1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.511 2007/06/14 13:27:11 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
55 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
56 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
59 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
62 PH/02 In an ACL statement such as
64 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
66 if a client was not listed at all, or was listed with a value other than
67 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
68 the condition was not true (as it should be), so access was not denied.
69 The bug was that the ! inversion was incorrectly passed on to the second
70 item. This has been fixed.
72 PH/03 Added additional dnslists conditions == and =& which are different from
73 = and & when the dns lookup returns more than one IP address.
75 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
76 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
78 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
79 FSYNC, which compiles an option called disable_fsync that allows for
80 bypassing fsync(). The documentation is heavily laced with warnings.
82 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
84 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
85 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
86 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
87 including adding "make clean"; (3) Added -fPIC when compiling the test
88 dynamically loaded module, to get rid of a warning.
90 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
91 message fails, move_frozen_messages = true and ignore_bounce_errors_after
92 = 0s. The bug is otherwise harmless.
94 PH/07 There was a bug in the dovecot authenticator such that the value of
95 $auth1 could be overwritten, and so not correctly preserved, after a
96 successful authentication. This usually meant that the value preserved by
97 the server_setid option was incorrect.
99 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
101 PH/09 Installed PCRE release 7.0.
103 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
104 run for batched SMTP input. It is now run at the start of every message
105 in the batch. While fixing this I discovered that the process information
106 (output by running exiwhat) was not always getting set for -bs and -bS
107 input. This is fixed, and it now also says "batched" for BSMTP.
109 PH/11 Added control=no_pipelining.
111 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
112 patch, slightly modified), and move the expansion of helo_data till after
113 the connection is made in the smtp transport (so it can use these
116 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
118 PH/14 Added log_selector = +pid.
120 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
122 PH/16 Add ${if forany and ${if forall.
124 PH/17 Added dsn_from option to vary the From: line in DSNs.
126 PH/18 Flush SMTP output before performing a callout, unless control =
127 no_callout_flush is set.
129 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
130 was true (the default) a successful delivery failed to delete the retry
131 item, thus causing premature timeout of the address. The bug is now
134 PH/20 Added hosts_avoid_pipelining to the smtp transport.
136 PH/21 Long custom messages for fakedefer and fakereject are now split up
137 into multiline reponses in the same way that messages for "deny" and
138 other ACL rejections are.
140 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
141 with slight modification.
143 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
144 draft, changing the syntax and factoring some duplicate code".
146 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
147 for deliveries of the second and subsequent messages over the same SMTP
150 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
151 ${reduce, with only minor "tidies".
153 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
155 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
156 expansion side effects.
158 PH/27 When a message times out after an over-quota error from an Exim-imposed
159 quota, the bounce message says "mailbox is full". This message was not
160 being given when it was a system quota that was exceeded. It now should
163 MH/03 Made $recipients available in local_scan(). local_scan() already has
164 better access to the recipient list through recipients_list[], but
165 $recipients can be useful in postmaster-provided expansion strings.
167 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
168 in the case of a MAIL command with additional options following the
169 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
170 were accidentally chopped off.
172 PH/29 SMTP synchronization checks are implemented when a command is read -
173 there is a check that no more input is waiting when there shouldn't be
174 any. However, for some commands, a delay in an ACL can mean that it is
175 some time before the response is written. In this time, more input might
176 arrive, invalidly. So now there are extra checks after an ACL has run for
177 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
178 pipelining has not been advertised.
180 PH/30 MH's patch to allow iscntrl() characters to be list separators.
182 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
183 returned in the SMTP response when smtp_return_error_details was false.
186 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
187 instead of the C I/O that was originally supplied, because problems were
190 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
191 Exim which did not show up earlier: it was assuming that a call to
192 SSL_CTX_set_info_callback() might give an error value. In fact, there is
193 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
194 was a macro that became an assignment, so it seemed to work. This has
195 changed to a proper function call with a void return, hence the compile
196 error. Exim's code has been fixed.
198 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
201 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
203 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
205 PH/37 If a message is not accepted after it has had an id assigned (e.g.
206 because it turns out to be too big or there is a timeout) there is no
207 "Completed" line in the log. When some messages of this type were
208 selected by exigrep, they were listed as "not completed". Others were
209 picked up by some special patterns. I have improved the selection
210 criteria to be more general.
212 PH/38 The host_find_failed option in the manualroute router can now be set
213 to "ignore", to completely ignore a host whose IP address cannot be
214 found. If all hosts are ignored, the behaviour is controlled by the new
215 host_all_ignored option.
217 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
218 homing or because of multiple MX records with /mx) generated more than
219 one IP address, and the following item turned out to be the local host,
220 all the secondary addresses of the first item were incorrectly removed
221 from the list, along with the local host and any following hosts (which
222 is what is supposed to happen).
224 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
225 whoever called Exim into the -H file. In the case of the daemon it was
226 behaving confusingly. When first started, it used values for whoever
227 started the daemon, but after a SIGHUP it used the Exim user (because it
228 calls itself on a restart). I have changed the code so that it now always
231 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
232 message are rejected with the same error (e.g. no authentication or bad
233 sender address), and a DATA command is nevertheless sent (as can happen
234 with PIPELINING or a stupid MUA), the error message that was given to the
235 RCPT commands is included in the rejection of the DATA command. This is
236 intended to be helpful for MUAs that show only the final error to their
239 PH/42 Another patch from the Sieve maintainer.
241 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
242 Eximstats - Fixed some broken HTML links and added missing column headers
244 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
245 columns for Rejects, Temp Rejects, Ham, and Spam rows.
247 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
249 PH/43 Yet another patch from the Sieve maintainer.
251 PH/44 I found a way to check for a TCP/IP connection going away before sending
252 the response to the final '.' that terminates a message, but only in the
253 case where the client has not sent further data following the '.'
254 (unfortunately, this is allowed). However, in many cases there won't be
255 any further data because there won't be any more messages to send. A call
256 to select() can be used: if it shows that the input is "ready", there is
257 either input waiting, or the socket has been closed. An attempt to read
258 the next input character can distinguish the two cases. Previously, Exim
259 would have sent an OK response which the client would never have see.
260 This could lead to message repetition. This fix should cure that, at
261 least in a lot of common cases.
263 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
264 advertised in response to EHLO.
270 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
271 fixed by 4.65/MH/01 (is this a record?) are fixed:
273 (i) An empty string was always treated as zero by the numeric comparison
274 operators. This behaviour has been restored.
276 (ii) It is documented that the numeric comparison operators always treat
277 their arguments as decimal numbers. This was broken in that numbers
278 starting with 0 were being interpreted as octal.
280 While fixing these problems I realized that there was another issue that
281 hadn't been noticed. Values of message_size_limit (both the global option
282 and the transport option) were treated as octal if they started with 0.
283 The documentation was vague. These values are now always treated as
284 decimal, and I will make that clear in the documentation.
290 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
291 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
294 MH/01 Don't check that the operands of numeric comparison operators are
295 integers when their expansion is in "skipping" mode (fixes bug
296 introduced by 4.64-PH/07).
298 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
299 child addresses, Exim now panics and dies. Previously, because the count
300 is held in a short int, deliveries were likely to be lost. As such a
301 large number of recipients for a single message is ridiculous
302 (performance will be very, very poor), I have chosen to impose a limit
303 rather than extend the field.
309 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
310 leftover -K file (the existence of which was triggered by #402).
311 While we were at it, introduced process PID as part of the -K
312 filename. This should rule out race conditions when creating
315 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
316 processing considerably. Previous code took too long for large mails,
317 triggering a timeout which in turn triggers #401.
319 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
320 in the DK code in transports.c. sendfile() is not really portable,
321 hence the _LINUX specificness.
323 TF/01 In the add_headers option to the mail command in an Exim filter,
324 there was a bug that Exim would claim a syntax error in any
325 header after the first one which had an odd number of characters
328 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
329 callout verification, Exim cached a "reject" for the entire domain. This
330 is correct for most verifications, but it is not correct for a recipient
331 verification with use_sender or use_postmaster set, because in that case
332 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
333 case of MAIL FROM:<> rejection from other early rejections (e.g.
334 rejection of HELO). When verifying a recipient using a non-null MAIL
335 address, the cache is ignored if it shows MAIL FROM:<> rejection.
336 Whatever the result of the callout, the value of the domain cache is
337 left unchanged (for any other kind of callout, getting as far as trying
338 RCPT means that the domain itself is ok).
340 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
343 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
344 manifest itself as EPIPE rather than ECONNECT. When tidying away a
345 session, the daemon ignores ECONNECT errors and logs others; it now
346 ignores EPIPE as well.
348 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
349 (quoted-printable decoding).
351 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
352 later the small subsequent patch to fix an introduced bug.
354 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
356 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
358 PH/08 An error is now given if message_size_limit is specified negative.
360 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
361 to be given (somewhat) arbitrary names.
363 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
366 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
367 miscellaneous code fixes
369 PH/10 Added the log_reject_target ACL modifier to specify where to log
372 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
373 hostname. This is wrong, because it relates to the incoming message (and
374 probably the interface on which it is arriving) and not to the outgoing
375 callout (which could be using a different interface). This has been
376 changed to use the value of the helo_data option from the smtp transport
377 instead - this is what is used when a message is actually being sent. If
378 there is no remote transport (possible with a router that sets up host
379 addresses), $smtp_active_hostname is used.
381 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
382 tweaks were necessary in order to get it to work (see also 21 below):
383 (a) The code assumed that strncpy() returns a negative number on buffer
384 overflow, which isn't the case. Replaced with Exim's string_format()
386 (b) There were several signed/unsigned issues. I just did the minimum
387 hacking in of casts. There is scope for a larger refactoring.
388 (c) The code used strcasecmp() which is not a standard C function.
389 Replaced with Exim's strcmpic() function.
390 (d) The code set only $1; it now sets $auth1 as well.
391 (e) A simple test gave the error "authentication client didn't specify
392 service in request". It would seem that Dovecot has changed its
393 interface. Fortunately there's a specification; I followed it and
394 changed what the client sends and it appears to be working now.
396 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
399 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
400 address A is aliased to B and C, where B exists and C does not. Without
401 -v the output is "A verified" because verification stops after a
402 successful redirection if more than one address is generated. However,
403 with -v the child addresses are also verified. Exim was outputting "A
404 failed to verify" and then showing the successful verification for C,
405 with its parentage. It now outputs "B failed to verify", showing B's
406 parentage before showing the successful verification of C.
408 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
409 look up a TXT record in a specific list after matching in a combined
412 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
413 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
414 they consult the DNS. I had assumed they would set it the way they
415 wanted; and indeed my experiments on Linux seem to show that in some
416 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
417 To be on the safe side, however, I have now made the interface to
418 host_find_byname() similar to host_find_bydns(), with an argument
419 containing the DNS resolver options. The host_find_byname() function now
420 sets these options at its start, just as host_find_bydns() does. The smtp
421 transport options dns_qualify_single and dns_search_parents are passed to
422 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
423 of host_find_byname() use the default settings of RES_DEFNAMES
424 (qualify_single) but not RES_DNSRCH (search_parents).
426 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
427 spool_read_header() do less string testing, by means of a preliminary
428 switch on the second character of optional "-foo" lines. (This is
429 overdue, caused by the large number of possibilities that now exist.
430 Originally there were few.) While I was there, I also converted the
431 str(n)cmp tests so they don't re-test the leading "-" and the first
432 character, in the hope this might squeeze out yet more improvement.
434 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
435 flag allowing group syntax was set by the header_syntax check but not
436 turned off, possible causing trouble later; (2) The flag was not being
437 set at all for the header_verify test, causing "group"-style headers to
438 be rejected. I have now set it in this case, and also caused header_
439 verify to ignore an empty address taken from a group. While doing this, I
440 came across some other cases where the code for allowing group syntax
441 while scanning a header line wasn't quite right (mostly, not resetting
442 the flag correctly in the right place). These bugs could have caused
443 trouble for malformed header lines. I hope it is now all correct.
445 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
446 with the "reply" argument non-NULL. The code, however (which originally
447 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
448 but it didn't always do it. This confused somebody who was copying the
449 code for some other use. I have removed all the tests.
451 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
452 feature that was used to support insecure browsers during the U.S. crypto
453 embargo. It requires special client support, and Exim is probably the
454 only MTA that supported it -- and would never use it because real RSA is
455 always available. This code has been removed, because it had the bad
456 effect of slowing Exim down by computing (never used) parameters for the
457 RSA_EXPORT functionality.
459 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
460 authenticator to fail if there's a tab character in the incoming data
461 (there should never be unless someone is messing about, as it's supposed
462 to be base64-encoded). Also added, on Timo's advice, the "secured" option
463 if the connection is using TLS or if the remote IP is the same as the
464 local IP, and the "valid-client-cert option" if a client certificate has
467 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
468 authenticators. This can be used for authorization after authentication
469 succeeds. (In the case of plaintext, it servers for both authentication
472 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
473 if any retry times were supplied.
475 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
476 connection, where there is no client IP address to check. In this
477 situation, the verify now always succeeds.
479 PH/25 Applied John Jetmore's -Mset patch.
481 PH/26 Added -bem to be like -Mset, but loading a message from a file.
483 PH/27 In a string expansion for a processed (not raw) header when multiple
484 headers of the same name were present, leading whitespace was being
485 removed from all of them, but trailing whitespace was being removed only
486 from the last one. Now trailing whitespace is removed from each header
487 before concatenation. Completely empty headers in a concatenation (as
490 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
491 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
493 PH/29 [Removed. This was a change that I later backed out, and forgot to
494 correct the ChangeLog entry (that I had efficiently created) before
495 committing the later change.]
497 PH/30 Exim was sometimes attempting to deliver messages that had suffered
498 address errors (4xx response to RCPT) over the same connection as other
499 messages routed to the same hosts. Such deliveries are always "forced",
500 so retry times are not inspected. This resulted in far too many retries
501 for the affected addresses. The effect occurred only when there were more
502 hosts than the hosts_max_try setting in the smtp transport when it had
503 the 4xx errors. Those hosts that it had tried were not added to the list
504 of hosts for which the message was waiting, so if all were tried, there
505 was no problem. Two fixes have been applied:
507 (i) If there are any address or message errors in an SMTP delivery, none
508 of the hosts (tried or untried) are now added to the list of hosts
509 for which the message is waiting, so the message should not be a
510 candidate for sending over the same connection that was used for a
511 successful delivery of some other message. This seems entirely
512 reasonable: after all the message is NOT "waiting for some host".
513 This is so "obvious" that I'm not sure why it wasn't done
514 previously. Hope I haven't missed anything, but it can't do any
515 harm, as the worst effect is to miss an optimization.
517 (ii) If, despite (i), such a delivery is accidentally attempted, the
518 routing retry time is respected, so at least it doesn't keep
519 hammering the server.
521 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
522 in ${readsocket because some servers need this prod.
524 PH/32 Added some extra debug output when updating a wait-xxx database.
526 PH/33 The hint "could be header name not terminated by colon", which has been
527 given for certain expansion errors for a long time, was not being given
528 for the ${if def:h_colon_omitted{... case.
530 PH/34 The spec says: "With one important exception, whenever a domain list is
531 being scanned, $domain contains the subject domain." There was at least
532 one case where this was not true.
534 PH/35 The error "getsockname() failed: connection reset by peer" was being
535 written to the panic log as well as the main log, but it isn't really
536 panic-worthy as it just means the connection died rather early on. I have
537 removed the panic log writing for the ECONNRESET error when getsockname()
540 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
541 runs only) independently of the message's sender address. This meant
542 that, if the 4xx error was in fact related to the sender, a different
543 message to the same recipient with a different sender could confuse
544 things. In particualar, this can happen when sending to a greylisting
545 server, but other circumstances could also provoke similar problems.
546 I have changed the default so that the retry time for these errors is now
547 based a combination of the sender and recipient addresses. This change
548 can be overridden by setting address_retry_include_sender=false in the
551 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
552 remote server are returned as part of bounce messages. This was not
553 happening for LMTP over a pipe (the lmtp transport), but now it is the
554 same for both kinds of LMTP.
556 PH/38 Despite being documented as not happening, Exim was rewriting addresses
557 in header lines that were in fact CNAMEs. This is no longer the case.
559 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
560 and queue runs started by the daemon processed all messages. This has
561 been fixed so that -R and -S can now usefully be given with -q<time>.
563 PH/40 Import PCRE release 6.7 (fixes some bugs).
565 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
567 PH/42 Give an error if -q is specified more than once.
569 PH/43 Renamed the variables $interface_address and $interface_port as
570 $received_ip_address and $received_port, to make it clear that these
571 values apply to message reception, and not to the outgoing interface when
572 a message is delivered. (The old names remain recognized, of course.)
574 PH/44 There was no timeout on the connect() call when using a Unix domain
575 socket in the ${readsocket expansion. There now is.
577 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
578 be meaningful with "accept".
580 SC/01 Eximstats V1.43
581 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
583 SC/02 Eximstats V1.44
584 Use a glob alias rather than an array ref in the generated
585 parser. This improves both readability and performance.
587 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
588 Collect SpamAssassin and rejection statistics.
589 Don't display local sender or destination tables unless
590 there is data to show.
591 Added average volumes into the top table text output.
593 SC/04 Eximstats V1.46
594 Collect data on the number of addresses (recipients)
595 as well as the number of messages.
597 SC/05 Eximstats V1.47
598 Added 'Message too big' to the list of mail rejection
599 reasons (thanks to Marco Gaiarin).
601 SC/06 Eximstats V1.48
602 Mainlog lines which have GMT offsets and are too short to
603 have a flag are now skipped.
605 SC/07 Eximstats V1.49 (Alain Williams)
606 Added the -emptyok flag.
608 SC/08 Eximstats V1.50
609 Fixes for obtaining the IP address from reject messages.
611 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
612 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
613 whitesspace changes from 4.64-PH/27
615 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
618 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
619 are found, allow negative numbers in numeric criteria)
621 JJ/06 exipick.20061117.2, added new $message_body_missing variable
623 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
624 to match changes made in 4.64-PH/43
626 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
628 PH/47 Put in an explicit test for a DNS lookup of an address record where the
629 "domain" is actually an IP address, and force a failure. This locks out
630 those revolvers/nameservers that support "A-for-A" lookups, in
631 contravention of the specifications.
633 PH/48 When a host name was looked up from an IP address, and the subsequent
634 forward lookup of the name timed out, the host name was left in
635 $sender_host_name, contrary to the specification.
637 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
638 restricted to single-key lookups, Exim was not diagnosing an error if
639 * or *@ was used with a query-style lookup.
641 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
643 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
644 long ago, but noone interested enough thought of it. Let's just say that
645 the "1.1" means that there are some new functions that weren't there at
646 some point in the past.
648 PH/51 Error processing for expansion failure of helo_data from an smtp
649 transport during callout processing was broken.
651 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
652 tested/used via the -bh/-bhc/-bs options.
654 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
655 bug, fixed in subsequent PCRE releases).
657 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
658 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
660 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
666 SC/01 Use a glob alias rather than an array ref in eximstats generated
667 parser. This improves both readability and performance.
669 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
670 Don't display local sender or destination tables in eximstats unless
671 there is data to show.
672 Added average volumes into the eximstats top table text output.
674 SC/03 Collect data on the number of addresses (recipients) as well
675 as the number of messages in eximstats.
677 TF/01 Correct an error in the documentation for the redirect router. Exim
678 does (usually) call initgroups() when daemonizing.
680 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
681 with consistent privilege compared to when running as a daemon.
683 TF/03 Note in the spec that $authenticated_id is not set for local
684 submissions from trusted users.
686 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
687 Thanks to Dean Brooks <dean@iglou.com> for the patch.
689 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
690 by adding some example configuration directives to the default
691 configuration file. A little bit of work is required to uncomment the
692 directives and define how usernames and passwords are checked, but
693 there is now a framework to start from.
695 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
696 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
697 without this. I don't know how relevant this is to other LDAP libraries.
699 PH/02 Add the verb name to the "unknown ACL verb" error.
701 PH/03 Magnus Holmgren's patch for filter_prepend_home.
703 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
705 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
706 directory not expanded when it should be if an expanded home directory
707 was set for the address (which is overridden by the transport).
709 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
712 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
713 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
714 because it is too late at that time, and has no effect.
716 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
717 security issue with \' (bugzilla #107). I could not use the
718 PQescapeStringConn() function, because it needs a PGconn value as one of
721 PH/08 When testing addresses using -bt, indicate those final addresses that
722 are duplicates that would not cause an additional delivery. At least one
723 person was confused, thinking that -bt output corresponded to deliveries.
724 (Suppressing duplicates isn't a good idea as you lose the information
725 about possibly different redirections that led to the duplicates.)
727 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
728 systems where poll() doesn't work, in particular OS X.
730 PH/10 Added more information to debugging output for retry time not reached.
732 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
733 operations in malware.c.
735 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
738 PH/13 If write_rejectlog was set false when logging was sent to syslog with
739 syslog_duplication set false, log lines that would normally be written
740 both the the main log and to the reject log were not written to syslog at
743 PH/14 In the default configuration, change the use of "message" in ACL warn
744 statements to "add_header".
746 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
747 not followed by a command (e.g. "seen endif").
749 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
750 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
753 PH/17 Added extra conditions to the default value of delay_warning_condition
757 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
758 { match{$h_precedence:}{(?i)bulk|list|junk} } \
759 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
762 The Auto-Submitted: and various List- headers are standardised, whereas I
763 don't think Precedence: ever was.
765 PH/18 Refactored debugging code in route_finduser() to show more information,
766 in particular, the error code if getpwnam() issues one.
768 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
769 This is apparently needed in addition to the PH/07 change above to avoid
770 any possible encoding problems.
772 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
773 but not after initializing Perl.
775 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
776 output them only if debugging. By default they are written stderr,
777 apparently, which is not desirable.
779 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
782 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
785 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
787 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
788 authenticated or an ident call has been made. Suppress the default
789 values for $authenticated_id and $authenticated_sender (but permit -oMai
790 and -oMas) when testing with -bh.
792 PH/24 Re-jigged the order of the tests in the default configuration so that the
793 tests for valid domains and recipients precede the DNS black list and CSA
794 tests, on the grounds that those ones are more expensive.
796 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
797 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
798 This bug exists in every version of Exim that I still have, right back to
801 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
802 However, an attempt to turn on case-sensitivity in a regex key by
803 including (?-i) didn't work because the subject string was already
804 lowercased, and the effects were non-intuitive. It turns out that a
805 one-line patch can be used to allow (?-i) to work as expected.
811 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
812 other effects) broke the use of negated acl sub-conditions.
814 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
817 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
818 "Deny" causes Exim to reject the incoming connection with a 554 error.
819 Unfortunately, if there is a major crisis, such as a disk failure,
820 tcp-wrappers gives "deny", whereas what one would like would be some
821 kind of temporary error. A kludge has been added to help with this.
822 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
823 554 error is used if errno is still zero or contains ENOENT (which occurs
824 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
827 PH/03 Add -lutil to the default FreeBSD LIBS setting.
829 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
830 errors. Otherwise a message that provokes a temporary error (when other
831 messages do not) can cause a whole host to time out.
833 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
834 addresses were routed directly to files or pipes from a redirect router.
835 File deliveries just didn't batch; pipe deliveries might have suffered
838 PH/06 A failure to get a lock for a hints database would erroneously always say
839 "Failed to get write lock", even when it was really a read lock.
841 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
842 of 0600. This has been changed to use the value of the lockfile_mode
843 option (which defaults to 0600).
845 PH/08 Applied small patch from the Sieve maintainer.
847 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
848 folder from quota calculations, a direct delivery into this folder messed
849 up the contents of the maildirsize file. This was because the regex was
850 used only to exclude .Trash (or whatever) when the size of the mailbox
851 was calculated. There was no check that a delivery was happening into an
852 excluded directory. This bug has been fixed by ignoring all quota
853 processing for deliveries into excluded directories.
855 PH/10 Added the maildirfolder_create_regex option to appendfile.
861 PH/01 The code for finding all the local interface addresses on a FreeBSD
862 system running IPv6 was broken. This may well have applied to all BSD
863 systems, as well as to others that have similar system calls. The broken
864 code found IPv4 interfaces correctly, but gave incorrect values for the
865 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
866 that it would not match correctly against @[] and not recognize the IPv6
869 PH/02 The ipliteral router was not recognizing addresses of the form user@
870 [ipv6:....] because it didn't know about the "ipv6:" prefix.
872 PH/03 Added disable_ipv6.
874 PH/04 Changed $reply_address to use the raw form of the headers instead of the
875 decoded form, because it is most often used to construct To: headers
876 lines in autoreplies, and the decoded form may well be syntactically
877 invalid. However, $reply_address has leading white space removed, and all
878 newlines turned into spaces so that the autoreply transport does not
881 PH/05 If group was specified without a user on a router, and no group or user
882 was specified on a transport, the group from the router was ignored.
884 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
885 for visible compile-time settings that can be used to change these
886 numbers, for those that want even more. Backwards compatibility with old
887 spool files has been maintained. However, going back to a previous Exim
888 release will lost any variables that are in spool files.
890 PH/07 Two small changes when running in the test harness: increase delay when
891 passing a TCP/IP connection to a new process, in case the original
892 process has to generate a bounce, and remove special handling of
893 127.0.0.2 (sic), which is no longer necessary.
895 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
896 be the same on different OS.
898 PH/09 Moved a debug statement in filter processing to avoid a race problem when
901 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
902 whether --show-vars was specified or not
904 JJ/02 exipick: Added support for new ACL variable spool format introduced
907 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
908 syntactically invalid From: or Reply-to: line, and a filter used this to
909 generate an autoreply, and therefore failed to obtain an address for the
910 autoreply, Exim could try to deliver to a non-existent relative file
911 name, causing unrelated and misleading errors. What now happens is that
912 it logs this as a hard delivery error, but does not attempt to create a
915 PH/11 The exinext utility has a -C option for testing purposes, but although
916 the given file was scanned by exinext itself; it wasn't being passed on
917 when Exim was called.
919 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
920 an end-of-file indication when reading a command response.
922 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
923 compiled. In many other places in Exim, IPv6 addresses are always
924 recognized, so I have changed this. It also means that IPv4 domain
925 literals of the form [IPV4:n.n.n.n] are now always recognized.
927 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
928 used if the router is not running as root, for example, when verifying at
929 ACL time, or when using -bh. The debugging output from this situation was
930 non-existent - all you got was a failure to exec. I have made two
933 (a) Failures to set uid/gid, the current directory, or a process leader
934 in a subprocess such as that created by queryprogram now generate
935 suitable debugging ouput when -d is set.
937 (b) The queryprogram router detects when it is not running as root,
938 outputs suitable debugging information if -d is set, and then runs
939 the subprocess without attempting to change uid/gid.
941 PH/15 Minor change to Makefile for building test_host (undocumented testing
944 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
945 additional section of a DNS packet that returns MX or SRV records.
946 Instead, it always explicitly searches for A/AAAA records. This avoids
947 major problems that occur when a DNS server includes only records of one
948 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
949 fixed another bug: if SRV records were looked up and the corresponding
950 address records were *not* found in the additional section, the port
951 values from the SRV records were lost.
953 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
954 using the correct key (the original address) when searching the retry
955 rules in order to find which one to use for generating the retry hint.
957 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
958 adding the default one. Similarly, if it contains a Reply-To: header, the
959 errors_reply_to option, if set, is not used.
961 PH/19 When calculating a retry time, Exim used to measure the "time since
962 failure" by looking at the "first failed" field in the retry record. Now
963 it does not use this if it is later than than the arrival time of the
964 message. Instead it uses the arrival time. This makes for better
965 behaviour in cases where some deliveries succeed, thus re-setting the
966 "first failed" field. An example is a quota failure for a huge message
967 when small messages continue to be delivered. Without this change, the
968 "time since failure" will always be short, possible causing more frequent
969 delivery attempts for the huge message than are intended.
970 [Note: This change was subsequently modified - see PH/04 for 4.62.]
972 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
973 $1, $2, $3) because the numerical variables can be reset during some
974 expansion items (e.g. "match"), thereby losing the authentication data.
976 PH/21 Make -bV show the size of off_t variables so that the test suite can
977 decide whether to run tests for quotas > 2G.
979 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
980 mailbox_size, and mailbox_filecount in the appendfile transport. If a
981 filecount value is greater than 2G or if a quota value is greater than 2G
982 on a system where the size of off_t is not greater than 4, a panic error
985 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
986 never match. The debug and -bh output now contains an explicit error
987 message indicating a malformed IPv4 address or mask.
989 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
990 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
993 PH/25 Do not write to syslog when running in the test harness. The only
994 occasion when this arises is a failure to open the main or panic logs
995 (for which there is an explicit test).
997 PH/26 Added the /no_tell option to "control=freeze".
999 PH/27 If a host name lookup failed very early in a connection, for example, if
1000 the IP address matched host_lookup and the reverse lookup yielded a name
1001 that did not have a forward lookup, an error message of the form "no IP
1002 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1003 could be logged. Now it outputs the IP address instead of "NULL".
1005 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1006 allows the sender and the authenticated sender to be set when
1007 submitting a message from within Exim. Since child_open_exim() is
1008 documented for local_scan(), the new function should be too.
1010 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1011 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1012 results in an empty string is now treated as unset.
1014 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1016 PH/31 Added sender_verify_fail logging option.
1018 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1019 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1020 tidied the source and removed it altogether.
1022 PH/33 When a queue run was abandoned because the load average was too high, a
1023 log line was always written; now it is written only if the queue_run log
1024 selector is set. In addition, the log line for abandonment now contains
1025 information about the queue run such as the pid. This is always present
1026 in "start" and "stop" lines but was omitted from the "abandon" line.
1028 PH/34 Omit spaces between a header name and the colon in the error message that
1029 is given when verify = headers_syntax fails (if there are lots of them,
1030 the message gets confusing).
1032 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1033 names, as there are now some PTR records that contain slashes. This check
1034 is only to protect against broken name servers that fall over on strange
1035 characters, so the fact that it applies to all lookups doesn't matter.
1037 PH/36 Now that the new test suite is complete, we can remove some of the
1038 special code in Exim that was needed for the old test suite. For example,
1039 sorting DNS records because real resolvers return them in an arbitrary
1040 order. The new test suite's fake resolver always returns records in the
1043 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1044 bounces) except when queue_only is set, to avoid logging races between
1045 the different processes.
1047 PH/38 Panic-die if .include specifies a non-absolute path.
1049 PH/39 A tweak to the "H" retry rule from its user.
1051 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1052 a label. They prevented compilation on older perls.
1054 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1055 a warning to be raised on newish perls.
1057 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1058 on queue. Changes to match documented behaviour of showing count of
1059 messages matching specified criteria.
1061 PH/40 Changed the default ident timeout from 30s to 5s.
1063 PH/41 Added support for the use of login_cap features, on those BSD systems
1064 that have them, for controlling the resources used by pipe deliveries.
1066 PH/42 The content-scanning code uses fopen() to create files in which to put
1067 message data. Previously it was not paying any attention to the mode of
1068 the files. Exim runs with umask(0) because the rest of the code creates
1069 files with open(), and sets the required mode explicitly. Thus, these
1070 files were ending up world-writeable. This was not a big issue, because,
1071 being within the spool directory, they were not world-accessible. I have
1072 created a function called modefopen, which takes an additional mode
1073 argument. It sets umask(777), creates the file, chmods it to the required
1074 mode, then resets the umask. All the relevant calls to fopen() in the
1075 content scanning code have been changed to use this function.
1077 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1078 to 24 hours. This avoids potential overflow problems when processing G
1079 and H retry rules. I suspect nobody ever tinkers with this value.
1081 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1083 PH/45 When the plaintext authenticator is running as a client, the server's
1084 challenges are checked to ensure they are valid base64 strings. By
1085 default, the authentication attempt is cancelled if an invalid string is
1086 received. Setting client_ignore_invalid_base64 true ignores these errors.
1087 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1088 they are received. Thus, the responses can be made to depend on the
1089 challenges. If an invalid string is ignored, an empty string is placed in
1092 PH/46 Messages that are created by the autoreply transport now contains a
1093 References: header, in accordance with RFCs 2822 and 3834.
1095 PH/47 Added authenticated_sender_force to the smtp transport.
1097 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1099 PH/49 Installed latest patch from the Sieve maintainer.
1101 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1102 was also set, the appendfile transport was unnecessarily scanning a
1103 directory of message files (e.g. for maildir delivery) to find the count
1104 of files (along with the size), even though it did not need this
1105 information. It now does the scan only if it needs to find either the
1106 size of the count of files.
1108 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1110 PH/52 Two bugs concerned with error handling when the smtp transport is
1113 (i) Exim was not creating retry information for temporary errors given
1114 for individual recipients after the DATA command when the smtp transport
1115 was used in LMTP mode. This meant that they could be retried too
1116 frequently, and not timed out correctly.
1118 (ii) Exim was setting the flag that allows error details to be returned
1119 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1120 recipients that were returned after the DATA command.
1122 PH/53 This is related to PH/52, but is more general: for any failing address,
1123 when detailed error information was permitted to be returned to the
1124 sender, but the error was temporary, then after the final timeout, only
1125 "retry timeout exceeded" was returned. Now it returns the full error as
1126 well as "retry timeout exceeded".
1128 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1129 do this, and (what is worse) MTAs that accept it.
1131 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1132 will now be deprecated.
1134 PH/56 New os.c-cygwin from the Cygwin maintainer.
1136 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1137 formats (previously only available in exim formats via -bpr, -bpru,
1138 and -bpra. Now also available in native and exiqgrep formats)
1140 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1141 with very large, slow to parse queues
1143 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1145 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1147 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1148 responses to authentication challenges, though it was showing the
1149 challenges; (ii) I've removed the CR characters from the debug output for
1152 PH/58 Allow for the insertion of a newline as well as a space when a string
1153 is turned into more than one encoded-word during RFC 2047 encoding. The
1154 Sieve code now uses this.
1156 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1157 data_4xx, lost_connection, tls_required.
1159 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1160 message was being sent as an SMTP response.
1162 PH/61 Add -l and -k options to exicyclog.
1164 PH/62 When verifying, if an address was redirected to one new address, so that
1165 verification continued, and the new address failed or deferred after
1166 having set something in $address_data, the value of $address_data was not
1167 passed back to the ACL. This was different to the case when no
1168 redirection occurred. The value is now passed back in both cases.
1170 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1171 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1172 use login_cap.h, so on its own it isn't the distinguishing feature. The
1173 new name refers directly to the setclassresources() function.
1175 PH/65 Added configuration files for NetBSD3.
1177 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1179 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1180 is preferred over IPv4.
1182 PH/68 The bounce_return_message and bounce_return_body options were not being
1183 honoured for bounces generated during the reception of non-SMTP messages.
1184 In particular, this applied to messages rejected by the ACL. This bug has
1185 been fixed. However, if bounce_return_message is true and bounce_return_
1186 body is false, the headers that are returned for a non-SMTP message
1187 include only those that have been read before the error was detected.
1188 (In the case of an ACL rejection, they have all been read.)
1190 PH/69 The HTML version of the specification is now built in a directory called
1191 spec_html instead of spec.html, because the latter looks like a path with
1192 a MIME-type, and this confuses some software.
1194 PH/70 Catch two compiler warnings in sieve.c.
1196 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1197 function verify_get_ident() calls ip_connect() to connect a socket, but
1198 if the "connect()" function timed out, ip_connect() used to close the
1199 socket. However, verify_get_ident() also closes the socket later, and in
1200 between Exim writes to the log, which may get opened at this point. When
1201 the socket was closed in ip_connect(), the log could get the same file
1202 descriptor number as the socket. This naturally causes chaos. The fix is
1203 not to close the socket in ip_connect(); the socket should be closed by
1204 the function that creates it. There was only one place in the code where
1205 this was missing, in the iplookup router, which I don't think anybody now
1206 uses, but I've fixed it anyway.
1208 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1209 well as to direct DNS lookups. Otherwise the handling of names in host
1210 lists is inconsistent and therefore confusing.
1216 PH/01 Two changes to the default runtime configuration:
1218 (1) Move the checks for relay_from_hosts and authenticated clients from
1219 after to before the (commented out) DNS black list checks.
1221 (2) Add control=submission to the relay_from_hosts and authenticated
1222 clients checks, on the grounds that messages accepted by these
1223 statements are most likely to be submissions.
1225 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1227 (1) Generate an error if the third argument for the ${prvs expansion is
1230 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1233 (3) Reset the variables that are obtained from the first argument of
1234 ${prvscheck and used in the second argument before leaving the code,
1235 because their memory is reclaimed, so using them afterwards may do
1238 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1239 one (it's much easier than Tom thought :-).
1241 (5) Because of (4), we can now allow for the use of $prvscheck_result
1242 inside the third argument.
1244 PH/03 For some reason, the default setting of PATH when running a command from
1245 a pipe transport was just "/usr/bin". I have changed it to
1248 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1249 anything to be listed in the output from -bV.
1251 PH/05 When a filter generated an autoreply, the entire To: header line was
1252 quoted in the delivery log line, like this:
1254 => >A.N.Other <ano@some.domain> <original@ddress> ...
1256 This has been changed so that it extracts the operative address. There
1257 may be more than one such address. If so, they are comma-separated, like
1260 => >ano@some.domain,ona@other.domain <original@ddress> ...
1262 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1263 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1264 not being looked up in the rDNS to get a host name, Exim was showing the
1265 IP address twice in Received: lines, even though the IP addresses were
1266 identical. For example:
1268 Received: from [1.2.3.4] (helo=[1.2.3.4])
1270 However, if the real host name was known, it was omitting the HELO data
1271 if it matched the actual IP address. This has been tidied up so that it
1272 doesn't show the same IP address twice.
1274 PH/07 When both +timestamp and +memory debugging was on, the value given by
1275 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1276 called by the debug printing, thereby overwriting the timestamp buffer.
1277 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1279 PH/08 When the original message was included in an autoreply transport, it
1280 always said "this is a copy of the message, including all the headers",
1281 even if body_only or headers_only was set. It now gives an appropriate
1284 PH/09 Applied a patch from the Sieve maintainer which:
1286 o fixes some comments
1287 o adds the (disabled) notify extension core
1288 o adds some debug output for the result of if/elsif tests
1289 o points to the current vacation draft in the documentation
1290 and documents the missing references header update
1294 o fixes a bug in processing the envelope test (when testing
1295 multiple envelope elements, the last element determinted the
1298 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1299 Electronic Mail") by including:
1301 Auto-submitted: auto-generated
1303 in the messages that it generates (bounce messages and others, such as
1304 warnings). In the case of bounce messages for non-SMTP mesages, there was
1305 also a typo: it was using "Auto_submitted" (underscore instead of
1306 hyphen). Since every message generated by Exim is necessarily in response
1307 to another message, thes have all been changed to:
1309 Auto-Submitted: auto-replied
1311 in accordance with these statements in the RFC:
1313 The auto-replied keyword:
1315 - SHOULD be used on messages sent in direct response to another
1316 message by an automatic process,
1318 - MUST NOT be used on manually-generated messages,
1320 - MAY be used on Delivery Status Notifications (DSNs) and Message
1321 Disposition Notifications (MDNs),
1323 - MUST NOT be used on messages generated by automatic or periodic
1324 processes, except for messages which are automatic responses to
1327 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1328 to the default Received: header definition.
1330 PH/12 Added log selector acl_warn_skipped (default on).
1332 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1333 variables because (a) they are in the wrong storage pool and (b) even if
1334 they were copied, it wouldn't work properly because of the caching.
1336 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1337 checking when decoding. Apparently there are clients that generate
1338 overlong encoded strings. Why am I not surprised?
1340 PH/15 If the first argument of "${if match_address" was not empty, but did not
1341 contain an "@" character, Exim crashed. Now it writes a panic log message
1342 and treats the condition as false.
1344 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1346 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1347 extension "envelope-auth". The code is finished and in agreement with
1348 other implementations, but there is no documentation so far and in fact,
1349 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1350 not changing the active code.
1352 Print executed "if" and "elsif" statements when debugging is used. This
1353 helps a great deal to understand what a filter does.
1355 Document more things not specified clearly in RFC3028. I had all this
1356 sorted out, when out of a sudden new issues came to my mind. Oops."
1358 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1361 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1362 canonical form (as documented). However, after a host name lookup from
1363 the IP address, check_host() was doing a simple string comparison with
1364 addresses acquired from the DNS when checking that the found name did
1365 have the original IP as one of its addresses. Since any found IPv6
1366 addresses are likely to be in abbreviated form, the comparison could
1367 fail. Luckily, there already exists a function for doing the comparison
1368 by converting both addresses to binary, so now that is used instead of
1369 the text comparison.
1371 PH/20 There was another similar case to PH/19, when a complete host name was
1372 given in a host list; looking up its IP address could give an abbreviated
1373 form, whereas the current host's name might or might not be abbreviated.
1374 The same fix has been applied.
1380 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1381 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1384 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1385 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1387 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1389 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1390 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1391 Also, for "Auto-Submitted", treat anything other than "no" as
1392 non-personal, in accordance with RFC 3834. (Previously it treated
1393 anything starting "auto-" as non-personal.)
1395 TF/01 The control=submission/name=... option had a problem with syntax
1396 errors if the name included a slash character. The /name= option
1397 now slurps the rest of the string, so it can include any characters
1398 but it must come last in the list of options (after /sender_retain
1401 PH/05 Some modifications to the interface to the fake nameserver for the new
1409 TK/01 Added the "success_on_redirect" address verification option. See
1410 NewStuff for rationale and an example.
1412 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1414 PH/02 Patch to exigrep to allow it to work on syslog lines.
1416 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1417 fread() to skip over the body file's header line, because in Cygwin the
1418 header line is locked and is inaccessible.
1420 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1421 co-exist for some time) to make it clear that it is the Exim ID that is
1422 referenced, not the Message-ID: header line.
1424 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1425 string_format() function, because snprintf() does not exist on all
1428 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1429 ${stat: expansion item.
1431 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1432 protocol synchronization error", to keep the pedants happy.
1434 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1435 well as for IRIX systems, when gcc is being used. See the host.c source
1438 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1440 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1443 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1444 transport and to the smtp transport in LMTP mode.
1446 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1448 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1449 run a filter in a subprocess. This could lead to confusion in subsequent
1450 lookups in the parent process. There should also be a search_tidyup() at
1451 the end of the subprocess.
1453 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1454 only if the host matched helo_try_verify_hosts, which caused the
1455 verification to occur when the EHLO/HELO command was issued. The ACL just
1456 tested the remembered result. Now, if a previous verification attempt has
1457 not happened, "verify = helo" does it there and then.
1459 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1461 TK/03 Fix log output including CR from clamd.
1463 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1464 exist provoked a memory error which could cause a segfault.
1466 PH/15 Installed PCRE 6.2
1468 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1470 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1471 of the problem. Specifically, suggested +O2 rather than +O1 for the
1474 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1476 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1477 record, the port was not correctly logged when the outgoing_port log
1478 selector was set (it logged the transort's default port).
1480 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1481 fallback_hosts, and "hosts" in the smtp transport.
1483 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1484 host errors such as "Connection refused".
1486 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1487 authentication with radiusclient 0.4.9:
1489 - Error returned from rc_read_config was caught wrongly
1490 - Username/password not passed on to radius server due to wrong length.
1492 The presumption is that some radiusclient API changes for 4.51/PH/17
1493 were not taken care of correctly. The code is still untested by me (my
1494 Linux distribution still has 0.3.2 of radiusclient), but it was
1495 contributed by a Radius user.
1497 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1498 expanding the "port" option of the smtp transport.
1500 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1501 while reading a MIME header. Thanks to Tom Hughes for a patch.
1503 PH/24 Include config.h inside local_scan.h so that configuration settings are
1506 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1507 that in an ACL for RCPT (for example), you can examine exactly what was
1510 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1511 commands, but it was not correctly comparing the address with the actual
1512 client host address. Thus, it would show the EHLO address in Received:
1513 header lines when this was not necessary.
1515 PH/27 Added the % operator to ${eval:}.
1517 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1518 it should be ignoring failures (because with -C, for example, it has lost
1519 privilege). It wasn't ignoring creation failures other than "already
1522 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1525 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1526 because an input line was too long, either on its own, or by virtue of
1527 too many continuations, the temporary file was not being removed, and the
1528 return code was incorrect.
1530 PH/31 Missing "BOOL" in function definition in filtertest.c.
1532 PH/32 Applied Sieve patches from the maintainer.
1534 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1536 PH/33 Added "verify = not_blind".
1538 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1539 Local/Makefile (with some defaults set). These are used in built scripts
1540 such as exicyclog, but they have never been used in the exim_install
1541 script (though there are many overriding facilities there). I have
1542 arranged that the exim_install script now takes note of these two
1545 PH/35 Installed configuration files for Dragonfly.
1547 PH/36 When a locally submitted message by a trusted user did not contain a
1548 From: header, and the sender address was obtained from -f or from an SMTP
1549 MAIL command, and the trusted user did not use -F to supply a sender
1550 name, $originator_name was incorrectly used when constructing a From:
1551 header. Furthermore, $originator_name was used for submission mode
1552 messages from external hosts without From: headers in a similar way,
1553 which is clearly wrong.
1555 PH/37 Added control=suppress_local_fixups.
1557 PH/38 When log_selector = +received_sender was set, and the addition of the
1558 sender made the log line's construction buffer exactly full, or one byte
1559 less than full, an overflow happened when the terminating "\n" was
1562 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1563 when the result of a list match is failure because a DNS lookup failed.
1565 PH/40 RM_COMMAND is now used in the building process.
1567 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1568 the "build-* directories that it finds.
1570 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1571 address in a domain literal was a prefix of an interface address.
1573 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1574 when verifying a sender address, unless rewrite_headers is false.
1576 PH/44 Wrote a long comment about why errors_to addresses are verified as
1577 recipients, not senders.
1579 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1580 the ratelimit ACL was added.
1582 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1584 PH/46 Added extra information about PostgreSQL errors to the error string.
1586 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1587 avoiding the need to install special zones in a real server. This is
1588 backwards compatible; if it can't find the fake resolver, it drops back.
1589 Thus, both old and new test suites can be run.
1591 TF/02 Added util/ratelimit.pl
1593 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1594 clock is set back in time.
1596 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1597 Candler <B.Candler@pobox.com>.
1599 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1600 OK for addresses that are the result of redirections.
1602 PH/48 A number of further additions for the benefit of the new test suite,
1603 including a fake gethostbyname() that interfaces to the fake DNS resolver
1606 TF/06 The fix for widen_domains has also been applied to qualify_single and
1607 search_parents which are the other dnslookup options that can cause
1610 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1613 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1615 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1616 service" if it encountered an MX record with an empty target hostname.
1617 The message is now "an MX or SRV record indicated no SMTP service".
1619 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1620 if the DNS of the sending site is misconfigured. This is quite a
1621 common situation. This change restores the behaviour of treating a
1622 helo verification defer as a failure.
1624 PH/51 If self=fail was set on a router, the bounce message did not include the
1625 actual error message.
1631 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1633 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1634 command itself ended in error, the underlying message about the transport
1635 filter timeout was being overwritten with the pipe command error. Now the
1636 underlying error message should be appended to the second error message.
1638 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1640 PH/02 Reduce the amount of output that "make" produces by default. Full output
1641 can still be requested.
1643 PH/03 The warning log line about a condition test deferring for a "warn" verb
1644 was being output only once per connection, rather than after each
1645 occurrence (because it was using the same function as for successful
1646 "warn" verbs). This seems wrong, so I have changed it.
1648 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1649 it should not have, which might have caused a crash in the right
1650 circumstances, but probably never did.
1652 PH/04 Installed a modified version of Tony Finch's patch to make submission
1653 mode fix the return path as well as the Sender: header line, and to
1654 add a /name= option so that you can make the user's friendly name appear
1657 TF/03 Added the control = fakedefer ACL modifier.
1659 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1660 Mark Lowes for thorough testing.
1662 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1664 TK/03 Merged latest SRS patch from Miles Wilton.
1666 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1667 with the definition in sysexits.h (which is #included earlier).
1668 Fortunately, Exim does not actually use EX_OK. The code used to try to
1669 preserve the sysexits.h value, by assumimg that macro definitions were
1670 scanned for macro replacements. I have been disabused of this notion,
1671 so now the code just undefines EX_OK before #including unistd.h.
1673 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1674 in the smtp transport. When a block could not be written in a single
1675 write() function, the timeout was being re-applied to each part-write.
1676 This seems wrong - if the receiver was accepting one byte at a time it
1677 would take for ever. The timeout is now adjusted when this happens. It
1678 doesn't have to be particularly precise.
1680 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1681 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1683 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1684 without a domain if the check to <postmaster@domain> fails.
1686 SC/01 Eximstats: added -xls and the ability to specify output files
1687 (patch written by Frank Heydlauf).
1689 SC/02 Eximstats: use FileHandles for outputing results.
1691 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1693 SC/04 Eximstats: fixed display of large numbers with -nvr option
1695 SC/05 Eximstats: fixed merging of reports with empty tables.
1697 SC/06 Eximstats: added the -include_original_destination flag
1699 SC/07 Eximstats: removed tabs and trailing whitespace.
1701 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1703 TK/06 MBOX spool code: Add real "From " MBOX separator line
1704 so the .eml file is really in mbox format (even though
1705 most programs do not really care). Patch from Alex Miller.
1707 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1708 The latter is generated from $received_to and is only set if the
1709 message has one envelope recipient. SA can use these headers,
1710 obviously out-of-the-box. Patch from Alex Miller.
1712 PH/08 The ${def test on a variable was returning false if the variable's
1713 value was "0", contrary to what the specification has always said!
1714 The result should be true unless the variable is empty.
1716 PH/09 The syntax error of a character other than { following "${if
1717 def:variable_name" (after optional whitespace) was not being diagnosed.
1718 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1719 accidental colon was present, for example, could give incorrect results.
1721 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1722 result is used (not including appendfile, where other changes are about
1725 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1726 This involved changing a lot of size variables from int to off_t. It
1727 should work with maildirs and everything.
1729 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1730 spamd dying while we are connected to it.
1732 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1735 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1736 function for BDB 4.3.
1738 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1740 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1741 the sources for PCRE have been re-arranged and more files are now
1744 PH/15 The code I had for printing potentially long long variables in PH/11
1745 above was not the best (it lost precision). The length of off_t variables
1746 is now inspected at build time, and an appropriate printing format (%ld
1747 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1748 to be "long long int" or "long int". This is needed for the internal
1749 formatting function string_vformat().
1751 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1752 the configuration file to be ":syslog", then the script "guesses" where
1753 the logs files are, rather than using the compiled in default. In our
1754 case the guess is not the same as the compiled default, so the script
1755 suddenly stopped working when I started to use syslog. The patch checks
1756 to see if log_file_path is "". If so, it attempts to read it from exim
1757 with no configuration file to get the compiled in version, before it
1758 falls back to the previous guessing code."
1760 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1761 implementing BATV in an Exim configuration. See NewStuff for the gory
1764 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1765 Makefile that are specific to HP-UX.
1767 PH/18 If the "use_postmaster" option was set for a recipient callout together
1768 with the "random" option, the postmaster address was used as the MAIL
1769 FROM address for the random test, but not for the subsequent recipient
1770 test. It is now used for both.
1772 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1773 patch removes a few documentation additions to RFC 3028, because the
1774 latest draft now contains them. It adds the new en;ascii-case comparator
1775 and a new error check for 8bit text in MIME parts. Comparator and
1776 require names are now matched exactly. I enabled the subaddress
1777 extension, but it is not well tested yet (read: it works for me)."
1779 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1780 rework some of the code of TK/09 above to avoid the hardwired use of
1781 "%lld" and "long long". Replaced the call to snprintf() with a call to
1784 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1785 records point to non-existent hosts", "retry timeout exceeded", and
1786 "retry time not reached for any host after a long failure period".
1788 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1789 experimental DomainKeys support:
1791 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1792 (2) On an error such as an illegally used "control", the wrong name for
1793 the control was given.
1795 These problems did NOT occur unless DomainKeys support was compiled.
1797 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1799 PH/24 Added ${if match_ip condition.
1801 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1802 databases so that it will be absolutely obvious if a crash occurs in the
1803 DB library. This is a regular occurrence (often caused by mis-matched
1806 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1807 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1808 picked up on a user's system that detects such things. There doesn't seem
1809 to be a gcc warning option for this - only an attribute that has to be
1810 put on the function's prototype. It seems that in Fedora Core 4 they have
1811 set this on a number of new functions. No doubt there will be more in due
1814 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1815 specify a transport. However, if an address that was verified by such a
1816 router was the subject of a callout, Exim crashed because it tried to
1817 read the rcpt_include_affixes from the non-existent transport. Now it
1818 just assumes that the setting of that option is false. This bug was
1819 introduced by 4.51/PH/31.
1821 PH/28 Changed -d+all to exclude +memory, because that information is very
1822 rarely of interest, but it makes the output a lot bigger. People tend to
1823 do -d+all out of habit.
1825 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1826 code in os-type was giving problems when libc.so lives in lib64, like on
1829 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1830 aren't the modern standard, and it seems that some systems' include files
1831 don't always have them. Exim was already checking for some of the newer
1832 ones like T_AAAA, and defining it itself. I've added checks for all the
1833 record types that Exim uses.
1835 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1836 not automatically generating a new one, as it is supposed to. This
1837 prevented TLS from working. If the file did exist, but contained invalid
1838 data, a new version was generated, as expected. It was only the case of a
1839 non-existent file that was broken.
1841 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1842 with a change in libdomainkeys > 0.64.
1844 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1845 from DNS. If the selector record carries the flag, it now has
1846 precedence over the domain-wide flag.
1848 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1850 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1851 the use of a port name that isn't defined in /etc/services) occurred, the
1852 message was deferred as in a normal delivery, and thus remained on the
1853 spool, instead of being failed because of the mua_wrapper setting. This
1854 is now fixed, and I tidied up some of the mua_wrapper messages at the
1857 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1858 the messages in a hash of arrays rather than using individual hashes.
1859 This is a bit cleaner and results in dramatic memory savings, albeit
1860 at a slight CPU cost.
1862 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1863 as requested by Marc Sherman.
1865 SC/10 Eximstats: added histograms for user specified patterns as requested
1868 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1870 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1871 fopen() in the content-scanning modules that did not already have it.
1877 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1878 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1880 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1882 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1884 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1885 to test Sieve filters that use "vacation".
1887 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1888 that changes the way the GnuTLS parameters are stored in the cache file.
1889 The new format can be generated externally. For backward compatibility,
1890 if the data in the cache doesn't make sense, Exim assumes it has read an
1891 old-format file, and it generates new data and writes a new file. This
1892 means that you can't go back to an older release without removing the
1895 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1896 work if there are any delivery delays because "one_time" forces the
1897 parent to be marked "delivered", so its unseen clone is never tried
1898 again. For this reason, Exim now forbids the simultaneous setting of
1901 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1902 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1903 entry if you want to know the details. The fix, however, neglected to
1904 consider the case when local delivery batching is involved. The test for
1905 "previously delivered" was not happening when checking to see if an
1906 address could be batched with a previous (undelivered) one; under
1907 certain circumstances this could lead to multiple deliveries to the same
1910 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1911 in its include files, and this causes problems building Exim.
1913 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1914 header_syntax) but Exim was just ignoring anything given after a slash.
1915 In particular, this caused confusion with an attempt to use "verify =
1916 reverse_host_lookup/defer_ok". An error is now given when options are
1917 supplied for verify items that do not have them. (Maybe reverse_host_
1918 lookup should have a defer_ok option, but that's a different point.)
1920 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1921 defined by RFC 821) to 2048, because there were problems with some AUTH
1922 commands, and RFC 1869 says the size should be increased for extended
1923 SMTP commands that take arguments.
1925 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1928 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1929 "unknown" error; now it says that the functionality isn't in the binary.
1931 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1932 an address' error message when a string expansion fails (syntax or
1933 whatever). Otherwise the password may appear in the log. Following change
1934 PH/42 below, there is no longer a chance of it appearing in a bounce
1937 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1939 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1940 of its addresses was ever tried. (Bugzilla bug #2.)
1942 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1943 the result incorrectly in the debug output. (It correctly added a newline
1944 to what was transported.)
1946 TF/01 Added $received_time.
1948 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1949 commented out examples of how to interface to a virus scanner and to
1950 SpamAssassin. Also added commented examples of av_scanner and
1951 spamd_address settings.
1953 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1954 and controls are allowed in which ACLs. There were a couple of minor
1955 errors. Some of the entries in the conditions table (which is a table of
1956 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1957 negation of where the condition IS allowed.
1959 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1961 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1962 header file does not have a version number, so I've had to invent a new
1963 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1964 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1965 radiusclient), but it was contributed by a Radius user.
1967 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1968 files or directories, for OpenSSL.
1970 PH/19 When an Exim process that is running as root has to create an Exim log
1971 file, it does so in a subprocess that runs as exim:exim so as to get the
1972 ownership right at creation (otherwise, other Exim processes might see
1973 the file with the wrong ownership). There was no test for failure of this
1974 fork() call, which would lead to the process getting stuck as it waited
1975 for a non-existent subprocess. Forks do occasionally fail when resources
1976 run out. I reviewed all the other calls to fork(); they all seem to check
1979 PH/20 When checking for unexpected SMTP input at connect time (before writing
1980 the banner), Exim was not dealing correctly with a non-positive return
1981 from the read() function. If the client had disconnected by this time,
1982 the result was a log entry for a synchronization error with an empty
1983 string after "input=" when read() returned zero. If read() returned -1
1984 (an event I could not check), uninitialized data bytes were printed.
1985 There were reports of junk text (parts of files, etc) appearing after
1988 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
1990 PH/22 Added support for macro redefinition, and (re)definition in between
1991 driver and ACL definitions.
1993 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
1994 forgetting to use the resulting value; it was using the unexpanded value.
1996 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
1997 hadn't been configured. The fix is from Juergen Kreileder, who
1998 understands it better than I do:
2000 "Here's what I see happening with three configured cyrus_sasl
2001 authenticators configured (plain, login, cram-md5):
2003 On startup auth_cyrus_sasl_init() gets called for each of these.
2004 This means three calls to sasl_listmech() without a specified mech_list.
2005 => SASL tests which mechs of all available mechs actually work
2006 => three warnings about OTP not working
2007 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2009 With the patch, sasl_listmech() also gets called three times. But now
2010 SASL's mech_list option is set to the server_mech specified in the the
2011 authenticator. Or in other words, the answer from sasl_listmech()
2012 gets limited to just the mech you're testing for (which is different
2014 => the return list contains just 'plain' or 'login', 'cram-md5' or
2015 nothing depending on the value of ob->server_mech.
2017 I've just tested the patch: Authentication still works fine,
2018 unavailable mechs specified in the exim configuration are still
2019 caught, and the auth.log warnings about OTP are gone."
2021 PH/25 When debugging is enabled, the contents of the command line are added
2022 to the debugging output, even when log_selector=+arguments is not
2025 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2026 answer is "GNU", and only if the return is "GNU/something" is the answer
2029 PH/27 $acl_verify_message is now set immediately after the failure of a
2030 verification in an ACL, and so is available in subsequent modifiers. In
2031 particular, the message can be preserved by coding like this:
2033 warn !verify = sender
2034 set acl_m0 = $acl_verify_message
2036 Previously, $acl_verify_message was set only while expanding "message"
2037 and "log_message" when a very denied access.
2039 PH/28 Modified OS/os.c-Linux with
2041 -#ifndef OS_LOAD_AVERAGE
2042 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2044 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2045 nomenclature these days.)
2047 PH/29 Installed patch from the Sieve maintainer that adds the options
2048 sieve_useraddress and sieve_subaddress to the redirect router.
2050 PH/30 In these circumstances:
2051 . Two addresses routed to the same list of hosts;
2052 . First host does not offer TLS;
2053 . First host accepts first address;
2054 . First host gives temporary error to second address;
2055 . Second host offers TLS and a TLS session is established;
2056 . Second host accepts second address.
2057 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2058 and peerdn, if requested) that were in fact used only for the second
2061 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2062 attention to any local part prefix or suffix that was matched by the
2063 router that accepted the address. It now behaves in the same way as it
2064 does for delivery: the affixes are removed from the local part unless
2065 rcpt_include_affixes is set on the transport.
2067 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2068 timeout during the DATA phase of an incoming message.
2070 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2071 applied a patch sanctioned by the Sieve maintainer.
2073 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2074 the uid or gid is negative. A case of a negative gid caused this to be
2075 noticed. The fix allows for either to be negative.
2077 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2078 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2081 PH/36 The change PH/12 above was broken. Fixed it.
2083 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2084 the grounds that routing the same address twice would always produce the
2085 same answer. This might have been true once, but it is certainly no
2086 longer true now. Routing a child address may depend on the previous
2087 routing that produced that child. Some complicated redirection strategies
2088 went wrong when messages had multiple recipients, and made Exim's
2089 behaviour dependent on the order in which the addresses were given.
2091 I have moved the duplicate checking until after the routing is complete.
2092 Exim scans the addresses that are assigned to local and remote
2093 transports, and removes any duplicates. This means that more work will be
2094 done, as duplicates will always all be routed, but duplicates are
2095 presumably rare, so I don't expect this is of any significance.
2097 For deliveries to pipes, files, and autoreplies, the duplicate checking
2098 still happens during the routing process, since they are not going to be
2101 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2102 It corrects a timeout issue with spamd. This is Ian's comment: "The
2103 background is that sometimes spamd either never reads data from a
2104 connection it has accepted, or it never writes response data. The exiscan
2105 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2106 blindly assumes that writes won't block so it may never time out."
2108 PH/39 Allow G after quota size as well as K and M.
2110 PH/40 The value set for $authenticated_id in an authenticator may not contain
2111 binary zeroes or newlines because the value is written to log lines and
2112 to spool files. There was no check on this. Now the value is run through
2113 the string_printing() function so that such characters are converted to
2114 printable escape sequences.
2116 PH/41 $message_linecount is a new variable that contains the total number of
2117 lines in the message. Compare $body_linecount, which is the count for the
2120 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2121 bounce and delay warning messages, except in certain special cases, which
2124 (a) An SMTP error message from a remote host;
2125 (b) A message specified in a :fail: redirection;
2126 (c) A message specified in a "fail" command in a system filter;
2127 (d) A message specified in a FAIL return from the queryprogram router;
2128 (e) A message specified by the cannot_route_message router option.
2130 In these cases only, Exim does include the error details in bounce and
2131 warning messages. There are also a few cases where bland messages such
2132 as "unrouteable address" or "local delivery error" are given.
2134 PH/43 $value is now also set for the "else" part of a ${run expansion.
2136 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2137 being worked on, but at least Exim now implements the latest version to
2140 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2141 process to complete was treated as a delivery failure, a timeout while
2142 writing the message to the pipe was logged, but erroneously treated as a
2143 successful delivery. Such timeouts include transport filter timeouts. For
2144 consistency with the overall process timeout, these timeouts are now
2145 treated as errors, giving rise to delivery failures by default. However,
2146 there is now a new Boolean option for the pipe transport called
2147 timeout_defer, which, if set TRUE, converts the failures into defers for
2148 both kinds of timeout. A transport filter timeout is now identified in
2151 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2152 systems where "make" and "gmake" are different, calling "gmake" at top
2153 level broke things. I've arranged for the value of $(MAKE) to be passed
2154 from the Makefile to this script so that it can call the same version of
2158 A note about Exim versions 4.44 and 4.50
2159 ----------------------------------------
2161 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2162 changes of various kinds. As a consequence, a big documentation update was
2163 needed. This delayed the release for rather longer than seemed good, especially
2164 in the light of a couple of (minor) security issues. Therefore, the changes
2165 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2166 release. So 4.44 and 4.50 are in effect two different branches that both start
2169 I have left the 4.50 change log unchanged; it contains all the changes since
2170 4.43. The change log for 4.44 is below; many of its items are identical to
2171 those for 4.50. This seems to be the most sensible way to preserve the
2172 historical information.
2178 1. Minor wording change to the doc/README.SIEVE file.
2180 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2181 computation of the current number of files was incorrect.
2183 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2184 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2187 4. Give more explanation in the error message when the command for a transport
2188 filter fails to execute.
2190 5. There are several places where Exim runs a non-Exim command in a
2191 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2192 was being done only for the command run by the queryprogram router. It is
2193 now done for all such subprocesses. The other cases are: ${run, transport
2194 filters, and the commands run by the lmtp and pipe transports.
2196 6. Added CONFIGURE_GROUP build-time option.
2198 7. Some older OS have a limit of 256 on the maximum number of file
2199 descriptors. Exim was using setrlimit() to set 1000 as a large value
2200 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2201 systems. I've change it so that if it can't get 1000, it tries for 256.
2203 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2204 was an oversight, and furthermore, ever since the addition of extra
2205 controls (e.g. 4.43/32), the checks on when to allow different forms of
2206 "control" were broken. There should now be diagnostics for all cases when a
2207 control that does not make sense is encountered.
2209 9. Added the /retain_sender option to "control=submission".
2211 10. $recipients is now available in the predata ACL (oversight).
2213 11. Tidy the search cache before the fork to do a delivery from a message
2214 received from the command line. Otherwise the child will trigger a lookup
2215 failure and thereby defer the delivery if it tries to use (for example) a
2216 cached ldap connection that the parent has called unbind on.
2218 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2219 of $address_data from the recipient verification was clobbered by the
2220 sender verification.
2222 13. The value of address_data from a sender verification is now available in
2223 $sender_address_data in subsequent conditions in the ACL statement.
2225 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2227 15. Added a new option "connect=<time>" to callout options, to set a different
2230 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2231 was its contents. (It was OK if the option was not defined at all.)
2233 17. A "Completed" log line is now written for messages that are removed from
2234 the spool by the -Mrm option.
2236 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2237 information about exactly what failed.
2239 19. Added -dd to debug only the daemon process.
2241 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2242 handles timeouts, both on the server side and network timeouts. Renamed the
2243 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2245 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2246 It is now set to "smtps".
2248 22. $host_address is now set to the target address during the checking of
2249 ignore_target_hosts.
2251 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2252 being passed; this would have caused $sender_host_name to have been used if
2253 matching the list had actually called for a host name (not very likely,
2254 since this list is usually IP addresses). A host name is now passed as
2257 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2258 code that specifies a non-restarting handler (typically sigaction() in
2259 modern systems) in an attempt to fix a rare and obscure crash bug.
2261 25. Narrowed the window for a race in the daemon that could cause it to ignore
2262 SIGCHLD signals. This is not a major problem, because they are used only to
2263 wake it up if nothing else does.
2265 26. A malformed maildirsize file could cause Exim to calculate negative values
2266 for the mailbox size or file count. Odd effects could occur as a result.
2267 The maildirsize information is now recalculated if the size or filecount
2270 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2271 support for a long time. Removed HAVE_SYS_VFS_H.
2273 28. Installed the latest version of exipick from John Jetmore.
2275 29. In an address list, if the pattern was not a regular expression, an empty
2276 subject address (from a bounce message) matched only if the pattern was an
2277 empty string. Non-empty patterns were not even tested. This was the wrong
2278 because it is perfectly reasonable to use an empty address as part of a
2279 database query. An empty address is now tested by patterns that are
2280 lookups. However, all the other forms of pattern expect the subject to
2281 contain a local part and a domain, and therefore, for them, an empty
2282 address still always fails if the pattern is not itself empty.
2284 30. Exim went into a mad DNS loop when attempting to do a callout where the
2285 host was specified on an smtp transport, and looking it up yielded more
2286 than one IP address.
2288 31. Re-factored the code for checking spool and log partition space into a
2289 function that finds that data and another that does the check. The former
2290 is then used to implement four new variables: $spool_space, $log_space,
2291 $spool_inodes, and $log_inodes.
2293 32. The RFC2047 encoding function was originally intended for short strings
2294 such as real names; it was not keeping to the 75-character limit for
2295 encoded words that the RFC imposes. It now respects the limit, and
2296 generates multiple encoded words if necessary. To be on the safe side, I
2297 have increased the buffer size for the ${rfc2047: expansion operator from
2300 33. It is now permitted to omit both strings after an "if" condition; if the
2301 condition is true, the result is "true". As before, when the second string
2302 is omitted, a false condition yields an empty string. This makes it less
2303 cumbersome to write custom ACL and router conditions.
2305 34. Failure to deliver a bounce message always caused it to be frozen, even if
2306 there was an errors_to setting on the router. The errors_to setting is now
2309 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2310 canonical form (fully expanded) before being placed in
2311 $sender_host_address.
2313 36. The table in the code that translates DNS record types into text (T_A to
2314 "A" for instance) was missing entries for NS and CNAME. It is just possible
2315 that this could have caused confusion if both these types were looked up
2316 for the same domain, because the text type is used as part of Exim's
2317 per-process caching. But the chance of anyone hitting this buglet seems
2320 37. The dnsdb lookup has been extended in a number of ways.
2322 (1) There is a new type, "zns", which walks up the domain tree until it
2323 finds some nameserver records. It should be used with care.
2325 (2) There is a new type, "mxh", which is like "mx" except that it returns
2326 just the host names, not the priorities.
2328 (3) It is now possible to give a list of domains (or IP addresses) to be
2329 looked up. The behaviour when one of the lookups defers can be
2330 controlled by a keyword.
2332 (4) It is now possible to specify the separator character for use when
2333 multiple records are returned.
2335 38. The dnslists ACL condition has been extended: it is now possible to supply
2336 a list of IP addresses and/or domains to be looked up in a particular DNS
2339 39. Added log_selector=+queue_time_overall.
2341 40. When running the queue in the test harness, wait just a tad after forking a
2342 delivery process, to get repeatability of debugging output.
2344 41. Include certificate and key file names in error message when GnuTLS fails
2345 to set them up, because the GnuTLS error message doesn't include the name
2346 of the failing file when there is a problem reading it.
2348 42. Allow both -bf and -bF in the same test run.
2350 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2352 44. The "Exiscan patch" is now merged into the mainline Exim source.
2354 45. Sometimes the final signoff response after QUIT could fail to get
2355 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2356 < 0 before doing a fflush(). This bug looks as though it goes back to the
2357 introduction of TLS in release 3.20, but "sometimes" must have been rare
2358 because the tests only now provoked it.
2360 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2361 (this can affect the format of dates).
2363 47. exim_tidydb, when checking for the continued existence of a message for
2364 which it has found a message-specific retry record, was not finding
2365 messages that were in split spool directories. Consequently, it was
2366 deleting retry records that should have stayed in existence.
2368 48. Steve fixed some bugs in eximstats.
2370 49. The SPA authentication driver was not abandoning authentication and moving
2371 on to the next authenticator when an expansion was forced to fail,
2372 contradicting the general specification for all authenticators. Instead it
2373 was generating a temporary error. It now behaves as specified.
2375 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2376 (the order specifies the preference for clients). The order is now AES256,
2377 AES128, 3DES, ARCFOUR128.
2379 51. Small patch to Sieve code - explicitly set From: when generating an
2382 52. Exim crashed if a remote delivery caused a very long error message to be
2383 recorded - for instance if somebody sent an entire SpamAssassin report back
2384 as a large number of 550 error lines. This bug was coincidentally fixed by
2385 increasing the size of one of Exim's internal buffers (big_buffer) that
2386 happened as part of the Exiscan merge. However, to be on the safe side, I
2387 have made the code more robust (and fixed the comments that describe what
2390 53. Now that there can be additional text after "Completed" in log lines (if
2391 the queue_time_overall log selector is set), a one-byte patch to exigrep
2392 was needed to allow it to recognize "Completed" as not the last thing in
2395 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2396 patch that reportedly fixes this has been added. I am not expert enough to
2397 create a test for it. This is what the patch creator wrote:
2399 "I found a little strange behaviour of ldap code when working with
2400 Windows 2003 AD Domain, where users was placed in more than one
2401 Organization Units. When I tried to give exim partial DN, the exit code
2402 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2403 But simultaneously result of request was absolutely normal ldap result,
2404 so I produce this patch..."
2406 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2407 REFERENCE, so I have modified the code to exclude the patch when that macro
2410 55. Some experimental protocols are using DNS PTR records for new purposes. The
2411 keys for these records are domain names, not reversed IP addresses. The
2412 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2413 leaves it alone. Component reversal etc. now happens only for IP addresses.
2416 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2418 57. Double the size of the debug message buffer (to 2048) so that more of very
2419 long debug lines gets shown.
2421 58. The exicyclog utility now does better if the number of log files to keep
2422 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2424 59. Two changes related to the smtp_active_hostname option:
2426 (1) $smtp_active_hostname is now available as a variable.
2427 (2) The default for smtp_banner uses $smtp_active_hostname instead
2428 of $primary_hostname.
2430 60. The host_aton() function is supposed to be passed a string that is known
2431 to be a valid IP address. However, in the case of IPv6 addresses, it was
2432 not checking this. This is a hostage to fortune. Exim now panics and dies
2433 if the condition is not met. A case was found where this could be provoked
2434 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2435 components; fortuitously, this particular loophole had already been fixed
2436 by change 4.50/55 above.
2438 If there are any other similar loopholes, the new check in host_aton()
2439 itself should stop them being exploited. The report I received stated that
2440 data on the command line could provoke the exploit when Exim was running as
2441 exim, but did not say which command line option was involved. All I could
2442 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2443 running as the user.
2446 61. There was a buffer overflow vulnerability in the SPA authentication code
2447 (which came originally from the Samba project). I have added a test to the
2448 spa_base64_to_bits() function which I hope fixes it.
2451 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2452 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2454 63. The daemon start-up calls getloadavg() while still root for those OS that
2455 need the first call to be done as root, but it missed one case: when
2456 deliver_queue_load_max is set with deliver_drop_privilege. This is
2457 necessary for the benefit of the queue runner, because there is no re-exec
2458 when deliver_drop_privilege is set.
2460 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2461 This has been fixed.
2463 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2464 was in use, was not putting the data itself into the right store pool;
2465 consequently, it could be overwritten for a subsequent message in the same
2466 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2469 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2471 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2472 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2473 the function were treating the return as a boolean value, which happened to
2474 work because 0=false and not-0=true, but is not correct code.
2476 68. The host_aton() function was not handling scoped IPv6 addresses (those
2477 with, for example, "%eth0" on the end) correctly.
2479 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2480 negated items (that is, ~something) in unsigned ints. Some compilers
2481 apparently mutter when there is no cast.
2483 70. If an address verification called from an ACL failed, and did not produce a
2484 user-specific message (i.e. there was only a "system" message), nothing was
2485 put in $acl_verify_message. In this situation, it now puts the system
2488 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2489 session; change 4.31/43 added the unwanted input to the log line - except
2490 that it did not do this in the start of session case. It now does.
2492 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2493 This is wrong and can cause the other end to generate a synchronization
2494 error if it is another Exim or anything else that does the synchronization
2495 check. A QUIT command is no longer sent after a timeout.
2497 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2498 during host lookups.
2500 74. The defer_ok option of callout verification was not working if it was used
2501 when verifying addresses in header lines, that is, for this case:
2503 verify = header_sender/callout=defer_ok
2505 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2506 those file descriptors could be used for SMTP connections. If anything
2507 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2508 could be sent to the SMTP client, causing chaos. The daemon now opens
2509 stdin, stdout, and stderr to /dev/null when it puts itself into the
2512 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2513 log by default. The user can override this with suitable Perl magic.
2515 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2516 the log message when discard triggers, was not working for the DATA ACL or
2517 for the non-SMTP ACL.
2519 78. Error message wording change in sieve.c.
2521 79. If smtp_accept_max_per_host was set, the number of connections could be
2522 restricted to fewer than expected, because the daemon was trying to set up
2523 a new connection before checking whether the processes handling previous
2524 connections had finished. The check for completed processes is now done
2525 earlier. On busy systems, this bug wouldn't be noticed because something
2526 else would have woken the daemon, and it would have reaped the completed
2529 80. If a message was submitted locally by a user whose login name contained one
2530 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2531 It caused a spool format error. I have fixed the spool reading code. A
2532 related problem was that the "from" clause in the Received: line became
2533 illegal because of the space(s). It is now covered by ${quote_local_part.
2535 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2538 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2540 83. Patch from Sieve maintainer to fix unterminated string problem in
2541 "vacation" handling.
2543 84. Some minor changes to the Linux configuration files to help with other
2544 OS variants using glibc.
2546 85. One more patch for Sieve to update vacation handling to latest spec.
2549 ----------------------------------------------------
2550 See the note above about the 4.44 and 4.50 releases.
2551 ----------------------------------------------------
2557 1. Change 4.43/35 introduced a bug that caused file counts to be
2558 incorrectly computed when quota_filecount was set in an appendfile
2561 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2562 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2565 3. Give more explanation in the error message when the command for a transport
2566 filter fails to execute.
2568 4. There are several places where Exim runs a non-Exim command in a
2569 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2570 was being done only for the command run by the queryprogram router. It is
2571 now done for all such subprocesses. The other cases are: ${run, transport
2572 filters, and the commands run by the lmtp and pipe transports.
2574 5. Some older OS have a limit of 256 on the maximum number of file
2575 descriptors. Exim was using setrlimit() to set 1000 as a large value
2576 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2577 systems. I've change it so that if it can't get 1000, it tries for 256.
2579 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2580 was an oversight, and furthermore, ever since the addition of extra
2581 controls (e.g. 4.43/32), the checks on when to allow different forms of
2582 "control" were broken. There should now be diagnostics for all cases when a
2583 control that does not make sense is encountered.
2585 7. $recipients is now available in the predata ACL (oversight).
2587 8. Tidy the search cache before the fork to do a delivery from a message
2588 received from the command line. Otherwise the child will trigger a lookup
2589 failure and thereby defer the delivery if it tries to use (for example) a
2590 cached ldap connection that the parent has called unbind on.
2592 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2593 of $address_data from the recipient verification was clobbered by the
2594 sender verification.
2596 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2597 was its contents. (It was OK if the option was not defined at all.)
2599 11. A "Completed" log line is now written for messages that are removed from
2600 the spool by the -Mrm option.
2602 12. $host_address is now set to the target address during the checking of
2603 ignore_target_hosts.
2605 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2606 being passed; this would have caused $sender_host_name to have been used if
2607 matching the list had actually called for a host name (not very likely,
2608 since this list is usually IP addresses). A host name is now passed as
2611 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2612 code that specifies a non-restarting handler (typically sigaction() in
2613 modern systems) in an attempt to fix a rare and obscure crash bug.
2615 15. Narrowed the window for a race in the daemon that could cause it to ignore
2616 SIGCHLD signals. This is not a major problem, because they are used only to
2617 wake it up if nothing else does.
2619 16. A malformed maildirsize file could cause Exim to calculate negative values
2620 for the mailbox size or file count. Odd effects could occur as a result.
2621 The maildirsize information is now recalculated if the size or filecount
2624 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2625 support for a long time. Removed HAVE_SYS_VFS_H.
2627 18. Updated exipick to current release from John Jetmore.
2629 19. Allow an empty sender to be matched against a lookup in an address list.
2630 Previously the only cases considered were a regular expression, or an
2633 20. Exim went into a mad DNS lookup loop when doing a callout where the
2634 host was specified on the transport, if the DNS lookup yielded more than
2637 21. The RFC2047 encoding function was originally intended for short strings
2638 such as real names; it was not keeping to the 75-character limit for
2639 encoded words that the RFC imposes. It now respects the limit, and
2640 generates multiple encoded words if necessary. To be on the safe side, I
2641 have increased the buffer size for the ${rfc2047: expansion operator from
2644 22. Failure to deliver a bounce message always caused it to be frozen, even if
2645 there was an errors_to setting on the router. The errors_to setting is now
2648 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2649 canonical form (fully expanded) before being placed in
2650 $sender_host_address.
2652 24. Updated eximstats to version 1.33
2654 25. Include certificate and key file names in error message when GnuTLS fails
2655 to set them up, because the GnuTLS error message doesn't include the name
2656 of the failing file when there is a problem reading it.
2658 26. Expand error message when OpenSSL has problems setting up cert/key files.
2661 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2662 (this can affect the format of dates).
2664 28. exim_tidydb, when checking for the continued existence of a message for
2665 which it has found a message-specific retry record, was not finding
2666 messages that were in split spool directories. Consequently, it was
2667 deleting retry records that should have stayed in existence.
2669 29. eximstats updated to version 1.35
2670 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2671 1.35 - bugfix such that pie charts by volume are generated correctly
2673 30. The SPA authentication driver was not abandoning authentication and moving
2674 on to the next authenticator when an expansion was forced to fail,
2675 contradicting the general specification for all authenticators. Instead it
2676 was generating a temporary error. It now behaves as specified.
2678 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2679 (the order specifies the preference for clients). The order is now AES256,
2680 AES128, 3DES, ARCFOUR128.
2682 31. Small patch to Sieve code - explicitly set From: when generating an
2685 32. Exim crashed if a remote delivery caused a very long error message to be
2686 recorded - for instance if somebody sent an entire SpamAssassin report back
2687 as a large number of 550 error lines. This bug was coincidentally fixed by
2688 increasing the size of one of Exim's internal buffers (big_buffer) that
2689 happened as part of the Exiscan merge. However, to be on the safe side, I
2690 have made the code more robust (and fixed the comments that describe what
2693 33. Some experimental protocols are using DNS PTR records for new purposes. The
2694 keys for these records are domain names, not reversed IP addresses. The
2695 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2696 leaves it alone. Component reversal etc. now happens only for IP addresses.
2699 34. The host_aton() function is supposed to be passed a string that is known
2700 to be a valid IP address. However, in the case of IPv6 addresses, it was
2701 not checking this. This is a hostage to fortune. Exim now panics and dies
2702 if the condition is not met. A case was found where this could be provoked
2703 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2704 components; fortuitously, this particular loophole had already been fixed
2705 by change 4.50/55 or 4.44/33 above.
2707 If there are any other similar loopholes, the new check in host_aton()
2708 itself should stop them being exploited. The report I received stated that
2709 data on the command line could provoke the exploit when Exim was running as
2710 exim, but did not say which command line option was involved. All I could
2711 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2712 running as the user.
2715 35. There was a buffer overflow vulnerability in the SPA authentication code
2716 (which came originally from the Samba project). I have added a test to the
2717 spa_base64_to_bits() function which I hope fixes it.
2720 36. The daemon start-up calls getloadavg() while still root for those OS that
2721 need the first call to be done as root, but it missed one case: when
2722 deliver_queue_load_max is set with deliver_drop_privilege. This is
2723 necessary for the benefit of the queue runner, because there is no re-exec
2724 when deliver_drop_privilege is set.
2726 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2727 was in use, was not putting the data itself into the right store pool;
2728 consequently, it could be overwritten for a subsequent message in the same
2729 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2732 38. Sometimes the final signoff response after QUIT could fail to get
2733 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2734 < 0 before doing a fflush(). This bug looks as though it goes back to the
2735 introduction of TLS in release 3.20, but "sometimes" must have been rare
2736 because the tests only now provoked it.
2742 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2743 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2744 Now it returns an integer. A number of calls were still expecting a T/F
2745 return. Fortuitously, in all cases, the tests worked in OK situations,
2746 which is the norm. However, things would have gone wrong on any write
2747 failures on the smtp file descriptor. This function is used when sending
2748 messages over SMTP and also when doing verify callouts.
2750 2. When Exim is called to do synchronous delivery of a locally submitted
2751 message (the -odf or -odi options), it no longer closes stderr before doing
2754 3. Implemented the mua_wrapper option.
2756 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2758 5. Implemented the functions header_remove(), header_testname(),
2759 header_add_at_position(), and receive_remove_recipient(), and exported them
2762 6. If an ACL "warn" statement specified the addition of headers, Exim already
2763 inserted X-ACL-Warn: at the start if there was no header name. However, it
2764 was not making this test for the second and subsequent header lines if
2765 there were newlines in the string. This meant that an invalid header could
2766 be inserted if Exim was badly configured.
2768 7. Allow an ACL "warn" statement to add header lines at the start or after all
2769 the Received: headers, as well as at the end.
2771 8. Added the rcpt_4xx retry error code.
2773 9. Added postmaster_mailfrom=xxx to callout verification option.
2775 10. Added mailfrom=xxxx to the callout verification option, for verify=
2778 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2779 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2781 12. Inserted some casts to stop certain compilers warning when using pointer
2782 differences as field lengths or precisions in printf-type calls (mostly
2783 affecting debugging statements).
2785 13. Added optional readline() support for -be (dynamically loaded).
2787 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2788 same clock tick as a message's arrival, so that its received time was the
2789 same as the "first fail" time on the retry record, and that message
2790 remained on the queue past the ultimate address timeout, every queue runner
2791 would try a delivery (because it was past the ultimate address timeout) but
2792 after another failure, the ultimate address timeout, which should have then
2793 bounced the address, did not kick in. This was a "< instead of <=" error;
2794 in most cases the first failure would have been in the next clock tick
2795 after the received time, and all would be well.
2797 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2798 being recognized when the domain list was tested by the match_domain
2799 condition in an expansion string.
2801 16. Added the ${str2b64: operator.
2803 17. Exim was always calling setrlimit() to set a large limit for the number of
2804 processes, without checking whether the existing limit was already
2805 adequate. (It did check for the limit on file descriptors.) Furthermore,
2806 errors from getrlimit() and setrlimit() were being ignored. Now they are
2807 logged to the main and panic logs, but Exim does carry on, to try to do its
2808 job under whatever limits there are.
2810 18. Imported PCRE 5.0.
2812 19. Trivial typo in log message " temporarily refused connection" (the leading
2815 20. If the log selector return_path_on_delivery was set and an address was
2816 redirected to /dev/null, the delivery process crashed because it assumed
2817 that a return path would always be set for a "successful" delivery. In this
2818 case, the whole delivery is bypassed as an optimization, and therefore no
2821 21. Internal re-arrangement: the function for sending a challenge and reading
2822 a response while authentication was assuming a zero-terminated challenge
2823 string. It's now changed to take a pointer and a length, to allow for
2824 binary data in such strings.
2826 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2828 23. Exim was not respecting finduser_retries when seeking the login of the
2829 uid under which it was called; it was always trying 10 times. (The default
2830 setting of finduser_retries is zero.) Also, it was sleeping after the final
2831 failure, which is pointless.
2833 24. Implemented tls_on_connect_ports.
2835 25. Implemented acl_smtp_predata.
2837 26. If the domain in control=submission is set empty, Exim assumes that the
2838 authenticated id is a complete email address when it generates From: or
2839 Sender: header lines.
2841 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2842 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2843 chown and chgrp in /bin and hostname in /usr/bin.
2845 28. Exim was keeping the "process log" file open after each use, just as it
2846 does for the main log. This opens the possibility of it remaining open for
2847 long periods when the USR1 signal hits a daemon. Occasional processlog
2848 errors were reported, that could have been caused by this. Anyway, it seems
2849 much more sensible not to leave this file open at all, so that is what now
2852 29. The long-running daemon process does not normally write to the log once it
2853 has entered its main loop, and it closes the log before doing so. This is
2854 so that log files can straightforwardly be renamed and moved. However,
2855 there are a couple of unusual error situations where the daemon does write
2856 log entries, and I had neglected to close the log afterwards.
2858 30. The text of an SMTP error response that was received during a remote
2859 delivery was being truncated at 512 bytes. This is too short for some of
2860 the long messages that one sometimes sees. I've increased the limit to
2863 31. It is now possible to make retry rules that apply only when a message has a
2864 specific sender, in particular, an empty sender.
2866 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2867 it possible to be selective about when SMTP synchronization is enforced.
2869 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2871 32. Implemented hosts_connection_nolog.
2873 33. Added an ACL for QUIT.
2875 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2878 35. Added mailbox_size and mailbox_filecount to appendfile.
2880 36. Added control = no_multiline_responses to ACLs.
2882 37. There was a bug in the logic of the code that waits for the clock to tick
2883 in the case where the clock went backwards by a substantial amount such
2884 that the microsecond fraction of "now" was more than the microsecond
2885 fraction of "then" (but the whole seconds number was less).
2887 38. Added support for the libradius Radius client library this is found on
2888 FreeBSD (previously only the radiusclient library was supported).
2894 1. When certain lookups returned multiple values in the form name=value, the
2895 quoting of the values was not always being done properly. Specifically:
2896 (a) If the value started with a double quote, but contained no whitespace,
2898 (b) If the value contained whitespace other than a space character (i.e.
2899 tabs or newlines or carriage returns) it was not quoted.
2900 This fix has been applied to the mysql and pgsql lookups by writing a
2901 separate quoting function and calling it from the lookup code. The fix
2902 should probably also be applied to nisplus, ibase and oracle lookups, but
2903 since I cannot test any of those, I have not disturbed their existing code.
2905 2. A hit in the callout cache for a specific address caused a log line with no
2906 reason for rejecting RCPT. Now it says "Previous (cached) callout
2907 verification failure".
2909 3. There was an off-by-one bug in the queryprogram router. An over-long
2910 return line was truncated at 256 instead of 255 characters, thereby
2911 overflowing its buffer with the terminating zero. As well as fixing this, I
2912 have increased the buffer size to 1024 (and made a note to document this).
2914 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2915 when Exim is waiting for an SMTP response from a remote server, Exim
2916 restarts its select() call on the socket, thereby resetting its timeout.
2917 This is not a problem when such interrupts are rare. Somebody set up a cron
2918 job to run exiwhat every 2 minutes, which is less than the normal select()
2919 timeout (5 or 10 minutes). This meant that the select() timeout never
2920 kicked in because it was always reset. I have fixed this by comparing the
2921 time when an interrupt arrives with the time at the start of the first call
2922 to select(). If more time than the timeout has elapsed, the interrupt is
2923 treated as a timeout.
2925 5. Some internal re-factoring in preparation for the addition of Sieve
2926 extensions (by MH). In particular, the "personal" test is moved to a
2927 separate function, and given an option for scanning Cc: and Bcc: (which is
2928 not set for Exim filters).
2930 6. When Exim created an email address using the login of the caller as the
2931 local part (e.g. when creating a From: or Sender: header line), it was not
2932 quoting the local part when it contained special characters such as @.
2934 7. Installed new OpenBSD configuration files.
2936 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2937 try to make them clearer.
2939 9. Callout options, other than the timeout value, were being ignored when
2940 verifying sender addresses in header lines. For example, when using
2942 verify = header_sender/callout=no_cache
2944 the cache was (incorrectly) being used.
2946 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2947 only the Cygwin environment.
2949 11. When return_path_on_delivery was set as a log selector, if different remote
2950 addresses in the same message used different return paths and parallel
2951 remote delivery occurred, the wrong values would sometimes be logged.
2952 (Whenever a remote delivery process finished, the return path value from
2953 the most recently started remote delivery process was logged.)
2955 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2956 lines when AUTH and/or TLS are in use. This is the "received protocol"
2957 field. Exim used to use "asmtp" for authenticated SMTP, without any
2958 indication (in the protocol name) for TLS use. Now it follows the RFC and
2959 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2960 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2961 names appear in log lines as well as in Received: header lines.
2963 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2964 extensions, and comparison tests, and to fix some bugs.
2966 14. Changes to the "personal" filter test:
2968 (1) The test was buggy in that it was just doing the equivalent of
2969 "contains" tests on header lines. For example, if a user's address was
2970 anne@some.where, the "personal" test would incorrectly be true for
2972 To: susanne@some.where
2974 This test is now done by extracting each address from the header in turn,
2975 and checking the entire address. Other tests that are part of "personal"
2976 are now done using regular expressions (for example, to check local parts
2977 of addresses in From: header lines).
2979 (2) The list of non-personal local parts in From: addresses has been
2980 extended to include "listserv", "majordomo", "*-request", and "owner-*",
2981 taken from the Sieve specification recommendations.
2983 (3) If the message contains any header line starting with "List-" it is
2984 treated as non-personal.
2986 (4) The test for "circular" in the Subject: header line has been removed
2987 because it now seems ill-conceived.
2989 15. Minor typos in src/EDITME comments corrected.
2991 16. Installed latest exipick from John Jetmore.
2993 17. If headers_add on a router specified a text string that was too long for
2994 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
2995 of string_sprintf() is now avoided.
2997 18. $message_body_size was not set (it was always zero) when running the DATA
2998 ACL and the local_scan() function.
3000 19. For the "mail" command in an Exim filter, no default was being set for
3001 the once_repeat time, causing a random time value to be used if "once" was
3002 specified. (If the value happened to be <= 0, no repeat happened.) The
3003 default is now 0s, meaning "never repeat". The "vacation" command was OK
3004 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3005 (I found it when inspecting the code).
3007 20. There is now an overall timeout for performing a callout verification. It
3008 defaults to 4 times the callout timeout, which applies to individual SMTP
3009 commands during the callout. The overall timeout applies when there is more
3010 than one host that can be tried. The timeout is checked before trying the
3011 next host. This prevents very long delays if there are a large number of
3012 hosts and all are timing out (e.g. when the network connections are timing
3013 out). The value of the overall timeout can be changed by specifying an
3014 additional sub-option for "callout", called "maxwait". For example:
3016 verify = sender/callout=5s,maxwait=20s
3018 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3019 to the end before writing, but this should make it even safer).
3021 22. Exim was forgetting that it had advertised PIPELINING for the second and
3022 subsequent messages on an SMTP connection. It was also not resetting its
3023 memory on STARTTLS and an internal HELO.
3025 23. When Exim logs an SMTP synchronization error within a session, it now
3026 records whether PIPELINING has been advertised or not.
3028 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3029 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3030 rather than long int.
3032 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3034 26. Added the never_mail option to autoreply.
3040 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3041 crash if the getsockname() call failed; this can happen if a connection is
3042 closed very soon after it is established. The problem was simply in the
3043 order in which certain operations were done, causing Exim to try to write
3044 to the SMTP stream before it had set up the file descriptor. The bug has
3045 been fixed by making things happen in the correct order.
3051 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3052 before the connection was closed, thus losing the rejection response.
3054 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3055 some early Solaris releases, but causes trouble in current releases where
3056 socklen_t is defined.
3058 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3061 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3064 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3065 that releasing the top of it at the end releases what was used for sub-
3066 expansions (unless the block got too big). However, discard this block if
3067 the first thing is a variable or header, so that we can use its block when
3068 it is dynamic (useful for very large $message_headers, for example).
3070 6. Lookups now cache *every* query, not just the most recent. A new, separate
3071 store pool is used for this. It can be recovered when all lookup caches are
3072 flushed. Lookups now release memory at the end of their result strings.
3073 This has involved some general refactoring of the lookup sources.
3075 7. Some code has been added to the store_xxx() functions to reduce the amount
3076 of flapping under certain conditions.
3078 8. log_incoming_interface used to affect only the <= reception log lines. Now
3079 it causes the local interface and port to be added to several more SMTP log
3080 lines, for example "SMTP connection from", and rejection lines.
3082 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3084 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3086 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3087 could be overwritten at the end of the current message (or the start of a
3088 new message if it was set in a HELO ACL). The value is now preserved for
3089 the duration of the SMTP connection.
3091 12. If a transport had a headers_rewrite setting, and a matching header line
3092 contained an unqualified address, that address was qualified, even if it
3093 did not match any rewriting rules. The underlying bug was that the values
3094 of the flags that permit the existence of unqualified sender and recipient
3095 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3096 non-local messages, and by -bnq for local messages) were not being
3097 preserved with the message after it was received.
3099 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3100 "next input=" as part of the text comprising the host identity instead of
3101 the correct text. The code was using the same buffer for two different
3102 strings. However, depending on which order the printing function evaluated
3103 its arguments, the bug did not always show up. Under Linux, for example, my
3104 test suite worked just fine.
3106 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3107 doesn't work with some older versions of Perl. It has been changed to "my",
3108 which in any case is probably the better facility to use.
3110 15. A really picky compiler found some instances of statements for creating
3111 error messages that either had too many or two few arguments for the format
3114 16. The size of the buffer for calls to the DNS resolver has been increased
3115 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3116 for addresses that have a lot of PTR records. This alleviates a problem; it
3117 does not fully solve it.
3119 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3120 buffer now truncates the list and logs the incident, which is the same
3121 action as happens when Exim is looking up a host name and its aliases.
3122 Previously in this situation something unpredictable would happen;
3123 sometimes it was "internal error: store_reset failed".
3125 18. If a server dropped the connection unexpectedly when an Exim client was
3126 using GnuTLS and trying to read a response, the client delivery process
3127 crashed while trying to generate an error log message.
3129 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3130 string, for example:
3132 warn message = H1: something\nH2: something
3134 the text was added as a single header line from Exim's point of view
3135 though it ended up OK in the delivered message. However, searching for the
3136 second and subsequent header lines using $h_h2: did not work. This has been
3137 fixed. Similarly, if a system filter added multiple headers in this way,
3138 the routers could not see them.
3140 20. Expanded the error message when iplsearch is called with an invalid key to
3141 suggest using net-iplsearch in a host list.
3143 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3144 ACLs are no longer actually imposed (and a message to that effect is
3147 22. If a "gecos" field in a passwd entry contained escaped characters, in
3148 particular, if it contained a \" sequence, Exim got it wrong when building
3149 a From: or a Sender: header from that name. A second bug also caused
3150 incorrect handling when an unquoted " was present following a character
3151 that needed quoting.
3153 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3154 was not being matched caselessly.
3156 24. Arranged for all hyphens in the exim.8 source to be escaped with
3159 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3160 properly. Recipient callout cache records were still being keyed to include
3161 the sender, even when use_sender was set false. This led to far more
3162 callouts that were necessary. The sender is no longer included in the key
3163 when use_sender is false.
3165 26. Added "control = submission" modifier to ACLs.
3167 27. Added the ${base62d: operator to decode base 62 numbers.
3169 28. dnsdb lookups can now access SRV records.
3171 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3172 the configuration file.
3174 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3175 (-v) mode. This makes the output for a verbose queue run more intelligible.
3177 31. Added a use_postmaster feature to recipient callouts.
3179 32. Added the $body_zerocount variable, containing the number of binary zero
3180 bytes in the message body.
3182 33. The time of last modification of the "new" subdirectory is now used as the
3183 "mailbox time last read" when there is a quota error for a maildir
3186 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3188 35. Added +ignore_unknown as a special item in host lists.
3190 36. Code for decoding IPv6 addresses in host lists is now included, even if
3191 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3192 address was recognized as an IP address, but was then not correctly decoded
3193 into binary, causing unexpected and incorrect effects when compared with
3200 1. Very minor rewording of debugging text in manualroute to say "list of
3201 hosts" instead of "hostlist".
3203 2. If verify=header_syntax was set, and a header line with an unqualified
3204 address (no domain) and a large number of spaces between the end of the
3205 name and the colon was received, the reception process suffered a buffer
3206 overflow, and (when I tested it) crashed. This was caused by some obsolete
3207 code that should have been removed. The fix is to remove it!
3209 3. When running in the test harness, delay a bit after writing a bounce
3210 message to get a bit more predictability in the log output.
3212 4. Added a call to search_tidyup() just before forking a reception process. In
3213 theory, someone could use a lookup in the expansion of smtp_accept_max_
3214 per_host which, without the tidyup, could leave open a database connection.
3216 5. Added the variables $recipient_data and $sender_data which get set from a
3217 lookup success in an ACL "recipients" or "senders" condition, or a router
3218 "senders" option, similar to $domain_data and $local_part_data.
3220 6. Moved the writing of debug_print from before to after the "senders" test
3223 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3224 problems for message scanning, either using a data ACL, or using
3225 local_scan() because the Received: header was not generated till after they
3226 were called (in order to set the time as the time of reception completion).
3227 I have revised the way this works. The header is now generated after the
3228 body is received, but before the ACL or local_scan() are called. After they
3229 are run, the timestamp in the header is updated.
3235 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3236 before starting a queue runner without re-exec. This happened only when
3237 deliver_drop_privilege was set or when the Exim user was set to root. The
3238 effect of the bug was that timeouts during subsequent deliveries caused
3239 crashes instead of being properly handled. The handler is now left at its
3240 default (and expected) setting.
3242 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3243 message, again when deliver_drop_privilege is set or Exim is run as root.
3244 The bug described in (1) was not present in this case, but the tidying up
3245 of the other signals was missing. I have made the two cases consistent.
3247 3. The ignore_target_hosts setting on a manualroute router was being ignored
3248 for hosts that were looked up using the /MX notation.
3250 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3253 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3254 operated on the sender address. After changing the $sender_address to <>
3255 for the sender address verify, Exim was re-instated it as the original
3256 (before rewriting) address, but remembering that it had rewritten it, so it
3257 wasn't rewriting it again. This bug also had the effect of breaking the
3258 sender address verification caching when the sender address was rewritten.
3260 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3261 This has been changed so that if the ip literal address matches
3262 ignore_target_hosts, the router declines.
3264 7. Added expansion conditions match_domain, match_address, and match_local_
3265 part (NOT match_host).
3267 8. The placeholder for the Received: header didn't have a length field set.
3269 9. Added code to Exim itself and to exim_lock to test for a specific race
3270 condition that could lead to file corruption when using MBX delivery. The
3271 issue is with the lockfile that is created in /tmp. If this file is removed
3272 after a process has opened it but before that process has acquired a lock,
3273 there is the potential for a second process to recreate the file and also
3274 acquire a lock. This could lead to two Exim processes writing to the file
3275 at the same time. The added code performs the same test as UW imapd; it
3276 checks after acquiring the lock that its file descriptor still refers to
3277 the same named file.
3279 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3280 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3283 11. Added the smtp_active_hostname option. If used, this will typically be made
3284 to depend on the incoming interface address. Because $interface_address is
3285 not set up until the daemon has forked a reception process, error responses
3286 that can happen earlier (such as "too many connections") no longer contain
3289 12. If an expansion in a condition on a "warn" statement fails because a lookup
3290 defers, the "warn" statement is abandoned, and the next ACL statement is
3291 processed. Previously this caused the whole ACL to be aborted.
3293 13. Added the iplsearch lookup type.
3295 14. Added ident_timeout as a log selector.
3297 15. Added tls_certificate_verified as a log selector.
3299 16. Added a global option tls_require_ciphers (compare the smtp transport
3300 option of the same name). This controls incoming TLS connections.
3302 17. I finally figured out how to make tls_require_ciphers do a similar thing
3303 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3304 before starting the TLS session.
3306 18. Tabs are now shown as \t in -bP output.
3308 19. If the log selector return_path_on_delivery was set, Exim crashed when
3309 bouncing a message because it had too many Received: header lines.
3311 20. If two routers both had headers_remove settings, and the first one included
3312 a superfluous trailing colon, the final name in the first list and the
3313 first name in the second list were incorrectly joined into one item (with a
3314 colon in the middle).
3320 1. Added -C and -D options to the exinext utility, mainly to make it easier
3321 to include in the automated testing, but these could be helpful when
3322 multiple configurations are in use.
3324 2. The exinext utility was not formatting the output nicely when there was
3325 an alternate port involved in the retry record key, nor when there was a
3326 message id as well (for retries that were specific to a specific message
3327 and a specific host). It was also confused by IPv6 addresses, because of
3328 the additional colons they contain. I have fixed the IPv4 problem, and
3329 patched it up to do a reasonable job for IPv6.
3331 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3332 delivery, the log line now contains "pipelined" if PIPELINING was used.
3334 4. An SMTP transport process used to panic and die if the bind() call to set
3335 an explicit outgoing interface failed. This has been changed; it is now
3336 treated in the same way as a connect() failure.
3338 5. A reference to $sender_host_name in the part of a conditional expansion
3339 that was being skipped was still causing a DNS lookup. This no longer
3342 6. The def: expansion condition was not recognizing references to header lines
3343 that used bh_ and bheader_.
3345 7. Added the _cache feature to named lists.
3347 8. The code for checking quota_filecount in the appendfile transport was
3348 allowing one more file than it should have been.
3350 9. For compatibility with Sendmail, the command line option
3358 and sets the incoming protocol and host name (for trusted callers). The
3359 host name and its colon can be omitted when only the protocol is to be set.
3360 Note the Exim already has two private options, -pd and -ps, that refer to
3361 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3362 "s", but I don't think that's a major issue.
3364 10. A number of refactoring changes to the code, none of which should affect
3367 (a) The number of logging options was getting close to filling up the
3368 32-bit word that was used as a bit map. I have split them into two classes:
3369 those that are passed in the argument to log_write(), and those that are
3370 only ever tested independently outside of that function. These are now in
3371 separate 32-bit words, so there is plenty of room for expansion again.
3372 There is no change in the user interface or the logging behaviour.
3374 (b) When building, for example, log lines, the code previously used a
3375 macro that called string_cat() twice, in order to add two strings. This is
3376 not really sufficiently general. Furthermore, there was one instance where
3377 it was actually wrong because one of the argument was used twice, and in
3378 one call a function was used. (As it happened, calling the function twice
3379 did not affect the overall behaviour.) The macro has been replaced by a
3380 function that can join an arbitrary number of extra strings onto a growing
3383 (c) The code for expansion conditions now uses a table and a binary chop
3384 instead of a serial search (which was left over from when there were very
3385 few conditions). Also, it now recognizes conditions like "pam" even when
3386 the relevant support is not compiled in: a suitably worded error message is
3387 given if an attempt is made to use such a condition.
3389 11. Added ${time_interval:xxxxx}.
3391 12. A bug was causing one of the ddress fields not to be passed back correctly
3392 from remote delivery subprocesses. The field in question was not being
3393 subsequently used, so this caused to problems in practice.
3395 13. Added new log selectors queue_time and deliver_time.
3397 14. Might have fixed a bug in maildirsizefile handling that threw up
3398 "unexpected character" debug warnings, and recalculated the data
3399 unnecessarily. In any case, I expanded the warning message to give more
3402 15. Added the message "Restricted characters in address" to the statements in
3403 the default ACL that block characters like @ and % in local parts.
3405 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3406 Three changes have been made:
3408 (a) There was a serious bug; a negative response to MAIL caused the whole
3409 recipient domain to be cached as invalid, thereby blocking all messages
3410 to all local parts at the same domain, from all senders. This bug has
3411 been fixed. The domain is no longer cached after a negative response to
3412 MAIL if the sender used is not empty.
3414 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3417 (c) A new callout option, "use_sender" has been added for people who want
3418 the modified behaviour.
3424 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3427 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3428 indeed breaks things for older releases.
3430 3. Added additional logging to the case where there is a problem reading data
3431 from a filter that is running in a subprocess using a pipe, in order to
3432 try to track down a specific problem.
3434 4. Testing facility fudge: when running in the test harness and attempting
3435 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3436 getting "No route to host". Convert this to a timeout.
3438 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3441 6. Some OS don't have socklen_t but use size_t instead. This affects the
3442 fifth argument of getsockopt() amongst other things. This is now
3443 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3444 can be set for individual OS. I have set it for SunOS5, OSF1, and
3445 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3446 some earlier ones do not.
3448 7. Change 4.30/15 was not doing the test caselessly.
3450 8. The standard form for an IPv6 address literal was being rejected by address
3451 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3452 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3453 this, as well as the form without the "IPv6" on the front (but only when
3454 address literals are enabled, of course).
3456 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3458 10. Exim crashed if a message with an empty sender address specified by -f
3459 encountered a router with an errors_to setting. This could be provoked only
3460 by a command such as
3464 where an empty string was supplied; "<>" did not hit this bug.
3466 11. Installed PCRE release 4.5.
3468 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3469 remained set. It is now erased.
3471 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3472 times from message ids (which are base 36 rather than the normal 62).
3474 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3475 were being counted as actual protocol errors, and logged if the log
3476 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3477 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3478 and DATA following a set of rejected RCPTs do not count as protocol errors.
3479 In other words, Exim assumes they were pipelined, though this may not
3480 actually be the case. Of course, in all cases the client gets an
3481 appropriate error code.
3483 15. If a lookup fails in an ACL condition, a message about the failure may
3484 be available; it is used if testing the ACL cannot continue, because most
3485 such messages specify what the cause of the deferral is. However, some
3486 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3487 that caused an old message to be retained and used if a later statement
3488 caused a defer, replacing the real cause of the deferral.
3490 16. If an IP address had so many PTR records that the DNS lookup buffer
3491 was not large enough to hold them, Exim could crash while trying to process
3492 the truncated data. It now detects and logs this case.
3494 17. Further to 4.21/58, another change has been made: if (and only if) the
3495 first line of a message (the first header line) ends with CRLF, a bare LF
3496 in a subsequent header line has a space inserted after it, so as not to
3497 terminate the header.
3499 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3500 unnecessarily, used atoi() instead of strtol(), and didn't check the
3501 termination when getting file sizes from file names by regex.
3503 19. Completely re-implemented the support for maildirsize files, in the light
3504 of a number of problems with the previous contributed implementation
3505 (4.30/29). In particular:
3507 . If the quota is zero, the maildirsize file is maintained, but no quota is
3510 . If the maildir directory does not exist, it is created before any attempt
3511 to write a maildirsize file.
3513 . The quota value in the file is just a cache; if the quota is changed in
3514 the transport, the new value overrides.
3516 . A regular expression is available for excluding directories from the
3519 20. The autoreply transport checks the characters in options that define the
3520 message's headers; it allows continued headers, but it was checking with
3521 isspace() after an embedded newline instead of explicitly looking for a
3524 21. If all the "regular" hosts to which an address was routed had passed their
3525 expiry times, and had not reached their retry times, the address was
3526 bounced, even if fallback hosts were defined. Now Exim should go on to try
3529 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3530 equivalent code in the SMTP transport. Some hosts send humungous responses
3531 to HELO/EHLO, more than 1024 it seems.
3533 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3534 gives compiler warnings in some environments. I've now done it "properly",
3537 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3538 (because of problems with the built-in one) was declared to return uschar *
3539 instead of char *, causing compiler failure.
3541 25. Fixed a file descriptor leak when processing alias/forward files.
3543 26. Fixed a minor format string issue in dbfn.c.
3545 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3547 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3548 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3551 29. When a sender address is verified, it is cached, to save repeating the test
3552 when there is more than one recipient in a message. However, when the
3553 verification involves a callout, it is possible for different callout
3554 options to be set for different recipients. It is too complicated to keep
3555 track of this in the cache, so now Exim always runs a verification when a
3556 callout is required, relying on the callout cache for the optimization.
3557 The overhead is duplication of the address routing, but this should not be
3560 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3561 to be verified with callout=postmaster, and the main callout worked but the
3562 postmaster check failed, the verification correctly failed. However, if a
3563 subsequent RCPT command asked for sender verification *without* the
3564 postmaster check, incorrect caching caused this verification also to fail,
3567 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3568 it was not caching the DNS options (qualify_single, search_parents) that
3569 were used when the lookup failed. A subsequent lookup with different
3570 options therefore always gave the same answer, though there were cases
3571 where it should not have. (Example: a "domains = !$mx_any" option on a
3572 dnslookup router: the "domains" option is always processed without any
3573 widening, but the router might have qualify_single set.) Now Exim uses the
3574 cached value only when the same options are set.
3576 32. Added John Jetmore's "exipick" utility to the distribution.
3578 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3579 than a timeout (e.g. a certificate is required, and is not provided), an
3580 Exim server now closes the connection immediately. Previously it waited for
3581 the client to close - but if the client is SSL, it seems that they each
3582 wait for each other, leading to a delay before one of them times out.
3584 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3585 maintained 0.8.x compatibility because I don't think many are using it, and
3586 it is clearly obsolete.
3588 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3591 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3592 client certificate was expired. A simple patch fixes this, though I don't
3593 understand the full logic of why the verify callback is called multiple
3596 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3597 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3598 which causes problems with some clients (such as the Certicom SSL Plus
3599 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3600 disables the coutermeasure allowing Eudora to connect."
3602 38. Exim was not checking that a write() to a log file succeeded. This could
3603 lead to Bad Things if a log got too big, in particular if it hit a file
3604 size limit. Exim now panics and dies if it cannot write to a log file, just
3605 as it does if it cannot open a log file.
3607 39. Modified OS/Makefile-Linux so that it now contains
3609 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3611 The two -D definitions ensure that Exim is compiled with large file
3612 support, which makes it possible to handle log files that are bigger than
3615 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3616 instance) a domain was checked against a named list that involved a lookup,
3617 causing $domain_data to be set, then another domain was checked against the
3618 same list, then the first domain was re-checked, the value of $domain_data
3619 after the final check could be wrong. In particular, if the second check
3620 failed, it could be set empty. This bug probably also applied to
3623 41. The strip_trailing_dot option was not being applied to the address given
3624 with the -f command-line option.
3626 42. The code for reading a message's header from the spool was incrementing
3627 $received_count, but never initializing it. This meant that the value was
3628 incorrect (doubled) while delivering a message in the same process in which
3629 it was received. In the most common configuration of Exim, this never
3630 happens - a fresh exec is done - but it can happen when
3631 deliver_drop_privilege is set.
3633 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3634 it now includes up to 150 characters of the unexpected data in the log
3637 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3638 and building data strings. The size of both of these buffers was 10 000
3639 bytes - far larger than anybody would *ever* want, thought I. Needless to
3640 say, somebody hit the limit. I have increased the maximum line length to
3641 20 000 and the maximum data length of concatenated lines to 100 000. I have
3642 also fixed two bugs, because there was no checking on these buffers. Tsk,
3643 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3644 buffer is too small.
3646 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3647 lsearch lookups. Now it does.
3649 46. When parsing a route_list item in a manualroute router, a fixed-length
3650 buffer was used for the list of hosts. I made this 1024 bytes long,
3651 thinking that nobody would ever have a list of hosts that long. Wrong.
3652 Somebody had a whole pile of complicated expansion conditions, and the
3653 string was silently truncated, leading to an expansion error. It turns out
3654 that it is easier to change to an unlimited length (owing to other changes
3655 that have happened since this code was originally written) than to build
3656 structure for giving a limitation error. The length of the item that
3657 expands into the list of hosts is now unlimited.
3659 47. The lsearch lookup could not handle data where the length of text line was
3660 more than 4095 characters. Such lines were truncated, leading to shortened
3661 data being returned. It should now handle lines of any length.
3663 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3664 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3667 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3668 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3669 no longer happens in comment lines. A list of replacements is now placed
3670 at the head of all of the source files, except those whose only change is
3671 to replace PERL_COMMAND in the very first #! line.
3673 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3674 messages on the queue, with a bottom-up merge sort, using code contributed
3675 by Michael Haardt. This should make operations like -bp somewhat faster on
3676 large queues. It won't affect queue runners, except when queue_run_in_order
3679 51. Installed eximstats 1.31 in the distribution.
3681 52. Added support for SRV lookups to the dnslookup router.
3683 53. If an ACL referred to $message_body or $message_body_end, the value was not
3684 reset for any messages that followed in the same SMTP session.
3686 54. The store-handling optimization for building very long strings was not
3687 differentiating between the different store pools. I don't think this
3688 actually made any difference in practice, but I've tidied it.
3690 55. While running the routers to verify a sender address, $sender_address
3691 was still set to the sender address. This is wrong, because when routing to
3692 send a bounce to the sender, it would be empty. Therefore, I have changed
3693 it so that, while verifying a sender address, $sender_address is set to <>.
3694 (There is no change to what happens when verifying a recipient address.)
3696 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3697 target A or AAAA records (if not already returned) without resetting the
3698 qualify_single or search_parents options of the DNS resolver. These are
3699 inappropriate in this case because the targets of MX and SRV records must
3700 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3701 target that, when qualified, matched something in the local domain. These
3702 two options are now turned off when doing these lookups.
3704 57. It seems that at least some releases of Reiserfs (which does not have the
3705 concept of a fixed number of inodes) returns zero and not -1 for the
3706 number of available inodes. This interacted badly with check_spool_inodes,
3707 which assumed that -1 was the "no such thing" setting. What I have done is
3708 to check that the total number of inodes is greater than zero before doing
3709 the test of how many are available.
3711 58. When a "warn" ACL statement has a log_message modifier, the message is
3712 remembered, and not repeated. This is to avoid a lot of repetition when a
3713 message has many recipients that cause the same warning to be written.
3714 Howewer, Exim was preserving the list of already written lines for an
3715 entire SMTP session, which doesn't seem right. The memory is now reset if a
3716 new message is started.
3718 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3719 debugging output unless log_rewrite was also set.
3721 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3722 of (int)(handle) when we know that handle contains (void *)(-1).
3724 61. The Exim daemon panic-logs an error return when it closes the incoming
3725 connection. However "connection reset by peer" seems to be common, and
3726 isn't really an error worthy of noting specially, so that particular error
3729 62. When Exim is trying to find all the local interfaces, it used to panic and
3730 die if the ioctl to get the interface flags failed. However, it seems that
3731 on at least one OS (Solaris 9) it is possible to have an interface that is
3732 included in the list of interfaces, but for which you get a failure error
3733 for this call. This happens when the interface is not "plumbed" into a
3734 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3735 failure of the "get flags" call assumes that the interface is down.
3737 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3738 makes life easier for people who are doing arithmetic on fields extracted
3739 from dates, where you often get leading zeros that should not be
3740 interpreted as octal.
3742 64. Added qualify_domain to the redirect router, to override the global
3745 65. If a pathologically long header line contained very many addresses (the
3746 report of this problem mentioned 10 000) and each of them was rewritten,
3747 Exim could use up a very large amount of memory. (It kept on making new
3748 copies of the header line as it rewrote, and never released the old ones.)
3749 At the expense of a bit more processing, the header rewriting function has
3750 been changed so that it no longer eats memory in this way.
3752 66. The generation of the Received: header has been moved from the time that a
3753 message starts to be received, to the time that it finishes. The timestamp
3754 in the Received: header should now be very close to that of the <= log
3755 line. There are two side-effects of this change:
3757 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3758 logged header lines no longer include the local Received: line, because
3759 it has not yet been created. The same applies to a copy of the message
3760 that is returned to a non-SMTP sender when a message is rejected.
3762 (b) When a filter file is tested using -bf, no additional Received: header
3763 is added to the test message. After some thought, I decided that this
3766 This change does not affect the value of $received_for. It is still set
3767 after address rewriting, but before local_scan() is called.
3769 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3771 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3772 gave an unhelpful panic error message, and a defer error. I have managed to
3773 change this behaviour so that it now rejects any supplied certificate,
3774 which seems right, as the list of acceptable certificates is empty.
3776 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3777 gave an unhelpful defer error. I have not managed to make this reject any
3778 supplied certificates, but the error message it gives is "no certificate
3779 supplied", which is not helpful.
3781 70. exigrep's output now also includes lines that are not associated with any
3782 message, but which match the given pattern. Implemented by a patch from
3783 Martin Sluka, which also tidied up the Perl a bit.
3785 71. Recipient callout verification, like sender verification, was using <> in
3786 the MAIL FROM command. This isn't really the right thing, since the actual
3787 sender may affect whether the remote host accepts the recipient or not. I
3788 have changed it to use the actual sender in the callout; this means that
3789 the cache record is now keyed on a recipient/sender pair, not just the
3790 recipient address. There doesn't seem to be a real danger of callout loops,
3791 since a callout by the remote host to check the sender would use <>.
3792 [SEE ABOVE: changed after hitting problems.]
3794 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3795 temporary errors. However, in the case of such a code being given after
3796 the end of a data transmission (i.e. after ".") Exim was failing to write
3797 a retry record for the message. (Yes, there was some broken host that was
3798 actually sending 8xx at this point.)
3800 73. An unknown lookup type in a host list could cause Exim to panic-die when
3801 the list was checked. (An example that provoked this was putting <; in the
3802 middle of a list instead of at the start.) If this happened during a DATA
3803 ACL check, a -D file could be left lying around. This kind of configuration
3804 error no longer causes Exim to die; instead it causes a defer errror. The
3805 incident is still logged to the main and panic logs.
3807 74. Buglet left over from Exim 3 conversion. The message "too many messages
3808 in one connection" was written to the rejectlog but not the mainlog, except
3809 when address rewriting (yes!) was being logged.
3811 75. Added write_rejectlog option.
3813 76. When a system filter was run not as root (that is, when system_filter_user
3814 was set), the values of the $n variables were not being returned to the
3815 main process; thus, they were not subsequently available in the $sn
3818 77. Added +return_path_on_delivery log selector.
3820 78. A connection timeout was being treated differently from recipients deferred
3821 when testing hosts_max_try with a message that was older than the host's
3822 retry timeout. (The host should not be counted, thus allowing all hosts to
3823 be tried at least once before bouncing.) This may have been the cause of an
3824 occasionally reported bug whereby a message would remain on the queue
3825 longer than the retry timeout, but would be bounced if a delivery was
3826 forced. I say "may" because I never totally pinned down the problem;
3827 setting up timeout/retry tests is difficult. See also the next item.
3829 79. The ultimate address timeout was not being applied to errors that involved
3830 a combination of host plus message (for example, a timeout on a MAIL
3831 command). When an address resolved to a number of possible hosts, and they
3832 were not all tried for each delivery (e.g. because of hosts_max_try), a
3833 message could remain on the queue longer than the retry timeout.
3835 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3838 81. Fixed an obscure SMTP outgoing bug which required at least the following
3839 conditions: (a) there was another message waiting for the same server;
3840 (b) the server returned 5xx to all RCPT commands in the first message so
3841 that the message was not completed; (c) the server dropped the connection
3842 or gave a negative response to the RSET that Exim sends to abort the
3843 transaction. The observed case was a dropped connection after DATA that had
3844 been sent in pipelining mode. That is, the server had advertised PIPELINING
3845 but was not implementing it correctly. The effect of the bug was incorrect
3846 behaviour, such as trying another host, and this could lead to a crash.
3852 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3853 and daemon.c were passed as pointers to ints; they should have been
3854 pointers to socklen_t variables (which are typically unsigned ints).
3856 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3859 3. Fixed a really odd bug that affected only the testing scheme; patching a
3860 certain fixed string in the binary changed the value of another string that
3861 happened to be identical to the end of the original first string.
3863 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3864 name", it returns that address as the IP address. On some operating
3865 systems (e.g. Solaris), it also passes back the IP address string as the
3866 "host name". However, on others (e.g. Linux), it passes back an empty
3867 string. Exim wasn't checking for this, and was changing the host name to an
3868 empty string, assuming it had been canonicized.
3870 5. Although rare, it is permitted to have more than one PTR record for a given
3871 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3872 all the names associated with an address, because they do in Solaris.
3873 However, it seems that they do not in Linux for data that comes from the
3874 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3875 I found this out when I moved to a new Linux workstation and tried to run
3876 the Exim test suite.
3878 To get round this problem I have changed the code so that it now does its
3879 own call to the DNS to look up PTR records when searching for a host name.
3880 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3881 addresses that are only in /etc/hosts are still found.
3883 This behaviour is, however, controlled by an option called host_lookup_
3884 order, which defaults to "bydns:byaddr". If people want to use the other
3885 order, or indeed, just use one or the other means of lookup, they can
3886 specify it in this variable.
3888 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3889 some operating systems, this comes back from gethostbyaddr() as an empty
3890 string, and this is what Exim used to test for. However, it seems that in
3891 other systems, "." is yielded. Exim now tests for this case too.
3893 7. The values of check_spool_space and check_log_space are now held internally
3894 as a number of kilobytes instead of an absolute number of bytes. If a
3895 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3896 kilobyte. This means that much larger values can be stored.
3898 8. Exim monitor: an attempt to get the action menu when not actually pointing
3899 at a message produces an empty menu entitled "No message selected". This
3900 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3901 no entries in it ("Shell widget menu has zero width and/or height"). So I
3902 have added a single, blank menu entry in this case.
3904 9. Added ${quote_local_part.
3906 10. MIME decoding is now applied to the contents of Subject: header lines when
3909 11. Now that a reference to $sender_host_address automatically causes a reverse
3910 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3911 host lookup before query-style lookups in lists that might use this
3912 variable. This has therefore been abolished, and the "net-" prefix is no
3913 longer necessary for query-style lookups.
3915 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3916 have been a typo for LFLAGS, so it has been changed.
3918 13. The install script calls Exim with "-C /dev/null" in order to find the
3919 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3920 to be output. Howeve, since Exim outputs its version number before the
3921 error, it didn't break the script. It just looked ugly. I fixed this by
3922 always allowing "-C /dev/null" if the caller is root.
3924 14. Ignore overlarge ACL variable number when reading spool file - insurance
3925 against a later release with more variables having written the file.
3927 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3928 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3929 this, as well as the form without the "IPv6" on the front.
3931 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3932 OS/Makefile-Darwin file.
3934 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3935 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3937 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3938 variable. However, some OS use ints for inodes. I've added cast to long int
3939 to get rid of the compiler warning.
3941 19. I had forgotten to lock out "/../" in configuration file names when
3942 ALT_CONFIG_PREFIX was set.
3944 20. Routers used for verification do not need to specify transports. However,
3945 if such a router generated a host list, and callout was configured, Exim
3946 crashed, because it could not find a port number from the (non-existent)
3947 transport. It now assumes port 25 in this circumstance.
3949 21. Added the -t option to exigrep.
3951 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3952 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3953 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3954 them is set, LOOKUP_LSEARCH is forced.
3956 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3958 24. Added sender and host information to the "rejected by local_scan()" log
3959 line; previously there was no indication of these.
3961 25. Added .include_if_exists.
3963 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3964 Linux. It turns out that not all file systems support this. Apparently some
3965 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3966 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3967 not supported on the file descriptor, is now ignored when Exim is trying to
3968 sync a directory. This applies only to Linux.
3970 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3972 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3973 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3976 29. Added support for maildirsize files from supplied patch (modified a bit).
3978 30. The use of :fail: followed by an empty string could lead Exim to respond to
3979 sender verification failures with (e.g.):
3981 550 Verification failed for <xxx>
3982 550 Sender verify failed
3984 where the first response line was missing the '-' that indicates it is not
3985 the final line of the response.
3987 31. The loop for finding the name of the user that called Exim had a hardwired
3988 limit of 10; it now uses the value of finduser_retries, which is used for
3989 all other user lookups.
3991 32. Added $received_count variable, available in data and not_smtp ACLs, and at
3994 33. Exim was neglecting to zero errno before one call of strtol() when
3995 expanding a string and expecting an integer value. On some systems this
3996 resulted in spurious "integer overflow" errors. Also, it was casting the
3997 result into an int without checking.
3999 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4000 did not work. The code looks as if it has *never* worked, though it appears
4001 to have been documented since at least releast 1.62. I have made it work.
4003 35. The "timeout_DNS" error in retry rules, also documented since at least
4004 1.62, also never worked. As it isn't clear exactly what this means, and
4005 clearly it isn't a major issue, I have abolished the feature by treating it
4006 as "timeout", and writing a warning to the main and panic logs.
4008 36. The display of retry rules for -brt wasn't always showing the error code
4011 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4012 timeout_connect_A, timeout_connect_MX.
4014 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4015 to the empty sender.
4017 39. The daemon was not analysing the content of -oX till after it had closed
4018 stderr and disconnected from the controlling terminal. This meant that any
4019 syntax errors were only noted on the panic log, and the return code from
4020 the command was 0. By re-arranging the code a little, I've made the
4021 decoding happen first, so such errors now appear on stderr, and the return
4022 code is 1. However, the actual setting up of the sockets still happens in
4023 the disconnected process, so errors there are still only recorded on the
4026 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4027 connections (as happens on some IP stacks) was logged at start up time as
4028 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4029 it from "IPv6 and IPv4", which means that two separate sockets are being
4032 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4033 says whether AF_INET or AF_INET6 was passed as an argument.
4035 42. Exiwhat output was messed up when time zones were included in log
4038 43. Exiwhat now gives more information about the daemon's listening ports,
4039 and whether -tls-on-connect was used.
4041 44. The "port" option of the smtp transport is now expanded.
4043 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4044 silently ignored. Now an error message is written to the main and panic
4047 46. There's a new ACL modifier called "logwrite" which writes to a log file
4048 as soon as it is encountered.
4050 47. Added $local_user_uid and $local_user_gid at routing time.
4052 48. Exim crashed when trying to verify a sender address that was being
4055 49. Exim was recognizing only a space character after ".include". It now also
4056 recognizes a tab character.
4058 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4059 extracting the relevant information from the specification. The man page no
4060 longer contains scrambled data for the -d option, and I've added a section
4061 at the front about calling Exim under different names.
4063 51. Added "extra_headers" argument to the "mail" command in filter files.
4065 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4068 53. Installed eximstats 1.29.
4070 54. Added transport_filter_timeout as a generic transport option.
4072 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4073 Cc: header lines. This was required by RFC 822, but it not required by RFC
4076 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4077 incoming messages that did not have them. Now it does so only if the
4078 message originates locally, that is, if there is no associated remote host
4079 address. When Resent- header lines are present, this applies to the Resent-
4080 lines rather than the non-Resent- lines.
4082 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4083 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4085 58. Messages for configuration errors now include the name of the main
4086 configuration file - useful now that there may be more than one file in a
4087 list (.included file names were always shown).
4089 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4090 for those rare installations that do not start the daemon as root or run it
4091 setuid root. I've cut out the call to initgroups() if the daemon is not
4094 60. The Exim user and group can now be bound into the binary as text strings
4095 that are looked up at the start of Exim's processing.
4097 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4099 62. Added $mailstore_basename variable.
4101 63. Installed patch to sieve.c from Michael Haardt.
4103 64. When Exim failed to open the panic log after failing to open the main log,
4104 the original message it was trying to log was written to stderr and debug
4105 output, but if they were not available (the usual case in production), it
4106 was lost. Now it is written to syslog before the two lines that record the
4107 failures to open the logs.
4109 65. Users' Exim filters run in subprocesses under the user's uid. It is
4110 possible for a "deliver" command or an alias in a "personal" command to
4111 provoke an address rewrite. If logging of address rewriting is configured,
4112 this fails because the process is not running as root or exim. There may be
4113 a better way of dealing with this, but for the moment (because 4.30 needs
4114 to be released), I have disabled address rewrite logging when running a
4115 filter in a non-root, non-exim process.
4121 1. The buildconfig auxiliary program wasn't quoting the value set for
4122 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4123 not defined. This bug was masked in 4.22 by the effect that was fixed in
4126 2. Some messages that were rejected after a message id was allocated were
4127 shown as "incomplete" by exigrep. It no longer does this for messages that
4128 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4130 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4131 have allow_domain_literals set, the ID did not get logged in the <= line.
4132 Domain literals are now always recognized in Message-ID: header lines.
4134 4. The first argument for a ${extract expansion item is the key name or field
4135 number. Leading and trailing spaces in this item were not being ignored,
4136 causing some misleading effects.
4138 5. When deliver_drop_privilege was set, single queue runner processes started
4139 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4140 same command in the process it spins off) were not dropping privilege.
4142 6. When the daemon running as "exim" started a queue runner, it always
4143 re-executed Exim in the spun-off process. This is a waste of effort when
4144 deliver_drop_privilege is set. The new process now just calls the
4145 queue-runner function directly.
4151 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4154 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4155 ignored. Though the use of -oP was forcing the writing of a pid file, it
4156 was always written to the default place.
4158 3. If the message "no IP address found for host xxxx" is generated during
4159 incoming verification, it is now followed by identification of the incoming
4160 connection (so you can more easily find what provoked it).
4162 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4164 5. Added some features to "harden" Exim a bit more against certain attacks:
4166 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4167 be put in Local/Makefile. This is like the never_users runtime option,
4168 but it cannot be overridden. The default setting is "root".
4170 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4171 prefix string with which any file named in a -C command line option
4174 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4175 is retained for -C and -D only if the caller of Exim is root. Without
4176 it, the exim user may also use -C and -D and retain privilege.
4178 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4179 command line option is disabled.
4181 6. Macro names set by the -D option must start with an upper case letter, just
4182 like macro names defined in the configuration file.
4184 7. Added "dereference=" facility to LDAP.
4186 8. Two instances of the typo "uknown" in the source files are fixed.
4188 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4189 the Configure-Makefile script screwed up while processing it.
4191 10. Incorporated PCRE 4.4.
4193 11. The SMTP synchronization check was not operating right at the start of an
4194 SMTP session. For example, it could not catch a HELO sent before the client
4195 waited for the greeting. There is now a check for outstanding input at the
4196 point when the greeting is written. Because of the duplex, asynchronous
4197 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4198 way, but not yet received, when the check is performed.
4200 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4201 on TCP/IP sockets, because this apparently causes some broken clients to
4204 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4205 unchanged) from the Cygwin maintainer.
4207 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4208 was supported instead of testing for mbx. Effectively a typo.
4210 15. The spa authenticator server code was not checking that the input it
4211 received was valid base64.
4213 16. The debug output line for the "set" modifier in ACLs was not showing the
4214 name of the variable that was being set.
4216 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4218 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4219 lookup on its own. Something else was needed to trigger the lookup. For
4220 example, a match in host_lookup or the need for a host name in a host list.
4221 Now, if $sender_host_name is referenced and the host name has not yet been
4222 looked up, a lookup is performed. If the lookup fails, the variable remains
4223 empty, and $host_lookup_failed is set to "1".
4225 19. Added "eqi" as a case-independent comparison operator.
4227 20. The saslauthd authentication condition could segfault if neither service
4228 nor realm was specified.
4230 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4231 error message that was logged was misleading, and incoming SMTP
4232 connections were dropped. The message is now more accurate, and temporary
4233 errors are given to SMTP connections.
4235 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4236 (except RSET) with a 421 error, until QUIT is received. However, it was
4237 failing to send a response to QUIT.
4239 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4240 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4241 ACL is now run after the helo_try_verify_hosts code.
4243 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4244 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4245 case-independent, so other case variants are also recognized). Apparently
4246 some systems use these upper case variants.
4248 25. If more than two messages were waiting for the same host, and a transport
4249 filter was specified for the transport, Exim sent two messages over the
4250 same TCP/IP connection, and then failed with "socket operation on non-
4251 socket" when it tried to send the third.
4253 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4255 27. The extern definition of crypt16() in expand.c was not being excluded when
4256 the OS had its own crypt16() function.
4258 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4259 as a preferred synonym for return_size_limit, both as an option and as an
4262 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4264 30. Changed the default configuration ACL to relax the local part checking rule
4265 for addresses that are not in any local domains. For these addresses,
4266 slashes and pipe symbols are allowed within local parts, but the sequence
4267 /../ is explicitly forbidden.
4269 31. SPA server authentication was not clearing the challenge buffer before
4272 32. log_message in a "warn" ACL statement was writing to the reject log as
4273 well as to the main log, which contradicts the documentation and doesn't
4274 seem right (because no rejection is happening). So I have stopped it.
4276 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4277 However, I am unable to do any testing of this.
4279 34. Fixed an infelicity in the appendfile transport. When checking directories
4280 for a mailbox, to see if any needed to be created, it was accidentally
4281 using path names with one or more superfluous leading slashes; tracing
4282 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4284 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4285 contents are added to the log line that is written for every discarded
4286 recipient. (Previously a log_message setting was ignored.)
4288 36. The ${quote: operator now quotes the string if it is empty.
4290 37. The install script runs exim in order to find its version number. If for
4291 some reason other than non-existence or emptiness, which it checks, it
4292 could not run './exim', it was installing it with an empty version number,
4293 i.e. as "exim-". This error state is now caught, and the installation is
4296 38. An argument was missing from the function that creates an error message
4297 when Exim fails to connect to the socket for saslauthd authentication.
4298 This could cause Exim to crash, or give a corrupted message.
4300 39. Added isip, isip4, and isip6 to ${if conditions.
4302 40. The ACL variables $acl_xx are now saved with the message, and can be
4303 accessed later in routers, transports, and filters.
4305 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4306 strings in the file are not string-expanded.
4308 42. If a MAIL command specified a SIZE value that was too large to fit into an
4309 int variable, the check against message_size_limit failed. Such values are
4310 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4311 day this will have to be increased, but I don't think I want to be around
4312 when emails are that large.
4319 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4320 iconv() is not standard in FreeBSD.
4322 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4323 IPv6 enabled. The observed symptom was a segmentation fault on return from
4324 the function os_common_find_running_interfaces() in src/os.c.
4326 3. In the check_special_case() function in daemon.c I had used "errno" as an
4327 argument name, which causes warnings on some systems. This was basically a
4328 typo, since it was named "eno" in the comments!
4330 4. The code that waits for the clock to tick (at a resolution of some fraction
4331 of a second) so as to ensure message-id uniqueness was always waiting for
4332 at least one whole tick, when it could have waited for less. [This is
4333 almost certainly not relevant at current processor speeds, where it is
4334 unlikely to ever wait at all. But we try to future-proof.]
4336 5. The function that sleeps for a time interval that includes fractions of a
4337 second contained a race. It did not block SIGALRM between setting the
4338 timer, and suspending (a couple of lines later). If the interval was short
4339 and the sigsuspend() was delayed until after it had expired, the suspension
4340 never ended. On busy systems this could lead to processes getting stuck for
4343 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4344 process, before it forks any delivery processes. The open lookup caching
4345 mechanism meant that the open file or database connection was passed into
4346 the delivery process. The problem was that delivery processes always tidy
4347 up cached lookup data. This could cause a problem for the next delivery
4348 process started by the queue runner, because the external queue runner
4349 process does not know about the closure. So the next delivery process
4350 still has data in the lookup cache. In the case of a file lookup, there was
4351 no problem because closing a file descriptor in a subprocess doesn't affect
4352 the parent. However, if the lookup was caching a connection to a database,
4353 the connection was closed, and the second delivery process was likely to
4354 see errors such as "PGSQL: query failed: server closed the connection
4355 unexpectedly". The problem has been fixed by closing all cached lookups
4356 in a queue runner before running a delivery process.
4358 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4359 seem to have the "const" qualifier which it has on other OS. I've
4362 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4363 *of the same type* (client or server) with the same public name that an
4364 error should be diagnosed.
4366 9. When Exim looked up a host name for an IP address, but failed to find the
4367 original IP address when looking up the host name (a safety check), it
4368 output the message "<ip address> does not match any IP for NULL", which was
4369 confusing, to say the least. The bug was that the host name should have
4370 appeared instead of "NULL".
4372 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4373 user that is built into the binary, and the -C or -D options is used, root
4374 privilege is dropped before the configuration file is read. In addition,
4375 logging is switched to stderr instead of the normal log files. If the
4376 configuration then re-defines the Exim user, the unprivileged environment
4377 is probably not what is expected, so Exim logs a panic warning message (but
4380 However, if deliver_drop_privilege is set, the unprivileged state may well
4381 be exactly what is intended, so the warning has been cut out in that case,
4382 and Exim is allowed to try to write to its normal log files.
4388 1. smtp_return_error_details was not giving details for temporary sender
4389 or receiver verification errors.
4391 2. Diagnose a configuration error if two authenticators have the same public
4394 3. Exim used not to create the message log file for a message until the first
4395 delivery attempt. This could be confusing when incoming messages were held
4396 for policy or load reasons. The message log file is now created at the time
4397 the message is received, and an initial "Received" line is written to it.
4399 4. The automatically generated man page for command line options had a minor
4400 bug that caused no ill effects; however, a more serious problem was that
4401 the procedure for building the man page automatically didn't always
4402 operate. Consequently, release 4.20 contains an out-of-date version. This
4403 shouldn't happen again.
4405 5. When building Exim with embedded Perl support, the script that builds the
4406 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4407 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4409 6. The freeze_tell option was not being used for messages that were frozen on
4410 arrival, either by an ACL or by local_scan().
4412 7. Added the smtp_incomplete_transaction log selector.
4414 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4415 was accepting AUTH without a new EHLO.
4417 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4418 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4421 10. Logging of TCP/IP connections (when configured) now happens in the main
4422 daemon process instead of the child process, so that the TCP/IP connection
4423 count is more accurate (but it can never be perfect).
4425 11. The use of "drop" in a nested ACL was not being handled correctly in the
4426 outer ACL. Now, if condition failure induced by the nested "drop" causes
4427 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4428 or "require"), the connection is dropped.
4430 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4431 that yield "discard" can now be used with an "accept" or a "discard" verb,
4432 but an error is generated for any others (because I can't see a useful way
4433 to define what should happen).
4435 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4436 are now processed in the same way as lines in the Exim configuration file.
4437 In particular, continuation lines are supported.
4439 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4441 15. Added -ti meaning -t -i.
4443 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4444 domains, and warn by logging if others are found.
4446 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4447 returned from GIFCONF when Exim is trying to find the list of interfaces on
4448 a host. The code in os.c has been modified to copy each ifreq to an aligned
4449 structure in all cases.
4451 Also, in some cases, the returned ifreq's were being copied to a 'struct
4452 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4453 means the last couple of bytes of an IPv6 address could be chopped if the
4454 ifreq contained only a normal sockaddr (14 bytes storage).
4456 18. Named domain lists were not supported in the hosts_treat_as_local option.
4457 An entry such as +xxxx was not recognized, and was treated as a literal
4460 19. Ensure that header lines added by a DATA ACL are included in the reject log
4461 if the ACL subsequently rejects the message.
4463 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4464 MD5 (which is deprecated).
4466 21. When testing a filter file using -bf, Exim was writing a message when it
4467 took the sender from a "From " line in the message, but it was not doing so
4468 when it took $return_path from a Return-Path: header line. It now does.
4470 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4471 with a valid header line field name (a series of printing characters
4472 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4474 23. Changed "disc" in the source to "disk" to conform to the documentation and
4475 the book and for uniformity.
4477 24. Ignore Sendmail's -Ooption=value command line item.
4479 25. When execve() failed while trying to run a command in a pipe transport,
4480 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4481 could be confused with a return value of 69 from the command itself. This
4482 has been changed to 127, the value the shell returns if it is asked to run
4483 a non-existent command. The wording for the related log line suggests a
4484 non-existent command as the problem.
4486 26. If received_header_text expands to an empty string, do not add a Received:
4487 header line to the message. (Well, it adds a token one on the spool, but
4488 marks it "old" so that it doesn't get used or transmitted.)
4490 27. Installed eximstats 1.28 (addition of -nt option).
4492 28. There was no check for failure on the call to getsockname() in the daemon
4493 code. This can fail if there is a shortage of resources on the system, with
4494 ENOMEM, for example. A temporary error is now given on failure.
4496 29. Contrary to the C standard, it seems that in some environments, the
4497 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4498 program. Exim now does this explicitly; it affects the formatting of
4499 timestamps using strftime().
4501 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4502 complaints. I've now initialized all the variables, to avoid this.
4504 32. Header lines added by a system filter were not being "seen" during
4505 transport-time rewrites.
4507 33. The info_callback() function passed to OpenSSL is set up with type void
4508 (*)(SSL *, int, int), as described somewhere. However, when calling the
4509 function (actually a macro) that sets it up, the type void(*)() is
4510 expected. I've put in a cast to prevent warnings from picky compilers.
4512 34. If a DNS black list lookup found a CNAME record, but there were no A
4513 records associated with the domain it pointed at, Exim crashed.
4515 35. If a DNS black list lookup returned more than one A record, Exim ignored
4516 all but the first. It now scans all returned addresses if a particular IP
4517 value is being sought. In this situation, the contents of the
4518 $dnslist_value variable are a list of all the addresses, separated by a
4521 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4522 to accept a host name and all its aliases if the forward lookup for any of
4523 them yielded the IP address of the incoming connection. Now it accepts only
4524 those names whose forward lookup yields the correct IP address. Any other
4525 names are discarded. This closes a loophole whereby a rogue DNS
4526 administrator could create reverse DNS records to break through a
4527 wildcarded host restriction in an ACL.
4529 37. If a user filter or a system filter that ran in a subprocess used any of
4530 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4531 the wrong values were passed to the pipe command ($thisaddress had the
4532 value of $0, $0 had the value of $1, etc). This bug was introduced by
4533 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4535 38. Improved the line breaking for long SMTP error messages from ACLs.
4536 Previously, if there was no break point between 40 and 75 characters, Exim
4537 left the rest of the message alone. Two changes have been made: (a) I've
4538 reduced the minimum length to 35 characters; (b) if it can't find a break
4539 point between 35 and 75 characters, it looks ahead and uses the first one
4540 that it finds. This may give the occasional overlong line, but at least the
4541 remaining text gets split now.
4543 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4544 file descriptors might be low, and that setting 1000 would always raise it.
4545 It turns out that in some environments, the limit is already over 1000 and
4546 that lowering it causes trouble. So now Exim takes care not to decrease it.
4548 40. When delivering a message, the value of $return_path is set to $sender_
4549 address at the start of routing (routers may change the value). By an
4550 oversight, this default was not being set up when an address was tested by
4551 -bt or -bv, which affected the outcome if any router or filter referred to
4554 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4555 log only when "message" or "log_message" are set. However, if one of the
4556 conditions was an address verification, or a call to a nested ACL, the
4557 messages generated by the underlying test were being passed through. This
4558 no longer happens. The underlying message is available in $acl_verify_
4559 message for both "message" and "log_message" expansions, so it can be
4560 passed through if needed.
4562 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4563 new expansion $bh_ to give the encoded byte string without charset
4564 translation. Translation happens only if iconv() is available; HAVE_ICONV
4565 indicates this at build time. HEADERS_CHARSET gives the charset to
4566 translate to; headers_charset can change it in the configuration, and
4567 "headers charset" can change it in an individual filter file.
4569 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4570 that creates RFC 2047 encoded "words" labels them as that charset instead
4571 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4572 expansion operator; (ii) when Exim creates a From: line for a local
4573 message; (iii) when a header line is rewritten to include a "phrase" part.
4575 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4576 buggy, causing it to skip the first lines of messages whose message ID
4577 ended in 'D'. This would not have bitten before Exim release 4.14, because
4578 message IDs were unlikely to end in 'D' before then. The effect was to have
4579 incorrect size information for certain domains.
4581 45. #include "config.h" was missing at the start of the crypt16.c module. This
4582 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4585 46. If there was a timeout during a "random" callout check, Exim treated it as
4586 a failure of the random address, and carried on sending RSET and the real
4587 address. If the delay was just some slowness somewhere, the response to the
4588 original RCPT would be taken as a response to RSET and so on, causing
4589 mayhem of various kinds.
4591 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4592 when I implemented it. It didn't allow for the fact that some option values
4593 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4594 the right test for positive values.
4596 48. Domain names in DNS records are case-independent. Exim always looks them up
4597 in lower case. Some resolvers return domain names in exactly the case they
4598 appear in the zone file, that is, they may contain uppercase letters. Not
4599 all resolvers do this - some return always lower case. Exim was treating a
4600 change of case by a resolver as a change of domain, similar to a widening
4601 of a domain abbreviation. This triggered its re-routing code and so it was
4602 trying to route what was effectively the same domain again. This normally
4603 caused routing to fail (because the router wouldn't handle the domain
4604 twice). Now Exim checks for this case specially, and just changes the
4605 casing of the domain that it ultimately uses when it transmits the message
4608 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4611 50. If a filter generated a file delivery with a non-absolute name (possible if
4612 no home directory exists for the router), the forbid_file option was not
4615 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4616 the existing equality matching.
4618 52. Exim was using ints instead of ino_t variables in some places where it was
4619 dealing with inode numbers.
4621 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4622 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4623 called TMPDIR, and if it finds it is different, it changes its value.
4625 54. The smtp_printf() function is now made available to local_scan() so
4626 additional output lines can be written before returning. There is also an
4627 smtp_fflush() function to enable the detection of a dropped connection.
4628 The variables smtp_input and smtp_batched_input are exported to
4631 55. Changed the default runtime configuration: the message "Unknown user"
4632 has been removed from the ACL, and instead placed on the localuser router,
4633 using the cannot_route_message feature. This means that any verification
4634 failures that generate their own messages won't get overridden. Similarly,
4635 the "Unrouteable address" message that was in the ACL for unverifiable
4636 relay addresses has also been removed.
4638 56. Added hosts_avoid_esmtp to the smtp transport.
4640 57. The exicyclog script was not checking for the esoteric option
4641 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4642 will work only if exicyclog is run under the appropriate euid.
4644 58. Following a discussion on the list, the rules by which Exim recognises line
4645 endings on incoming messages have been changed. The -dropcr and drop_cr
4646 options are now no-ops, retained only for backwards compatibility. The
4647 following line terminators are recognized: LF CRLF CR. However, special
4648 processing applies to CR:
4650 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4651 nor a local message in the state where . is a terminator.
4653 (ii) If a bare CR is encountered in a header line, an extra space is added
4654 after the line terminator so as not to end the header. The reasoning
4655 behind this is that bare CRs in header lines are most likely either
4656 to be mistakes, or people trying to play silly games.
4658 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4659 was being incorrectly given as 18 bytes larger than it should have been.
4660 This is a VOB (very old bug).
4662 60. This may never have affected anything current, but just in case it has:
4663 When the local host is found other than at the start of a list of hosts,
4664 the local host, those with the same MX, and any that follow, are discarded.
4665 When the list in question was part of a longer list of hosts, the following
4666 hosts (not currently being processed) were also being discarded. This no
4667 longer happens. I'm not sure if this situation could ever has previously
4670 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4673 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4674 header. This is something that is recommended in a new Internet Draft, and
4675 is something that is documented as being done by Sendmail. There are two
4676 possible values. For messages generated by the autoreply transport, Exim
4679 Auto-Submitted: auto-replied
4681 whereas for all other generated messages (e.g. bounces) it adds
4683 Auto-Submitted: auto-generated
4685 63. The "personal" condition in filters now includes a test for the
4686 Auto-Submitted: header. If it contains the string "auto-" the message it
4687 not considered personal.
4689 64. Added rcpt_include_affixes as a generic transport option.
4691 65. Added queue_only_override (default true).
4693 66. Added the syslog_duplication option.
4695 67. If what should have been the first header line of a message consisted of
4696 a space followed by a colon, Exim was mis-interpreting it as a header line.
4697 It isn't of course - it is syntactically invalid and should therefore be
4698 treated as the start of the message body. The misbehaviour could have
4699 caused a number of strange effects, including loss of data in subsequent
4700 header lines, and spool format errors.
4702 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4703 client host had authenticated. This control can now be exercised by an ACL
4704 for more flexibility.
4706 69. By default, callouts do not happen when testing with -bh. There is now a
4707 variant, -bhc, which does actually run the callout code, including
4708 consulting and updating the callout cache.
4710 70. Added support for saslauthd authentication, courtesy of Alexander
4713 71. If statvfs() failed on the spool or log directories while checking their
4714 size for availability, Exim confusingly gave the error "space shortage".
4715 Furthermore, in debugging mode it crashed with a floating point exception.
4716 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4717 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4718 serious problem, Exim now writes to the main and panic logs when this
4719 happens, with details of the failure. It then refuses to accept the
4720 incoming message, giving the message "spool directory problem" or "log
4721 directory problem" with a 421 code for SMTP messages.
4723 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4724 0, 1, and 2 exist, because some OS complain for execs without them (see
4725 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4726 descriptors. However, the code omitted to check that the open succeeded,
4727 causing mysterious errors if for some reason the permissions on /dev/null
4728 got screwed. Now Exim writes a message to the main and panic logs, and
4729 bombs out if it can't open /dev/null.
4731 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4732 interact so that it is all more flexible. It is supposed to remain
4733 backwards compatible. Also added extra_local_interfaces.
4735 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4736 to bomb out with an assertion failure - to the client this appears as a
4737 connection drop. This problem occurs in the part of the code that was taken
4738 from the Samba project. Fortunately, the assertion is in a very simple
4739 function, so I have fixed this by reproducing the function inline in the
4740 one place where it is called, and arranging for authentication to fail
4741 instead of killing the process with assert().
4743 75. The SPA client code was not working when the server requested OEM rather
4744 than Unicode encoding.
4746 76. Added code to make require_files with a specific uid setting more usable in
4747 the case where statting the file as root fails - usually a non-root-mounted
4748 NFS file system. When this happens and the failure is EACCES, Exim now
4749 forks a subprocess and does the per-uid checking as the relevant uid.
4751 77. Added process_log_path.
4753 78. If log_file_path was not explicitly set, a setting of check_log_space or
4754 check_log_inodes was ignored.
4756 79. If a space check for the spool or log partitions fails, the incident is now
4757 logged. Of course, in the latter case the data may get lost...
4759 80. Added the %p formatting code to string_format() so that it can be used to
4760 print addresses in debug_print(). Adjusted all the address printing in the
4761 debugging in store.c to use %p rather than %d.
4763 81. There was a concern that a line of code in smtp_in.c could overflow a
4764 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4765 initially expressed, the concern was not well-founded, because trailing
4766 spaces are removed early. However, if the trailing spaces were followed by
4767 a NULL, they did not get removed, so the overflow was possible. Two fixes
4770 (a) I re-wrote the offending code in a cleaner fashion.
4771 (b) If an incoming SMTP command contains a NULL character, it is rejected
4774 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4775 runs initgroups(), so that if the Exim user is in any additional groups,
4776 they will be used during message reception.
4782 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob