cast.
2514 70. If an address verification called from an ACL failed, and did not produce a
2515 user-specific message (i.e. there was only a "system" message), nothing was
2516 put in $acl_verify_message. In this situation, it now puts the system
2519 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2520 session; change 4.31/43 added the unwanted input to the log line - except
2521 that it did not do this in the start of session case. It now does.
2523 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2524 This is wrong and can cause the other end to generate a synchronization
2525 error if it is another Exim or anything else that does the synchronization
2526 check. A QUIT command is no longer sent after a timeout.
2528 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2529 during host lookups.
2531 74. The defer_ok option of callout verification was not working if it was used
2532 when verifying addresses in header lines, that is, for this case:
2534 verify = header_sender/callout=defer_ok
2536 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2537 those file descriptors could be used for SMTP connections. If anything
2538 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2539 could be sent to the SMTP client, causing chaos. The daemon now opens
2540 stdin, stdout, and stderr to /dev/null when it puts itself into the
2543 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2544 log by default. The user can override this with suitable Perl magic.
2546 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2547 the log message when discard triggers, was not working for the DATA ACL or
2548 for the non-SMTP ACL.
2550 78. Error message wording change in sieve.c.
2552 79. If smtp_accept_max_per_host was set, the number of connections could be
2553 restricted to fewer than expected, because the daemon was trying to set up
2554 a new connection before checking whether the processes handling previous
2555 connections had finished. The check for completed processes is now done
2556 earlier. On busy systems, this bug wouldn't be noticed because something
2557 else would have woken the daemon, and it would have reaped the completed
2560 80. If a message was submitted locally by a user whose login name contained one
2561 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2562 It caused a spool format error. I have fixed the spool reading code. A
2563 related problem was that the "from" clause in the Received: line became
2564 illegal because of the space(s). It is now covered by ${quote_local_part.
2566 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2569 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2571 83. Patch from Sieve maintainer to fix unterminated string problem in
2572 "vacation" handling.
2574 84. Some minor changes to the Linux configuration files to help with other
2575 OS variants using glibc.
2577 85. One more patch for Sieve to update vacation handling to latest spec.
2580 ----------------------------------------------------
2581 See the note above about the 4.44 and 4.50 releases.
2582 ----------------------------------------------------
2588 1. Change 4.43/35 introduced a bug that caused file counts to be
2589 incorrectly computed when quota_filecount was set in an appendfile
2592 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2593 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2596 3. Give more explanation in the error message when the command for a transport
2597 filter fails to execute.
2599 4. There are several places where Exim runs a non-Exim command in a
2600 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2601 was being done only for the command run by the queryprogram router. It is
2602 now done for all such subprocesses. The other cases are: ${run, transport
2603 filters, and the commands run by the lmtp and pipe transports.
2605 5. Some older OS have a limit of 256 on the maximum number of file
2606 descriptors. Exim was using setrlimit() to set 1000 as a large value
2607 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2608 systems. I've change it so that if it can't get 1000, it tries for 256.
2610 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2611 was an oversight, and furthermore, ever since the addition of extra
2612 controls (e.g. 4.43/32), the checks on when to allow different forms of
2613 "control" were broken. There should now be diagnostics for all cases when a
2614 control that does not make sense is encountered.
2616 7. $recipients is now available in the predata ACL (oversight).
2618 8. Tidy the search cache before the fork to do a delivery from a message
2619 received from the command line. Otherwise the child will trigger a lookup
2620 failure and thereby defer the delivery if it tries to use (for example) a
2621 cached ldap connection that the parent has called unbind on.
2623 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2624 of $address_data from the recipient verification was clobbered by the
2625 sender verification.
2627 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2628 was its contents. (It was OK if the option was not defined at all.)
2630 11. A "Completed" log line is now written for messages that are removed from
2631 the spool by the -Mrm option.
2633 12. $host_address is now set to the target address during the checking of
2634 ignore_target_hosts.
2636 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2637 being passed; this would have caused $sender_host_name to have been used if
2638 matching the list had actually called for a host name (not very likely,
2639 since this list is usually IP addresses). A host name is now passed as
2642 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2643 code that specifies a non-restarting handler (typically sigaction() in
2644 modern systems) in an attempt to fix a rare and obscure crash bug.
2646 15. Narrowed the window for a race in the daemon that could cause it to ignore
2647 SIGCHLD signals. This is not a major problem, because they are used only to
2648 wake it up if nothing else does.
2650 16. A malformed maildirsize file could cause Exim to calculate negative values
2651 for the mailbox size or file count. Odd effects could occur as a result.
2652 The maildirsize information is now recalculated if the size or filecount
2655 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2656 support for a long time. Removed HAVE_SYS_VFS_H.
2658 18. Updated exipick to current release from John Jetmore.
2660 19. Allow an empty sender to be matched against a lookup in an address list.
2661 Previously the only cases considered were a regular expression, or an
2664 20. Exim went into a mad DNS lookup loop when doing a callout where the
2665 host was specified on the transport, if the DNS lookup yielded more than
2668 21. The RFC2047 encoding function was originally intended for short strings
2669 such as real names; it was not keeping to the 75-character limit for
2670 encoded words that the RFC imposes. It now respects the limit, and
2671 generates multiple encoded words if necessary. To be on the safe side, I
2672 have increased the buffer size for the ${rfc2047: expansion operator from
2675 22. Failure to deliver a bounce message always caused it to be frozen, even if
2676 there was an errors_to setting on the router. The errors_to setting is now
2679 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2680 canonical form (fully expanded) before being placed in
2681 $sender_host_address.
2683 24. Updated eximstats to version 1.33
2685 25. Include certificate and key file names in error message when GnuTLS fails
2686 to set them up, because the GnuTLS error message doesn't include the name
2687 of the failing file when there is a problem reading it.
2689 26. Expand error message when OpenSSL has problems setting up cert/key files.
2692 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2693 (this can affect the format of dates).
2695 28. exim_tidydb, when checking for the continued existence of a message for
2696 which it has found a message-specific retry record, was not finding
2697 messages that were in split spool directories. Consequently, it was
2698 deleting retry records that should have stayed in existence.
2700 29. eximstats updated to version 1.35
2701 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2702 1.35 - bugfix such that pie charts by volume are generated correctly
2704 30. The SPA authentication driver was not abandoning authentication and moving
2705 on to the next authenticator when an expansion was forced to fail,
2706 contradicting the general specification for all authenticators. Instead it
2707 was generating a temporary error. It now behaves as specified.
2709 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2710 (the order specifies the preference for clients). The order is now AES256,
2711 AES128, 3DES, ARCFOUR128.
2713 31. Small patch to Sieve code - explicitly set From: when generating an
2716 32. Exim crashed if a remote delivery caused a very long error message to be
2717 recorded - for instance if somebody sent an entire SpamAssassin report back
2718 as a large number of 550 error lines. This bug was coincidentally fixed by
2719 increasing the size of one of Exim's internal buffers (big_buffer) that
2720 happened as part of the Exiscan merge. However, to be on the safe side, I
2721 have made the code more robust (and fixed the comments that describe what
2724 33. Some experimental protocols are using DNS PTR records for new purposes. The
2725 keys for these records are domain names, not reversed IP addresses. The
2726 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2727 leaves it alone. Component reversal etc. now happens only for IP addresses.
2730 34. The host_aton() function is supposed to be passed a string that is known
2731 to be a valid IP address. However, in the case of IPv6 addresses, it was
2732 not checking this. This is a hostage to fortune. Exim now panics and dies
2733 if the condition is not met. A case was found where this could be provoked
2734 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2735 components; fortuitously, this particular loophole had already been fixed
2736 by change 4.50/55 or 4.44/33 above.
2738 If there are any other similar loopholes, the new check in host_aton()
2739 itself should stop them being exploited. The report I received stated that
2740 data on the command line could provoke the exploit when Exim was running as
2741 exim, but did not say which command line option was involved. All I could
2742 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2743 running as the user.
2746 35. There was a buffer overflow vulnerability in the SPA authentication code
2747 (which came originally from the Samba project). I have added a test to the
2748 spa_base64_to_bits() function which I hope fixes it.
2751 36. The daemon start-up calls getloadavg() while still root for those OS that
2752 need the first call to be done as root, but it missed one case: when
2753 deliver_queue_load_max is set with deliver_drop_privilege. This is
2754 necessary for the benefit of the queue runner, because there is no re-exec
2755 when deliver_drop_privilege is set.
2757 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2758 was in use, was not putting the data itself into the right store pool;
2759 consequently, it could be overwritten for a subsequent message in the same
2760 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2763 38. Sometimes the final signoff response after QUIT could fail to get
2764 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2765 < 0 before doing a fflush(). This bug looks as though it goes back to the
2766 introduction of TLS in release 3.20, but "sometimes" must have been rare
2767 because the tests only now provoked it.
2773 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2774 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2775 Now it returns an integer. A number of calls were still expecting a T/F
2776 return. Fortuitously, in all cases, the tests worked in OK situations,
2777 which is the norm. However, things would have gone wrong on any write
2778 failures on the smtp file descriptor. This function is used when sending
2779 messages over SMTP and also when doing verify callouts.
2781 2. When Exim is called to do synchronous delivery of a locally submitted
2782 message (the -odf or -odi options), it no longer closes stderr before doing
2785 3. Implemented the mua_wrapper option.
2787 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2789 5. Implemented the functions header_remove(), header_testname(),
2790 header_add_at_position(), and receive_remove_recipient(), and exported them
2793 6. If an ACL "warn" statement specified the addition of headers, Exim already
2794 inserted X-ACL-Warn: at the start if there was no header name. However, it
2795 was not making this test for the second and subsequent header lines if
2796 there were newlines in the string. This meant that an invalid header could
2797 be inserted if Exim was badly configured.
2799 7. Allow an ACL "warn" statement to add header lines at the start or after all
2800 the Received: headers, as well as at the end.
2802 8. Added the rcpt_4xx retry error code.
2804 9. Added postmaster_mailfrom=xxx to callout verification option.
2806 10. Added mailfrom=xxxx to the callout verification option, for verify=
2809 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2810 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2812 12. Inserted some casts to stop certain compilers warning when using pointer
2813 differences as field lengths or precisions in printf-type calls (mostly
2814 affecting debugging statements).
2816 13. Added optional readline() support for -be (dynamically loaded).
2818 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2819 same clock tick as a message's arrival, so that its received time was the
2820 same as the "first fail" time on the retry record, and that message
2821 remained on the queue past the ultimate address timeout, every queue runner
2822 would try a delivery (because it was past the ultimate address timeout) but
2823 after another failure, the ultimate address timeout, which should have then
2824 bounced the address, did not kick in. This was a "< instead of <=" error;
2825 in most cases the first failure would have been in the next clock tick
2826 after the received time, and all would be well.
2828 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2829 being recognized when the domain list was tested by the match_domain
2830 condition in an expansion string.
2832 16. Added the ${str2b64: operator.
2834 17. Exim was always calling setrlimit() to set a large limit for the number of
2835 processes, without checking whether the existing limit was already
2836 adequate. (It did check for the limit on file descriptors.) Furthermore,
2837 errors from getrlimit() and setrlimit() were being ignored. Now they are
2838 logged to the main and panic logs, but Exim does carry on, to try to do its
2839 job under whatever limits there are.
2841 18. Imported PCRE 5.0.
2843 19. Trivial typo in log message " temporarily refused connection" (the leading
2846 20. If the log selector return_path_on_delivery was set and an address was
2847 redirected to /dev/null, the delivery process crashed because it assumed
2848 that a return path would always be set for a "successful" delivery. In this
2849 case, the whole delivery is bypassed as an optimization, and therefore no
2852 21. Internal re-arrangement: the function for sending a challenge and reading
2853 a response while authentication was assuming a zero-terminated challenge
2854 string. It's now changed to take a pointer and a length, to allow for
2855 binary data in such strings.
2857 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2859 23. Exim was not respecting finduser_retries when seeking the login of the
2860 uid under which it was called; it was always trying 10 times. (The default
2861 setting of finduser_retries is zero.) Also, it was sleeping after the final
2862 failure, which is pointless.
2864 24. Implemented tls_on_connect_ports.
2866 25. Implemented acl_smtp_predata.
2868 26. If the domain in control=submission is set empty, Exim assumes that the
2869 authenticated id is a complete email address when it generates From: or
2870 Sender: header lines.
2872 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2873 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2874 chown and chgrp in /bin and hostname in /usr/bin.
2876 28. Exim was keeping the "process log" file open after each use, just as it
2877 does for the main log. This opens the possibility of it remaining open for
2878 long periods when the USR1 signal hits a daemon. Occasional processlog
2879 errors were reported, that could have been caused by this. Anyway, it seems
2880 much more sensible not to leave this file open at all, so that is what now
2883 29. The long-running daemon process does not normally write to the log once it
2884 has entered its main loop, and it closes the log before doing so. This is
2885 so that log files can straightforwardly be renamed and moved. However,
2886 there are a couple of unusual error situations where the daemon does write
2887 log entries, and I had neglected to close the log afterwards.
2889 30. The text of an SMTP error response that was received during a remote
2890 delivery was being truncated at 512 bytes. This is too short for some of
2891 the long messages that one sometimes sees. I've increased the limit to
2894 31. It is now possible to make retry rules that apply only when a message has a
2895 specific sender, in particular, an empty sender.
2897 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2898 it possible to be selective about when SMTP synchronization is enforced.
2900 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2902 32. Implemented hosts_connection_nolog.
2904 33. Added an ACL for QUIT.
2906 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2909 35. Added mailbox_size and mailbox_filecount to appendfile.
2911 36. Added control = no_multiline_responses to ACLs.
2913 37. There was a bug in the logic of the code that waits for the clock to tick
2914 in the case where the clock went backwards by a substantial amount such
2915 that the microsecond fraction of "now" was more than the microsecond
2916 fraction of "then" (but the whole seconds number was less).
2918 38. Added support for the libradius Radius client library this is found on
2919 FreeBSD (previously only the radiusclient library was supported).
2925 1. When certain lookups returned multiple values in the form name=value, the
2926 quoting of the values was not always being done properly. Specifically:
2927 (a) If the value started with a double quote, but contained no whitespace,
2929 (b) If the value contained whitespace other than a space character (i.e.
2930 tabs or newlines or carriage returns) it was not quoted.
2931 This fix has been applied to the mysql and pgsql lookups by writing a
2932 separate quoting function and calling it from the lookup code. The fix
2933 should probably also be applied to nisplus, ibase and oracle lookups, but
2934 since I cannot test any of those, I have not disturbed their existing code.
2936 2. A hit in the callout cache for a specific address caused a log line with no
2937 reason for rejecting RCPT. Now it says "Previous (cached) callout
2938 verification failure".
2940 3. There was an off-by-one bug in the queryprogram router. An over-long
2941 return line was truncated at 256 instead of 255 characters, thereby
2942 overflowing its buffer with the terminating zero. As well as fixing this, I
2943 have increased the buffer size to 1024 (and made a note to document this).
2945 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2946 when Exim is waiting for an SMTP response from a remote server, Exim
2947 restarts its select() call on the socket, thereby resetting its timeout.
2948 This is not a problem when such interrupts are rare. Somebody set up a cron
2949 job to run exiwhat every 2 minutes, which is less than the normal select()
2950 timeout (5 or 10 minutes). This meant that the select() timeout never
2951 kicked in because it was always reset. I have fixed this by comparing the
2952 time when an interrupt arrives with the time at the start of the first call
2953 to select(). If more time than the timeout has elapsed, the interrupt is
2954 treated as a timeout.
2956 5. Some internal re-factoring in preparation for the addition of Sieve
2957 extensions (by MH). In particular, the "personal" test is moved to a
2958 separate function, and given an option for scanning Cc: and Bcc: (which is
2959 not set for Exim filters).
2961 6. When Exim created an email address using the login of the caller as the
2962 local part (e.g. when creating a From: or Sender: header line), it was not
2963 quoting the local part when it contained special characters such as @.
2965 7. Installed new OpenBSD configuration files.
2967 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2968 try to make them clearer.
2970 9. Callout options, other than the timeout value, were being ignored when
2971 verifying sender addresses in header lines. For example, when using
2973 verify = header_sender/callout=no_cache
2975 the cache was (incorrectly) being used.
2977 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2978 only the Cygwin environment.
2980 11. When return_path_on_delivery was set as a log selector, if different remote
2981 addresses in the same message used different return paths and parallel
2982 remote delivery occurred, the wrong values would sometimes be logged.
2983 (Whenever a remote delivery process finished, the return path value from
2984 the most recently started remote delivery process was logged.)
2986 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2987 lines when AUTH and/or TLS are in use. This is the "received protocol"
2988 field. Exim used to use "asmtp" for authenticated SMTP, without any
2989 indication (in the protocol name) for TLS use. Now it follows the RFC and
2990 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2991 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2992 names appear in log lines as well as in Received: header lines.
2994 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2995 extensions, and comparison tests, and to fix some bugs.
2997 14. Changes to the "personal" filter test:
2999 (1) The test was buggy in that it was just doing the equivalent of
3000 "contains" tests on header lines. For example, if a user's address was
3001 anne@some.where, the "personal" test would incorrectly be true for
3003 To: susanne@some.where
3005 This test is now done by extracting each address from the header in turn,
3006 and checking the entire address. Other tests that are part of "personal"
3007 are now done using regular expressions (for example, to check local parts
3008 of addresses in From: header lines).
3010 (2) The list of non-personal local parts in From: addresses has been
3011 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3012 taken from the Sieve specification recommendations.
3014 (3) If the message contains any header line starting with "List-" it is
3015 treated as non-personal.
3017 (4) The test for "circular" in the Subject: header line has been removed
3018 because it now seems ill-conceived.
3020 15. Minor typos in src/EDITME comments corrected.
3022 16. Installed latest exipick from John Jetmore.
3024 17. If headers_add on a router specified a text string that was too long for
3025 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3026 of string_sprintf() is now avoided.
3028 18. $message_body_size was not set (it was always zero) when running the DATA
3029 ACL and the local_scan() function.
3031 19. For the "mail" command in an Exim filter, no default was being set for
3032 the once_repeat time, causing a random time value to be used if "once" was
3033 specified. (If the value happened to be <= 0, no repeat happened.) The
3034 default is now 0s, meaning "never repeat". The "vacation" command was OK
3035 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3036 (I found it when inspecting the code).
3038 20. There is now an overall timeout for performing a callout verification. It
3039 defaults to 4 times the callout timeout, which applies to individual SMTP
3040 commands during the callout. The overall timeout applies when there is more
3041 than one host that can be tried. The timeout is checked before trying the
3042 next host. This prevents very long delays if there are a large number of
3043 hosts and all are timing out (e.g. when the network connections are timing
3044 out). The value of the overall timeout can be changed by specifying an
3045 additional sub-option for "callout", called "maxwait". For example:
3047 verify = sender/callout=5s,maxwait=20s
3049 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3050 to the end before writing, but this should make it even safer).
3052 22. Exim was forgetting that it had advertised PIPELINING for the second and
3053 subsequent messages on an SMTP connection. It was also not resetting its
3054 memory on STARTTLS and an internal HELO.
3056 23. When Exim logs an SMTP synchronization error within a session, it now
3057 records whether PIPELINING has been advertised or not.
3059 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3060 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3061 rather than long int.
3063 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3065 26. Added the never_mail option to autoreply.
3071 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3072 crash if the getsockname() call failed; this can happen if a connection is
3073 closed very soon after it is established. The problem was simply in the
3074 order in which certain operations were done, causing Exim to try to write
3075 to the SMTP stream before it had set up the file descriptor. The bug has
3076 been fixed by making things happen in the correct order.
3082 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3083 before the connection was closed, thus losing the rejection response.
3085 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3086 some early Solaris releases, but causes trouble in current releases where
3087 socklen_t is defined.
3089 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3092 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3095 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3096 that releasing the top of it at the end releases what was used for sub-
3097 expansions (unless the block got too big). However, discard this block if
3098 the first thing is a variable or header, so that we can use its block when
3099 it is dynamic (useful for very large $message_headers, for example).
3101 6. Lookups now cache *every* query, not just the most recent. A new, separate
3102 store pool is used for this. It can be recovered when all lookup caches are
3103 flushed. Lookups now release memory at the end of their result strings.
3104 This has involved some general refactoring of the lookup sources.
3106 7. Some code has been added to the store_xxx() functions to reduce the amount
3107 of flapping under certain conditions.
3109 8. log_incoming_interface used to affect only the <= reception log lines. Now
3110 it causes the local interface and port to be added to several more SMTP log
3111 lines, for example "SMTP connection from", and rejection lines.
3113 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3115 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3117 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3118 could be overwritten at the end of the current message (or the start of a
3119 new message if it was set in a HELO ACL). The value is now preserved for
3120 the duration of the SMTP connection.
3122 12. If a transport had a headers_rewrite setting, and a matching header line
3123 contained an unqualified address, that address was qualified, even if it
3124 did not match any rewriting rules. The underlying bug was that the values
3125 of the flags that permit the existence of unqualified sender and recipient
3126 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3127 non-local messages, and by -bnq for local messages) were not being
3128 preserved with the message after it was received.
3130 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3131 "next input=" as part of the text comprising the host identity instead of
3132 the correct text. The code was using the same buffer for two different
3133 strings. However, depending on which order the printing function evaluated
3134 its arguments, the bug did not always show up. Under Linux, for example, my
3135 test suite worked just fine.
3137 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3138 doesn't work with some older versions of Perl. It has been changed to "my",
3139 which in any case is probably the better facility to use.
3141 15. A really picky compiler found some instances of statements for creating
3142 error messages that either had too many or two few arguments for the format
3145 16. The size of the buffer for calls to the DNS resolver has been increased
3146 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3147 for addresses that have a lot of PTR records. This alleviates a problem; it
3148 does not fully solve it.
3150 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3151 buffer now truncates the list and logs the incident, which is the same
3152 action as happens when Exim is looking up a host name and its aliases.
3153 Previously in this situation something unpredictable would happen;
3154 sometimes it was "internal error: store_reset failed".
3156 18. If a server dropped the connection unexpectedly when an Exim client was
3157 using GnuTLS and trying to read a response, the client delivery process
3158 crashed while trying to generate an error log message.
3160 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3161 string, for example:
3163 warn message = H1: something\nH2: something
3165 the text was added as a single header line from Exim's point of view
3166 though it ended up OK in the delivered message. However, searching for the
3167 second and subsequent header lines using $h_h2: did not work. This has been
3168 fixed. Similarly, if a system filter added multiple headers in this way,
3169 the routers could not see them.
3171 20. Expanded the error message when iplsearch is called with an invalid key to
3172 suggest using net-iplsearch in a host list.
3174 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3175 ACLs are no longer actually imposed (and a message to that effect is
3178 22. If a "gecos" field in a passwd entry contained escaped characters, in
3179 particular, if it contained a \" sequence, Exim got it wrong when building
3180 a From: or a Sender: header from that name. A second bug also caused
3181 incorrect handling when an unquoted " was present following a character
3182 that needed quoting.
3184 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3185 was not being matched caselessly.
3187 24. Arranged for all hyphens in the exim.8 source to be escaped with
3190 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3191 properly. Recipient callout cache records were still being keyed to include
3192 the sender, even when use_sender was set false. This led to far more
3193 callouts that were necessary. The sender is no longer included in the key
3194 when use_sender is false.
3196 26. Added "control = submission" modifier to ACLs.
3198 27. Added the ${base62d: operator to decode base 62 numbers.
3200 28. dnsdb lookups can now access SRV records.
3202 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3203 the configuration file.
3205 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3206 (-v) mode. This makes the output for a verbose queue run more intelligible.
3208 31. Added a use_postmaster feature to recipient callouts.
3210 32. Added the $body_zerocount variable, containing the number of binary zero
3211 bytes in the message body.
3213 33. The time of last modification of the "new" subdirectory is now used as the
3214 "mailbox time last read" when there is a quota error for a maildir
3217 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3219 35. Added +ignore_unknown as a special item in host lists.
3221 36. Code for decoding IPv6 addresses in host lists is now included, even if
3222 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3223 address was recognized as an IP address, but was then not correctly decoded
3224 into binary, causing unexpected and incorrect effects when compared with
3231 1. Very minor rewording of debugging text in manualroute to say "list of
3232 hosts" instead of "hostlist".
3234 2. If verify=header_syntax was set, and a header line with an unqualified
3235 address (no domain) and a large number of spaces between the end of the
3236 name and the colon was received, the reception process suffered a buffer
3237 overflow, and (when I tested it) crashed. This was caused by some obsolete
3238 code that should have been removed. The fix is to remove it!
3240 3. When running in the test harness, delay a bit after writing a bounce
3241 message to get a bit more predictability in the log output.
3243 4. Added a call to search_tidyup() just before forking a reception process. In
3244 theory, someone could use a lookup in the expansion of smtp_accept_max_
3245 per_host which, without the tidyup, could leave open a database connection.
3247 5. Added the variables $recipient_data and $sender_data which get set from a
3248 lookup success in an ACL "recipients" or "senders" condition, or a router
3249 "senders" option, similar to $domain_data and $local_part_data.
3251 6. Moved the writing of debug_print from before to after the "senders" test
3254 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3255 problems for message scanning, either using a data ACL, or using
3256 local_scan() because the Received: header was not generated till after they
3257 were called (in order to set the time as the time of reception completion).
3258 I have revised the way this works. The header is now generated after the
3259 body is received, but before the ACL or local_scan() are called. After they
3260 are run, the timestamp in the header is updated.
3266 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3267 before starting a queue runner without re-exec. This happened only when
3268 deliver_drop_privilege was set or when the Exim user was set to root. The
3269 effect of the bug was that timeouts during subsequent deliveries caused
3270 crashes instead of being properly handled. The handler is now left at its
3271 default (and expected) setting.
3273 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3274 message, again when deliver_drop_privilege is set or Exim is run as root.
3275 The bug described in (1) was not present in this case, but the tidying up
3276 of the other signals was missing. I have made the two cases consistent.
3278 3. The ignore_target_hosts setting on a manualroute router was being ignored
3279 for hosts that were looked up using the /MX notation.
3281 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3284 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3285 operated on the sender address. After changing the $sender_address to <>
3286 for the sender address verify, Exim was re-instated it as the original
3287 (before rewriting) address, but remembering that it had rewritten it, so it
3288 wasn't rewriting it again. This bug also had the effect of breaking the
3289 sender address verification caching when the sender address was rewritten.
3291 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3292 This has been changed so that if the ip literal address matches
3293 ignore_target_hosts, the router declines.
3295 7. Added expansion conditions match_domain, match_address, and match_local_
3296 part (NOT match_host).
3298 8. The placeholder for the Received: header didn't have a length field set.
3300 9. Added code to Exim itself and to exim_lock to test for a specific race
3301 condition that could lead to file corruption when using MBX delivery. The
3302 issue is with the lockfile that is created in /tmp. If this file is removed
3303 after a process has opened it but before that process has acquired a lock,
3304 there is the potential for a second process to recreate the file and also
3305 acquire a lock. This could lead to two Exim processes writing to the file
3306 at the same time. The added code performs the same test as UW imapd; it
3307 checks after acquiring the lock that its file descriptor still refers to
3308 the same named file.
3310 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3311 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3314 11. Added the smtp_active_hostname option. If used, this will typically be made
3315 to depend on the incoming interface address. Because $interface_address is
3316 not set up until the daemon has forked a reception process, error responses
3317 that can happen earlier (such as "too many connections") no longer contain
3320 12. If an expansion in a condition on a "warn" statement fails because a lookup
3321 defers, the "warn" statement is abandoned, and the next ACL statement is
3322 processed. Previously this caused the whole ACL to be aborted.
3324 13. Added the iplsearch lookup type.
3326 14. Added ident_timeout as a log selector.
3328 15. Added tls_certificate_verified as a log selector.
3330 16. Added a global option tls_require_ciphers (compare the smtp transport
3331 option of the same name). This controls incoming TLS connections.
3333 17. I finally figured out how to make tls_require_ciphers do a similar thing
3334 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3335 before starting the TLS session.
3337 18. Tabs are now shown as \t in -bP output.
3339 19. If the log selector return_path_on_delivery was set, Exim crashed when
3340 bouncing a message because it had too many Received: header lines.
3342 20. If two routers both had headers_remove settings, and the first one included
3343 a superfluous trailing colon, the final name in the first list and the
3344 first name in the second list were incorrectly joined into one item (with a
3345 colon in the middle).
3351 1. Added -C and -D options to the exinext utility, mainly to make it easier
3352 to include in the automated testing, but these could be helpful when
3353 multiple configurations are in use.
3355 2. The exinext utility was not formatting the output nicely when there was
3356 an alternate port involved in the retry record key, nor when there was a
3357 message id as well (for retries that were specific to a specific message
3358 and a specific host). It was also confused by IPv6 addresses, because of
3359 the additional colons they contain. I have fixed the IPv4 problem, and
3360 patched it up to do a reasonable job for IPv6.
3362 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3363 delivery, the log line now contains "pipelined" if PIPELINING was used.
3365 4. An SMTP transport process used to panic and die if the bind() call to set
3366 an explicit outgoing interface failed. This has been changed; it is now
3367 treated in the same way as a connect() failure.
3369 5. A reference to $sender_host_name in the part of a conditional expansion
3370 that was being skipped was still causing a DNS lookup. This no longer
3373 6. The def: expansion condition was not recognizing references to header lines
3374 that used bh_ and bheader_.
3376 7. Added the _cache feature to named lists.
3378 8. The code for checking quota_filecount in the appendfile transport was
3379 allowing one more file than it should have been.
3381 9. For compatibility with Sendmail, the command line option
3389 and sets the incoming protocol and host name (for trusted callers). The
3390 host name and its colon can be omitted when only the protocol is to be set.
3391 Note the Exim already has two private options, -pd and -ps, that refer to
3392 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3393 "s", but I don't think that's a major issue.
3395 10. A number of refactoring changes to the code, none of which should affect
3398 (a) The number of logging options was getting close to filling up the
3399 32-bit word that was used as a bit map. I have split them into two classes:
3400 those that are passed in the argument to log_write(), and those that are
3401 only ever tested independently outside of that function. These are now in
3402 separate 32-bit words, so there is plenty of room for expansion again.
3403 There is no change in the user interface or the logging behaviour.
3405 (b) When building, for example, log lines, the code previously used a
3406 macro that called string_cat() twice, in order to add two strings. This is
3407 not really sufficiently general. Furthermore, there was one instance where
3408 it was actually wrong because one of the argument was used twice, and in
3409 one call a function was used. (As it happened, calling the function twice
3410 did not affect the overall behaviour.) The macro has been replaced by a
3411 function that can join an arbitrary number of extra strings onto a growing
3414 (c) The code for expansion conditions now uses a table and a binary chop
3415 instead of a serial search (which was left over from when there were very
3416 few conditions). Also, it now recognizes conditions like "pam" even when
3417 the relevant support is not compiled in: a suitably worded error message is
3418 given if an attempt is made to use such a condition.
3420 11. Added ${time_interval:xxxxx}.
3422 12. A bug was causing one of the ddress fields not to be passed back correctly
3423 from remote delivery subprocesses. The field in question was not being
3424 subsequently used, so this caused to problems in practice.
3426 13. Added new log selectors queue_time and deliver_time.
3428 14. Might have fixed a bug in maildirsizefile handling that threw up
3429 "unexpected character" debug warnings, and recalculated the data
3430 unnecessarily. In any case, I expanded the warning message to give more
3433 15. Added the message "Restricted characters in address" to the statements in
3434 the default ACL that block characters like @ and % in local parts.
3436 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3437 Three changes have been made:
3439 (a) There was a serious bug; a negative response to MAIL caused the whole
3440 recipient domain to be cached as invalid, thereby blocking all messages
3441 to all local parts at the same domain, from all senders. This bug has
3442 been fixed. The domain is no longer cached after a negative response to
3443 MAIL if the sender used is not empty.
3445 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3448 (c) A new callout option, "use_sender" has been added for people who want
3449 the modified behaviour.
3455 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3458 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3459 indeed breaks things for older releases.
3461 3. Added additional logging to the case where there is a problem reading data
3462 from a filter that is running in a subprocess using a pipe, in order to
3463 try to track down a specific problem.
3465 4. Testing facility fudge: when running in the test harness and attempting
3466 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3467 getting "No route to host". Convert this to a timeout.
3469 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3472 6. Some OS don't have socklen_t but use size_t instead. This affects the
3473 fifth argument of getsockopt() amongst other things. This is now
3474 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3475 can be set for individual OS. I have set it for SunOS5, OSF1, and
3476 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3477 some earlier ones do not.
3479 7. Change 4.30/15 was not doing the test caselessly.
3481 8. The standard form for an IPv6 address literal was being rejected by address
3482 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3483 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3484 this, as well as the form without the "IPv6" on the front (but only when
3485 address literals are enabled, of course).
3487 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3489 10. Exim crashed if a message with an empty sender address specified by -f
3490 encountered a router with an errors_to setting. This could be provoked only
3491 by a command such as
3495 where an empty string was supplied; "<>" did not hit this bug.
3497 11. Installed PCRE release 4.5.
3499 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3500 remained set. It is now erased.
3502 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3503 times from message ids (which are base 36 rather than the normal 62).
3505 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3506 were being counted as actual protocol errors, and logged if the log
3507 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3508 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3509 and DATA following a set of rejected RCPTs do not count as protocol errors.
3510 In other words, Exim assumes they were pipelined, though this may not
3511 actually be the case. Of course, in all cases the client gets an
3512 appropriate error code.
3514 15. If a lookup fails in an ACL condition, a message about the failure may
3515 be available; it is used if testing the ACL cannot continue, because most
3516 such messages specify what the cause of the deferral is. However, some
3517 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3518 that caused an old message to be retained and used if a later statement
3519 caused a defer, replacing the real cause of the deferral.
3521 16. If an IP address had so many PTR records that the DNS lookup buffer
3522 was not large enough to hold them, Exim could crash while trying to process
3523 the truncated data. It now detects and logs this case.
3525 17. Further to 4.21/58, another change has been made: if (and only if) the
3526 first line of a message (the first header line) ends with CRLF, a bare LF
3527 in a subsequent header line has a space inserted after it, so as not to
3528 terminate the header.
3530 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3531 unnecessarily, used atoi() instead of strtol(), and didn't check the
3532 termination when getting file sizes from file names by regex.
3534 19. Completely re-implemented the support for maildirsize files, in the light
3535 of a number of problems with the previous contributed implementation
3536 (4.30/29). In particular:
3538 . If the quota is zero, the maildirsize file is maintained, but no quota is
3541 . If the maildir directory does not exist, it is created before any attempt
3542 to write a maildirsize file.
3544 . The quota value in the file is just a cache; if the quota is changed in
3545 the transport, the new value overrides.
3547 . A regular expression is available for excluding directories from the
3550 20. The autoreply transport checks the characters in options that define the
3551 message's headers; it allows continued headers, but it was checking with
3552 isspace() after an embedded newline instead of explicitly looking for a
3555 21. If all the "regular" hosts to which an address was routed had passed their
3556 expiry times, and had not reached their retry times, the address was
3557 bounced, even if fallback hosts were defined. Now Exim should go on to try
3560 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3561 equivalent code in the SMTP transport. Some hosts send humungous responses
3562 to HELO/EHLO, more than 1024 it seems.
3564 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3565 gives compiler warnings in some environments. I've now done it "properly",
3568 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3569 (because of problems with the built-in one) was declared to return uschar *
3570 instead of char *, causing compiler failure.
3572 25. Fixed a file descriptor leak when processing alias/forward files.
3574 26. Fixed a minor format string issue in dbfn.c.
3576 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3578 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3579 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3582 29. When a sender address is verified, it is cached, to save repeating the test
3583 when there is more than one recipient in a message. However, when the
3584 verification involves a callout, it is possible for different callout
3585 options to be set for different recipients. It is too complicated to keep
3586 track of this in the cache, so now Exim always runs a verification when a
3587 callout is required, relying on the callout cache for the optimization.
3588 The overhead is duplication of the address routing, but this should not be
3591 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3592 to be verified with callout=postmaster, and the main callout worked but the
3593 postmaster check failed, the verification correctly failed. However, if a
3594 subsequent RCPT command asked for sender verification *without* the
3595 postmaster check, incorrect caching caused this verification also to fail,
3598 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3599 it was not caching the DNS options (qualify_single, search_parents) that
3600 were used when the lookup failed. A subsequent lookup with different
3601 options therefore always gave the same answer, though there were cases
3602 where it should not have. (Example: a "domains = !$mx_any" option on a
3603 dnslookup router: the "domains" option is always processed without any
3604 widening, but the router might have qualify_single set.) Now Exim uses the
3605 cached value only when the same options are set.
3607 32. Added John Jetmore's "exipick" utility to the distribution.
3609 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3610 than a timeout (e.g. a certificate is required, and is not provided), an
3611 Exim server now closes the connection immediately. Previously it waited for
3612 the client to close - but if the client is SSL, it seems that they each
3613 wait for each other, leading to a delay before one of them times out.
3615 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3616 maintained 0.8.x compatibility because I don't think many are using it, and
3617 it is clearly obsolete.
3619 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3622 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3623 client certificate was expired. A simple patch fixes this, though I don't
3624 understand the full logic of why the verify callback is called multiple
3627 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3628 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3629 which causes problems with some clients (such as the Certicom SSL Plus
3630 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3631 disables the coutermeasure allowing Eudora to connect."
3633 38. Exim was not checking that a write() to a log file succeeded. This could
3634 lead to Bad Things if a log got too big, in particular if it hit a file
3635 size limit. Exim now panics and dies if it cannot write to a log file, just
3636 as it does if it cannot open a log file.
3638 39. Modified OS/Makefile-Linux so that it now contains
3640 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3642 The two -D definitions ensure that Exim is compiled with large file
3643 support, which makes it possible to handle log files that are bigger than
3646 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3647 instance) a domain was checked against a named list that involved a lookup,
3648 causing $domain_data to be set, then another domain was checked against the
3649 same list, then the first domain was re-checked, the value of $domain_data
3650 after the final check could be wrong. In particular, if the second check
3651 failed, it could be set empty. This bug probably also applied to
3654 41. The strip_trailing_dot option was not being applied to the address given
3655 with the -f command-line option.
3657 42. The code for reading a message's header from the spool was incrementing
3658 $received_count, but never initializing it. This meant that the value was
3659 incorrect (doubled) while delivering a message in the same process in which
3660 it was received. In the most common configuration of Exim, this never
3661 happens - a fresh exec is done - but it can happen when
3662 deliver_drop_privilege is set.
3664 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3665 it now includes up to 150 characters of the unexpected data in the log
3668 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3669 and building data strings. The size of both of these buffers was 10 000
3670 bytes - far larger than anybody would *ever* want, thought I. Needless to
3671 say, somebody hit the limit. I have increased the maximum line length to
3672 20 000 and the maximum data length of concatenated lines to 100 000. I have
3673 also fixed two bugs, because there was no checking on these buffers. Tsk,
3674 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3675 buffer is too small.
3677 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3678 lsearch lookups. Now it does.
3680 46. When parsing a route_list item in a manualroute router, a fixed-length
3681 buffer was used for the list of hosts. I made this 1024 bytes long,
3682 thinking that nobody would ever have a list of hosts that long. Wrong.
3683 Somebody had a whole pile of complicated expansion conditions, and the
3684 string was silently truncated, leading to an expansion error. It turns out
3685 that it is easier to change to an unlimited length (owing to other changes
3686 that have happened since this code was originally written) than to build
3687 structure for giving a limitation error. The length of the item that
3688 expands into the list of hosts is now unlimited.
3690 47. The lsearch lookup could not handle data where the length of text line was
3691 more than 4095 characters. Such lines were truncated, leading to shortened
3692 data being returned. It should now handle lines of any length.
3694 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3695 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3698 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3699 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3700 no longer happens in comment lines. A list of replacements is now placed
3701 at the head of all of the source files, except those whose only change is
3702 to replace PERL_COMMAND in the very first #! line.
3704 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3705 messages on the queue, with a bottom-up merge sort, using code contributed
3706 by Michael Haardt. This should make operations like -bp somewhat faster on
3707 large queues. It won't affect queue runners, except when queue_run_in_order
3710 51. Installed eximstats 1.31 in the distribution.
3712 52. Added support for SRV lookups to the dnslookup router.
3714 53. If an ACL referred to $message_body or $message_body_end, the value was not
3715 reset for any messages that followed in the same SMTP session.
3717 54. The store-handling optimization for building very long strings was not
3718 differentiating between the different store pools. I don't think this
3719 actually made any difference in practice, but I've tidied it.
3721 55. While running the routers to verify a sender address, $sender_address
3722 was still set to the sender address. This is wrong, because when routing to
3723 send a bounce to the sender, it would be empty. Therefore, I have changed
3724 it so that, while verifying a sender address, $sender_address is set to <>.
3725 (There is no change to what happens when verifying a recipient address.)
3727 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3728 target A or AAAA records (if not already returned) without resetting the
3729 qualify_single or search_parents options of the DNS resolver. These are
3730 inappropriate in this case because the targets of MX and SRV records must
3731 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3732 target that, when qualified, matched something in the local domain. These
3733 two options are now turned off when doing these lookups.
3735 57. It seems that at least some releases of Reiserfs (which does not have the
3736 concept of a fixed number of inodes) returns zero and not -1 for the
3737 number of available inodes. This interacted badly with check_spool_inodes,
3738 which assumed that -1 was the "no such thing" setting. What I have done is
3739 to check that the total number of inodes is greater than zero before doing
3740 the test of how many are available.
3742 58. When a "warn" ACL statement has a log_message modifier, the message is
3743 remembered, and not repeated. This is to avoid a lot of repetition when a
3744 message has many recipients that cause the same warning to be written.
3745 Howewer, Exim was preserving the list of already written lines for an
3746 entire SMTP session, which doesn't seem right. The memory is now reset if a
3747 new message is started.
3749 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3750 debugging output unless log_rewrite was also set.
3752 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3753 of (int)(handle) when we know that handle contains (void *)(-1).
3755 61. The Exim daemon panic-logs an error return when it closes the incoming
3756 connection. However "connection reset by peer" seems to be common, and
3757 isn't really an error worthy of noting specially, so that particular error
3760 62. When Exim is trying to find all the local interfaces, it used to panic and
3761 die if the ioctl to get the interface flags failed. However, it seems that
3762 on at least one OS (Solaris 9) it is possible to have an interface that is
3763 included in the list of interfaces, but for which you get a failure error
3764 for this call. This happens when the interface is not "plumbed" into a
3765 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3766 failure of the "get flags" call assumes that the interface is down.
3768 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3769 makes life easier for people who are doing arithmetic on fields extracted
3770 from dates, where you often get leading zeros that should not be
3771 interpreted as octal.
3773 64. Added qualify_domain to the redirect router, to override the global
3776 65. If a pathologically long header line contained very many addresses (the
3777 report of this problem mentioned 10 000) and each of them was rewritten,
3778 Exim could use up a very large amount of memory. (It kept on making new
3779 copies of the header line as it rewrote, and never released the old ones.)
3780 At the expense of a bit more processing, the header rewriting function has
3781 been changed so that it no longer eats memory in this way.
3783 66. The generation of the Received: header has been moved from the time that a
3784 message starts to be received, to the time that it finishes. The timestamp
3785 in the Received: header should now be very close to that of the <= log
3786 line. There are two side-effects of this change:
3788 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3789 logged header lines no longer include the local Received: line, because
3790 it has not yet been created. The same applies to a copy of the message
3791 that is returned to a non-SMTP sender when a message is rejected.
3793 (b) When a filter file is tested using -bf, no additional Received: header
3794 is added to the test message. After some thought, I decided that this
3797 This change does not affect the value of $received_for. It is still set
3798 after address rewriting, but before local_scan() is called.
3800 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3802 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3803 gave an unhelpful panic error message, and a defer error. I have managed to
3804 change this behaviour so that it now rejects any supplied certificate,
3805 which seems right, as the list of acceptable certificates is empty.
3807 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3808 gave an unhelpful defer error. I have not managed to make this reject any
3809 supplied certificates, but the error message it gives is "no certificate
3810 supplied", which is not helpful.
3812 70. exigrep's output now also includes lines that are not associated with any
3813 message, but which match the given pattern. Implemented by a patch from
3814 Martin Sluka, which also tidied up the Perl a bit.
3816 71. Recipient callout verification, like sender verification, was using <> in
3817 the MAIL FROM command. This isn't really the right thing, since the actual
3818 sender may affect whether the remote host accepts the recipient or not. I
3819 have changed it to use the actual sender in the callout; this means that
3820 the cache record is now keyed on a recipient/sender pair, not just the
3821 recipient address. There doesn't seem to be a real danger of callout loops,
3822 since a callout by the remote host to check the sender would use <>.
3823 [SEE ABOVE: changed after hitting problems.]
3825 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3826 temporary errors. However, in the case of such a code being given after
3827 the end of a data transmission (i.e. after ".") Exim was failing to write
3828 a retry record for the message. (Yes, there was some broken host that was
3829 actually sending 8xx at this point.)
3831 73. An unknown lookup type in a host list could cause Exim to panic-die when
3832 the list was checked. (An example that provoked this was putting <; in the
3833 middle of a list instead of at the start.) If this happened during a DATA
3834 ACL check, a -D file could be left lying around. This kind of configuration
3835 error no longer causes Exim to die; instead it causes a defer errror. The
3836 incident is still logged to the main and panic logs.
3838 74. Buglet left over from Exim 3 conversion. The message "too many messages
3839 in one connection" was written to the rejectlog but not the mainlog, except
3840 when address rewriting (yes!) was being logged.
3842 75. Added write_rejectlog option.
3844 76. When a system filter was run not as root (that is, when system_filter_user
3845 was set), the values of the $n variables were not being returned to the
3846 main process; thus, they were not subsequently available in the $sn
3849 77. Added +return_path_on_delivery log selector.
3851 78. A connection timeout was being treated differently from recipients deferred
3852 when testing hosts_max_try with a message that was older than the host's
3853 retry timeout. (The host should not be counted, thus allowing all hosts to
3854 be tried at least once before bouncing.) This may have been the cause of an
3855 occasionally reported bug whereby a message would remain on the queue
3856 longer than the retry timeout, but would be bounced if a delivery was
3857 forced. I say "may" because I never totally pinned down the problem;
3858 setting up timeout/retry tests is difficult. See also the next item.
3860 79. The ultimate address timeout was not being applied to errors that involved
3861 a combination of host plus message (for example, a timeout on a MAIL
3862 command). When an address resolved to a number of possible hosts, and they
3863 were not all tried for each delivery (e.g. because of hosts_max_try), a
3864 message could remain on the queue longer than the retry timeout.
3866 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3869 81. Fixed an obscure SMTP outgoing bug which required at least the following
3870 conditions: (a) there was another message waiting for the same server;
3871 (b) the server returned 5xx to all RCPT commands in the first message so
3872 that the message was not completed; (c) the server dropped the connection
3873 or gave a negative response to the RSET that Exim sends to abort the
3874 transaction. The observed case was a dropped connection after DATA that had
3875 been sent in pipelining mode. That is, the server had advertised PIPELINING
3876 but was not implementing it correctly. The effect of the bug was incorrect
3877 behaviour, such as trying another host, and this could lead to a crash.
3883 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3884 and daemon.c were passed as pointers to ints; they should have been
3885 pointers to socklen_t variables (which are typically unsigned ints).
3887 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3890 3. Fixed a really odd bug that affected only the testing scheme; patching a
3891 certain fixed string in the binary changed the value of another string that
3892 happened to be identical to the end of the original first string.
3894 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3895 name", it returns that address as the IP address. On some operating
3896 systems (e.g. Solaris), it also passes back the IP address string as the
3897 "host name". However, on others (e.g. Linux), it passes back an empty
3898 string. Exim wasn't checking for this, and was changing the host name to an
3899 empty string, assuming it had been canonicized.
3901 5. Although rare, it is permitted to have more than one PTR record for a given
3902 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3903 all the names associated with an address, because they do in Solaris.
3904 However, it seems that they do not in Linux for data that comes from the
3905 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3906 I found this out when I moved to a new Linux workstation and tried to run
3907 the Exim test suite.
3909 To get round this problem I have changed the code so that it now does its
3910 own call to the DNS to look up PTR records when searching for a host name.
3911 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3912 addresses that are only in /etc/hosts are still found.
3914 This behaviour is, however, controlled by an option called host_lookup_
3915 order, which defaults to "bydns:byaddr". If people want to use the other
3916 order, or indeed, just use one or the other means of lookup, they can
3917 specify it in this variable.
3919 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3920 some operating systems, this comes back from gethostbyaddr() as an empty
3921 string, and this is what Exim used to test for. However, it seems that in
3922 other systems, "." is yielded. Exim now tests for this case too.
3924 7. The values of check_spool_space and check_log_space are now held internally
3925 as a number of kilobytes instead of an absolute number of bytes. If a
3926 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3927 kilobyte. This means that much larger values can be stored.
3929 8. Exim monitor: an attempt to get the action menu when not actually pointing
3930 at a message produces an empty menu entitled "No message selected". This
3931 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3932 no entries in it ("Shell widget menu has zero width and/or height"). So I
3933 have added a single, blank menu entry in this case.
3935 9. Added ${quote_local_part.
3937 10. MIME decoding is now applied to the contents of Subject: header lines when
3940 11. Now that a reference to $sender_host_address automatically causes a reverse
3941 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3942 host lookup before query-style lookups in lists that might use this
3943 variable. This has therefore been abolished, and the "net-" prefix is no
3944 longer necessary for query-style lookups.
3946 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3947 have been a typo for LFLAGS, so it has been changed.
3949 13. The install script calls Exim with "-C /dev/null" in order to find the
3950 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3951 to be output. Howeve, since Exim outputs its version number before the
3952 error, it didn't break the script. It just looked ugly. I fixed this by
3953 always allowing "-C /dev/null" if the caller is root.
3955 14. Ignore overlarge ACL variable number when reading spool file - insurance
3956 against a later release with more variables having written the file.
3958 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3959 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3960 this, as well as the form without the "IPv6" on the front.
3962 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3963 OS/Makefile-Darwin file.
3965 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3966 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3968 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3969 variable. However, some OS use ints for inodes. I've added cast to long int
3970 to get rid of the compiler warning.
3972 19. I had forgotten to lock out "/../" in configuration file names when
3973 ALT_CONFIG_PREFIX was set.
3975 20. Routers used for verification do not need to specify transports. However,
3976 if such a router generated a host list, and callout was configured, Exim
3977 crashed, because it could not find a port number from the (non-existent)
3978 transport. It now assumes port 25 in this circumstance.
3980 21. Added the -t option to exigrep.
3982 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3983 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3984 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3985 them is set, LOOKUP_LSEARCH is forced.
3987 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3989 24. Added sender and host information to the "rejected by local_scan()" log
3990 line; previously there was no indication of these.
3992 25. Added .include_if_exists.
3994 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3995 Linux. It turns out that not all file systems support this. Apparently some
3996 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3997 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3998 not supported on the file descriptor, is now ignored when Exim is trying to
3999 sync a directory. This applies only to Linux.
4001 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4003 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4004 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4007 29. Added support for maildirsize files from supplied patch (modified a bit).
4009 30. The use of :fail: followed by an empty string could lead Exim to respond to
4010 sender verification failures with (e.g.):
4012 550 Verification failed for <xxx>
4013 550 Sender verify failed
4015 where the first response line was missing the '-' that indicates it is not
4016 the final line of the response.
4018 31. The loop for finding the name of the user that called Exim had a hardwired
4019 limit of 10; it now uses the value of finduser_retries, which is used for
4020 all other user lookups.
4022 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4025 33. Exim was neglecting to zero errno before one call of strtol() when
4026 expanding a string and expecting an integer value. On some systems this
4027 resulted in spurious "integer overflow" errors. Also, it was casting the
4028 result into an int without checking.
4030 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4031 did not work. The code looks as if it has *never* worked, though it appears
4032 to have been documented since at least releast 1.62. I have made it work.
4034 35. The "timeout_DNS" error in retry rules, also documented since at least
4035 1.62, also never worked. As it isn't clear exactly what this means, and
4036 clearly it isn't a major issue, I have abolished the feature by treating it
4037 as "timeout", and writing a warning to the main and panic logs.
4039 36. The display of retry rules for -brt wasn't always showing the error code
4042 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4043 timeout_connect_A, timeout_connect_MX.
4045 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4046 to the empty sender.
4048 39. The daemon was not analysing the content of -oX till after it had closed
4049 stderr and disconnected from the controlling terminal. This meant that any
4050 syntax errors were only noted on the panic log, and the return code from
4051 the command was 0. By re-arranging the code a little, I've made the
4052 decoding happen first, so such errors now appear on stderr, and the return
4053 code is 1. However, the actual setting up of the sockets still happens in
4054 the disconnected process, so errors there are still only recorded on the
4057 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4058 connections (as happens on some IP stacks) was logged at start up time as
4059 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4060 it from "IPv6 and IPv4", which means that two separate sockets are being
4063 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4064 says whether AF_INET or AF_INET6 was passed as an argument.
4066 42. Exiwhat output was messed up when time zones were included in log
4069 43. Exiwhat now gives more information about the daemon's listening ports,
4070 and whether -tls-on-connect was used.
4072 44. The "port" option of the smtp transport is now expanded.
4074 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4075 silently ignored. Now an error message is written to the main and panic
4078 46. There's a new ACL modifier called "logwrite" which writes to a log file
4079 as soon as it is encountered.
4081 47. Added $local_user_uid and $local_user_gid at routing time.
4083 48. Exim crashed when trying to verify a sender address that was being
4086 49. Exim was recognizing only a space character after ".include". It now also
4087 recognizes a tab character.
4089 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4090 extracting the relevant information from the specification. The man page no
4091 longer contains scrambled data for the -d option, and I've added a section
4092 at the front about calling Exim under different names.
4094 51. Added "extra_headers" argument to the "mail" command in filter files.
4096 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4099 53. Installed eximstats 1.29.
4101 54. Added transport_filter_timeout as a generic transport option.
4103 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4104 Cc: header lines. This was required by RFC 822, but it not required by RFC
4107 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4108 incoming messages that did not have them. Now it does so only if the
4109 message originates locally, that is, if there is no associated remote host
4110 address. When Resent- header lines are present, this applies to the Resent-
4111 lines rather than the non-Resent- lines.
4113 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4114 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4116 58. Messages for configuration errors now include the name of the main
4117 configuration file - useful now that there may be more than one file in a
4118 list (.included file names were always shown).
4120 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4121 for those rare installations that do not start the daemon as root or run it
4122 setuid root. I've cut out the call to initgroups() if the daemon is not
4125 60. The Exim user and group can now be bound into the binary as text strings
4126 that are looked up at the start of Exim's processing.
4128 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4130 62. Added $mailstore_basename variable.
4132 63. Installed patch to sieve.c from Michael Haardt.
4134 64. When Exim failed to open the panic log after failing to open the main log,
4135 the original message it was trying to log was written to stderr and debug
4136 output, but if they were not available (the usual case in production), it
4137 was lost. Now it is written to syslog before the two lines that record the
4138 failures to open the logs.
4140 65. Users' Exim filters run in subprocesses under the user's uid. It is
4141 possible for a "deliver" command or an alias in a "personal" command to
4142 provoke an address rewrite. If logging of address rewriting is configured,
4143 this fails because the process is not running as root or exim. There may be
4144 a better way of dealing with this, but for the moment (because 4.30 needs
4145 to be released), I have disabled address rewrite logging when running a
4146 filter in a non-root, non-exim process.
4152 1. The buildconfig auxiliary program wasn't quoting the value set for
4153 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4154 not defined. This bug was masked in 4.22 by the effect that was fixed in
4157 2. Some messages that were rejected after a message id was allocated were
4158 shown as "incomplete" by exigrep. It no longer does this for messages that
4159 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4161 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4162 have allow_domain_literals set, the ID did not get logged in the <= line.
4163 Domain literals are now always recognized in Message-ID: header lines.
4165 4. The first argument for a ${extract expansion item is the key name or field
4166 number. Leading and trailing spaces in this item were not being ignored,
4167 causing some misleading effects.
4169 5. When deliver_drop_privilege was set, single queue runner processes started
4170 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4171 same command in the process it spins off) were not dropping privilege.
4173 6. When the daemon running as "exim" started a queue runner, it always
4174 re-executed Exim in the spun-off process. This is a waste of effort when
4175 deliver_drop_privilege is set. The new process now just calls the
4176 queue-runner function directly.
4182 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4185 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4186 ignored. Though the use of -oP was forcing the writing of a pid file, it
4187 was always written to the default place.
4189 3. If the message "no IP address found for host xxxx" is generated during
4190 incoming verification, it is now followed by identification of the incoming
4191 connection (so you can more easily find what provoked it).
4193 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4195 5. Added some features to "harden" Exim a bit more against certain attacks:
4197 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4198 be put in Local/Makefile. This is like the never_users runtime option,
4199 but it cannot be overridden. The default setting is "root".
4201 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4202 prefix string with which any file named in a -C command line option
4205 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4206 is retained for -C and -D only if the caller of Exim is root. Without
4207 it, the exim user may also use -C and -D and retain privilege.
4209 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4210 command line option is disabled.
4212 6. Macro names set by the -D option must start with an upper case letter, just
4213 like macro names defined in the configuration file.
4215 7. Added "dereference=" facility to LDAP.
4217 8. Two instances of the typo "uknown" in the source files are fixed.
4219 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4220 the Configure-Makefile script screwed up while processing it.
4222 10. Incorporated PCRE 4.4.
4224 11. The SMTP synchronization check was not operating right at the start of an
4225 SMTP session. For example, it could not catch a HELO sent before the client
4226 waited for the greeting. There is now a check for outstanding input at the
4227 point when the greeting is written. Because of the duplex, asynchronous
4228 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4229 way, but not yet received, when the check is performed.
4231 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4232 on TCP/IP sockets, because this apparently causes some broken clients to
4235 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4236 unchanged) from the Cygwin maintainer.
4238 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4239 was supported instead of testing for mbx. Effectively a typo.
4241 15. The spa authenticator server code was not checking that the input it
4242 received was valid base64.
4244 16. The debug output line for the "set" modifier in ACLs was not showing the
4245 name of the variable that was being set.
4247 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4249 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4250 lookup on its own. Something else was needed to trigger the lookup. For
4251 example, a match in host_lookup or the need for a host name in a host list.
4252 Now, if $sender_host_name is referenced and the host name has not yet been
4253 looked up, a lookup is performed. If the lookup fails, the variable remains
4254 empty, and $host_lookup_failed is set to "1".
4256 19. Added "eqi" as a case-independent comparison operator.
4258 20. The saslauthd authentication condition could segfault if neither service
4259 nor realm was specified.
4261 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4262 error message that was logged was misleading, and incoming SMTP
4263 connections were dropped. The message is now more accurate, and temporary
4264 errors are given to SMTP connections.
4266 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4267 (except RSET) with a 421 error, until QUIT is received. However, it was
4268 failing to send a response to QUIT.
4270 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4271 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4272 ACL is now run after the helo_try_verify_hosts code.
4274 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4275 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4276 case-independent, so other case variants are also recognized). Apparently
4277 some systems use these upper case variants.
4279 25. If more than two messages were waiting for the same host, and a transport
4280 filter was specified for the transport, Exim sent two messages over the
4281 same TCP/IP connection, and then failed with "socket operation on non-
4282 socket" when it tried to send the third.
4284 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4286 27. The extern definition of crypt16() in expand.c was not being excluded when
4287 the OS had its own crypt16() function.
4289 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4290 as a preferred synonym for return_size_limit, both as an option and as an
4293 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4295 30. Changed the default configuration ACL to relax the local part checking rule
4296 for addresses that are not in any local domains. For these addresses,
4297 slashes and pipe symbols are allowed within local parts, but the sequence
4298 /../ is explicitly forbidden.
4300 31. SPA server authentication was not clearing the challenge buffer before
4303 32. log_message in a "warn" ACL statement was writing to the reject log as
4304 well as to the main log, which contradicts the documentation and doesn't
4305 seem right (because no rejection is happening). So I have stopped it.
4307 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4308 However, I am unable to do any testing of this.
4310 34. Fixed an infelicity in the appendfile transport. When checking directories
4311 for a mailbox, to see if any needed to be created, it was accidentally
4312 using path names with one or more superfluous leading slashes; tracing
4313 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4315 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4316 contents are added to the log line that is written for every discarded
4317 recipient. (Previously a log_message setting was ignored.)
4319 36. The ${quote: operator now quotes the string if it is empty.
4321 37. The install script runs exim in order to find its version number. If for
4322 some reason other than non-existence or emptiness, which it checks, it
4323 could not run './exim', it was installing it with an empty version number,
4324 i.e. as "exim-". This error state is now caught, and the installation is
4327 38. An argument was missing from the function that creates an error message
4328 when Exim fails to connect to the socket for saslauthd authentication.
4329 This could cause Exim to crash, or give a corrupted message.
4331 39. Added isip, isip4, and isip6 to ${if conditions.
4333 40. The ACL variables $acl_xx are now saved with the message, and can be
4334 accessed later in routers, transports, and filters.
4336 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4337 strings in the file are not string-expanded.
4339 42. If a MAIL command specified a SIZE value that was too large to fit into an
4340 int variable, the check against message_size_limit failed. Such values are
4341 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4342 day this will have to be increased, but I don't think I want to be around
4343 when emails are that large.
4350 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4351 iconv() is not standard in FreeBSD.
4353 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4354 IPv6 enabled. The observed symptom was a segmentation fault on return from
4355 the function os_common_find_running_interfaces() in src/os.c.
4357 3. In the check_special_case() function in daemon.c I had used "errno" as an
4358 argument name, which causes warnings on some systems. This was basically a
4359 typo, since it was named "eno" in the comments!
4361 4. The code that waits for the clock to tick (at a resolution of some fraction
4362 of a second) so as to ensure message-id uniqueness was always waiting for
4363 at least one whole tick, when it could have waited for less. [This is
4364 almost certainly not relevant at current processor speeds, where it is
4365 unlikely to ever wait at all. But we try to future-proof.]
4367 5. The function that sleeps for a time interval that includes fractions of a
4368 second contained a race. It did not block SIGALRM between setting the
4369 timer, and suspending (a couple of lines later). If the interval was short
4370 and the sigsuspend() was delayed until after it had expired, the suspension
4371 never ended. On busy systems this could lead to processes getting stuck for
4374 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4375 process, before it forks any delivery processes. The open lookup caching
4376 mechanism meant that the open file or database connection was passed into
4377 the delivery process. The problem was that delivery processes always tidy
4378 up cached lookup data. This could cause a problem for the next delivery
4379 process started by the queue runner, because the external queue runner
4380 process does not know about the closure. So the next delivery process
4381 still has data in the lookup cache. In the case of a file lookup, there was
4382 no problem because closing a file descriptor in a subprocess doesn't affect
4383 the parent. However, if the lookup was caching a connection to a database,
4384 the connection was closed, and the second delivery process was likely to
4385 see errors such as "PGSQL: query failed: server closed the connection
4386 unexpectedly". The problem has been fixed by closing all cached lookups
4387 in a queue runner before running a delivery process.
4389 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4390 seem to have the "const" qualifier which it has on other OS. I've
4393 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4394 *of the same type* (client or server) with the same public name that an
4395 error should be diagnosed.
4397 9. When Exim looked up a host name for an IP address, but failed to find the
4398 original IP address when looking up the host name (a safety check), it
4399 output the message "<ip address> does not match any IP for NULL", which was
4400 confusing, to say the least. The bug was that the host name should have
4401 appeared instead of "NULL".
4403 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4404 user that is built into the binary, and the -C or -D options is used, root
4405 privilege is dropped before the configuration file is read. In addition,
4406 logging is switched to stderr instead of the normal log files. If the
4407 configuration then re-defines the Exim user, the unprivileged environment
4408 is probably not what is expected, so Exim logs a panic warning message (but
4411 However, if deliver_drop_privilege is set, the unprivileged state may well
4412 be exactly what is intended, so the warning has been cut out in that case,
4413 and Exim is allowed to try to write to its normal log files.
4419 1. smtp_return_error_details was not giving details for temporary sender
4420 or receiver verification errors.
4422 2. Diagnose a configuration error if two authenticators have the same public
4425 3. Exim used not to create the message log file for a message until the first
4426 delivery attempt. This could be confusing when incoming messages were held
4427 for policy or load reasons. The message log file is now created at the time
4428 the message is received, and an initial "Received" line is written to it.
4430 4. The automatically generated man page for command line options had a minor
4431 bug that caused no ill effects; however, a more serious problem was that
4432 the procedure for building the man page automatically didn't always
4433 operate. Consequently, release 4.20 contains an out-of-date version. This
4434 shouldn't happen again.
4436 5. When building Exim with embedded Perl support, the script that builds the
4437 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4438 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4440 6. The freeze_tell option was not being used for messages that were frozen on
4441 arrival, either by an ACL or by local_scan().
4443 7. Added the smtp_incomplete_transaction log selector.
4445 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4446 was accepting AUTH without a new EHLO.
4448 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4449 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4452 10. Logging of TCP/IP connections (when configured) now happens in the main
4453 daemon process instead of the child process, so that the TCP/IP connection
4454 count is more accurate (but it can never be perfect).
4456 11. The use of "drop" in a nested ACL was not being handled correctly in the
4457 outer ACL. Now, if condition failure induced by the nested "drop" causes
4458 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4459 or "require"), the connection is dropped.
4461 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4462 that yield "discard" can now be used with an "accept" or a "discard" verb,
4463 but an error is generated for any others (because I can't see a useful way
4464 to define what should happen).
4466 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4467 are now processed in the same way as lines in the Exim configuration file.
4468 In particular, continuation lines are supported.
4470 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4472 15. Added -ti meaning -t -i.
4474 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4475 domains, and warn by logging if others are found.
4477 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4478 returned from GIFCONF when Exim is trying to find the list of interfaces on
4479 a host. The code in os.c has been modified to copy each ifreq to an aligned
4480 structure in all cases.
4482 Also, in some cases, the returned ifreq's were being copied to a 'struct
4483 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4484 means the last couple of bytes of an IPv6 address could be chopped if the
4485 ifreq contained only a normal sockaddr (14 bytes storage).
4487 18. Named domain lists were not supported in the hosts_treat_as_local option.
4488 An entry such as +xxxx was not recognized, and was treated as a literal
4491 19. Ensure that header lines added by a DATA ACL are included in the reject log
4492 if the ACL subsequently rejects the message.
4494 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4495 MD5 (which is deprecated).
4497 21. When testing a filter file using -bf, Exim was writing a message when it
4498 took the sender from a "From " line in the message, but it was not doing so
4499 when it took $return_path from a Return-Path: header line. It now does.
4501 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4502 with a valid header line field name (a series of printing characters
4503 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4505 23. Changed "disc" in the source to "disk" to conform to the documentation and
4506 the book and for uniformity.
4508 24. Ignore Sendmail's -Ooption=value command line item.
4510 25. When execve() failed while trying to run a command in a pipe transport,
4511 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4512 could be confused with a return value of 69 from the command itself. This
4513 has been changed to 127, the value the shell returns if it is asked to run
4514 a non-existent command. The wording for the related log line suggests a
4515 non-existent command as the problem.
4517 26. If received_header_text expands to an empty string, do not add a Received:
4518 header line to the message. (Well, it adds a token one on the spool, but
4519 marks it "old" so that it doesn't get used or transmitted.)
4521 27. Installed eximstats 1.28 (addition of -nt option).
4523 28. There was no check for failure on the call to getsockname() in the daemon
4524 code. This can fail if there is a shortage of resources on the system, with
4525 ENOMEM, for example. A temporary error is now given on failure.
4527 29. Contrary to the C standard, it seems that in some environments, the
4528 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4529 program. Exim now does this explicitly; it affects the formatting of
4530 timestamps using strftime().
4532 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4533 complaints. I've now initialized all the variables, to avoid this.
4535 32. Header lines added by a system filter were not being "seen" during
4536 transport-time rewrites.
4538 33. The info_callback() function passed to OpenSSL is set up with type void
4539 (*)(SSL *, int, int), as described somewhere. However, when calling the
4540 function (actually a macro) that sets it up, the type void(*)() is
4541 expected. I've put in a cast to prevent warnings from picky compilers.
4543 34. If a DNS black list lookup found a CNAME record, but there were no A
4544 records associated with the domain it pointed at, Exim crashed.
4546 35. If a DNS black list lookup returned more than one A record, Exim ignored
4547 all but the first. It now scans all returned addresses if a particular IP
4548 value is being sought. In this situation, the contents of the
4549 $dnslist_value variable are a list of all the addresses, separated by a
4552 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4553 to accept a host name and all its aliases if the forward lookup for any of
4554 them yielded the IP address of the incoming connection. Now it accepts only
4555 those names whose forward lookup yields the correct IP address. Any other
4556 names are discarded. This closes a loophole whereby a rogue DNS
4557 administrator could create reverse DNS records to break through a
4558 wildcarded host restriction in an ACL.
4560 37. If a user filter or a system filter that ran in a subprocess used any of
4561 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4562 the wrong values were passed to the pipe command ($thisaddress had the
4563 value of $0, $0 had the value of $1, etc). This bug was introduced by
4564 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4566 38. Improved the line breaking for long SMTP error messages from ACLs.
4567 Previously, if there was no break point between 40 and 75 characters, Exim
4568 left the rest of the message alone. Two changes have been made: (a) I've
4569 reduced the minimum length to 35 characters; (b) if it can't find a break
4570 point between 35 and 75 characters, it looks ahead and uses the first one
4571 that it finds. This may give the occasional overlong line, but at least the
4572 remaining text gets split now.
4574 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4575 file descriptors might be low, and that setting 1000 would always raise it.
4576 It turns out that in some environments, the limit is already over 1000 and
4577 that lowering it causes trouble. So now Exim takes care not to decrease it.
4579 40. When delivering a message, the value of $return_path is set to $sender_
4580 address at the start of routing (routers may change the value). By an
4581 oversight, this default was not being set up when an address was tested by
4582 -bt or -bv, which affected the outcome if any router or filter referred to
4585 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4586 log only when "message" or "log_message" are set. However, if one of the
4587 conditions was an address verification, or a call to a nested ACL, the
4588 messages generated by the underlying test were being passed through. This
4589 no longer happens. The underlying message is available in $acl_verify_
4590 message for both "message" and "log_message" expansions, so it can be
4591 passed through if needed.
4593 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4594 new expansion $bh_ to give the encoded byte string without charset
4595 translation. Translation happens only if iconv() is available; HAVE_ICONV
4596 indicates this at build time. HEADERS_CHARSET gives the charset to
4597 translate to; headers_charset can change it in the configuration, and
4598 "headers charset" can change it in an individual filter file.
4600 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4601 that creates RFC 2047 encoded "words" labels them as that charset instead
4602 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4603 expansion operator; (ii) when Exim creates a From: line for a local
4604 message; (iii) when a header line is rewritten to include a "phrase" part.
4606 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4607 buggy, causing it to skip the first lines of messages whose message ID
4608 ended in 'D'. This would not have bitten before Exim release 4.14, because
4609 message IDs were unlikely to end in 'D' before then. The effect was to have
4610 incorrect size information for certain domains.
4612 45. #include "config.h" was missing at the start of the crypt16.c module. This
4613 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4616 46. If there was a timeout during a "random" callout check, Exim treated it as
4617 a failure of the random address, and carried on sending RSET and the real
4618 address. If the delay was just some slowness somewhere, the response to the
4619 original RCPT would be taken as a response to RSET and so on, causing
4620 mayhem of various kinds.
4622 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4623 when I implemented it. It didn't allow for the fact that some option values
4624 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4625 the right test for positive values.
4627 48. Domain names in DNS records are case-independent. Exim always looks them up
4628 in lower case. Some resolvers return domain names in exactly the case they
4629 appear in the zone file, that is, they may contain uppercase letters. Not
4630 all resolvers do this - some return always lower case. Exim was treating a
4631 change of case by a resolver as a change of domain, similar to a widening
4632 of a domain abbreviation. This triggered its re-routing code and so it was
4633 trying to route what was effectively the same domain again. This normally
4634 caused routing to fail (because the router wouldn't handle the domain
4635 twice). Now Exim checks for this case specially, and just changes the
4636 casing of the domain that it ultimately uses when it transmits the message
4639 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4642 50. If a filter generated a file delivery with a non-absolute name (possible if
4643 no home directory exists for the router), the forbid_file option was not
4646 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4647 the existing equality matching.
4649 52. Exim was using ints instead of ino_t variables in some places where it was
4650 dealing with inode numbers.
4652 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4653 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4654 called TMPDIR, and if it finds it is different, it changes its value.
4656 54. The smtp_printf() function is now made available to local_scan() so
4657 additional output lines can be written before returning. There is also an
4658 smtp_fflush() function to enable the detection of a dropped connection.
4659 The variables smtp_input and smtp_batched_input are exported to
4662 55. Changed the default runtime configuration: the message "Unknown user"
4663 has been removed from the ACL, and instead placed on the localuser router,
4664 using the cannot_route_message feature. This means that any verification
4665 failures that generate their own messages won't get overridden. Similarly,
4666 the "Unrouteable address" message that was in the ACL for unverifiable
4667 relay addresses has also been removed.
4669 56. Added hosts_avoid_esmtp to the smtp transport.
4671 57. The exicyclog script was not checking for the esoteric option
4672 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4673 will work only if exicyclog is run under the appropriate euid.
4675 58. Following a discussion on the list, the rules by which Exim recognises line
4676 endings on incoming messages have been changed. The -dropcr and drop_cr
4677 options are now no-ops, retained only for backwards compatibility. The
4678 following line terminators are recognized: LF CRLF CR. However, special
4679 processing applies to CR:
4681 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4682 nor a local message in the state where . is a terminator.
4684 (ii) If a bare CR is encountered in a header line, an extra space is added
4685 after the line terminator so as not to end the header. The reasoning
4686 behind this is that bare CRs in header lines are most likely either
4687 to be mistakes, or people trying to play silly games.
4689 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4690 was being incorrectly given as 18 bytes larger than it should have been.
4691 This is a VOB (very old bug).
4693 60. This may never have affected anything current, but just in case it has:
4694 When the local host is found other than at the start of a list of hosts,
4695 the local host, those with the same MX, and any that follow, are discarded.
4696 When the list in question was part of a longer list of hosts, the following
4697 hosts (not currently being processed) were also being discarded. This no
4698 longer happens. I'm not sure if this situation could ever has previously
4701 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4704 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4705 header. This is something that is recommended in a new Internet Draft, and
4706 is something that is documented as being done by Sendmail. There are two
4707 possible values. For messages generated by the autoreply transport, Exim
4710 Auto-Submitted: auto-replied
4712 whereas for all other generated messages (e.g. bounces) it adds
4714 Auto-Submitted: auto-generated
4716 63. The "personal" condition in filters now includes a test for the
4717 Auto-Submitted: header. If it contains the string "auto-" the message it
4718 not considered personal.
4720 64. Added rcpt_include_affixes as a generic transport option.
4722 65. Added queue_only_override (default true).
4724 66. Added the syslog_duplication option.
4726 67. If what should have been the first header line of a message consisted of
4727 a space followed by a colon, Exim was mis-interpreting it as a header line.
4728 It isn't of course - it is syntactically invalid and should therefore be
4729 treated as the start of the message body. The misbehaviour could have
4730 caused a number of strange effects, including loss of data in subsequent
4731 header lines, and spool format errors.
4733 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4734 client host had authenticated. This control can now be exercised by an ACL
4735 for more flexibility.
4737 69. By default, callouts do not happen when testing with -bh. There is now a
4738 variant, -bhc, which does actually run the callout code, including
4739 consulting and updating the callout cache.
4741 70. Added support for saslauthd authentication, courtesy of Alexander
4744 71. If statvfs() failed on the spool or log directories while checking their
4745 size for availability, Exim confusingly gave the error "space shortage".
4746 Furthermore, in debugging mode it crashed with a floating point exception.
4747 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4748 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4749 serious problem, Exim now writes to the main and panic logs when this
4750 happens, with details of the failure. It then refuses to accept the
4751 incoming message, giving the message "spool directory problem" or "log
4752 directory problem" with a 421 code for SMTP messages.
4754 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4755 0, 1, and 2 exist, because some OS complain for execs without them (see
4756 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4757 descriptors. However, the code omitted to check that the open succeeded,
4758 causing mysterious errors if for some reason the permissions on /dev/null
4759 got screwed. Now Exim writes a message to the main and panic logs, and
4760 bombs out if it can't open /dev/null.
4762 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4763 interact so that it is all more flexible. It is supposed to remain
4764 backwards compatible. Also added extra_local_interfaces.
4766 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4767 to bomb out with an assertion failure - to the client this appears as a
4768 connection drop. This problem occurs in the part of the code that was taken
4769 from the Samba project. Fortunately, the assertion is in a very simple
4770 function, so I have fixed this by reproducing the function inline in the
4771 one place where it is called, and arranging for authentication to fail
4772 instead of killing the process with assert().
4774 75. The SPA client code was not working when the server requested OEM rather
4775 than Unicode encoding.
4777 76. Added code to make require_files with a specific uid setting more usable in
4778 the case where statting the file as root fails - usually a non-root-mounted
4779 NFS file system. When this happens and the failure is EACCES, Exim now
4780 forks a subprocess and does the per-uid checking as the relevant uid.
4782 77. Added process_log_path.
4784 78. If log_file_path was not explicitly set, a setting of check_log_space or
4785 check_log_inodes was ignored.
4787 79. If a space check for the spool or log partitions fails, the incident is now
4788 logged. Of course, in the latter case the data may get lost...
4790 80. Added the %p formatting code to string_format() so that it can be used to
4791 print addresses in debug_print(). Adjusted all the address printing in the
4792 debugging in store.c to use %p rather than %d.
4794 81. There was a concern that a line of code in smtp_in.c could overflow a
4795 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4796 initially expressed, the concern was not well-founded, because trailing
4797 spaces are removed early. However, if the trailing spaces were followed by
4798 a NULL, they did not get removed, so the overflow was possible. Two fixes
4801 (a) I re-wrote the offending code in a cleaner fashion.
4802 (b) If an incoming SMTP command contains a NULL character, it is rejected
4805 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4806 runs initgroups(), so that if the Exim user is in any additional groups,
4807 they will be used during message reception.
4813 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob