1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.530 2007/08/29 15:06:47 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Another patch from the Sieve maintainer.
11 PH/02 When an IPv6 address is converted to a string for single-key lookup
12 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
13 dots are used instead of colons so that keys in lsearch files need not
14 contain colons. This was done some time before quoting was made available
15 in lsearch files. However, iplsearch files do require colons in IPv6 keys
16 (notated using the quote facility) so as to distinguish them from IPv4
17 keys. This meant that lookups for IP addresses in host lists did not work
18 for iplsearch lookups.
20 This has been fixed by arranging for IPv6 addresses to be expressed with
21 colons if the lookup type is iplsearch. This is not incompatible, because
22 previously such lookups could never work.
24 The situation is now rather anomolous, since one *can* have colons in
25 ordinary lsearch keys. However, making the change in all cases is
26 incompatible and would probably break a number of configurations.
28 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
31 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
32 conversion specification without a maximum field width, thereby enabling
33 a rogue spamd server to cause a buffer overflow. While nobody in their
34 right mind would setup Exim to query an untrusted spamd server, an
35 attacker that gains access to a server running spamd could potentially
36 exploit this vulnerability to run arbitrary code as the Exim user.
38 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
39 $primary_hostname instead of what libspf2 thinks the hosts name is.
41 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
42 a directory entry by the name of the lookup key. Previously, if a
43 symlink pointed to a non-existing file or a file in a directory that
44 Exim lacked permissions to read, a lookup for a key matching that
45 symlink would fail. Now it is enough that a matching directory entry
46 exists, symlink or not. (Bugzilla 503.)
48 PH/03 The body_linecount and body_zerocount variables are now exported in the
51 PH/04 Added the $dnslist_matched variable.
53 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
54 This means they are set thereafter only if the connection becomes
57 PH/06 Added the client_condition to authenticators so that some can be skipped
58 by clients under certain conditions.
60 PH/07 The error message for a badly-placed control=no_multiline_responses left
61 "_responses" off the end of the name.
63 PH/08 Added -Mvc to output a copy of a message in RFC 2822 format.
65 PH/09 Tidied the code for creating ratelimiting keys, creating them explicitly
66 (without spaces) instead of just copying the configuration text.
68 PH/10 Added the /noupdate option to the ratelimit ACL condition.
70 PH/11 Added $max_received_linelength.
72 PH/12 Added +ignore_defer and +include_defer to host lists.
74 PH/13 Installed PCRE version 7.2. This needed some changes because of the new
75 way in which PCRE > 7.0 is built.
77 PH/14 Implemented queue_only_load_latch.
79 PH/15 Removed an incorrect (int) cast when reading the value of SIZE in a
80 MAIL command. The effect was to mangle the value on 64-bit systems.
82 PH/16 Another patch from the Sieve maintainer.
84 PH/17 Added the NOTQUIT ACL, based on a patch from Ted Cooper.
86 PH/18 If a system quota error occurred while trying to create the file for
87 a maildir delivery, the message "Mailbox is full" was not appended to the
88 bounce if the delivery eventually timed out. Change 4.67/27 below applied
89 only to a quota excession during the actual writing of the file.
91 PH/19 It seems that peer DN values may contain newlines (and other non-printing
92 characters?) which causes problems in log lines. The DN values are now
93 passed through string_printing() before being added to log lines.
95 PH/20 Added the "servers=" facility to MySQL and PostgreSQL lookups. (Oracle
96 and InterBase are left for another time.)
98 PH/21 Added message_body_newlines option.
100 PH/22 Guard against possible overflow in moan_check_errorcopy().
102 PH/23 POSIX allows open() to be a macro; guard against that.
104 PH/24 If the recipient of an error message contained an @ in the local part
105 (suitably quoted, of course), incorrect values were put in $domain and
106 $local_part during the evaluation of errors_copy.
112 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
113 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
116 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
117 issue a MAIL command.
119 PH/02 In an ACL statement such as
121 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
123 if a client was not listed at all, or was listed with a value other than
124 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
125 the condition was not true (as it should be), so access was not denied.
126 The bug was that the ! inversion was incorrectly passed on to the second
127 item. This has been fixed.
129 PH/03 Added additional dnslists conditions == and =& which are different from
130 = and & when the dns lookup returns more than one IP address.
132 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
133 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
135 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
136 FSYNC, which compiles an option called disable_fsync that allows for
137 bypassing fsync(). The documentation is heavily laced with warnings.
139 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
141 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
142 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
143 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
144 including adding "make clean"; (3) Added -fPIC when compiling the test
145 dynamically loaded module, to get rid of a warning.
147 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
148 message fails, move_frozen_messages = true and ignore_bounce_errors_after
149 = 0s. The bug is otherwise harmless.
151 PH/07 There was a bug in the dovecot authenticator such that the value of
152 $auth1 could be overwritten, and so not correctly preserved, after a
153 successful authentication. This usually meant that the value preserved by
154 the server_setid option was incorrect.
156 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
158 PH/09 Installed PCRE release 7.0.
160 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
161 run for batched SMTP input. It is now run at the start of every message
162 in the batch. While fixing this I discovered that the process information
163 (output by running exiwhat) was not always getting set for -bs and -bS
164 input. This is fixed, and it now also says "batched" for BSMTP.
166 PH/11 Added control=no_pipelining.
168 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
169 patch, slightly modified), and move the expansion of helo_data till after
170 the connection is made in the smtp transport (so it can use these
173 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
175 PH/14 Added log_selector = +pid.
177 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
179 PH/16 Add ${if forany and ${if forall.
181 PH/17 Added dsn_from option to vary the From: line in DSNs.
183 PH/18 Flush SMTP output before performing a callout, unless control =
184 no_callout_flush is set.
186 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
187 was true (the default) a successful delivery failed to delete the retry
188 item, thus causing premature timeout of the address. The bug is now
191 PH/20 Added hosts_avoid_pipelining to the smtp transport.
193 PH/21 Long custom messages for fakedefer and fakereject are now split up
194 into multiline reponses in the same way that messages for "deny" and
195 other ACL rejections are.
197 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
198 with slight modification.
200 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
201 draft, changing the syntax and factoring some duplicate code".
203 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
204 for deliveries of the second and subsequent messages over the same SMTP
207 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
208 ${reduce, with only minor "tidies".
210 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
212 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
213 expansion side effects.
215 PH/27 When a message times out after an over-quota error from an Exim-imposed
216 quota, the bounce message says "mailbox is full". This message was not
217 being given when it was a system quota that was exceeded. It now should
220 MH/03 Made $recipients available in local_scan(). local_scan() already has
221 better access to the recipient list through recipients_list[], but
222 $recipients can be useful in postmaster-provided expansion strings.
224 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
225 in the case of a MAIL command with additional options following the
226 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
227 were accidentally chopped off.
229 PH/29 SMTP synchronization checks are implemented when a command is read -
230 there is a check that no more input is waiting when there shouldn't be
231 any. However, for some commands, a delay in an ACL can mean that it is
232 some time before the response is written. In this time, more input might
233 arrive, invalidly. So now there are extra checks after an ACL has run for
234 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
235 pipelining has not been advertised.
237 PH/30 MH's patch to allow iscntrl() characters to be list separators.
239 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
240 returned in the SMTP response when smtp_return_error_details was false.
243 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
244 instead of the C I/O that was originally supplied, because problems were
247 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
248 Exim which did not show up earlier: it was assuming that a call to
249 SSL_CTX_set_info_callback() might give an error value. In fact, there is
250 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
251 was a macro that became an assignment, so it seemed to work. This has
252 changed to a proper function call with a void return, hence the compile
253 error. Exim's code has been fixed.
255 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
258 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
260 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
262 PH/37 If a message is not accepted after it has had an id assigned (e.g.
263 because it turns out to be too big or there is a timeout) there is no
264 "Completed" line in the log. When some messages of this type were
265 selected by exigrep, they were listed as "not completed". Others were
266 picked up by some special patterns. I have improved the selection
267 criteria to be more general.
269 PH/38 The host_find_failed option in the manualroute router can now be set
270 to "ignore", to completely ignore a host whose IP address cannot be
271 found. If all hosts are ignored, the behaviour is controlled by the new
272 host_all_ignored option.
274 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
275 homing or because of multiple MX records with /mx) generated more than
276 one IP address, and the following item turned out to be the local host,
277 all the secondary addresses of the first item were incorrectly removed
278 from the list, along with the local host and any following hosts (which
279 is what is supposed to happen).
281 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
282 whoever called Exim into the -H file. In the case of the daemon it was
283 behaving confusingly. When first started, it used values for whoever
284 started the daemon, but after a SIGHUP it used the Exim user (because it
285 calls itself on a restart). I have changed the code so that it now always
288 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
289 message are rejected with the same error (e.g. no authentication or bad
290 sender address), and a DATA command is nevertheless sent (as can happen
291 with PIPELINING or a stupid MUA), the error message that was given to the
292 RCPT commands is included in the rejection of the DATA command. This is
293 intended to be helpful for MUAs that show only the final error to their
296 PH/42 Another patch from the Sieve maintainer.
298 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
299 Eximstats - Fixed some broken HTML links and added missing column headers
301 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
302 columns for Rejects, Temp Rejects, Ham, and Spam rows.
304 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
306 PH/43 Yet another patch from the Sieve maintainer.
308 PH/44 I found a way to check for a TCP/IP connection going away before sending
309 the response to the final '.' that terminates a message, but only in the
310 case where the client has not sent further data following the '.'
311 (unfortunately, this is allowed). However, in many cases there won't be
312 any further data because there won't be any more messages to send. A call
313 to select() can be used: if it shows that the input is "ready", there is
314 either input waiting, or the socket has been closed. An attempt to read
315 the next input character can distinguish the two cases. Previously, Exim
316 would have sent an OK response which the client would never have see.
317 This could lead to message repetition. This fix should cure that, at
318 least in a lot of common cases.
320 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
321 advertised in response to EHLO.
327 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
328 fixed by 4.65/MH/01 (is this a record?) are fixed:
330 (i) An empty string was always treated as zero by the numeric comparison
331 operators. This behaviour has been restored.
333 (ii) It is documented that the numeric comparison operators always treat
334 their arguments as decimal numbers. This was broken in that numbers
335 starting with 0 were being interpreted as octal.
337 While fixing these problems I realized that there was another issue that
338 hadn't been noticed. Values of message_size_limit (both the global option
339 and the transport option) were treated as octal if they started with 0.
340 The documentation was vague. These values are now always treated as
341 decimal, and I will make that clear in the documentation.
347 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
348 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
351 MH/01 Don't check that the operands of numeric comparison operators are
352 integers when their expansion is in "skipping" mode (fixes bug
353 introduced by 4.64-PH/07).
355 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
356 child addresses, Exim now panics and dies. Previously, because the count
357 is held in a short int, deliveries were likely to be lost. As such a
358 large number of recipients for a single message is ridiculous
359 (performance will be very, very poor), I have chosen to impose a limit
360 rather than extend the field.
366 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
367 leftover -K file (the existence of which was triggered by #402).
368 While we were at it, introduced process PID as part of the -K
369 filename. This should rule out race conditions when creating
372 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
373 processing considerably. Previous code took too long for large mails,
374 triggering a timeout which in turn triggers #401.
376 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
377 in the DK code in transports.c. sendfile() is not really portable,
378 hence the _LINUX specificness.
380 TF/01 In the add_headers option to the mail command in an Exim filter,
381 there was a bug that Exim would claim a syntax error in any
382 header after the first one which had an odd number of characters
385 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
386 callout verification, Exim cached a "reject" for the entire domain. This
387 is correct for most verifications, but it is not correct for a recipient
388 verification with use_sender or use_postmaster set, because in that case
389 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
390 case of MAIL FROM:<> rejection from other early rejections (e.g.
391 rejection of HELO). When verifying a recipient using a non-null MAIL
392 address, the cache is ignored if it shows MAIL FROM:<> rejection.
393 Whatever the result of the callout, the value of the domain cache is
394 left unchanged (for any other kind of callout, getting as far as trying
395 RCPT means that the domain itself is ok).
397 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
400 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
401 manifest itself as EPIPE rather than ECONNECT. When tidying away a
402 session, the daemon ignores ECONNECT errors and logs others; it now
403 ignores EPIPE as well.
405 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
406 (quoted-printable decoding).
408 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
409 later the small subsequent patch to fix an introduced bug.
411 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
413 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
415 PH/08 An error is now given if message_size_limit is specified negative.
417 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
418 to be given (somewhat) arbitrary names.
420 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
423 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
424 miscellaneous code fixes
426 PH/10 Added the log_reject_target ACL modifier to specify where to log
429 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
430 hostname. This is wrong, because it relates to the incoming message (and
431 probably the interface on which it is arriving) and not to the outgoing
432 callout (which could be using a different interface). This has been
433 changed to use the value of the helo_data option from the smtp transport
434 instead - this is what is used when a message is actually being sent. If
435 there is no remote transport (possible with a router that sets up host
436 addresses), $smtp_active_hostname is used.
438 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
439 tweaks were necessary in order to get it to work (see also 21 below):
440 (a) The code assumed that strncpy() returns a negative number on buffer
441 overflow, which isn't the case. Replaced with Exim's string_format()
443 (b) There were several signed/unsigned issues. I just did the minimum
444 hacking in of casts. There is scope for a larger refactoring.
445 (c) The code used strcasecmp() which is not a standard C function.
446 Replaced with Exim's strcmpic() function.
447 (d) The code set only $1; it now sets $auth1 as well.
448 (e) A simple test gave the error "authentication client didn't specify
449 service in request". It would seem that Dovecot has changed its
450 interface. Fortunately there's a specification; I followed it and
451 changed what the client sends and it appears to be working now.
453 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
456 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
457 address A is aliased to B and C, where B exists and C does not. Without
458 -v the output is "A verified" because verification stops after a
459 successful redirection if more than one address is generated. However,
460 with -v the child addresses are also verified. Exim was outputting "A
461 failed to verify" and then showing the successful verification for C,
462 with its parentage. It now outputs "B failed to verify", showing B's
463 parentage before showing the successful verification of C.
465 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
466 look up a TXT record in a specific list after matching in a combined
469 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
470 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
471 they consult the DNS. I had assumed they would set it the way they
472 wanted; and indeed my experiments on Linux seem to show that in some
473 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
474 To be on the safe side, however, I have now made the interface to
475 host_find_byname() similar to host_find_bydns(), with an argument
476 containing the DNS resolver options. The host_find_byname() function now
477 sets these options at its start, just as host_find_bydns() does. The smtp
478 transport options dns_qualify_single and dns_search_parents are passed to
479 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
480 of host_find_byname() use the default settings of RES_DEFNAMES
481 (qualify_single) but not RES_DNSRCH (search_parents).
483 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
484 spool_read_header() do less string testing, by means of a preliminary
485 switch on the second character of optional "-foo" lines. (This is
486 overdue, caused by the large number of possibilities that now exist.
487 Originally there were few.) While I was there, I also converted the
488 str(n)cmp tests so they don't re-test the leading "-" and the first
489 character, in the hope this might squeeze out yet more improvement.
491 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
492 flag allowing group syntax was set by the header_syntax check but not
493 turned off, possible causing trouble later; (2) The flag was not being
494 set at all for the header_verify test, causing "group"-style headers to
495 be rejected. I have now set it in this case, and also caused header_
496 verify to ignore an empty address taken from a group. While doing this, I
497 came across some other cases where the code for allowing group syntax
498 while scanning a header line wasn't quite right (mostly, not resetting
499 the flag correctly in the right place). These bugs could have caused
500 trouble for malformed header lines. I hope it is now all correct.
502 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
503 with the "reply" argument non-NULL. The code, however (which originally
504 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
505 but it didn't always do it. This confused somebody who was copying the
506 code for some other use. I have removed all the tests.
508 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
509 feature that was used to support insecure browsers during the U.S. crypto
510 embargo. It requires special client support, and Exim is probably the
511 only MTA that supported it -- and would never use it because real RSA is
512 always available. This code has been removed, because it had the bad
513 effect of slowing Exim down by computing (never used) parameters for the
514 RSA_EXPORT functionality.
516 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
517 authenticator to fail if there's a tab character in the incoming data
518 (there should never be unless someone is messing about, as it's supposed
519 to be base64-encoded). Also added, on Timo's advice, the "secured" option
520 if the connection is using TLS or if the remote IP is the same as the
521 local IP, and the "valid-client-cert option" if a client certificate has
524 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
525 authenticators. This can be used for authorization after authentication
526 succeeds. (In the case of plaintext, it servers for both authentication
529 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
530 if any retry times were supplied.
532 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
533 connection, where there is no client IP address to check. In this
534 situation, the verify now always succeeds.
536 PH/25 Applied John Jetmore's -Mset patch.
538 PH/26 Added -bem to be like -Mset, but loading a message from a file.
540 PH/27 In a string expansion for a processed (not raw) header when multiple
541 headers of the same name were present, leading whitespace was being
542 removed from all of them, but trailing whitespace was being removed only
543 from the last one. Now trailing whitespace is removed from each header
544 before concatenation. Completely empty headers in a concatenation (as
547 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
548 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
550 PH/29 [Removed. This was a change that I later backed out, and forgot to
551 correct the ChangeLog entry (that I had efficiently created) before
552 committing the later change.]
554 PH/30 Exim was sometimes attempting to deliver messages that had suffered
555 address errors (4xx response to RCPT) over the same connection as other
556 messages routed to the same hosts. Such deliveries are always "forced",
557 so retry times are not inspected. This resulted in far too many retries
558 for the affected addresses. The effect occurred only when there were more
559 hosts than the hosts_max_try setting in the smtp transport when it had
560 the 4xx errors. Those hosts that it had tried were not added to the list
561 of hosts for which the message was waiting, so if all were tried, there
562 was no problem. Two fixes have been applied:
564 (i) If there are any address or message errors in an SMTP delivery, none
565 of the hosts (tried or untried) are now added to the list of hosts
566 for which the message is waiting, so the message should not be a
567 candidate for sending over the same connection that was used for a
568 successful delivery of some other message. This seems entirely
569 reasonable: after all the message is NOT "waiting for some host".
570 This is so "obvious" that I'm not sure why it wasn't done
571 previously. Hope I haven't missed anything, but it can't do any
572 harm, as the worst effect is to miss an optimization.
574 (ii) If, despite (i), such a delivery is accidentally attempted, the
575 routing retry time is respected, so at least it doesn't keep
576 hammering the server.
578 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
579 in ${readsocket because some servers need this prod.
581 PH/32 Added some extra debug output when updating a wait-xxx database.
583 PH/33 The hint "could be header name not terminated by colon", which has been
584 given for certain expansion errors for a long time, was not being given
585 for the ${if def:h_colon_omitted{... case.
587 PH/34 The spec says: "With one important exception, whenever a domain list is
588 being scanned, $domain contains the subject domain." There was at least
589 one case where this was not true.
591 PH/35 The error "getsockname() failed: connection reset by peer" was being
592 written to the panic log as well as the main log, but it isn't really
593 panic-worthy as it just means the connection died rather early on. I have
594 removed the panic log writing for the ECONNRESET error when getsockname()
597 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
598 runs only) independently of the message's sender address. This meant
599 that, if the 4xx error was in fact related to the sender, a different
600 message to the same recipient with a different sender could confuse
601 things. In particualar, this can happen when sending to a greylisting
602 server, but other circumstances could also provoke similar problems.
603 I have changed the default so that the retry time for these errors is now
604 based a combination of the sender and recipient addresses. This change
605 can be overridden by setting address_retry_include_sender=false in the
608 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
609 remote server are returned as part of bounce messages. This was not
610 happening for LMTP over a pipe (the lmtp transport), but now it is the
611 same for both kinds of LMTP.
613 PH/38 Despite being documented as not happening, Exim was rewriting addresses
614 in header lines that were in fact CNAMEs. This is no longer the case.
616 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
617 and queue runs started by the daemon processed all messages. This has
618 been fixed so that -R and -S can now usefully be given with -q<time>.
620 PH/40 Import PCRE release 6.7 (fixes some bugs).
622 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
624 PH/42 Give an error if -q is specified more than once.
626 PH/43 Renamed the variables $interface_address and $interface_port as
627 $received_ip_address and $received_port, to make it clear that these
628 values apply to message reception, and not to the outgoing interface when
629 a message is delivered. (The old names remain recognized, of course.)
631 PH/44 There was no timeout on the connect() call when using a Unix domain
632 socket in the ${readsocket expansion. There now is.
634 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
635 be meaningful with "accept".
637 SC/01 Eximstats V1.43
638 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
640 SC/02 Eximstats V1.44
641 Use a glob alias rather than an array ref in the generated
642 parser. This improves both readability and performance.
644 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
645 Collect SpamAssassin and rejection statistics.
646 Don't display local sender or destination tables unless
647 there is data to show.
648 Added average volumes into the top table text output.
650 SC/04 Eximstats V1.46
651 Collect data on the number of addresses (recipients)
652 as well as the number of messages.
654 SC/05 Eximstats V1.47
655 Added 'Message too big' to the list of mail rejection
656 reasons (thanks to Marco Gaiarin).
658 SC/06 Eximstats V1.48
659 Mainlog lines which have GMT offsets and are too short to
660 have a flag are now skipped.
662 SC/07 Eximstats V1.49 (Alain Williams)
663 Added the -emptyok flag.
665 SC/08 Eximstats V1.50
666 Fixes for obtaining the IP address from reject messages.
668 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
669 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
670 whitesspace changes from 4.64-PH/27
672 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
675 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
676 are found, allow negative numbers in numeric criteria)
678 JJ/06 exipick.20061117.2, added new $message_body_missing variable
680 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
681 to match changes made in 4.64-PH/43
683 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
685 PH/47 Put in an explicit test for a DNS lookup of an address record where the
686 "domain" is actually an IP address, and force a failure. This locks out
687 those revolvers/nameservers that support "A-for-A" lookups, in
688 contravention of the specifications.
690 PH/48 When a host name was looked up from an IP address, and the subsequent
691 forward lookup of the name timed out, the host name was left in
692 $sender_host_name, contrary to the specification.
694 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
695 restricted to single-key lookups, Exim was not diagnosing an error if
696 * or *@ was used with a query-style lookup.
698 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
700 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
701 long ago, but noone interested enough thought of it. Let's just say that
702 the "1.1" means that there are some new functions that weren't there at
703 some point in the past.
705 PH/51 Error processing for expansion failure of helo_data from an smtp
706 transport during callout processing was broken.
708 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
709 tested/used via the -bh/-bhc/-bs options.
711 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
712 bug, fixed in subsequent PCRE releases).
714 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
715 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
717 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
723 SC/01 Use a glob alias rather than an array ref in eximstats generated
724 parser. This improves both readability and performance.
726 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
727 Don't display local sender or destination tables in eximstats unless
728 there is data to show.
729 Added average volumes into the eximstats top table text output.
731 SC/03 Collect data on the number of addresses (recipients) as well
732 as the number of messages in eximstats.
734 TF/01 Correct an error in the documentation for the redirect router. Exim
735 does (usually) call initgroups() when daemonizing.
737 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
738 with consistent privilege compared to when running as a daemon.
740 TF/03 Note in the spec that $authenticated_id is not set for local
741 submissions from trusted users.
743 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
744 Thanks to Dean Brooks <dean@iglou.com> for the patch.
746 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
747 by adding some example configuration directives to the default
748 configuration file. A little bit of work is required to uncomment the
749 directives and define how usernames and passwords are checked, but
750 there is now a framework to start from.
752 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
753 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
754 without this. I don't know how relevant this is to other LDAP libraries.
756 PH/02 Add the verb name to the "unknown ACL verb" error.
758 PH/03 Magnus Holmgren's patch for filter_prepend_home.
760 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
762 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
763 directory not expanded when it should be if an expanded home directory
764 was set for the address (which is overridden by the transport).
766 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
769 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
770 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
771 because it is too late at that time, and has no effect.
773 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
774 security issue with \' (bugzilla #107). I could not use the
775 PQescapeStringConn() function, because it needs a PGconn value as one of
778 PH/08 When testing addresses using -bt, indicate those final addresses that
779 are duplicates that would not cause an additional delivery. At least one
780 person was confused, thinking that -bt output corresponded to deliveries.
781 (Suppressing duplicates isn't a good idea as you lose the information
782 about possibly different redirections that led to the duplicates.)
784 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
785 systems where poll() doesn't work, in particular OS X.
787 PH/10 Added more information to debugging output for retry time not reached.
789 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
790 operations in malware.c.
792 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
795 PH/13 If write_rejectlog was set false when logging was sent to syslog with
796 syslog_duplication set false, log lines that would normally be written
797 both the the main log and to the reject log were not written to syslog at
800 PH/14 In the default configuration, change the use of "message" in ACL warn
801 statements to "add_header".
803 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
804 not followed by a command (e.g. "seen endif").
806 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
807 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
810 PH/17 Added extra conditions to the default value of delay_warning_condition
814 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
815 { match{$h_precedence:}{(?i)bulk|list|junk} } \
816 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
819 The Auto-Submitted: and various List- headers are standardised, whereas I
820 don't think Precedence: ever was.
822 PH/18 Refactored debugging code in route_finduser() to show more information,
823 in particular, the error code if getpwnam() issues one.
825 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
826 This is apparently needed in addition to the PH/07 change above to avoid
827 any possible encoding problems.
829 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
830 but not after initializing Perl.
832 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
833 output them only if debugging. By default they are written stderr,
834 apparently, which is not desirable.
836 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
839 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
842 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
844 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
845 authenticated or an ident call has been made. Suppress the default
846 values for $authenticated_id and $authenticated_sender (but permit -oMai
847 and -oMas) when testing with -bh.
849 PH/24 Re-jigged the order of the tests in the default configuration so that the
850 tests for valid domains and recipients precede the DNS black list and CSA
851 tests, on the grounds that those ones are more expensive.
853 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
854 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
855 This bug exists in every version of Exim that I still have, right back to
858 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
859 However, an attempt to turn on case-sensitivity in a regex key by
860 including (?-i) didn't work because the subject string was already
861 lowercased, and the effects were non-intuitive. It turns out that a
862 one-line patch can be used to allow (?-i) to work as expected.
868 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
869 other effects) broke the use of negated acl sub-conditions.
871 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
874 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
875 "Deny" causes Exim to reject the incoming connection with a 554 error.
876 Unfortunately, if there is a major crisis, such as a disk failure,
877 tcp-wrappers gives "deny", whereas what one would like would be some
878 kind of temporary error. A kludge has been added to help with this.
879 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
880 554 error is used if errno is still zero or contains ENOENT (which occurs
881 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
884 PH/03 Add -lutil to the default FreeBSD LIBS setting.
886 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
887 errors. Otherwise a message that provokes a temporary error (when other
888 messages do not) can cause a whole host to time out.
890 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
891 addresses were routed directly to files or pipes from a redirect router.
892 File deliveries just didn't batch; pipe deliveries might have suffered
895 PH/06 A failure to get a lock for a hints database would erroneously always say
896 "Failed to get write lock", even when it was really a read lock.
898 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
899 of 0600. This has been changed to use the value of the lockfile_mode
900 option (which defaults to 0600).
902 PH/08 Applied small patch from the Sieve maintainer.
904 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
905 folder from quota calculations, a direct delivery into this folder messed
906 up the contents of the maildirsize file. This was because the regex was
907 used only to exclude .Trash (or whatever) when the size of the mailbox
908 was calculated. There was no check that a delivery was happening into an
909 excluded directory. This bug has been fixed by ignoring all quota
910 processing for deliveries into excluded directories.
912 PH/10 Added the maildirfolder_create_regex option to appendfile.
918 PH/01 The code for finding all the local interface addresses on a FreeBSD
919 system running IPv6 was broken. This may well have applied to all BSD
920 systems, as well as to others that have similar system calls. The broken
921 code found IPv4 interfaces correctly, but gave incorrect values for the
922 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
923 that it would not match correctly against @[] and not recognize the IPv6
926 PH/02 The ipliteral router was not recognizing addresses of the form user@
927 [ipv6:....] because it didn't know about the "ipv6:" prefix.
929 PH/03 Added disable_ipv6.
931 PH/04 Changed $reply_address to use the raw form of the headers instead of the
932 decoded form, because it is most often used to construct To: headers
933 lines in autoreplies, and the decoded form may well be syntactically
934 invalid. However, $reply_address has leading white space removed, and all
935 newlines turned into spaces so that the autoreply transport does not
938 PH/05 If group was specified without a user on a router, and no group or user
939 was specified on a transport, the group from the router was ignored.
941 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
942 for visible compile-time settings that can be used to change these
943 numbers, for those that want even more. Backwards compatibility with old
944 spool files has been maintained. However, going back to a previous Exim
945 release will lost any variables that are in spool files.
947 PH/07 Two small changes when running in the test harness: increase delay when
948 passing a TCP/IP connection to a new process, in case the original
949 process has to generate a bounce, and remove special handling of
950 127.0.0.2 (sic), which is no longer necessary.
952 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
953 be the same on different OS.
955 PH/09 Moved a debug statement in filter processing to avoid a race problem when
958 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
959 whether --show-vars was specified or not
961 JJ/02 exipick: Added support for new ACL variable spool format introduced
964 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
965 syntactically invalid From: or Reply-to: line, and a filter used this to
966 generate an autoreply, and therefore failed to obtain an address for the
967 autoreply, Exim could try to deliver to a non-existent relative file
968 name, causing unrelated and misleading errors. What now happens is that
969 it logs this as a hard delivery error, but does not attempt to create a
972 PH/11 The exinext utility has a -C option for testing purposes, but although
973 the given file was scanned by exinext itself; it wasn't being passed on
974 when Exim was called.
976 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
977 an end-of-file indication when reading a command response.
979 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
980 compiled. In many other places in Exim, IPv6 addresses are always
981 recognized, so I have changed this. It also means that IPv4 domain
982 literals of the form [IPV4:n.n.n.n] are now always recognized.
984 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
985 used if the router is not running as root, for example, when verifying at
986 ACL time, or when using -bh. The debugging output from this situation was
987 non-existent - all you got was a failure to exec. I have made two
990 (a) Failures to set uid/gid, the current directory, or a process leader
991 in a subprocess such as that created by queryprogram now generate
992 suitable debugging ouput when -d is set.
994 (b) The queryprogram router detects when it is not running as root,
995 outputs suitable debugging information if -d is set, and then runs
996 the subprocess without attempting to change uid/gid.
998 PH/15 Minor change to Makefile for building test_host (undocumented testing
1001 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
1002 additional section of a DNS packet that returns MX or SRV records.
1003 Instead, it always explicitly searches for A/AAAA records. This avoids
1004 major problems that occur when a DNS server includes only records of one
1005 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
1006 fixed another bug: if SRV records were looked up and the corresponding
1007 address records were *not* found in the additional section, the port
1008 values from the SRV records were lost.
1010 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1011 using the correct key (the original address) when searching the retry
1012 rules in order to find which one to use for generating the retry hint.
1014 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1015 adding the default one. Similarly, if it contains a Reply-To: header, the
1016 errors_reply_to option, if set, is not used.
1018 PH/19 When calculating a retry time, Exim used to measure the "time since
1019 failure" by looking at the "first failed" field in the retry record. Now
1020 it does not use this if it is later than than the arrival time of the
1021 message. Instead it uses the arrival time. This makes for better
1022 behaviour in cases where some deliveries succeed, thus re-setting the
1023 "first failed" field. An example is a quota failure for a huge message
1024 when small messages continue to be delivered. Without this change, the
1025 "time since failure" will always be short, possible causing more frequent
1026 delivery attempts for the huge message than are intended.
1027 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1029 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1030 $1, $2, $3) because the numerical variables can be reset during some
1031 expansion items (e.g. "match"), thereby losing the authentication data.
1033 PH/21 Make -bV show the size of off_t variables so that the test suite can
1034 decide whether to run tests for quotas > 2G.
1036 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1037 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1038 filecount value is greater than 2G or if a quota value is greater than 2G
1039 on a system where the size of off_t is not greater than 4, a panic error
1042 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1043 never match. The debug and -bh output now contains an explicit error
1044 message indicating a malformed IPv4 address or mask.
1046 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1047 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1048 PH/23 above applies.
1050 PH/25 Do not write to syslog when running in the test harness. The only
1051 occasion when this arises is a failure to open the main or panic logs
1052 (for which there is an explicit test).
1054 PH/26 Added the /no_tell option to "control=freeze".
1056 PH/27 If a host name lookup failed very early in a connection, for example, if
1057 the IP address matched host_lookup and the reverse lookup yielded a name
1058 that did not have a forward lookup, an error message of the form "no IP
1059 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1060 could be logged. Now it outputs the IP address instead of "NULL".
1062 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1063 allows the sender and the authenticated sender to be set when
1064 submitting a message from within Exim. Since child_open_exim() is
1065 documented for local_scan(), the new function should be too.
1067 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1068 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1069 results in an empty string is now treated as unset.
1071 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1073 PH/31 Added sender_verify_fail logging option.
1075 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1076 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1077 tidied the source and removed it altogether.
1079 PH/33 When a queue run was abandoned because the load average was too high, a
1080 log line was always written; now it is written only if the queue_run log
1081 selector is set. In addition, the log line for abandonment now contains
1082 information about the queue run such as the pid. This is always present
1083 in "start" and "stop" lines but was omitted from the "abandon" line.
1085 PH/34 Omit spaces between a header name and the colon in the error message that
1086 is given when verify = headers_syntax fails (if there are lots of them,
1087 the message gets confusing).
1089 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1090 names, as there are now some PTR records that contain slashes. This check
1091 is only to protect against broken name servers that fall over on strange
1092 characters, so the fact that it applies to all lookups doesn't matter.
1094 PH/36 Now that the new test suite is complete, we can remove some of the
1095 special code in Exim that was needed for the old test suite. For example,
1096 sorting DNS records because real resolvers return them in an arbitrary
1097 order. The new test suite's fake resolver always returns records in the
1100 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1101 bounces) except when queue_only is set, to avoid logging races between
1102 the different processes.
1104 PH/38 Panic-die if .include specifies a non-absolute path.
1106 PH/39 A tweak to the "H" retry rule from its user.
1108 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1109 a label. They prevented compilation on older perls.
1111 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1112 a warning to be raised on newish perls.
1114 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1115 on queue. Changes to match documented behaviour of showing count of
1116 messages matching specified criteria.
1118 PH/40 Changed the default ident timeout from 30s to 5s.
1120 PH/41 Added support for the use of login_cap features, on those BSD systems
1121 that have them, for controlling the resources used by pipe deliveries.
1123 PH/42 The content-scanning code uses fopen() to create files in which to put
1124 message data. Previously it was not paying any attention to the mode of
1125 the files. Exim runs with umask(0) because the rest of the code creates
1126 files with open(), and sets the required mode explicitly. Thus, these
1127 files were ending up world-writeable. This was not a big issue, because,
1128 being within the spool directory, they were not world-accessible. I have
1129 created a function called modefopen, which takes an additional mode
1130 argument. It sets umask(777), creates the file, chmods it to the required
1131 mode, then resets the umask. All the relevant calls to fopen() in the
1132 content scanning code have been changed to use this function.
1134 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1135 to 24 hours. This avoids potential overflow problems when processing G
1136 and H retry rules. I suspect nobody ever tinkers with this value.
1138 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1140 PH/45 When the plaintext authenticator is running as a client, the server's
1141 challenges are checked to ensure they are valid base64 strings. By
1142 default, the authentication attempt is cancelled if an invalid string is
1143 received. Setting client_ignore_invalid_base64 true ignores these errors.
1144 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1145 they are received. Thus, the responses can be made to depend on the
1146 challenges. If an invalid string is ignored, an empty string is placed in
1149 PH/46 Messages that are created by the autoreply transport now contains a
1150 References: header, in accordance with RFCs 2822 and 3834.
1152 PH/47 Added authenticated_sender_force to the smtp transport.
1154 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1156 PH/49 Installed latest patch from the Sieve maintainer.
1158 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1159 was also set, the appendfile transport was unnecessarily scanning a
1160 directory of message files (e.g. for maildir delivery) to find the count
1161 of files (along with the size), even though it did not need this
1162 information. It now does the scan only if it needs to find either the
1163 size of the count of files.
1165 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1167 PH/52 Two bugs concerned with error handling when the smtp transport is
1170 (i) Exim was not creating retry information for temporary errors given
1171 for individual recipients after the DATA command when the smtp transport
1172 was used in LMTP mode. This meant that they could be retried too
1173 frequently, and not timed out correctly.
1175 (ii) Exim was setting the flag that allows error details to be returned
1176 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1177 recipients that were returned after the DATA command.
1179 PH/53 This is related to PH/52, but is more general: for any failing address,
1180 when detailed error information was permitted to be returned to the
1181 sender, but the error was temporary, then after the final timeout, only
1182 "retry timeout exceeded" was returned. Now it returns the full error as
1183 well as "retry timeout exceeded".
1185 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1186 do this, and (what is worse) MTAs that accept it.
1188 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1189 will now be deprecated.
1191 PH/56 New os.c-cygwin from the Cygwin maintainer.
1193 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1194 formats (previously only available in exim formats via -bpr, -bpru,
1195 and -bpra. Now also available in native and exiqgrep formats)
1197 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1198 with very large, slow to parse queues
1200 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1202 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1204 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1205 responses to authentication challenges, though it was showing the
1206 challenges; (ii) I've removed the CR characters from the debug output for
1209 PH/58 Allow for the insertion of a newline as well as a space when a string
1210 is turned into more than one encoded-word during RFC 2047 encoding. The
1211 Sieve code now uses this.
1213 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1214 data_4xx, lost_connection, tls_required.
1216 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1217 message was being sent as an SMTP response.
1219 PH/61 Add -l and -k options to exicyclog.
1221 PH/62 When verifying, if an address was redirected to one new address, so that
1222 verification continued, and the new address failed or deferred after
1223 having set something in $address_data, the value of $address_data was not
1224 passed back to the ACL. This was different to the case when no
1225 redirection occurred. The value is now passed back in both cases.
1227 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1228 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1229 use login_cap.h, so on its own it isn't the distinguishing feature. The
1230 new name refers directly to the setclassresources() function.
1232 PH/65 Added configuration files for NetBSD3.
1234 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1236 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1237 is preferred over IPv4.
1239 PH/68 The bounce_return_message and bounce_return_body options were not being
1240 honoured for bounces generated during the reception of non-SMTP messages.
1241 In particular, this applied to messages rejected by the ACL. This bug has
1242 been fixed. However, if bounce_return_message is true and bounce_return_
1243 body is false, the headers that are returned for a non-SMTP message
1244 include only those that have been read before the error was detected.
1245 (In the case of an ACL rejection, they have all been read.)
1247 PH/69 The HTML version of the specification is now built in a directory called
1248 spec_html instead of spec.html, because the latter looks like a path with
1249 a MIME-type, and this confuses some software.
1251 PH/70 Catch two compiler warnings in sieve.c.
1253 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1254 function verify_get_ident() calls ip_connect() to connect a socket, but
1255 if the "connect()" function timed out, ip_connect() used to close the
1256 socket. However, verify_get_ident() also closes the socket later, and in
1257 between Exim writes to the log, which may get opened at this point. When
1258 the socket was closed in ip_connect(), the log could get the same file
1259 descriptor number as the socket. This naturally causes chaos. The fix is
1260 not to close the socket in ip_connect(); the socket should be closed by
1261 the function that creates it. There was only one place in the code where
1262 this was missing, in the iplookup router, which I don't think anybody now
1263 uses, but I've fixed it anyway.
1265 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1266 well as to direct DNS lookups. Otherwise the handling of names in host
1267 lists is inconsistent and therefore confusing.
1273 PH/01 Two changes to the default runtime configuration:
1275 (1) Move the checks for relay_from_hosts and authenticated clients from
1276 after to before the (commented out) DNS black list checks.
1278 (2) Add control=submission to the relay_from_hosts and authenticated
1279 clients checks, on the grounds that messages accepted by these
1280 statements are most likely to be submissions.
1282 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1284 (1) Generate an error if the third argument for the ${prvs expansion is
1287 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1290 (3) Reset the variables that are obtained from the first argument of
1291 ${prvscheck and used in the second argument before leaving the code,
1292 because their memory is reclaimed, so using them afterwards may do
1295 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1296 one (it's much easier than Tom thought :-).
1298 (5) Because of (4), we can now allow for the use of $prvscheck_result
1299 inside the third argument.
1301 PH/03 For some reason, the default setting of PATH when running a command from
1302 a pipe transport was just "/usr/bin". I have changed it to
1305 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1306 anything to be listed in the output from -bV.
1308 PH/05 When a filter generated an autoreply, the entire To: header line was
1309 quoted in the delivery log line, like this:
1311 => >A.N.Other <ano@some.domain> <original@ddress> ...
1313 This has been changed so that it extracts the operative address. There
1314 may be more than one such address. If so, they are comma-separated, like
1317 => >ano@some.domain,ona@other.domain <original@ddress> ...
1319 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1320 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1321 not being looked up in the rDNS to get a host name, Exim was showing the
1322 IP address twice in Received: lines, even though the IP addresses were
1323 identical. For example:
1325 Received: from [1.2.3.4] (helo=[1.2.3.4])
1327 However, if the real host name was known, it was omitting the HELO data
1328 if it matched the actual IP address. This has been tidied up so that it
1329 doesn't show the same IP address twice.
1331 PH/07 When both +timestamp and +memory debugging was on, the value given by
1332 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1333 called by the debug printing, thereby overwriting the timestamp buffer.
1334 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1336 PH/08 When the original message was included in an autoreply transport, it
1337 always said "this is a copy of the message, including all the headers",
1338 even if body_only or headers_only was set. It now gives an appropriate
1341 PH/09 Applied a patch from the Sieve maintainer which:
1343 o fixes some comments
1344 o adds the (disabled) notify extension core
1345 o adds some debug output for the result of if/elsif tests
1346 o points to the current vacation draft in the documentation
1347 and documents the missing references header update
1351 o fixes a bug in processing the envelope test (when testing
1352 multiple envelope elements, the last element determinted the
1355 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1356 Electronic Mail") by including:
1358 Auto-submitted: auto-generated
1360 in the messages that it generates (bounce messages and others, such as
1361 warnings). In the case of bounce messages for non-SMTP mesages, there was
1362 also a typo: it was using "Auto_submitted" (underscore instead of
1363 hyphen). Since every message generated by Exim is necessarily in response
1364 to another message, thes have all been changed to:
1366 Auto-Submitted: auto-replied
1368 in accordance with these statements in the RFC:
1370 The auto-replied keyword:
1372 - SHOULD be used on messages sent in direct response to another
1373 message by an automatic process,
1375 - MUST NOT be used on manually-generated messages,
1377 - MAY be used on Delivery Status Notifications (DSNs) and Message
1378 Disposition Notifications (MDNs),
1380 - MUST NOT be used on messages generated by automatic or periodic
1381 processes, except for messages which are automatic responses to
1384 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1385 to the default Received: header definition.
1387 PH/12 Added log selector acl_warn_skipped (default on).
1389 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1390 variables because (a) they are in the wrong storage pool and (b) even if
1391 they were copied, it wouldn't work properly because of the caching.
1393 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1394 checking when decoding. Apparently there are clients that generate
1395 overlong encoded strings. Why am I not surprised?
1397 PH/15 If the first argument of "${if match_address" was not empty, but did not
1398 contain an "@" character, Exim crashed. Now it writes a panic log message
1399 and treats the condition as false.
1401 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1403 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1404 extension "envelope-auth". The code is finished and in agreement with
1405 other implementations, but there is no documentation so far and in fact,
1406 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1407 not changing the active code.
1409 Print executed "if" and "elsif" statements when debugging is used. This
1410 helps a great deal to understand what a filter does.
1412 Document more things not specified clearly in RFC3028. I had all this
1413 sorted out, when out of a sudden new issues came to my mind. Oops."
1415 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1418 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1419 canonical form (as documented). However, after a host name lookup from
1420 the IP address, check_host() was doing a simple string comparison with
1421 addresses acquired from the DNS when checking that the found name did
1422 have the original IP as one of its addresses. Since any found IPv6
1423 addresses are likely to be in abbreviated form, the comparison could
1424 fail. Luckily, there already exists a function for doing the comparison
1425 by converting both addresses to binary, so now that is used instead of
1426 the text comparison.
1428 PH/20 There was another similar case to PH/19, when a complete host name was
1429 given in a host list; looking up its IP address could give an abbreviated
1430 form, whereas the current host's name might or might not be abbreviated.
1431 The same fix has been applied.
1437 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1438 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1441 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1442 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1444 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1446 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1447 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1448 Also, for "Auto-Submitted", treat anything other than "no" as
1449 non-personal, in accordance with RFC 3834. (Previously it treated
1450 anything starting "auto-" as non-personal.)
1452 TF/01 The control=submission/name=... option had a problem with syntax
1453 errors if the name included a slash character. The /name= option
1454 now slurps the rest of the string, so it can include any characters
1455 but it must come last in the list of options (after /sender_retain
1458 PH/05 Some modifications to the interface to the fake nameserver for the new
1466 TK/01 Added the "success_on_redirect" address verification option. See
1467 NewStuff for rationale and an example.
1469 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1471 PH/02 Patch to exigrep to allow it to work on syslog lines.
1473 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1474 fread() to skip over the body file's header line, because in Cygwin the
1475 header line is locked and is inaccessible.
1477 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1478 co-exist for some time) to make it clear that it is the Exim ID that is
1479 referenced, not the Message-ID: header line.
1481 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1482 string_format() function, because snprintf() does not exist on all
1485 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1486 ${stat: expansion item.
1488 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1489 protocol synchronization error", to keep the pedants happy.
1491 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1492 well as for IRIX systems, when gcc is being used. See the host.c source
1495 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1497 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1500 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1501 transport and to the smtp transport in LMTP mode.
1503 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1505 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1506 run a filter in a subprocess. This could lead to confusion in subsequent
1507 lookups in the parent process. There should also be a search_tidyup() at
1508 the end of the subprocess.
1510 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1511 only if the host matched helo_try_verify_hosts, which caused the
1512 verification to occur when the EHLO/HELO command was issued. The ACL just
1513 tested the remembered result. Now, if a previous verification attempt has
1514 not happened, "verify = helo" does it there and then.
1516 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1518 TK/03 Fix log output including CR from clamd.
1520 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1521 exist provoked a memory error which could cause a segfault.
1523 PH/15 Installed PCRE 6.2
1525 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1527 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1528 of the problem. Specifically, suggested +O2 rather than +O1 for the
1531 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1533 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1534 record, the port was not correctly logged when the outgoing_port log
1535 selector was set (it logged the transort's default port).
1537 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1538 fallback_hosts, and "hosts" in the smtp transport.
1540 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1541 host errors such as "Connection refused".
1543 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1544 authentication with radiusclient 0.4.9:
1546 - Error returned from rc_read_config was caught wrongly
1547 - Username/password not passed on to radius server due to wrong length.
1549 The presumption is that some radiusclient API changes for 4.51/PH/17
1550 were not taken care of correctly. The code is still untested by me (my
1551 Linux distribution still has 0.3.2 of radiusclient), but it was
1552 contributed by a Radius user.
1554 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1555 expanding the "port" option of the smtp transport.
1557 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1558 while reading a MIME header. Thanks to Tom Hughes for a patch.
1560 PH/24 Include config.h inside local_scan.h so that configuration settings are
1563 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1564 that in an ACL for RCPT (for example), you can examine exactly what was
1567 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1568 commands, but it was not correctly comparing the address with the actual
1569 client host address. Thus, it would show the EHLO address in Received:
1570 header lines when this was not necessary.
1572 PH/27 Added the % operator to ${eval:}.
1574 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1575 it should be ignoring failures (because with -C, for example, it has lost
1576 privilege). It wasn't ignoring creation failures other than "already
1579 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1582 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1583 because an input line was too long, either on its own, or by virtue of
1584 too many continuations, the temporary file was not being removed, and the
1585 return code was incorrect.
1587 PH/31 Missing "BOOL" in function definition in filtertest.c.
1589 PH/32 Applied Sieve patches from the maintainer.
1591 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1593 PH/33 Added "verify = not_blind".
1595 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1596 Local/Makefile (with some defaults set). These are used in built scripts
1597 such as exicyclog, but they have never been used in the exim_install
1598 script (though there are many overriding facilities there). I have
1599 arranged that the exim_install script now takes note of these two
1602 PH/35 Installed configuration files for Dragonfly.
1604 PH/36 When a locally submitted message by a trusted user did not contain a
1605 From: header, and the sender address was obtained from -f or from an SMTP
1606 MAIL command, and the trusted user did not use -F to supply a sender
1607 name, $originator_name was incorrectly used when constructing a From:
1608 header. Furthermore, $originator_name was used for submission mode
1609 messages from external hosts without From: headers in a similar way,
1610 which is clearly wrong.
1612 PH/37 Added control=suppress_local_fixups.
1614 PH/38 When log_selector = +received_sender was set, and the addition of the
1615 sender made the log line's construction buffer exactly full, or one byte
1616 less than full, an overflow happened when the terminating "\n" was
1619 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1620 when the result of a list match is failure because a DNS lookup failed.
1622 PH/40 RM_COMMAND is now used in the building process.
1624 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1625 the "build-* directories that it finds.
1627 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1628 address in a domain literal was a prefix of an interface address.
1630 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1631 when verifying a sender address, unless rewrite_headers is false.
1633 PH/44 Wrote a long comment about why errors_to addresses are verified as
1634 recipients, not senders.
1636 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1637 the ratelimit ACL was added.
1639 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1641 PH/46 Added extra information about PostgreSQL errors to the error string.
1643 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1644 avoiding the need to install special zones in a real server. This is
1645 backwards compatible; if it can't find the fake resolver, it drops back.
1646 Thus, both old and new test suites can be run.
1648 TF/02 Added util/ratelimit.pl
1650 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1651 clock is set back in time.
1653 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1654 Candler <B.Candler@pobox.com>.
1656 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1657 OK for addresses that are the result of redirections.
1659 PH/48 A number of further additions for the benefit of the new test suite,
1660 including a fake gethostbyname() that interfaces to the fake DNS resolver
1663 TF/06 The fix for widen_domains has also been applied to qualify_single and
1664 search_parents which are the other dnslookup options that can cause
1667 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1670 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1672 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1673 service" if it encountered an MX record with an empty target hostname.
1674 The message is now "an MX or SRV record indicated no SMTP service".
1676 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1677 if the DNS of the sending site is misconfigured. This is quite a
1678 common situation. This change restores the behaviour of treating a
1679 helo verification defer as a failure.
1681 PH/51 If self=fail was set on a router, the bounce message did not include the
1682 actual error message.
1688 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1690 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1691 command itself ended in error, the underlying message about the transport
1692 filter timeout was being overwritten with the pipe command error. Now the
1693 underlying error message should be appended to the second error message.
1695 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1697 PH/02 Reduce the amount of output that "make" produces by default. Full output
1698 can still be requested.
1700 PH/03 The warning log line about a condition test deferring for a "warn" verb
1701 was being output only once per connection, rather than after each
1702 occurrence (because it was using the same function as for successful
1703 "warn" verbs). This seems wrong, so I have changed it.
1705 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1706 it should not have, which might have caused a crash in the right
1707 circumstances, but probably never did.
1709 PH/04 Installed a modified version of Tony Finch's patch to make submission
1710 mode fix the return path as well as the Sender: header line, and to
1711 add a /name= option so that you can make the user's friendly name appear
1714 TF/03 Added the control = fakedefer ACL modifier.
1716 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1717 Mark Lowes for thorough testing.
1719 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1721 TK/03 Merged latest SRS patch from Miles Wilton.
1723 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1724 with the definition in sysexits.h (which is #included earlier).
1725 Fortunately, Exim does not actually use EX_OK. The code used to try to
1726 preserve the sysexits.h value, by assumimg that macro definitions were
1727 scanned for macro replacements. I have been disabused of this notion,
1728 so now the code just undefines EX_OK before #including unistd.h.
1730 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1731 in the smtp transport. When a block could not be written in a single
1732 write() function, the timeout was being re-applied to each part-write.
1733 This seems wrong - if the receiver was accepting one byte at a time it
1734 would take for ever. The timeout is now adjusted when this happens. It
1735 doesn't have to be particularly precise.
1737 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1738 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1740 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1741 without a domain if the check to <postmaster@domain> fails.
1743 SC/01 Eximstats: added -xls and the ability to specify output files
1744 (patch written by Frank Heydlauf).
1746 SC/02 Eximstats: use FileHandles for outputing results.
1748 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1750 SC/04 Eximstats: fixed display of large numbers with -nvr option
1752 SC/05 Eximstats: fixed merging of reports with empty tables.
1754 SC/06 Eximstats: added the -include_original_destination flag
1756 SC/07 Eximstats: removed tabs and trailing whitespace.
1758 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1760 TK/06 MBOX spool code: Add real "From " MBOX separator line
1761 so the .eml file is really in mbox format (even though
1762 most programs do not really care). Patch from Alex Miller.
1764 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1765 The latter is generated from $received_to and is only set if the
1766 message has one envelope recipient. SA can use these headers,
1767 obviously out-of-the-box. Patch from Alex Miller.
1769 PH/08 The ${def test on a variable was returning false if the variable's
1770 value was "0", contrary to what the specification has always said!
1771 The result should be true unless the variable is empty.
1773 PH/09 The syntax error of a character other than { following "${if
1774 def:variable_name" (after optional whitespace) was not being diagnosed.
1775 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1776 accidental colon was present, for example, could give incorrect results.
1778 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1779 result is used (not including appendfile, where other changes are about
1782 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1783 This involved changing a lot of size variables from int to off_t. It
1784 should work with maildirs and everything.
1786 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1787 spamd dying while we are connected to it.
1789 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1792 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1793 function for BDB 4.3.
1795 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1797 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1798 the sources for PCRE have been re-arranged and more files are now
1801 PH/15 The code I had for printing potentially long long variables in PH/11
1802 above was not the best (it lost precision). The length of off_t variables
1803 is now inspected at build time, and an appropriate printing format (%ld
1804 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1805 to be "long long int" or "long int". This is needed for the internal
1806 formatting function string_vformat().
1808 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1809 the configuration file to be ":syslog", then the script "guesses" where
1810 the logs files are, rather than using the compiled in default. In our
1811 case the guess is not the same as the compiled default, so the script
1812 suddenly stopped working when I started to use syslog. The patch checks
1813 to see if log_file_path is "". If so, it attempts to read it from exim
1814 with no configuration file to get the compiled in version, before it
1815 falls back to the previous guessing code."
1817 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1818 implementing BATV in an Exim configuration. See NewStuff for the gory
1821 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1822 Makefile that are specific to HP-UX.
1824 PH/18 If the "use_postmaster" option was set for a recipient callout together
1825 with the "random" option, the postmaster address was used as the MAIL
1826 FROM address for the random test, but not for the subsequent recipient
1827 test. It is now used for both.
1829 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1830 patch removes a few documentation additions to RFC 3028, because the
1831 latest draft now contains them. It adds the new en;ascii-case comparator
1832 and a new error check for 8bit text in MIME parts. Comparator and
1833 require names are now matched exactly. I enabled the subaddress
1834 extension, but it is not well tested yet (read: it works for me)."
1836 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1837 rework some of the code of TK/09 above to avoid the hardwired use of
1838 "%lld" and "long long". Replaced the call to snprintf() with a call to
1841 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1842 records point to non-existent hosts", "retry timeout exceeded", and
1843 "retry time not reached for any host after a long failure period".
1845 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1846 experimental DomainKeys support:
1848 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1849 (2) On an error such as an illegally used "control", the wrong name for
1850 the control was given.
1852 These problems did NOT occur unless DomainKeys support was compiled.
1854 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1856 PH/24 Added ${if match_ip condition.
1858 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1859 databases so that it will be absolutely obvious if a crash occurs in the
1860 DB library. This is a regular occurrence (often caused by mis-matched
1863 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1864 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1865 picked up on a user's system that detects such things. There doesn't seem
1866 to be a gcc warning option for this - only an attribute that has to be
1867 put on the function's prototype. It seems that in Fedora Core 4 they have
1868 set this on a number of new functions. No doubt there will be more in due
1871 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1872 specify a transport. However, if an address that was verified by such a
1873 router was the subject of a callout, Exim crashed because it tried to
1874 read the rcpt_include_affixes from the non-existent transport. Now it
1875 just assumes that the setting of that option is false. This bug was
1876 introduced by 4.51/PH/31.
1878 PH/28 Changed -d+all to exclude +memory, because that information is very
1879 rarely of interest, but it makes the output a lot bigger. People tend to
1880 do -d+all out of habit.
1882 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1883 code in os-type was giving problems when libc.so lives in lib64, like on
1886 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1887 aren't the modern standard, and it seems that some systems' include files
1888 don't always have them. Exim was already checking for some of the newer
1889 ones like T_AAAA, and defining it itself. I've added checks for all the
1890 record types that Exim uses.
1892 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1893 not automatically generating a new one, as it is supposed to. This
1894 prevented TLS from working. If the file did exist, but contained invalid
1895 data, a new version was generated, as expected. It was only the case of a
1896 non-existent file that was broken.
1898 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1899 with a change in libdomainkeys > 0.64.
1901 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1902 from DNS. If the selector record carries the flag, it now has
1903 precedence over the domain-wide flag.
1905 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1907 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1908 the use of a port name that isn't defined in /etc/services) occurred, the
1909 message was deferred as in a normal delivery, and thus remained on the
1910 spool, instead of being failed because of the mua_wrapper setting. This
1911 is now fixed, and I tidied up some of the mua_wrapper messages at the
1914 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1915 the messages in a hash of arrays rather than using individual hashes.
1916 This is a bit cleaner and results in dramatic memory savings, albeit
1917 at a slight CPU cost.
1919 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1920 as requested by Marc Sherman.
1922 SC/10 Eximstats: added histograms for user specified patterns as requested
1925 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1927 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1928 fopen() in the content-scanning modules that did not already have it.
1934 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1935 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1937 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1939 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1941 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1942 to test Sieve filters that use "vacation".
1944 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1945 that changes the way the GnuTLS parameters are stored in the cache file.
1946 The new format can be generated externally. For backward compatibility,
1947 if the data in the cache doesn't make sense, Exim assumes it has read an
1948 old-format file, and it generates new data and writes a new file. This
1949 means that you can't go back to an older release without removing the
1952 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1953 work if there are any delivery delays because "one_time" forces the
1954 parent to be marked "delivered", so its unseen clone is never tried
1955 again. For this reason, Exim now forbids the simultaneous setting of
1958 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1959 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1960 entry if you want to know the details. The fix, however, neglected to
1961 consider the case when local delivery batching is involved. The test for
1962 "previously delivered" was not happening when checking to see if an
1963 address could be batched with a previous (undelivered) one; under
1964 certain circumstances this could lead to multiple deliveries to the same
1967 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1968 in its include files, and this causes problems building Exim.
1970 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1971 header_syntax) but Exim was just ignoring anything given after a slash.
1972 In particular, this caused confusion with an attempt to use "verify =
1973 reverse_host_lookup/defer_ok". An error is now given when options are
1974 supplied for verify items that do not have them. (Maybe reverse_host_
1975 lookup should have a defer_ok option, but that's a different point.)
1977 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1978 defined by RFC 821) to 2048, because there were problems with some AUTH
1979 commands, and RFC 1869 says the size should be increased for extended
1980 SMTP commands that take arguments.
1982 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1985 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1986 "unknown" error; now it says that the functionality isn't in the binary.
1988 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1989 an address' error message when a string expansion fails (syntax or
1990 whatever). Otherwise the password may appear in the log. Following change
1991 PH/42 below, there is no longer a chance of it appearing in a bounce
1994 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1996 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1997 of its addresses was ever tried. (Bugzilla bug #2.)
1999 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
2000 the result incorrectly in the debug output. (It correctly added a newline
2001 to what was transported.)
2003 TF/01 Added $received_time.
2005 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
2006 commented out examples of how to interface to a virus scanner and to
2007 SpamAssassin. Also added commented examples of av_scanner and
2008 spamd_address settings.
2010 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2011 and controls are allowed in which ACLs. There were a couple of minor
2012 errors. Some of the entries in the conditions table (which is a table of
2013 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2014 negation of where the condition IS allowed.
2016 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2018 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2019 header file does not have a version number, so I've had to invent a new
2020 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2021 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2022 radiusclient), but it was contributed by a Radius user.
2024 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2025 files or directories, for OpenSSL.
2027 PH/19 When an Exim process that is running as root has to create an Exim log
2028 file, it does so in a subprocess that runs as exim:exim so as to get the
2029 ownership right at creation (otherwise, other Exim processes might see
2030 the file with the wrong ownership). There was no test for failure of this
2031 fork() call, which would lead to the process getting stuck as it waited
2032 for a non-existent subprocess. Forks do occasionally fail when resources
2033 run out. I reviewed all the other calls to fork(); they all seem to check
2036 PH/20 When checking for unexpected SMTP input at connect time (before writing
2037 the banner), Exim was not dealing correctly with a non-positive return
2038 from the read() function. If the client had disconnected by this time,
2039 the result was a log entry for a synchronization error with an empty
2040 string after "input=" when read() returned zero. If read() returned -1
2041 (an event I could not check), uninitialized data bytes were printed.
2042 There were reports of junk text (parts of files, etc) appearing after
2045 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2047 PH/22 Added support for macro redefinition, and (re)definition in between
2048 driver and ACL definitions.
2050 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2051 forgetting to use the resulting value; it was using the unexpanded value.
2053 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2054 hadn't been configured. The fix is from Juergen Kreileder, who
2055 understands it better than I do:
2057 "Here's what I see happening with three configured cyrus_sasl
2058 authenticators configured (plain, login, cram-md5):
2060 On startup auth_cyrus_sasl_init() gets called for each of these.
2061 This means three calls to sasl_listmech() without a specified mech_list.
2062 => SASL tests which mechs of all available mechs actually work
2063 => three warnings about OTP not working
2064 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2066 With the patch, sasl_listmech() also gets called three times. But now
2067 SASL's mech_list option is set to the server_mech specified in the the
2068 authenticator. Or in other words, the answer from sasl_listmech()
2069 gets limited to just the mech you're testing for (which is different
2071 => the return list contains just 'plain' or 'login', 'cram-md5' or
2072 nothing depending on the value of ob->server_mech.
2074 I've just tested the patch: Authentication still works fine,
2075 unavailable mechs specified in the exim configuration are still
2076 caught, and the auth.log warnings about OTP are gone."
2078 PH/25 When debugging is enabled, the contents of the command line are added
2079 to the debugging output, even when log_selector=+arguments is not
2082 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2083 answer is "GNU", and only if the return is "GNU/something" is the answer
2086 PH/27 $acl_verify_message is now set immediately after the failure of a
2087 verification in an ACL, and so is available in subsequent modifiers. In
2088 particular, the message can be preserved by coding like this:
2090 warn !verify = sender
2091 set acl_m0 = $acl_verify_message
2093 Previously, $acl_verify_message was set only while expanding "message"
2094 and "log_message" when a very denied access.
2096 PH/28 Modified OS/os.c-Linux with
2098 -#ifndef OS_LOAD_AVERAGE
2099 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2101 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2102 nomenclature these days.)
2104 PH/29 Installed patch from the Sieve maintainer that adds the options
2105 sieve_useraddress and sieve_subaddress to the redirect router.
2107 PH/30 In these circumstances:
2108 . Two addresses routed to the same list of hosts;
2109 . First host does not offer TLS;
2110 . First host accepts first address;
2111 . First host gives temporary error to second address;
2112 . Second host offers TLS and a TLS session is established;
2113 . Second host accepts second address.
2114 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2115 and peerdn, if requested) that were in fact used only for the second
2118 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2119 attention to any local part prefix or suffix that was matched by the
2120 router that accepted the address. It now behaves in the same way as it
2121 does for delivery: the affixes are removed from the local part unless
2122 rcpt_include_affixes is set on the transport.
2124 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2125 timeout during the DATA phase of an incoming message.
2127 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2128 applied a patch sanctioned by the Sieve maintainer.
2130 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2131 the uid or gid is negative. A case of a negative gid caused this to be
2132 noticed. The fix allows for either to be negative.
2134 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2135 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2138 PH/36 The change PH/12 above was broken. Fixed it.
2140 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2141 the grounds that routing the same address twice would always produce the
2142 same answer. This might have been true once, but it is certainly no
2143 longer true now. Routing a child address may depend on the previous
2144 routing that produced that child. Some complicated redirection strategies
2145 went wrong when messages had multiple recipients, and made Exim's
2146 behaviour dependent on the order in which the addresses were given.
2148 I have moved the duplicate checking until after the routing is complete.
2149 Exim scans the addresses that are assigned to local and remote
2150 transports, and removes any duplicates. This means that more work will be
2151 done, as duplicates will always all be routed, but duplicates are
2152 presumably rare, so I don't expect this is of any significance.
2154 For deliveries to pipes, files, and autoreplies, the duplicate checking
2155 still happens during the routing process, since they are not going to be
2158 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2159 It corrects a timeout issue with spamd. This is Ian's comment: "The
2160 background is that sometimes spamd either never reads data from a
2161 connection it has accepted, or it never writes response data. The exiscan
2162 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2163 blindly assumes that writes won't block so it may never time out."
2165 PH/39 Allow G after quota size as well as K and M.
2167 PH/40 The value set for $authenticated_id in an authenticator may not contain
2168 binary zeroes or newlines because the value is written to log lines and
2169 to spool files. There was no check on this. Now the value is run through
2170 the string_printing() function so that such characters are converted to
2171 printable escape sequences.
2173 PH/41 $message_linecount is a new variable that contains the total number of
2174 lines in the message. Compare $body_linecount, which is the count for the
2177 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2178 bounce and delay warning messages, except in certain special cases, which
2181 (a) An SMTP error message from a remote host;
2182 (b) A message specified in a :fail: redirection;
2183 (c) A message specified in a "fail" command in a system filter;
2184 (d) A message specified in a FAIL return from the queryprogram router;
2185 (e) A message specified by the cannot_route_message router option.
2187 In these cases only, Exim does include the error details in bounce and
2188 warning messages. There are also a few cases where bland messages such
2189 as "unrouteable address" or "local delivery error" are given.
2191 PH/43 $value is now also set for the "else" part of a ${run expansion.
2193 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2194 being worked on, but at least Exim now implements the latest version to
2197 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2198 process to complete was treated as a delivery failure, a timeout while
2199 writing the message to the pipe was logged, but erroneously treated as a
2200 successful delivery. Such timeouts include transport filter timeouts. For
2201 consistency with the overall process timeout, these timeouts are now
2202 treated as errors, giving rise to delivery failures by default. However,
2203 there is now a new Boolean option for the pipe transport called
2204 timeout_defer, which, if set TRUE, converts the failures into defers for
2205 both kinds of timeout. A transport filter timeout is now identified in
2208 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2209 systems where "make" and "gmake" are different, calling "gmake" at top
2210 level broke things. I've arranged for the value of $(MAKE) to be passed
2211 from the Makefile to this script so that it can call the same version of
2215 A note about Exim versions 4.44 and 4.50
2216 ----------------------------------------
2218 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2219 changes of various kinds. As a consequence, a big documentation update was
2220 needed. This delayed the release for rather longer than seemed good, especially
2221 in the light of a couple of (minor) security issues. Therefore, the changes
2222 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2223 release. So 4.44 and 4.50 are in effect two different branches that both start
2226 I have left the 4.50 change log unchanged; it contains all the changes since
2227 4.43. The change log for 4.44 is below; many of its items are identical to
2228 those for 4.50. This seems to be the most sensible way to preserve the
2229 historical information.
2235 1. Minor wording change to the doc/README.SIEVE file.
2237 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2238 computation of the current number of files was incorrect.
2240 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2241 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2244 4. Give more explanation in the error message when the command for a transport
2245 filter fails to execute.
2247 5. There are several places where Exim runs a non-Exim command in a
2248 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2249 was being done only for the command run by the queryprogram router. It is
2250 now done for all such subprocesses. The other cases are: ${run, transport
2251 filters, and the commands run by the lmtp and pipe transports.
2253 6. Added CONFIGURE_GROUP build-time option.
2255 7. Some older OS have a limit of 256 on the maximum number of file
2256 descriptors. Exim was using setrlimit() to set 1000 as a large value
2257 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2258 systems. I've change it so that if it can't get 1000, it tries for 256.
2260 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2261 was an oversight, and furthermore, ever since the addition of extra
2262 controls (e.g. 4.43/32), the checks on when to allow different forms of
2263 "control" were broken. There should now be diagnostics for all cases when a
2264 control that does not make sense is encountered.
2266 9. Added the /retain_sender option to "control=submission".
2268 10. $recipients is now available in the predata ACL (oversight).
2270 11. Tidy the search cache before the fork to do a delivery from a message
2271 received from the command line. Otherwise the child will trigger a lookup
2272 failure and thereby defer the delivery if it tries to use (for example) a
2273 cached ldap connection that the parent has called unbind on.
2275 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2276 of $address_data from the recipient verification was clobbered by the
2277 sender verification.
2279 13. The value of address_data from a sender verification is now available in
2280 $sender_address_data in subsequent conditions in the ACL statement.
2282 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2284 15. Added a new option "connect=<time>" to callout options, to set a different
2287 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2288 was its contents. (It was OK if the option was not defined at all.)
2290 17. A "Completed" log line is now written for messages that are removed from
2291 the spool by the -Mrm option.
2293 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2294 information about exactly what failed.
2296 19. Added -dd to debug only the daemon process.
2298 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2299 handles timeouts, both on the server side and network timeouts. Renamed the
2300 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2302 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2303 It is now set to "smtps".
2305 22. $host_address is now set to the target address during the checking of
2306 ignore_target_hosts.
2308 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2309 being passed; this would have caused $sender_host_name to have been used if
2310 matching the list had actually called for a host name (not very likely,
2311 since this list is usually IP addresses). A host name is now passed as
2314 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2315 code that specifies a non-restarting handler (typically sigaction() in
2316 modern systems) in an attempt to fix a rare and obscure crash bug.
2318 25. Narrowed the window for a race in the daemon that could cause it to ignore
2319 SIGCHLD signals. This is not a major problem, because they are used only to
2320 wake it up if nothing else does.
2322 26. A malformed maildirsize file could cause Exim to calculate negative values
2323 for the mailbox size or file count. Odd effects could occur as a result.
2324 The maildirsize information is now recalculated if the size or filecount
2327 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2328 support for a long time. Removed HAVE_SYS_VFS_H.
2330 28. Installed the latest version of exipick from John Jetmore.
2332 29. In an address list, if the pattern was not a regular expression, an empty
2333 subject address (from a bounce message) matched only if the pattern was an
2334 empty string. Non-empty patterns were not even tested. This was the wrong
2335 because it is perfectly reasonable to use an empty address as part of a
2336 database query. An empty address is now tested by patterns that are
2337 lookups. However, all the other forms of pattern expect the subject to
2338 contain a local part and a domain, and therefore, for them, an empty
2339 address still always fails if the pattern is not itself empty.
2341 30. Exim went into a mad DNS loop when attempting to do a callout where the
2342 host was specified on an smtp transport, and looking it up yielded more
2343 than one IP address.
2345 31. Re-factored the code for checking spool and log partition space into a
2346 function that finds that data and another that does the check. The former
2347 is then used to implement four new variables: $spool_space, $log_space,
2348 $spool_inodes, and $log_inodes.
2350 32. The RFC2047 encoding function was originally intended for short strings
2351 such as real names; it was not keeping to the 75-character limit for
2352 encoded words that the RFC imposes. It now respects the limit, and
2353 generates multiple encoded words if necessary. To be on the safe side, I
2354 have increased the buffer size for the ${rfc2047: expansion operator from
2357 33. It is now permitted to omit both strings after an "if" condition; if the
2358 condition is true, the result is "true". As before, when the second string
2359 is omitted, a false condition yields an empty string. This makes it less
2360 cumbersome to write custom ACL and router conditions.
2362 34. Failure to deliver a bounce message always caused it to be frozen, even if
2363 there was an errors_to setting on the router. The errors_to setting is now
2366 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2367 canonical form (fully expanded) before being placed in
2368 $sender_host_address.
2370 36. The table in the code that translates DNS record types into text (T_A to
2371 "A" for instance) was missing entries for NS and CNAME. It is just possible
2372 that this could have caused confusion if both these types were looked up
2373 for the same domain, because the text type is used as part of Exim's
2374 per-process caching. But the chance of anyone hitting this buglet seems
2377 37. The dnsdb lookup has been extended in a number of ways.
2379 (1) There is a new type, "zns", which walks up the domain tree until it
2380 finds some nameserver records. It should be used with care.
2382 (2) There is a new type, "mxh", which is like "mx" except that it returns
2383 just the host names, not the priorities.
2385 (3) It is now possible to give a list of domains (or IP addresses) to be
2386 looked up. The behaviour when one of the lookups defers can be
2387 controlled by a keyword.
2389 (4) It is now possible to specify the separator character for use when
2390 multiple records are returned.
2392 38. The dnslists ACL condition has been extended: it is now possible to supply
2393 a list of IP addresses and/or domains to be looked up in a particular DNS
2396 39. Added log_selector=+queue_time_overall.
2398 40. When running the queue in the test harness, wait just a tad after forking a
2399 delivery process, to get repeatability of debugging output.
2401 41. Include certificate and key file names in error message when GnuTLS fails
2402 to set them up, because the GnuTLS error message doesn't include the name
2403 of the failing file when there is a problem reading it.
2405 42. Allow both -bf and -bF in the same test run.
2407 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2409 44. The "Exiscan patch" is now merged into the mainline Exim source.
2411 45. Sometimes the final signoff response after QUIT could fail to get
2412 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2413 < 0 before doing a fflush(). This bug looks as though it goes back to the
2414 introduction of TLS in release 3.20, but "sometimes" must have been rare
2415 because the tests only now provoked it.
2417 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2418 (this can affect the format of dates).
2420 47. exim_tidydb, when checking for the continued existence of a message for
2421 which it has found a message-specific retry record, was not finding
2422 messages that were in split spool directories. Consequently, it was
2423 deleting retry records that should have stayed in existence.
2425 48. Steve fixed some bugs in eximstats.
2427 49. The SPA authentication driver was not abandoning authentication and moving
2428 on to the next authenticator when an expansion was forced to fail,
2429 contradicting the general specification for all authenticators. Instead it
2430 was generating a temporary error. It now behaves as specified.
2432 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2433 (the order specifies the preference for clients). The order is now AES256,
2434 AES128, 3DES, ARCFOUR128.
2436 51. Small patch to Sieve code - explicitly set From: when generating an
2439 52. Exim crashed if a remote delivery caused a very long error message to be
2440 recorded - for instance if somebody sent an entire SpamAssassin report back
2441 as a large number of 550 error lines. This bug was coincidentally fixed by
2442 increasing the size of one of Exim's internal buffers (big_buffer) that
2443 happened as part of the Exiscan merge. However, to be on the safe side, I
2444 have made the code more robust (and fixed the comments that describe what
2447 53. Now that there can be additional text after "Completed" in log lines (if
2448 the queue_time_overall log selector is set), a one-byte patch to exigrep
2449 was needed to allow it to recognize "Completed" as not the last thing in
2452 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2453 patch that reportedly fixes this has been added. I am not expert enough to
2454 create a test for it. This is what the patch creator wrote:
2456 "I found a little strange behaviour of ldap code when working with
2457 Windows 2003 AD Domain, where users was placed in more than one
2458 Organization Units. When I tried to give exim partial DN, the exit code
2459 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2460 But simultaneously result of request was absolutely normal ldap result,
2461 so I produce this patch..."
2463 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2464 REFERENCE, so I have modified the code to exclude the patch when that macro
2467 55. Some experimental protocols are using DNS PTR records for new purposes. The
2468 keys for these records are domain names, not reversed IP addresses. The
2469 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2470 leaves it alone. Component reversal etc. now happens only for IP addresses.
2473 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2475 57. Double the size of the debug message buffer (to 2048) so that more of very
2476 long debug lines gets shown.
2478 58. The exicyclog utility now does better if the number of log files to keep
2479 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2481 59. Two changes related to the smtp_active_hostname option:
2483 (1) $smtp_active_hostname is now available as a variable.
2484 (2) The default for smtp_banner uses $smtp_active_hostname instead
2485 of $primary_hostname.
2487 60. The host_aton() function is supposed to be passed a string that is known
2488 to be a valid IP address. However, in the case of IPv6 addresses, it was
2489 not checking this. This is a hostage to fortune. Exim now panics and dies
2490 if the condition is not met. A case was found where this could be provoked
2491 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2492 components; fortuitously, this particular loophole had already been fixed
2493 by change 4.50/55 above.
2495 If there are any other similar loopholes, the new check in host_aton()
2496 itself should stop them being exploited. The report I received stated that
2497 data on the command line could provoke the exploit when Exim was running as
2498 exim, but did not say which command line option was involved. All I could
2499 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2500 running as the user.
2503 61. There was a buffer overflow vulnerability in the SPA authentication code
2504 (which came originally from the Samba project). I have added a test to the
2505 spa_base64_to_bits() function which I hope fixes it.
2508 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2509 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2511 63. The daemon start-up calls getloadavg() while still root for those OS that
2512 need the first call to be done as root, but it missed one case: when
2513 deliver_queue_load_max is set with deliver_drop_privilege. This is
2514 necessary for the benefit of the queue runner, because there is no re-exec
2515 when deliver_drop_privilege is set.
2517 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2518 This has been fixed.
2520 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2521 was in use, was not putting the data itself into the right store pool;
2522 consequently, it could be overwritten for a subsequent message in the same
2523 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2526 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2528 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2529 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2530 the function were treating the return as a boolean value, which happened to
2531 work because 0=false and not-0=true, but is not correct code.
2533 68. The host_aton() function was not handling scoped IPv6 addresses (those
2534 with, for example, "%eth0" on the end) correctly.
2536 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2537 negated items (that is, ~something) in unsigned ints. Some compilers
2538 apparently mutter when there is no cast.
2540 70. If an address verification called from an ACL failed, and did not produce a
2541 user-specific message (i.e. there was only a "system" message), nothing was
2542 put in $acl_verify_message. In this situation, it now puts the system
2545 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2546 session; change 4.31/43 added the unwanted input to the log line - except
2547 that it did not do this in the start of session case. It now does.
2549 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2550 This is wrong and can cause the other end to generate a synchronization
2551 error if it is another Exim or anything else that does the synchronization
2552 check. A QUIT command is no longer sent after a timeout.
2554 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2555 during host lookups.
2557 74. The defer_ok option of callout verification was not working if it was used
2558 when verifying addresses in header lines, that is, for this case:
2560 verify = header_sender/callout=defer_ok
2562 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2563 those file descriptors could be used for SMTP connections. If anything
2564 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2565 could be sent to the SMTP client, causing chaos. The daemon now opens
2566 stdin, stdout, and stderr to /dev/null when it puts itself into the
2569 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2570 log by default. The user can override this with suitable Perl magic.
2572 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2573 the log message when discard triggers, was not working for the DATA ACL or
2574 for the non-SMTP ACL.
2576 78. Error message wording change in sieve.c.
2578 79. If smtp_accept_max_per_host was set, the number of connections could be
2579 restricted to fewer than expected, because the daemon was trying to set up
2580 a new connection before checking whether the processes handling previous
2581 connections had finished. The check for completed processes is now done
2582 earlier. On busy systems, this bug wouldn't be noticed because something
2583 else would have woken the daemon, and it would have reaped the completed
2586 80. If a message was submitted locally by a user whose login name contained one
2587 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2588 It caused a spool format error. I have fixed the spool reading code. A
2589 related problem was that the "from" clause in the Received: line became
2590 illegal because of the space(s). It is now covered by ${quote_local_part.
2592 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2595 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2597 83. Patch from Sieve maintainer to fix unterminated string problem in
2598 "vacation" handling.
2600 84. Some minor changes to the Linux configuration files to help with other
2601 OS variants using glibc.
2603 85. One more patch for Sieve to update vacation handling to latest spec.
2606 ----------------------------------------------------
2607 See the note above about the 4.44 and 4.50 releases.
2608 ----------------------------------------------------
2614 1. Change 4.43/35 introduced a bug that caused file counts to be
2615 incorrectly computed when quota_filecount was set in an appendfile
2618 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2619 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2622 3. Give more explanation in the error message when the command for a transport
2623 filter fails to execute.
2625 4. There are several places where Exim runs a non-Exim command in a
2626 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2627 was being done only for the command run by the queryprogram router. It is
2628 now done for all such subprocesses. The other cases are: ${run, transport
2629 filters, and the commands run by the lmtp and pipe transports.
2631 5. Some older OS have a limit of 256 on the maximum number of file
2632 descriptors. Exim was using setrlimit() to set 1000 as a large value
2633 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2634 systems. I've change it so that if it can't get 1000, it tries for 256.
2636 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2637 was an oversight, and furthermore, ever since the addition of extra
2638 controls (e.g. 4.43/32), the checks on when to allow different forms of
2639 "control" were broken. There should now be diagnostics for all cases when a
2640 control that does not make sense is encountered.
2642 7. $recipients is now available in the predata ACL (oversight).
2644 8. Tidy the search cache before the fork to do a delivery from a message
2645 received from the command line. Otherwise the child will trigger a lookup
2646 failure and thereby defer the delivery if it tries to use (for example) a
2647 cached ldap connection that the parent has called unbind on.
2649 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2650 of $address_data from the recipient verification was clobbered by the
2651 sender verification.
2653 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2654 was its contents. (It was OK if the option was not defined at all.)
2656 11. A "Completed" log line is now written for messages that are removed from
2657 the spool by the -Mrm option.
2659 12. $host_address is now set to the target address during the checking of
2660 ignore_target_hosts.
2662 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2663 being passed; this would have caused $sender_host_name to have been used if
2664 matching the list had actually called for a host name (not very likely,
2665 since this list is usually IP addresses). A host name is now passed as
2668 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2669 code that specifies a non-restarting handler (typically sigaction() in
2670 modern systems) in an attempt to fix a rare and obscure crash bug.
2672 15. Narrowed the window for a race in the daemon that could cause it to ignore
2673 SIGCHLD signals. This is not a major problem, because they are used only to
2674 wake it up if nothing else does.
2676 16. A malformed maildirsize file could cause Exim to calculate negative values
2677 for the mailbox size or file count. Odd effects could occur as a result.
2678 The maildirsize information is now recalculated if the size or filecount
2681 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2682 support for a long time. Removed HAVE_SYS_VFS_H.
2684 18. Updated exipick to current release from John Jetmore.
2686 19. Allow an empty sender to be matched against a lookup in an address list.
2687 Previously the only cases considered were a regular expression, or an
2690 20. Exim went into a mad DNS lookup loop when doing a callout where the
2691 host was specified on the transport, if the DNS lookup yielded more than
2694 21. The RFC2047 encoding function was originally intended for short strings
2695 such as real names; it was not keeping to the 75-character limit for
2696 encoded words that the RFC imposes. It now respects the limit, and
2697 generates multiple encoded words if necessary. To be on the safe side, I
2698 have increased the buffer size for the ${rfc2047: expansion operator from
2701 22. Failure to deliver a bounce message always caused it to be frozen, even if
2702 there was an errors_to setting on the router. The errors_to setting is now
2705 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2706 canonical form (fully expanded) before being placed in
2707 $sender_host_address.
2709 24. Updated eximstats to version 1.33
2711 25. Include certificate and key file names in error message when GnuTLS fails
2712 to set them up, because the GnuTLS error message doesn't include the name
2713 of the failing file when there is a problem reading it.
2715 26. Expand error message when OpenSSL has problems setting up cert/key files.
2718 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2719 (this can affect the format of dates).
2721 28. exim_tidydb, when checking for the continued existence of a message for
2722 which it has found a message-specific retry record, was not finding
2723 messages that were in split spool directories. Consequently, it was
2724 deleting retry records that should have stayed in existence.
2726 29. eximstats updated to version 1.35
2727 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2728 1.35 - bugfix such that pie charts by volume are generated correctly
2730 30. The SPA authentication driver was not abandoning authentication and moving
2731 on to the next authenticator when an expansion was forced to fail,
2732 contradicting the general specification for all authenticators. Instead it
2733 was generating a temporary error. It now behaves as specified.
2735 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2736 (the order specifies the preference for clients). The order is now AES256,
2737 AES128, 3DES, ARCFOUR128.
2739 31. Small patch to Sieve code - explicitly set From: when generating an
2742 32. Exim crashed if a remote delivery caused a very long error message to be
2743 recorded - for instance if somebody sent an entire SpamAssassin report back
2744 as a large number of 550 error lines. This bug was coincidentally fixed by
2745 increasing the size of one of Exim's internal buffers (big_buffer) that
2746 happened as part of the Exiscan merge. However, to be on the safe side, I
2747 have made the code more robust (and fixed the comments that describe what
2750 33. Some experimental protocols are using DNS PTR records for new purposes. The
2751 keys for these records are domain names, not reversed IP addresses. The
2752 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2753 leaves it alone. Component reversal etc. now happens only for IP addresses.
2756 34. The host_aton() function is supposed to be passed a string that is known
2757 to be a valid IP address. However, in the case of IPv6 addresses, it was
2758 not checking this. This is a hostage to fortune. Exim now panics and dies
2759 if the condition is not met. A case was found where this could be provoked
2760 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2761 components; fortuitously, this particular loophole had already been fixed
2762 by change 4.50/55 or 4.44/33 above.
2764 If there are any other similar loopholes, the new check in host_aton()
2765 itself should stop them being exploited. The report I received stated that
2766 data on the command line could provoke the exploit when Exim was running as
2767 exim, but did not say which command line option was involved. All I could
2768 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2769 running as the user.
2772 35. There was a buffer overflow vulnerability in the SPA authentication code
2773 (which came originally from the Samba project). I have added a test to the
2774 spa_base64_to_bits() function which I hope fixes it.
2777 36. The daemon start-up calls getloadavg() while still root for those OS that
2778 need the first call to be done as root, but it missed one case: when
2779 deliver_queue_load_max is set with deliver_drop_privilege. This is
2780 necessary for the benefit of the queue runner, because there is no re-exec
2781 when deliver_drop_privilege is set.
2783 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2784 was in use, was not putting the data itself into the right store pool;
2785 consequently, it could be overwritten for a subsequent message in the same
2786 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2789 38. Sometimes the final signoff response after QUIT could fail to get
2790 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2791 < 0 before doing a fflush(). This bug looks as though it goes back to the
2792 introduction of TLS in release 3.20, but "sometimes" must have been rare
2793 because the tests only now provoked it.
2799 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2800 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2801 Now it returns an integer. A number of calls were still expecting a T/F
2802 return. Fortuitously, in all cases, the tests worked in OK situations,
2803 which is the norm. However, things would have gone wrong on any write
2804 failures on the smtp file descriptor. This function is used when sending
2805 messages over SMTP and also when doing verify callouts.
2807 2. When Exim is called to do synchronous delivery of a locally submitted
2808 message (the -odf or -odi options), it no longer closes stderr before doing
2811 3. Implemented the mua_wrapper option.
2813 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2815 5. Implemented the functions header_remove(), header_testname(),
2816 header_add_at_position(), and receive_remove_recipient(), and exported them
2819 6. If an ACL "warn" statement specified the addition of headers, Exim already
2820 inserted X-ACL-Warn: at the start if there was no header name. However, it
2821 was not making this test for the second and subsequent header lines if
2822 there were newlines in the string. This meant that an invalid header could
2823 be inserted if Exim was badly configured.
2825 7. Allow an ACL "warn" statement to add header lines at the start or after all
2826 the Received: headers, as well as at the end.
2828 8. Added the rcpt_4xx retry error code.
2830 9. Added postmaster_mailfrom=xxx to callout verification option.
2832 10. Added mailfrom=xxxx to the callout verification option, for verify=
2835 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2836 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2838 12. Inserted some casts to stop certain compilers warning when using pointer
2839 differences as field lengths or precisions in printf-type calls (mostly
2840 affecting debugging statements).
2842 13. Added optional readline() support for -be (dynamically loaded).
2844 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2845 same clock tick as a message's arrival, so that its received time was the
2846 same as the "first fail" time on the retry record, and that message
2847 remained on the queue past the ultimate address timeout, every queue runner
2848 would try a delivery (because it was past the ultimate address timeout) but
2849 after another failure, the ultimate address timeout, which should have then
2850 bounced the address, did not kick in. This was a "< instead of <=" error;
2851 in most cases the first failure would have been in the next clock tick
2852 after the received time, and all would be well.
2854 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2855 being recognized when the domain list was tested by the match_domain
2856 condition in an expansion string.
2858 16. Added the ${str2b64: operator.
2860 17. Exim was always calling setrlimit() to set a large limit for the number of
2861 processes, without checking whether the existing limit was already
2862 adequate. (It did check for the limit on file descriptors.) Furthermore,
2863 errors from getrlimit() and setrlimit() were being ignored. Now they are
2864 logged to the main and panic logs, but Exim does carry on, to try to do its
2865 job under whatever limits there are.
2867 18. Imported PCRE 5.0.
2869 19. Trivial typo in log message " temporarily refused connection" (the leading
2872 20. If the log selector return_path_on_delivery was set and an address was
2873 redirected to /dev/null, the delivery process crashed because it assumed
2874 that a return path would always be set for a "successful" delivery. In this
2875 case, the whole delivery is bypassed as an optimization, and therefore no
2878 21. Internal re-arrangement: the function for sending a challenge and reading
2879 a response while authentication was assuming a zero-terminated challenge
2880 string. It's now changed to take a pointer and a length, to allow for
2881 binary data in such strings.
2883 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2885 23. Exim was not respecting finduser_retries when seeking the login of the
2886 uid under which it was called; it was always trying 10 times. (The default
2887 setting of finduser_retries is zero.) Also, it was sleeping after the final
2888 failure, which is pointless.
2890 24. Implemented tls_on_connect_ports.
2892 25. Implemented acl_smtp_predata.
2894 26. If the domain in control=submission is set empty, Exim assumes that the
2895 authenticated id is a complete email address when it generates From: or
2896 Sender: header lines.
2898 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2899 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2900 chown and chgrp in /bin and hostname in /usr/bin.
2902 28. Exim was keeping the "process log" file open after each use, just as it
2903 does for the main log. This opens the possibility of it remaining open for
2904 long periods when the USR1 signal hits a daemon. Occasional processlog
2905 errors were reported, that could have been caused by this. Anyway, it seems
2906 much more sensible not to leave this file open at all, so that is what now
2909 29. The long-running daemon process does not normally write to the log once it
2910 has entered its main loop, and it closes the log before doing so. This is
2911 so that log files can straightforwardly be renamed and moved. However,
2912 there are a couple of unusual error situations where the daemon does write
2913 log entries, and I had neglected to close the log afterwards.
2915 30. The text of an SMTP error response that was received during a remote
2916 delivery was being truncated at 512 bytes. This is too short for some of
2917 the long messages that one sometimes sees. I've increased the limit to
2920 31. It is now possible to make retry rules that apply only when a message has a
2921 specific sender, in particular, an empty sender.
2923 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2924 it possible to be selective about when SMTP synchronization is enforced.
2926 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2928 32. Implemented hosts_connection_nolog.
2930 33. Added an ACL for QUIT.
2932 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2935 35. Added mailbox_size and mailbox_filecount to appendfile.
2937 36. Added control = no_multiline_responses to ACLs.
2939 37. There was a bug in the logic of the code that waits for the clock to tick
2940 in the case where the clock went backwards by a substantial amount such
2941 that the microsecond fraction of "now" was more than the microsecond
2942 fraction of "then" (but the whole seconds number was less).
2944 38. Added support for the libradius Radius client library this is found on
2945 FreeBSD (previously only the radiusclient library was supported).
2951 1. When certain lookups returned multiple values in the form name=value, the
2952 quoting of the values was not always being done properly. Specifically:
2953 (a) If the value started with a double quote, but contained no whitespace,
2955 (b) If the value contained whitespace other than a space character (i.e.
2956 tabs or newlines or carriage returns) it was not quoted.
2957 This fix has been applied to the mysql and pgsql lookups by writing a
2958 separate quoting function and calling it from the lookup code. The fix
2959 should probably also be applied to nisplus, ibase and oracle lookups, but
2960 since I cannot test any of those, I have not disturbed their existing code.
2962 2. A hit in the callout cache for a specific address caused a log line with no
2963 reason for rejecting RCPT. Now it says "Previous (cached) callout
2964 verification failure".
2966 3. There was an off-by-one bug in the queryprogram router. An over-long
2967 return line was truncated at 256 instead of 255 characters, thereby
2968 overflowing its buffer with the terminating zero. As well as fixing this, I
2969 have increased the buffer size to 1024 (and made a note to document this).
2971 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2972 when Exim is waiting for an SMTP response from a remote server, Exim
2973 restarts its select() call on the socket, thereby resetting its timeout.
2974 This is not a problem when such interrupts are rare. Somebody set up a cron
2975 job to run exiwhat every 2 minutes, which is less than the normal select()
2976 timeout (5 or 10 minutes). This meant that the select() timeout never
2977 kicked in because it was always reset. I have fixed this by comparing the
2978 time when an interrupt arrives with the time at the start of the first call
2979 to select(). If more time than the timeout has elapsed, the interrupt is
2980 treated as a timeout.
2982 5. Some internal re-factoring in preparation for the addition of Sieve
2983 extensions (by MH). In particular, the "personal" test is moved to a
2984 separate function, and given an option for scanning Cc: and Bcc: (which is
2985 not set for Exim filters).
2987 6. When Exim created an email address using the login of the caller as the
2988 local part (e.g. when creating a From: or Sender: header line), it was not
2989 quoting the local part when it contained special characters such as @.
2991 7. Installed new OpenBSD configuration files.
2993 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2994 try to make them clearer.
2996 9. Callout options, other than the timeout value, were being ignored when
2997 verifying sender addresses in header lines. For example, when using
2999 verify = header_sender/callout=no_cache
3001 the cache was (incorrectly) being used.
3003 10. Added a missing instance of ${EXE} to the exim_install script; this affects
3004 only the Cygwin environment.
3006 11. When return_path_on_delivery was set as a log selector, if different remote
3007 addresses in the same message used different return paths and parallel
3008 remote delivery occurred, the wrong values would sometimes be logged.
3009 (Whenever a remote delivery process finished, the return path value from
3010 the most recently started remote delivery process was logged.)
3012 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3013 lines when AUTH and/or TLS are in use. This is the "received protocol"
3014 field. Exim used to use "asmtp" for authenticated SMTP, without any
3015 indication (in the protocol name) for TLS use. Now it follows the RFC and
3016 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3017 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3018 names appear in log lines as well as in Received: header lines.
3020 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3021 extensions, and comparison tests, and to fix some bugs.
3023 14. Changes to the "personal" filter test:
3025 (1) The test was buggy in that it was just doing the equivalent of
3026 "contains" tests on header lines. For example, if a user's address was
3027 anne@some.where, the "personal" test would incorrectly be true for
3029 To: susanne@some.where
3031 This test is now done by extracting each address from the header in turn,
3032 and checking the entire address. Other tests that are part of "personal"
3033 are now done using regular expressions (for example, to check local parts
3034 of addresses in From: header lines).
3036 (2) The list of non-personal local parts in From: addresses has been
3037 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3038 taken from the Sieve specification recommendations.
3040 (3) If the message contains any header line starting with "List-" it is
3041 treated as non-personal.
3043 (4) The test for "circular" in the Subject: header line has been removed
3044 because it now seems ill-conceived.
3046 15. Minor typos in src/EDITME comments corrected.
3048 16. Installed latest exipick from John Jetmore.
3050 17. If headers_add on a router specified a text string that was too long for
3051 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3052 of string_sprintf() is now avoided.
3054 18. $message_body_size was not set (it was always zero) when running the DATA
3055 ACL and the local_scan() function.
3057 19. For the "mail" command in an Exim filter, no default was being set for
3058 the once_repeat time, causing a random time value to be used if "once" was
3059 specified. (If the value happened to be <= 0, no repeat happened.) The
3060 default is now 0s, meaning "never repeat". The "vacation" command was OK
3061 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3062 (I found it when inspecting the code).
3064 20. There is now an overall timeout for performing a callout verification. It
3065 defaults to 4 times the callout timeout, which applies to individual SMTP
3066 commands during the callout. The overall timeout applies when there is more
3067 than one host that can be tried. The timeout is checked before trying the
3068 next host. This prevents very long delays if there are a large number of
3069 hosts and all are timing out (e.g. when the network connections are timing
3070 out). The value of the overall timeout can be changed by specifying an
3071 additional sub-option for "callout", called "maxwait". For example:
3073 verify = sender/callout=5s,maxwait=20s
3075 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3076 to the end before writing, but this should make it even safer).
3078 22. Exim was forgetting that it had advertised PIPELINING for the second and
3079 subsequent messages on an SMTP connection. It was also not resetting its
3080 memory on STARTTLS and an internal HELO.
3082 23. When Exim logs an SMTP synchronization error within a session, it now
3083 records whether PIPELINING has been advertised or not.
3085 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3086 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3087 rather than long int.
3089 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3091 26. Added the never_mail option to autoreply.
3097 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3098 crash if the getsockname() call failed; this can happen if a connection is
3099 closed very soon after it is established. The problem was simply in the
3100 order in which certain operations were done, causing Exim to try to write
3101 to the SMTP stream before it had set up the file descriptor. The bug has
3102 been fixed by making things happen in the correct order.
3108 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3109 before the connection was closed, thus losing the rejection response.
3111 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3112 some early Solaris releases, but causes trouble in current releases where
3113 socklen_t is defined.
3115 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3118 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3121 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3122 that releasing the top of it at the end releases what was used for sub-
3123 expansions (unless the block got too big). However, discard this block if
3124 the first thing is a variable or header, so that we can use its block when
3125 it is dynamic (useful for very large $message_headers, for example).
3127 6. Lookups now cache *every* query, not just the most recent. A new, separate
3128 store pool is used for this. It can be recovered when all lookup caches are
3129 flushed. Lookups now release memory at the end of their result strings.
3130 This has involved some general refactoring of the lookup sources.
3132 7. Some code has been added to the store_xxx() functions to reduce the amount
3133 of flapping under certain conditions.
3135 8. log_incoming_interface used to affect only the <= reception log lines. Now
3136 it causes the local interface and port to be added to several more SMTP log
3137 lines, for example "SMTP connection from", and rejection lines.
3139 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3141 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3143 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3144 could be overwritten at the end of the current message (or the start of a
3145 new message if it was set in a HELO ACL). The value is now preserved for
3146 the duration of the SMTP connection.
3148 12. If a transport had a headers_rewrite setting, and a matching header line
3149 contained an unqualified address, that address was qualified, even if it
3150 did not match any rewriting rules. The underlying bug was that the values
3151 of the flags that permit the existence of unqualified sender and recipient
3152 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3153 non-local messages, and by -bnq for local messages) were not being
3154 preserved with the message after it was received.
3156 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3157 "next input=" as part of the text comprising the host identity instead of
3158 the correct text. The code was using the same buffer for two different
3159 strings. However, depending on which order the printing function evaluated
3160 its arguments, the bug did not always show up. Under Linux, for example, my
3161 test suite worked just fine.
3163 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3164 doesn't work with some older versions of Perl. It has been changed to "my",
3165 which in any case is probably the better facility to use.
3167 15. A really picky compiler found some instances of statements for creating
3168 error messages that either had too many or two few arguments for the format
3171 16. The size of the buffer for calls to the DNS resolver has been increased
3172 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3173 for addresses that have a lot of PTR records. This alleviates a problem; it
3174 does not fully solve it.
3176 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3177 buffer now truncates the list and logs the incident, which is the same
3178 action as happens when Exim is looking up a host name and its aliases.
3179 Previously in this situation something unpredictable would happen;
3180 sometimes it was "internal error: store_reset failed".
3182 18. If a server dropped the connection unexpectedly when an Exim client was
3183 using GnuTLS and trying to read a response, the client delivery process
3184 crashed while trying to generate an error log message.
3186 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3187 string, for example:
3189 warn message = H1: something\nH2: something
3191 the text was added as a single header line from Exim's point of view
3192 though it ended up OK in the delivered message. However, searching for the
3193 second and subsequent header lines using $h_h2: did not work. This has been
3194 fixed. Similarly, if a system filter added multiple headers in this way,
3195 the routers could not see them.
3197 20. Expanded the error message when iplsearch is called with an invalid key to
3198 suggest using net-iplsearch in a host list.
3200 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3201 ACLs are no longer actually imposed (and a message to that effect is
3204 22. If a "gecos" field in a passwd entry contained escaped characters, in
3205 particular, if it contained a \" sequence, Exim got it wrong when building
3206 a From: or a Sender: header from that name. A second bug also caused
3207 incorrect handling when an unquoted " was present following a character
3208 that needed quoting.
3210 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3211 was not being matched caselessly.
3213 24. Arranged for all hyphens in the exim.8 source to be escaped with
3216 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3217 properly. Recipient callout cache records were still being keyed to include
3218 the sender, even when use_sender was set false. This led to far more
3219 callouts that were necessary. The sender is no longer included in the key
3220 when use_sender is false.
3222 26. Added "control = submission" modifier to ACLs.
3224 27. Added the ${base62d: operator to decode base 62 numbers.
3226 28. dnsdb lookups can now access SRV records.
3228 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3229 the configuration file.
3231 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3232 (-v) mode. This makes the output for a verbose queue run more intelligible.
3234 31. Added a use_postmaster feature to recipient callouts.
3236 32. Added the $body_zerocount variable, containing the number of binary zero
3237 bytes in the message body.
3239 33. The time of last modification of the "new" subdirectory is now used as the
3240 "mailbox time last read" when there is a quota error for a maildir
3243 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3245 35. Added +ignore_unknown as a special item in host lists.
3247 36. Code for decoding IPv6 addresses in host lists is now included, even if
3248 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3249 address was recognized as an IP address, but was then not correctly decoded
3250 into binary, causing unexpected and incorrect effects when compared with
3257 1. Very minor rewording of debugging text in manualroute to say "list of
3258 hosts" instead of "hostlist".
3260 2. If verify=header_syntax was set, and a header line with an unqualified
3261 address (no domain) and a large number of spaces between the end of the
3262 name and the colon was received, the reception process suffered a buffer
3263 overflow, and (when I tested it) crashed. This was caused by some obsolete
3264 code that should have been removed. The fix is to remove it!
3266 3. When running in the test harness, delay a bit after writing a bounce
3267 message to get a bit more predictability in the log output.
3269 4. Added a call to search_tidyup() just before forking a reception process. In
3270 theory, someone could use a lookup in the expansion of smtp_accept_max_
3271 per_host which, without the tidyup, could leave open a database connection.
3273 5. Added the variables $recipient_data and $sender_data which get set from a
3274 lookup success in an ACL "recipients" or "senders" condition, or a router
3275 "senders" option, similar to $domain_data and $local_part_data.
3277 6. Moved the writing of debug_print from before to after the "senders" test
3280 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3281 problems for message scanning, either using a data ACL, or using
3282 local_scan() because the Received: header was not generated till after they
3283 were called (in order to set the time as the time of reception completion).
3284 I have revised the way this works. The header is now generated after the
3285 body is received, but before the ACL or local_scan() are called. After they
3286 are run, the timestamp in the header is updated.
3292 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3293 before starting a queue runner without re-exec. This happened only when
3294 deliver_drop_privilege was set or when the Exim user was set to root. The
3295 effect of the bug was that timeouts during subsequent deliveries caused
3296 crashes instead of being properly handled. The handler is now left at its
3297 default (and expected) setting.
3299 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3300 message, again when deliver_drop_privilege is set or Exim is run as root.
3301 The bug described in (1) was not present in this case, but the tidying up
3302 of the other signals was missing. I have made the two cases consistent.
3304 3. The ignore_target_hosts setting on a manualroute router was being ignored
3305 for hosts that were looked up using the /MX notation.
3307 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3310 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3311 operated on the sender address. After changing the $sender_address to <>
3312 for the sender address verify, Exim was re-instated it as the original
3313 (before rewriting) address, but remembering that it had rewritten it, so it
3314 wasn't rewriting it again. This bug also had the effect of breaking the
3315 sender address verification caching when the sender address was rewritten.
3317 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3318 This has been changed so that if the ip literal address matches
3319 ignore_target_hosts, the router declines.
3321 7. Added expansion conditions match_domain, match_address, and match_local_
3322 part (NOT match_host).
3324 8. The placeholder for the Received: header didn't have a length field set.
3326 9. Added code to Exim itself and to exim_lock to test for a specific race
3327 condition that could lead to file corruption when using MBX delivery. The
3328 issue is with the lockfile that is created in /tmp. If this file is removed
3329 after a process has opened it but before that process has acquired a lock,
3330 there is the potential for a second process to recreate the file and also
3331 acquire a lock. This could lead to two Exim processes writing to the file
3332 at the same time. The added code performs the same test as UW imapd; it
3333 checks after acquiring the lock that its file descriptor still refers to
3334 the same named file.
3336 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3337 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3340 11. Added the smtp_active_hostname option. If used, this will typically be made
3341 to depend on the incoming interface address. Because $interface_address is
3342 not set up until the daemon has forked a reception process, error responses
3343 that can happen earlier (such as "too many connections") no longer contain
3346 12. If an expansion in a condition on a "warn" statement fails because a lookup
3347 defers, the "warn" statement is abandoned, and the next ACL statement is
3348 processed. Previously this caused the whole ACL to be aborted.
3350 13. Added the iplsearch lookup type.
3352 14. Added ident_timeout as a log selector.
3354 15. Added tls_certificate_verified as a log selector.
3356 16. Added a global option tls_require_ciphers (compare the smtp transport
3357 option of the same name). This controls incoming TLS connections.
3359 17. I finally figured out how to make tls_require_ciphers do a similar thing
3360 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3361 before starting the TLS session.
3363 18. Tabs are now shown as \t in -bP output.
3365 19. If the log selector return_path_on_delivery was set, Exim crashed when
3366 bouncing a message because it had too many Received: header lines.
3368 20. If two routers both had headers_remove settings, and the first one included
3369 a superfluous trailing colon, the final name in the first list and the
3370 first name in the second list were incorrectly joined into one item (with a
3371 colon in the middle).
3377 1. Added -C and -D options to the exinext utility, mainly to make it easier
3378 to include in the automated testing, but these could be helpful when
3379 multiple configurations are in use.
3381 2. The exinext utility was not formatting the output nicely when there was
3382 an alternate port involved in the retry record key, nor when there was a
3383 message id as well (for retries that were specific to a specific message
3384 and a specific host). It was also confused by IPv6 addresses, because of
3385 the additional colons they contain. I have fixed the IPv4 problem, and
3386 patched it up to do a reasonable job for IPv6.
3388 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3389 delivery, the log line now contains "pipelined" if PIPELINING was used.
3391 4. An SMTP transport process used to panic and die if the bind() call to set
3392 an explicit outgoing interface failed. This has been changed; it is now
3393 treated in the same way as a connect() failure.
3395 5. A reference to $sender_host_name in the part of a conditional expansion
3396 that was being skipped was still causing a DNS lookup. This no longer
3399 6. The def: expansion condition was not recognizing references to header lines
3400 that used bh_ and bheader_.
3402 7. Added the _cache feature to named lists.
3404 8. The code for checking quota_filecount in the appendfile transport was
3405 allowing one more file than it should have been.
3407 9. For compatibility with Sendmail, the command line option
3415 and sets the incoming protocol and host name (for trusted callers). The
3416 host name and its colon can be omitted when only the protocol is to be set.
3417 Note the Exim already has two private options, -pd and -ps, that refer to
3418 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3419 "s", but I don't think that's a major issue.
3421 10. A number of refactoring changes to the code, none of which should affect
3424 (a) The number of logging options was getting close to filling up the
3425 32-bit word that was used as a bit map. I have split them into two classes:
3426 those that are passed in the argument to log_write(), and those that are
3427 only ever tested independently outside of that function. These are now in
3428 separate 32-bit words, so there is plenty of room for expansion again.
3429 There is no change in the user interface or the logging behaviour.
3431 (b) When building, for example, log lines, the code previously used a
3432 macro that called string_cat() twice, in order to add two strings. This is
3433 not really sufficiently general. Furthermore, there was one instance where
3434 it was actually wrong because one of the argument was used twice, and in
3435 one call a function was used. (As it happened, calling the function twice
3436 did not affect the overall behaviour.) The macro has been replaced by a
3437 function that can join an arbitrary number of extra strings onto a growing
3440 (c) The code for expansion conditions now uses a table and a binary chop
3441 instead of a serial search (which was left over from when there were very
3442 few conditions). Also, it now recognizes conditions like "pam" even when
3443 the relevant support is not compiled in: a suitably worded error message is
3444 given if an attempt is made to use such a condition.
3446 11. Added ${time_interval:xxxxx}.
3448 12. A bug was causing one of the ddress fields not to be passed back correctly
3449 from remote delivery subprocesses. The field in question was not being
3450 subsequently used, so this caused to problems in practice.
3452 13. Added new log selectors queue_time and deliver_time.
3454 14. Might have fixed a bug in maildirsizefile handling that threw up
3455 "unexpected character" debug warnings, and recalculated the data
3456 unnecessarily. In any case, I expanded the warning message to give more
3459 15. Added the message "Restricted characters in address" to the statements in
3460 the default ACL that block characters like @ and % in local parts.
3462 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3463 Three changes have been made:
3465 (a) There was a serious bug; a negative response to MAIL caused the whole
3466 recipient domain to be cached as invalid, thereby blocking all messages
3467 to all local parts at the same domain, from all senders. This bug has
3468 been fixed. The domain is no longer cached after a negative response to
3469 MAIL if the sender used is not empty.
3471 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3474 (c) A new callout option, "use_sender" has been added for people who want
3475 the modified behaviour.
3481 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3484 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3485 indeed breaks things for older releases.
3487 3. Added additional logging to the case where there is a problem reading data
3488 from a filter that is running in a subprocess using a pipe, in order to
3489 try to track down a specific problem.
3491 4. Testing facility fudge: when running in the test harness and attempting
3492 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3493 getting "No route to host". Convert this to a timeout.
3495 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3498 6. Some OS don't have socklen_t but use size_t instead. This affects the
3499 fifth argument of getsockopt() amongst other things. This is now
3500 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3501 can be set for individual OS. I have set it for SunOS5, OSF1, and
3502 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3503 some earlier ones do not.
3505 7. Change 4.30/15 was not doing the test caselessly.
3507 8. The standard form for an IPv6 address literal was being rejected by address
3508 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3509 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3510 this, as well as the form without the "IPv6" on the front (but only when
3511 address literals are enabled, of course).
3513 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3515 10. Exim crashed if a message with an empty sender address specified by -f
3516 encountered a router with an errors_to setting. This could be provoked only
3517 by a command such as
3521 where an empty string was supplied; "<>" did not hit this bug.
3523 11. Installed PCRE release 4.5.
3525 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3526 remained set. It is now erased.
3528 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3529 times from message ids (which are base 36 rather than the normal 62).
3531 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3532 were being counted as actual protocol errors, and logged if the log
3533 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3534 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3535 and DATA following a set of rejected RCPTs do not count as protocol errors.
3536 In other words, Exim assumes they were pipelined, though this may not
3537 actually be the case. Of course, in all cases the client gets an
3538 appropriate error code.
3540 15. If a lookup fails in an ACL condition, a message about the failure may
3541 be available; it is used if testing the ACL cannot continue, because most
3542 such messages specify what the cause of the deferral is. However, some
3543 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3544 that caused an old message to be retained and used if a later statement
3545 caused a defer, replacing the real cause of the deferral.
3547 16. If an IP address had so many PTR records that the DNS lookup buffer
3548 was not large enough to hold them, Exim could crash while trying to process
3549 the truncated data. It now detects and logs this case.
3551 17. Further to 4.21/58, another change has been made: if (and only if) the
3552 first line of a message (the first header line) ends with CRLF, a bare LF
3553 in a subsequent header line has a space inserted after it, so as not to
3554 terminate the header.
3556 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3557 unnecessarily, used atoi() instead of strtol(), and didn't check the
3558 termination when getting file sizes from file names by regex.
3560 19. Completely re-implemented the support for maildirsize files, in the light
3561 of a number of problems with the previous contributed implementation
3562 (4.30/29). In particular:
3564 . If the quota is zero, the maildirsize file is maintained, but no quota is
3567 . If the maildir directory does not exist, it is created before any attempt
3568 to write a maildirsize file.
3570 . The quota value in the file is just a cache; if the quota is changed in
3571 the transport, the new value overrides.
3573 . A regular expression is available for excluding directories from the
3576 20. The autoreply transport checks the characters in options that define the
3577 message's headers; it allows continued headers, but it was checking with
3578 isspace() after an embedded newline instead of explicitly looking for a
3581 21. If all the "regular" hosts to which an address was routed had passed their
3582 expiry times, and had not reached their retry times, the address was
3583 bounced, even if fallback hosts were defined. Now Exim should go on to try
3586 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3587 equivalent code in the SMTP transport. Some hosts send humungous responses
3588 to HELO/EHLO, more than 1024 it seems.
3590 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3591 gives compiler warnings in some environments. I've now done it "properly",
3594 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3595 (because of problems with the built-in one) was declared to return uschar *
3596 instead of char *, causing compiler failure.
3598 25. Fixed a file descriptor leak when processing alias/forward files.
3600 26. Fixed a minor format string issue in dbfn.c.
3602 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3604 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3605 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3608 29. When a sender address is verified, it is cached, to save repeating the test
3609 when there is more than one recipient in a message. However, when the
3610 verification involves a callout, it is possible for different callout
3611 options to be set for different recipients. It is too complicated to keep
3612 track of this in the cache, so now Exim always runs a verification when a
3613 callout is required, relying on the callout cache for the optimization.
3614 The overhead is duplication of the address routing, but this should not be
3617 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3618 to be verified with callout=postmaster, and the main callout worked but the
3619 postmaster check failed, the verification correctly failed. However, if a
3620 subsequent RCPT command asked for sender verification *without* the
3621 postmaster check, incorrect caching caused this verification also to fail,
3624 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3625 it was not caching the DNS options (qualify_single, search_parents) that
3626 were used when the lookup failed. A subsequent lookup with different
3627 options therefore always gave the same answer, though there were cases
3628 where it should not have. (Example: a "domains = !$mx_any" option on a
3629 dnslookup router: the "domains" option is always processed without any
3630 widening, but the router might have qualify_single set.) Now Exim uses the
3631 cached value only when the same options are set.
3633 32. Added John Jetmore's "exipick" utility to the distribution.
3635 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3636 than a timeout (e.g. a certificate is required, and is not provided), an
3637 Exim server now closes the connection immediately. Previously it waited for
3638 the client to close - but if the client is SSL, it seems that they each
3639 wait for each other, leading to a delay before one of them times out.
3641 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3642 maintained 0.8.x compatibility because I don't think many are using it, and
3643 it is clearly obsolete.
3645 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3648 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3649 client certificate was expired. A simple patch fixes this, though I don't
3650 understand the full logic of why the verify callback is called multiple
3653 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3654 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3655 which causes problems with some clients (such as the Certicom SSL Plus
3656 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3657 disables the coutermeasure allowing Eudora to connect."
3659 38. Exim was not checking that a write() to a log file succeeded. This could
3660 lead to Bad Things if a log got too big, in particular if it hit a file
3661 size limit. Exim now panics and dies if it cannot write to a log file, just
3662 as it does if it cannot open a log file.
3664 39. Modified OS/Makefile-Linux so that it now contains
3666 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3668 The two -D definitions ensure that Exim is compiled with large file
3669 support, which makes it possible to handle log files that are bigger than
3672 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3673 instance) a domain was checked against a named list that involved a lookup,
3674 causing $domain_data to be set, then another domain was checked against the
3675 same list, then the first domain was re-checked, the value of $domain_data
3676 after the final check could be wrong. In particular, if the second check
3677 failed, it could be set empty. This bug probably also applied to
3680 41. The strip_trailing_dot option was not being applied to the address given
3681 with the -f command-line option.
3683 42. The code for reading a message's header from the spool was incrementing
3684 $received_count, but never initializing it. This meant that the value was
3685 incorrect (doubled) while delivering a message in the same process in which
3686 it was received. In the most common configuration of Exim, this never
3687 happens - a fresh exec is done - but it can happen when
3688 deliver_drop_privilege is set.
3690 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3691 it now includes up to 150 characters of the unexpected data in the log
3694 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3695 and building data strings. The size of both of these buffers was 10 000
3696 bytes - far larger than anybody would *ever* want, thought I. Needless to
3697 say, somebody hit the limit. I have increased the maximum line length to
3698 20 000 and the maximum data length of concatenated lines to 100 000. I have
3699 also fixed two bugs, because there was no checking on these buffers. Tsk,
3700 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3701 buffer is too small.
3703 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3704 lsearch lookups. Now it does.
3706 46. When parsing a route_list item in a manualroute router, a fixed-length
3707 buffer was used for the list of hosts. I made this 1024 bytes long,
3708 thinking that nobody would ever have a list of hosts that long. Wrong.
3709 Somebody had a whole pile of complicated expansion conditions, and the
3710 string was silently truncated, leading to an expansion error. It turns out
3711 that it is easier to change to an unlimited length (owing to other changes
3712 that have happened since this code was originally written) than to build
3713 structure for giving a limitation error. The length of the item that
3714 expands into the list of hosts is now unlimited.
3716 47. The lsearch lookup could not handle data where the length of text line was
3717 more than 4095 characters. Such lines were truncated, leading to shortened
3718 data being returned. It should now handle lines of any length.
3720 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3721 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3724 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3725 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3726 no longer happens in comment lines. A list of replacements is now placed
3727 at the head of all of the source files, except those whose only change is
3728 to replace PERL_COMMAND in the very first #! line.
3730 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3731 messages on the queue, with a bottom-up merge sort, using code contributed
3732 by Michael Haardt. This should make operations like -bp somewhat faster on
3733 large queues. It won't affect queue runners, except when queue_run_in_order
3736 51. Installed eximstats 1.31 in the distribution.
3738 52. Added support for SRV lookups to the dnslookup router.
3740 53. If an ACL referred to $message_body or $message_body_end, the value was not
3741 reset for any messages that followed in the same SMTP session.
3743 54. The store-handling optimization for building very long strings was not
3744 differentiating between the different store pools. I don't think this
3745 actually made any difference in practice, but I've tidied it.
3747 55. While running the routers to verify a sender address, $sender_address
3748 was still set to the sender address. This is wrong, because when routing to
3749 send a bounce to the sender, it would be empty. Therefore, I have changed
3750 it so that, while verifying a sender address, $sender_address is set to <>.
3751 (There is no change to what happens when verifying a recipient address.)
3753 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3754 target A or AAAA records (if not already returned) without resetting the
3755 qualify_single or search_parents options of the DNS resolver. These are
3756 inappropriate in this case because the targets of MX and SRV records must
3757 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3758 target that, when qualified, matched something in the local domain. These
3759 two options are now turned off when doing these lookups.
3761 57. It seems that at least some releases of Reiserfs (which does not have the
3762 concept of a fixed number of inodes) returns zero and not -1 for the
3763 number of available inodes. This interacted badly with check_spool_inodes,
3764 which assumed that -1 was the "no such thing" setting. What I have done is
3765 to check that the total number of inodes is greater than zero before doing
3766 the test of how many are available.
3768 58. When a "warn" ACL statement has a log_message modifier, the message is
3769 remembered, and not repeated. This is to avoid a lot of repetition when a
3770 message has many recipients that cause the same warning to be written.
3771 Howewer, Exim was preserving the list of already written lines for an
3772 entire SMTP session, which doesn't seem right. The memory is now reset if a
3773 new message is started.
3775 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3776 debugging output unless log_rewrite was also set.
3778 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3779 of (int)(handle) when we know that handle contains (void *)(-1).
3781 61. The Exim daemon panic-logs an error return when it closes the incoming
3782 connection. However "connection reset by peer" seems to be common, and
3783 isn't really an error worthy of noting specially, so that particular error
3786 62. When Exim is trying to find all the local interfaces, it used to panic and
3787 die if the ioctl to get the interface flags failed. However, it seems that
3788 on at least one OS (Solaris 9) it is possible to have an interface that is
3789 included in the list of interfaces, but for which you get a failure error
3790 for this call. This happens when the interface is not "plumbed" into a
3791 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3792 failure of the "get flags" call assumes that the interface is down.
3794 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3795 makes life easier for people who are doing arithmetic on fields extracted
3796 from dates, where you often get leading zeros that should not be
3797 interpreted as octal.
3799 64. Added qualify_domain to the redirect router, to override the global
3802 65. If a pathologically long header line contained very many addresses (the
3803 report of this problem mentioned 10 000) and each of them was rewritten,
3804 Exim could use up a very large amount of memory. (It kept on making new
3805 copies of the header line as it rewrote, and never released the old ones.)
3806 At the expense of a bit more processing, the header rewriting function has
3807 been changed so that it no longer eats memory in this way.
3809 66. The generation of the Received: header has been moved from the time that a
3810 message starts to be received, to the time that it finishes. The timestamp
3811 in the Received: header should now be very close to that of the <= log
3812 line. There are two side-effects of this change:
3814 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3815 logged header lines no longer include the local Received: line, because
3816 it has not yet been created. The same applies to a copy of the message
3817 that is returned to a non-SMTP sender when a message is rejected.
3819 (b) When a filter file is tested using -bf, no additional Received: header
3820 is added to the test message. After some thought, I decided that this
3823 This change does not affect the value of $received_for. It is still set
3824 after address rewriting, but before local_scan() is called.
3826 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3828 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3829 gave an unhelpful panic error message, and a defer error. I have managed to
3830 change this behaviour so that it now rejects any supplied certificate,
3831 which seems right, as the list of acceptable certificates is empty.
3833 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3834 gave an unhelpful defer error. I have not managed to make this reject any
3835 supplied certificates, but the error message it gives is "no certificate
3836 supplied", which is not helpful.
3838 70. exigrep's output now also includes lines that are not associated with any
3839 message, but which match the given pattern. Implemented by a patch from
3840 Martin Sluka, which also tidied up the Perl a bit.
3842 71. Recipient callout verification, like sender verification, was using <> in
3843 the MAIL FROM command. This isn't really the right thing, since the actual
3844 sender may affect whether the remote host accepts the recipient or not. I
3845 have changed it to use the actual sender in the callout; this means that
3846 the cache record is now keyed on a recipient/sender pair, not just the
3847 recipient address. There doesn't seem to be a real danger of callout loops,
3848 since a callout by the remote host to check the sender would use <>.
3849 [SEE ABOVE: changed after hitting problems.]
3851 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3852 temporary errors. However, in the case of such a code being given after
3853 the end of a data transmission (i.e. after ".") Exim was failing to write
3854 a retry record for the message. (Yes, there was some broken host that was
3855 actually sending 8xx at this point.)
3857 73. An unknown lookup type in a host list could cause Exim to panic-die when
3858 the list was checked. (An example that provoked this was putting <; in the
3859 middle of a list instead of at the start.) If this happened during a DATA
3860 ACL check, a -D file could be left lying around. This kind of configuration
3861 error no longer causes Exim to die; instead it causes a defer errror. The
3862 incident is still logged to the main and panic logs.
3864 74. Buglet left over from Exim 3 conversion. The message "too many messages
3865 in one connection" was written to the rejectlog but not the mainlog, except
3866 when address rewriting (yes!) was being logged.
3868 75. Added write_rejectlog option.
3870 76. When a system filter was run not as root (that is, when system_filter_user
3871 was set), the values of the $n variables were not being returned to the
3872 main process; thus, they were not subsequently available in the $sn
3875 77. Added +return_path_on_delivery log selector.
3877 78. A connection timeout was being treated differently from recipients deferred
3878 when testing hosts_max_try with a message that was older than the host's
3879 retry timeout. (The host should not be counted, thus allowing all hosts to
3880 be tried at least once before bouncing.) This may have been the cause of an
3881 occasionally reported bug whereby a message would remain on the queue
3882 longer than the retry timeout, but would be bounced if a delivery was
3883 forced. I say "may" because I never totally pinned down the problem;
3884 setting up timeout/retry tests is difficult. See also the next item.
3886 79. The ultimate address timeout was not being applied to errors that involved
3887 a combination of host plus message (for example, a timeout on a MAIL
3888 command). When an address resolved to a number of possible hosts, and they
3889 were not all tried for each delivery (e.g. because of hosts_max_try), a
3890 message could remain on the queue longer than the retry timeout.
3892 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3895 81. Fixed an obscure SMTP outgoing bug which required at least the following
3896 conditions: (a) there was another message waiting for the same server;
3897 (b) the server returned 5xx to all RCPT commands in the first message so
3898 that the message was not completed; (c) the server dropped the connection
3899 or gave a negative response to the RSET that Exim sends to abort the
3900 transaction. The observed case was a dropped connection after DATA that had
3901 been sent in pipelining mode. That is, the server had advertised PIPELINING
3902 but was not implementing it correctly. The effect of the bug was incorrect
3903 behaviour, such as trying another host, and this could lead to a crash.
3909 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3910 and daemon.c were passed as pointers to ints; they should have been
3911 pointers to socklen_t variables (which are typically unsigned ints).
3913 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3916 3. Fixed a really odd bug that affected only the testing scheme; patching a
3917 certain fixed string in the binary changed the value of another string that
3918 happened to be identical to the end of the original first string.
3920 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3921 name", it returns that address as the IP address. On some operating
3922 systems (e.g. Solaris), it also passes back the IP address string as the
3923 "host name". However, on others (e.g. Linux), it passes back an empty
3924 string. Exim wasn't checking for this, and was changing the host name to an
3925 empty string, assuming it had been canonicized.
3927 5. Although rare, it is permitted to have more than one PTR record for a given
3928 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3929 all the names associated with an address, because they do in Solaris.
3930 However, it seems that they do not in Linux for data that comes from the
3931 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3932 I found this out when I moved to a new Linux workstation and tried to run
3933 the Exim test suite.
3935 To get round this problem I have changed the code so that it now does its
3936 own call to the DNS to look up PTR records when searching for a host name.
3937 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3938 addresses that are only in /etc/hosts are still found.
3940 This behaviour is, however, controlled by an option called host_lookup_
3941 order, which defaults to "bydns:byaddr". If people want to use the other
3942 order, or indeed, just use one or the other means of lookup, they can
3943 specify it in this variable.
3945 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3946 some operating systems, this comes back from gethostbyaddr() as an empty
3947 string, and this is what Exim used to test for. However, it seems that in
3948 other systems, "." is yielded. Exim now tests for this case too.
3950 7. The values of check_spool_space and check_log_space are now held internally
3951 as a number of kilobytes instead of an absolute number of bytes. If a
3952 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3953 kilobyte. This means that much larger values can be stored.
3955 8. Exim monitor: an attempt to get the action menu when not actually pointing
3956 at a message produces an empty menu entitled "No message selected". This
3957 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3958 no entries in it ("Shell widget menu has zero width and/or height"). So I
3959 have added a single, blank menu entry in this case.
3961 9. Added ${quote_local_part.
3963 10. MIME decoding is now applied to the contents of Subject: header lines when
3966 11. Now that a reference to $sender_host_address automatically causes a reverse
3967 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3968 host lookup before query-style lookups in lists that might use this
3969 variable. This has therefore been abolished, and the "net-" prefix is no
3970 longer necessary for query-style lookups.
3972 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3973 have been a typo for LFLAGS, so it has been changed.
3975 13. The install script calls Exim with "-C /dev/null" in order to find the
3976 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3977 to be output. Howeve, since Exim outputs its version number before the
3978 error, it didn't break the script. It just looked ugly. I fixed this by
3979 always allowing "-C /dev/null" if the caller is root.
3981 14. Ignore overlarge ACL variable number when reading spool file - insurance
3982 against a later release with more variables having written the file.
3984 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3985 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3986 this, as well as the form without the "IPv6" on the front.
3988 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3989 OS/Makefile-Darwin file.
3991 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3992 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3994 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3995 variable. However, some OS use ints for inodes. I've added cast to long int
3996 to get rid of the compiler warning.
3998 19. I had forgotten to lock out "/../" in configuration file names when
3999 ALT_CONFIG_PREFIX was set.
4001 20. Routers used for verification do not need to specify transports. However,
4002 if such a router generated a host list, and callout was configured, Exim
4003 crashed, because it could not find a port number from the (non-existent)
4004 transport. It now assumes port 25 in this circumstance.
4006 21. Added the -t option to exigrep.
4008 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4009 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4010 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4011 them is set, LOOKUP_LSEARCH is forced.
4013 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4015 24. Added sender and host information to the "rejected by local_scan()" log
4016 line; previously there was no indication of these.
4018 25. Added .include_if_exists.
4020 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4021 Linux. It turns out that not all file systems support this. Apparently some
4022 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4023 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4024 not supported on the file descriptor, is now ignored when Exim is trying to
4025 sync a directory. This applies only to Linux.
4027 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4029 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4030 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4033 29. Added support for maildirsize files from supplied patch (modified a bit).
4035 30. The use of :fail: followed by an empty string could lead Exim to respond to
4036 sender verification failures with (e.g.):
4038 550 Verification failed for <xxx>
4039 550 Sender verify failed
4041 where the first response line was missing the '-' that indicates it is not
4042 the final line of the response.
4044 31. The loop for finding the name of the user that called Exim had a hardwired
4045 limit of 10; it now uses the value of finduser_retries, which is used for
4046 all other user lookups.
4048 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4051 33. Exim was neglecting to zero errno before one call of strtol() when
4052 expanding a string and expecting an integer value. On some systems this
4053 resulted in spurious "integer overflow" errors. Also, it was casting the
4054 result into an int without checking.
4056 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4057 did not work. The code looks as if it has *never* worked, though it appears
4058 to have been documented since at least releast 1.62. I have made it work.
4060 35. The "timeout_DNS" error in retry rules, also documented since at least
4061 1.62, also never worked. As it isn't clear exactly what this means, and
4062 clearly it isn't a major issue, I have abolished the feature by treating it
4063 as "timeout", and writing a warning to the main and panic logs.
4065 36. The display of retry rules for -brt wasn't always showing the error code
4068 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4069 timeout_connect_A, timeout_connect_MX.
4071 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4072 to the empty sender.
4074 39. The daemon was not analysing the content of -oX till after it had closed
4075 stderr and disconnected from the controlling terminal. This meant that any
4076 syntax errors were only noted on the panic log, and the return code from
4077 the command was 0. By re-arranging the code a little, I've made the
4078 decoding happen first, so such errors now appear on stderr, and the return
4079 code is 1. However, the actual setting up of the sockets still happens in
4080 the disconnected process, so errors there are still only recorded on the
4083 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4084 connections (as happens on some IP stacks) was logged at start up time as
4085 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4086 it from "IPv6 and IPv4", which means that two separate sockets are being
4089 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4090 says whether AF_INET or AF_INET6 was passed as an argument.
4092 42. Exiwhat output was messed up when time zones were included in log
4095 43. Exiwhat now gives more information about the daemon's listening ports,
4096 and whether -tls-on-connect was used.
4098 44. The "port" option of the smtp transport is now expanded.
4100 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4101 silently ignored. Now an error message is written to the main and panic
4104 46. There's a new ACL modifier called "logwrite" which writes to a log file
4105 as soon as it is encountered.
4107 47. Added $local_user_uid and $local_user_gid at routing time.
4109 48. Exim crashed when trying to verify a sender address that was being
4112 49. Exim was recognizing only a space character after ".include". It now also
4113 recognizes a tab character.
4115 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4116 extracting the relevant information from the specification. The man page no
4117 longer contains scrambled data for the -d option, and I've added a section
4118 at the front about calling Exim under different names.
4120 51. Added "extra_headers" argument to the "mail" command in filter files.
4122 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4125 53. Installed eximstats 1.29.
4127 54. Added transport_filter_timeout as a generic transport option.
4129 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4130 Cc: header lines. This was required by RFC 822, but it not required by RFC
4133 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4134 incoming messages that did not have them. Now it does so only if the
4135 message originates locally, that is, if there is no associated remote host
4136 address. When Resent- header lines are present, this applies to the Resent-
4137 lines rather than the non-Resent- lines.
4139 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4140 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4142 58. Messages for configuration errors now include the name of the main
4143 configuration file - useful now that there may be more than one file in a
4144 list (.included file names were always shown).
4146 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4147 for those rare installations that do not start the daemon as root or run it
4148 setuid root. I've cut out the call to initgroups() if the daemon is not
4151 60. The Exim user and group can now be bound into the binary as text strings
4152 that are looked up at the start of Exim's processing.
4154 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4156 62. Added $mailstore_basename variable.
4158 63. Installed patch to sieve.c from Michael Haardt.
4160 64. When Exim failed to open the panic log after failing to open the main log,
4161 the original message it was trying to log was written to stderr and debug
4162 output, but if they were not available (the usual case in production), it
4163 was lost. Now it is written to syslog before the two lines that record the
4164 failures to open the logs.
4166 65. Users' Exim filters run in subprocesses under the user's uid. It is
4167 possible for a "deliver" command or an alias in a "personal" command to
4168 provoke an address rewrite. If logging of address rewriting is configured,
4169 this fails because the process is not running as root or exim. There may be
4170 a better way of dealing with this, but for the moment (because 4.30 needs
4171 to be released), I have disabled address rewrite logging when running a
4172 filter in a non-root, non-exim process.
4178 1. The buildconfig auxiliary program wasn't quoting the value set for
4179 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4180 not defined. This bug was masked in 4.22 by the effect that was fixed in
4183 2. Some messages that were rejected after a message id was allocated were
4184 shown as "incomplete" by exigrep. It no longer does this for messages that
4185 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4187 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4188 have allow_domain_literals set, the ID did not get logged in the <= line.
4189 Domain literals are now always recognized in Message-ID: header lines.
4191 4. The first argument for a ${extract expansion item is the key name or field
4192 number. Leading and trailing spaces in this item were not being ignored,
4193 causing some misleading effects.
4195 5. When deliver_drop_privilege was set, single queue runner processes started
4196 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4197 same command in the process it spins off) were not dropping privilege.
4199 6. When the daemon running as "exim" started a queue runner, it always
4200 re-executed Exim in the spun-off process. This is a waste of effort when
4201 deliver_drop_privilege is set. The new process now just calls the
4202 queue-runner function directly.
4208 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4211 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4212 ignored. Though the use of -oP was forcing the writing of a pid file, it
4213 was always written to the default place.
4215 3. If the message "no IP address found for host xxxx" is generated during
4216 incoming verification, it is now followed by identification of the incoming
4217 connection (so you can more easily find what provoked it).
4219 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4221 5. Added some features to "harden" Exim a bit more against certain attacks:
4223 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4224 be put in Local/Makefile. This is like the never_users runtime option,
4225 but it cannot be overridden. The default setting is "root".
4227 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4228 prefix string with which any file named in a -C command line option
4231 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4232 is retained for -C and -D only if the caller of Exim is root. Without
4233 it, the exim user may also use -C and -D and retain privilege.
4235 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4236 command line option is disabled.
4238 6. Macro names set by the -D option must start with an upper case letter, just
4239 like macro names defined in the configuration file.
4241 7. Added "dereference=" facility to LDAP.
4243 8. Two instances of the typo "uknown" in the source files are fixed.
4245 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4246 the Configure-Makefile script screwed up while processing it.
4248 10. Incorporated PCRE 4.4.
4250 11. The SMTP synchronization check was not operating right at the start of an
4251 SMTP session. For example, it could not catch a HELO sent before the client
4252 waited for the greeting. There is now a check for outstanding input at the
4253 point when the greeting is written. Because of the duplex, asynchronous
4254 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4255 way, but not yet received, when the check is performed.
4257 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4258 on TCP/IP sockets, because this apparently causes some broken clients to
4261 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4262 unchanged) from the Cygwin maintainer.
4264 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4265 was supported instead of testing for mbx. Effectively a typo.
4267 15. The spa authenticator server code was not checking that the input it
4268 received was valid base64.
4270 16. The debug output line for the "set" modifier in ACLs was not showing the
4271 name of the variable that was being set.
4273 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4275 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4276 lookup on its own. Something else was needed to trigger the lookup. For
4277 example, a match in host_lookup or the need for a host name in a host list.
4278 Now, if $sender_host_name is referenced and the host name has not yet been
4279 looked up, a lookup is performed. If the lookup fails, the variable remains
4280 empty, and $host_lookup_failed is set to "1".
4282 19. Added "eqi" as a case-independent comparison operator.
4284 20. The saslauthd authentication condition could segfault if neither service
4285 nor realm was specified.
4287 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4288 error message that was logged was misleading, and incoming SMTP
4289 connections were dropped. The message is now more accurate, and temporary
4290 errors are given to SMTP connections.
4292 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4293 (except RSET) with a 421 error, until QUIT is received. However, it was
4294 failing to send a response to QUIT.
4296 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4297 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4298 ACL is now run after the helo_try_verify_hosts code.
4300 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4301 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4302 case-independent, so other case variants are also recognized). Apparently
4303 some systems use these upper case variants.
4305 25. If more than two messages were waiting for the same host, and a transport
4306 filter was specified for the transport, Exim sent two messages over the
4307 same TCP/IP connection, and then failed with "socket operation on non-
4308 socket" when it tried to send the third.
4310 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4312 27. The extern definition of crypt16() in expand.c was not being excluded when
4313 the OS had its own crypt16() function.
4315 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4316 as a preferred synonym for return_size_limit, both as an option and as an
4319 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4321 30. Changed the default configuration ACL to relax the local part checking rule
4322 for addresses that are not in any local domains. For these addresses,
4323 slashes and pipe symbols are allowed within local parts, but the sequence
4324 /../ is explicitly forbidden.
4326 31. SPA server authentication was not clearing the challenge buffer before
4329 32. log_message in a "warn" ACL statement was writing to the reject log as
4330 well as to the main log, which contradicts the documentation and doesn't
4331 seem right (because no rejection is happening). So I have stopped it.
4333 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4334 However, I am unable to do any testing of this.
4336 34. Fixed an infelicity in the appendfile transport. When checking directories
4337 for a mailbox, to see if any needed to be created, it was accidentally
4338 using path names with one or more superfluous leading slashes; tracing
4339 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4341 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4342 contents are added to the log line that is written for every discarded
4343 recipient. (Previously a log_message setting was ignored.)
4345 36. The ${quote: operator now quotes the string if it is empty.
4347 37. The install script runs exim in order to find its version number. If for
4348 some reason other than non-existence or emptiness, which it checks, it
4349 could not run './exim', it was installing it with an empty version number,
4350 i.e. as "exim-". This error state is now caught, and the installation is
4353 38. An argument was missing from the function that creates an error message
4354 when Exim fails to connect to the socket for saslauthd authentication.
4355 This could cause Exim to crash, or give a corrupted message.
4357 39. Added isip, isip4, and isip6 to ${if conditions.
4359 40. The ACL variables $acl_xx are now saved with the message, and can be
4360 accessed later in routers, transports, and filters.
4362 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4363 strings in the file are not string-expanded.
4365 42. If a MAIL command specified a SIZE value that was too large to fit into an
4366 int variable, the check against message_size_limit failed. Such values are
4367 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4368 day this will have to be increased, but I don't think I want to be around
4369 when emails are that large.
4376 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4377 iconv() is not standard in FreeBSD.
4379 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4380 IPv6 enabled. The observed symptom was a segmentation fault on return from
4381 the function os_common_find_running_interfaces() in src/os.c.
4383 3. In the check_special_case() function in daemon.c I had used "errno" as an
4384 argument name, which causes warnings on some systems. This was basically a
4385 typo, since it was named "eno" in the comments!
4387 4. The code that waits for the clock to tick (at a resolution of some fraction
4388 of a second) so as to ensure message-id uniqueness was always waiting for
4389 at least one whole tick, when it could have waited for less. [This is
4390 almost certainly not relevant at current processor speeds, where it is
4391 unlikely to ever wait at all. But we try to future-proof.]
4393 5. The function that sleeps for a time interval that includes fractions of a
4394 second contained a race. It did not block SIGALRM between setting the
4395 timer, and suspending (a couple of lines later). If the interval was short
4396 and the sigsuspend() was delayed until after it had expired, the suspension
4397 never ended. On busy systems this could lead to processes getting stuck for
4400 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4401 process, before it forks any delivery processes. The open lookup caching
4402 mechanism meant that the open file or database connection was passed into
4403 the delivery process. The problem was that delivery processes always tidy
4404 up cached lookup data. This could cause a problem for the next delivery
4405 process started by the queue runner, because the external queue runner
4406 process does not know about the closure. So the next delivery process
4407 still has data in the lookup cache. In the case of a file lookup, there was
4408 no problem because closing a file descriptor in a subprocess doesn't affect
4409 the parent. However, if the lookup was caching a connection to a database,
4410 the connection was closed, and the second delivery process was likely to
4411 see errors such as "PGSQL: query failed: server closed the connection
4412 unexpectedly". The problem has been fixed by closing all cached lookups
4413 in a queue runner before running a delivery process.
4415 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4416 seem to have the "const" qualifier which it has on other OS. I've
4419 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4420 *of the same type* (client or server) with the same public name that an
4421 error should be diagnosed.
4423 9. When Exim looked up a host name for an IP address, but failed to find the
4424 original IP address when looking up the host name (a safety check), it
4425 output the message "<ip address> does not match any IP for NULL", which was
4426 confusing, to say the least. The bug was that the host name should have
4427 appeared instead of "NULL".
4429 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4430 user that is built into the binary, and the -C or -D options is used, root
4431 privilege is dropped before the configuration file is read. In addition,
4432 logging is switched to stderr instead of the normal log files. If the
4433 configuration then re-defines the Exim user, the unprivileged environment
4434 is probably not what is expected, so Exim logs a panic warning message (but
4437 However, if deliver_drop_privilege is set, the unprivileged state may well
4438 be exactly what is intended, so the warning has been cut out in that case,
4439 and Exim is allowed to try to write to its normal log files.
4445 1. smtp_return_error_details was not giving details for temporary sender
4446 or receiver verification errors.
4448 2. Diagnose a configuration error if two authenticators have the same public
4451 3. Exim used not to create the message log file for a message until the first
4452 delivery attempt. This could be confusing when incoming messages were held
4453 for policy or load reasons. The message log file is now created at the time
4454 the message is received, and an initial "Received" line is written to it.
4456 4. The automatically generated man page for command line options had a minor
4457 bug that caused no ill effects; however, a more serious problem was that
4458 the procedure for building the man page automatically didn't always
4459 operate. Consequently, release 4.20 contains an out-of-date version. This
4460 shouldn't happen again.
4462 5. When building Exim with embedded Perl support, the script that builds the
4463 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4464 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4466 6. The freeze_tell option was not being used for messages that were frozen on
4467 arrival, either by an ACL or by local_scan().
4469 7. Added the smtp_incomplete_transaction log selector.
4471 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4472 was accepting AUTH without a new EHLO.
4474 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4475 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4478 10. Logging of TCP/IP connections (when configured) now happens in the main
4479 daemon process instead of the child process, so that the TCP/IP connection
4480 count is more accurate (but it can never be perfect).
4482 11. The use of "drop" in a nested ACL was not being handled correctly in the
4483 outer ACL. Now, if condition failure induced by the nested "drop" causes
4484 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4485 or "require"), the connection is dropped.
4487 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4488 that yield "discard" can now be used with an "accept" or a "discard" verb,
4489 but an error is generated for any others (because I can't see a useful way
4490 to define what should happen).
4492 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4493 are now processed in the same way as lines in the Exim configuration file.
4494 In particular, continuation lines are supported.
4496 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4498 15. Added -ti meaning -t -i.
4500 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4501 domains, and warn by logging if others are found.
4503 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4504 returned from GIFCONF when Exim is trying to find the list of interfaces on
4505 a host. The code in os.c has been modified to copy each ifreq to an aligned
4506 structure in all cases.
4508 Also, in some cases, the returned ifreq's were being copied to a 'struct
4509 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4510 means the last couple of bytes of an IPv6 address could be chopped if the
4511 ifreq contained only a normal sockaddr (14 bytes storage).
4513 18. Named domain lists were not supported in the hosts_treat_as_local option.
4514 An entry such as +xxxx was not recognized, and was treated as a literal
4517 19. Ensure that header lines added by a DATA ACL are included in the reject log
4518 if the ACL subsequently rejects the message.
4520 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4521 MD5 (which is deprecated).
4523 21. When testing a filter file using -bf, Exim was writing a message when it
4524 took the sender from a "From " line in the message, but it was not doing so
4525 when it took $return_path from a Return-Path: header line. It now does.
4527 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4528 with a valid header line field name (a series of printing characters
4529 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4531 23. Changed "disc" in the source to "disk" to conform to the documentation and
4532 the book and for uniformity.
4534 24. Ignore Sendmail's -Ooption=value command line item.
4536 25. When execve() failed while trying to run a command in a pipe transport,
4537 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4538 could be confused with a return value of 69 from the command itself. This
4539 has been changed to 127, the value the shell returns if it is asked to run
4540 a non-existent command. The wording for the related log line suggests a
4541 non-existent command as the problem.
4543 26. If received_header_text expands to an empty string, do not add a Received:
4544 header line to the message. (Well, it adds a token one on the spool, but
4545 marks it "old" so that it doesn't get used or transmitted.)
4547 27. Installed eximstats 1.28 (addition of -nt option).
4549 28. There was no check for failure on the call to getsockname() in the daemon
4550 code. This can fail if there is a shortage of resources on the system, with
4551 ENOMEM, for example. A temporary error is now given on failure.
4553 29. Contrary to the C standard, it seems that in some environments, the
4554 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4555 program. Exim now does this explicitly; it affects the formatting of
4556 timestamps using strftime().
4558 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4559 complaints. I've now initialized all the variables, to avoid this.
4561 32. Header lines added by a system filter were not being "seen" during
4562 transport-time rewrites.
4564 33. The info_callback() function passed to OpenSSL is set up with type void
4565 (*)(SSL *, int, int), as described somewhere. However, when calling the
4566 function (actually a macro) that sets it up, the type void(*)() is
4567 expected. I've put in a cast to prevent warnings from picky compilers.
4569 34. If a DNS black list lookup found a CNAME record, but there were no A
4570 records associated with the domain it pointed at, Exim crashed.
4572 35. If a DNS black list lookup returned more than one A record, Exim ignored
4573 all but the first. It now scans all returned addresses if a particular IP
4574 value is being sought. In this situation, the contents of the
4575 $dnslist_value variable are a list of all the addresses, separated by a
4578 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4579 to accept a host name and all its aliases if the forward lookup for any of
4580 them yielded the IP address of the incoming connection. Now it accepts only
4581 those names whose forward lookup yields the correct IP address. Any other
4582 names are discarded. This closes a loophole whereby a rogue DNS
4583 administrator could create reverse DNS records to break through a
4584 wildcarded host restriction in an ACL.
4586 37. If a user filter or a system filter that ran in a subprocess used any of
4587 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4588 the wrong values were passed to the pipe command ($thisaddress had the
4589 value of $0, $0 had the value of $1, etc). This bug was introduced by
4590 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4592 38. Improved the line breaking for long SMTP error messages from ACLs.
4593 Previously, if there was no break point between 40 and 75 characters, Exim
4594 left the rest of the message alone. Two changes have been made: (a) I've
4595 reduced the minimum length to 35 characters; (b) if it can't find a break
4596 point between 35 and 75 characters, it looks ahead and uses the first one
4597 that it finds. This may give the occasional overlong line, but at least the
4598 remaining text gets split now.
4600 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4601 file descriptors might be low, and that setting 1000 would always raise it.
4602 It turns out that in some environments, the limit is already over 1000 and
4603 that lowering it causes trouble. So now Exim takes care not to decrease it.
4605 40. When delivering a message, the value of $return_path is set to $sender_
4606 address at the start of routing (routers may change the value). By an
4607 oversight, this default was not being set up when an address was tested by
4608 -bt or -bv, which affected the outcome if any router or filter referred to
4611 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4612 log only when "message" or "log_message" are set. However, if one of the
4613 conditions was an address verification, or a call to a nested ACL, the
4614 messages generated by the underlying test were being passed through. This
4615 no longer happens. The underlying message is available in $acl_verify_
4616 message for both "message" and "log_message" expansions, so it can be
4617 passed through if needed.
4619 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4620 new expansion $bh_ to give the encoded byte string without charset
4621 translation. Translation happens only if iconv() is available; HAVE_ICONV
4622 indicates this at build time. HEADERS_CHARSET gives the charset to
4623 translate to; headers_charset can change it in the configuration, and
4624 "headers charset" can change it in an individual filter file.
4626 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4627 that creates RFC 2047 encoded "words" labels them as that charset instead
4628 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4629 expansion operator; (ii) when Exim creates a From: line for a local
4630 message; (iii) when a header line is rewritten to include a "phrase" part.
4632 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4633 buggy, causing it to skip the first lines of messages whose message ID
4634 ended in 'D'. This would not have bitten before Exim release 4.14, because
4635 message IDs were unlikely to end in 'D' before then. The effect was to have
4636 incorrect size information for certain domains.
4638 45. #include "config.h" was missing at the start of the crypt16.c module. This
4639 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4642 46. If there was a timeout during a "random" callout check, Exim treated it as
4643 a failure of the random address, and carried on sending RSET and the real
4644 address. If the delay was just some slowness somewhere, the response to the
4645 original RCPT would be taken as a response to RSET and so on, causing
4646 mayhem of various kinds.
4648 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4649 when I implemented it. It didn't allow for the fact that some option values
4650 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4651 the right test for positive values.
4653 48. Domain names in DNS records are case-independent. Exim always looks them up
4654 in lower case. Some resolvers return domain names in exactly the case they
4655 appear in the zone file, that is, they may contain uppercase letters. Not
4656 all resolvers do this - some return always lower case. Exim was treating a
4657 change of case by a resolver as a change of domain, similar to a widening
4658 of a domain abbreviation. This triggered its re-routing code and so it was
4659 trying to route what was effectively the same domain again. This normally
4660 caused routing to fail (because the router wouldn't handle the domain
4661 twice). Now Exim checks for this case specially, and just changes the
4662 casing of the domain that it ultimately uses when it transmits the message
4665 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4668 50. If a filter generated a file delivery with a non-absolute name (possible if
4669 no home directory exists for the router), the forbid_file option was not
4672 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4673 the existing equality matching.
4675 52. Exim was using ints instead of ino_t variables in some places where it was
4676 dealing with inode numbers.
4678 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4679 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4680 called TMPDIR, and if it finds it is different, it changes its value.
4682 54. The smtp_printf() function is now made available to local_scan() so
4683 additional output lines can be written before returning. There is also an
4684 smtp_fflush() function to enable the detection of a dropped connection.
4685 The variables smtp_input and smtp_batched_input are exported to
4688 55. Changed the default runtime configuration: the message "Unknown user"
4689 has been removed from the ACL, and instead placed on the localuser router,
4690 using the cannot_route_message feature. This means that any verification
4691 failures that generate their own messages won't get overridden. Similarly,
4692 the "Unrouteable address" message that was in the ACL for unverifiable
4693 relay addresses has also been removed.
4695 56. Added hosts_avoid_esmtp to the smtp transport.
4697 57. The exicyclog script was not checking for the esoteric option
4698 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4699 will work only if exicyclog is run under the appropriate euid.
4701 58. Following a discussion on the list, the rules by which Exim recognises line
4702 endings on incoming messages have been changed. The -dropcr and drop_cr
4703 options are now no-ops, retained only for backwards compatibility. The
4704 following line terminators are recognized: LF CRLF CR. However, special
4705 processing applies to CR:
4707 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4708 nor a local message in the state where . is a terminator.
4710 (ii) If a bare CR is encountered in a header line, an extra space is added
4711 after the line terminator so as not to end the header. The reasoning
4712 behind this is that bare CRs in header lines are most likely either
4713 to be mistakes, or people trying to play silly games.
4715 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4716 was being incorrectly given as 18 bytes larger than it should have been.
4717 This is a VOB (very old bug).
4719 60. This may never have affected anything current, but just in case it has:
4720 When the local host is found other than at the start of a list of hosts,
4721 the local host, those with the same MX, and any that follow, are discarded.
4722 When the list in question was part of a longer list of hosts, the following
4723 hosts (not currently being processed) were also being discarded. This no
4724 longer happens. I'm not sure if this situation could ever has previously
4727 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4730 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4731 header. This is something that is recommended in a new Internet Draft, and
4732 is something that is documented as being done by Sendmail. There are two
4733 possible values. For messages generated by the autoreply transport, Exim
4736 Auto-Submitted: auto-replied
4738 whereas for all other generated messages (e.g. bounces) it adds
4740 Auto-Submitted: auto-generated
4742 63. The "personal" condition in filters now includes a test for the
4743 Auto-Submitted: header. If it contains the string "auto-" the message it
4744 not considered personal.
4746 64. Added rcpt_include_affixes as a generic transport option.
4748 65. Added queue_only_override (default true).
4750 66. Added the syslog_duplication option.
4752 67. If what should have been the first header line of a message consisted of
4753 a space followed by a colon, Exim was mis-interpreting it as a header line.
4754 It isn't of course - it is syntactically invalid and should therefore be
4755 treated as the start of the message body. The misbehaviour could have
4756 caused a number of strange effects, including loss of data in subsequent
4757 header lines, and spool format errors.
4759 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4760 client host had authenticated. This control can now be exercised by an ACL
4761 for more flexibility.
4763 69. By default, callouts do not happen when testing with -bh. There is now a
4764 variant, -bhc, which does actually run the callout code, including
4765 consulting and updating the callout cache.
4767 70. Added support for saslauthd authentication, courtesy of Alexander
4770 71. If statvfs() failed on the spool or log directories while checking their
4771 size for availability, Exim confusingly gave the error "space shortage".
4772 Furthermore, in debugging mode it crashed with a floating point exception.
4773 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4774 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4775 serious problem, Exim now writes to the main and panic logs when this
4776 happens, with details of the failure. It then refuses to accept the
4777 incoming message, giving the message "spool directory problem" or "log
4778 directory problem" with a 421 code for SMTP messages.
4780 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4781 0, 1, and 2 exist, because some OS complain for execs without them (see
4782 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4783 descriptors. However, the code omitted to check that the open succeeded,
4784 causing mysterious errors if for some reason the permissions on /dev/null
4785 got screwed. Now Exim writes a message to the main and panic logs, and
4786 bombs out if it can't open /dev/null.
4788 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4789 interact so that it is all more flexible. It is supposed to remain
4790 backwards compatible. Also added extra_local_interfaces.
4792 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4793 to bomb out with an assertion failure - to the client this appears as a
4794 connection drop. This problem occurs in the part of the code that was taken
4795 from the Samba project. Fortunately, the assertion is in a very simple
4796 function, so I have fixed this by reproducing the function inline in the
4797 one place where it is called, and arranging for authentication to fail
4798 instead of killing the process with assert().
4800 75. The SPA client code was not working when the server requested OEM rather
4801 than Unicode encoding.
4803 76. Added code to make require_files with a specific uid setting more usable in
4804 the case where statting the file as root fails - usually a non-root-mounted
4805 NFS file system. When this happens and the failure is EACCES, Exim now
4806 forks a subprocess and does the per-uid checking as the relevant uid.
4808 77. Added process_log_path.
4810 78. If log_file_path was not explicitly set, a setting of check_log_space or
4811 check_log_inodes was ignored.
4813 79. If a space check for the spool or log partitions fails, the incident is now
4814 logged. Of course, in the latter case the data may get lost...
4816 80. Added the %p formatting code to string_format() so that it can be used to
4817 print addresses in debug_print(). Adjusted all the address printing in the
4818 debugging in store.c to use %p rather than %d.
4820 81. There was a concern that a line of code in smtp_in.c could overflow a
4821 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4822 initially expressed, the concern was not well-founded, because trailing
4823 spaces are removed early. However, if the trailing spaces were followed by
4824 a NULL, they did not get removed, so the overflow was possible. Two fixes
4827 (a) I re-wrote the offending code in a cleaner fashion.
4828 (b) If an incoming SMTP command contains a NULL character, it is rejected
4831 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4832 runs initgroups(), so that if the Exim user is in any additional groups,
4833 they will be used during message reception.
4839 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob