1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.527 2007/08/23 11:01:49 ph10 Exp $
3 Change log file for Exim from version 4.21
4 -------------------------------------------
9 PH/01 Another patch from the Sieve maintainer.
11 PH/02 When an IPv6 address is converted to a string for single-key lookup
12 in an address list (e.g. for an item such as "net24-dbm;/net/works"),
13 dots are used instead of colons so that keys in lsearch files need not
14 contain colons. This was done some time before quoting was made available
15 in lsearch files. However, iplsearch files do require colons in IPv6 keys
16 (notated using the quote facility) so as to distinguish them from IPv4
17 keys. This meant that lookups for IP addresses in host lists did not work
18 for iplsearch lookups.
20 This has been fixed by arranging for IPv6 addresses to be expressed with
21 colons if the lookup type is iplsearch. This is not incompatible, because
22 previously such lookups could never work.
24 The situation is now rather anomolous, since one *can* have colons in
25 ordinary lsearch keys. However, making the change in all cases is
26 incompatible and would probably break a number of configurations.
28 TK/01 Change PRVS address formatting scheme to reflect latests BATV draft
31 MH/01 The "spam" ACL condition code contained a sscanf() call with a %s
32 conversion specification without a maximum field width, thereby enabling
33 a rogue spamd server to cause a buffer overflow. While nobody in their
34 right mind would setup Exim to query an untrusted spamd server, an
35 attacker that gains access to a server running spamd could potentially
36 exploit this vulnerability to run arbitrary code as the Exim user.
38 TK/02 Bugzilla 502: Apply patch to make the SPF-Received: header use
39 $primary_hostname instead of what libspf2 thinks the hosts name is.
41 MH/02 The dsearch lookup now uses lstat(2) instead of stat(2) to look for
42 a directory entry by the name of the lookup key. Previously, if a
43 symlink pointed to a non-existing file or a file in a directory that
44 Exim lacked permissions to read, a lookup for a key matching that
45 symlink would fail. Now it is enough that a matching directory entry
46 exists, symlink or not. (Bugzilla 503.)
48 PH/03 The body_linecount and body_zerocount variables are now exported in the
51 PH/04 Added the $dnslist_matched variable.
53 PH/05 Unset $tls_cipher and $tls_peerdn before making a connection as a client.
54 This means they are set thereafter only if the connection becomes
57 PH/06 Added the client_condition to authenticators so that some can be skipped
58 by clients under certain conditions.
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.
104 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
105 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
108 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
109 issue a MAIL command.
111 PH/02 In an ACL statement such as
113 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
115 if a client was not listed at all, or was listed with a value other than
116 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
117 the condition was not true (as it should be), so access was not denied.
118 The bug was that the ! inversion was incorrectly passed on to the second
119 item. This has been fixed.
121 PH/03 Added additional dnslists conditions == and =& which are different from
122 = and & when the dns lookup returns more than one IP address.
124 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
125 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
127 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
128 FSYNC, which compiles an option called disable_fsync that allows for
129 bypassing fsync(). The documentation is heavily laced with warnings.
131 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
133 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
134 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
135 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
136 including adding "make clean"; (3) Added -fPIC when compiling the test
137 dynamically loaded module, to get rid of a warning.
139 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
140 message fails, move_frozen_messages = true and ignore_bounce_errors_after
141 = 0s. The bug is otherwise harmless.
143 PH/07 There was a bug in the dovecot authenticator such that the value of
144 $auth1 could be overwritten, and so not correctly preserved, after a
145 successful authentication. This usually meant that the value preserved by
146 the server_setid option was incorrect.
148 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
150 PH/09 Installed PCRE release 7.0.
152 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
153 run for batched SMTP input. It is now run at the start of every message
154 in the batch. While fixing this I discovered that the process information
155 (output by running exiwhat) was not always getting set for -bs and -bS
156 input. This is fixed, and it now also says "batched" for BSMTP.
158 PH/11 Added control=no_pipelining.
160 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
161 patch, slightly modified), and move the expansion of helo_data till after
162 the connection is made in the smtp transport (so it can use these
165 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
167 PH/14 Added log_selector = +pid.
169 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
171 PH/16 Add ${if forany and ${if forall.
173 PH/17 Added dsn_from option to vary the From: line in DSNs.
175 PH/18 Flush SMTP output before performing a callout, unless control =
176 no_callout_flush is set.
178 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
179 was true (the default) a successful delivery failed to delete the retry
180 item, thus causing premature timeout of the address. The bug is now
183 PH/20 Added hosts_avoid_pipelining to the smtp transport.
185 PH/21 Long custom messages for fakedefer and fakereject are now split up
186 into multiline reponses in the same way that messages for "deny" and
187 other ACL rejections are.
189 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
190 with slight modification.
192 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
193 draft, changing the syntax and factoring some duplicate code".
195 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
196 for deliveries of the second and subsequent messages over the same SMTP
199 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
200 ${reduce, with only minor "tidies".
202 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
204 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
205 expansion side effects.
207 PH/27 When a message times out after an over-quota error from an Exim-imposed
208 quota, the bounce message says "mailbox is full". This message was not
209 being given when it was a system quota that was exceeded. It now should
212 MH/03 Made $recipients available in local_scan(). local_scan() already has
213 better access to the recipient list through recipients_list[], but
214 $recipients can be useful in postmaster-provided expansion strings.
216 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
217 in the case of a MAIL command with additional options following the
218 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
219 were accidentally chopped off.
221 PH/29 SMTP synchronization checks are implemented when a command is read -
222 there is a check that no more input is waiting when there shouldn't be
223 any. However, for some commands, a delay in an ACL can mean that it is
224 some time before the response is written. In this time, more input might
225 arrive, invalidly. So now there are extra checks after an ACL has run for
226 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
227 pipelining has not been advertised.
229 PH/30 MH's patch to allow iscntrl() characters to be list separators.
231 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
232 returned in the SMTP response when smtp_return_error_details was false.
235 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
236 instead of the C I/O that was originally supplied, because problems were
239 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
240 Exim which did not show up earlier: it was assuming that a call to
241 SSL_CTX_set_info_callback() might give an error value. In fact, there is
242 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
243 was a macro that became an assignment, so it seemed to work. This has
244 changed to a proper function call with a void return, hence the compile
245 error. Exim's code has been fixed.
247 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
250 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
252 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
254 PH/37 If a message is not accepted after it has had an id assigned (e.g.
255 because it turns out to be too big or there is a timeout) there is no
256 "Completed" line in the log. When some messages of this type were
257 selected by exigrep, they were listed as "not completed". Others were
258 picked up by some special patterns. I have improved the selection
259 criteria to be more general.
261 PH/38 The host_find_failed option in the manualroute router can now be set
262 to "ignore", to completely ignore a host whose IP address cannot be
263 found. If all hosts are ignored, the behaviour is controlled by the new
264 host_all_ignored option.
266 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
267 homing or because of multiple MX records with /mx) generated more than
268 one IP address, and the following item turned out to be the local host,
269 all the secondary addresses of the first item were incorrectly removed
270 from the list, along with the local host and any following hosts (which
271 is what is supposed to happen).
273 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
274 whoever called Exim into the -H file. In the case of the daemon it was
275 behaving confusingly. When first started, it used values for whoever
276 started the daemon, but after a SIGHUP it used the Exim user (because it
277 calls itself on a restart). I have changed the code so that it now always
280 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
281 message are rejected with the same error (e.g. no authentication or bad
282 sender address), and a DATA command is nevertheless sent (as can happen
283 with PIPELINING or a stupid MUA), the error message that was given to the
284 RCPT commands is included in the rejection of the DATA command. This is
285 intended to be helpful for MUAs that show only the final error to their
288 PH/42 Another patch from the Sieve maintainer.
290 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
291 Eximstats - Fixed some broken HTML links and added missing column headers
293 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
294 columns for Rejects, Temp Rejects, Ham, and Spam rows.
296 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
298 PH/43 Yet another patch from the Sieve maintainer.
300 PH/44 I found a way to check for a TCP/IP connection going away before sending
301 the response to the final '.' that terminates a message, but only in the
302 case where the client has not sent further data following the '.'
303 (unfortunately, this is allowed). However, in many cases there won't be
304 any further data because there won't be any more messages to send. A call
305 to select() can be used: if it shows that the input is "ready", there is
306 either input waiting, or the socket has been closed. An attempt to read
307 the next input character can distinguish the two cases. Previously, Exim
308 would have sent an OK response which the client would never have see.
309 This could lead to message repetition. This fix should cure that, at
310 least in a lot of common cases.
312 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
313 advertised in response to EHLO.
319 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
320 fixed by 4.65/MH/01 (is this a record?) are fixed:
322 (i) An empty string was always treated as zero by the numeric comparison
323 operators. This behaviour has been restored.
325 (ii) It is documented that the numeric comparison operators always treat
326 their arguments as decimal numbers. This was broken in that numbers
327 starting with 0 were being interpreted as octal.
329 While fixing these problems I realized that there was another issue that
330 hadn't been noticed. Values of message_size_limit (both the global option
331 and the transport option) were treated as octal if they started with 0.
332 The documentation was vague. These values are now always treated as
333 decimal, and I will make that clear in the documentation.
339 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
340 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
343 MH/01 Don't check that the operands of numeric comparison operators are
344 integers when their expansion is in "skipping" mode (fixes bug
345 introduced by 4.64-PH/07).
347 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
348 child addresses, Exim now panics and dies. Previously, because the count
349 is held in a short int, deliveries were likely to be lost. As such a
350 large number of recipients for a single message is ridiculous
351 (performance will be very, very poor), I have chosen to impose a limit
352 rather than extend the field.
358 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
359 leftover -K file (the existence of which was triggered by #402).
360 While we were at it, introduced process PID as part of the -K
361 filename. This should rule out race conditions when creating
364 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
365 processing considerably. Previous code took too long for large mails,
366 triggering a timeout which in turn triggers #401.
368 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
369 in the DK code in transports.c. sendfile() is not really portable,
370 hence the _LINUX specificness.
372 TF/01 In the add_headers option to the mail command in an Exim filter,
373 there was a bug that Exim would claim a syntax error in any
374 header after the first one which had an odd number of characters
377 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
378 callout verification, Exim cached a "reject" for the entire domain. This
379 is correct for most verifications, but it is not correct for a recipient
380 verification with use_sender or use_postmaster set, because in that case
381 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
382 case of MAIL FROM:<> rejection from other early rejections (e.g.
383 rejection of HELO). When verifying a recipient using a non-null MAIL
384 address, the cache is ignored if it shows MAIL FROM:<> rejection.
385 Whatever the result of the callout, the value of the domain cache is
386 left unchanged (for any other kind of callout, getting as far as trying
387 RCPT means that the domain itself is ok).
389 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
392 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
393 manifest itself as EPIPE rather than ECONNECT. When tidying away a
394 session, the daemon ignores ECONNECT errors and logs others; it now
395 ignores EPIPE as well.
397 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
398 (quoted-printable decoding).
400 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
401 later the small subsequent patch to fix an introduced bug.
403 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
405 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
407 PH/08 An error is now given if message_size_limit is specified negative.
409 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
410 to be given (somewhat) arbitrary names.
412 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
415 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
416 miscellaneous code fixes
418 PH/10 Added the log_reject_target ACL modifier to specify where to log
421 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
422 hostname. This is wrong, because it relates to the incoming message (and
423 probably the interface on which it is arriving) and not to the outgoing
424 callout (which could be using a different interface). This has been
425 changed to use the value of the helo_data option from the smtp transport
426 instead - this is what is used when a message is actually being sent. If
427 there is no remote transport (possible with a router that sets up host
428 addresses), $smtp_active_hostname is used.
430 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
431 tweaks were necessary in order to get it to work (see also 21 below):
432 (a) The code assumed that strncpy() returns a negative number on buffer
433 overflow, which isn't the case. Replaced with Exim's string_format()
435 (b) There were several signed/unsigned issues. I just did the minimum
436 hacking in of casts. There is scope for a larger refactoring.
437 (c) The code used strcasecmp() which is not a standard C function.
438 Replaced with Exim's strcmpic() function.
439 (d) The code set only $1; it now sets $auth1 as well.
440 (e) A simple test gave the error "authentication client didn't specify
441 service in request". It would seem that Dovecot has changed its
442 interface. Fortunately there's a specification; I followed it and
443 changed what the client sends and it appears to be working now.
445 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
448 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
449 address A is aliased to B and C, where B exists and C does not. Without
450 -v the output is "A verified" because verification stops after a
451 successful redirection if more than one address is generated. However,
452 with -v the child addresses are also verified. Exim was outputting "A
453 failed to verify" and then showing the successful verification for C,
454 with its parentage. It now outputs "B failed to verify", showing B's
455 parentage before showing the successful verification of C.
457 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
458 look up a TXT record in a specific list after matching in a combined
461 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
462 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
463 they consult the DNS. I had assumed they would set it the way they
464 wanted; and indeed my experiments on Linux seem to show that in some
465 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
466 To be on the safe side, however, I have now made the interface to
467 host_find_byname() similar to host_find_bydns(), with an argument
468 containing the DNS resolver options. The host_find_byname() function now
469 sets these options at its start, just as host_find_bydns() does. The smtp
470 transport options dns_qualify_single and dns_search_parents are passed to
471 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
472 of host_find_byname() use the default settings of RES_DEFNAMES
473 (qualify_single) but not RES_DNSRCH (search_parents).
475 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
476 spool_read_header() do less string testing, by means of a preliminary
477 switch on the second character of optional "-foo" lines. (This is
478 overdue, caused by the large number of possibilities that now exist.
479 Originally there were few.) While I was there, I also converted the
480 str(n)cmp tests so they don't re-test the leading "-" and the first
481 character, in the hope this might squeeze out yet more improvement.
483 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
484 flag allowing group syntax was set by the header_syntax check but not
485 turned off, possible causing trouble later; (2) The flag was not being
486 set at all for the header_verify test, causing "group"-style headers to
487 be rejected. I have now set it in this case, and also caused header_
488 verify to ignore an empty address taken from a group. While doing this, I
489 came across some other cases where the code for allowing group syntax
490 while scanning a header line wasn't quite right (mostly, not resetting
491 the flag correctly in the right place). These bugs could have caused
492 trouble for malformed header lines. I hope it is now all correct.
494 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
495 with the "reply" argument non-NULL. The code, however (which originally
496 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
497 but it didn't always do it. This confused somebody who was copying the
498 code for some other use. I have removed all the tests.
500 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
501 feature that was used to support insecure browsers during the U.S. crypto
502 embargo. It requires special client support, and Exim is probably the
503 only MTA that supported it -- and would never use it because real RSA is
504 always available. This code has been removed, because it had the bad
505 effect of slowing Exim down by computing (never used) parameters for the
506 RSA_EXPORT functionality.
508 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
509 authenticator to fail if there's a tab character in the incoming data
510 (there should never be unless someone is messing about, as it's supposed
511 to be base64-encoded). Also added, on Timo's advice, the "secured" option
512 if the connection is using TLS or if the remote IP is the same as the
513 local IP, and the "valid-client-cert option" if a client certificate has
516 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
517 authenticators. This can be used for authorization after authentication
518 succeeds. (In the case of plaintext, it servers for both authentication
521 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
522 if any retry times were supplied.
524 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
525 connection, where there is no client IP address to check. In this
526 situation, the verify now always succeeds.
528 PH/25 Applied John Jetmore's -Mset patch.
530 PH/26 Added -bem to be like -Mset, but loading a message from a file.
532 PH/27 In a string expansion for a processed (not raw) header when multiple
533 headers of the same name were present, leading whitespace was being
534 removed from all of them, but trailing whitespace was being removed only
535 from the last one. Now trailing whitespace is removed from each header
536 before concatenation. Completely empty headers in a concatenation (as
539 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
540 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
542 PH/29 [Removed. This was a change that I later backed out, and forgot to
543 correct the ChangeLog entry (that I had efficiently created) before
544 committing the later change.]
546 PH/30 Exim was sometimes attempting to deliver messages that had suffered
547 address errors (4xx response to RCPT) over the same connection as other
548 messages routed to the same hosts. Such deliveries are always "forced",
549 so retry times are not inspected. This resulted in far too many retries
550 for the affected addresses. The effect occurred only when there were more
551 hosts than the hosts_max_try setting in the smtp transport when it had
552 the 4xx errors. Those hosts that it had tried were not added to the list
553 of hosts for which the message was waiting, so if all were tried, there
554 was no problem. Two fixes have been applied:
556 (i) If there are any address or message errors in an SMTP delivery, none
557 of the hosts (tried or untried) are now added to the list of hosts
558 for which the message is waiting, so the message should not be a
559 candidate for sending over the same connection that was used for a
560 successful delivery of some other message. This seems entirely
561 reasonable: after all the message is NOT "waiting for some host".
562 This is so "obvious" that I'm not sure why it wasn't done
563 previously. Hope I haven't missed anything, but it can't do any
564 harm, as the worst effect is to miss an optimization.
566 (ii) If, despite (i), such a delivery is accidentally attempted, the
567 routing retry time is respected, so at least it doesn't keep
568 hammering the server.
570 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
571 in ${readsocket because some servers need this prod.
573 PH/32 Added some extra debug output when updating a wait-xxx database.
575 PH/33 The hint "could be header name not terminated by colon", which has been
576 given for certain expansion errors for a long time, was not being given
577 for the ${if def:h_colon_omitted{... case.
579 PH/34 The spec says: "With one important exception, whenever a domain list is
580 being scanned, $domain contains the subject domain." There was at least
581 one case where this was not true.
583 PH/35 The error "getsockname() failed: connection reset by peer" was being
584 written to the panic log as well as the main log, but it isn't really
585 panic-worthy as it just means the connection died rather early on. I have
586 removed the panic log writing for the ECONNRESET error when getsockname()
589 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
590 runs only) independently of the message's sender address. This meant
591 that, if the 4xx error was in fact related to the sender, a different
592 message to the same recipient with a different sender could confuse
593 things. In particualar, this can happen when sending to a greylisting
594 server, but other circumstances could also provoke similar problems.
595 I have changed the default so that the retry time for these errors is now
596 based a combination of the sender and recipient addresses. This change
597 can be overridden by setting address_retry_include_sender=false in the
600 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
601 remote server are returned as part of bounce messages. This was not
602 happening for LMTP over a pipe (the lmtp transport), but now it is the
603 same for both kinds of LMTP.
605 PH/38 Despite being documented as not happening, Exim was rewriting addresses
606 in header lines that were in fact CNAMEs. This is no longer the case.
608 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
609 and queue runs started by the daemon processed all messages. This has
610 been fixed so that -R and -S can now usefully be given with -q<time>.
612 PH/40 Import PCRE release 6.7 (fixes some bugs).
614 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
616 PH/42 Give an error if -q is specified more than once.
618 PH/43 Renamed the variables $interface_address and $interface_port as
619 $received_ip_address and $received_port, to make it clear that these
620 values apply to message reception, and not to the outgoing interface when
621 a message is delivered. (The old names remain recognized, of course.)
623 PH/44 There was no timeout on the connect() call when using a Unix domain
624 socket in the ${readsocket expansion. There now is.
626 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
627 be meaningful with "accept".
629 SC/01 Eximstats V1.43
630 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
632 SC/02 Eximstats V1.44
633 Use a glob alias rather than an array ref in the generated
634 parser. This improves both readability and performance.
636 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
637 Collect SpamAssassin and rejection statistics.
638 Don't display local sender or destination tables unless
639 there is data to show.
640 Added average volumes into the top table text output.
642 SC/04 Eximstats V1.46
643 Collect data on the number of addresses (recipients)
644 as well as the number of messages.
646 SC/05 Eximstats V1.47
647 Added 'Message too big' to the list of mail rejection
648 reasons (thanks to Marco Gaiarin).
650 SC/06 Eximstats V1.48
651 Mainlog lines which have GMT offsets and are too short to
652 have a flag are now skipped.
654 SC/07 Eximstats V1.49 (Alain Williams)
655 Added the -emptyok flag.
657 SC/08 Eximstats V1.50
658 Fixes for obtaining the IP address from reject messages.
660 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
661 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
662 whitesspace changes from 4.64-PH/27
664 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
667 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
668 are found, allow negative numbers in numeric criteria)
670 JJ/06 exipick.20061117.2, added new $message_body_missing variable
672 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
673 to match changes made in 4.64-PH/43
675 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
677 PH/47 Put in an explicit test for a DNS lookup of an address record where the
678 "domain" is actually an IP address, and force a failure. This locks out
679 those revolvers/nameservers that support "A-for-A" lookups, in
680 contravention of the specifications.
682 PH/48 When a host name was looked up from an IP address, and the subsequent
683 forward lookup of the name timed out, the host name was left in
684 $sender_host_name, contrary to the specification.
686 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
687 restricted to single-key lookups, Exim was not diagnosing an error if
688 * or *@ was used with a query-style lookup.
690 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
692 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
693 long ago, but noone interested enough thought of it. Let's just say that
694 the "1.1" means that there are some new functions that weren't there at
695 some point in the past.
697 PH/51 Error processing for expansion failure of helo_data from an smtp
698 transport during callout processing was broken.
700 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
701 tested/used via the -bh/-bhc/-bs options.
703 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
704 bug, fixed in subsequent PCRE releases).
706 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
707 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
709 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
715 SC/01 Use a glob alias rather than an array ref in eximstats generated
716 parser. This improves both readability and performance.
718 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
719 Don't display local sender or destination tables in eximstats unless
720 there is data to show.
721 Added average volumes into the eximstats top table text output.
723 SC/03 Collect data on the number of addresses (recipients) as well
724 as the number of messages in eximstats.
726 TF/01 Correct an error in the documentation for the redirect router. Exim
727 does (usually) call initgroups() when daemonizing.
729 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
730 with consistent privilege compared to when running as a daemon.
732 TF/03 Note in the spec that $authenticated_id is not set for local
733 submissions from trusted users.
735 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
736 Thanks to Dean Brooks <dean@iglou.com> for the patch.
738 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
739 by adding some example configuration directives to the default
740 configuration file. A little bit of work is required to uncomment the
741 directives and define how usernames and passwords are checked, but
742 there is now a framework to start from.
744 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
745 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
746 without this. I don't know how relevant this is to other LDAP libraries.
748 PH/02 Add the verb name to the "unknown ACL verb" error.
750 PH/03 Magnus Holmgren's patch for filter_prepend_home.
752 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
754 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
755 directory not expanded when it should be if an expanded home directory
756 was set for the address (which is overridden by the transport).
758 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
761 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
762 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
763 because it is too late at that time, and has no effect.
765 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
766 security issue with \' (bugzilla #107). I could not use the
767 PQescapeStringConn() function, because it needs a PGconn value as one of
770 PH/08 When testing addresses using -bt, indicate those final addresses that
771 are duplicates that would not cause an additional delivery. At least one
772 person was confused, thinking that -bt output corresponded to deliveries.
773 (Suppressing duplicates isn't a good idea as you lose the information
774 about possibly different redirections that led to the duplicates.)
776 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
777 systems where poll() doesn't work, in particular OS X.
779 PH/10 Added more information to debugging output for retry time not reached.
781 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
782 operations in malware.c.
784 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
787 PH/13 If write_rejectlog was set false when logging was sent to syslog with
788 syslog_duplication set false, log lines that would normally be written
789 both the the main log and to the reject log were not written to syslog at
792 PH/14 In the default configuration, change the use of "message" in ACL warn
793 statements to "add_header".
795 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
796 not followed by a command (e.g. "seen endif").
798 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
799 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
802 PH/17 Added extra conditions to the default value of delay_warning_condition
806 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
807 { match{$h_precedence:}{(?i)bulk|list|junk} } \
808 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
811 The Auto-Submitted: and various List- headers are standardised, whereas I
812 don't think Precedence: ever was.
814 PH/18 Refactored debugging code in route_finduser() to show more information,
815 in particular, the error code if getpwnam() issues one.
817 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
818 This is apparently needed in addition to the PH/07 change above to avoid
819 any possible encoding problems.
821 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
822 but not after initializing Perl.
824 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
825 output them only if debugging. By default they are written stderr,
826 apparently, which is not desirable.
828 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
831 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
834 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
836 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
837 authenticated or an ident call has been made. Suppress the default
838 values for $authenticated_id and $authenticated_sender (but permit -oMai
839 and -oMas) when testing with -bh.
841 PH/24 Re-jigged the order of the tests in the default configuration so that the
842 tests for valid domains and recipients precede the DNS black list and CSA
843 tests, on the grounds that those ones are more expensive.
845 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
846 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
847 This bug exists in every version of Exim that I still have, right back to
850 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
851 However, an attempt to turn on case-sensitivity in a regex key by
852 including (?-i) didn't work because the subject string was already
853 lowercased, and the effects were non-intuitive. It turns out that a
854 one-line patch can be used to allow (?-i) to work as expected.
860 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
861 other effects) broke the use of negated acl sub-conditions.
863 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
866 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
867 "Deny" causes Exim to reject the incoming connection with a 554 error.
868 Unfortunately, if there is a major crisis, such as a disk failure,
869 tcp-wrappers gives "deny", whereas what one would like would be some
870 kind of temporary error. A kludge has been added to help with this.
871 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
872 554 error is used if errno is still zero or contains ENOENT (which occurs
873 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
876 PH/03 Add -lutil to the default FreeBSD LIBS setting.
878 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
879 errors. Otherwise a message that provokes a temporary error (when other
880 messages do not) can cause a whole host to time out.
882 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
883 addresses were routed directly to files or pipes from a redirect router.
884 File deliveries just didn't batch; pipe deliveries might have suffered
887 PH/06 A failure to get a lock for a hints database would erroneously always say
888 "Failed to get write lock", even when it was really a read lock.
890 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
891 of 0600. This has been changed to use the value of the lockfile_mode
892 option (which defaults to 0600).
894 PH/08 Applied small patch from the Sieve maintainer.
896 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
897 folder from quota calculations, a direct delivery into this folder messed
898 up the contents of the maildirsize file. This was because the regex was
899 used only to exclude .Trash (or whatever) when the size of the mailbox
900 was calculated. There was no check that a delivery was happening into an
901 excluded directory. This bug has been fixed by ignoring all quota
902 processing for deliveries into excluded directories.
904 PH/10 Added the maildirfolder_create_regex option to appendfile.
910 PH/01 The code for finding all the local interface addresses on a FreeBSD
911 system running IPv6 was broken. This may well have applied to all BSD
912 systems, as well as to others that have similar system calls. The broken
913 code found IPv4 interfaces correctly, but gave incorrect values for the
914 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
915 that it would not match correctly against @[] and not recognize the IPv6
918 PH/02 The ipliteral router was not recognizing addresses of the form user@
919 [ipv6:....] because it didn't know about the "ipv6:" prefix.
921 PH/03 Added disable_ipv6.
923 PH/04 Changed $reply_address to use the raw form of the headers instead of the
924 decoded form, because it is most often used to construct To: headers
925 lines in autoreplies, and the decoded form may well be syntactically
926 invalid. However, $reply_address has leading white space removed, and all
927 newlines turned into spaces so that the autoreply transport does not
930 PH/05 If group was specified without a user on a router, and no group or user
931 was specified on a transport, the group from the router was ignored.
933 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
934 for visible compile-time settings that can be used to change these
935 numbers, for those that want even more. Backwards compatibility with old
936 spool files has been maintained. However, going back to a previous Exim
937 release will lost any variables that are in spool files.
939 PH/07 Two small changes when running in the test harness: increase delay when
940 passing a TCP/IP connection to a new process, in case the original
941 process has to generate a bounce, and remove special handling of
942 127.0.0.2 (sic), which is no longer necessary.
944 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
945 be the same on different OS.
947 PH/09 Moved a debug statement in filter processing to avoid a race problem when
950 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
951 whether --show-vars was specified or not
953 JJ/02 exipick: Added support for new ACL variable spool format introduced
956 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
957 syntactically invalid From: or Reply-to: line, and a filter used this to
958 generate an autoreply, and therefore failed to obtain an address for the
959 autoreply, Exim could try to deliver to a non-existent relative file
960 name, causing unrelated and misleading errors. What now happens is that
961 it logs this as a hard delivery error, but does not attempt to create a
964 PH/11 The exinext utility has a -C option for testing purposes, but although
965 the given file was scanned by exinext itself; it wasn't being passed on
966 when Exim was called.
968 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
969 an end-of-file indication when reading a command response.
971 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
972 compiled. In many other places in Exim, IPv6 addresses are always
973 recognized, so I have changed this. It also means that IPv4 domain
974 literals of the form [IPV4:n.n.n.n] are now always recognized.
976 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
977 used if the router is not running as root, for example, when verifying at
978 ACL time, or when using -bh. The debugging output from this situation was
979 non-existent - all you got was a failure to exec. I have made two
982 (a) Failures to set uid/gid, the current directory, or a process leader
983 in a subprocess such as that created by queryprogram now generate
984 suitable debugging ouput when -d is set.
986 (b) The queryprogram router detects when it is not running as root,
987 outputs suitable debugging information if -d is set, and then runs
988 the subprocess without attempting to change uid/gid.
990 PH/15 Minor change to Makefile for building test_host (undocumented testing
993 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
994 additional section of a DNS packet that returns MX or SRV records.
995 Instead, it always explicitly searches for A/AAAA records. This avoids
996 major problems that occur when a DNS server includes only records of one
997 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
998 fixed another bug: if SRV records were looked up and the corresponding
999 address records were *not* found in the additional section, the port
1000 values from the SRV records were lost.
1002 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
1003 using the correct key (the original address) when searching the retry
1004 rules in order to find which one to use for generating the retry hint.
1006 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
1007 adding the default one. Similarly, if it contains a Reply-To: header, the
1008 errors_reply_to option, if set, is not used.
1010 PH/19 When calculating a retry time, Exim used to measure the "time since
1011 failure" by looking at the "first failed" field in the retry record. Now
1012 it does not use this if it is later than than the arrival time of the
1013 message. Instead it uses the arrival time. This makes for better
1014 behaviour in cases where some deliveries succeed, thus re-setting the
1015 "first failed" field. An example is a quota failure for a huge message
1016 when small messages continue to be delivered. Without this change, the
1017 "time since failure" will always be short, possible causing more frequent
1018 delivery attempts for the huge message than are intended.
1019 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1021 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1022 $1, $2, $3) because the numerical variables can be reset during some
1023 expansion items (e.g. "match"), thereby losing the authentication data.
1025 PH/21 Make -bV show the size of off_t variables so that the test suite can
1026 decide whether to run tests for quotas > 2G.
1028 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1029 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1030 filecount value is greater than 2G or if a quota value is greater than 2G
1031 on a system where the size of off_t is not greater than 4, a panic error
1034 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1035 never match. The debug and -bh output now contains an explicit error
1036 message indicating a malformed IPv4 address or mask.
1038 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1039 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1040 PH/23 above applies.
1042 PH/25 Do not write to syslog when running in the test harness. The only
1043 occasion when this arises is a failure to open the main or panic logs
1044 (for which there is an explicit test).
1046 PH/26 Added the /no_tell option to "control=freeze".
1048 PH/27 If a host name lookup failed very early in a connection, for example, if
1049 the IP address matched host_lookup and the reverse lookup yielded a name
1050 that did not have a forward lookup, an error message of the form "no IP
1051 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1052 could be logged. Now it outputs the IP address instead of "NULL".
1054 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1055 allows the sender and the authenticated sender to be set when
1056 submitting a message from within Exim. Since child_open_exim() is
1057 documented for local_scan(), the new function should be too.
1059 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1060 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1061 results in an empty string is now treated as unset.
1063 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1065 PH/31 Added sender_verify_fail logging option.
1067 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1068 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1069 tidied the source and removed it altogether.
1071 PH/33 When a queue run was abandoned because the load average was too high, a
1072 log line was always written; now it is written only if the queue_run log
1073 selector is set. In addition, the log line for abandonment now contains
1074 information about the queue run such as the pid. This is always present
1075 in "start" and "stop" lines but was omitted from the "abandon" line.
1077 PH/34 Omit spaces between a header name and the colon in the error message that
1078 is given when verify = headers_syntax fails (if there are lots of them,
1079 the message gets confusing).
1081 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1082 names, as there are now some PTR records that contain slashes. This check
1083 is only to protect against broken name servers that fall over on strange
1084 characters, so the fact that it applies to all lookups doesn't matter.
1086 PH/36 Now that the new test suite is complete, we can remove some of the
1087 special code in Exim that was needed for the old test suite. For example,
1088 sorting DNS records because real resolvers return them in an arbitrary
1089 order. The new test suite's fake resolver always returns records in the
1092 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1093 bounces) except when queue_only is set, to avoid logging races between
1094 the different processes.
1096 PH/38 Panic-die if .include specifies a non-absolute path.
1098 PH/39 A tweak to the "H" retry rule from its user.
1100 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1101 a label. They prevented compilation on older perls.
1103 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1104 a warning to be raised on newish perls.
1106 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1107 on queue. Changes to match documented behaviour of showing count of
1108 messages matching specified criteria.
1110 PH/40 Changed the default ident timeout from 30s to 5s.
1112 PH/41 Added support for the use of login_cap features, on those BSD systems
1113 that have them, for controlling the resources used by pipe deliveries.
1115 PH/42 The content-scanning code uses fopen() to create files in which to put
1116 message data. Previously it was not paying any attention to the mode of
1117 the files. Exim runs with umask(0) because the rest of the code creates
1118 files with open(), and sets the required mode explicitly. Thus, these
1119 files were ending up world-writeable. This was not a big issue, because,
1120 being within the spool directory, they were not world-accessible. I have
1121 created a function called modefopen, which takes an additional mode
1122 argument. It sets umask(777), creates the file, chmods it to the required
1123 mode, then resets the umask. All the relevant calls to fopen() in the
1124 content scanning code have been changed to use this function.
1126 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1127 to 24 hours. This avoids potential overflow problems when processing G
1128 and H retry rules. I suspect nobody ever tinkers with this value.
1130 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1132 PH/45 When the plaintext authenticator is running as a client, the server's
1133 challenges are checked to ensure they are valid base64 strings. By
1134 default, the authentication attempt is cancelled if an invalid string is
1135 received. Setting client_ignore_invalid_base64 true ignores these errors.
1136 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1137 they are received. Thus, the responses can be made to depend on the
1138 challenges. If an invalid string is ignored, an empty string is placed in
1141 PH/46 Messages that are created by the autoreply transport now contains a
1142 References: header, in accordance with RFCs 2822 and 3834.
1144 PH/47 Added authenticated_sender_force to the smtp transport.
1146 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1148 PH/49 Installed latest patch from the Sieve maintainer.
1150 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1151 was also set, the appendfile transport was unnecessarily scanning a
1152 directory of message files (e.g. for maildir delivery) to find the count
1153 of files (along with the size), even though it did not need this
1154 information. It now does the scan only if it needs to find either the
1155 size of the count of files.
1157 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1159 PH/52 Two bugs concerned with error handling when the smtp transport is
1162 (i) Exim was not creating retry information for temporary errors given
1163 for individual recipients after the DATA command when the smtp transport
1164 was used in LMTP mode. This meant that they could be retried too
1165 frequently, and not timed out correctly.
1167 (ii) Exim was setting the flag that allows error details to be returned
1168 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1169 recipients that were returned after the DATA command.
1171 PH/53 This is related to PH/52, but is more general: for any failing address,
1172 when detailed error information was permitted to be returned to the
1173 sender, but the error was temporary, then after the final timeout, only
1174 "retry timeout exceeded" was returned. Now it returns the full error as
1175 well as "retry timeout exceeded".
1177 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1178 do this, and (what is worse) MTAs that accept it.
1180 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1181 will now be deprecated.
1183 PH/56 New os.c-cygwin from the Cygwin maintainer.
1185 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1186 formats (previously only available in exim formats via -bpr, -bpru,
1187 and -bpra. Now also available in native and exiqgrep formats)
1189 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1190 with very large, slow to parse queues
1192 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1194 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1196 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1197 responses to authentication challenges, though it was showing the
1198 challenges; (ii) I've removed the CR characters from the debug output for
1201 PH/58 Allow for the insertion of a newline as well as a space when a string
1202 is turned into more than one encoded-word during RFC 2047 encoding. The
1203 Sieve code now uses this.
1205 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1206 data_4xx, lost_connection, tls_required.
1208 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1209 message was being sent as an SMTP response.
1211 PH/61 Add -l and -k options to exicyclog.
1213 PH/62 When verifying, if an address was redirected to one new address, so that
1214 verification continued, and the new address failed or deferred after
1215 having set something in $address_data, the value of $address_data was not
1216 passed back to the ACL. This was different to the case when no
1217 redirection occurred. The value is now passed back in both cases.
1219 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1220 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1221 use login_cap.h, so on its own it isn't the distinguishing feature. The
1222 new name refers directly to the setclassresources() function.
1224 PH/65 Added configuration files for NetBSD3.
1226 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1228 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1229 is preferred over IPv4.
1231 PH/68 The bounce_return_message and bounce_return_body options were not being
1232 honoured for bounces generated during the reception of non-SMTP messages.
1233 In particular, this applied to messages rejected by the ACL. This bug has
1234 been fixed. However, if bounce_return_message is true and bounce_return_
1235 body is false, the headers that are returned for a non-SMTP message
1236 include only those that have been read before the error was detected.
1237 (In the case of an ACL rejection, they have all been read.)
1239 PH/69 The HTML version of the specification is now built in a directory called
1240 spec_html instead of spec.html, because the latter looks like a path with
1241 a MIME-type, and this confuses some software.
1243 PH/70 Catch two compiler warnings in sieve.c.
1245 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1246 function verify_get_ident() calls ip_connect() to connect a socket, but
1247 if the "connect()" function timed out, ip_connect() used to close the
1248 socket. However, verify_get_ident() also closes the socket later, and in
1249 between Exim writes to the log, which may get opened at this point. When
1250 the socket was closed in ip_connect(), the log could get the same file
1251 descriptor number as the socket. This naturally causes chaos. The fix is
1252 not to close the socket in ip_connect(); the socket should be closed by
1253 the function that creates it. There was only one place in the code where
1254 this was missing, in the iplookup router, which I don't think anybody now
1255 uses, but I've fixed it anyway.
1257 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1258 well as to direct DNS lookups. Otherwise the handling of names in host
1259 lists is inconsistent and therefore confusing.
1265 PH/01 Two changes to the default runtime configuration:
1267 (1) Move the checks for relay_from_hosts and authenticated clients from
1268 after to before the (commented out) DNS black list checks.
1270 (2) Add control=submission to the relay_from_hosts and authenticated
1271 clients checks, on the grounds that messages accepted by these
1272 statements are most likely to be submissions.
1274 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1276 (1) Generate an error if the third argument for the ${prvs expansion is
1279 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1282 (3) Reset the variables that are obtained from the first argument of
1283 ${prvscheck and used in the second argument before leaving the code,
1284 because their memory is reclaimed, so using them afterwards may do
1287 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1288 one (it's much easier than Tom thought :-).
1290 (5) Because of (4), we can now allow for the use of $prvscheck_result
1291 inside the third argument.
1293 PH/03 For some reason, the default setting of PATH when running a command from
1294 a pipe transport was just "/usr/bin". I have changed it to
1297 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1298 anything to be listed in the output from -bV.
1300 PH/05 When a filter generated an autoreply, the entire To: header line was
1301 quoted in the delivery log line, like this:
1303 => >A.N.Other <ano@some.domain> <original@ddress> ...
1305 This has been changed so that it extracts the operative address. There
1306 may be more than one such address. If so, they are comma-separated, like
1309 => >ano@some.domain,ona@other.domain <original@ddress> ...
1311 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1312 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1313 not being looked up in the rDNS to get a host name, Exim was showing the
1314 IP address twice in Received: lines, even though the IP addresses were
1315 identical. For example:
1317 Received: from [1.2.3.4] (helo=[1.2.3.4])
1319 However, if the real host name was known, it was omitting the HELO data
1320 if it matched the actual IP address. This has been tidied up so that it
1321 doesn't show the same IP address twice.
1323 PH/07 When both +timestamp and +memory debugging was on, the value given by
1324 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1325 called by the debug printing, thereby overwriting the timestamp buffer.
1326 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1328 PH/08 When the original message was included in an autoreply transport, it
1329 always said "this is a copy of the message, including all the headers",
1330 even if body_only or headers_only was set. It now gives an appropriate
1333 PH/09 Applied a patch from the Sieve maintainer which:
1335 o fixes some comments
1336 o adds the (disabled) notify extension core
1337 o adds some debug output for the result of if/elsif tests
1338 o points to the current vacation draft in the documentation
1339 and documents the missing references header update
1343 o fixes a bug in processing the envelope test (when testing
1344 multiple envelope elements, the last element determinted the
1347 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1348 Electronic Mail") by including:
1350 Auto-submitted: auto-generated
1352 in the messages that it generates (bounce messages and others, such as
1353 warnings). In the case of bounce messages for non-SMTP mesages, there was
1354 also a typo: it was using "Auto_submitted" (underscore instead of
1355 hyphen). Since every message generated by Exim is necessarily in response
1356 to another message, thes have all been changed to:
1358 Auto-Submitted: auto-replied
1360 in accordance with these statements in the RFC:
1362 The auto-replied keyword:
1364 - SHOULD be used on messages sent in direct response to another
1365 message by an automatic process,
1367 - MUST NOT be used on manually-generated messages,
1369 - MAY be used on Delivery Status Notifications (DSNs) and Message
1370 Disposition Notifications (MDNs),
1372 - MUST NOT be used on messages generated by automatic or periodic
1373 processes, except for messages which are automatic responses to
1376 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1377 to the default Received: header definition.
1379 PH/12 Added log selector acl_warn_skipped (default on).
1381 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1382 variables because (a) they are in the wrong storage pool and (b) even if
1383 they were copied, it wouldn't work properly because of the caching.
1385 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1386 checking when decoding. Apparently there are clients that generate
1387 overlong encoded strings. Why am I not surprised?
1389 PH/15 If the first argument of "${if match_address" was not empty, but did not
1390 contain an "@" character, Exim crashed. Now it writes a panic log message
1391 and treats the condition as false.
1393 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1395 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1396 extension "envelope-auth". The code is finished and in agreement with
1397 other implementations, but there is no documentation so far and in fact,
1398 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1399 not changing the active code.
1401 Print executed "if" and "elsif" statements when debugging is used. This
1402 helps a great deal to understand what a filter does.
1404 Document more things not specified clearly in RFC3028. I had all this
1405 sorted out, when out of a sudden new issues came to my mind. Oops."
1407 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1410 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1411 canonical form (as documented). However, after a host name lookup from
1412 the IP address, check_host() was doing a simple string comparison with
1413 addresses acquired from the DNS when checking that the found name did
1414 have the original IP as one of its addresses. Since any found IPv6
1415 addresses are likely to be in abbreviated form, the comparison could
1416 fail. Luckily, there already exists a function for doing the comparison
1417 by converting both addresses to binary, so now that is used instead of
1418 the text comparison.
1420 PH/20 There was another similar case to PH/19, when a complete host name was
1421 given in a host list; looking up its IP address could give an abbreviated
1422 form, whereas the current host's name might or might not be abbreviated.
1423 The same fix has been applied.
1429 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1430 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1433 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1434 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1436 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1438 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1439 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1440 Also, for "Auto-Submitted", treat anything other than "no" as
1441 non-personal, in accordance with RFC 3834. (Previously it treated
1442 anything starting "auto-" as non-personal.)
1444 TF/01 The control=submission/name=... option had a problem with syntax
1445 errors if the name included a slash character. The /name= option
1446 now slurps the rest of the string, so it can include any characters
1447 but it must come last in the list of options (after /sender_retain
1450 PH/05 Some modifications to the interface to the fake nameserver for the new
1458 TK/01 Added the "success_on_redirect" address verification option. See
1459 NewStuff for rationale and an example.
1461 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1463 PH/02 Patch to exigrep to allow it to work on syslog lines.
1465 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1466 fread() to skip over the body file's header line, because in Cygwin the
1467 header line is locked and is inaccessible.
1469 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1470 co-exist for some time) to make it clear that it is the Exim ID that is
1471 referenced, not the Message-ID: header line.
1473 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1474 string_format() function, because snprintf() does not exist on all
1477 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1478 ${stat: expansion item.
1480 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1481 protocol synchronization error", to keep the pedants happy.
1483 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1484 well as for IRIX systems, when gcc is being used. See the host.c source
1487 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1489 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1492 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1493 transport and to the smtp transport in LMTP mode.
1495 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1497 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1498 run a filter in a subprocess. This could lead to confusion in subsequent
1499 lookups in the parent process. There should also be a search_tidyup() at
1500 the end of the subprocess.
1502 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1503 only if the host matched helo_try_verify_hosts, which caused the
1504 verification to occur when the EHLO/HELO command was issued. The ACL just
1505 tested the remembered result. Now, if a previous verification attempt has
1506 not happened, "verify = helo" does it there and then.
1508 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1510 TK/03 Fix log output including CR from clamd.
1512 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1513 exist provoked a memory error which could cause a segfault.
1515 PH/15 Installed PCRE 6.2
1517 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1519 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1520 of the problem. Specifically, suggested +O2 rather than +O1 for the
1523 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1525 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1526 record, the port was not correctly logged when the outgoing_port log
1527 selector was set (it logged the transort's default port).
1529 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1530 fallback_hosts, and "hosts" in the smtp transport.
1532 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1533 host errors such as "Connection refused".
1535 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1536 authentication with radiusclient 0.4.9:
1538 - Error returned from rc_read_config was caught wrongly
1539 - Username/password not passed on to radius server due to wrong length.
1541 The presumption is that some radiusclient API changes for 4.51/PH/17
1542 were not taken care of correctly. The code is still untested by me (my
1543 Linux distribution still has 0.3.2 of radiusclient), but it was
1544 contributed by a Radius user.
1546 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1547 expanding the "port" option of the smtp transport.
1549 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1550 while reading a MIME header. Thanks to Tom Hughes for a patch.
1552 PH/24 Include config.h inside local_scan.h so that configuration settings are
1555 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1556 that in an ACL for RCPT (for example), you can examine exactly what was
1559 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1560 commands, but it was not correctly comparing the address with the actual
1561 client host address. Thus, it would show the EHLO address in Received:
1562 header lines when this was not necessary.
1564 PH/27 Added the % operator to ${eval:}.
1566 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1567 it should be ignoring failures (because with -C, for example, it has lost
1568 privilege). It wasn't ignoring creation failures other than "already
1571 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1574 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1575 because an input line was too long, either on its own, or by virtue of
1576 too many continuations, the temporary file was not being removed, and the
1577 return code was incorrect.
1579 PH/31 Missing "BOOL" in function definition in filtertest.c.
1581 PH/32 Applied Sieve patches from the maintainer.
1583 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1585 PH/33 Added "verify = not_blind".
1587 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1588 Local/Makefile (with some defaults set). These are used in built scripts
1589 such as exicyclog, but they have never been used in the exim_install
1590 script (though there are many overriding facilities there). I have
1591 arranged that the exim_install script now takes note of these two
1594 PH/35 Installed configuration files for Dragonfly.
1596 PH/36 When a locally submitted message by a trusted user did not contain a
1597 From: header, and the sender address was obtained from -f or from an SMTP
1598 MAIL command, and the trusted user did not use -F to supply a sender
1599 name, $originator_name was incorrectly used when constructing a From:
1600 header. Furthermore, $originator_name was used for submission mode
1601 messages from external hosts without From: headers in a similar way,
1602 which is clearly wrong.
1604 PH/37 Added control=suppress_local_fixups.
1606 PH/38 When log_selector = +received_sender was set, and the addition of the
1607 sender made the log line's construction buffer exactly full, or one byte
1608 less than full, an overflow happened when the terminating "\n" was
1611 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1612 when the result of a list match is failure because a DNS lookup failed.
1614 PH/40 RM_COMMAND is now used in the building process.
1616 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1617 the "build-* directories that it finds.
1619 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1620 address in a domain literal was a prefix of an interface address.
1622 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1623 when verifying a sender address, unless rewrite_headers is false.
1625 PH/44 Wrote a long comment about why errors_to addresses are verified as
1626 recipients, not senders.
1628 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1629 the ratelimit ACL was added.
1631 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1633 PH/46 Added extra information about PostgreSQL errors to the error string.
1635 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1636 avoiding the need to install special zones in a real server. This is
1637 backwards compatible; if it can't find the fake resolver, it drops back.
1638 Thus, both old and new test suites can be run.
1640 TF/02 Added util/ratelimit.pl
1642 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1643 clock is set back in time.
1645 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1646 Candler <B.Candler@pobox.com>.
1648 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1649 OK for addresses that are the result of redirections.
1651 PH/48 A number of further additions for the benefit of the new test suite,
1652 including a fake gethostbyname() that interfaces to the fake DNS resolver
1655 TF/06 The fix for widen_domains has also been applied to qualify_single and
1656 search_parents which are the other dnslookup options that can cause
1659 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1662 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1664 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1665 service" if it encountered an MX record with an empty target hostname.
1666 The message is now "an MX or SRV record indicated no SMTP service".
1668 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1669 if the DNS of the sending site is misconfigured. This is quite a
1670 common situation. This change restores the behaviour of treating a
1671 helo verification defer as a failure.
1673 PH/51 If self=fail was set on a router, the bounce message did not include the
1674 actual error message.
1680 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1682 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1683 command itself ended in error, the underlying message about the transport
1684 filter timeout was being overwritten with the pipe command error. Now the
1685 underlying error message should be appended to the second error message.
1687 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1689 PH/02 Reduce the amount of output that "make" produces by default. Full output
1690 can still be requested.
1692 PH/03 The warning log line about a condition test deferring for a "warn" verb
1693 was being output only once per connection, rather than after each
1694 occurrence (because it was using the same function as for successful
1695 "warn" verbs). This seems wrong, so I have changed it.
1697 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1698 it should not have, which might have caused a crash in the right
1699 circumstances, but probably never did.
1701 PH/04 Installed a modified version of Tony Finch's patch to make submission
1702 mode fix the return path as well as the Sender: header line, and to
1703 add a /name= option so that you can make the user's friendly name appear
1706 TF/03 Added the control = fakedefer ACL modifier.
1708 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1709 Mark Lowes for thorough testing.
1711 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1713 TK/03 Merged latest SRS patch from Miles Wilton.
1715 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1716 with the definition in sysexits.h (which is #included earlier).
1717 Fortunately, Exim does not actually use EX_OK. The code used to try to
1718 preserve the sysexits.h value, by assumimg that macro definitions were
1719 scanned for macro replacements. I have been disabused of this notion,
1720 so now the code just undefines EX_OK before #including unistd.h.
1722 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1723 in the smtp transport. When a block could not be written in a single
1724 write() function, the timeout was being re-applied to each part-write.
1725 This seems wrong - if the receiver was accepting one byte at a time it
1726 would take for ever. The timeout is now adjusted when this happens. It
1727 doesn't have to be particularly precise.
1729 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1730 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1732 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1733 without a domain if the check to <postmaster@domain> fails.
1735 SC/01 Eximstats: added -xls and the ability to specify output files
1736 (patch written by Frank Heydlauf).
1738 SC/02 Eximstats: use FileHandles for outputing results.
1740 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1742 SC/04 Eximstats: fixed display of large numbers with -nvr option
1744 SC/05 Eximstats: fixed merging of reports with empty tables.
1746 SC/06 Eximstats: added the -include_original_destination flag
1748 SC/07 Eximstats: removed tabs and trailing whitespace.
1750 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1752 TK/06 MBOX spool code: Add real "From " MBOX separator line
1753 so the .eml file is really in mbox format (even though
1754 most programs do not really care). Patch from Alex Miller.
1756 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1757 The latter is generated from $received_to and is only set if the
1758 message has one envelope recipient. SA can use these headers,
1759 obviously out-of-the-box. Patch from Alex Miller.
1761 PH/08 The ${def test on a variable was returning false if the variable's
1762 value was "0", contrary to what the specification has always said!
1763 The result should be true unless the variable is empty.
1765 PH/09 The syntax error of a character other than { following "${if
1766 def:variable_name" (after optional whitespace) was not being diagnosed.
1767 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1768 accidental colon was present, for example, could give incorrect results.
1770 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1771 result is used (not including appendfile, where other changes are about
1774 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1775 This involved changing a lot of size variables from int to off_t. It
1776 should work with maildirs and everything.
1778 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1779 spamd dying while we are connected to it.
1781 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1784 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1785 function for BDB 4.3.
1787 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1789 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1790 the sources for PCRE have been re-arranged and more files are now
1793 PH/15 The code I had for printing potentially long long variables in PH/11
1794 above was not the best (it lost precision). The length of off_t variables
1795 is now inspected at build time, and an appropriate printing format (%ld
1796 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1797 to be "long long int" or "long int". This is needed for the internal
1798 formatting function string_vformat().
1800 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1801 the configuration file to be ":syslog", then the script "guesses" where
1802 the logs files are, rather than using the compiled in default. In our
1803 case the guess is not the same as the compiled default, so the script
1804 suddenly stopped working when I started to use syslog. The patch checks
1805 to see if log_file_path is "". If so, it attempts to read it from exim
1806 with no configuration file to get the compiled in version, before it
1807 falls back to the previous guessing code."
1809 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1810 implementing BATV in an Exim configuration. See NewStuff for the gory
1813 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1814 Makefile that are specific to HP-UX.
1816 PH/18 If the "use_postmaster" option was set for a recipient callout together
1817 with the "random" option, the postmaster address was used as the MAIL
1818 FROM address for the random test, but not for the subsequent recipient
1819 test. It is now used for both.
1821 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1822 patch removes a few documentation additions to RFC 3028, because the
1823 latest draft now contains them. It adds the new en;ascii-case comparator
1824 and a new error check for 8bit text in MIME parts. Comparator and
1825 require names are now matched exactly. I enabled the subaddress
1826 extension, but it is not well tested yet (read: it works for me)."
1828 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1829 rework some of the code of TK/09 above to avoid the hardwired use of
1830 "%lld" and "long long". Replaced the call to snprintf() with a call to
1833 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1834 records point to non-existent hosts", "retry timeout exceeded", and
1835 "retry time not reached for any host after a long failure period".
1837 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1838 experimental DomainKeys support:
1840 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1841 (2) On an error such as an illegally used "control", the wrong name for
1842 the control was given.
1844 These problems did NOT occur unless DomainKeys support was compiled.
1846 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1848 PH/24 Added ${if match_ip condition.
1850 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1851 databases so that it will be absolutely obvious if a crash occurs in the
1852 DB library. This is a regular occurrence (often caused by mis-matched
1855 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1856 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1857 picked up on a user's system that detects such things. There doesn't seem
1858 to be a gcc warning option for this - only an attribute that has to be
1859 put on the function's prototype. It seems that in Fedora Core 4 they have
1860 set this on a number of new functions. No doubt there will be more in due
1863 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1864 specify a transport. However, if an address that was verified by such a
1865 router was the subject of a callout, Exim crashed because it tried to
1866 read the rcpt_include_affixes from the non-existent transport. Now it
1867 just assumes that the setting of that option is false. This bug was
1868 introduced by 4.51/PH/31.
1870 PH/28 Changed -d+all to exclude +memory, because that information is very
1871 rarely of interest, but it makes the output a lot bigger. People tend to
1872 do -d+all out of habit.
1874 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1875 code in os-type was giving problems when libc.so lives in lib64, like on
1878 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1879 aren't the modern standard, and it seems that some systems' include files
1880 don't always have them. Exim was already checking for some of the newer
1881 ones like T_AAAA, and defining it itself. I've added checks for all the
1882 record types that Exim uses.
1884 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1885 not automatically generating a new one, as it is supposed to. This
1886 prevented TLS from working. If the file did exist, but contained invalid
1887 data, a new version was generated, as expected. It was only the case of a
1888 non-existent file that was broken.
1890 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1891 with a change in libdomainkeys > 0.64.
1893 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1894 from DNS. If the selector record carries the flag, it now has
1895 precedence over the domain-wide flag.
1897 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1899 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1900 the use of a port name that isn't defined in /etc/services) occurred, the
1901 message was deferred as in a normal delivery, and thus remained on the
1902 spool, instead of being failed because of the mua_wrapper setting. This
1903 is now fixed, and I tidied up some of the mua_wrapper messages at the
1906 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1907 the messages in a hash of arrays rather than using individual hashes.
1908 This is a bit cleaner and results in dramatic memory savings, albeit
1909 at a slight CPU cost.
1911 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1912 as requested by Marc Sherman.
1914 SC/10 Eximstats: added histograms for user specified patterns as requested
1917 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1919 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1920 fopen() in the content-scanning modules that did not already have it.
1926 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1927 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1929 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1931 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1933 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1934 to test Sieve filters that use "vacation".
1936 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1937 that changes the way the GnuTLS parameters are stored in the cache file.
1938 The new format can be generated externally. For backward compatibility,
1939 if the data in the cache doesn't make sense, Exim assumes it has read an
1940 old-format file, and it generates new data and writes a new file. This
1941 means that you can't go back to an older release without removing the
1944 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1945 work if there are any delivery delays because "one_time" forces the
1946 parent to be marked "delivered", so its unseen clone is never tried
1947 again. For this reason, Exim now forbids the simultaneous setting of
1950 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1951 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1952 entry if you want to know the details. The fix, however, neglected to
1953 consider the case when local delivery batching is involved. The test for
1954 "previously delivered" was not happening when checking to see if an
1955 address could be batched with a previous (undelivered) one; under
1956 certain circumstances this could lead to multiple deliveries to the same
1959 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1960 in its include files, and this causes problems building Exim.
1962 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1963 header_syntax) but Exim was just ignoring anything given after a slash.
1964 In particular, this caused confusion with an attempt to use "verify =
1965 reverse_host_lookup/defer_ok". An error is now given when options are
1966 supplied for verify items that do not have them. (Maybe reverse_host_
1967 lookup should have a defer_ok option, but that's a different point.)
1969 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1970 defined by RFC 821) to 2048, because there were problems with some AUTH
1971 commands, and RFC 1869 says the size should be increased for extended
1972 SMTP commands that take arguments.
1974 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1977 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1978 "unknown" error; now it says that the functionality isn't in the binary.
1980 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1981 an address' error message when a string expansion fails (syntax or
1982 whatever). Otherwise the password may appear in the log. Following change
1983 PH/42 below, there is no longer a chance of it appearing in a bounce
1986 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1988 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1989 of its addresses was ever tried. (Bugzilla bug #2.)
1991 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1992 the result incorrectly in the debug output. (It correctly added a newline
1993 to what was transported.)
1995 TF/01 Added $received_time.
1997 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1998 commented out examples of how to interface to a virus scanner and to
1999 SpamAssassin. Also added commented examples of av_scanner and
2000 spamd_address settings.
2002 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
2003 and controls are allowed in which ACLs. There were a couple of minor
2004 errors. Some of the entries in the conditions table (which is a table of
2005 where they are NOT allowed) were getting very unwieldy; rewrote them as a
2006 negation of where the condition IS allowed.
2008 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
2010 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
2011 header file does not have a version number, so I've had to invent a new
2012 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2013 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2014 radiusclient), but it was contributed by a Radius user.
2016 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2017 files or directories, for OpenSSL.
2019 PH/19 When an Exim process that is running as root has to create an Exim log
2020 file, it does so in a subprocess that runs as exim:exim so as to get the
2021 ownership right at creation (otherwise, other Exim processes might see
2022 the file with the wrong ownership). There was no test for failure of this
2023 fork() call, which would lead to the process getting stuck as it waited
2024 for a non-existent subprocess. Forks do occasionally fail when resources
2025 run out. I reviewed all the other calls to fork(); they all seem to check
2028 PH/20 When checking for unexpected SMTP input at connect time (before writing
2029 the banner), Exim was not dealing correctly with a non-positive return
2030 from the read() function. If the client had disconnected by this time,
2031 the result was a log entry for a synchronization error with an empty
2032 string after "input=" when read() returned zero. If read() returned -1
2033 (an event I could not check), uninitialized data bytes were printed.
2034 There were reports of junk text (parts of files, etc) appearing after
2037 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2039 PH/22 Added support for macro redefinition, and (re)definition in between
2040 driver and ACL definitions.
2042 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2043 forgetting to use the resulting value; it was using the unexpanded value.
2045 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2046 hadn't been configured. The fix is from Juergen Kreileder, who
2047 understands it better than I do:
2049 "Here's what I see happening with three configured cyrus_sasl
2050 authenticators configured (plain, login, cram-md5):
2052 On startup auth_cyrus_sasl_init() gets called for each of these.
2053 This means three calls to sasl_listmech() without a specified mech_list.
2054 => SASL tests which mechs of all available mechs actually work
2055 => three warnings about OTP not working
2056 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2058 With the patch, sasl_listmech() also gets called three times. But now
2059 SASL's mech_list option is set to the server_mech specified in the the
2060 authenticator. Or in other words, the answer from sasl_listmech()
2061 gets limited to just the mech you're testing for (which is different
2063 => the return list contains just 'plain' or 'login', 'cram-md5' or
2064 nothing depending on the value of ob->server_mech.
2066 I've just tested the patch: Authentication still works fine,
2067 unavailable mechs specified in the exim configuration are still
2068 caught, and the auth.log warnings about OTP are gone."
2070 PH/25 When debugging is enabled, the contents of the command line are added
2071 to the debugging output, even when log_selector=+arguments is not
2074 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2075 answer is "GNU", and only if the return is "GNU/something" is the answer
2078 PH/27 $acl_verify_message is now set immediately after the failure of a
2079 verification in an ACL, and so is available in subsequent modifiers. In
2080 particular, the message can be preserved by coding like this:
2082 warn !verify = sender
2083 set acl_m0 = $acl_verify_message
2085 Previously, $acl_verify_message was set only while expanding "message"
2086 and "log_message" when a very denied access.
2088 PH/28 Modified OS/os.c-Linux with
2090 -#ifndef OS_LOAD_AVERAGE
2091 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2093 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2094 nomenclature these days.)
2096 PH/29 Installed patch from the Sieve maintainer that adds the options
2097 sieve_useraddress and sieve_subaddress to the redirect router.
2099 PH/30 In these circumstances:
2100 . Two addresses routed to the same list of hosts;
2101 . First host does not offer TLS;
2102 . First host accepts first address;
2103 . First host gives temporary error to second address;
2104 . Second host offers TLS and a TLS session is established;
2105 . Second host accepts second address.
2106 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2107 and peerdn, if requested) that were in fact used only for the second
2110 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2111 attention to any local part prefix or suffix that was matched by the
2112 router that accepted the address. It now behaves in the same way as it
2113 does for delivery: the affixes are removed from the local part unless
2114 rcpt_include_affixes is set on the transport.
2116 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2117 timeout during the DATA phase of an incoming message.
2119 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2120 applied a patch sanctioned by the Sieve maintainer.
2122 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2123 the uid or gid is negative. A case of a negative gid caused this to be
2124 noticed. The fix allows for either to be negative.
2126 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2127 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2130 PH/36 The change PH/12 above was broken. Fixed it.
2132 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2133 the grounds that routing the same address twice would always produce the
2134 same answer. This might have been true once, but it is certainly no
2135 longer true now. Routing a child address may depend on the previous
2136 routing that produced that child. Some complicated redirection strategies
2137 went wrong when messages had multiple recipients, and made Exim's
2138 behaviour dependent on the order in which the addresses were given.
2140 I have moved the duplicate checking until after the routing is complete.
2141 Exim scans the addresses that are assigned to local and remote
2142 transports, and removes any duplicates. This means that more work will be
2143 done, as duplicates will always all be routed, but duplicates are
2144 presumably rare, so I don't expect this is of any significance.
2146 For deliveries to pipes, files, and autoreplies, the duplicate checking
2147 still happens during the routing process, since they are not going to be
2150 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2151 It corrects a timeout issue with spamd. This is Ian's comment: "The
2152 background is that sometimes spamd either never reads data from a
2153 connection it has accepted, or it never writes response data. The exiscan
2154 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2155 blindly assumes that writes won't block so it may never time out."
2157 PH/39 Allow G after quota size as well as K and M.
2159 PH/40 The value set for $authenticated_id in an authenticator may not contain
2160 binary zeroes or newlines because the value is written to log lines and
2161 to spool files. There was no check on this. Now the value is run through
2162 the string_printing() function so that such characters are converted to
2163 printable escape sequences.
2165 PH/41 $message_linecount is a new variable that contains the total number of
2166 lines in the message. Compare $body_linecount, which is the count for the
2169 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2170 bounce and delay warning messages, except in certain special cases, which
2173 (a) An SMTP error message from a remote host;
2174 (b) A message specified in a :fail: redirection;
2175 (c) A message specified in a "fail" command in a system filter;
2176 (d) A message specified in a FAIL return from the queryprogram router;
2177 (e) A message specified by the cannot_route_message router option.
2179 In these cases only, Exim does include the error details in bounce and
2180 warning messages. There are also a few cases where bland messages such
2181 as "unrouteable address" or "local delivery error" are given.
2183 PH/43 $value is now also set for the "else" part of a ${run expansion.
2185 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2186 being worked on, but at least Exim now implements the latest version to
2189 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2190 process to complete was treated as a delivery failure, a timeout while
2191 writing the message to the pipe was logged, but erroneously treated as a
2192 successful delivery. Such timeouts include transport filter timeouts. For
2193 consistency with the overall process timeout, these timeouts are now
2194 treated as errors, giving rise to delivery failures by default. However,
2195 there is now a new Boolean option for the pipe transport called
2196 timeout_defer, which, if set TRUE, converts the failures into defers for
2197 both kinds of timeout. A transport filter timeout is now identified in
2200 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2201 systems where "make" and "gmake" are different, calling "gmake" at top
2202 level broke things. I've arranged for the value of $(MAKE) to be passed
2203 from the Makefile to this script so that it can call the same version of
2207 A note about Exim versions 4.44 and 4.50
2208 ----------------------------------------
2210 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2211 changes of various kinds. As a consequence, a big documentation update was
2212 needed. This delayed the release for rather longer than seemed good, especially
2213 in the light of a couple of (minor) security issues. Therefore, the changes
2214 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2215 release. So 4.44 and 4.50 are in effect two different branches that both start
2218 I have left the 4.50 change log unchanged; it contains all the changes since
2219 4.43. The change log for 4.44 is below; many of its items are identical to
2220 those for 4.50. This seems to be the most sensible way to preserve the
2221 historical information.
2227 1. Minor wording change to the doc/README.SIEVE file.
2229 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2230 computation of the current number of files was incorrect.
2232 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2233 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2236 4. Give more explanation in the error message when the command for a transport
2237 filter fails to execute.
2239 5. There are several places where Exim runs a non-Exim command in a
2240 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2241 was being done only for the command run by the queryprogram router. It is
2242 now done for all such subprocesses. The other cases are: ${run, transport
2243 filters, and the commands run by the lmtp and pipe transports.
2245 6. Added CONFIGURE_GROUP build-time option.
2247 7. Some older OS have a limit of 256 on the maximum number of file
2248 descriptors. Exim was using setrlimit() to set 1000 as a large value
2249 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2250 systems. I've change it so that if it can't get 1000, it tries for 256.
2252 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2253 was an oversight, and furthermore, ever since the addition of extra
2254 controls (e.g. 4.43/32), the checks on when to allow different forms of
2255 "control" were broken. There should now be diagnostics for all cases when a
2256 control that does not make sense is encountered.
2258 9. Added the /retain_sender option to "control=submission".
2260 10. $recipients is now available in the predata ACL (oversight).
2262 11. Tidy the search cache before the fork to do a delivery from a message
2263 received from the command line. Otherwise the child will trigger a lookup
2264 failure and thereby defer the delivery if it tries to use (for example) a
2265 cached ldap connection that the parent has called unbind on.
2267 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2268 of $address_data from the recipient verification was clobbered by the
2269 sender verification.
2271 13. The value of address_data from a sender verification is now available in
2272 $sender_address_data in subsequent conditions in the ACL statement.
2274 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2276 15. Added a new option "connect=<time>" to callout options, to set a different
2279 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2280 was its contents. (It was OK if the option was not defined at all.)
2282 17. A "Completed" log line is now written for messages that are removed from
2283 the spool by the -Mrm option.
2285 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2286 information about exactly what failed.
2288 19. Added -dd to debug only the daemon process.
2290 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2291 handles timeouts, both on the server side and network timeouts. Renamed the
2292 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2294 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2295 It is now set to "smtps".
2297 22. $host_address is now set to the target address during the checking of
2298 ignore_target_hosts.
2300 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2301 being passed; this would have caused $sender_host_name to have been used if
2302 matching the list had actually called for a host name (not very likely,
2303 since this list is usually IP addresses). A host name is now passed as
2306 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2307 code that specifies a non-restarting handler (typically sigaction() in
2308 modern systems) in an attempt to fix a rare and obscure crash bug.
2310 25. Narrowed the window for a race in the daemon that could cause it to ignore
2311 SIGCHLD signals. This is not a major problem, because they are used only to
2312 wake it up if nothing else does.
2314 26. A malformed maildirsize file could cause Exim to calculate negative values
2315 for the mailbox size or file count. Odd effects could occur as a result.
2316 The maildirsize information is now recalculated if the size or filecount
2319 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2320 support for a long time. Removed HAVE_SYS_VFS_H.
2322 28. Installed the latest version of exipick from John Jetmore.
2324 29. In an address list, if the pattern was not a regular expression, an empty
2325 subject address (from a bounce message) matched only if the pattern was an
2326 empty string. Non-empty patterns were not even tested. This was the wrong
2327 because it is perfectly reasonable to use an empty address as part of a
2328 database query. An empty address is now tested by patterns that are
2329 lookups. However, all the other forms of pattern expect the subject to
2330 contain a local part and a domain, and therefore, for them, an empty
2331 address still always fails if the pattern is not itself empty.
2333 30. Exim went into a mad DNS loop when attempting to do a callout where the
2334 host was specified on an smtp transport, and looking it up yielded more
2335 than one IP address.
2337 31. Re-factored the code for checking spool and log partition space into a
2338 function that finds that data and another that does the check. The former
2339 is then used to implement four new variables: $spool_space, $log_space,
2340 $spool_inodes, and $log_inodes.
2342 32. The RFC2047 encoding function was originally intended for short strings
2343 such as real names; it was not keeping to the 75-character limit for
2344 encoded words that the RFC imposes. It now respects the limit, and
2345 generates multiple encoded words if necessary. To be on the safe side, I
2346 have increased the buffer size for the ${rfc2047: expansion operator from
2349 33. It is now permitted to omit both strings after an "if" condition; if the
2350 condition is true, the result is "true". As before, when the second string
2351 is omitted, a false condition yields an empty string. This makes it less
2352 cumbersome to write custom ACL and router conditions.
2354 34. Failure to deliver a bounce message always caused it to be frozen, even if
2355 there was an errors_to setting on the router. The errors_to setting is now
2358 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2359 canonical form (fully expanded) before being placed in
2360 $sender_host_address.
2362 36. The table in the code that translates DNS record types into text (T_A to
2363 "A" for instance) was missing entries for NS and CNAME. It is just possible
2364 that this could have caused confusion if both these types were looked up
2365 for the same domain, because the text type is used as part of Exim's
2366 per-process caching. But the chance of anyone hitting this buglet seems
2369 37. The dnsdb lookup has been extended in a number of ways.
2371 (1) There is a new type, "zns", which walks up the domain tree until it
2372 finds some nameserver records. It should be used with care.
2374 (2) There is a new type, "mxh", which is like "mx" except that it returns
2375 just the host names, not the priorities.
2377 (3) It is now possible to give a list of domains (or IP addresses) to be
2378 looked up. The behaviour when one of the lookups defers can be
2379 controlled by a keyword.
2381 (4) It is now possible to specify the separator character for use when
2382 multiple records are returned.
2384 38. The dnslists ACL condition has been extended: it is now possible to supply
2385 a list of IP addresses and/or domains to be looked up in a particular DNS
2388 39. Added log_selector=+queue_time_overall.
2390 40. When running the queue in the test harness, wait just a tad after forking a
2391 delivery process, to get repeatability of debugging output.
2393 41. Include certificate and key file names in error message when GnuTLS fails
2394 to set them up, because the GnuTLS error message doesn't include the name
2395 of the failing file when there is a problem reading it.
2397 42. Allow both -bf and -bF in the same test run.
2399 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2401 44. The "Exiscan patch" is now merged into the mainline Exim source.
2403 45. Sometimes the final signoff response after QUIT could fail to get
2404 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2405 < 0 before doing a fflush(). This bug looks as though it goes back to the
2406 introduction of TLS in release 3.20, but "sometimes" must have been rare
2407 because the tests only now provoked it.
2409 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2410 (this can affect the format of dates).
2412 47. exim_tidydb, when checking for the continued existence of a message for
2413 which it has found a message-specific retry record, was not finding
2414 messages that were in split spool directories. Consequently, it was
2415 deleting retry records that should have stayed in existence.
2417 48. Steve fixed some bugs in eximstats.
2419 49. The SPA authentication driver was not abandoning authentication and moving
2420 on to the next authenticator when an expansion was forced to fail,
2421 contradicting the general specification for all authenticators. Instead it
2422 was generating a temporary error. It now behaves as specified.
2424 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2425 (the order specifies the preference for clients). The order is now AES256,
2426 AES128, 3DES, ARCFOUR128.
2428 51. Small patch to Sieve code - explicitly set From: when generating an
2431 52. Exim crashed if a remote delivery caused a very long error message to be
2432 recorded - for instance if somebody sent an entire SpamAssassin report back
2433 as a large number of 550 error lines. This bug was coincidentally fixed by
2434 increasing the size of one of Exim's internal buffers (big_buffer) that
2435 happened as part of the Exiscan merge. However, to be on the safe side, I
2436 have made the code more robust (and fixed the comments that describe what
2439 53. Now that there can be additional text after "Completed" in log lines (if
2440 the queue_time_overall log selector is set), a one-byte patch to exigrep
2441 was needed to allow it to recognize "Completed" as not the last thing in
2444 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2445 patch that reportedly fixes this has been added. I am not expert enough to
2446 create a test for it. This is what the patch creator wrote:
2448 "I found a little strange behaviour of ldap code when working with
2449 Windows 2003 AD Domain, where users was placed in more than one
2450 Organization Units. When I tried to give exim partial DN, the exit code
2451 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2452 But simultaneously result of request was absolutely normal ldap result,
2453 so I produce this patch..."
2455 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2456 REFERENCE, so I have modified the code to exclude the patch when that macro
2459 55. Some experimental protocols are using DNS PTR records for new purposes. The
2460 keys for these records are domain names, not reversed IP addresses. The
2461 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2462 leaves it alone. Component reversal etc. now happens only for IP addresses.
2465 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2467 57. Double the size of the debug message buffer (to 2048) so that more of very
2468 long debug lines gets shown.
2470 58. The exicyclog utility now does better if the number of log files to keep
2471 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2473 59. Two changes related to the smtp_active_hostname option:
2475 (1) $smtp_active_hostname is now available as a variable.
2476 (2) The default for smtp_banner uses $smtp_active_hostname instead
2477 of $primary_hostname.
2479 60. The host_aton() function is supposed to be passed a string that is known
2480 to be a valid IP address. However, in the case of IPv6 addresses, it was
2481 not checking this. This is a hostage to fortune. Exim now panics and dies
2482 if the condition is not met. A case was found where this could be provoked
2483 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2484 components; fortuitously, this particular loophole had already been fixed
2485 by change 4.50/55 above.
2487 If there are any other similar loopholes, the new check in host_aton()
2488 itself should stop them being exploited. The report I received stated that
2489 data on the command line could provoke the exploit when Exim was running as
2490 exim, but did not say which command line option was involved. All I could
2491 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2492 running as the user.
2495 61. There was a buffer overflow vulnerability in the SPA authentication code
2496 (which came originally from the Samba project). I have added a test to the
2497 spa_base64_to_bits() function which I hope fixes it.
2500 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2501 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2503 63. The daemon start-up calls getloadavg() while still root for those OS that
2504 need the first call to be done as root, but it missed one case: when
2505 deliver_queue_load_max is set with deliver_drop_privilege. This is
2506 necessary for the benefit of the queue runner, because there is no re-exec
2507 when deliver_drop_privilege is set.
2509 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2510 This has been fixed.
2512 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2513 was in use, was not putting the data itself into the right store pool;
2514 consequently, it could be overwritten for a subsequent message in the same
2515 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2518 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2520 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2521 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2522 the function were treating the return as a boolean value, which happened to
2523 work because 0=false and not-0=true, but is not correct code.
2525 68. The host_aton() function was not handling scoped IPv6 addresses (those
2526 with, for example, "%eth0" on the end) correctly.
2528 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2529 negated items (that is, ~something) in unsigned ints. Some compilers
2530 apparently mutter when there is no cast.
2532 70. If an address verification called from an ACL failed, and did not produce a
2533 user-specific message (i.e. there was only a "system" message), nothing was
2534 put in $acl_verify_message. In this situation, it now puts the system
2537 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2538 session; change 4.31/43 added the unwanted input to the log line - except
2539 that it did not do this in the start of session case. It now does.
2541 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2542 This is wrong and can cause the other end to generate a synchronization
2543 error if it is another Exim or anything else that does the synchronization
2544 check. A QUIT command is no longer sent after a timeout.
2546 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2547 during host lookups.
2549 74. The defer_ok option of callout verification was not working if it was used
2550 when verifying addresses in header lines, that is, for this case:
2552 verify = header_sender/callout=defer_ok
2554 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2555 those file descriptors could be used for SMTP connections. If anything
2556 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2557 could be sent to the SMTP client, causing chaos. The daemon now opens
2558 stdin, stdout, and stderr to /dev/null when it puts itself into the
2561 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2562 log by default. The user can override this with suitable Perl magic.
2564 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2565 the log message when discard triggers, was not working for the DATA ACL or
2566 for the non-SMTP ACL.
2568 78. Error message wording change in sieve.c.
2570 79. If smtp_accept_max_per_host was set, the number of connections could be
2571 restricted to fewer than expected, because the daemon was trying to set up
2572 a new connection before checking whether the processes handling previous
2573 connections had finished. The check for completed processes is now done
2574 earlier. On busy systems, this bug wouldn't be noticed because something
2575 else would have woken the daemon, and it would have reaped the completed
2578 80. If a message was submitted locally by a user whose login name contained one
2579 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2580 It caused a spool format error. I have fixed the spool reading code. A
2581 related problem was that the "from" clause in the Received: line became
2582 illegal because of the space(s). It is now covered by ${quote_local_part.
2584 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2587 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2589 83. Patch from Sieve maintainer to fix unterminated string problem in
2590 "vacation" handling.
2592 84. Some minor changes to the Linux configuration files to help with other
2593 OS variants using glibc.
2595 85. One more patch for Sieve to update vacation handling to latest spec.
2598 ----------------------------------------------------
2599 See the note above about the 4.44 and 4.50 releases.
2600 ----------------------------------------------------
2606 1. Change 4.43/35 introduced a bug that caused file counts to be
2607 incorrectly computed when quota_filecount was set in an appendfile
2610 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2611 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2614 3. Give more explanation in the error message when the command for a transport
2615 filter fails to execute.
2617 4. There are several places where Exim runs a non-Exim command in a
2618 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2619 was being done only for the command run by the queryprogram router. It is
2620 now done for all such subprocesses. The other cases are: ${run, transport
2621 filters, and the commands run by the lmtp and pipe transports.
2623 5. Some older OS have a limit of 256 on the maximum number of file
2624 descriptors. Exim was using setrlimit() to set 1000 as a large value
2625 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2626 systems. I've change it so that if it can't get 1000, it tries for 256.
2628 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2629 was an oversight, and furthermore, ever since the addition of extra
2630 controls (e.g. 4.43/32), the checks on when to allow different forms of
2631 "control" were broken. There should now be diagnostics for all cases when a
2632 control that does not make sense is encountered.
2634 7. $recipients is now available in the predata ACL (oversight).
2636 8. Tidy the search cache before the fork to do a delivery from a message
2637 received from the command line. Otherwise the child will trigger a lookup
2638 failure and thereby defer the delivery if it tries to use (for example) a
2639 cached ldap connection that the parent has called unbind on.
2641 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2642 of $address_data from the recipient verification was clobbered by the
2643 sender verification.
2645 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2646 was its contents. (It was OK if the option was not defined at all.)
2648 11. A "Completed" log line is now written for messages that are removed from
2649 the spool by the -Mrm option.
2651 12. $host_address is now set to the target address during the checking of
2652 ignore_target_hosts.
2654 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2655 being passed; this would have caused $sender_host_name to have been used if
2656 matching the list had actually called for a host name (not very likely,
2657 since this list is usually IP addresses). A host name is now passed as
2660 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2661 code that specifies a non-restarting handler (typically sigaction() in
2662 modern systems) in an attempt to fix a rare and obscure crash bug.
2664 15. Narrowed the window for a race in the daemon that could cause it to ignore
2665 SIGCHLD signals. This is not a major problem, because they are used only to
2666 wake it up if nothing else does.
2668 16. A malformed maildirsize file could cause Exim to calculate negative values
2669 for the mailbox size or file count. Odd effects could occur as a result.
2670 The maildirsize information is now recalculated if the size or filecount
2673 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2674 support for a long time. Removed HAVE_SYS_VFS_H.
2676 18. Updated exipick to current release from John Jetmore.
2678 19. Allow an empty sender to be matched against a lookup in an address list.
2679 Previously the only cases considered were a regular expression, or an
2682 20. Exim went into a mad DNS lookup loop when doing a callout where the
2683 host was specified on the transport, if the DNS lookup yielded more than
2686 21. The RFC2047 encoding function was originally intended for short strings
2687 such as real names; it was not keeping to the 75-character limit for
2688 encoded words that the RFC imposes. It now respects the limit, and
2689 generates multiple encoded words if necessary. To be on the safe side, I
2690 have increased the buffer size for the ${rfc2047: expansion operator from
2693 22. Failure to deliver a bounce message always caused it to be frozen, even if
2694 there was an errors_to setting on the router. The errors_to setting is now
2697 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2698 canonical form (fully expanded) before being placed in
2699 $sender_host_address.
2701 24. Updated eximstats to version 1.33
2703 25. Include certificate and key file names in error message when GnuTLS fails
2704 to set them up, because the GnuTLS error message doesn't include the name
2705 of the failing file when there is a problem reading it.
2707 26. Expand error message when OpenSSL has problems setting up cert/key files.
2710 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2711 (this can affect the format of dates).
2713 28. exim_tidydb, when checking for the continued existence of a message for
2714 which it has found a message-specific retry record, was not finding
2715 messages that were in split spool directories. Consequently, it was
2716 deleting retry records that should have stayed in existence.
2718 29. eximstats updated to version 1.35
2719 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2720 1.35 - bugfix such that pie charts by volume are generated correctly
2722 30. The SPA authentication driver was not abandoning authentication and moving
2723 on to the next authenticator when an expansion was forced to fail,
2724 contradicting the general specification for all authenticators. Instead it
2725 was generating a temporary error. It now behaves as specified.
2727 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2728 (the order specifies the preference for clients). The order is now AES256,
2729 AES128, 3DES, ARCFOUR128.
2731 31. Small patch to Sieve code - explicitly set From: when generating an
2734 32. Exim crashed if a remote delivery caused a very long error message to be
2735 recorded - for instance if somebody sent an entire SpamAssassin report back
2736 as a large number of 550 error lines. This bug was coincidentally fixed by
2737 increasing the size of one of Exim's internal buffers (big_buffer) that
2738 happened as part of the Exiscan merge. However, to be on the safe side, I
2739 have made the code more robust (and fixed the comments that describe what
2742 33. Some experimental protocols are using DNS PTR records for new purposes. The
2743 keys for these records are domain names, not reversed IP addresses. The
2744 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2745 leaves it alone. Component reversal etc. now happens only for IP addresses.
2748 34. The host_aton() function is supposed to be passed a string that is known
2749 to be a valid IP address. However, in the case of IPv6 addresses, it was
2750 not checking this. This is a hostage to fortune. Exim now panics and dies
2751 if the condition is not met. A case was found where this could be provoked
2752 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2753 components; fortuitously, this particular loophole had already been fixed
2754 by change 4.50/55 or 4.44/33 above.
2756 If there are any other similar loopholes, the new check in host_aton()
2757 itself should stop them being exploited. The report I received stated that
2758 data on the command line could provoke the exploit when Exim was running as
2759 exim, but did not say which command line option was involved. All I could
2760 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2761 running as the user.
2764 35. There was a buffer overflow vulnerability in the SPA authentication code
2765 (which came originally from the Samba project). I have added a test to the
2766 spa_base64_to_bits() function which I hope fixes it.
2769 36. The daemon start-up calls getloadavg() while still root for those OS that
2770 need the first call to be done as root, but it missed one case: when
2771 deliver_queue_load_max is set with deliver_drop_privilege. This is
2772 necessary for the benefit of the queue runner, because there is no re-exec
2773 when deliver_drop_privilege is set.
2775 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2776 was in use, was not putting the data itself into the right store pool;
2777 consequently, it could be overwritten for a subsequent message in the same
2778 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2781 38. Sometimes the final signoff response after QUIT could fail to get
2782 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2783 < 0 before doing a fflush(). This bug looks as though it goes back to the
2784 introduction of TLS in release 3.20, but "sometimes" must have been rare
2785 because the tests only now provoked it.
2791 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2792 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2793 Now it returns an integer. A number of calls were still expecting a T/F
2794 return. Fortuitously, in all cases, the tests worked in OK situations,
2795 which is the norm. However, things would have gone wrong on any write
2796 failures on the smtp file descriptor. This function is used when sending
2797 messages over SMTP and also when doing verify callouts.
2799 2. When Exim is called to do synchronous delivery of a locally submitted
2800 message (the -odf or -odi options), it no longer closes stderr before doing
2803 3. Implemented the mua_wrapper option.
2805 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2807 5. Implemented the functions header_remove(), header_testname(),
2808 header_add_at_position(), and receive_remove_recipient(), and exported them
2811 6. If an ACL "warn" statement specified the addition of headers, Exim already
2812 inserted X-ACL-Warn: at the start if there was no header name. However, it
2813 was not making this test for the second and subsequent header lines if
2814 there were newlines in the string. This meant that an invalid header could
2815 be inserted if Exim was badly configured.
2817 7. Allow an ACL "warn" statement to add header lines at the start or after all
2818 the Received: headers, as well as at the end.
2820 8. Added the rcpt_4xx retry error code.
2822 9. Added postmaster_mailfrom=xxx to callout verification option.
2824 10. Added mailfrom=xxxx to the callout verification option, for verify=
2827 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2828 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2830 12. Inserted some casts to stop certain compilers warning when using pointer
2831 differences as field lengths or precisions in printf-type calls (mostly
2832 affecting debugging statements).
2834 13. Added optional readline() support for -be (dynamically loaded).
2836 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2837 same clock tick as a message's arrival, so that its received time was the
2838 same as the "first fail" time on the retry record, and that message
2839 remained on the queue past the ultimate address timeout, every queue runner
2840 would try a delivery (because it was past the ultimate address timeout) but
2841 after another failure, the ultimate address timeout, which should have then
2842 bounced the address, did not kick in. This was a "< instead of <=" error;
2843 in most cases the first failure would have been in the next clock tick
2844 after the received time, and all would be well.
2846 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2847 being recognized when the domain list was tested by the match_domain
2848 condition in an expansion string.
2850 16. Added the ${str2b64: operator.
2852 17. Exim was always calling setrlimit() to set a large limit for the number of
2853 processes, without checking whether the existing limit was already
2854 adequate. (It did check for the limit on file descriptors.) Furthermore,
2855 errors from getrlimit() and setrlimit() were being ignored. Now they are
2856 logged to the main and panic logs, but Exim does carry on, to try to do its
2857 job under whatever limits there are.
2859 18. Imported PCRE 5.0.
2861 19. Trivial typo in log message " temporarily refused connection" (the leading
2864 20. If the log selector return_path_on_delivery was set and an address was
2865 redirected to /dev/null, the delivery process crashed because it assumed
2866 that a return path would always be set for a "successful" delivery. In this
2867 case, the whole delivery is bypassed as an optimization, and therefore no
2870 21. Internal re-arrangement: the function for sending a challenge and reading
2871 a response while authentication was assuming a zero-terminated challenge
2872 string. It's now changed to take a pointer and a length, to allow for
2873 binary data in such strings.
2875 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2877 23. Exim was not respecting finduser_retries when seeking the login of the
2878 uid under which it was called; it was always trying 10 times. (The default
2879 setting of finduser_retries is zero.) Also, it was sleeping after the final
2880 failure, which is pointless.
2882 24. Implemented tls_on_connect_ports.
2884 25. Implemented acl_smtp_predata.
2886 26. If the domain in control=submission is set empty, Exim assumes that the
2887 authenticated id is a complete email address when it generates From: or
2888 Sender: header lines.
2890 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2891 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2892 chown and chgrp in /bin and hostname in /usr/bin.
2894 28. Exim was keeping the "process log" file open after each use, just as it
2895 does for the main log. This opens the possibility of it remaining open for
2896 long periods when the USR1 signal hits a daemon. Occasional processlog
2897 errors were reported, that could have been caused by this. Anyway, it seems
2898 much more sensible not to leave this file open at all, so that is what now
2901 29. The long-running daemon process does not normally write to the log once it
2902 has entered its main loop, and it closes the log before doing so. This is
2903 so that log files can straightforwardly be renamed and moved. However,
2904 there are a couple of unusual error situations where the daemon does write
2905 log entries, and I had neglected to close the log afterwards.
2907 30. The text of an SMTP error response that was received during a remote
2908 delivery was being truncated at 512 bytes. This is too short for some of
2909 the long messages that one sometimes sees. I've increased the limit to
2912 31. It is now possible to make retry rules that apply only when a message has a
2913 specific sender, in particular, an empty sender.
2915 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2916 it possible to be selective about when SMTP synchronization is enforced.
2918 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2920 32. Implemented hosts_connection_nolog.
2922 33. Added an ACL for QUIT.
2924 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2927 35. Added mailbox_size and mailbox_filecount to appendfile.
2929 36. Added control = no_multiline_responses to ACLs.
2931 37. There was a bug in the logic of the code that waits for the clock to tick
2932 in the case where the clock went backwards by a substantial amount such
2933 that the microsecond fraction of "now" was more than the microsecond
2934 fraction of "then" (but the whole seconds number was less).
2936 38. Added support for the libradius Radius client library this is found on
2937 FreeBSD (previously only the radiusclient library was supported).
2943 1. When certain lookups returned multiple values in the form name=value, the
2944 quoting of the values was not always being done properly. Specifically:
2945 (a) If the value started with a double quote, but contained no whitespace,
2947 (b) If the value contained whitespace other than a space character (i.e.
2948 tabs or newlines or carriage returns) it was not quoted.
2949 This fix has been applied to the mysql and pgsql lookups by writing a
2950 separate quoting function and calling it from the lookup code. The fix
2951 should probably also be applied to nisplus, ibase and oracle lookups, but
2952 since I cannot test any of those, I have not disturbed their existing code.
2954 2. A hit in the callout cache for a specific address caused a log line with no
2955 reason for rejecting RCPT. Now it says "Previous (cached) callout
2956 verification failure".
2958 3. There was an off-by-one bug in the queryprogram router. An over-long
2959 return line was truncated at 256 instead of 255 characters, thereby
2960 overflowing its buffer with the terminating zero. As well as fixing this, I
2961 have increased the buffer size to 1024 (and made a note to document this).
2963 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2964 when Exim is waiting for an SMTP response from a remote server, Exim
2965 restarts its select() call on the socket, thereby resetting its timeout.
2966 This is not a problem when such interrupts are rare. Somebody set up a cron
2967 job to run exiwhat every 2 minutes, which is less than the normal select()
2968 timeout (5 or 10 minutes). This meant that the select() timeout never
2969 kicked in because it was always reset. I have fixed this by comparing the
2970 time when an interrupt arrives with the time at the start of the first call
2971 to select(). If more time than the timeout has elapsed, the interrupt is
2972 treated as a timeout.
2974 5. Some internal re-factoring in preparation for the addition of Sieve
2975 extensions (by MH). In particular, the "personal" test is moved to a
2976 separate function, and given an option for scanning Cc: and Bcc: (which is
2977 not set for Exim filters).
2979 6. When Exim created an email address using the login of the caller as the
2980 local part (e.g. when creating a From: or Sender: header line), it was not
2981 quoting the local part when it contained special characters such as @.
2983 7. Installed new OpenBSD configuration files.
2985 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2986 try to make them clearer.
2988 9. Callout options, other than the timeout value, were being ignored when
2989 verifying sender addresses in header lines. For example, when using
2991 verify = header_sender/callout=no_cache
2993 the cache was (incorrectly) being used.
2995 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2996 only the Cygwin environment.
2998 11. When return_path_on_delivery was set as a log selector, if different remote
2999 addresses in the same message used different return paths and parallel
3000 remote delivery occurred, the wrong values would sometimes be logged.
3001 (Whenever a remote delivery process finished, the return path value from
3002 the most recently started remote delivery process was logged.)
3004 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
3005 lines when AUTH and/or TLS are in use. This is the "received protocol"
3006 field. Exim used to use "asmtp" for authenticated SMTP, without any
3007 indication (in the protocol name) for TLS use. Now it follows the RFC and
3008 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
3009 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
3010 names appear in log lines as well as in Received: header lines.
3012 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3013 extensions, and comparison tests, and to fix some bugs.
3015 14. Changes to the "personal" filter test:
3017 (1) The test was buggy in that it was just doing the equivalent of
3018 "contains" tests on header lines. For example, if a user's address was
3019 anne@some.where, the "personal" test would incorrectly be true for
3021 To: susanne@some.where
3023 This test is now done by extracting each address from the header in turn,
3024 and checking the entire address. Other tests that are part of "personal"
3025 are now done using regular expressions (for example, to check local parts
3026 of addresses in From: header lines).
3028 (2) The list of non-personal local parts in From: addresses has been
3029 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3030 taken from the Sieve specification recommendations.
3032 (3) If the message contains any header line starting with "List-" it is
3033 treated as non-personal.
3035 (4) The test for "circular" in the Subject: header line has been removed
3036 because it now seems ill-conceived.
3038 15. Minor typos in src/EDITME comments corrected.
3040 16. Installed latest exipick from John Jetmore.
3042 17. If headers_add on a router specified a text string that was too long for
3043 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3044 of string_sprintf() is now avoided.
3046 18. $message_body_size was not set (it was always zero) when running the DATA
3047 ACL and the local_scan() function.
3049 19. For the "mail" command in an Exim filter, no default was being set for
3050 the once_repeat time, causing a random time value to be used if "once" was
3051 specified. (If the value happened to be <= 0, no repeat happened.) The
3052 default is now 0s, meaning "never repeat". The "vacation" command was OK
3053 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3054 (I found it when inspecting the code).
3056 20. There is now an overall timeout for performing a callout verification. It
3057 defaults to 4 times the callout timeout, which applies to individual SMTP
3058 commands during the callout. The overall timeout applies when there is more
3059 than one host that can be tried. The timeout is checked before trying the
3060 next host. This prevents very long delays if there are a large number of
3061 hosts and all are timing out (e.g. when the network connections are timing
3062 out). The value of the overall timeout can be changed by specifying an
3063 additional sub-option for "callout", called "maxwait". For example:
3065 verify = sender/callout=5s,maxwait=20s
3067 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3068 to the end before writing, but this should make it even safer).
3070 22. Exim was forgetting that it had advertised PIPELINING for the second and
3071 subsequent messages on an SMTP connection. It was also not resetting its
3072 memory on STARTTLS and an internal HELO.
3074 23. When Exim logs an SMTP synchronization error within a session, it now
3075 records whether PIPELINING has been advertised or not.
3077 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3078 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3079 rather than long int.
3081 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3083 26. Added the never_mail option to autoreply.
3089 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3090 crash if the getsockname() call failed; this can happen if a connection is
3091 closed very soon after it is established. The problem was simply in the
3092 order in which certain operations were done, causing Exim to try to write
3093 to the SMTP stream before it had set up the file descriptor. The bug has
3094 been fixed by making things happen in the correct order.
3100 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3101 before the connection was closed, thus losing the rejection response.
3103 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3104 some early Solaris releases, but causes trouble in current releases where
3105 socklen_t is defined.
3107 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3110 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3113 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3114 that releasing the top of it at the end releases what was used for sub-
3115 expansions (unless the block got too big). However, discard this block if
3116 the first thing is a variable or header, so that we can use its block when
3117 it is dynamic (useful for very large $message_headers, for example).
3119 6. Lookups now cache *every* query, not just the most recent. A new, separate
3120 store pool is used for this. It can be recovered when all lookup caches are
3121 flushed. Lookups now release memory at the end of their result strings.
3122 This has involved some general refactoring of the lookup sources.
3124 7. Some code has been added to the store_xxx() functions to reduce the amount
3125 of flapping under certain conditions.
3127 8. log_incoming_interface used to affect only the <= reception log lines. Now
3128 it causes the local interface and port to be added to several more SMTP log
3129 lines, for example "SMTP connection from", and rejection lines.
3131 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3133 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3135 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3136 could be overwritten at the end of the current message (or the start of a
3137 new message if it was set in a HELO ACL). The value is now preserved for
3138 the duration of the SMTP connection.
3140 12. If a transport had a headers_rewrite setting, and a matching header line
3141 contained an unqualified address, that address was qualified, even if it
3142 did not match any rewriting rules. The underlying bug was that the values
3143 of the flags that permit the existence of unqualified sender and recipient
3144 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3145 non-local messages, and by -bnq for local messages) were not being
3146 preserved with the message after it was received.
3148 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3149 "next input=" as part of the text comprising the host identity instead of
3150 the correct text. The code was using the same buffer for two different
3151 strings. However, depending on which order the printing function evaluated
3152 its arguments, the bug did not always show up. Under Linux, for example, my
3153 test suite worked just fine.
3155 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3156 doesn't work with some older versions of Perl. It has been changed to "my",
3157 which in any case is probably the better facility to use.
3159 15. A really picky compiler found some instances of statements for creating
3160 error messages that either had too many or two few arguments for the format
3163 16. The size of the buffer for calls to the DNS resolver has been increased
3164 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3165 for addresses that have a lot of PTR records. This alleviates a problem; it
3166 does not fully solve it.
3168 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3169 buffer now truncates the list and logs the incident, which is the same
3170 action as happens when Exim is looking up a host name and its aliases.
3171 Previously in this situation something unpredictable would happen;
3172 sometimes it was "internal error: store_reset failed".
3174 18. If a server dropped the connection unexpectedly when an Exim client was
3175 using GnuTLS and trying to read a response, the client delivery process
3176 crashed while trying to generate an error log message.
3178 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3179 string, for example:
3181 warn message = H1: something\nH2: something
3183 the text was added as a single header line from Exim's point of view
3184 though it ended up OK in the delivered message. However, searching for the
3185 second and subsequent header lines using $h_h2: did not work. This has been
3186 fixed. Similarly, if a system filter added multiple headers in this way,
3187 the routers could not see them.
3189 20. Expanded the error message when iplsearch is called with an invalid key to
3190 suggest using net-iplsearch in a host list.
3192 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3193 ACLs are no longer actually imposed (and a message to that effect is
3196 22. If a "gecos" field in a passwd entry contained escaped characters, in
3197 particular, if it contained a \" sequence, Exim got it wrong when building
3198 a From: or a Sender: header from that name. A second bug also caused
3199 incorrect handling when an unquoted " was present following a character
3200 that needed quoting.
3202 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3203 was not being matched caselessly.
3205 24. Arranged for all hyphens in the exim.8 source to be escaped with
3208 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3209 properly. Recipient callout cache records were still being keyed to include
3210 the sender, even when use_sender was set false. This led to far more
3211 callouts that were necessary. The sender is no longer included in the key
3212 when use_sender is false.
3214 26. Added "control = submission" modifier to ACLs.
3216 27. Added the ${base62d: operator to decode base 62 numbers.
3218 28. dnsdb lookups can now access SRV records.
3220 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3221 the configuration file.
3223 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3224 (-v) mode. This makes the output for a verbose queue run more intelligible.
3226 31. Added a use_postmaster feature to recipient callouts.
3228 32. Added the $body_zerocount variable, containing the number of binary zero
3229 bytes in the message body.
3231 33. The time of last modification of the "new" subdirectory is now used as the
3232 "mailbox time last read" when there is a quota error for a maildir
3235 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3237 35. Added +ignore_unknown as a special item in host lists.
3239 36. Code for decoding IPv6 addresses in host lists is now included, even if
3240 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3241 address was recognized as an IP address, but was then not correctly decoded
3242 into binary, causing unexpected and incorrect effects when compared with
3249 1. Very minor rewording of debugging text in manualroute to say "list of
3250 hosts" instead of "hostlist".
3252 2. If verify=header_syntax was set, and a header line with an unqualified
3253 address (no domain) and a large number of spaces between the end of the
3254 name and the colon was received, the reception process suffered a buffer
3255 overflow, and (when I tested it) crashed. This was caused by some obsolete
3256 code that should have been removed. The fix is to remove it!
3258 3. When running in the test harness, delay a bit after writing a bounce
3259 message to get a bit more predictability in the log output.
3261 4. Added a call to search_tidyup() just before forking a reception process. In
3262 theory, someone could use a lookup in the expansion of smtp_accept_max_
3263 per_host which, without the tidyup, could leave open a database connection.
3265 5. Added the variables $recipient_data and $sender_data which get set from a
3266 lookup success in an ACL "recipients" or "senders" condition, or a router
3267 "senders" option, similar to $domain_data and $local_part_data.
3269 6. Moved the writing of debug_print from before to after the "senders" test
3272 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3273 problems for message scanning, either using a data ACL, or using
3274 local_scan() because the Received: header was not generated till after they
3275 were called (in order to set the time as the time of reception completion).
3276 I have revised the way this works. The header is now generated after the
3277 body is received, but before the ACL or local_scan() are called. After they
3278 are run, the timestamp in the header is updated.
3284 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3285 before starting a queue runner without re-exec. This happened only when
3286 deliver_drop_privilege was set or when the Exim user was set to root. The
3287 effect of the bug was that timeouts during subsequent deliveries caused
3288 crashes instead of being properly handled. The handler is now left at its
3289 default (and expected) setting.
3291 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3292 message, again when deliver_drop_privilege is set or Exim is run as root.
3293 The bug described in (1) was not present in this case, but the tidying up
3294 of the other signals was missing. I have made the two cases consistent.
3296 3. The ignore_target_hosts setting on a manualroute router was being ignored
3297 for hosts that were looked up using the /MX notation.
3299 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3302 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3303 operated on the sender address. After changing the $sender_address to <>
3304 for the sender address verify, Exim was re-instated it as the original
3305 (before rewriting) address, but remembering that it had rewritten it, so it
3306 wasn't rewriting it again. This bug also had the effect of breaking the
3307 sender address verification caching when the sender address was rewritten.
3309 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3310 This has been changed so that if the ip literal address matches
3311 ignore_target_hosts, the router declines.
3313 7. Added expansion conditions match_domain, match_address, and match_local_
3314 part (NOT match_host).
3316 8. The placeholder for the Received: header didn't have a length field set.
3318 9. Added code to Exim itself and to exim_lock to test for a specific race
3319 condition that could lead to file corruption when using MBX delivery. The
3320 issue is with the lockfile that is created in /tmp. If this file is removed
3321 after a process has opened it but before that process has acquired a lock,
3322 there is the potential for a second process to recreate the file and also
3323 acquire a lock. This could lead to two Exim processes writing to the file
3324 at the same time. The added code performs the same test as UW imapd; it
3325 checks after acquiring the lock that its file descriptor still refers to
3326 the same named file.
3328 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3329 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3332 11. Added the smtp_active_hostname option. If used, this will typically be made
3333 to depend on the incoming interface address. Because $interface_address is
3334 not set up until the daemon has forked a reception process, error responses
3335 that can happen earlier (such as "too many connections") no longer contain
3338 12. If an expansion in a condition on a "warn" statement fails because a lookup
3339 defers, the "warn" statement is abandoned, and the next ACL statement is
3340 processed. Previously this caused the whole ACL to be aborted.
3342 13. Added the iplsearch lookup type.
3344 14. Added ident_timeout as a log selector.
3346 15. Added tls_certificate_verified as a log selector.
3348 16. Added a global option tls_require_ciphers (compare the smtp transport
3349 option of the same name). This controls incoming TLS connections.
3351 17. I finally figured out how to make tls_require_ciphers do a similar thing
3352 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3353 before starting the TLS session.
3355 18. Tabs are now shown as \t in -bP output.
3357 19. If the log selector return_path_on_delivery was set, Exim crashed when
3358 bouncing a message because it had too many Received: header lines.
3360 20. If two routers both had headers_remove settings, and the first one included
3361 a superfluous trailing colon, the final name in the first list and the
3362 first name in the second list were incorrectly joined into one item (with a
3363 colon in the middle).
3369 1. Added -C and -D options to the exinext utility, mainly to make it easier
3370 to include in the automated testing, but these could be helpful when
3371 multiple configurations are in use.
3373 2. The exinext utility was not formatting the output nicely when there was
3374 an alternate port involved in the retry record key, nor when there was a
3375 message id as well (for retries that were specific to a specific message
3376 and a specific host). It was also confused by IPv6 addresses, because of
3377 the additional colons they contain. I have fixed the IPv4 problem, and
3378 patched it up to do a reasonable job for IPv6.
3380 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3381 delivery, the log line now contains "pipelined" if PIPELINING was used.
3383 4. An SMTP transport process used to panic and die if the bind() call to set
3384 an explicit outgoing interface failed. This has been changed; it is now
3385 treated in the same way as a connect() failure.
3387 5. A reference to $sender_host_name in the part of a conditional expansion
3388 that was being skipped was still causing a DNS lookup. This no longer
3391 6. The def: expansion condition was not recognizing references to header lines
3392 that used bh_ and bheader_.
3394 7. Added the _cache feature to named lists.
3396 8. The code for checking quota_filecount in the appendfile transport was
3397 allowing one more file than it should have been.
3399 9. For compatibility with Sendmail, the command line option
3407 and sets the incoming protocol and host name (for trusted callers). The
3408 host name and its colon can be omitted when only the protocol is to be set.
3409 Note the Exim already has two private options, -pd and -ps, that refer to
3410 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3411 "s", but I don't think that's a major issue.
3413 10. A number of refactoring changes to the code, none of which should affect
3416 (a) The number of logging options was getting close to filling up the
3417 32-bit word that was used as a bit map. I have split them into two classes:
3418 those that are passed in the argument to log_write(), and those that are
3419 only ever tested independently outside of that function. These are now in
3420 separate 32-bit words, so there is plenty of room for expansion again.
3421 There is no change in the user interface or the logging behaviour.
3423 (b) When building, for example, log lines, the code previously used a
3424 macro that called string_cat() twice, in order to add two strings. This is
3425 not really sufficiently general. Furthermore, there was one instance where
3426 it was actually wrong because one of the argument was used twice, and in
3427 one call a function was used. (As it happened, calling the function twice
3428 did not affect the overall behaviour.) The macro has been replaced by a
3429 function that can join an arbitrary number of extra strings onto a growing
3432 (c) The code for expansion conditions now uses a table and a binary chop
3433 instead of a serial search (which was left over from when there were very
3434 few conditions). Also, it now recognizes conditions like "pam" even when
3435 the relevant support is not compiled in: a suitably worded error message is
3436 given if an attempt is made to use such a condition.
3438 11. Added ${time_interval:xxxxx}.
3440 12. A bug was causing one of the ddress fields not to be passed back correctly
3441 from remote delivery subprocesses. The field in question was not being
3442 subsequently used, so this caused to problems in practice.
3444 13. Added new log selectors queue_time and deliver_time.
3446 14. Might have fixed a bug in maildirsizefile handling that threw up
3447 "unexpected character" debug warnings, and recalculated the data
3448 unnecessarily. In any case, I expanded the warning message to give more
3451 15. Added the message "Restricted characters in address" to the statements in
3452 the default ACL that block characters like @ and % in local parts.
3454 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3455 Three changes have been made:
3457 (a) There was a serious bug; a negative response to MAIL caused the whole
3458 recipient domain to be cached as invalid, thereby blocking all messages
3459 to all local parts at the same domain, from all senders. This bug has
3460 been fixed. The domain is no longer cached after a negative response to
3461 MAIL if the sender used is not empty.
3463 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3466 (c) A new callout option, "use_sender" has been added for people who want
3467 the modified behaviour.
3473 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3476 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3477 indeed breaks things for older releases.
3479 3. Added additional logging to the case where there is a problem reading data
3480 from a filter that is running in a subprocess using a pipe, in order to
3481 try to track down a specific problem.
3483 4. Testing facility fudge: when running in the test harness and attempting
3484 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3485 getting "No route to host". Convert this to a timeout.
3487 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3490 6. Some OS don't have socklen_t but use size_t instead. This affects the
3491 fifth argument of getsockopt() amongst other things. This is now
3492 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3493 can be set for individual OS. I have set it for SunOS5, OSF1, and
3494 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3495 some earlier ones do not.
3497 7. Change 4.30/15 was not doing the test caselessly.
3499 8. The standard form for an IPv6 address literal was being rejected by address
3500 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3501 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3502 this, as well as the form without the "IPv6" on the front (but only when
3503 address literals are enabled, of course).
3505 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3507 10. Exim crashed if a message with an empty sender address specified by -f
3508 encountered a router with an errors_to setting. This could be provoked only
3509 by a command such as
3513 where an empty string was supplied; "<>" did not hit this bug.
3515 11. Installed PCRE release 4.5.
3517 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3518 remained set. It is now erased.
3520 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3521 times from message ids (which are base 36 rather than the normal 62).
3523 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3524 were being counted as actual protocol errors, and logged if the log
3525 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3526 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3527 and DATA following a set of rejected RCPTs do not count as protocol errors.
3528 In other words, Exim assumes they were pipelined, though this may not
3529 actually be the case. Of course, in all cases the client gets an
3530 appropriate error code.
3532 15. If a lookup fails in an ACL condition, a message about the failure may
3533 be available; it is used if testing the ACL cannot continue, because most
3534 such messages specify what the cause of the deferral is. However, some
3535 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3536 that caused an old message to be retained and used if a later statement
3537 caused a defer, replacing the real cause of the deferral.
3539 16. If an IP address had so many PTR records that the DNS lookup buffer
3540 was not large enough to hold them, Exim could crash while trying to process
3541 the truncated data. It now detects and logs this case.
3543 17. Further to 4.21/58, another change has been made: if (and only if) the
3544 first line of a message (the first header line) ends with CRLF, a bare LF
3545 in a subsequent header line has a space inserted after it, so as not to
3546 terminate the header.
3548 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3549 unnecessarily, used atoi() instead of strtol(), and didn't check the
3550 termination when getting file sizes from file names by regex.
3552 19. Completely re-implemented the support for maildirsize files, in the light
3553 of a number of problems with the previous contributed implementation
3554 (4.30/29). In particular:
3556 . If the quota is zero, the maildirsize file is maintained, but no quota is
3559 . If the maildir directory does not exist, it is created before any attempt
3560 to write a maildirsize file.
3562 . The quota value in the file is just a cache; if the quota is changed in
3563 the transport, the new value overrides.
3565 . A regular expression is available for excluding directories from the
3568 20. The autoreply transport checks the characters in options that define the
3569 message's headers; it allows continued headers, but it was checking with
3570 isspace() after an embedded newline instead of explicitly looking for a
3573 21. If all the "regular" hosts to which an address was routed had passed their
3574 expiry times, and had not reached their retry times, the address was
3575 bounced, even if fallback hosts were defined. Now Exim should go on to try
3578 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3579 equivalent code in the SMTP transport. Some hosts send humungous responses
3580 to HELO/EHLO, more than 1024 it seems.
3582 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3583 gives compiler warnings in some environments. I've now done it "properly",
3586 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3587 (because of problems with the built-in one) was declared to return uschar *
3588 instead of char *, causing compiler failure.
3590 25. Fixed a file descriptor leak when processing alias/forward files.
3592 26. Fixed a minor format string issue in dbfn.c.
3594 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3596 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3597 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3600 29. When a sender address is verified, it is cached, to save repeating the test
3601 when there is more than one recipient in a message. However, when the
3602 verification involves a callout, it is possible for different callout
3603 options to be set for different recipients. It is too complicated to keep
3604 track of this in the cache, so now Exim always runs a verification when a
3605 callout is required, relying on the callout cache for the optimization.
3606 The overhead is duplication of the address routing, but this should not be
3609 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3610 to be verified with callout=postmaster, and the main callout worked but the
3611 postmaster check failed, the verification correctly failed. However, if a
3612 subsequent RCPT command asked for sender verification *without* the
3613 postmaster check, incorrect caching caused this verification also to fail,
3616 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3617 it was not caching the DNS options (qualify_single, search_parents) that
3618 were used when the lookup failed. A subsequent lookup with different
3619 options therefore always gave the same answer, though there were cases
3620 where it should not have. (Example: a "domains = !$mx_any" option on a
3621 dnslookup router: the "domains" option is always processed without any
3622 widening, but the router might have qualify_single set.) Now Exim uses the
3623 cached value only when the same options are set.
3625 32. Added John Jetmore's "exipick" utility to the distribution.
3627 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3628 than a timeout (e.g. a certificate is required, and is not provided), an
3629 Exim server now closes the connection immediately. Previously it waited for
3630 the client to close - but if the client is SSL, it seems that they each
3631 wait for each other, leading to a delay before one of them times out.
3633 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3634 maintained 0.8.x compatibility because I don't think many are using it, and
3635 it is clearly obsolete.
3637 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3640 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3641 client certificate was expired. A simple patch fixes this, though I don't
3642 understand the full logic of why the verify callback is called multiple
3645 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3646 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3647 which causes problems with some clients (such as the Certicom SSL Plus
3648 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3649 disables the coutermeasure allowing Eudora to connect."
3651 38. Exim was not checking that a write() to a log file succeeded. This could
3652 lead to Bad Things if a log got too big, in particular if it hit a file
3653 size limit. Exim now panics and dies if it cannot write to a log file, just
3654 as it does if it cannot open a log file.
3656 39. Modified OS/Makefile-Linux so that it now contains
3658 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3660 The two -D definitions ensure that Exim is compiled with large file
3661 support, which makes it possible to handle log files that are bigger than
3664 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3665 instance) a domain was checked against a named list that involved a lookup,
3666 causing $domain_data to be set, then another domain was checked against the
3667 same list, then the first domain was re-checked, the value of $domain_data
3668 after the final check could be wrong. In particular, if the second check
3669 failed, it could be set empty. This bug probably also applied to
3672 41. The strip_trailing_dot option was not being applied to the address given
3673 with the -f command-line option.
3675 42. The code for reading a message's header from the spool was incrementing
3676 $received_count, but never initializing it. This meant that the value was
3677 incorrect (doubled) while delivering a message in the same process in which
3678 it was received. In the most common configuration of Exim, this never
3679 happens - a fresh exec is done - but it can happen when
3680 deliver_drop_privilege is set.
3682 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3683 it now includes up to 150 characters of the unexpected data in the log
3686 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3687 and building data strings. The size of both of these buffers was 10 000
3688 bytes - far larger than anybody would *ever* want, thought I. Needless to
3689 say, somebody hit the limit. I have increased the maximum line length to
3690 20 000 and the maximum data length of concatenated lines to 100 000. I have
3691 also fixed two bugs, because there was no checking on these buffers. Tsk,
3692 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3693 buffer is too small.
3695 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3696 lsearch lookups. Now it does.
3698 46. When parsing a route_list item in a manualroute router, a fixed-length
3699 buffer was used for the list of hosts. I made this 1024 bytes long,
3700 thinking that nobody would ever have a list of hosts that long. Wrong.
3701 Somebody had a whole pile of complicated expansion conditions, and the
3702 string was silently truncated, leading to an expansion error. It turns out
3703 that it is easier to change to an unlimited length (owing to other changes
3704 that have happened since this code was originally written) than to build
3705 structure for giving a limitation error. The length of the item that
3706 expands into the list of hosts is now unlimited.
3708 47. The lsearch lookup could not handle data where the length of text line was
3709 more than 4095 characters. Such lines were truncated, leading to shortened
3710 data being returned. It should now handle lines of any length.
3712 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3713 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3716 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3717 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3718 no longer happens in comment lines. A list of replacements is now placed
3719 at the head of all of the source files, except those whose only change is
3720 to replace PERL_COMMAND in the very first #! line.
3722 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3723 messages on the queue, with a bottom-up merge sort, using code contributed
3724 by Michael Haardt. This should make operations like -bp somewhat faster on
3725 large queues. It won't affect queue runners, except when queue_run_in_order
3728 51. Installed eximstats 1.31 in the distribution.
3730 52. Added support for SRV lookups to the dnslookup router.
3732 53. If an ACL referred to $message_body or $message_body_end, the value was not
3733 reset for any messages that followed in the same SMTP session.
3735 54. The store-handling optimization for building very long strings was not
3736 differentiating between the different store pools. I don't think this
3737 actually made any difference in practice, but I've tidied it.
3739 55. While running the routers to verify a sender address, $sender_address
3740 was still set to the sender address. This is wrong, because when routing to
3741 send a bounce to the sender, it would be empty. Therefore, I have changed
3742 it so that, while verifying a sender address, $sender_address is set to <>.
3743 (There is no change to what happens when verifying a recipient address.)
3745 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3746 target A or AAAA records (if not already returned) without resetting the
3747 qualify_single or search_parents options of the DNS resolver. These are
3748 inappropriate in this case because the targets of MX and SRV records must
3749 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3750 target that, when qualified, matched something in the local domain. These
3751 two options are now turned off when doing these lookups.
3753 57. It seems that at least some releases of Reiserfs (which does not have the
3754 concept of a fixed number of inodes) returns zero and not -1 for the
3755 number of available inodes. This interacted badly with check_spool_inodes,
3756 which assumed that -1 was the "no such thing" setting. What I have done is
3757 to check that the total number of inodes is greater than zero before doing
3758 the test of how many are available.
3760 58. When a "warn" ACL statement has a log_message modifier, the message is
3761 remembered, and not repeated. This is to avoid a lot of repetition when a
3762 message has many recipients that cause the same warning to be written.
3763 Howewer, Exim was preserving the list of already written lines for an
3764 entire SMTP session, which doesn't seem right. The memory is now reset if a
3765 new message is started.
3767 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3768 debugging output unless log_rewrite was also set.
3770 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3771 of (int)(handle) when we know that handle contains (void *)(-1).
3773 61. The Exim daemon panic-logs an error return when it closes the incoming
3774 connection. However "connection reset by peer" seems to be common, and
3775 isn't really an error worthy of noting specially, so that particular error
3778 62. When Exim is trying to find all the local interfaces, it used to panic and
3779 die if the ioctl to get the interface flags failed. However, it seems that
3780 on at least one OS (Solaris 9) it is possible to have an interface that is
3781 included in the list of interfaces, but for which you get a failure error
3782 for this call. This happens when the interface is not "plumbed" into a
3783 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3784 failure of the "get flags" call assumes that the interface is down.
3786 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3787 makes life easier for people who are doing arithmetic on fields extracted
3788 from dates, where you often get leading zeros that should not be
3789 interpreted as octal.
3791 64. Added qualify_domain to the redirect router, to override the global
3794 65. If a pathologically long header line contained very many addresses (the
3795 report of this problem mentioned 10 000) and each of them was rewritten,
3796 Exim could use up a very large amount of memory. (It kept on making new
3797 copies of the header line as it rewrote, and never released the old ones.)
3798 At the expense of a bit more processing, the header rewriting function has
3799 been changed so that it no longer eats memory in this way.
3801 66. The generation of the Received: header has been moved from the time that a
3802 message starts to be received, to the time that it finishes. The timestamp
3803 in the Received: header should now be very close to that of the <= log
3804 line. There are two side-effects of this change:
3806 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3807 logged header lines no longer include the local Received: line, because
3808 it has not yet been created. The same applies to a copy of the message
3809 that is returned to a non-SMTP sender when a message is rejected.
3811 (b) When a filter file is tested using -bf, no additional Received: header
3812 is added to the test message. After some thought, I decided that this
3815 This change does not affect the value of $received_for. It is still set
3816 after address rewriting, but before local_scan() is called.
3818 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3820 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3821 gave an unhelpful panic error message, and a defer error. I have managed to
3822 change this behaviour so that it now rejects any supplied certificate,
3823 which seems right, as the list of acceptable certificates is empty.
3825 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3826 gave an unhelpful defer error. I have not managed to make this reject any
3827 supplied certificates, but the error message it gives is "no certificate
3828 supplied", which is not helpful.
3830 70. exigrep's output now also includes lines that are not associated with any
3831 message, but which match the given pattern. Implemented by a patch from
3832 Martin Sluka, which also tidied up the Perl a bit.
3834 71. Recipient callout verification, like sender verification, was using <> in
3835 the MAIL FROM command. This isn't really the right thing, since the actual
3836 sender may affect whether the remote host accepts the recipient or not. I
3837 have changed it to use the actual sender in the callout; this means that
3838 the cache record is now keyed on a recipient/sender pair, not just the
3839 recipient address. There doesn't seem to be a real danger of callout loops,
3840 since a callout by the remote host to check the sender would use <>.
3841 [SEE ABOVE: changed after hitting problems.]
3843 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3844 temporary errors. However, in the case of such a code being given after
3845 the end of a data transmission (i.e. after ".") Exim was failing to write
3846 a retry record for the message. (Yes, there was some broken host that was
3847 actually sending 8xx at this point.)
3849 73. An unknown lookup type in a host list could cause Exim to panic-die when
3850 the list was checked. (An example that provoked this was putting <; in the
3851 middle of a list instead of at the start.) If this happened during a DATA
3852 ACL check, a -D file could be left lying around. This kind of configuration
3853 error no longer causes Exim to die; instead it causes a defer errror. The
3854 incident is still logged to the main and panic logs.
3856 74. Buglet left over from Exim 3 conversion. The message "too many messages
3857 in one connection" was written to the rejectlog but not the mainlog, except
3858 when address rewriting (yes!) was being logged.
3860 75. Added write_rejectlog option.
3862 76. When a system filter was run not as root (that is, when system_filter_user
3863 was set), the values of the $n variables were not being returned to the
3864 main process; thus, they were not subsequently available in the $sn
3867 77. Added +return_path_on_delivery log selector.
3869 78. A connection timeout was being treated differently from recipients deferred
3870 when testing hosts_max_try with a message that was older than the host's
3871 retry timeout. (The host should not be counted, thus allowing all hosts to
3872 be tried at least once before bouncing.) This may have been the cause of an
3873 occasionally reported bug whereby a message would remain on the queue
3874 longer than the retry timeout, but would be bounced if a delivery was
3875 forced. I say "may" because I never totally pinned down the problem;
3876 setting up timeout/retry tests is difficult. See also the next item.
3878 79. The ultimate address timeout was not being applied to errors that involved
3879 a combination of host plus message (for example, a timeout on a MAIL
3880 command). When an address resolved to a number of possible hosts, and they
3881 were not all tried for each delivery (e.g. because of hosts_max_try), a
3882 message could remain on the queue longer than the retry timeout.
3884 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3887 81. Fixed an obscure SMTP outgoing bug which required at least the following
3888 conditions: (a) there was another message waiting for the same server;
3889 (b) the server returned 5xx to all RCPT commands in the first message so
3890 that the message was not completed; (c) the server dropped the connection
3891 or gave a negative response to the RSET that Exim sends to abort the
3892 transaction. The observed case was a dropped connection after DATA that had
3893 been sent in pipelining mode. That is, the server had advertised PIPELINING
3894 but was not implementing it correctly. The effect of the bug was incorrect
3895 behaviour, such as trying another host, and this could lead to a crash.
3901 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3902 and daemon.c were passed as pointers to ints; they should have been
3903 pointers to socklen_t variables (which are typically unsigned ints).
3905 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3908 3. Fixed a really odd bug that affected only the testing scheme; patching a
3909 certain fixed string in the binary changed the value of another string that
3910 happened to be identical to the end of the original first string.
3912 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3913 name", it returns that address as the IP address. On some operating
3914 systems (e.g. Solaris), it also passes back the IP address string as the
3915 "host name". However, on others (e.g. Linux), it passes back an empty
3916 string. Exim wasn't checking for this, and was changing the host name to an
3917 empty string, assuming it had been canonicized.
3919 5. Although rare, it is permitted to have more than one PTR record for a given
3920 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3921 all the names associated with an address, because they do in Solaris.
3922 However, it seems that they do not in Linux for data that comes from the
3923 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3924 I found this out when I moved to a new Linux workstation and tried to run
3925 the Exim test suite.
3927 To get round this problem I have changed the code so that it now does its
3928 own call to the DNS to look up PTR records when searching for a host name.
3929 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3930 addresses that are only in /etc/hosts are still found.
3932 This behaviour is, however, controlled by an option called host_lookup_
3933 order, which defaults to "bydns:byaddr". If people want to use the other
3934 order, or indeed, just use one or the other means of lookup, they can
3935 specify it in this variable.
3937 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3938 some operating systems, this comes back from gethostbyaddr() as an empty
3939 string, and this is what Exim used to test for. However, it seems that in
3940 other systems, "." is yielded. Exim now tests for this case too.
3942 7. The values of check_spool_space and check_log_space are now held internally
3943 as a number of kilobytes instead of an absolute number of bytes. If a
3944 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3945 kilobyte. This means that much larger values can be stored.
3947 8. Exim monitor: an attempt to get the action menu when not actually pointing
3948 at a message produces an empty menu entitled "No message selected". This
3949 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3950 no entries in it ("Shell widget menu has zero width and/or height"). So I
3951 have added a single, blank menu entry in this case.
3953 9. Added ${quote_local_part.
3955 10. MIME decoding is now applied to the contents of Subject: header lines when
3958 11. Now that a reference to $sender_host_address automatically causes a reverse
3959 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3960 host lookup before query-style lookups in lists that might use this
3961 variable. This has therefore been abolished, and the "net-" prefix is no
3962 longer necessary for query-style lookups.
3964 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3965 have been a typo for LFLAGS, so it has been changed.
3967 13. The install script calls Exim with "-C /dev/null" in order to find the
3968 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3969 to be output. Howeve, since Exim outputs its version number before the
3970 error, it didn't break the script. It just looked ugly. I fixed this by
3971 always allowing "-C /dev/null" if the caller is root.
3973 14. Ignore overlarge ACL variable number when reading spool file - insurance
3974 against a later release with more variables having written the file.
3976 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3977 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3978 this, as well as the form without the "IPv6" on the front.
3980 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3981 OS/Makefile-Darwin file.
3983 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3984 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3986 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3987 variable. However, some OS use ints for inodes. I've added cast to long int
3988 to get rid of the compiler warning.
3990 19. I had forgotten to lock out "/../" in configuration file names when
3991 ALT_CONFIG_PREFIX was set.
3993 20. Routers used for verification do not need to specify transports. However,
3994 if such a router generated a host list, and callout was configured, Exim
3995 crashed, because it could not find a port number from the (non-existent)
3996 transport. It now assumes port 25 in this circumstance.
3998 21. Added the -t option to exigrep.
4000 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
4001 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
4002 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
4003 them is set, LOOKUP_LSEARCH is forced.
4005 23. "exim -bV" now outputs a list of lookups that are included in the binary.
4007 24. Added sender and host information to the "rejected by local_scan()" log
4008 line; previously there was no indication of these.
4010 25. Added .include_if_exists.
4012 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4013 Linux. It turns out that not all file systems support this. Apparently some
4014 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4015 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4016 not supported on the file descriptor, is now ignored when Exim is trying to
4017 sync a directory. This applies only to Linux.
4019 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4021 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4022 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4025 29. Added support for maildirsize files from supplied patch (modified a bit).
4027 30. The use of :fail: followed by an empty string could lead Exim to respond to
4028 sender verification failures with (e.g.):
4030 550 Verification failed for <xxx>
4031 550 Sender verify failed
4033 where the first response line was missing the '-' that indicates it is not
4034 the final line of the response.
4036 31. The loop for finding the name of the user that called Exim had a hardwired
4037 limit of 10; it now uses the value of finduser_retries, which is used for
4038 all other user lookups.
4040 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4043 33. Exim was neglecting to zero errno before one call of strtol() when
4044 expanding a string and expecting an integer value. On some systems this
4045 resulted in spurious "integer overflow" errors. Also, it was casting the
4046 result into an int without checking.
4048 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4049 did not work. The code looks as if it has *never* worked, though it appears
4050 to have been documented since at least releast 1.62. I have made it work.
4052 35. The "timeout_DNS" error in retry rules, also documented since at least
4053 1.62, also never worked. As it isn't clear exactly what this means, and
4054 clearly it isn't a major issue, I have abolished the feature by treating it
4055 as "timeout", and writing a warning to the main and panic logs.
4057 36. The display of retry rules for -brt wasn't always showing the error code
4060 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4061 timeout_connect_A, timeout_connect_MX.
4063 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4064 to the empty sender.
4066 39. The daemon was not analysing the content of -oX till after it had closed
4067 stderr and disconnected from the controlling terminal. This meant that any
4068 syntax errors were only noted on the panic log, and the return code from
4069 the command was 0. By re-arranging the code a little, I've made the
4070 decoding happen first, so such errors now appear on stderr, and the return
4071 code is 1. However, the actual setting up of the sockets still happens in
4072 the disconnected process, so errors there are still only recorded on the
4075 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4076 connections (as happens on some IP stacks) was logged at start up time as
4077 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4078 it from "IPv6 and IPv4", which means that two separate sockets are being
4081 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4082 says whether AF_INET or AF_INET6 was passed as an argument.
4084 42. Exiwhat output was messed up when time zones were included in log
4087 43. Exiwhat now gives more information about the daemon's listening ports,
4088 and whether -tls-on-connect was used.
4090 44. The "port" option of the smtp transport is now expanded.
4092 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4093 silently ignored. Now an error message is written to the main and panic
4096 46. There's a new ACL modifier called "logwrite" which writes to a log file
4097 as soon as it is encountered.
4099 47. Added $local_user_uid and $local_user_gid at routing time.
4101 48. Exim crashed when trying to verify a sender address that was being
4104 49. Exim was recognizing only a space character after ".include". It now also
4105 recognizes a tab character.
4107 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4108 extracting the relevant information from the specification. The man page no
4109 longer contains scrambled data for the -d option, and I've added a section
4110 at the front about calling Exim under different names.
4112 51. Added "extra_headers" argument to the "mail" command in filter files.
4114 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4117 53. Installed eximstats 1.29.
4119 54. Added transport_filter_timeout as a generic transport option.
4121 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4122 Cc: header lines. This was required by RFC 822, but it not required by RFC
4125 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4126 incoming messages that did not have them. Now it does so only if the
4127 message originates locally, that is, if there is no associated remote host
4128 address. When Resent- header lines are present, this applies to the Resent-
4129 lines rather than the non-Resent- lines.
4131 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4132 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4134 58. Messages for configuration errors now include the name of the main
4135 configuration file - useful now that there may be more than one file in a
4136 list (.included file names were always shown).
4138 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4139 for those rare installations that do not start the daemon as root or run it
4140 setuid root. I've cut out the call to initgroups() if the daemon is not
4143 60. The Exim user and group can now be bound into the binary as text strings
4144 that are looked up at the start of Exim's processing.
4146 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4148 62. Added $mailstore_basename variable.
4150 63. Installed patch to sieve.c from Michael Haardt.
4152 64. When Exim failed to open the panic log after failing to open the main log,
4153 the original message it was trying to log was written to stderr and debug
4154 output, but if they were not available (the usual case in production), it
4155 was lost. Now it is written to syslog before the two lines that record the
4156 failures to open the logs.
4158 65. Users' Exim filters run in subprocesses under the user's uid. It is
4159 possible for a "deliver" command or an alias in a "personal" command to
4160 provoke an address rewrite. If logging of address rewriting is configured,
4161 this fails because the process is not running as root or exim. There may be
4162 a better way of dealing with this, but for the moment (because 4.30 needs
4163 to be released), I have disabled address rewrite logging when running a
4164 filter in a non-root, non-exim process.
4170 1. The buildconfig auxiliary program wasn't quoting the value set for
4171 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4172 not defined. This bug was masked in 4.22 by the effect that was fixed in
4175 2. Some messages that were rejected after a message id was allocated were
4176 shown as "incomplete" by exigrep. It no longer does this for messages that
4177 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4179 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4180 have allow_domain_literals set, the ID did not get logged in the <= line.
4181 Domain literals are now always recognized in Message-ID: header lines.
4183 4. The first argument for a ${extract expansion item is the key name or field
4184 number. Leading and trailing spaces in this item were not being ignored,
4185 causing some misleading effects.
4187 5. When deliver_drop_privilege was set, single queue runner processes started
4188 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4189 same command in the process it spins off) were not dropping privilege.
4191 6. When the daemon running as "exim" started a queue runner, it always
4192 re-executed Exim in the spun-off process. This is a waste of effort when
4193 deliver_drop_privilege is set. The new process now just calls the
4194 queue-runner function directly.
4200 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4203 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4204 ignored. Though the use of -oP was forcing the writing of a pid file, it
4205 was always written to the default place.
4207 3. If the message "no IP address found for host xxxx" is generated during
4208 incoming verification, it is now followed by identification of the incoming
4209 connection (so you can more easily find what provoked it).
4211 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4213 5. Added some features to "harden" Exim a bit more against certain attacks:
4215 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4216 be put in Local/Makefile. This is like the never_users runtime option,
4217 but it cannot be overridden. The default setting is "root".
4219 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4220 prefix string with which any file named in a -C command line option
4223 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4224 is retained for -C and -D only if the caller of Exim is root. Without
4225 it, the exim user may also use -C and -D and retain privilege.
4227 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4228 command line option is disabled.
4230 6. Macro names set by the -D option must start with an upper case letter, just
4231 like macro names defined in the configuration file.
4233 7. Added "dereference=" facility to LDAP.
4235 8. Two instances of the typo "uknown" in the source files are fixed.
4237 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4238 the Configure-Makefile script screwed up while processing it.
4240 10. Incorporated PCRE 4.4.
4242 11. The SMTP synchronization check was not operating right at the start of an
4243 SMTP session. For example, it could not catch a HELO sent before the client
4244 waited for the greeting. There is now a check for outstanding input at the
4245 point when the greeting is written. Because of the duplex, asynchronous
4246 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4247 way, but not yet received, when the check is performed.
4249 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4250 on TCP/IP sockets, because this apparently causes some broken clients to
4253 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4254 unchanged) from the Cygwin maintainer.
4256 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4257 was supported instead of testing for mbx. Effectively a typo.
4259 15. The spa authenticator server code was not checking that the input it
4260 received was valid base64.
4262 16. The debug output line for the "set" modifier in ACLs was not showing the
4263 name of the variable that was being set.
4265 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4267 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4268 lookup on its own. Something else was needed to trigger the lookup. For
4269 example, a match in host_lookup or the need for a host name in a host list.
4270 Now, if $sender_host_name is referenced and the host name has not yet been
4271 looked up, a lookup is performed. If the lookup fails, the variable remains
4272 empty, and $host_lookup_failed is set to "1".
4274 19. Added "eqi" as a case-independent comparison operator.
4276 20. The saslauthd authentication condition could segfault if neither service
4277 nor realm was specified.
4279 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4280 error message that was logged was misleading, and incoming SMTP
4281 connections were dropped. The message is now more accurate, and temporary
4282 errors are given to SMTP connections.
4284 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4285 (except RSET) with a 421 error, until QUIT is received. However, it was
4286 failing to send a response to QUIT.
4288 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4289 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4290 ACL is now run after the helo_try_verify_hosts code.
4292 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4293 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4294 case-independent, so other case variants are also recognized). Apparently
4295 some systems use these upper case variants.
4297 25. If more than two messages were waiting for the same host, and a transport
4298 filter was specified for the transport, Exim sent two messages over the
4299 same TCP/IP connection, and then failed with "socket operation on non-
4300 socket" when it tried to send the third.
4302 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4304 27. The extern definition of crypt16() in expand.c was not being excluded when
4305 the OS had its own crypt16() function.
4307 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4308 as a preferred synonym for return_size_limit, both as an option and as an
4311 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4313 30. Changed the default configuration ACL to relax the local part checking rule
4314 for addresses that are not in any local domains. For these addresses,
4315 slashes and pipe symbols are allowed within local parts, but the sequence
4316 /../ is explicitly forbidden.
4318 31. SPA server authentication was not clearing the challenge buffer before
4321 32. log_message in a "warn" ACL statement was writing to the reject log as
4322 well as to the main log, which contradicts the documentation and doesn't
4323 seem right (because no rejection is happening). So I have stopped it.
4325 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4326 However, I am unable to do any testing of this.
4328 34. Fixed an infelicity in the appendfile transport. When checking directories
4329 for a mailbox, to see if any needed to be created, it was accidentally
4330 using path names with one or more superfluous leading slashes; tracing
4331 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4333 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4334 contents are added to the log line that is written for every discarded
4335 recipient. (Previously a log_message setting was ignored.)
4337 36. The ${quote: operator now quotes the string if it is empty.
4339 37. The install script runs exim in order to find its version number. If for
4340 some reason other than non-existence or emptiness, which it checks, it
4341 could not run './exim', it was installing it with an empty version number,
4342 i.e. as "exim-". This error state is now caught, and the installation is
4345 38. An argument was missing from the function that creates an error message
4346 when Exim fails to connect to the socket for saslauthd authentication.
4347 This could cause Exim to crash, or give a corrupted message.
4349 39. Added isip, isip4, and isip6 to ${if conditions.
4351 40. The ACL variables $acl_xx are now saved with the message, and can be
4352 accessed later in routers, transports, and filters.
4354 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4355 strings in the file are not string-expanded.
4357 42. If a MAIL command specified a SIZE value that was too large to fit into an
4358 int variable, the check against message_size_limit failed. Such values are
4359 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4360 day this will have to be increased, but I don't think I want to be around
4361 when emails are that large.
4368 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4369 iconv() is not standard in FreeBSD.
4371 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4372 IPv6 enabled. The observed symptom was a segmentation fault on return from
4373 the function os_common_find_running_interfaces() in src/os.c.
4375 3. In the check_special_case() function in daemon.c I had used "errno" as an
4376 argument name, which causes warnings on some systems. This was basically a
4377 typo, since it was named "eno" in the comments!
4379 4. The code that waits for the clock to tick (at a resolution of some fraction
4380 of a second) so as to ensure message-id uniqueness was always waiting for
4381 at least one whole tick, when it could have waited for less. [This is
4382 almost certainly not relevant at current processor speeds, where it is
4383 unlikely to ever wait at all. But we try to future-proof.]
4385 5. The function that sleeps for a time interval that includes fractions of a
4386 second contained a race. It did not block SIGALRM between setting the
4387 timer, and suspending (a couple of lines later). If the interval was short
4388 and the sigsuspend() was delayed until after it had expired, the suspension
4389 never ended. On busy systems this could lead to processes getting stuck for
4392 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4393 process, before it forks any delivery processes. The open lookup caching
4394 mechanism meant that the open file or database connection was passed into
4395 the delivery process. The problem was that delivery processes always tidy
4396 up cached lookup data. This could cause a problem for the next delivery
4397 process started by the queue runner, because the external queue runner
4398 process does not know about the closure. So the next delivery process
4399 still has data in the lookup cache. In the case of a file lookup, there was
4400 no problem because closing a file descriptor in a subprocess doesn't affect
4401 the parent. However, if the lookup was caching a connection to a database,
4402 the connection was closed, and the second delivery process was likely to
4403 see errors such as "PGSQL: query failed: server closed the connection
4404 unexpectedly". The problem has been fixed by closing all cached lookups
4405 in a queue runner before running a delivery process.
4407 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4408 seem to have the "const" qualifier which it has on other OS. I've
4411 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4412 *of the same type* (client or server) with the same public name that an
4413 error should be diagnosed.
4415 9. When Exim looked up a host name for an IP address, but failed to find the
4416 original IP address when looking up the host name (a safety check), it
4417 output the message "<ip address> does not match any IP for NULL", which was
4418 confusing, to say the least. The bug was that the host name should have
4419 appeared instead of "NULL".
4421 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4422 user that is built into the binary, and the -C or -D options is used, root
4423 privilege is dropped before the configuration file is read. In addition,
4424 logging is switched to stderr instead of the normal log files. If the
4425 configuration then re-defines the Exim user, the unprivileged environment
4426 is probably not what is expected, so Exim logs a panic warning message (but
4429 However, if deliver_drop_privilege is set, the unprivileged state may well
4430 be exactly what is intended, so the warning has been cut out in that case,
4431 and Exim is allowed to try to write to its normal log files.
4437 1. smtp_return_error_details was not giving details for temporary sender
4438 or receiver verification errors.
4440 2. Diagnose a configuration error if two authenticators have the same public
4443 3. Exim used not to create the message log file for a message until the first
4444 delivery attempt. This could be confusing when incoming messages were held
4445 for policy or load reasons. The message log file is now created at the time
4446 the message is received, and an initial "Received" line is written to it.
4448 4. The automatically generated man page for command line options had a minor
4449 bug that caused no ill effects; however, a more serious problem was that
4450 the procedure for building the man page automatically didn't always
4451 operate. Consequently, release 4.20 contains an out-of-date version. This
4452 shouldn't happen again.
4454 5. When building Exim with embedded Perl support, the script that builds the
4455 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4456 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4458 6. The freeze_tell option was not being used for messages that were frozen on
4459 arrival, either by an ACL or by local_scan().
4461 7. Added the smtp_incomplete_transaction log selector.
4463 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4464 was accepting AUTH without a new EHLO.
4466 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4467 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4470 10. Logging of TCP/IP connections (when configured) now happens in the main
4471 daemon process instead of the child process, so that the TCP/IP connection
4472 count is more accurate (but it can never be perfect).
4474 11. The use of "drop" in a nested ACL was not being handled correctly in the
4475 outer ACL. Now, if condition failure induced by the nested "drop" causes
4476 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4477 or "require"), the connection is dropped.
4479 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4480 that yield "discard" can now be used with an "accept" or a "discard" verb,
4481 but an error is generated for any others (because I can't see a useful way
4482 to define what should happen).
4484 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4485 are now processed in the same way as lines in the Exim configuration file.
4486 In particular, continuation lines are supported.
4488 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4490 15. Added -ti meaning -t -i.
4492 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4493 domains, and warn by logging if others are found.
4495 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4496 returned from GIFCONF when Exim is trying to find the list of interfaces on
4497 a host. The code in os.c has been modified to copy each ifreq to an aligned
4498 structure in all cases.
4500 Also, in some cases, the returned ifreq's were being copied to a 'struct
4501 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4502 means the last couple of bytes of an IPv6 address could be chopped if the
4503 ifreq contained only a normal sockaddr (14 bytes storage).
4505 18. Named domain lists were not supported in the hosts_treat_as_local option.
4506 An entry such as +xxxx was not recognized, and was treated as a literal
4509 19. Ensure that header lines added by a DATA ACL are included in the reject log
4510 if the ACL subsequently rejects the message.
4512 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4513 MD5 (which is deprecated).
4515 21. When testing a filter file using -bf, Exim was writing a message when it
4516 took the sender from a "From " line in the message, but it was not doing so
4517 when it took $return_path from a Return-Path: header line. It now does.
4519 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4520 with a valid header line field name (a series of printing characters
4521 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4523 23. Changed "disc" in the source to "disk" to conform to the documentation and
4524 the book and for uniformity.
4526 24. Ignore Sendmail's -Ooption=value command line item.
4528 25. When execve() failed while trying to run a command in a pipe transport,
4529 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4530 could be confused with a return value of 69 from the command itself. This
4531 has been changed to 127, the value the shell returns if it is asked to run
4532 a non-existent command. The wording for the related log line suggests a
4533 non-existent command as the problem.
4535 26. If received_header_text expands to an empty string, do not add a Received:
4536 header line to the message. (Well, it adds a token one on the spool, but
4537 marks it "old" so that it doesn't get used or transmitted.)
4539 27. Installed eximstats 1.28 (addition of -nt option).
4541 28. There was no check for failure on the call to getsockname() in the daemon
4542 code. This can fail if there is a shortage of resources on the system, with
4543 ENOMEM, for example. A temporary error is now given on failure.
4545 29. Contrary to the C standard, it seems that in some environments, the
4546 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4547 program. Exim now does this explicitly; it affects the formatting of
4548 timestamps using strftime().
4550 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4551 complaints. I've now initialized all the variables, to avoid this.
4553 32. Header lines added by a system filter were not being "seen" during
4554 transport-time rewrites.
4556 33. The info_callback() function passed to OpenSSL is set up with type void
4557 (*)(SSL *, int, int), as described somewhere. However, when calling the
4558 function (actually a macro) that sets it up, the type void(*)() is
4559 expected. I've put in a cast to prevent warnings from picky compilers.
4561 34. If a DNS black list lookup found a CNAME record, but there were no A
4562 records associated with the domain it pointed at, Exim crashed.
4564 35. If a DNS black list lookup returned more than one A record, Exim ignored
4565 all but the first. It now scans all returned addresses if a particular IP
4566 value is being sought. In this situation, the contents of the
4567 $dnslist_value variable are a list of all the addresses, separated by a
4570 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4571 to accept a host name and all its aliases if the forward lookup for any of
4572 them yielded the IP address of the incoming connection. Now it accepts only
4573 those names whose forward lookup yields the correct IP address. Any other
4574 names are discarded. This closes a loophole whereby a rogue DNS
4575 administrator could create reverse DNS records to break through a
4576 wildcarded host restriction in an ACL.
4578 37. If a user filter or a system filter that ran in a subprocess used any of
4579 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4580 the wrong values were passed to the pipe command ($thisaddress had the
4581 value of $0, $0 had the value of $1, etc). This bug was introduced by
4582 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4584 38. Improved the line breaking for long SMTP error messages from ACLs.
4585 Previously, if there was no break point between 40 and 75 characters, Exim
4586 left the rest of the message alone. Two changes have been made: (a) I've
4587 reduced the minimum length to 35 characters; (b) if it can't find a break
4588 point between 35 and 75 characters, it looks ahead and uses the first one
4589 that it finds. This may give the occasional overlong line, but at least the
4590 remaining text gets split now.
4592 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4593 file descriptors might be low, and that setting 1000 would always raise it.
4594 It turns out that in some environments, the limit is already over 1000 and
4595 that lowering it causes trouble. So now Exim takes care not to decrease it.
4597 40. When delivering a message, the value of $return_path is set to $sender_
4598 address at the start of routing (routers may change the value). By an
4599 oversight, this default was not being set up when an address was tested by
4600 -bt or -bv, which affected the outcome if any router or filter referred to
4603 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4604 log only when "message" or "log_message" are set. However, if one of the
4605 conditions was an address verification, or a call to a nested ACL, the
4606 messages generated by the underlying test were being passed through. This
4607 no longer happens. The underlying message is available in $acl_verify_
4608 message for both "message" and "log_message" expansions, so it can be
4609 passed through if needed.
4611 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4612 new expansion $bh_ to give the encoded byte string without charset
4613 translation. Translation happens only if iconv() is available; HAVE_ICONV
4614 indicates this at build time. HEADERS_CHARSET gives the charset to
4615 translate to; headers_charset can change it in the configuration, and
4616 "headers charset" can change it in an individual filter file.
4618 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4619 that creates RFC 2047 encoded "words" labels them as that charset instead
4620 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4621 expansion operator; (ii) when Exim creates a From: line for a local
4622 message; (iii) when a header line is rewritten to include a "phrase" part.
4624 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4625 buggy, causing it to skip the first lines of messages whose message ID
4626 ended in 'D'. This would not have bitten before Exim release 4.14, because
4627 message IDs were unlikely to end in 'D' before then. The effect was to have
4628 incorrect size information for certain domains.
4630 45. #include "config.h" was missing at the start of the crypt16.c module. This
4631 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4634 46. If there was a timeout during a "random" callout check, Exim treated it as
4635 a failure of the random address, and carried on sending RSET and the real
4636 address. If the delay was just some slowness somewhere, the response to the
4637 original RCPT would be taken as a response to RSET and so on, causing
4638 mayhem of various kinds.
4640 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4641 when I implemented it. It didn't allow for the fact that some option values
4642 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4643 the right test for positive values.
4645 48. Domain names in DNS records are case-independent. Exim always looks them up
4646 in lower case. Some resolvers return domain names in exactly the case they
4647 appear in the zone file, that is, they may contain uppercase letters. Not
4648 all resolvers do this - some return always lower case. Exim was treating a
4649 change of case by a resolver as a change of domain, similar to a widening
4650 of a domain abbreviation. This triggered its re-routing code and so it was
4651 trying to route what was effectively the same domain again. This normally
4652 caused routing to fail (because the router wouldn't handle the domain
4653 twice). Now Exim checks for this case specially, and just changes the
4654 casing of the domain that it ultimately uses when it transmits the message
4657 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4660 50. If a filter generated a file delivery with a non-absolute name (possible if
4661 no home directory exists for the router), the forbid_file option was not
4664 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4665 the existing equality matching.
4667 52. Exim was using ints instead of ino_t variables in some places where it was
4668 dealing with inode numbers.
4670 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4671 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4672 called TMPDIR, and if it finds it is different, it changes its value.
4674 54. The smtp_printf() function is now made available to local_scan() so
4675 additional output lines can be written before returning. There is also an
4676 smtp_fflush() function to enable the detection of a dropped connection.
4677 The variables smtp_input and smtp_batched_input are exported to
4680 55. Changed the default runtime configuration: the message "Unknown user"
4681 has been removed from the ACL, and instead placed on the localuser router,
4682 using the cannot_route_message feature. This means that any verification
4683 failures that generate their own messages won't get overridden. Similarly,
4684 the "Unrouteable address" message that was in the ACL for unverifiable
4685 relay addresses has also been removed.
4687 56. Added hosts_avoid_esmtp to the smtp transport.
4689 57. The exicyclog script was not checking for the esoteric option
4690 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4691 will work only if exicyclog is run under the appropriate euid.
4693 58. Following a discussion on the list, the rules by which Exim recognises line
4694 endings on incoming messages have been changed. The -dropcr and drop_cr
4695 options are now no-ops, retained only for backwards compatibility. The
4696 following line terminators are recognized: LF CRLF CR. However, special
4697 processing applies to CR:
4699 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4700 nor a local message in the state where . is a terminator.
4702 (ii) If a bare CR is encountered in a header line, an extra space is added
4703 after the line terminator so as not to end the header. The reasoning
4704 behind this is that bare CRs in header lines are most likely either
4705 to be mistakes, or people trying to play silly games.
4707 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4708 was being incorrectly given as 18 bytes larger than it should have been.
4709 This is a VOB (very old bug).
4711 60. This may never have affected anything current, but just in case it has:
4712 When the local host is found other than at the start of a list of hosts,
4713 the local host, those with the same MX, and any that follow, are discarded.
4714 When the list in question was part of a longer list of hosts, the following
4715 hosts (not currently being processed) were also being discarded. This no
4716 longer happens. I'm not sure if this situation could ever has previously
4719 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4722 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4723 header. This is something that is recommended in a new Internet Draft, and
4724 is something that is documented as being done by Sendmail. There are two
4725 possible values. For messages generated by the autoreply transport, Exim
4728 Auto-Submitted: auto-replied
4730 whereas for all other generated messages (e.g. bounces) it adds
4732 Auto-Submitted: auto-generated
4734 63. The "personal" condition in filters now includes a test for the
4735 Auto-Submitted: header. If it contains the string "auto-" the message it
4736 not considered personal.
4738 64. Added rcpt_include_affixes as a generic transport option.
4740 65. Added queue_only_override (default true).
4742 66. Added the syslog_duplication option.
4744 67. If what should have been the first header line of a message consisted of
4745 a space followed by a colon, Exim was mis-interpreting it as a header line.
4746 It isn't of course - it is syntactically invalid and should therefore be
4747 treated as the start of the message body. The misbehaviour could have
4748 caused a number of strange effects, including loss of data in subsequent
4749 header lines, and spool format errors.
4751 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4752 client host had authenticated. This control can now be exercised by an ACL
4753 for more flexibility.
4755 69. By default, callouts do not happen when testing with -bh. There is now a
4756 variant, -bhc, which does actually run the callout code, including
4757 consulting and updating the callout cache.
4759 70. Added support for saslauthd authentication, courtesy of Alexander
4762 71. If statvfs() failed on the spool or log directories while checking their
4763 size for availability, Exim confusingly gave the error "space shortage".
4764 Furthermore, in debugging mode it crashed with a floating point exception.
4765 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4766 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4767 serious problem, Exim now writes to the main and panic logs when this
4768 happens, with details of the failure. It then refuses to accept the
4769 incoming message, giving the message "spool directory problem" or "log
4770 directory problem" with a 421 code for SMTP messages.
4772 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4773 0, 1, and 2 exist, because some OS complain for execs without them (see
4774 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4775 descriptors. However, the code omitted to check that the open succeeded,
4776 causing mysterious errors if for some reason the permissions on /dev/null
4777 got screwed. Now Exim writes a message to the main and panic logs, and
4778 bombs out if it can't open /dev/null.
4780 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4781 interact so that it is all more flexible. It is supposed to remain
4782 backwards compatible. Also added extra_local_interfaces.
4784 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4785 to bomb out with an assertion failure - to the client this appears as a
4786 connection drop. This problem occurs in the part of the code that was taken
4787 from the Samba project. Fortunately, the assertion is in a very simple
4788 function, so I have fixed this by reproducing the function inline in the
4789 one place where it is called, and arranging for authentication to fail
4790 instead of killing the process with assert().
4792 75. The SPA client code was not working when the server requested OEM rather
4793 than Unicode encoding.
4795 76. Added code to make require_files with a specific uid setting more usable in
4796 the case where statting the file as root fails - usually a non-root-mounted
4797 NFS file system. When this happens and the failure is EACCES, Exim now
4798 forks a subprocess and does the per-uid checking as the relevant uid.
4800 77. Added process_log_path.
4802 78. If log_file_path was not explicitly set, a setting of check_log_space or
4803 check_log_inodes was ignored.
4805 79. If a space check for the spool or log partitions fails, the incident is now
4806 logged. Of course, in the latter case the data may get lost...
4808 80. Added the %p formatting code to string_format() so that it can be used to
4809 print addresses in debug_print(). Adjusted all the address printing in the
4810 debugging in store.c to use %p rather than %d.
4812 81. There was a concern that a line of code in smtp_in.c could overflow a
4813 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4814 initially expressed, the concern was not well-founded, because trailing
4815 spaces are removed early. However, if the trailing spaces were followed by
4816 a NULL, they did not get removed, so the overflow was possible. Two fixes
4819 (a) I re-wrote the offending code in a cleaner fashion.
4820 (b) If an incoming SMTP command contains a NULL character, it is rejected
4823 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4824 runs initgroups(), so that if the Exim user is in any additional groups,
4825 they will be used during message reception.
4831 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob