1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.523 2007/08/22 10:10:23 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.
91 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
92 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
95 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
98 PH/02 In an ACL statement such as
100 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
102 if a client was not listed at all, or was listed with a value other than
103 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
104 the condition was not true (as it should be), so access was not denied.
105 The bug was that the ! inversion was incorrectly passed on to the second
106 item. This has been fixed.
108 PH/03 Added additional dnslists conditions == and =& which are different from
109 = and & when the dns lookup returns more than one IP address.
111 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
112 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
114 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
115 FSYNC, which compiles an option called disable_fsync that allows for
116 bypassing fsync(). The documentation is heavily laced with warnings.
118 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
120 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
121 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
122 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
123 including adding "make clean"; (3) Added -fPIC when compiling the test
124 dynamically loaded module, to get rid of a warning.
126 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
127 message fails, move_frozen_messages = true and ignore_bounce_errors_after
128 = 0s. The bug is otherwise harmless.
130 PH/07 There was a bug in the dovecot authenticator such that the value of
131 $auth1 could be overwritten, and so not correctly preserved, after a
132 successful authentication. This usually meant that the value preserved by
133 the server_setid option was incorrect.
135 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
137 PH/09 Installed PCRE release 7.0.
139 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
140 run for batched SMTP input. It is now run at the start of every message
141 in the batch. While fixing this I discovered that the process information
142 (output by running exiwhat) was not always getting set for -bs and -bS
143 input. This is fixed, and it now also says "batched" for BSMTP.
145 PH/11 Added control=no_pipelining.
147 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
148 patch, slightly modified), and move the expansion of helo_data till after
149 the connection is made in the smtp transport (so it can use these
152 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
154 PH/14 Added log_selector = +pid.
156 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
158 PH/16 Add ${if forany and ${if forall.
160 PH/17 Added dsn_from option to vary the From: line in DSNs.
162 PH/18 Flush SMTP output before performing a callout, unless control =
163 no_callout_flush is set.
165 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
166 was true (the default) a successful delivery failed to delete the retry
167 item, thus causing premature timeout of the address. The bug is now
170 PH/20 Added hosts_avoid_pipelining to the smtp transport.
172 PH/21 Long custom messages for fakedefer and fakereject are now split up
173 into multiline reponses in the same way that messages for "deny" and
174 other ACL rejections are.
176 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
177 with slight modification.
179 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
180 draft, changing the syntax and factoring some duplicate code".
182 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
183 for deliveries of the second and subsequent messages over the same SMTP
186 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
187 ${reduce, with only minor "tidies".
189 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
191 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
192 expansion side effects.
194 PH/27 When a message times out after an over-quota error from an Exim-imposed
195 quota, the bounce message says "mailbox is full". This message was not
196 being given when it was a system quota that was exceeded. It now should
199 MH/03 Made $recipients available in local_scan(). local_scan() already has
200 better access to the recipient list through recipients_list[], but
201 $recipients can be useful in postmaster-provided expansion strings.
203 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
204 in the case of a MAIL command with additional options following the
205 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
206 were accidentally chopped off.
208 PH/29 SMTP synchronization checks are implemented when a command is read -
209 there is a check that no more input is waiting when there shouldn't be
210 any. However, for some commands, a delay in an ACL can mean that it is
211 some time before the response is written. In this time, more input might
212 arrive, invalidly. So now there are extra checks after an ACL has run for
213 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
214 pipelining has not been advertised.
216 PH/30 MH's patch to allow iscntrl() characters to be list separators.
218 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
219 returned in the SMTP response when smtp_return_error_details was false.
222 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
223 instead of the C I/O that was originally supplied, because problems were
226 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
227 Exim which did not show up earlier: it was assuming that a call to
228 SSL_CTX_set_info_callback() might give an error value. In fact, there is
229 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
230 was a macro that became an assignment, so it seemed to work. This has
231 changed to a proper function call with a void return, hence the compile
232 error. Exim's code has been fixed.
234 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
237 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
239 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
241 PH/37 If a message is not accepted after it has had an id assigned (e.g.
242 because it turns out to be too big or there is a timeout) there is no
243 "Completed" line in the log. When some messages of this type were
244 selected by exigrep, they were listed as "not completed". Others were
245 picked up by some special patterns. I have improved the selection
246 criteria to be more general.
248 PH/38 The host_find_failed option in the manualroute router can now be set
249 to "ignore", to completely ignore a host whose IP address cannot be
250 found. If all hosts are ignored, the behaviour is controlled by the new
251 host_all_ignored option.
253 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
254 homing or because of multiple MX records with /mx) generated more than
255 one IP address, and the following item turned out to be the local host,
256 all the secondary addresses of the first item were incorrectly removed
257 from the list, along with the local host and any following hosts (which
258 is what is supposed to happen).
260 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
261 whoever called Exim into the -H file. In the case of the daemon it was
262 behaving confusingly. When first started, it used values for whoever
263 started the daemon, but after a SIGHUP it used the Exim user (because it
264 calls itself on a restart). I have changed the code so that it now always
267 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
268 message are rejected with the same error (e.g. no authentication or bad
269 sender address), and a DATA command is nevertheless sent (as can happen
270 with PIPELINING or a stupid MUA), the error message that was given to the
271 RCPT commands is included in the rejection of the DATA command. This is
272 intended to be helpful for MUAs that show only the final error to their
275 PH/42 Another patch from the Sieve maintainer.
277 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
278 Eximstats - Fixed some broken HTML links and added missing column headers
280 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
281 columns for Rejects, Temp Rejects, Ham, and Spam rows.
283 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
285 PH/43 Yet another patch from the Sieve maintainer.
287 PH/44 I found a way to check for a TCP/IP connection going away before sending
288 the response to the final '.' that terminates a message, but only in the
289 case where the client has not sent further data following the '.'
290 (unfortunately, this is allowed). However, in many cases there won't be
291 any further data because there won't be any more messages to send. A call
292 to select() can be used: if it shows that the input is "ready", there is
293 either input waiting, or the socket has been closed. An attempt to read
294 the next input character can distinguish the two cases. Previously, Exim
295 would have sent an OK response which the client would never have see.
296 This could lead to message repetition. This fix should cure that, at
297 least in a lot of common cases.
299 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
300 advertised in response to EHLO.
306 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
307 fixed by 4.65/MH/01 (is this a record?) are fixed:
309 (i) An empty string was always treated as zero by the numeric comparison
310 operators. This behaviour has been restored.
312 (ii) It is documented that the numeric comparison operators always treat
313 their arguments as decimal numbers. This was broken in that numbers
314 starting with 0 were being interpreted as octal.
316 While fixing these problems I realized that there was another issue that
317 hadn't been noticed. Values of message_size_limit (both the global option
318 and the transport option) were treated as octal if they started with 0.
319 The documentation was vague. These values are now always treated as
320 decimal, and I will make that clear in the documentation.
326 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
327 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
330 MH/01 Don't check that the operands of numeric comparison operators are
331 integers when their expansion is in "skipping" mode (fixes bug
332 introduced by 4.64-PH/07).
334 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
335 child addresses, Exim now panics and dies. Previously, because the count
336 is held in a short int, deliveries were likely to be lost. As such a
337 large number of recipients for a single message is ridiculous
338 (performance will be very, very poor), I have chosen to impose a limit
339 rather than extend the field.
345 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
346 leftover -K file (the existence of which was triggered by #402).
347 While we were at it, introduced process PID as part of the -K
348 filename. This should rule out race conditions when creating
351 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
352 processing considerably. Previous code took too long for large mails,
353 triggering a timeout which in turn triggers #401.
355 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
356 in the DK code in transports.c. sendfile() is not really portable,
357 hence the _LINUX specificness.
359 TF/01 In the add_headers option to the mail command in an Exim filter,
360 there was a bug that Exim would claim a syntax error in any
361 header after the first one which had an odd number of characters
364 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
365 callout verification, Exim cached a "reject" for the entire domain. This
366 is correct for most verifications, but it is not correct for a recipient
367 verification with use_sender or use_postmaster set, because in that case
368 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
369 case of MAIL FROM:<> rejection from other early rejections (e.g.
370 rejection of HELO). When verifying a recipient using a non-null MAIL
371 address, the cache is ignored if it shows MAIL FROM:<> rejection.
372 Whatever the result of the callout, the value of the domain cache is
373 left unchanged (for any other kind of callout, getting as far as trying
374 RCPT means that the domain itself is ok).
376 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
379 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
380 manifest itself as EPIPE rather than ECONNECT. When tidying away a
381 session, the daemon ignores ECONNECT errors and logs others; it now
382 ignores EPIPE as well.
384 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
385 (quoted-printable decoding).
387 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
388 later the small subsequent patch to fix an introduced bug.
390 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
392 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
394 PH/08 An error is now given if message_size_limit is specified negative.
396 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
397 to be given (somewhat) arbitrary names.
399 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
402 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
403 miscellaneous code fixes
405 PH/10 Added the log_reject_target ACL modifier to specify where to log
408 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
409 hostname. This is wrong, because it relates to the incoming message (and
410 probably the interface on which it is arriving) and not to the outgoing
411 callout (which could be using a different interface). This has been
412 changed to use the value of the helo_data option from the smtp transport
413 instead - this is what is used when a message is actually being sent. If
414 there is no remote transport (possible with a router that sets up host
415 addresses), $smtp_active_hostname is used.
417 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
418 tweaks were necessary in order to get it to work (see also 21 below):
419 (a) The code assumed that strncpy() returns a negative number on buffer
420 overflow, which isn't the case. Replaced with Exim's string_format()
422 (b) There were several signed/unsigned issues. I just did the minimum
423 hacking in of casts. There is scope for a larger refactoring.
424 (c) The code used strcasecmp() which is not a standard C function.
425 Replaced with Exim's strcmpic() function.
426 (d) The code set only $1; it now sets $auth1 as well.
427 (e) A simple test gave the error "authentication client didn't specify
428 service in request". It would seem that Dovecot has changed its
429 interface. Fortunately there's a specification; I followed it and
430 changed what the client sends and it appears to be working now.
432 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
435 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
436 address A is aliased to B and C, where B exists and C does not. Without
437 -v the output is "A verified" because verification stops after a
438 successful redirection if more than one address is generated. However,
439 with -v the child addresses are also verified. Exim was outputting "A
440 failed to verify" and then showing the successful verification for C,
441 with its parentage. It now outputs "B failed to verify", showing B's
442 parentage before showing the successful verification of C.
444 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
445 look up a TXT record in a specific list after matching in a combined
448 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
449 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
450 they consult the DNS. I had assumed they would set it the way they
451 wanted; and indeed my experiments on Linux seem to show that in some
452 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
453 To be on the safe side, however, I have now made the interface to
454 host_find_byname() similar to host_find_bydns(), with an argument
455 containing the DNS resolver options. The host_find_byname() function now
456 sets these options at its start, just as host_find_bydns() does. The smtp
457 transport options dns_qualify_single and dns_search_parents are passed to
458 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
459 of host_find_byname() use the default settings of RES_DEFNAMES
460 (qualify_single) but not RES_DNSRCH (search_parents).
462 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
463 spool_read_header() do less string testing, by means of a preliminary
464 switch on the second character of optional "-foo" lines. (This is
465 overdue, caused by the large number of possibilities that now exist.
466 Originally there were few.) While I was there, I also converted the
467 str(n)cmp tests so they don't re-test the leading "-" and the first
468 character, in the hope this might squeeze out yet more improvement.
470 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
471 flag allowing group syntax was set by the header_syntax check but not
472 turned off, possible causing trouble later; (2) The flag was not being
473 set at all for the header_verify test, causing "group"-style headers to
474 be rejected. I have now set it in this case, and also caused header_
475 verify to ignore an empty address taken from a group. While doing this, I
476 came across some other cases where the code for allowing group syntax
477 while scanning a header line wasn't quite right (mostly, not resetting
478 the flag correctly in the right place). These bugs could have caused
479 trouble for malformed header lines. I hope it is now all correct.
481 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
482 with the "reply" argument non-NULL. The code, however (which originally
483 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
484 but it didn't always do it. This confused somebody who was copying the
485 code for some other use. I have removed all the tests.
487 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
488 feature that was used to support insecure browsers during the U.S. crypto
489 embargo. It requires special client support, and Exim is probably the
490 only MTA that supported it -- and would never use it because real RSA is
491 always available. This code has been removed, because it had the bad
492 effect of slowing Exim down by computing (never used) parameters for the
493 RSA_EXPORT functionality.
495 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
496 authenticator to fail if there's a tab character in the incoming data
497 (there should never be unless someone is messing about, as it's supposed
498 to be base64-encoded). Also added, on Timo's advice, the "secured" option
499 if the connection is using TLS or if the remote IP is the same as the
500 local IP, and the "valid-client-cert option" if a client certificate has
503 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
504 authenticators. This can be used for authorization after authentication
505 succeeds. (In the case of plaintext, it servers for both authentication
508 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
509 if any retry times were supplied.
511 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
512 connection, where there is no client IP address to check. In this
513 situation, the verify now always succeeds.
515 PH/25 Applied John Jetmore's -Mset patch.
517 PH/26 Added -bem to be like -Mset, but loading a message from a file.
519 PH/27 In a string expansion for a processed (not raw) header when multiple
520 headers of the same name were present, leading whitespace was being
521 removed from all of them, but trailing whitespace was being removed only
522 from the last one. Now trailing whitespace is removed from each header
523 before concatenation. Completely empty headers in a concatenation (as
526 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
527 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
529 PH/29 [Removed. This was a change that I later backed out, and forgot to
530 correct the ChangeLog entry (that I had efficiently created) before
531 committing the later change.]
533 PH/30 Exim was sometimes attempting to deliver messages that had suffered
534 address errors (4xx response to RCPT) over the same connection as other
535 messages routed to the same hosts. Such deliveries are always "forced",
536 so retry times are not inspected. This resulted in far too many retries
537 for the affected addresses. The effect occurred only when there were more
538 hosts than the hosts_max_try setting in the smtp transport when it had
539 the 4xx errors. Those hosts that it had tried were not added to the list
540 of hosts for which the message was waiting, so if all were tried, there
541 was no problem. Two fixes have been applied:
543 (i) If there are any address or message errors in an SMTP delivery, none
544 of the hosts (tried or untried) are now added to the list of hosts
545 for which the message is waiting, so the message should not be a
546 candidate for sending over the same connection that was used for a
547 successful delivery of some other message. This seems entirely
548 reasonable: after all the message is NOT "waiting for some host".
549 This is so "obvious" that I'm not sure why it wasn't done
550 previously. Hope I haven't missed anything, but it can't do any
551 harm, as the worst effect is to miss an optimization.
553 (ii) If, despite (i), such a delivery is accidentally attempted, the
554 routing retry time is respected, so at least it doesn't keep
555 hammering the server.
557 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
558 in ${readsocket because some servers need this prod.
560 PH/32 Added some extra debug output when updating a wait-xxx database.
562 PH/33 The hint "could be header name not terminated by colon", which has been
563 given for certain expansion errors for a long time, was not being given
564 for the ${if def:h_colon_omitted{... case.
566 PH/34 The spec says: "With one important exception, whenever a domain list is
567 being scanned, $domain contains the subject domain." There was at least
568 one case where this was not true.
570 PH/35 The error "getsockname() failed: connection reset by peer" was being
571 written to the panic log as well as the main log, but it isn't really
572 panic-worthy as it just means the connection died rather early on. I have
573 removed the panic log writing for the ECONNRESET error when getsockname()
576 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
577 runs only) independently of the message's sender address. This meant
578 that, if the 4xx error was in fact related to the sender, a different
579 message to the same recipient with a different sender could confuse
580 things. In particualar, this can happen when sending to a greylisting
581 server, but other circumstances could also provoke similar problems.
582 I have changed the default so that the retry time for these errors is now
583 based a combination of the sender and recipient addresses. This change
584 can be overridden by setting address_retry_include_sender=false in the
587 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
588 remote server are returned as part of bounce messages. This was not
589 happening for LMTP over a pipe (the lmtp transport), but now it is the
590 same for both kinds of LMTP.
592 PH/38 Despite being documented as not happening, Exim was rewriting addresses
593 in header lines that were in fact CNAMEs. This is no longer the case.
595 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
596 and queue runs started by the daemon processed all messages. This has
597 been fixed so that -R and -S can now usefully be given with -q<time>.
599 PH/40 Import PCRE release 6.7 (fixes some bugs).
601 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
603 PH/42 Give an error if -q is specified more than once.
605 PH/43 Renamed the variables $interface_address and $interface_port as
606 $received_ip_address and $received_port, to make it clear that these
607 values apply to message reception, and not to the outgoing interface when
608 a message is delivered. (The old names remain recognized, of course.)
610 PH/44 There was no timeout on the connect() call when using a Unix domain
611 socket in the ${readsocket expansion. There now is.
613 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
614 be meaningful with "accept".
616 SC/01 Eximstats V1.43
617 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
619 SC/02 Eximstats V1.44
620 Use a glob alias rather than an array ref in the generated
621 parser. This improves both readability and performance.
623 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
624 Collect SpamAssassin and rejection statistics.
625 Don't display local sender or destination tables unless
626 there is data to show.
627 Added average volumes into the top table text output.
629 SC/04 Eximstats V1.46
630 Collect data on the number of addresses (recipients)
631 as well as the number of messages.
633 SC/05 Eximstats V1.47
634 Added 'Message too big' to the list of mail rejection
635 reasons (thanks to Marco Gaiarin).
637 SC/06 Eximstats V1.48
638 Mainlog lines which have GMT offsets and are too short to
639 have a flag are now skipped.
641 SC/07 Eximstats V1.49 (Alain Williams)
642 Added the -emptyok flag.
644 SC/08 Eximstats V1.50
645 Fixes for obtaining the IP address from reject messages.
647 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
648 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
649 whitesspace changes from 4.64-PH/27
651 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
654 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
655 are found, allow negative numbers in numeric criteria)
657 JJ/06 exipick.20061117.2, added new $message_body_missing variable
659 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
660 to match changes made in 4.64-PH/43
662 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
664 PH/47 Put in an explicit test for a DNS lookup of an address record where the
665 "domain" is actually an IP address, and force a failure. This locks out
666 those revolvers/nameservers that support "A-for-A" lookups, in
667 contravention of the specifications.
669 PH/48 When a host name was looked up from an IP address, and the subsequent
670 forward lookup of the name timed out, the host name was left in
671 $sender_host_name, contrary to the specification.
673 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
674 restricted to single-key lookups, Exim was not diagnosing an error if
675 * or *@ was used with a query-style lookup.
677 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
679 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
680 long ago, but noone interested enough thought of it. Let's just say that
681 the "1.1" means that there are some new functions that weren't there at
682 some point in the past.
684 PH/51 Error processing for expansion failure of helo_data from an smtp
685 transport during callout processing was broken.
687 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
688 tested/used via the -bh/-bhc/-bs options.
690 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
691 bug, fixed in subsequent PCRE releases).
693 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
694 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
696 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
702 SC/01 Use a glob alias rather than an array ref in eximstats generated
703 parser. This improves both readability and performance.
705 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
706 Don't display local sender or destination tables in eximstats unless
707 there is data to show.
708 Added average volumes into the eximstats top table text output.
710 SC/03 Collect data on the number of addresses (recipients) as well
711 as the number of messages in eximstats.
713 TF/01 Correct an error in the documentation for the redirect router. Exim
714 does (usually) call initgroups() when daemonizing.
716 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
717 with consistent privilege compared to when running as a daemon.
719 TF/03 Note in the spec that $authenticated_id is not set for local
720 submissions from trusted users.
722 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
723 Thanks to Dean Brooks <dean@iglou.com> for the patch.
725 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
726 by adding some example configuration directives to the default
727 configuration file. A little bit of work is required to uncomment the
728 directives and define how usernames and passwords are checked, but
729 there is now a framework to start from.
731 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
732 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
733 without this. I don't know how relevant this is to other LDAP libraries.
735 PH/02 Add the verb name to the "unknown ACL verb" error.
737 PH/03 Magnus Holmgren's patch for filter_prepend_home.
739 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
741 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
742 directory not expanded when it should be if an expanded home directory
743 was set for the address (which is overridden by the transport).
745 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
748 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
749 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
750 because it is too late at that time, and has no effect.
752 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
753 security issue with \' (bugzilla #107). I could not use the
754 PQescapeStringConn() function, because it needs a PGconn value as one of
757 PH/08 When testing addresses using -bt, indicate those final addresses that
758 are duplicates that would not cause an additional delivery. At least one
759 person was confused, thinking that -bt output corresponded to deliveries.
760 (Suppressing duplicates isn't a good idea as you lose the information
761 about possibly different redirections that led to the duplicates.)
763 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
764 systems where poll() doesn't work, in particular OS X.
766 PH/10 Added more information to debugging output for retry time not reached.
768 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
769 operations in malware.c.
771 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
774 PH/13 If write_rejectlog was set false when logging was sent to syslog with
775 syslog_duplication set false, log lines that would normally be written
776 both the the main log and to the reject log were not written to syslog at
779 PH/14 In the default configuration, change the use of "message" in ACL warn
780 statements to "add_header".
782 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
783 not followed by a command (e.g. "seen endif").
785 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
786 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
789 PH/17 Added extra conditions to the default value of delay_warning_condition
793 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
794 { match{$h_precedence:}{(?i)bulk|list|junk} } \
795 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
798 The Auto-Submitted: and various List- headers are standardised, whereas I
799 don't think Precedence: ever was.
801 PH/18 Refactored debugging code in route_finduser() to show more information,
802 in particular, the error code if getpwnam() issues one.
804 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
805 This is apparently needed in addition to the PH/07 change above to avoid
806 any possible encoding problems.
808 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
809 but not after initializing Perl.
811 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
812 output them only if debugging. By default they are written stderr,
813 apparently, which is not desirable.
815 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
818 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
821 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
823 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
824 authenticated or an ident call has been made. Suppress the default
825 values for $authenticated_id and $authenticated_sender (but permit -oMai
826 and -oMas) when testing with -bh.
828 PH/24 Re-jigged the order of the tests in the default configuration so that the
829 tests for valid domains and recipients precede the DNS black list and CSA
830 tests, on the grounds that those ones are more expensive.
832 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
833 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
834 This bug exists in every version of Exim that I still have, right back to
837 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
838 However, an attempt to turn on case-sensitivity in a regex key by
839 including (?-i) didn't work because the subject string was already
840 lowercased, and the effects were non-intuitive. It turns out that a
841 one-line patch can be used to allow (?-i) to work as expected.
847 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
848 other effects) broke the use of negated acl sub-conditions.
850 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
853 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
854 "Deny" causes Exim to reject the incoming connection with a 554 error.
855 Unfortunately, if there is a major crisis, such as a disk failure,
856 tcp-wrappers gives "deny", whereas what one would like would be some
857 kind of temporary error. A kludge has been added to help with this.
858 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
859 554 error is used if errno is still zero or contains ENOENT (which occurs
860 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
863 PH/03 Add -lutil to the default FreeBSD LIBS setting.
865 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
866 errors. Otherwise a message that provokes a temporary error (when other
867 messages do not) can cause a whole host to time out.
869 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
870 addresses were routed directly to files or pipes from a redirect router.
871 File deliveries just didn't batch; pipe deliveries might have suffered
874 PH/06 A failure to get a lock for a hints database would erroneously always say
875 "Failed to get write lock", even when it was really a read lock.
877 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
878 of 0600. This has been changed to use the value of the lockfile_mode
879 option (which defaults to 0600).
881 PH/08 Applied small patch from the Sieve maintainer.
883 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
884 folder from quota calculations, a direct delivery into this folder messed
885 up the contents of the maildirsize file. This was because the regex was
886 used only to exclude .Trash (or whatever) when the size of the mailbox
887 was calculated. There was no check that a delivery was happening into an
888 excluded directory. This bug has been fixed by ignoring all quota
889 processing for deliveries into excluded directories.
891 PH/10 Added the maildirfolder_create_regex option to appendfile.
897 PH/01 The code for finding all the local interface addresses on a FreeBSD
898 system running IPv6 was broken. This may well have applied to all BSD
899 systems, as well as to others that have similar system calls. The broken
900 code found IPv4 interfaces correctly, but gave incorrect values for the
901 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
902 that it would not match correctly against @[] and not recognize the IPv6
905 PH/02 The ipliteral router was not recognizing addresses of the form user@
906 [ipv6:....] because it didn't know about the "ipv6:" prefix.
908 PH/03 Added disable_ipv6.
910 PH/04 Changed $reply_address to use the raw form of the headers instead of the
911 decoded form, because it is most often used to construct To: headers
912 lines in autoreplies, and the decoded form may well be syntactically
913 invalid. However, $reply_address has leading white space removed, and all
914 newlines turned into spaces so that the autoreply transport does not
917 PH/05 If group was specified without a user on a router, and no group or user
918 was specified on a transport, the group from the router was ignored.
920 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
921 for visible compile-time settings that can be used to change these
922 numbers, for those that want even more. Backwards compatibility with old
923 spool files has been maintained. However, going back to a previous Exim
924 release will lost any variables that are in spool files.
926 PH/07 Two small changes when running in the test harness: increase delay when
927 passing a TCP/IP connection to a new process, in case the original
928 process has to generate a bounce, and remove special handling of
929 127.0.0.2 (sic), which is no longer necessary.
931 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
932 be the same on different OS.
934 PH/09 Moved a debug statement in filter processing to avoid a race problem when
937 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
938 whether --show-vars was specified or not
940 JJ/02 exipick: Added support for new ACL variable spool format introduced
943 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
944 syntactically invalid From: or Reply-to: line, and a filter used this to
945 generate an autoreply, and therefore failed to obtain an address for the
946 autoreply, Exim could try to deliver to a non-existent relative file
947 name, causing unrelated and misleading errors. What now happens is that
948 it logs this as a hard delivery error, but does not attempt to create a
951 PH/11 The exinext utility has a -C option for testing purposes, but although
952 the given file was scanned by exinext itself; it wasn't being passed on
953 when Exim was called.
955 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
956 an end-of-file indication when reading a command response.
958 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
959 compiled. In many other places in Exim, IPv6 addresses are always
960 recognized, so I have changed this. It also means that IPv4 domain
961 literals of the form [IPV4:n.n.n.n] are now always recognized.
963 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
964 used if the router is not running as root, for example, when verifying at
965 ACL time, or when using -bh. The debugging output from this situation was
966 non-existent - all you got was a failure to exec. I have made two
969 (a) Failures to set uid/gid, the current directory, or a process leader
970 in a subprocess such as that created by queryprogram now generate
971 suitable debugging ouput when -d is set.
973 (b) The queryprogram router detects when it is not running as root,
974 outputs suitable debugging information if -d is set, and then runs
975 the subprocess without attempting to change uid/gid.
977 PH/15 Minor change to Makefile for building test_host (undocumented testing
980 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
981 additional section of a DNS packet that returns MX or SRV records.
982 Instead, it always explicitly searches for A/AAAA records. This avoids
983 major problems that occur when a DNS server includes only records of one
984 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
985 fixed another bug: if SRV records were looked up and the corresponding
986 address records were *not* found in the additional section, the port
987 values from the SRV records were lost.
989 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
990 using the correct key (the original address) when searching the retry
991 rules in order to find which one to use for generating the retry hint.
993 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
994 adding the default one. Similarly, if it contains a Reply-To: header, the
995 errors_reply_to option, if set, is not used.
997 PH/19 When calculating a retry time, Exim used to measure the "time since
998 failure" by looking at the "first failed" field in the retry record. Now
999 it does not use this if it is later than than the arrival time of the
1000 message. Instead it uses the arrival time. This makes for better
1001 behaviour in cases where some deliveries succeed, thus re-setting the
1002 "first failed" field. An example is a quota failure for a huge message
1003 when small messages continue to be delivered. Without this change, the
1004 "time since failure" will always be short, possible causing more frequent
1005 delivery attempts for the huge message than are intended.
1006 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1008 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1009 $1, $2, $3) because the numerical variables can be reset during some
1010 expansion items (e.g. "match"), thereby losing the authentication data.
1012 PH/21 Make -bV show the size of off_t variables so that the test suite can
1013 decide whether to run tests for quotas > 2G.
1015 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1016 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1017 filecount value is greater than 2G or if a quota value is greater than 2G
1018 on a system where the size of off_t is not greater than 4, a panic error
1021 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1022 never match. The debug and -bh output now contains an explicit error
1023 message indicating a malformed IPv4 address or mask.
1025 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1026 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1027 PH/23 above applies.
1029 PH/25 Do not write to syslog when running in the test harness. The only
1030 occasion when this arises is a failure to open the main or panic logs
1031 (for which there is an explicit test).
1033 PH/26 Added the /no_tell option to "control=freeze".
1035 PH/27 If a host name lookup failed very early in a connection, for example, if
1036 the IP address matched host_lookup and the reverse lookup yielded a name
1037 that did not have a forward lookup, an error message of the form "no IP
1038 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1039 could be logged. Now it outputs the IP address instead of "NULL".
1041 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1042 allows the sender and the authenticated sender to be set when
1043 submitting a message from within Exim. Since child_open_exim() is
1044 documented for local_scan(), the new function should be too.
1046 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1047 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1048 results in an empty string is now treated as unset.
1050 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1052 PH/31 Added sender_verify_fail logging option.
1054 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1055 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1056 tidied the source and removed it altogether.
1058 PH/33 When a queue run was abandoned because the load average was too high, a
1059 log line was always written; now it is written only if the queue_run log
1060 selector is set. In addition, the log line for abandonment now contains
1061 information about the queue run such as the pid. This is always present
1062 in "start" and "stop" lines but was omitted from the "abandon" line.
1064 PH/34 Omit spaces between a header name and the colon in the error message that
1065 is given when verify = headers_syntax fails (if there are lots of them,
1066 the message gets confusing).
1068 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1069 names, as there are now some PTR records that contain slashes. This check
1070 is only to protect against broken name servers that fall over on strange
1071 characters, so the fact that it applies to all lookups doesn't matter.
1073 PH/36 Now that the new test suite is complete, we can remove some of the
1074 special code in Exim that was needed for the old test suite. For example,
1075 sorting DNS records because real resolvers return them in an arbitrary
1076 order. The new test suite's fake resolver always returns records in the
1079 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1080 bounces) except when queue_only is set, to avoid logging races between
1081 the different processes.
1083 PH/38 Panic-die if .include specifies a non-absolute path.
1085 PH/39 A tweak to the "H" retry rule from its user.
1087 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1088 a label. They prevented compilation on older perls.
1090 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1091 a warning to be raised on newish perls.
1093 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1094 on queue. Changes to match documented behaviour of showing count of
1095 messages matching specified criteria.
1097 PH/40 Changed the default ident timeout from 30s to 5s.
1099 PH/41 Added support for the use of login_cap features, on those BSD systems
1100 that have them, for controlling the resources used by pipe deliveries.
1102 PH/42 The content-scanning code uses fopen() to create files in which to put
1103 message data. Previously it was not paying any attention to the mode of
1104 the files. Exim runs with umask(0) because the rest of the code creates
1105 files with open(), and sets the required mode explicitly. Thus, these
1106 files were ending up world-writeable. This was not a big issue, because,
1107 being within the spool directory, they were not world-accessible. I have
1108 created a function called modefopen, which takes an additional mode
1109 argument. It sets umask(777), creates the file, chmods it to the required
1110 mode, then resets the umask. All the relevant calls to fopen() in the
1111 content scanning code have been changed to use this function.
1113 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1114 to 24 hours. This avoids potential overflow problems when processing G
1115 and H retry rules. I suspect nobody ever tinkers with this value.
1117 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1119 PH/45 When the plaintext authenticator is running as a client, the server's
1120 challenges are checked to ensure they are valid base64 strings. By
1121 default, the authentication attempt is cancelled if an invalid string is
1122 received. Setting client_ignore_invalid_base64 true ignores these errors.
1123 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1124 they are received. Thus, the responses can be made to depend on the
1125 challenges. If an invalid string is ignored, an empty string is placed in
1128 PH/46 Messages that are created by the autoreply transport now contains a
1129 References: header, in accordance with RFCs 2822 and 3834.
1131 PH/47 Added authenticated_sender_force to the smtp transport.
1133 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1135 PH/49 Installed latest patch from the Sieve maintainer.
1137 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1138 was also set, the appendfile transport was unnecessarily scanning a
1139 directory of message files (e.g. for maildir delivery) to find the count
1140 of files (along with the size), even though it did not need this
1141 information. It now does the scan only if it needs to find either the
1142 size of the count of files.
1144 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1146 PH/52 Two bugs concerned with error handling when the smtp transport is
1149 (i) Exim was not creating retry information for temporary errors given
1150 for individual recipients after the DATA command when the smtp transport
1151 was used in LMTP mode. This meant that they could be retried too
1152 frequently, and not timed out correctly.
1154 (ii) Exim was setting the flag that allows error details to be returned
1155 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1156 recipients that were returned after the DATA command.
1158 PH/53 This is related to PH/52, but is more general: for any failing address,
1159 when detailed error information was permitted to be returned to the
1160 sender, but the error was temporary, then after the final timeout, only
1161 "retry timeout exceeded" was returned. Now it returns the full error as
1162 well as "retry timeout exceeded".
1164 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1165 do this, and (what is worse) MTAs that accept it.
1167 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1168 will now be deprecated.
1170 PH/56 New os.c-cygwin from the Cygwin maintainer.
1172 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1173 formats (previously only available in exim formats via -bpr, -bpru,
1174 and -bpra. Now also available in native and exiqgrep formats)
1176 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1177 with very large, slow to parse queues
1179 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1181 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1183 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1184 responses to authentication challenges, though it was showing the
1185 challenges; (ii) I've removed the CR characters from the debug output for
1188 PH/58 Allow for the insertion of a newline as well as a space when a string
1189 is turned into more than one encoded-word during RFC 2047 encoding. The
1190 Sieve code now uses this.
1192 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1193 data_4xx, lost_connection, tls_required.
1195 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1196 message was being sent as an SMTP response.
1198 PH/61 Add -l and -k options to exicyclog.
1200 PH/62 When verifying, if an address was redirected to one new address, so that
1201 verification continued, and the new address failed or deferred after
1202 having set something in $address_data, the value of $address_data was not
1203 passed back to the ACL. This was different to the case when no
1204 redirection occurred. The value is now passed back in both cases.
1206 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1207 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1208 use login_cap.h, so on its own it isn't the distinguishing feature. The
1209 new name refers directly to the setclassresources() function.
1211 PH/65 Added configuration files for NetBSD3.
1213 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1215 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1216 is preferred over IPv4.
1218 PH/68 The bounce_return_message and bounce_return_body options were not being
1219 honoured for bounces generated during the reception of non-SMTP messages.
1220 In particular, this applied to messages rejected by the ACL. This bug has
1221 been fixed. However, if bounce_return_message is true and bounce_return_
1222 body is false, the headers that are returned for a non-SMTP message
1223 include only those that have been read before the error was detected.
1224 (In the case of an ACL rejection, they have all been read.)
1226 PH/69 The HTML version of the specification is now built in a directory called
1227 spec_html instead of spec.html, because the latter looks like a path with
1228 a MIME-type, and this confuses some software.
1230 PH/70 Catch two compiler warnings in sieve.c.
1232 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1233 function verify_get_ident() calls ip_connect() to connect a socket, but
1234 if the "connect()" function timed out, ip_connect() used to close the
1235 socket. However, verify_get_ident() also closes the socket later, and in
1236 between Exim writes to the log, which may get opened at this point. When
1237 the socket was closed in ip_connect(), the log could get the same file
1238 descriptor number as the socket. This naturally causes chaos. The fix is
1239 not to close the socket in ip_connect(); the socket should be closed by
1240 the function that creates it. There was only one place in the code where
1241 this was missing, in the iplookup router, which I don't think anybody now
1242 uses, but I've fixed it anyway.
1244 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1245 well as to direct DNS lookups. Otherwise the handling of names in host
1246 lists is inconsistent and therefore confusing.
1252 PH/01 Two changes to the default runtime configuration:
1254 (1) Move the checks for relay_from_hosts and authenticated clients from
1255 after to before the (commented out) DNS black list checks.
1257 (2) Add control=submission to the relay_from_hosts and authenticated
1258 clients checks, on the grounds that messages accepted by these
1259 statements are most likely to be submissions.
1261 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1263 (1) Generate an error if the third argument for the ${prvs expansion is
1266 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1269 (3) Reset the variables that are obtained from the first argument of
1270 ${prvscheck and used in the second argument before leaving the code,
1271 because their memory is reclaimed, so using them afterwards may do
1274 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1275 one (it's much easier than Tom thought :-).
1277 (5) Because of (4), we can now allow for the use of $prvscheck_result
1278 inside the third argument.
1280 PH/03 For some reason, the default setting of PATH when running a command from
1281 a pipe transport was just "/usr/bin". I have changed it to
1284 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1285 anything to be listed in the output from -bV.
1287 PH/05 When a filter generated an autoreply, the entire To: header line was
1288 quoted in the delivery log line, like this:
1290 => >A.N.Other <ano@some.domain> <original@ddress> ...
1292 This has been changed so that it extracts the operative address. There
1293 may be more than one such address. If so, they are comma-separated, like
1296 => >ano@some.domain,ona@other.domain <original@ddress> ...
1298 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1299 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1300 not being looked up in the rDNS to get a host name, Exim was showing the
1301 IP address twice in Received: lines, even though the IP addresses were
1302 identical. For example:
1304 Received: from [1.2.3.4] (helo=[1.2.3.4])
1306 However, if the real host name was known, it was omitting the HELO data
1307 if it matched the actual IP address. This has been tidied up so that it
1308 doesn't show the same IP address twice.
1310 PH/07 When both +timestamp and +memory debugging was on, the value given by
1311 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1312 called by the debug printing, thereby overwriting the timestamp buffer.
1313 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1315 PH/08 When the original message was included in an autoreply transport, it
1316 always said "this is a copy of the message, including all the headers",
1317 even if body_only or headers_only was set. It now gives an appropriate
1320 PH/09 Applied a patch from the Sieve maintainer which:
1322 o fixes some comments
1323 o adds the (disabled) notify extension core
1324 o adds some debug output for the result of if/elsif tests
1325 o points to the current vacation draft in the documentation
1326 and documents the missing references header update
1330 o fixes a bug in processing the envelope test (when testing
1331 multiple envelope elements, the last element determinted the
1334 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1335 Electronic Mail") by including:
1337 Auto-submitted: auto-generated
1339 in the messages that it generates (bounce messages and others, such as
1340 warnings). In the case of bounce messages for non-SMTP mesages, there was
1341 also a typo: it was using "Auto_submitted" (underscore instead of
1342 hyphen). Since every message generated by Exim is necessarily in response
1343 to another message, thes have all been changed to:
1345 Auto-Submitted: auto-replied
1347 in accordance with these statements in the RFC:
1349 The auto-replied keyword:
1351 - SHOULD be used on messages sent in direct response to another
1352 message by an automatic process,
1354 - MUST NOT be used on manually-generated messages,
1356 - MAY be used on Delivery Status Notifications (DSNs) and Message
1357 Disposition Notifications (MDNs),
1359 - MUST NOT be used on messages generated by automatic or periodic
1360 processes, except for messages which are automatic responses to
1363 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1364 to the default Received: header definition.
1366 PH/12 Added log selector acl_warn_skipped (default on).
1368 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1369 variables because (a) they are in the wrong storage pool and (b) even if
1370 they were copied, it wouldn't work properly because of the caching.
1372 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1373 checking when decoding. Apparently there are clients that generate
1374 overlong encoded strings. Why am I not surprised?
1376 PH/15 If the first argument of "${if match_address" was not empty, but did not
1377 contain an "@" character, Exim crashed. Now it writes a panic log message
1378 and treats the condition as false.
1380 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1382 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1383 extension "envelope-auth". The code is finished and in agreement with
1384 other implementations, but there is no documentation so far and in fact,
1385 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1386 not changing the active code.
1388 Print executed "if" and "elsif" statements when debugging is used. This
1389 helps a great deal to understand what a filter does.
1391 Document more things not specified clearly in RFC3028. I had all this
1392 sorted out, when out of a sudden new issues came to my mind. Oops."
1394 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1397 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1398 canonical form (as documented). However, after a host name lookup from
1399 the IP address, check_host() was doing a simple string comparison with
1400 addresses acquired from the DNS when checking that the found name did
1401 have the original IP as one of its addresses. Since any found IPv6
1402 addresses are likely to be in abbreviated form, the comparison could
1403 fail. Luckily, there already exists a function for doing the comparison
1404 by converting both addresses to binary, so now that is used instead of
1405 the text comparison.
1407 PH/20 There was another similar case to PH/19, when a complete host name was
1408 given in a host list; looking up its IP address could give an abbreviated
1409 form, whereas the current host's name might or might not be abbreviated.
1410 The same fix has been applied.
1416 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1417 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1420 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1421 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1423 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1425 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1426 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1427 Also, for "Auto-Submitted", treat anything other than "no" as
1428 non-personal, in accordance with RFC 3834. (Previously it treated
1429 anything starting "auto-" as non-personal.)
1431 TF/01 The control=submission/name=... option had a problem with syntax
1432 errors if the name included a slash character. The /name= option
1433 now slurps the rest of the string, so it can include any characters
1434 but it must come last in the list of options (after /sender_retain
1437 PH/05 Some modifications to the interface to the fake nameserver for the new
1445 TK/01 Added the "success_on_redirect" address verification option. See
1446 NewStuff for rationale and an example.
1448 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1450 PH/02 Patch to exigrep to allow it to work on syslog lines.
1452 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1453 fread() to skip over the body file's header line, because in Cygwin the
1454 header line is locked and is inaccessible.
1456 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1457 co-exist for some time) to make it clear that it is the Exim ID that is
1458 referenced, not the Message-ID: header line.
1460 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1461 string_format() function, because snprintf() does not exist on all
1464 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1465 ${stat: expansion item.
1467 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1468 protocol synchronization error", to keep the pedants happy.
1470 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1471 well as for IRIX systems, when gcc is being used. See the host.c source
1474 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1476 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1479 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1480 transport and to the smtp transport in LMTP mode.
1482 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1484 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1485 run a filter in a subprocess. This could lead to confusion in subsequent
1486 lookups in the parent process. There should also be a search_tidyup() at
1487 the end of the subprocess.
1489 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1490 only if the host matched helo_try_verify_hosts, which caused the
1491 verification to occur when the EHLO/HELO command was issued. The ACL just
1492 tested the remembered result. Now, if a previous verification attempt has
1493 not happened, "verify = helo" does it there and then.
1495 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1497 TK/03 Fix log output including CR from clamd.
1499 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1500 exist provoked a memory error which could cause a segfault.
1502 PH/15 Installed PCRE 6.2
1504 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1506 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1507 of the problem. Specifically, suggested +O2 rather than +O1 for the
1510 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1512 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1513 record, the port was not correctly logged when the outgoing_port log
1514 selector was set (it logged the transort's default port).
1516 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1517 fallback_hosts, and "hosts" in the smtp transport.
1519 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1520 host errors such as "Connection refused".
1522 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1523 authentication with radiusclient 0.4.9:
1525 - Error returned from rc_read_config was caught wrongly
1526 - Username/password not passed on to radius server due to wrong length.
1528 The presumption is that some radiusclient API changes for 4.51/PH/17
1529 were not taken care of correctly. The code is still untested by me (my
1530 Linux distribution still has 0.3.2 of radiusclient), but it was
1531 contributed by a Radius user.
1533 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1534 expanding the "port" option of the smtp transport.
1536 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1537 while reading a MIME header. Thanks to Tom Hughes for a patch.
1539 PH/24 Include config.h inside local_scan.h so that configuration settings are
1542 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1543 that in an ACL for RCPT (for example), you can examine exactly what was
1546 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1547 commands, but it was not correctly comparing the address with the actual
1548 client host address. Thus, it would show the EHLO address in Received:
1549 header lines when this was not necessary.
1551 PH/27 Added the % operator to ${eval:}.
1553 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1554 it should be ignoring failures (because with -C, for example, it has lost
1555 privilege). It wasn't ignoring creation failures other than "already
1558 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1561 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1562 because an input line was too long, either on its own, or by virtue of
1563 too many continuations, the temporary file was not being removed, and the
1564 return code was incorrect.
1566 PH/31 Missing "BOOL" in function definition in filtertest.c.
1568 PH/32 Applied Sieve patches from the maintainer.
1570 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1572 PH/33 Added "verify = not_blind".
1574 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1575 Local/Makefile (with some defaults set). These are used in built scripts
1576 such as exicyclog, but they have never been used in the exim_install
1577 script (though there are many overriding facilities there). I have
1578 arranged that the exim_install script now takes note of these two
1581 PH/35 Installed configuration files for Dragonfly.
1583 PH/36 When a locally submitted message by a trusted user did not contain a
1584 From: header, and the sender address was obtained from -f or from an SMTP
1585 MAIL command, and the trusted user did not use -F to supply a sender
1586 name, $originator_name was incorrectly used when constructing a From:
1587 header. Furthermore, $originator_name was used for submission mode
1588 messages from external hosts without From: headers in a similar way,
1589 which is clearly wrong.
1591 PH/37 Added control=suppress_local_fixups.
1593 PH/38 When log_selector = +received_sender was set, and the addition of the
1594 sender made the log line's construction buffer exactly full, or one byte
1595 less than full, an overflow happened when the terminating "\n" was
1598 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1599 when the result of a list match is failure because a DNS lookup failed.
1601 PH/40 RM_COMMAND is now used in the building process.
1603 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1604 the "build-* directories that it finds.
1606 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1607 address in a domain literal was a prefix of an interface address.
1609 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1610 when verifying a sender address, unless rewrite_headers is false.
1612 PH/44 Wrote a long comment about why errors_to addresses are verified as
1613 recipients, not senders.
1615 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1616 the ratelimit ACL was added.
1618 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1620 PH/46 Added extra information about PostgreSQL errors to the error string.
1622 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1623 avoiding the need to install special zones in a real server. This is
1624 backwards compatible; if it can't find the fake resolver, it drops back.
1625 Thus, both old and new test suites can be run.
1627 TF/02 Added util/ratelimit.pl
1629 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1630 clock is set back in time.
1632 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1633 Candler <B.Candler@pobox.com>.
1635 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1636 OK for addresses that are the result of redirections.
1638 PH/48 A number of further additions for the benefit of the new test suite,
1639 including a fake gethostbyname() that interfaces to the fake DNS resolver
1642 TF/06 The fix for widen_domains has also been applied to qualify_single and
1643 search_parents which are the other dnslookup options that can cause
1646 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1649 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1651 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1652 service" if it encountered an MX record with an empty target hostname.
1653 The message is now "an MX or SRV record indicated no SMTP service".
1655 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1656 if the DNS of the sending site is misconfigured. This is quite a
1657 common situation. This change restores the behaviour of treating a
1658 helo verification defer as a failure.
1660 PH/51 If self=fail was set on a router, the bounce message did not include the
1661 actual error message.
1667 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1669 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1670 command itself ended in error, the underlying message about the transport
1671 filter timeout was being overwritten with the pipe command error. Now the
1672 underlying error message should be appended to the second error message.
1674 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1676 PH/02 Reduce the amount of output that "make" produces by default. Full output
1677 can still be requested.
1679 PH/03 The warning log line about a condition test deferring for a "warn" verb
1680 was being output only once per connection, rather than after each
1681 occurrence (because it was using the same function as for successful
1682 "warn" verbs). This seems wrong, so I have changed it.
1684 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1685 it should not have, which might have caused a crash in the right
1686 circumstances, but probably never did.
1688 PH/04 Installed a modified version of Tony Finch's patch to make submission
1689 mode fix the return path as well as the Sender: header line, and to
1690 add a /name= option so that you can make the user's friendly name appear
1693 TF/03 Added the control = fakedefer ACL modifier.
1695 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1696 Mark Lowes for thorough testing.
1698 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1700 TK/03 Merged latest SRS patch from Miles Wilton.
1702 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1703 with the definition in sysexits.h (which is #included earlier).
1704 Fortunately, Exim does not actually use EX_OK. The code used to try to
1705 preserve the sysexits.h value, by assumimg that macro definitions were
1706 scanned for macro replacements. I have been disabused of this notion,
1707 so now the code just undefines EX_OK before #including unistd.h.
1709 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1710 in the smtp transport. When a block could not be written in a single
1711 write() function, the timeout was being re-applied to each part-write.
1712 This seems wrong - if the receiver was accepting one byte at a time it
1713 would take for ever. The timeout is now adjusted when this happens. It
1714 doesn't have to be particularly precise.
1716 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1717 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1719 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1720 without a domain if the check to <postmaster@domain> fails.
1722 SC/01 Eximstats: added -xls and the ability to specify output files
1723 (patch written by Frank Heydlauf).
1725 SC/02 Eximstats: use FileHandles for outputing results.
1727 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1729 SC/04 Eximstats: fixed display of large numbers with -nvr option
1731 SC/05 Eximstats: fixed merging of reports with empty tables.
1733 SC/06 Eximstats: added the -include_original_destination flag
1735 SC/07 Eximstats: removed tabs and trailing whitespace.
1737 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1739 TK/06 MBOX spool code: Add real "From " MBOX separator line
1740 so the .eml file is really in mbox format (even though
1741 most programs do not really care). Patch from Alex Miller.
1743 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1744 The latter is generated from $received_to and is only set if the
1745 message has one envelope recipient. SA can use these headers,
1746 obviously out-of-the-box. Patch from Alex Miller.
1748 PH/08 The ${def test on a variable was returning false if the variable's
1749 value was "0", contrary to what the specification has always said!
1750 The result should be true unless the variable is empty.
1752 PH/09 The syntax error of a character other than { following "${if
1753 def:variable_name" (after optional whitespace) was not being diagnosed.
1754 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1755 accidental colon was present, for example, could give incorrect results.
1757 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1758 result is used (not including appendfile, where other changes are about
1761 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1762 This involved changing a lot of size variables from int to off_t. It
1763 should work with maildirs and everything.
1765 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1766 spamd dying while we are connected to it.
1768 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1771 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1772 function for BDB 4.3.
1774 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1776 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1777 the sources for PCRE have been re-arranged and more files are now
1780 PH/15 The code I had for printing potentially long long variables in PH/11
1781 above was not the best (it lost precision). The length of off_t variables
1782 is now inspected at build time, and an appropriate printing format (%ld
1783 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1784 to be "long long int" or "long int". This is needed for the internal
1785 formatting function string_vformat().
1787 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1788 the configuration file to be ":syslog", then the script "guesses" where
1789 the logs files are, rather than using the compiled in default. In our
1790 case the guess is not the same as the compiled default, so the script
1791 suddenly stopped working when I started to use syslog. The patch checks
1792 to see if log_file_path is "". If so, it attempts to read it from exim
1793 with no configuration file to get the compiled in version, before it
1794 falls back to the previous guessing code."
1796 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1797 implementing BATV in an Exim configuration. See NewStuff for the gory
1800 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1801 Makefile that are specific to HP-UX.
1803 PH/18 If the "use_postmaster" option was set for a recipient callout together
1804 with the "random" option, the postmaster address was used as the MAIL
1805 FROM address for the random test, but not for the subsequent recipient
1806 test. It is now used for both.
1808 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1809 patch removes a few documentation additions to RFC 3028, because the
1810 latest draft now contains them. It adds the new en;ascii-case comparator
1811 and a new error check for 8bit text in MIME parts. Comparator and
1812 require names are now matched exactly. I enabled the subaddress
1813 extension, but it is not well tested yet (read: it works for me)."
1815 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1816 rework some of the code of TK/09 above to avoid the hardwired use of
1817 "%lld" and "long long". Replaced the call to snprintf() with a call to
1820 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1821 records point to non-existent hosts", "retry timeout exceeded", and
1822 "retry time not reached for any host after a long failure period".
1824 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1825 experimental DomainKeys support:
1827 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1828 (2) On an error such as an illegally used "control", the wrong name for
1829 the control was given.
1831 These problems did NOT occur unless DomainKeys support was compiled.
1833 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1835 PH/24 Added ${if match_ip condition.
1837 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1838 databases so that it will be absolutely obvious if a crash occurs in the
1839 DB library. This is a regular occurrence (often caused by mis-matched
1842 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1843 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1844 picked up on a user's system that detects such things. There doesn't seem
1845 to be a gcc warning option for this - only an attribute that has to be
1846 put on the function's prototype. It seems that in Fedora Core 4 they have
1847 set this on a number of new functions. No doubt there will be more in due
1850 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1851 specify a transport. However, if an address that was verified by such a
1852 router was the subject of a callout, Exim crashed because it tried to
1853 read the rcpt_include_affixes from the non-existent transport. Now it
1854 just assumes that the setting of that option is false. This bug was
1855 introduced by 4.51/PH/31.
1857 PH/28 Changed -d+all to exclude +memory, because that information is very
1858 rarely of interest, but it makes the output a lot bigger. People tend to
1859 do -d+all out of habit.
1861 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1862 code in os-type was giving problems when libc.so lives in lib64, like on
1865 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1866 aren't the modern standard, and it seems that some systems' include files
1867 don't always have them. Exim was already checking for some of the newer
1868 ones like T_AAAA, and defining it itself. I've added checks for all the
1869 record types that Exim uses.
1871 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1872 not automatically generating a new one, as it is supposed to. This
1873 prevented TLS from working. If the file did exist, but contained invalid
1874 data, a new version was generated, as expected. It was only the case of a
1875 non-existent file that was broken.
1877 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1878 with a change in libdomainkeys > 0.64.
1880 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1881 from DNS. If the selector record carries the flag, it now has
1882 precedence over the domain-wide flag.
1884 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1886 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1887 the use of a port name that isn't defined in /etc/services) occurred, the
1888 message was deferred as in a normal delivery, and thus remained on the
1889 spool, instead of being failed because of the mua_wrapper setting. This
1890 is now fixed, and I tidied up some of the mua_wrapper messages at the
1893 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1894 the messages in a hash of arrays rather than using individual hashes.
1895 This is a bit cleaner and results in dramatic memory savings, albeit
1896 at a slight CPU cost.
1898 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1899 as requested by Marc Sherman.
1901 SC/10 Eximstats: added histograms for user specified patterns as requested
1904 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1906 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1907 fopen() in the content-scanning modules that did not already have it.
1913 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1914 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1916 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1918 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1920 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1921 to test Sieve filters that use "vacation".
1923 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1924 that changes the way the GnuTLS parameters are stored in the cache file.
1925 The new format can be generated externally. For backward compatibility,
1926 if the data in the cache doesn't make sense, Exim assumes it has read an
1927 old-format file, and it generates new data and writes a new file. This
1928 means that you can't go back to an older release without removing the
1931 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1932 work if there are any delivery delays because "one_time" forces the
1933 parent to be marked "delivered", so its unseen clone is never tried
1934 again. For this reason, Exim now forbids the simultaneous setting of
1937 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1938 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1939 entry if you want to know the details. The fix, however, neglected to
1940 consider the case when local delivery batching is involved. The test for
1941 "previously delivered" was not happening when checking to see if an
1942 address could be batched with a previous (undelivered) one; under
1943 certain circumstances this could lead to multiple deliveries to the same
1946 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1947 in its include files, and this causes problems building Exim.
1949 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1950 header_syntax) but Exim was just ignoring anything given after a slash.
1951 In particular, this caused confusion with an attempt to use "verify =
1952 reverse_host_lookup/defer_ok". An error is now given when options are
1953 supplied for verify items that do not have them. (Maybe reverse_host_
1954 lookup should have a defer_ok option, but that's a different point.)
1956 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1957 defined by RFC 821) to 2048, because there were problems with some AUTH
1958 commands, and RFC 1869 says the size should be increased for extended
1959 SMTP commands that take arguments.
1961 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1964 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1965 "unknown" error; now it says that the functionality isn't in the binary.
1967 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1968 an address' error message when a string expansion fails (syntax or
1969 whatever). Otherwise the password may appear in the log. Following change
1970 PH/42 below, there is no longer a chance of it appearing in a bounce
1973 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1975 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1976 of its addresses was ever tried. (Bugzilla bug #2.)
1978 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1979 the result incorrectly in the debug output. (It correctly added a newline
1980 to what was transported.)
1982 TF/01 Added $received_time.
1984 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1985 commented out examples of how to interface to a virus scanner and to
1986 SpamAssassin. Also added commented examples of av_scanner and
1987 spamd_address settings.
1989 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1990 and controls are allowed in which ACLs. There were a couple of minor
1991 errors. Some of the entries in the conditions table (which is a table of
1992 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1993 negation of where the condition IS allowed.
1995 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1997 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1998 header file does not have a version number, so I've had to invent a new
1999 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
2000 API. The code is untested by me (my Linux distribution still has 0.3.2 of
2001 radiusclient), but it was contributed by a Radius user.
2003 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
2004 files or directories, for OpenSSL.
2006 PH/19 When an Exim process that is running as root has to create an Exim log
2007 file, it does so in a subprocess that runs as exim:exim so as to get the
2008 ownership right at creation (otherwise, other Exim processes might see
2009 the file with the wrong ownership). There was no test for failure of this
2010 fork() call, which would lead to the process getting stuck as it waited
2011 for a non-existent subprocess. Forks do occasionally fail when resources
2012 run out. I reviewed all the other calls to fork(); they all seem to check
2015 PH/20 When checking for unexpected SMTP input at connect time (before writing
2016 the banner), Exim was not dealing correctly with a non-positive return
2017 from the read() function. If the client had disconnected by this time,
2018 the result was a log entry for a synchronization error with an empty
2019 string after "input=" when read() returned zero. If read() returned -1
2020 (an event I could not check), uninitialized data bytes were printed.
2021 There were reports of junk text (parts of files, etc) appearing after
2024 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2026 PH/22 Added support for macro redefinition, and (re)definition in between
2027 driver and ACL definitions.
2029 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2030 forgetting to use the resulting value; it was using the unexpanded value.
2032 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2033 hadn't been configured. The fix is from Juergen Kreileder, who
2034 understands it better than I do:
2036 "Here's what I see happening with three configured cyrus_sasl
2037 authenticators configured (plain, login, cram-md5):
2039 On startup auth_cyrus_sasl_init() gets called for each of these.
2040 This means three calls to sasl_listmech() without a specified mech_list.
2041 => SASL tests which mechs of all available mechs actually work
2042 => three warnings about OTP not working
2043 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2045 With the patch, sasl_listmech() also gets called three times. But now
2046 SASL's mech_list option is set to the server_mech specified in the the
2047 authenticator. Or in other words, the answer from sasl_listmech()
2048 gets limited to just the mech you're testing for (which is different
2050 => the return list contains just 'plain' or 'login', 'cram-md5' or
2051 nothing depending on the value of ob->server_mech.
2053 I've just tested the patch: Authentication still works fine,
2054 unavailable mechs specified in the exim configuration are still
2055 caught, and the auth.log warnings about OTP are gone."
2057 PH/25 When debugging is enabled, the contents of the command line are added
2058 to the debugging output, even when log_selector=+arguments is not
2061 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2062 answer is "GNU", and only if the return is "GNU/something" is the answer
2065 PH/27 $acl_verify_message is now set immediately after the failure of a
2066 verification in an ACL, and so is available in subsequent modifiers. In
2067 particular, the message can be preserved by coding like this:
2069 warn !verify = sender
2070 set acl_m0 = $acl_verify_message
2072 Previously, $acl_verify_message was set only while expanding "message"
2073 and "log_message" when a very denied access.
2075 PH/28 Modified OS/os.c-Linux with
2077 -#ifndef OS_LOAD_AVERAGE
2078 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2080 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2081 nomenclature these days.)
2083 PH/29 Installed patch from the Sieve maintainer that adds the options
2084 sieve_useraddress and sieve_subaddress to the redirect router.
2086 PH/30 In these circumstances:
2087 . Two addresses routed to the same list of hosts;
2088 . First host does not offer TLS;
2089 . First host accepts first address;
2090 . First host gives temporary error to second address;
2091 . Second host offers TLS and a TLS session is established;
2092 . Second host accepts second address.
2093 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2094 and peerdn, if requested) that were in fact used only for the second
2097 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2098 attention to any local part prefix or suffix that was matched by the
2099 router that accepted the address. It now behaves in the same way as it
2100 does for delivery: the affixes are removed from the local part unless
2101 rcpt_include_affixes is set on the transport.
2103 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2104 timeout during the DATA phase of an incoming message.
2106 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2107 applied a patch sanctioned by the Sieve maintainer.
2109 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2110 the uid or gid is negative. A case of a negative gid caused this to be
2111 noticed. The fix allows for either to be negative.
2113 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2114 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2117 PH/36 The change PH/12 above was broken. Fixed it.
2119 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2120 the grounds that routing the same address twice would always produce the
2121 same answer. This might have been true once, but it is certainly no
2122 longer true now. Routing a child address may depend on the previous
2123 routing that produced that child. Some complicated redirection strategies
2124 went wrong when messages had multiple recipients, and made Exim's
2125 behaviour dependent on the order in which the addresses were given.
2127 I have moved the duplicate checking until after the routing is complete.
2128 Exim scans the addresses that are assigned to local and remote
2129 transports, and removes any duplicates. This means that more work will be
2130 done, as duplicates will always all be routed, but duplicates are
2131 presumably rare, so I don't expect this is of any significance.
2133 For deliveries to pipes, files, and autoreplies, the duplicate checking
2134 still happens during the routing process, since they are not going to be
2137 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2138 It corrects a timeout issue with spamd. This is Ian's comment: "The
2139 background is that sometimes spamd either never reads data from a
2140 connection it has accepted, or it never writes response data. The exiscan
2141 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2142 blindly assumes that writes won't block so it may never time out."
2144 PH/39 Allow G after quota size as well as K and M.
2146 PH/40 The value set for $authenticated_id in an authenticator may not contain
2147 binary zeroes or newlines because the value is written to log lines and
2148 to spool files. There was no check on this. Now the value is run through
2149 the string_printing() function so that such characters are converted to
2150 printable escape sequences.
2152 PH/41 $message_linecount is a new variable that contains the total number of
2153 lines in the message. Compare $body_linecount, which is the count for the
2156 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2157 bounce and delay warning messages, except in certain special cases, which
2160 (a) An SMTP error message from a remote host;
2161 (b) A message specified in a :fail: redirection;
2162 (c) A message specified in a "fail" command in a system filter;
2163 (d) A message specified in a FAIL return from the queryprogram router;
2164 (e) A message specified by the cannot_route_message router option.
2166 In these cases only, Exim does include the error details in bounce and
2167 warning messages. There are also a few cases where bland messages such
2168 as "unrouteable address" or "local delivery error" are given.
2170 PH/43 $value is now also set for the "else" part of a ${run expansion.
2172 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2173 being worked on, but at least Exim now implements the latest version to
2176 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2177 process to complete was treated as a delivery failure, a timeout while
2178 writing the message to the pipe was logged, but erroneously treated as a
2179 successful delivery. Such timeouts include transport filter timeouts. For
2180 consistency with the overall process timeout, these timeouts are now
2181 treated as errors, giving rise to delivery failures by default. However,
2182 there is now a new Boolean option for the pipe transport called
2183 timeout_defer, which, if set TRUE, converts the failures into defers for
2184 both kinds of timeout. A transport filter timeout is now identified in
2187 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2188 systems where "make" and "gmake" are different, calling "gmake" at top
2189 level broke things. I've arranged for the value of $(MAKE) to be passed
2190 from the Makefile to this script so that it can call the same version of
2194 A note about Exim versions 4.44 and 4.50
2195 ----------------------------------------
2197 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2198 changes of various kinds. As a consequence, a big documentation update was
2199 needed. This delayed the release for rather longer than seemed good, especially
2200 in the light of a couple of (minor) security issues. Therefore, the changes
2201 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2202 release. So 4.44 and 4.50 are in effect two different branches that both start
2205 I have left the 4.50 change log unchanged; it contains all the changes since
2206 4.43. The change log for 4.44 is below; many of its items are identical to
2207 those for 4.50. This seems to be the most sensible way to preserve the
2208 historical information.
2214 1. Minor wording change to the doc/README.SIEVE file.
2216 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2217 computation of the current number of files was incorrect.
2219 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2220 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2223 4. Give more explanation in the error message when the command for a transport
2224 filter fails to execute.
2226 5. There are several places where Exim runs a non-Exim command in a
2227 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2228 was being done only for the command run by the queryprogram router. It is
2229 now done for all such subprocesses. The other cases are: ${run, transport
2230 filters, and the commands run by the lmtp and pipe transports.
2232 6. Added CONFIGURE_GROUP build-time option.
2234 7. Some older OS have a limit of 256 on the maximum number of file
2235 descriptors. Exim was using setrlimit() to set 1000 as a large value
2236 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2237 systems. I've change it so that if it can't get 1000, it tries for 256.
2239 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2240 was an oversight, and furthermore, ever since the addition of extra
2241 controls (e.g. 4.43/32), the checks on when to allow different forms of
2242 "control" were broken. There should now be diagnostics for all cases when a
2243 control that does not make sense is encountered.
2245 9. Added the /retain_sender option to "control=submission".
2247 10. $recipients is now available in the predata ACL (oversight).
2249 11. Tidy the search cache before the fork to do a delivery from a message
2250 received from the command line. Otherwise the child will trigger a lookup
2251 failure and thereby defer the delivery if it tries to use (for example) a
2252 cached ldap connection that the parent has called unbind on.
2254 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2255 of $address_data from the recipient verification was clobbered by the
2256 sender verification.
2258 13. The value of address_data from a sender verification is now available in
2259 $sender_address_data in subsequent conditions in the ACL statement.
2261 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2263 15. Added a new option "connect=<time>" to callout options, to set a different
2266 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2267 was its contents. (It was OK if the option was not defined at all.)
2269 17. A "Completed" log line is now written for messages that are removed from
2270 the spool by the -Mrm option.
2272 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2273 information about exactly what failed.
2275 19. Added -dd to debug only the daemon process.
2277 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2278 handles timeouts, both on the server side and network timeouts. Renamed the
2279 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2281 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2282 It is now set to "smtps".
2284 22. $host_address is now set to the target address during the checking of
2285 ignore_target_hosts.
2287 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2288 being passed; this would have caused $sender_host_name to have been used if
2289 matching the list had actually called for a host name (not very likely,
2290 since this list is usually IP addresses). A host name is now passed as
2293 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2294 code that specifies a non-restarting handler (typically sigaction() in
2295 modern systems) in an attempt to fix a rare and obscure crash bug.
2297 25. Narrowed the window for a race in the daemon that could cause it to ignore
2298 SIGCHLD signals. This is not a major problem, because they are used only to
2299 wake it up if nothing else does.
2301 26. A malformed maildirsize file could cause Exim to calculate negative values
2302 for the mailbox size or file count. Odd effects could occur as a result.
2303 The maildirsize information is now recalculated if the size or filecount
2306 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2307 support for a long time. Removed HAVE_SYS_VFS_H.
2309 28. Installed the latest version of exipick from John Jetmore.
2311 29. In an address list, if the pattern was not a regular expression, an empty
2312 subject address (from a bounce message) matched only if the pattern was an
2313 empty string. Non-empty patterns were not even tested. This was the wrong
2314 because it is perfectly reasonable to use an empty address as part of a
2315 database query. An empty address is now tested by patterns that are
2316 lookups. However, all the other forms of pattern expect the subject to
2317 contain a local part and a domain, and therefore, for them, an empty
2318 address still always fails if the pattern is not itself empty.
2320 30. Exim went into a mad DNS loop when attempting to do a callout where the
2321 host was specified on an smtp transport, and looking it up yielded more
2322 than one IP address.
2324 31. Re-factored the code for checking spool and log partition space into a
2325 function that finds that data and another that does the check. The former
2326 is then used to implement four new variables: $spool_space, $log_space,
2327 $spool_inodes, and $log_inodes.
2329 32. The RFC2047 encoding function was originally intended for short strings
2330 such as real names; it was not keeping to the 75-character limit for
2331 encoded words that the RFC imposes. It now respects the limit, and
2332 generates multiple encoded words if necessary. To be on the safe side, I
2333 have increased the buffer size for the ${rfc2047: expansion operator from
2336 33. It is now permitted to omit both strings after an "if" condition; if the
2337 condition is true, the result is "true". As before, when the second string
2338 is omitted, a false condition yields an empty string. This makes it less
2339 cumbersome to write custom ACL and router conditions.
2341 34. Failure to deliver a bounce message always caused it to be frozen, even if
2342 there was an errors_to setting on the router. The errors_to setting is now
2345 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2346 canonical form (fully expanded) before being placed in
2347 $sender_host_address.
2349 36. The table in the code that translates DNS record types into text (T_A to
2350 "A" for instance) was missing entries for NS and CNAME. It is just possible
2351 that this could have caused confusion if both these types were looked up
2352 for the same domain, because the text type is used as part of Exim's
2353 per-process caching. But the chance of anyone hitting this buglet seems
2356 37. The dnsdb lookup has been extended in a number of ways.
2358 (1) There is a new type, "zns", which walks up the domain tree until it
2359 finds some nameserver records. It should be used with care.
2361 (2) There is a new type, "mxh", which is like "mx" except that it returns
2362 just the host names, not the priorities.
2364 (3) It is now possible to give a list of domains (or IP addresses) to be
2365 looked up. The behaviour when one of the lookups defers can be
2366 controlled by a keyword.
2368 (4) It is now possible to specify the separator character for use when
2369 multiple records are returned.
2371 38. The dnslists ACL condition has been extended: it is now possible to supply
2372 a list of IP addresses and/or domains to be looked up in a particular DNS
2375 39. Added log_selector=+queue_time_overall.
2377 40. When running the queue in the test harness, wait just a tad after forking a
2378 delivery process, to get repeatability of debugging output.
2380 41. Include certificate and key file names in error message when GnuTLS fails
2381 to set them up, because the GnuTLS error message doesn't include the name
2382 of the failing file when there is a problem reading it.
2384 42. Allow both -bf and -bF in the same test run.
2386 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2388 44. The "Exiscan patch" is now merged into the mainline Exim source.
2390 45. Sometimes the final signoff response after QUIT could fail to get
2391 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2392 < 0 before doing a fflush(). This bug looks as though it goes back to the
2393 introduction of TLS in release 3.20, but "sometimes" must have been rare
2394 because the tests only now provoked it.
2396 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2397 (this can affect the format of dates).
2399 47. exim_tidydb, when checking for the continued existence of a message for
2400 which it has found a message-specific retry record, was not finding
2401 messages that were in split spool directories. Consequently, it was
2402 deleting retry records that should have stayed in existence.
2404 48. Steve fixed some bugs in eximstats.
2406 49. The SPA authentication driver was not abandoning authentication and moving
2407 on to the next authenticator when an expansion was forced to fail,
2408 contradicting the general specification for all authenticators. Instead it
2409 was generating a temporary error. It now behaves as specified.
2411 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2412 (the order specifies the preference for clients). The order is now AES256,
2413 AES128, 3DES, ARCFOUR128.
2415 51. Small patch to Sieve code - explicitly set From: when generating an
2418 52. Exim crashed if a remote delivery caused a very long error message to be
2419 recorded - for instance if somebody sent an entire SpamAssassin report back
2420 as a large number of 550 error lines. This bug was coincidentally fixed by
2421 increasing the size of one of Exim's internal buffers (big_buffer) that
2422 happened as part of the Exiscan merge. However, to be on the safe side, I
2423 have made the code more robust (and fixed the comments that describe what
2426 53. Now that there can be additional text after "Completed" in log lines (if
2427 the queue_time_overall log selector is set), a one-byte patch to exigrep
2428 was needed to allow it to recognize "Completed" as not the last thing in
2431 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2432 patch that reportedly fixes this has been added. I am not expert enough to
2433 create a test for it. This is what the patch creator wrote:
2435 "I found a little strange behaviour of ldap code when working with
2436 Windows 2003 AD Domain, where users was placed in more than one
2437 Organization Units. When I tried to give exim partial DN, the exit code
2438 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2439 But simultaneously result of request was absolutely normal ldap result,
2440 so I produce this patch..."
2442 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2443 REFERENCE, so I have modified the code to exclude the patch when that macro
2446 55. Some experimental protocols are using DNS PTR records for new purposes. The
2447 keys for these records are domain names, not reversed IP addresses. The
2448 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2449 leaves it alone. Component reversal etc. now happens only for IP addresses.
2452 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2454 57. Double the size of the debug message buffer (to 2048) so that more of very
2455 long debug lines gets shown.
2457 58. The exicyclog utility now does better if the number of log files to keep
2458 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2460 59. Two changes related to the smtp_active_hostname option:
2462 (1) $smtp_active_hostname is now available as a variable.
2463 (2) The default for smtp_banner uses $smtp_active_hostname instead
2464 of $primary_hostname.
2466 60. The host_aton() function is supposed to be passed a string that is known
2467 to be a valid IP address. However, in the case of IPv6 addresses, it was
2468 not checking this. This is a hostage to fortune. Exim now panics and dies
2469 if the condition is not met. A case was found where this could be provoked
2470 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2471 components; fortuitously, this particular loophole had already been fixed
2472 by change 4.50/55 above.
2474 If there are any other similar loopholes, the new check in host_aton()
2475 itself should stop them being exploited. The report I received stated that
2476 data on the command line could provoke the exploit when Exim was running as
2477 exim, but did not say which command line option was involved. All I could
2478 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2479 running as the user.
2482 61. There was a buffer overflow vulnerability in the SPA authentication code
2483 (which came originally from the Samba project). I have added a test to the
2484 spa_base64_to_bits() function which I hope fixes it.
2487 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2488 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2490 63. The daemon start-up calls getloadavg() while still root for those OS that
2491 need the first call to be done as root, but it missed one case: when
2492 deliver_queue_load_max is set with deliver_drop_privilege. This is
2493 necessary for the benefit of the queue runner, because there is no re-exec
2494 when deliver_drop_privilege is set.
2496 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2497 This has been fixed.
2499 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2500 was in use, was not putting the data itself into the right store pool;
2501 consequently, it could be overwritten for a subsequent message in the same
2502 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2505 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2507 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2508 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2509 the function were treating the return as a boolean value, which happened to
2510 work because 0=false and not-0=true, but is not correct code.
2512 68. The host_aton() function was not handling scoped IPv6 addresses (those
2513 with, for example, "%eth0" on the end) correctly.
2515 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2516 negated items (that is, ~something) in unsigned ints. Some compilers
2517 apparently mutter when there is no cast.
2519 70. If an address verification called from an ACL failed, and did not produce a
2520 user-specific message (i.e. there was only a "system" message), nothing was
2521 put in $acl_verify_message. In this situation, it now puts the system
2524 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2525 session; change 4.31/43 added the unwanted input to the log line - except
2526 that it did not do this in the start of session case. It now does.
2528 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2529 This is wrong and can cause the other end to generate a synchronization
2530 error if it is another Exim or anything else that does the synchronization
2531 check. A QUIT command is no longer sent after a timeout.
2533 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2534 during host lookups.
2536 74. The defer_ok option of callout verification was not working if it was used
2537 when verifying addresses in header lines, that is, for this case:
2539 verify = header_sender/callout=defer_ok
2541 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2542 those file descriptors could be used for SMTP connections. If anything
2543 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2544 could be sent to the SMTP client, causing chaos. The daemon now opens
2545 stdin, stdout, and stderr to /dev/null when it puts itself into the
2548 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2549 log by default. The user can override this with suitable Perl magic.
2551 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2552 the log message when discard triggers, was not working for the DATA ACL or
2553 for the non-SMTP ACL.
2555 78. Error message wording change in sieve.c.
2557 79. If smtp_accept_max_per_host was set, the number of connections could be
2558 restricted to fewer than expected, because the daemon was trying to set up
2559 a new connection before checking whether the processes handling previous
2560 connections had finished. The check for completed processes is now done
2561 earlier. On busy systems, this bug wouldn't be noticed because something
2562 else would have woken the daemon, and it would have reaped the completed
2565 80. If a message was submitted locally by a user whose login name contained one
2566 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2567 It caused a spool format error. I have fixed the spool reading code. A
2568 related problem was that the "from" clause in the Received: line became
2569 illegal because of the space(s). It is now covered by ${quote_local_part.
2571 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2574 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2576 83. Patch from Sieve maintainer to fix unterminated string problem in
2577 "vacation" handling.
2579 84. Some minor changes to the Linux configuration files to help with other
2580 OS variants using glibc.
2582 85. One more patch for Sieve to update vacation handling to latest spec.
2585 ----------------------------------------------------
2586 See the note above about the 4.44 and 4.50 releases.
2587 ----------------------------------------------------
2593 1. Change 4.43/35 introduced a bug that caused file counts to be
2594 incorrectly computed when quota_filecount was set in an appendfile
2597 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2598 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2601 3. Give more explanation in the error message when the command for a transport
2602 filter fails to execute.
2604 4. There are several places where Exim runs a non-Exim command in a
2605 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2606 was being done only for the command run by the queryprogram router. It is
2607 now done for all such subprocesses. The other cases are: ${run, transport
2608 filters, and the commands run by the lmtp and pipe transports.
2610 5. Some older OS have a limit of 256 on the maximum number of file
2611 descriptors. Exim was using setrlimit() to set 1000 as a large value
2612 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2613 systems. I've change it so that if it can't get 1000, it tries for 256.
2615 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2616 was an oversight, and furthermore, ever since the addition of extra
2617 controls (e.g. 4.43/32), the checks on when to allow different forms of
2618 "control" were broken. There should now be diagnostics for all cases when a
2619 control that does not make sense is encountered.
2621 7. $recipients is now available in the predata ACL (oversight).
2623 8. Tidy the search cache before the fork to do a delivery from a message
2624 received from the command line. Otherwise the child will trigger a lookup
2625 failure and thereby defer the delivery if it tries to use (for example) a
2626 cached ldap connection that the parent has called unbind on.
2628 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2629 of $address_data from the recipient verification was clobbered by the
2630 sender verification.
2632 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2633 was its contents. (It was OK if the option was not defined at all.)
2635 11. A "Completed" log line is now written for messages that are removed from
2636 the spool by the -Mrm option.
2638 12. $host_address is now set to the target address during the checking of
2639 ignore_target_hosts.
2641 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2642 being passed; this would have caused $sender_host_name to have been used if
2643 matching the list had actually called for a host name (not very likely,
2644 since this list is usually IP addresses). A host name is now passed as
2647 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2648 code that specifies a non-restarting handler (typically sigaction() in
2649 modern systems) in an attempt to fix a rare and obscure crash bug.
2651 15. Narrowed the window for a race in the daemon that could cause it to ignore
2652 SIGCHLD signals. This is not a major problem, because they are used only to
2653 wake it up if nothing else does.
2655 16. A malformed maildirsize file could cause Exim to calculate negative values
2656 for the mailbox size or file count. Odd effects could occur as a result.
2657 The maildirsize information is now recalculated if the size or filecount
2660 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2661 support for a long time. Removed HAVE_SYS_VFS_H.
2663 18. Updated exipick to current release from John Jetmore.
2665 19. Allow an empty sender to be matched against a lookup in an address list.
2666 Previously the only cases considered were a regular expression, or an
2669 20. Exim went into a mad DNS lookup loop when doing a callout where the
2670 host was specified on the transport, if the DNS lookup yielded more than
2673 21. The RFC2047 encoding function was originally intended for short strings
2674 such as real names; it was not keeping to the 75-character limit for
2675 encoded words that the RFC imposes. It now respects the limit, and
2676 generates multiple encoded words if necessary. To be on the safe side, I
2677 have increased the buffer size for the ${rfc2047: expansion operator from
2680 22. Failure to deliver a bounce message always caused it to be frozen, even if
2681 there was an errors_to setting on the router. The errors_to setting is now
2684 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2685 canonical form (fully expanded) before being placed in
2686 $sender_host_address.
2688 24. Updated eximstats to version 1.33
2690 25. Include certificate and key file names in error message when GnuTLS fails
2691 to set them up, because the GnuTLS error message doesn't include the name
2692 of the failing file when there is a problem reading it.
2694 26. Expand error message when OpenSSL has problems setting up cert/key files.
2697 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2698 (this can affect the format of dates).
2700 28. exim_tidydb, when checking for the continued existence of a message for
2701 which it has found a message-specific retry record, was not finding
2702 messages that were in split spool directories. Consequently, it was
2703 deleting retry records that should have stayed in existence.
2705 29. eximstats updated to version 1.35
2706 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2707 1.35 - bugfix such that pie charts by volume are generated correctly
2709 30. The SPA authentication driver was not abandoning authentication and moving
2710 on to the next authenticator when an expansion was forced to fail,
2711 contradicting the general specification for all authenticators. Instead it
2712 was generating a temporary error. It now behaves as specified.
2714 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2715 (the order specifies the preference for clients). The order is now AES256,
2716 AES128, 3DES, ARCFOUR128.
2718 31. Small patch to Sieve code - explicitly set From: when generating an
2721 32. Exim crashed if a remote delivery caused a very long error message to be
2722 recorded - for instance if somebody sent an entire SpamAssassin report back
2723 as a large number of 550 error lines. This bug was coincidentally fixed by
2724 increasing the size of one of Exim's internal buffers (big_buffer) that
2725 happened as part of the Exiscan merge. However, to be on the safe side, I
2726 have made the code more robust (and fixed the comments that describe what
2729 33. Some experimental protocols are using DNS PTR records for new purposes. The
2730 keys for these records are domain names, not reversed IP addresses. The
2731 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2732 leaves it alone. Component reversal etc. now happens only for IP addresses.
2735 34. The host_aton() function is supposed to be passed a string that is known
2736 to be a valid IP address. However, in the case of IPv6 addresses, it was
2737 not checking this. This is a hostage to fortune. Exim now panics and dies
2738 if the condition is not met. A case was found where this could be provoked
2739 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2740 components; fortuitously, this particular loophole had already been fixed
2741 by change 4.50/55 or 4.44/33 above.
2743 If there are any other similar loopholes, the new check in host_aton()
2744 itself should stop them being exploited. The report I received stated that
2745 data on the command line could provoke the exploit when Exim was running as
2746 exim, but did not say which command line option was involved. All I could
2747 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2748 running as the user.
2751 35. There was a buffer overflow vulnerability in the SPA authentication code
2752 (which came originally from the Samba project). I have added a test to the
2753 spa_base64_to_bits() function which I hope fixes it.
2756 36. The daemon start-up calls getloadavg() while still root for those OS that
2757 need the first call to be done as root, but it missed one case: when
2758 deliver_queue_load_max is set with deliver_drop_privilege. This is
2759 necessary for the benefit of the queue runner, because there is no re-exec
2760 when deliver_drop_privilege is set.
2762 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2763 was in use, was not putting the data itself into the right store pool;
2764 consequently, it could be overwritten for a subsequent message in the same
2765 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2768 38. Sometimes the final signoff response after QUIT could fail to get
2769 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2770 < 0 before doing a fflush(). This bug looks as though it goes back to the
2771 introduction of TLS in release 3.20, but "sometimes" must have been rare
2772 because the tests only now provoked it.
2778 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2779 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2780 Now it returns an integer. A number of calls were still expecting a T/F
2781 return. Fortuitously, in all cases, the tests worked in OK situations,
2782 which is the norm. However, things would have gone wrong on any write
2783 failures on the smtp file descriptor. This function is used when sending
2784 messages over SMTP and also when doing verify callouts.
2786 2. When Exim is called to do synchronous delivery of a locally submitted
2787 message (the -odf or -odi options), it no longer closes stderr before doing
2790 3. Implemented the mua_wrapper option.
2792 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2794 5. Implemented the functions header_remove(), header_testname(),
2795 header_add_at_position(), and receive_remove_recipient(), and exported them
2798 6. If an ACL "warn" statement specified the addition of headers, Exim already
2799 inserted X-ACL-Warn: at the start if there was no header name. However, it
2800 was not making this test for the second and subsequent header lines if
2801 there were newlines in the string. This meant that an invalid header could
2802 be inserted if Exim was badly configured.
2804 7. Allow an ACL "warn" statement to add header lines at the start or after all
2805 the Received: headers, as well as at the end.
2807 8. Added the rcpt_4xx retry error code.
2809 9. Added postmaster_mailfrom=xxx to callout verification option.
2811 10. Added mailfrom=xxxx to the callout verification option, for verify=
2814 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2815 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2817 12. Inserted some casts to stop certain compilers warning when using pointer
2818 differences as field lengths or precisions in printf-type calls (mostly
2819 affecting debugging statements).
2821 13. Added optional readline() support for -be (dynamically loaded).
2823 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2824 same clock tick as a message's arrival, so that its received time was the
2825 same as the "first fail" time on the retry record, and that message
2826 remained on the queue past the ultimate address timeout, every queue runner
2827 would try a delivery (because it was past the ultimate address timeout) but
2828 after another failure, the ultimate address timeout, which should have then
2829 bounced the address, did not kick in. This was a "< instead of <=" error;
2830 in most cases the first failure would have been in the next clock tick
2831 after the received time, and all would be well.
2833 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2834 being recognized when the domain list was tested by the match_domain
2835 condition in an expansion string.
2837 16. Added the ${str2b64: operator.
2839 17. Exim was always calling setrlimit() to set a large limit for the number of
2840 processes, without checking whether the existing limit was already
2841 adequate. (It did check for the limit on file descriptors.) Furthermore,
2842 errors from getrlimit() and setrlimit() were being ignored. Now they are
2843 logged to the main and panic logs, but Exim does carry on, to try to do its
2844 job under whatever limits there are.
2846 18. Imported PCRE 5.0.
2848 19. Trivial typo in log message " temporarily refused connection" (the leading
2851 20. If the log selector return_path_on_delivery was set and an address was
2852 redirected to /dev/null, the delivery process crashed because it assumed
2853 that a return path would always be set for a "successful" delivery. In this
2854 case, the whole delivery is bypassed as an optimization, and therefore no
2857 21. Internal re-arrangement: the function for sending a challenge and reading
2858 a response while authentication was assuming a zero-terminated challenge
2859 string. It's now changed to take a pointer and a length, to allow for
2860 binary data in such strings.
2862 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2864 23. Exim was not respecting finduser_retries when seeking the login of the
2865 uid under which it was called; it was always trying 10 times. (The default
2866 setting of finduser_retries is zero.) Also, it was sleeping after the final
2867 failure, which is pointless.
2869 24. Implemented tls_on_connect_ports.
2871 25. Implemented acl_smtp_predata.
2873 26. If the domain in control=submission is set empty, Exim assumes that the
2874 authenticated id is a complete email address when it generates From: or
2875 Sender: header lines.
2877 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2878 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2879 chown and chgrp in /bin and hostname in /usr/bin.
2881 28. Exim was keeping the "process log" file open after each use, just as it
2882 does for the main log. This opens the possibility of it remaining open for
2883 long periods when the USR1 signal hits a daemon. Occasional processlog
2884 errors were reported, that could have been caused by this. Anyway, it seems
2885 much more sensible not to leave this file open at all, so that is what now
2888 29. The long-running daemon process does not normally write to the log once it
2889 has entered its main loop, and it closes the log before doing so. This is
2890 so that log files can straightforwardly be renamed and moved. However,
2891 there are a couple of unusual error situations where the daemon does write
2892 log entries, and I had neglected to close the log afterwards.
2894 30. The text of an SMTP error response that was received during a remote
2895 delivery was being truncated at 512 bytes. This is too short for some of
2896 the long messages that one sometimes sees. I've increased the limit to
2899 31. It is now possible to make retry rules that apply only when a message has a
2900 specific sender, in particular, an empty sender.
2902 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2903 it possible to be selective about when SMTP synchronization is enforced.
2905 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2907 32. Implemented hosts_connection_nolog.
2909 33. Added an ACL for QUIT.
2911 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2914 35. Added mailbox_size and mailbox_filecount to appendfile.
2916 36. Added control = no_multiline_responses to ACLs.
2918 37. There was a bug in the logic of the code that waits for the clock to tick
2919 in the case where the clock went backwards by a substantial amount such
2920 that the microsecond fraction of "now" was more than the microsecond
2921 fraction of "then" (but the whole seconds number was less).
2923 38. Added support for the libradius Radius client library this is found on
2924 FreeBSD (previously only the radiusclient library was supported).
2930 1. When certain lookups returned multiple values in the form name=value, the
2931 quoting of the values was not always being done properly. Specifically:
2932 (a) If the value started with a double quote, but contained no whitespace,
2934 (b) If the value contained whitespace other than a space character (i.e.
2935 tabs or newlines or carriage returns) it was not quoted.
2936 This fix has been applied to the mysql and pgsql lookups by writing a
2937 separate quoting function and calling it from the lookup code. The fix
2938 should probably also be applied to nisplus, ibase and oracle lookups, but
2939 since I cannot test any of those, I have not disturbed their existing code.
2941 2. A hit in the callout cache for a specific address caused a log line with no
2942 reason for rejecting RCPT. Now it says "Previous (cached) callout
2943 verification failure".
2945 3. There was an off-by-one bug in the queryprogram router. An over-long
2946 return line was truncated at 256 instead of 255 characters, thereby
2947 overflowing its buffer with the terminating zero. As well as fixing this, I
2948 have increased the buffer size to 1024 (and made a note to document this).
2950 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2951 when Exim is waiting for an SMTP response from a remote server, Exim
2952 restarts its select() call on the socket, thereby resetting its timeout.
2953 This is not a problem when such interrupts are rare. Somebody set up a cron
2954 job to run exiwhat every 2 minutes, which is less than the normal select()
2955 timeout (5 or 10 minutes). This meant that the select() timeout never
2956 kicked in because it was always reset. I have fixed this by comparing the
2957 time when an interrupt arrives with the time at the start of the first call
2958 to select(). If more time than the timeout has elapsed, the interrupt is
2959 treated as a timeout.
2961 5. Some internal re-factoring in preparation for the addition of Sieve
2962 extensions (by MH). In particular, the "personal" test is moved to a
2963 separate function, and given an option for scanning Cc: and Bcc: (which is
2964 not set for Exim filters).
2966 6. When Exim created an email address using the login of the caller as the
2967 local part (e.g. when creating a From: or Sender: header line), it was not
2968 quoting the local part when it contained special characters such as @.
2970 7. Installed new OpenBSD configuration files.
2972 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2973 try to make them clearer.
2975 9. Callout options, other than the timeout value, were being ignored when
2976 verifying sender addresses in header lines. For example, when using
2978 verify = header_sender/callout=no_cache
2980 the cache was (incorrectly) being used.
2982 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2983 only the Cygwin environment.
2985 11. When return_path_on_delivery was set as a log selector, if different remote
2986 addresses in the same message used different return paths and parallel
2987 remote delivery occurred, the wrong values would sometimes be logged.
2988 (Whenever a remote delivery process finished, the return path value from
2989 the most recently started remote delivery process was logged.)
2991 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2992 lines when AUTH and/or TLS are in use. This is the "received protocol"
2993 field. Exim used to use "asmtp" for authenticated SMTP, without any
2994 indication (in the protocol name) for TLS use. Now it follows the RFC and
2995 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2996 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2997 names appear in log lines as well as in Received: header lines.
2999 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
3000 extensions, and comparison tests, and to fix some bugs.
3002 14. Changes to the "personal" filter test:
3004 (1) The test was buggy in that it was just doing the equivalent of
3005 "contains" tests on header lines. For example, if a user's address was
3006 anne@some.where, the "personal" test would incorrectly be true for
3008 To: susanne@some.where
3010 This test is now done by extracting each address from the header in turn,
3011 and checking the entire address. Other tests that are part of "personal"
3012 are now done using regular expressions (for example, to check local parts
3013 of addresses in From: header lines).
3015 (2) The list of non-personal local parts in From: addresses has been
3016 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3017 taken from the Sieve specification recommendations.
3019 (3) If the message contains any header line starting with "List-" it is
3020 treated as non-personal.
3022 (4) The test for "circular" in the Subject: header line has been removed
3023 because it now seems ill-conceived.
3025 15. Minor typos in src/EDITME comments corrected.
3027 16. Installed latest exipick from John Jetmore.
3029 17. If headers_add on a router specified a text string that was too long for
3030 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3031 of string_sprintf() is now avoided.
3033 18. $message_body_size was not set (it was always zero) when running the DATA
3034 ACL and the local_scan() function.
3036 19. For the "mail" command in an Exim filter, no default was being set for
3037 the once_repeat time, causing a random time value to be used if "once" was
3038 specified. (If the value happened to be <= 0, no repeat happened.) The
3039 default is now 0s, meaning "never repeat". The "vacation" command was OK
3040 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3041 (I found it when inspecting the code).
3043 20. There is now an overall timeout for performing a callout verification. It
3044 defaults to 4 times the callout timeout, which applies to individual SMTP
3045 commands during the callout. The overall timeout applies when there is more
3046 than one host that can be tried. The timeout is checked before trying the
3047 next host. This prevents very long delays if there are a large number of
3048 hosts and all are timing out (e.g. when the network connections are timing
3049 out). The value of the overall timeout can be changed by specifying an
3050 additional sub-option for "callout", called "maxwait". For example:
3052 verify = sender/callout=5s,maxwait=20s
3054 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3055 to the end before writing, but this should make it even safer).
3057 22. Exim was forgetting that it had advertised PIPELINING for the second and
3058 subsequent messages on an SMTP connection. It was also not resetting its
3059 memory on STARTTLS and an internal HELO.
3061 23. When Exim logs an SMTP synchronization error within a session, it now
3062 records whether PIPELINING has been advertised or not.
3064 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3065 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3066 rather than long int.
3068 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3070 26. Added the never_mail option to autoreply.
3076 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3077 crash if the getsockname() call failed; this can happen if a connection is
3078 closed very soon after it is established. The problem was simply in the
3079 order in which certain operations were done, causing Exim to try to write
3080 to the SMTP stream before it had set up the file descriptor. The bug has
3081 been fixed by making things happen in the correct order.
3087 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3088 before the connection was closed, thus losing the rejection response.
3090 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3091 some early Solaris releases, but causes trouble in current releases where
3092 socklen_t is defined.
3094 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3097 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3100 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3101 that releasing the top of it at the end releases what was used for sub-
3102 expansions (unless the block got too big). However, discard this block if
3103 the first thing is a variable or header, so that we can use its block when
3104 it is dynamic (useful for very large $message_headers, for example).
3106 6. Lookups now cache *every* query, not just the most recent. A new, separate
3107 store pool is used for this. It can be recovered when all lookup caches are
3108 flushed. Lookups now release memory at the end of their result strings.
3109 This has involved some general refactoring of the lookup sources.
3111 7. Some code has been added to the store_xxx() functions to reduce the amount
3112 of flapping under certain conditions.
3114 8. log_incoming_interface used to affect only the <= reception log lines. Now
3115 it causes the local interface and port to be added to several more SMTP log
3116 lines, for example "SMTP connection from", and rejection lines.
3118 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3120 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3122 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3123 could be overwritten at the end of the current message (or the start of a
3124 new message if it was set in a HELO ACL). The value is now preserved for
3125 the duration of the SMTP connection.
3127 12. If a transport had a headers_rewrite setting, and a matching header line
3128 contained an unqualified address, that address was qualified, even if it
3129 did not match any rewriting rules. The underlying bug was that the values
3130 of the flags that permit the existence of unqualified sender and recipient
3131 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3132 non-local messages, and by -bnq for local messages) were not being
3133 preserved with the message after it was received.
3135 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3136 "next input=" as part of the text comprising the host identity instead of
3137 the correct text. The code was using the same buffer for two different
3138 strings. However, depending on which order the printing function evaluated
3139 its arguments, the bug did not always show up. Under Linux, for example, my
3140 test suite worked just fine.
3142 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3143 doesn't work with some older versions of Perl. It has been changed to "my",
3144 which in any case is probably the better facility to use.
3146 15. A really picky compiler found some instances of statements for creating
3147 error messages that either had too many or two few arguments for the format
3150 16. The size of the buffer for calls to the DNS resolver has been increased
3151 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3152 for addresses that have a lot of PTR records. This alleviates a problem; it
3153 does not fully solve it.
3155 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3156 buffer now truncates the list and logs the incident, which is the same
3157 action as happens when Exim is looking up a host name and its aliases.
3158 Previously in this situation something unpredictable would happen;
3159 sometimes it was "internal error: store_reset failed".
3161 18. If a server dropped the connection unexpectedly when an Exim client was
3162 using GnuTLS and trying to read a response, the client delivery process
3163 crashed while trying to generate an error log message.
3165 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3166 string, for example:
3168 warn message = H1: something\nH2: something
3170 the text was added as a single header line from Exim's point of view
3171 though it ended up OK in the delivered message. However, searching for the
3172 second and subsequent header lines using $h_h2: did not work. This has been
3173 fixed. Similarly, if a system filter added multiple headers in this way,
3174 the routers could not see them.
3176 20. Expanded the error message when iplsearch is called with an invalid key to
3177 suggest using net-iplsearch in a host list.
3179 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3180 ACLs are no longer actually imposed (and a message to that effect is
3183 22. If a "gecos" field in a passwd entry contained escaped characters, in
3184 particular, if it contained a \" sequence, Exim got it wrong when building
3185 a From: or a Sender: header from that name. A second bug also caused
3186 incorrect handling when an unquoted " was present following a character
3187 that needed quoting.
3189 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3190 was not being matched caselessly.
3192 24. Arranged for all hyphens in the exim.8 source to be escaped with
3195 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3196 properly. Recipient callout cache records were still being keyed to include
3197 the sender, even when use_sender was set false. This led to far more
3198 callouts that were necessary. The sender is no longer included in the key
3199 when use_sender is false.
3201 26. Added "control = submission" modifier to ACLs.
3203 27. Added the ${base62d: operator to decode base 62 numbers.
3205 28. dnsdb lookups can now access SRV records.
3207 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3208 the configuration file.
3210 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3211 (-v) mode. This makes the output for a verbose queue run more intelligible.
3213 31. Added a use_postmaster feature to recipient callouts.
3215 32. Added the $body_zerocount variable, containing the number of binary zero
3216 bytes in the message body.
3218 33. The time of last modification of the "new" subdirectory is now used as the
3219 "mailbox time last read" when there is a quota error for a maildir
3222 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3224 35. Added +ignore_unknown as a special item in host lists.
3226 36. Code for decoding IPv6 addresses in host lists is now included, even if
3227 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3228 address was recognized as an IP address, but was then not correctly decoded
3229 into binary, causing unexpected and incorrect effects when compared with
3236 1. Very minor rewording of debugging text in manualroute to say "list of
3237 hosts" instead of "hostlist".
3239 2. If verify=header_syntax was set, and a header line with an unqualified
3240 address (no domain) and a large number of spaces between the end of the
3241 name and the colon was received, the reception process suffered a buffer
3242 overflow, and (when I tested it) crashed. This was caused by some obsolete
3243 code that should have been removed. The fix is to remove it!
3245 3. When running in the test harness, delay a bit after writing a bounce
3246 message to get a bit more predictability in the log output.
3248 4. Added a call to search_tidyup() just before forking a reception process. In
3249 theory, someone could use a lookup in the expansion of smtp_accept_max_
3250 per_host which, without the tidyup, could leave open a database connection.
3252 5. Added the variables $recipient_data and $sender_data which get set from a
3253 lookup success in an ACL "recipients" or "senders" condition, or a router
3254 "senders" option, similar to $domain_data and $local_part_data.
3256 6. Moved the writing of debug_print from before to after the "senders" test
3259 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3260 problems for message scanning, either using a data ACL, or using
3261 local_scan() because the Received: header was not generated till after they
3262 were called (in order to set the time as the time of reception completion).
3263 I have revised the way this works. The header is now generated after the
3264 body is received, but before the ACL or local_scan() are called. After they
3265 are run, the timestamp in the header is updated.
3271 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3272 before starting a queue runner without re-exec. This happened only when
3273 deliver_drop_privilege was set or when the Exim user was set to root. The
3274 effect of the bug was that timeouts during subsequent deliveries caused
3275 crashes instead of being properly handled. The handler is now left at its
3276 default (and expected) setting.
3278 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3279 message, again when deliver_drop_privilege is set or Exim is run as root.
3280 The bug described in (1) was not present in this case, but the tidying up
3281 of the other signals was missing. I have made the two cases consistent.
3283 3. The ignore_target_hosts setting on a manualroute router was being ignored
3284 for hosts that were looked up using the /MX notation.
3286 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3289 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3290 operated on the sender address. After changing the $sender_address to <>
3291 for the sender address verify, Exim was re-instated it as the original
3292 (before rewriting) address, but remembering that it had rewritten it, so it
3293 wasn't rewriting it again. This bug also had the effect of breaking the
3294 sender address verification caching when the sender address was rewritten.
3296 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3297 This has been changed so that if the ip literal address matches
3298 ignore_target_hosts, the router declines.
3300 7. Added expansion conditions match_domain, match_address, and match_local_
3301 part (NOT match_host).
3303 8. The placeholder for the Received: header didn't have a length field set.
3305 9. Added code to Exim itself and to exim_lock to test for a specific race
3306 condition that could lead to file corruption when using MBX delivery. The
3307 issue is with the lockfile that is created in /tmp. If this file is removed
3308 after a process has opened it but before that process has acquired a lock,
3309 there is the potential for a second process to recreate the file and also
3310 acquire a lock. This could lead to two Exim processes writing to the file
3311 at the same time. The added code performs the same test as UW imapd; it
3312 checks after acquiring the lock that its file descriptor still refers to
3313 the same named file.
3315 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3316 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3319 11. Added the smtp_active_hostname option. If used, this will typically be made
3320 to depend on the incoming interface address. Because $interface_address is
3321 not set up until the daemon has forked a reception process, error responses
3322 that can happen earlier (such as "too many connections") no longer contain
3325 12. If an expansion in a condition on a "warn" statement fails because a lookup
3326 defers, the "warn" statement is abandoned, and the next ACL statement is
3327 processed. Previously this caused the whole ACL to be aborted.
3329 13. Added the iplsearch lookup type.
3331 14. Added ident_timeout as a log selector.
3333 15. Added tls_certificate_verified as a log selector.
3335 16. Added a global option tls_require_ciphers (compare the smtp transport
3336 option of the same name). This controls incoming TLS connections.
3338 17. I finally figured out how to make tls_require_ciphers do a similar thing
3339 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3340 before starting the TLS session.
3342 18. Tabs are now shown as \t in -bP output.
3344 19. If the log selector return_path_on_delivery was set, Exim crashed when
3345 bouncing a message because it had too many Received: header lines.
3347 20. If two routers both had headers_remove settings, and the first one included
3348 a superfluous trailing colon, the final name in the first list and the
3349 first name in the second list were incorrectly joined into one item (with a
3350 colon in the middle).
3356 1. Added -C and -D options to the exinext utility, mainly to make it easier
3357 to include in the automated testing, but these could be helpful when
3358 multiple configurations are in use.
3360 2. The exinext utility was not formatting the output nicely when there was
3361 an alternate port involved in the retry record key, nor when there was a
3362 message id as well (for retries that were specific to a specific message
3363 and a specific host). It was also confused by IPv6 addresses, because of
3364 the additional colons they contain. I have fixed the IPv4 problem, and
3365 patched it up to do a reasonable job for IPv6.
3367 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3368 delivery, the log line now contains "pipelined" if PIPELINING was used.
3370 4. An SMTP transport process used to panic and die if the bind() call to set
3371 an explicit outgoing interface failed. This has been changed; it is now
3372 treated in the same way as a connect() failure.
3374 5. A reference to $sender_host_name in the part of a conditional expansion
3375 that was being skipped was still causing a DNS lookup. This no longer
3378 6. The def: expansion condition was not recognizing references to header lines
3379 that used bh_ and bheader_.
3381 7. Added the _cache feature to named lists.
3383 8. The code for checking quota_filecount in the appendfile transport was
3384 allowing one more file than it should have been.
3386 9. For compatibility with Sendmail, the command line option
3394 and sets the incoming protocol and host name (for trusted callers). The
3395 host name and its colon can be omitted when only the protocol is to be set.
3396 Note the Exim already has two private options, -pd and -ps, that refer to
3397 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3398 "s", but I don't think that's a major issue.
3400 10. A number of refactoring changes to the code, none of which should affect
3403 (a) The number of logging options was getting close to filling up the
3404 32-bit word that was used as a bit map. I have split them into two classes:
3405 those that are passed in the argument to log_write(), and those that are
3406 only ever tested independently outside of that function. These are now in
3407 separate 32-bit words, so there is plenty of room for expansion again.
3408 There is no change in the user interface or the logging behaviour.
3410 (b) When building, for example, log lines, the code previously used a
3411 macro that called string_cat() twice, in order to add two strings. This is
3412 not really sufficiently general. Furthermore, there was one instance where
3413 it was actually wrong because one of the argument was used twice, and in
3414 one call a function was used. (As it happened, calling the function twice
3415 did not affect the overall behaviour.) The macro has been replaced by a
3416 function that can join an arbitrary number of extra strings onto a growing
3419 (c) The code for expansion conditions now uses a table and a binary chop
3420 instead of a serial search (which was left over from when there were very
3421 few conditions). Also, it now recognizes conditions like "pam" even when
3422 the relevant support is not compiled in: a suitably worded error message is
3423 given if an attempt is made to use such a condition.
3425 11. Added ${time_interval:xxxxx}.
3427 12. A bug was causing one of the ddress fields not to be passed back correctly
3428 from remote delivery subprocesses. The field in question was not being
3429 subsequently used, so this caused to problems in practice.
3431 13. Added new log selectors queue_time and deliver_time.
3433 14. Might have fixed a bug in maildirsizefile handling that threw up
3434 "unexpected character" debug warnings, and recalculated the data
3435 unnecessarily. In any case, I expanded the warning message to give more
3438 15. Added the message "Restricted characters in address" to the statements in
3439 the default ACL that block characters like @ and % in local parts.
3441 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3442 Three changes have been made:
3444 (a) There was a serious bug; a negative response to MAIL caused the whole
3445 recipient domain to be cached as invalid, thereby blocking all messages
3446 to all local parts at the same domain, from all senders. This bug has
3447 been fixed. The domain is no longer cached after a negative response to
3448 MAIL if the sender used is not empty.
3450 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3453 (c) A new callout option, "use_sender" has been added for people who want
3454 the modified behaviour.
3460 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3463 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3464 indeed breaks things for older releases.
3466 3. Added additional logging to the case where there is a problem reading data
3467 from a filter that is running in a subprocess using a pipe, in order to
3468 try to track down a specific problem.
3470 4. Testing facility fudge: when running in the test harness and attempting
3471 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3472 getting "No route to host". Convert this to a timeout.
3474 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3477 6. Some OS don't have socklen_t but use size_t instead. This affects the
3478 fifth argument of getsockopt() amongst other things. This is now
3479 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3480 can be set for individual OS. I have set it for SunOS5, OSF1, and
3481 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3482 some earlier ones do not.
3484 7. Change 4.30/15 was not doing the test caselessly.
3486 8. The standard form for an IPv6 address literal was being rejected by address
3487 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3488 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3489 this, as well as the form without the "IPv6" on the front (but only when
3490 address literals are enabled, of course).
3492 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3494 10. Exim crashed if a message with an empty sender address specified by -f
3495 encountered a router with an errors_to setting. This could be provoked only
3496 by a command such as
3500 where an empty string was supplied; "<>" did not hit this bug.
3502 11. Installed PCRE release 4.5.
3504 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3505 remained set. It is now erased.
3507 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3508 times from message ids (which are base 36 rather than the normal 62).
3510 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3511 were being counted as actual protocol errors, and logged if the log
3512 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3513 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3514 and DATA following a set of rejected RCPTs do not count as protocol errors.
3515 In other words, Exim assumes they were pipelined, though this may not
3516 actually be the case. Of course, in all cases the client gets an
3517 appropriate error code.
3519 15. If a lookup fails in an ACL condition, a message about the failure may
3520 be available; it is used if testing the ACL cannot continue, because most
3521 such messages specify what the cause of the deferral is. However, some
3522 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3523 that caused an old message to be retained and used if a later statement
3524 caused a defer, replacing the real cause of the deferral.
3526 16. If an IP address had so many PTR records that the DNS lookup buffer
3527 was not large enough to hold them, Exim could crash while trying to process
3528 the truncated data. It now detects and logs this case.
3530 17. Further to 4.21/58, another change has been made: if (and only if) the
3531 first line of a message (the first header line) ends with CRLF, a bare LF
3532 in a subsequent header line has a space inserted after it, so as not to
3533 terminate the header.
3535 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3536 unnecessarily, used atoi() instead of strtol(), and didn't check the
3537 termination when getting file sizes from file names by regex.
3539 19. Completely re-implemented the support for maildirsize files, in the light
3540 of a number of problems with the previous contributed implementation
3541 (4.30/29). In particular:
3543 . If the quota is zero, the maildirsize file is maintained, but no quota is
3546 . If the maildir directory does not exist, it is created before any attempt
3547 to write a maildirsize file.
3549 . The quota value in the file is just a cache; if the quota is changed in
3550 the transport, the new value overrides.
3552 . A regular expression is available for excluding directories from the
3555 20. The autoreply transport checks the characters in options that define the
3556 message's headers; it allows continued headers, but it was checking with
3557 isspace() after an embedded newline instead of explicitly looking for a
3560 21. If all the "regular" hosts to which an address was routed had passed their
3561 expiry times, and had not reached their retry times, the address was
3562 bounced, even if fallback hosts were defined. Now Exim should go on to try
3565 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3566 equivalent code in the SMTP transport. Some hosts send humungous responses
3567 to HELO/EHLO, more than 1024 it seems.
3569 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3570 gives compiler warnings in some environments. I've now done it "properly",
3573 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3574 (because of problems with the built-in one) was declared to return uschar *
3575 instead of char *, causing compiler failure.
3577 25. Fixed a file descriptor leak when processing alias/forward files.
3579 26. Fixed a minor format string issue in dbfn.c.
3581 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3583 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3584 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3587 29. When a sender address is verified, it is cached, to save repeating the test
3588 when there is more than one recipient in a message. However, when the
3589 verification involves a callout, it is possible for different callout
3590 options to be set for different recipients. It is too complicated to keep
3591 track of this in the cache, so now Exim always runs a verification when a
3592 callout is required, relying on the callout cache for the optimization.
3593 The overhead is duplication of the address routing, but this should not be
3596 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3597 to be verified with callout=postmaster, and the main callout worked but the
3598 postmaster check failed, the verification correctly failed. However, if a
3599 subsequent RCPT command asked for sender verification *without* the
3600 postmaster check, incorrect caching caused this verification also to fail,
3603 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3604 it was not caching the DNS options (qualify_single, search_parents) that
3605 were used when the lookup failed. A subsequent lookup with different
3606 options therefore always gave the same answer, though there were cases
3607 where it should not have. (Example: a "domains = !$mx_any" option on a
3608 dnslookup router: the "domains" option is always processed without any
3609 widening, but the router might have qualify_single set.) Now Exim uses the
3610 cached value only when the same options are set.
3612 32. Added John Jetmore's "exipick" utility to the distribution.
3614 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3615 than a timeout (e.g. a certificate is required, and is not provided), an
3616 Exim server now closes the connection immediately. Previously it waited for
3617 the client to close - but if the client is SSL, it seems that they each
3618 wait for each other, leading to a delay before one of them times out.
3620 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3621 maintained 0.8.x compatibility because I don't think many are using it, and
3622 it is clearly obsolete.
3624 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3627 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3628 client certificate was expired. A simple patch fixes this, though I don't
3629 understand the full logic of why the verify callback is called multiple
3632 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3633 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3634 which causes problems with some clients (such as the Certicom SSL Plus
3635 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3636 disables the coutermeasure allowing Eudora to connect."
3638 38. Exim was not checking that a write() to a log file succeeded. This could
3639 lead to Bad Things if a log got too big, in particular if it hit a file
3640 size limit. Exim now panics and dies if it cannot write to a log file, just
3641 as it does if it cannot open a log file.
3643 39. Modified OS/Makefile-Linux so that it now contains
3645 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3647 The two -D definitions ensure that Exim is compiled with large file
3648 support, which makes it possible to handle log files that are bigger than
3651 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3652 instance) a domain was checked against a named list that involved a lookup,
3653 causing $domain_data to be set, then another domain was checked against the
3654 same list, then the first domain was re-checked, the value of $domain_data
3655 after the final check could be wrong. In particular, if the second check
3656 failed, it could be set empty. This bug probably also applied to
3659 41. The strip_trailing_dot option was not being applied to the address given
3660 with the -f command-line option.
3662 42. The code for reading a message's header from the spool was incrementing
3663 $received_count, but never initializing it. This meant that the value was
3664 incorrect (doubled) while delivering a message in the same process in which
3665 it was received. In the most common configuration of Exim, this never
3666 happens - a fresh exec is done - but it can happen when
3667 deliver_drop_privilege is set.
3669 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3670 it now includes up to 150 characters of the unexpected data in the log
3673 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3674 and building data strings. The size of both of these buffers was 10 000
3675 bytes - far larger than anybody would *ever* want, thought I. Needless to
3676 say, somebody hit the limit. I have increased the maximum line length to
3677 20 000 and the maximum data length of concatenated lines to 100 000. I have
3678 also fixed two bugs, because there was no checking on these buffers. Tsk,
3679 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3680 buffer is too small.
3682 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3683 lsearch lookups. Now it does.
3685 46. When parsing a route_list item in a manualroute router, a fixed-length
3686 buffer was used for the list of hosts. I made this 1024 bytes long,
3687 thinking that nobody would ever have a list of hosts that long. Wrong.
3688 Somebody had a whole pile of complicated expansion conditions, and the
3689 string was silently truncated, leading to an expansion error. It turns out
3690 that it is easier to change to an unlimited length (owing to other changes
3691 that have happened since this code was originally written) than to build
3692 structure for giving a limitation error. The length of the item that
3693 expands into the list of hosts is now unlimited.
3695 47. The lsearch lookup could not handle data where the length of text line was
3696 more than 4095 characters. Such lines were truncated, leading to shortened
3697 data being returned. It should now handle lines of any length.
3699 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3700 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3703 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3704 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3705 no longer happens in comment lines. A list of replacements is now placed
3706 at the head of all of the source files, except those whose only change is
3707 to replace PERL_COMMAND in the very first #! line.
3709 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3710 messages on the queue, with a bottom-up merge sort, using code contributed
3711 by Michael Haardt. This should make operations like -bp somewhat faster on
3712 large queues. It won't affect queue runners, except when queue_run_in_order
3715 51. Installed eximstats 1.31 in the distribution.
3717 52. Added support for SRV lookups to the dnslookup router.
3719 53. If an ACL referred to $message_body or $message_body_end, the value was not
3720 reset for any messages that followed in the same SMTP session.
3722 54. The store-handling optimization for building very long strings was not
3723 differentiating between the different store pools. I don't think this
3724 actually made any difference in practice, but I've tidied it.
3726 55. While running the routers to verify a sender address, $sender_address
3727 was still set to the sender address. This is wrong, because when routing to
3728 send a bounce to the sender, it would be empty. Therefore, I have changed
3729 it so that, while verifying a sender address, $sender_address is set to <>.
3730 (There is no change to what happens when verifying a recipient address.)
3732 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3733 target A or AAAA records (if not already returned) without resetting the
3734 qualify_single or search_parents options of the DNS resolver. These are
3735 inappropriate in this case because the targets of MX and SRV records must
3736 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3737 target that, when qualified, matched something in the local domain. These
3738 two options are now turned off when doing these lookups.
3740 57. It seems that at least some releases of Reiserfs (which does not have the
3741 concept of a fixed number of inodes) returns zero and not -1 for the
3742 number of available inodes. This interacted badly with check_spool_inodes,
3743 which assumed that -1 was the "no such thing" setting. What I have done is
3744 to check that the total number of inodes is greater than zero before doing
3745 the test of how many are available.
3747 58. When a "warn" ACL statement has a log_message modifier, the message is
3748 remembered, and not repeated. This is to avoid a lot of repetition when a
3749 message has many recipients that cause the same warning to be written.
3750 Howewer, Exim was preserving the list of already written lines for an
3751 entire SMTP session, which doesn't seem right. The memory is now reset if a
3752 new message is started.
3754 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3755 debugging output unless log_rewrite was also set.
3757 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3758 of (int)(handle) when we know that handle contains (void *)(-1).
3760 61. The Exim daemon panic-logs an error return when it closes the incoming
3761 connection. However "connection reset by peer" seems to be common, and
3762 isn't really an error worthy of noting specially, so that particular error
3765 62. When Exim is trying to find all the local interfaces, it used to panic and
3766 die if the ioctl to get the interface flags failed. However, it seems that
3767 on at least one OS (Solaris 9) it is possible to have an interface that is
3768 included in the list of interfaces, but for which you get a failure error
3769 for this call. This happens when the interface is not "plumbed" into a
3770 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3771 failure of the "get flags" call assumes that the interface is down.
3773 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3774 makes life easier for people who are doing arithmetic on fields extracted
3775 from dates, where you often get leading zeros that should not be
3776 interpreted as octal.
3778 64. Added qualify_domain to the redirect router, to override the global
3781 65. If a pathologically long header line contained very many addresses (the
3782 report of this problem mentioned 10 000) and each of them was rewritten,
3783 Exim could use up a very large amount of memory. (It kept on making new
3784 copies of the header line as it rewrote, and never released the old ones.)
3785 At the expense of a bit more processing, the header rewriting function has
3786 been changed so that it no longer eats memory in this way.
3788 66. The generation of the Received: header has been moved from the time that a
3789 message starts to be received, to the time that it finishes. The timestamp
3790 in the Received: header should now be very close to that of the <= log
3791 line. There are two side-effects of this change:
3793 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3794 logged header lines no longer include the local Received: line, because
3795 it has not yet been created. The same applies to a copy of the message
3796 that is returned to a non-SMTP sender when a message is rejected.
3798 (b) When a filter file is tested using -bf, no additional Received: header
3799 is added to the test message. After some thought, I decided that this
3802 This change does not affect the value of $received_for. It is still set
3803 after address rewriting, but before local_scan() is called.
3805 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3807 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3808 gave an unhelpful panic error message, and a defer error. I have managed to
3809 change this behaviour so that it now rejects any supplied certificate,
3810 which seems right, as the list of acceptable certificates is empty.
3812 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3813 gave an unhelpful defer error. I have not managed to make this reject any
3814 supplied certificates, but the error message it gives is "no certificate
3815 supplied", which is not helpful.
3817 70. exigrep's output now also includes lines that are not associated with any
3818 message, but which match the given pattern. Implemented by a patch from
3819 Martin Sluka, which also tidied up the Perl a bit.
3821 71. Recipient callout verification, like sender verification, was using <> in
3822 the MAIL FROM command. This isn't really the right thing, since the actual
3823 sender may affect whether the remote host accepts the recipient or not. I
3824 have changed it to use the actual sender in the callout; this means that
3825 the cache record is now keyed on a recipient/sender pair, not just the
3826 recipient address. There doesn't seem to be a real danger of callout loops,
3827 since a callout by the remote host to check the sender would use <>.
3828 [SEE ABOVE: changed after hitting problems.]
3830 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3831 temporary errors. However, in the case of such a code being given after
3832 the end of a data transmission (i.e. after ".") Exim was failing to write
3833 a retry record for the message. (Yes, there was some broken host that was
3834 actually sending 8xx at this point.)
3836 73. An unknown lookup type in a host list could cause Exim to panic-die when
3837 the list was checked. (An example that provoked this was putting <; in the
3838 middle of a list instead of at the start.) If this happened during a DATA
3839 ACL check, a -D file could be left lying around. This kind of configuration
3840 error no longer causes Exim to die; instead it causes a defer errror. The
3841 incident is still logged to the main and panic logs.
3843 74. Buglet left over from Exim 3 conversion. The message "too many messages
3844 in one connection" was written to the rejectlog but not the mainlog, except
3845 when address rewriting (yes!) was being logged.
3847 75. Added write_rejectlog option.
3849 76. When a system filter was run not as root (that is, when system_filter_user
3850 was set), the values of the $n variables were not being returned to the
3851 main process; thus, they were not subsequently available in the $sn
3854 77. Added +return_path_on_delivery log selector.
3856 78. A connection timeout was being treated differently from recipients deferred
3857 when testing hosts_max_try with a message that was older than the host's
3858 retry timeout. (The host should not be counted, thus allowing all hosts to
3859 be tried at least once before bouncing.) This may have been the cause of an
3860 occasionally reported bug whereby a message would remain on the queue
3861 longer than the retry timeout, but would be bounced if a delivery was
3862 forced. I say "may" because I never totally pinned down the problem;
3863 setting up timeout/retry tests is difficult. See also the next item.
3865 79. The ultimate address timeout was not being applied to errors that involved
3866 a combination of host plus message (for example, a timeout on a MAIL
3867 command). When an address resolved to a number of possible hosts, and they
3868 were not all tried for each delivery (e.g. because of hosts_max_try), a
3869 message could remain on the queue longer than the retry timeout.
3871 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3874 81. Fixed an obscure SMTP outgoing bug which required at least the following
3875 conditions: (a) there was another message waiting for the same server;
3876 (b) the server returned 5xx to all RCPT commands in the first message so
3877 that the message was not completed; (c) the server dropped the connection
3878 or gave a negative response to the RSET that Exim sends to abort the
3879 transaction. The observed case was a dropped connection after DATA that had
3880 been sent in pipelining mode. That is, the server had advertised PIPELINING
3881 but was not implementing it correctly. The effect of the bug was incorrect
3882 behaviour, such as trying another host, and this could lead to a crash.
3888 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3889 and daemon.c were passed as pointers to ints; they should have been
3890 pointers to socklen_t variables (which are typically unsigned ints).
3892 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3895 3. Fixed a really odd bug that affected only the testing scheme; patching a
3896 certain fixed string in the binary changed the value of another string that
3897 happened to be identical to the end of the original first string.
3899 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3900 name", it returns that address as the IP address. On some operating
3901 systems (e.g. Solaris), it also passes back the IP address string as the
3902 "host name". However, on others (e.g. Linux), it passes back an empty
3903 string. Exim wasn't checking for this, and was changing the host name to an
3904 empty string, assuming it had been canonicized.
3906 5. Although rare, it is permitted to have more than one PTR record for a given
3907 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3908 all the names associated with an address, because they do in Solaris.
3909 However, it seems that they do not in Linux for data that comes from the
3910 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3911 I found this out when I moved to a new Linux workstation and tried to run
3912 the Exim test suite.
3914 To get round this problem I have changed the code so that it now does its
3915 own call to the DNS to look up PTR records when searching for a host name.
3916 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3917 addresses that are only in /etc/hosts are still found.
3919 This behaviour is, however, controlled by an option called host_lookup_
3920 order, which defaults to "bydns:byaddr". If people want to use the other
3921 order, or indeed, just use one or the other means of lookup, they can
3922 specify it in this variable.
3924 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3925 some operating systems, this comes back from gethostbyaddr() as an empty
3926 string, and this is what Exim used to test for. However, it seems that in
3927 other systems, "." is yielded. Exim now tests for this case too.
3929 7. The values of check_spool_space and check_log_space are now held internally
3930 as a number of kilobytes instead of an absolute number of bytes. If a
3931 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3932 kilobyte. This means that much larger values can be stored.
3934 8. Exim monitor: an attempt to get the action menu when not actually pointing
3935 at a message produces an empty menu entitled "No message selected". This
3936 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3937 no entries in it ("Shell widget menu has zero width and/or height"). So I
3938 have added a single, blank menu entry in this case.
3940 9. Added ${quote_local_part.
3942 10. MIME decoding is now applied to the contents of Subject: header lines when
3945 11. Now that a reference to $sender_host_address automatically causes a reverse
3946 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3947 host lookup before query-style lookups in lists that might use this
3948 variable. This has therefore been abolished, and the "net-" prefix is no
3949 longer necessary for query-style lookups.
3951 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3952 have been a typo for LFLAGS, so it has been changed.
3954 13. The install script calls Exim with "-C /dev/null" in order to find the
3955 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3956 to be output. Howeve, since Exim outputs its version number before the
3957 error, it didn't break the script. It just looked ugly. I fixed this by
3958 always allowing "-C /dev/null" if the caller is root.
3960 14. Ignore overlarge ACL variable number when reading spool file - insurance
3961 against a later release with more variables having written the file.
3963 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3964 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3965 this, as well as the form without the "IPv6" on the front.
3967 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3968 OS/Makefile-Darwin file.
3970 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3971 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3973 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3974 variable. However, some OS use ints for inodes. I've added cast to long int
3975 to get rid of the compiler warning.
3977 19. I had forgotten to lock out "/../" in configuration file names when
3978 ALT_CONFIG_PREFIX was set.
3980 20. Routers used for verification do not need to specify transports. However,
3981 if such a router generated a host list, and callout was configured, Exim
3982 crashed, because it could not find a port number from the (non-existent)
3983 transport. It now assumes port 25 in this circumstance.
3985 21. Added the -t option to exigrep.
3987 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3988 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3989 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3990 them is set, LOOKUP_LSEARCH is forced.
3992 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3994 24. Added sender and host information to the "rejected by local_scan()" log
3995 line; previously there was no indication of these.
3997 25. Added .include_if_exists.
3999 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
4000 Linux. It turns out that not all file systems support this. Apparently some
4001 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
4002 do it.) To cope with this, the error EINVAL, which means that sync-ing is
4003 not supported on the file descriptor, is now ignored when Exim is trying to
4004 sync a directory. This applies only to Linux.
4006 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4008 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4009 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4012 29. Added support for maildirsize files from supplied patch (modified a bit).
4014 30. The use of :fail: followed by an empty string could lead Exim to respond to
4015 sender verification failures with (e.g.):
4017 550 Verification failed for <xxx>
4018 550 Sender verify failed
4020 where the first response line was missing the '-' that indicates it is not
4021 the final line of the response.
4023 31. The loop for finding the name of the user that called Exim had a hardwired
4024 limit of 10; it now uses the value of finduser_retries, which is used for
4025 all other user lookups.
4027 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4030 33. Exim was neglecting to zero errno before one call of strtol() when
4031 expanding a string and expecting an integer value. On some systems this
4032 resulted in spurious "integer overflow" errors. Also, it was casting the
4033 result into an int without checking.
4035 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4036 did not work. The code looks as if it has *never* worked, though it appears
4037 to have been documented since at least releast 1.62. I have made it work.
4039 35. The "timeout_DNS" error in retry rules, also documented since at least
4040 1.62, also never worked. As it isn't clear exactly what this means, and
4041 clearly it isn't a major issue, I have abolished the feature by treating it
4042 as "timeout", and writing a warning to the main and panic logs.
4044 36. The display of retry rules for -brt wasn't always showing the error code
4047 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4048 timeout_connect_A, timeout_connect_MX.
4050 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4051 to the empty sender.
4053 39. The daemon was not analysing the content of -oX till after it had closed
4054 stderr and disconnected from the controlling terminal. This meant that any
4055 syntax errors were only noted on the panic log, and the return code from
4056 the command was 0. By re-arranging the code a little, I've made the
4057 decoding happen first, so such errors now appear on stderr, and the return
4058 code is 1. However, the actual setting up of the sockets still happens in
4059 the disconnected process, so errors there are still only recorded on the
4062 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4063 connections (as happens on some IP stacks) was logged at start up time as
4064 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4065 it from "IPv6 and IPv4", which means that two separate sockets are being
4068 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4069 says whether AF_INET or AF_INET6 was passed as an argument.
4071 42. Exiwhat output was messed up when time zones were included in log
4074 43. Exiwhat now gives more information about the daemon's listening ports,
4075 and whether -tls-on-connect was used.
4077 44. The "port" option of the smtp transport is now expanded.
4079 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4080 silently ignored. Now an error message is written to the main and panic
4083 46. There's a new ACL modifier called "logwrite" which writes to a log file
4084 as soon as it is encountered.
4086 47. Added $local_user_uid and $local_user_gid at routing time.
4088 48. Exim crashed when trying to verify a sender address that was being
4091 49. Exim was recognizing only a space character after ".include". It now also
4092 recognizes a tab character.
4094 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4095 extracting the relevant information from the specification. The man page no
4096 longer contains scrambled data for the -d option, and I've added a section
4097 at the front about calling Exim under different names.
4099 51. Added "extra_headers" argument to the "mail" command in filter files.
4101 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4104 53. Installed eximstats 1.29.
4106 54. Added transport_filter_timeout as a generic transport option.
4108 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4109 Cc: header lines. This was required by RFC 822, but it not required by RFC
4112 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4113 incoming messages that did not have them. Now it does so only if the
4114 message originates locally, that is, if there is no associated remote host
4115 address. When Resent- header lines are present, this applies to the Resent-
4116 lines rather than the non-Resent- lines.
4118 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4119 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4121 58. Messages for configuration errors now include the name of the main
4122 configuration file - useful now that there may be more than one file in a
4123 list (.included file names were always shown).
4125 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4126 for those rare installations that do not start the daemon as root or run it
4127 setuid root. I've cut out the call to initgroups() if the daemon is not
4130 60. The Exim user and group can now be bound into the binary as text strings
4131 that are looked up at the start of Exim's processing.
4133 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4135 62. Added $mailstore_basename variable.
4137 63. Installed patch to sieve.c from Michael Haardt.
4139 64. When Exim failed to open the panic log after failing to open the main log,
4140 the original message it was trying to log was written to stderr and debug
4141 output, but if they were not available (the usual case in production), it
4142 was lost. Now it is written to syslog before the two lines that record the
4143 failures to open the logs.
4145 65. Users' Exim filters run in subprocesses under the user's uid. It is
4146 possible for a "deliver" command or an alias in a "personal" command to
4147 provoke an address rewrite. If logging of address rewriting is configured,
4148 this fails because the process is not running as root or exim. There may be
4149 a better way of dealing with this, but for the moment (because 4.30 needs
4150 to be released), I have disabled address rewrite logging when running a
4151 filter in a non-root, non-exim process.
4157 1. The buildconfig auxiliary program wasn't quoting the value set for
4158 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4159 not defined. This bug was masked in 4.22 by the effect that was fixed in
4162 2. Some messages that were rejected after a message id was allocated were
4163 shown as "incomplete" by exigrep. It no longer does this for messages that
4164 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4166 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4167 have allow_domain_literals set, the ID did not get logged in the <= line.
4168 Domain literals are now always recognized in Message-ID: header lines.
4170 4. The first argument for a ${extract expansion item is the key name or field
4171 number. Leading and trailing spaces in this item were not being ignored,
4172 causing some misleading effects.
4174 5. When deliver_drop_privilege was set, single queue runner processes started
4175 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4176 same command in the process it spins off) were not dropping privilege.
4178 6. When the daemon running as "exim" started a queue runner, it always
4179 re-executed Exim in the spun-off process. This is a waste of effort when
4180 deliver_drop_privilege is set. The new process now just calls the
4181 queue-runner function directly.
4187 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4190 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4191 ignored. Though the use of -oP was forcing the writing of a pid file, it
4192 was always written to the default place.
4194 3. If the message "no IP address found for host xxxx" is generated during
4195 incoming verification, it is now followed by identification of the incoming
4196 connection (so you can more easily find what provoked it).
4198 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4200 5. Added some features to "harden" Exim a bit more against certain attacks:
4202 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4203 be put in Local/Makefile. This is like the never_users runtime option,
4204 but it cannot be overridden. The default setting is "root".
4206 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4207 prefix string with which any file named in a -C command line option
4210 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4211 is retained for -C and -D only if the caller of Exim is root. Without
4212 it, the exim user may also use -C and -D and retain privilege.
4214 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4215 command line option is disabled.
4217 6. Macro names set by the -D option must start with an upper case letter, just
4218 like macro names defined in the configuration file.
4220 7. Added "dereference=" facility to LDAP.
4222 8. Two instances of the typo "uknown" in the source files are fixed.
4224 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4225 the Configure-Makefile script screwed up while processing it.
4227 10. Incorporated PCRE 4.4.
4229 11. The SMTP synchronization check was not operating right at the start of an
4230 SMTP session. For example, it could not catch a HELO sent before the client
4231 waited for the greeting. There is now a check for outstanding input at the
4232 point when the greeting is written. Because of the duplex, asynchronous
4233 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4234 way, but not yet received, when the check is performed.
4236 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4237 on TCP/IP sockets, because this apparently causes some broken clients to
4240 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4241 unchanged) from the Cygwin maintainer.
4243 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4244 was supported instead of testing for mbx. Effectively a typo.
4246 15. The spa authenticator server code was not checking that the input it
4247 received was valid base64.
4249 16. The debug output line for the "set" modifier in ACLs was not showing the
4250 name of the variable that was being set.
4252 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4254 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4255 lookup on its own. Something else was needed to trigger the lookup. For
4256 example, a match in host_lookup or the need for a host name in a host list.
4257 Now, if $sender_host_name is referenced and the host name has not yet been
4258 looked up, a lookup is performed. If the lookup fails, the variable remains
4259 empty, and $host_lookup_failed is set to "1".
4261 19. Added "eqi" as a case-independent comparison operator.
4263 20. The saslauthd authentication condition could segfault if neither service
4264 nor realm was specified.
4266 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4267 error message that was logged was misleading, and incoming SMTP
4268 connections were dropped. The message is now more accurate, and temporary
4269 errors are given to SMTP connections.
4271 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4272 (except RSET) with a 421 error, until QUIT is received. However, it was
4273 failing to send a response to QUIT.
4275 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4276 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4277 ACL is now run after the helo_try_verify_hosts code.
4279 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4280 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4281 case-independent, so other case variants are also recognized). Apparently
4282 some systems use these upper case variants.
4284 25. If more than two messages were waiting for the same host, and a transport
4285 filter was specified for the transport, Exim sent two messages over the
4286 same TCP/IP connection, and then failed with "socket operation on non-
4287 socket" when it tried to send the third.
4289 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4291 27. The extern definition of crypt16() in expand.c was not being excluded when
4292 the OS had its own crypt16() function.
4294 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4295 as a preferred synonym for return_size_limit, both as an option and as an
4298 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4300 30. Changed the default configuration ACL to relax the local part checking rule
4301 for addresses that are not in any local domains. For these addresses,
4302 slashes and pipe symbols are allowed within local parts, but the sequence
4303 /../ is explicitly forbidden.
4305 31. SPA server authentication was not clearing the challenge buffer before
4308 32. log_message in a "warn" ACL statement was writing to the reject log as
4309 well as to the main log, which contradicts the documentation and doesn't
4310 seem right (because no rejection is happening). So I have stopped it.
4312 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4313 However, I am unable to do any testing of this.
4315 34. Fixed an infelicity in the appendfile transport. When checking directories
4316 for a mailbox, to see if any needed to be created, it was accidentally
4317 using path names with one or more superfluous leading slashes; tracing
4318 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4320 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4321 contents are added to the log line that is written for every discarded
4322 recipient. (Previously a log_message setting was ignored.)
4324 36. The ${quote: operator now quotes the string if it is empty.
4326 37. The install script runs exim in order to find its version number. If for
4327 some reason other than non-existence or emptiness, which it checks, it
4328 could not run './exim', it was installing it with an empty version number,
4329 i.e. as "exim-". This error state is now caught, and the installation is
4332 38. An argument was missing from the function that creates an error message
4333 when Exim fails to connect to the socket for saslauthd authentication.
4334 This could cause Exim to crash, or give a corrupted message.
4336 39. Added isip, isip4, and isip6 to ${if conditions.
4338 40. The ACL variables $acl_xx are now saved with the message, and can be
4339 accessed later in routers, transports, and filters.
4341 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4342 strings in the file are not string-expanded.
4344 42. If a MAIL command specified a SIZE value that was too large to fit into an
4345 int variable, the check against message_size_limit failed. Such values are
4346 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4347 day this will have to be increased, but I don't think I want to be around
4348 when emails are that large.
4355 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4356 iconv() is not standard in FreeBSD.
4358 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4359 IPv6 enabled. The observed symptom was a segmentation fault on return from
4360 the function os_common_find_running_interfaces() in src/os.c.
4362 3. In the check_special_case() function in daemon.c I had used "errno" as an
4363 argument name, which causes warnings on some systems. This was basically a
4364 typo, since it was named "eno" in the comments!
4366 4. The code that waits for the clock to tick (at a resolution of some fraction
4367 of a second) so as to ensure message-id uniqueness was always waiting for
4368 at least one whole tick, when it could have waited for less. [This is
4369 almost certainly not relevant at current processor speeds, where it is
4370 unlikely to ever wait at all. But we try to future-proof.]
4372 5. The function that sleeps for a time interval that includes fractions of a
4373 second contained a race. It did not block SIGALRM between setting the
4374 timer, and suspending (a couple of lines later). If the interval was short
4375 and the sigsuspend() was delayed until after it had expired, the suspension
4376 never ended. On busy systems this could lead to processes getting stuck for
4379 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4380 process, before it forks any delivery processes. The open lookup caching
4381 mechanism meant that the open file or database connection was passed into
4382 the delivery process. The problem was that delivery processes always tidy
4383 up cached lookup data. This could cause a problem for the next delivery
4384 process started by the queue runner, because the external queue runner
4385 process does not know about the closure. So the next delivery process
4386 still has data in the lookup cache. In the case of a file lookup, there was
4387 no problem because closing a file descriptor in a subprocess doesn't affect
4388 the parent. However, if the lookup was caching a connection to a database,
4389 the connection was closed, and the second delivery process was likely to
4390 see errors such as "PGSQL: query failed: server closed the connection
4391 unexpectedly". The problem has been fixed by closing all cached lookups
4392 in a queue runner before running a delivery process.
4394 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4395 seem to have the "const" qualifier which it has on other OS. I've
4398 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4399 *of the same type* (client or server) with the same public name that an
4400 error should be diagnosed.
4402 9. When Exim looked up a host name for an IP address, but failed to find the
4403 original IP address when looking up the host name (a safety check), it
4404 output the message "<ip address> does not match any IP for NULL", which was
4405 confusing, to say the least. The bug was that the host name should have
4406 appeared instead of "NULL".
4408 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4409 user that is built into the binary, and the -C or -D options is used, root
4410 privilege is dropped before the configuration file is read. In addition,
4411 logging is switched to stderr instead of the normal log files. If the
4412 configuration then re-defines the Exim user, the unprivileged environment
4413 is probably not what is expected, so Exim logs a panic warning message (but
4416 However, if deliver_drop_privilege is set, the unprivileged state may well
4417 be exactly what is intended, so the warning has been cut out in that case,
4418 and Exim is allowed to try to write to its normal log files.
4424 1. smtp_return_error_details was not giving details for temporary sender
4425 or receiver verification errors.
4427 2. Diagnose a configuration error if two authenticators have the same public
4430 3. Exim used not to create the message log file for a message until the first
4431 delivery attempt. This could be confusing when incoming messages were held
4432 for policy or load reasons. The message log file is now created at the time
4433 the message is received, and an initial "Received" line is written to it.
4435 4. The automatically generated man page for command line options had a minor
4436 bug that caused no ill effects; however, a more serious problem was that
4437 the procedure for building the man page automatically didn't always
4438 operate. Consequently, release 4.20 contains an out-of-date version. This
4439 shouldn't happen again.
4441 5. When building Exim with embedded Perl support, the script that builds the
4442 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4443 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4445 6. The freeze_tell option was not being used for messages that were frozen on
4446 arrival, either by an ACL or by local_scan().
4448 7. Added the smtp_incomplete_transaction log selector.
4450 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4451 was accepting AUTH without a new EHLO.
4453 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4454 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4457 10. Logging of TCP/IP connections (when configured) now happens in the main
4458 daemon process instead of the child process, so that the TCP/IP connection
4459 count is more accurate (but it can never be perfect).
4461 11. The use of "drop" in a nested ACL was not being handled correctly in the
4462 outer ACL. Now, if condition failure induced by the nested "drop" causes
4463 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4464 or "require"), the connection is dropped.
4466 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4467 that yield "discard" can now be used with an "accept" or a "discard" verb,
4468 but an error is generated for any others (because I can't see a useful way
4469 to define what should happen).
4471 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4472 are now processed in the same way as lines in the Exim configuration file.
4473 In particular, continuation lines are supported.
4475 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4477 15. Added -ti meaning -t -i.
4479 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4480 domains, and warn by logging if others are found.
4482 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4483 returned from GIFCONF when Exim is trying to find the list of interfaces on
4484 a host. The code in os.c has been modified to copy each ifreq to an aligned
4485 structure in all cases.
4487 Also, in some cases, the returned ifreq's were being copied to a 'struct
4488 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4489 means the last couple of bytes of an IPv6 address could be chopped if the
4490 ifreq contained only a normal sockaddr (14 bytes storage).
4492 18. Named domain lists were not supported in the hosts_treat_as_local option.
4493 An entry such as +xxxx was not recognized, and was treated as a literal
4496 19. Ensure that header lines added by a DATA ACL are included in the reject log
4497 if the ACL subsequently rejects the message.
4499 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4500 MD5 (which is deprecated).
4502 21. When testing a filter file using -bf, Exim was writing a message when it
4503 took the sender from a "From " line in the message, but it was not doing so
4504 when it took $return_path from a Return-Path: header line. It now does.
4506 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4507 with a valid header line field name (a series of printing characters
4508 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4510 23. Changed "disc" in the source to "disk" to conform to the documentation and
4511 the book and for uniformity.
4513 24. Ignore Sendmail's -Ooption=value command line item.
4515 25. When execve() failed while trying to run a command in a pipe transport,
4516 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4517 could be confused with a return value of 69 from the command itself. This
4518 has been changed to 127, the value the shell returns if it is asked to run
4519 a non-existent command. The wording for the related log line suggests a
4520 non-existent command as the problem.
4522 26. If received_header_text expands to an empty string, do not add a Received:
4523 header line to the message. (Well, it adds a token one on the spool, but
4524 marks it "old" so that it doesn't get used or transmitted.)
4526 27. Installed eximstats 1.28 (addition of -nt option).
4528 28. There was no check for failure on the call to getsockname() in the daemon
4529 code. This can fail if there is a shortage of resources on the system, with
4530 ENOMEM, for example. A temporary error is now given on failure.
4532 29. Contrary to the C standard, it seems that in some environments, the
4533 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4534 program. Exim now does this explicitly; it affects the formatting of
4535 timestamps using strftime().
4537 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4538 complaints. I've now initialized all the variables, to avoid this.
4540 32. Header lines added by a system filter were not being "seen" during
4541 transport-time rewrites.
4543 33. The info_callback() function passed to OpenSSL is set up with type void
4544 (*)(SSL *, int, int), as described somewhere. However, when calling the
4545 function (actually a macro) that sets it up, the type void(*)() is
4546 expected. I've put in a cast to prevent warnings from picky compilers.
4548 34. If a DNS black list lookup found a CNAME record, but there were no A
4549 records associated with the domain it pointed at, Exim crashed.
4551 35. If a DNS black list lookup returned more than one A record, Exim ignored
4552 all but the first. It now scans all returned addresses if a particular IP
4553 value is being sought. In this situation, the contents of the
4554 $dnslist_value variable are a list of all the addresses, separated by a
4557 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4558 to accept a host name and all its aliases if the forward lookup for any of
4559 them yielded the IP address of the incoming connection. Now it accepts only
4560 those names whose forward lookup yields the correct IP address. Any other
4561 names are discarded. This closes a loophole whereby a rogue DNS
4562 administrator could create reverse DNS records to break through a
4563 wildcarded host restriction in an ACL.
4565 37. If a user filter or a system filter that ran in a subprocess used any of
4566 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4567 the wrong values were passed to the pipe command ($thisaddress had the
4568 value of $0, $0 had the value of $1, etc). This bug was introduced by
4569 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4571 38. Improved the line breaking for long SMTP error messages from ACLs.
4572 Previously, if there was no break point between 40 and 75 characters, Exim
4573 left the rest of the message alone. Two changes have been made: (a) I've
4574 reduced the minimum length to 35 characters; (b) if it can't find a break
4575 point between 35 and 75 characters, it looks ahead and uses the first one
4576 that it finds. This may give the occasional overlong line, but at least the
4577 remaining text gets split now.
4579 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4580 file descriptors might be low, and that setting 1000 would always raise it.
4581 It turns out that in some environments, the limit is already over 1000 and
4582 that lowering it causes trouble. So now Exim takes care not to decrease it.
4584 40. When delivering a message, the value of $return_path is set to $sender_
4585 address at the start of routing (routers may change the value). By an
4586 oversight, this default was not being set up when an address was tested by
4587 -bt or -bv, which affected the outcome if any router or filter referred to
4590 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4591 log only when "message" or "log_message" are set. However, if one of the
4592 conditions was an address verification, or a call to a nested ACL, the
4593 messages generated by the underlying test were being passed through. This
4594 no longer happens. The underlying message is available in $acl_verify_
4595 message for both "message" and "log_message" expansions, so it can be
4596 passed through if needed.
4598 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4599 new expansion $bh_ to give the encoded byte string without charset
4600 translation. Translation happens only if iconv() is available; HAVE_ICONV
4601 indicates this at build time. HEADERS_CHARSET gives the charset to
4602 translate to; headers_charset can change it in the configuration, and
4603 "headers charset" can change it in an individual filter file.
4605 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4606 that creates RFC 2047 encoded "words" labels them as that charset instead
4607 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4608 expansion operator; (ii) when Exim creates a From: line for a local
4609 message; (iii) when a header line is rewritten to include a "phrase" part.
4611 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4612 buggy, causing it to skip the first lines of messages whose message ID
4613 ended in 'D'. This would not have bitten before Exim release 4.14, because
4614 message IDs were unlikely to end in 'D' before then. The effect was to have
4615 incorrect size information for certain domains.
4617 45. #include "config.h" was missing at the start of the crypt16.c module. This
4618 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4621 46. If there was a timeout during a "random" callout check, Exim treated it as
4622 a failure of the random address, and carried on sending RSET and the real
4623 address. If the delay was just some slowness somewhere, the response to the
4624 original RCPT would be taken as a response to RSET and so on, causing
4625 mayhem of various kinds.
4627 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4628 when I implemented it. It didn't allow for the fact that some option values
4629 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4630 the right test for positive values.
4632 48. Domain names in DNS records are case-independent. Exim always looks them up
4633 in lower case. Some resolvers return domain names in exactly the case they
4634 appear in the zone file, that is, they may contain uppercase letters. Not
4635 all resolvers do this - some return always lower case. Exim was treating a
4636 change of case by a resolver as a change of domain, similar to a widening
4637 of a domain abbreviation. This triggered its re-routing code and so it was
4638 trying to route what was effectively the same domain again. This normally
4639 caused routing to fail (because the router wouldn't handle the domain
4640 twice). Now Exim checks for this case specially, and just changes the
4641 casing of the domain that it ultimately uses when it transmits the message
4644 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4647 50. If a filter generated a file delivery with a non-absolute name (possible if
4648 no home directory exists for the router), the forbid_file option was not
4651 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4652 the existing equality matching.
4654 52. Exim was using ints instead of ino_t variables in some places where it was
4655 dealing with inode numbers.
4657 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4658 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4659 called TMPDIR, and if it finds it is different, it changes its value.
4661 54. The smtp_printf() function is now made available to local_scan() so
4662 additional output lines can be written before returning. There is also an
4663 smtp_fflush() function to enable the detection of a dropped connection.
4664 The variables smtp_input and smtp_batched_input are exported to
4667 55. Changed the default runtime configuration: the message "Unknown user"
4668 has been removed from the ACL, and instead placed on the localuser router,
4669 using the cannot_route_message feature. This means that any verification
4670 failures that generate their own messages won't get overridden. Similarly,
4671 the "Unrouteable address" message that was in the ACL for unverifiable
4672 relay addresses has also been removed.
4674 56. Added hosts_avoid_esmtp to the smtp transport.
4676 57. The exicyclog script was not checking for the esoteric option
4677 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4678 will work only if exicyclog is run under the appropriate euid.
4680 58. Following a discussion on the list, the rules by which Exim recognises line
4681 endings on incoming messages have been changed. The -dropcr and drop_cr
4682 options are now no-ops, retained only for backwards compatibility. The
4683 following line terminators are recognized: LF CRLF CR. However, special
4684 processing applies to CR:
4686 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4687 nor a local message in the state where . is a terminator.
4689 (ii) If a bare CR is encountered in a header line, an extra space is added
4690 after the line terminator so as not to end the header. The reasoning
4691 behind this is that bare CRs in header lines are most likely either
4692 to be mistakes, or people trying to play silly games.
4694 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4695 was being incorrectly given as 18 bytes larger than it should have been.
4696 This is a VOB (very old bug).
4698 60. This may never have affected anything current, but just in case it has:
4699 When the local host is found other than at the start of a list of hosts,
4700 the local host, those with the same MX, and any that follow, are discarded.
4701 When the list in question was part of a longer list of hosts, the following
4702 hosts (not currently being processed) were also being discarded. This no
4703 longer happens. I'm not sure if this situation could ever has previously
4706 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4709 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4710 header. This is something that is recommended in a new Internet Draft, and
4711 is something that is documented as being done by Sendmail. There are two
4712 possible values. For messages generated by the autoreply transport, Exim
4715 Auto-Submitted: auto-replied
4717 whereas for all other generated messages (e.g. bounces) it adds
4719 Auto-Submitted: auto-generated
4721 63. The "personal" condition in filters now includes a test for the
4722 Auto-Submitted: header. If it contains the string "auto-" the message it
4723 not considered personal.
4725 64. Added rcpt_include_affixes as a generic transport option.
4727 65. Added queue_only_override (default true).
4729 66. Added the syslog_duplication option.
4731 67. If what should have been the first header line of a message consisted of
4732 a space followed by a colon, Exim was mis-interpreting it as a header line.
4733 It isn't of course - it is syntactically invalid and should therefore be
4734 treated as the start of the message body. The misbehaviour could have
4735 caused a number of strange effects, including loss of data in subsequent
4736 header lines, and spool format errors.
4738 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4739 client host had authenticated. This control can now be exercised by an ACL
4740 for more flexibility.
4742 69. By default, callouts do not happen when testing with -bh. There is now a
4743 variant, -bhc, which does actually run the callout code, including
4744 consulting and updating the callout cache.
4746 70. Added support for saslauthd authentication, courtesy of Alexander
4749 71. If statvfs() failed on the spool or log directories while checking their
4750 size for availability, Exim confusingly gave the error "space shortage".
4751 Furthermore, in debugging mode it crashed with a floating point exception.
4752 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4753 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4754 serious problem, Exim now writes to the main and panic logs when this
4755 happens, with details of the failure. It then refuses to accept the
4756 incoming message, giving the message "spool directory problem" or "log
4757 directory problem" with a 421 code for SMTP messages.
4759 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4760 0, 1, and 2 exist, because some OS complain for execs without them (see
4761 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4762 descriptors. However, the code omitted to check that the open succeeded,
4763 causing mysterious errors if for some reason the permissions on /dev/null
4764 got screwed. Now Exim writes a message to the main and panic logs, and
4765 bombs out if it can't open /dev/null.
4767 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4768 interact so that it is all more flexible. It is supposed to remain
4769 backwards compatible. Also added extra_local_interfaces.
4771 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4772 to bomb out with an assertion failure - to the client this appears as a
4773 connection drop. This problem occurs in the part of the code that was taken
4774 from the Samba project. Fortunately, the assertion is in a very simple
4775 function, so I have fixed this by reproducing the function inline in the
4776 one place where it is called, and arranging for authentication to fail
4777 instead of killing the process with assert().
4779 75. The SPA client code was not working when the server requested OEM rather
4780 than Unicode encoding.
4782 76. Added code to make require_files with a specific uid setting more usable in
4783 the case where statting the file as root fails - usually a non-root-mounted
4784 NFS file system. When this happens and the failure is EACCES, Exim now
4785 forks a subprocess and does the per-uid checking as the relevant uid.
4787 77. Added process_log_path.
4789 78. If log_file_path was not explicitly set, a setting of check_log_space or
4790 check_log_inodes was ignored.
4792 79. If a space check for the spool or log partitions fails, the incident is now
4793 logged. Of course, in the latter case the data may get lost...
4795 80. Added the %p formatting code to string_format() so that it can be used to
4796 print addresses in debug_print(). Adjusted all the address printing in the
4797 debugging in store.c to use %p rather than %d.
4799 81. There was a concern that a line of code in smtp_in.c could overflow a
4800 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4801 initially expressed, the concern was not well-founded, because trailing
4802 spaces are removed early. However, if the trailing spaces were followed by
4803 a NULL, they did not get removed, so the overflow was possible. Two fixes
4806 (a) I re-wrote the offending code in a cleaner fashion.
4807 (b) If an incoming SMTP command contains a NULL character, it is rejected
4810 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4811 runs initgroups(), so that if the Exim user is in any additional groups,
4812 they will be used during message reception.
4818 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / users / jgh / exim.git / blob