1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.520 2007/06/27 11:01:51 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.
83 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
84 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
87 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
90 PH/02 In an ACL statement such as
92 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
94 if a client was not listed at all, or was listed with a value other than
95 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
96 the condition was not true (as it should be), so access was not denied.
97 The bug was that the ! inversion was incorrectly passed on to the second
98 item. This has been fixed.
100 PH/03 Added additional dnslists conditions == and =& which are different from
101 = and & when the dns lookup returns more than one IP address.
103 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
104 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
106 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
107 FSYNC, which compiles an option called disable_fsync that allows for
108 bypassing fsync(). The documentation is heavily laced with warnings.
110 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
112 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
113 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
114 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
115 including adding "make clean"; (3) Added -fPIC when compiling the test
116 dynamically loaded module, to get rid of a warning.
118 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
119 message fails, move_frozen_messages = true and ignore_bounce_errors_after
120 = 0s. The bug is otherwise harmless.
122 PH/07 There was a bug in the dovecot authenticator such that the value of
123 $auth1 could be overwritten, and so not correctly preserved, after a
124 successful authentication. This usually meant that the value preserved by
125 the server_setid option was incorrect.
127 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
129 PH/09 Installed PCRE release 7.0.
131 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
132 run for batched SMTP input. It is now run at the start of every message
133 in the batch. While fixing this I discovered that the process information
134 (output by running exiwhat) was not always getting set for -bs and -bS
135 input. This is fixed, and it now also says "batched" for BSMTP.
137 PH/11 Added control=no_pipelining.
139 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
140 patch, slightly modified), and move the expansion of helo_data till after
141 the connection is made in the smtp transport (so it can use these
144 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
146 PH/14 Added log_selector = +pid.
148 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
150 PH/16 Add ${if forany and ${if forall.
152 PH/17 Added dsn_from option to vary the From: line in DSNs.
154 PH/18 Flush SMTP output before performing a callout, unless control =
155 no_callout_flush is set.
157 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
158 was true (the default) a successful delivery failed to delete the retry
159 item, thus causing premature timeout of the address. The bug is now
162 PH/20 Added hosts_avoid_pipelining to the smtp transport.
164 PH/21 Long custom messages for fakedefer and fakereject are now split up
165 into multiline reponses in the same way that messages for "deny" and
166 other ACL rejections are.
168 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
169 with slight modification.
171 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
172 draft, changing the syntax and factoring some duplicate code".
174 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
175 for deliveries of the second and subsequent messages over the same SMTP
178 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
179 ${reduce, with only minor "tidies".
181 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
183 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
184 expansion side effects.
186 PH/27 When a message times out after an over-quota error from an Exim-imposed
187 quota, the bounce message says "mailbox is full". This message was not
188 being given when it was a system quota that was exceeded. It now should
191 MH/03 Made $recipients available in local_scan(). local_scan() already has
192 better access to the recipient list through recipients_list[], but
193 $recipients can be useful in postmaster-provided expansion strings.
195 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
196 in the case of a MAIL command with additional options following the
197 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
198 were accidentally chopped off.
200 PH/29 SMTP synchronization checks are implemented when a command is read -
201 there is a check that no more input is waiting when there shouldn't be
202 any. However, for some commands, a delay in an ACL can mean that it is
203 some time before the response is written. In this time, more input might
204 arrive, invalidly. So now there are extra checks after an ACL has run for
205 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
206 pipelining has not been advertised.
208 PH/30 MH's patch to allow iscntrl() characters to be list separators.
210 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
211 returned in the SMTP response when smtp_return_error_details was false.
214 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
215 instead of the C I/O that was originally supplied, because problems were
218 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
219 Exim which did not show up earlier: it was assuming that a call to
220 SSL_CTX_set_info_callback() might give an error value. In fact, there is
221 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
222 was a macro that became an assignment, so it seemed to work. This has
223 changed to a proper function call with a void return, hence the compile
224 error. Exim's code has been fixed.
226 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
229 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
231 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
233 PH/37 If a message is not accepted after it has had an id assigned (e.g.
234 because it turns out to be too big or there is a timeout) there is no
235 "Completed" line in the log. When some messages of this type were
236 selected by exigrep, they were listed as "not completed". Others were
237 picked up by some special patterns. I have improved the selection
238 criteria to be more general.
240 PH/38 The host_find_failed option in the manualroute router can now be set
241 to "ignore", to completely ignore a host whose IP address cannot be
242 found. If all hosts are ignored, the behaviour is controlled by the new
243 host_all_ignored option.
245 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
246 homing or because of multiple MX records with /mx) generated more than
247 one IP address, and the following item turned out to be the local host,
248 all the secondary addresses of the first item were incorrectly removed
249 from the list, along with the local host and any following hosts (which
250 is what is supposed to happen).
252 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
253 whoever called Exim into the -H file. In the case of the daemon it was
254 behaving confusingly. When first started, it used values for whoever
255 started the daemon, but after a SIGHUP it used the Exim user (because it
256 calls itself on a restart). I have changed the code so that it now always
259 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
260 message are rejected with the same error (e.g. no authentication or bad
261 sender address), and a DATA command is nevertheless sent (as can happen
262 with PIPELINING or a stupid MUA), the error message that was given to the
263 RCPT commands is included in the rejection of the DATA command. This is
264 intended to be helpful for MUAs that show only the final error to their
267 PH/42 Another patch from the Sieve maintainer.
269 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
270 Eximstats - Fixed some broken HTML links and added missing column headers
272 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
273 columns for Rejects, Temp Rejects, Ham, and Spam rows.
275 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
277 PH/43 Yet another patch from the Sieve maintainer.
279 PH/44 I found a way to check for a TCP/IP connection going away before sending
280 the response to the final '.' that terminates a message, but only in the
281 case where the client has not sent further data following the '.'
282 (unfortunately, this is allowed). However, in many cases there won't be
283 any further data because there won't be any more messages to send. A call
284 to select() can be used: if it shows that the input is "ready", there is
285 either input waiting, or the socket has been closed. An attempt to read
286 the next input character can distinguish the two cases. Previously, Exim
287 would have sent an OK response which the client would never have see.
288 This could lead to message repetition. This fix should cure that, at
289 least in a lot of common cases.
291 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
292 advertised in response to EHLO.
298 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
299 fixed by 4.65/MH/01 (is this a record?) are fixed:
301 (i) An empty string was always treated as zero by the numeric comparison
302 operators. This behaviour has been restored.
304 (ii) It is documented that the numeric comparison operators always treat
305 their arguments as decimal numbers. This was broken in that numbers
306 starting with 0 were being interpreted as octal.
308 While fixing these problems I realized that there was another issue that
309 hadn't been noticed. Values of message_size_limit (both the global option
310 and the transport option) were treated as octal if they started with 0.
311 The documentation was vague. These values are now always treated as
312 decimal, and I will make that clear in the documentation.
318 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
319 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
322 MH/01 Don't check that the operands of numeric comparison operators are
323 integers when their expansion is in "skipping" mode (fixes bug
324 introduced by 4.64-PH/07).
326 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
327 child addresses, Exim now panics and dies. Previously, because the count
328 is held in a short int, deliveries were likely to be lost. As such a
329 large number of recipients for a single message is ridiculous
330 (performance will be very, very poor), I have chosen to impose a limit
331 rather than extend the field.
337 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
338 leftover -K file (the existence of which was triggered by #402).
339 While we were at it, introduced process PID as part of the -K
340 filename. This should rule out race conditions when creating
343 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
344 processing considerably. Previous code took too long for large mails,
345 triggering a timeout which in turn triggers #401.
347 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
348 in the DK code in transports.c. sendfile() is not really portable,
349 hence the _LINUX specificness.
351 TF/01 In the add_headers option to the mail command in an Exim filter,
352 there was a bug that Exim would claim a syntax error in any
353 header after the first one which had an odd number of characters
356 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
357 callout verification, Exim cached a "reject" for the entire domain. This
358 is correct for most verifications, but it is not correct for a recipient
359 verification with use_sender or use_postmaster set, because in that case
360 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
361 case of MAIL FROM:<> rejection from other early rejections (e.g.
362 rejection of HELO). When verifying a recipient using a non-null MAIL
363 address, the cache is ignored if it shows MAIL FROM:<> rejection.
364 Whatever the result of the callout, the value of the domain cache is
365 left unchanged (for any other kind of callout, getting as far as trying
366 RCPT means that the domain itself is ok).
368 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
371 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
372 manifest itself as EPIPE rather than ECONNECT. When tidying away a
373 session, the daemon ignores ECONNECT errors and logs others; it now
374 ignores EPIPE as well.
376 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
377 (quoted-printable decoding).
379 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
380 later the small subsequent patch to fix an introduced bug.
382 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
384 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
386 PH/08 An error is now given if message_size_limit is specified negative.
388 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
389 to be given (somewhat) arbitrary names.
391 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
394 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
395 miscellaneous code fixes
397 PH/10 Added the log_reject_target ACL modifier to specify where to log
400 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
401 hostname. This is wrong, because it relates to the incoming message (and
402 probably the interface on which it is arriving) and not to the outgoing
403 callout (which could be using a different interface). This has been
404 changed to use the value of the helo_data option from the smtp transport
405 instead - this is what is used when a message is actually being sent. If
406 there is no remote transport (possible with a router that sets up host
407 addresses), $smtp_active_hostname is used.
409 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
410 tweaks were necessary in order to get it to work (see also 21 below):
411 (a) The code assumed that strncpy() returns a negative number on buffer
412 overflow, which isn't the case. Replaced with Exim's string_format()
414 (b) There were several signed/unsigned issues. I just did the minimum
415 hacking in of casts. There is scope for a larger refactoring.
416 (c) The code used strcasecmp() which is not a standard C function.
417 Replaced with Exim's strcmpic() function.
418 (d) The code set only $1; it now sets $auth1 as well.
419 (e) A simple test gave the error "authentication client didn't specify
420 service in request". It would seem that Dovecot has changed its
421 interface. Fortunately there's a specification; I followed it and
422 changed what the client sends and it appears to be working now.
424 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
427 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
428 address A is aliased to B and C, where B exists and C does not. Without
429 -v the output is "A verified" because verification stops after a
430 successful redirection if more than one address is generated. However,
431 with -v the child addresses are also verified. Exim was outputting "A
432 failed to verify" and then showing the successful verification for C,
433 with its parentage. It now outputs "B failed to verify", showing B's
434 parentage before showing the successful verification of C.
436 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
437 look up a TXT record in a specific list after matching in a combined
440 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
441 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
442 they consult the DNS. I had assumed they would set it the way they
443 wanted; and indeed my experiments on Linux seem to show that in some
444 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
445 To be on the safe side, however, I have now made the interface to
446 host_find_byname() similar to host_find_bydns(), with an argument
447 containing the DNS resolver options. The host_find_byname() function now
448 sets these options at its start, just as host_find_bydns() does. The smtp
449 transport options dns_qualify_single and dns_search_parents are passed to
450 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
451 of host_find_byname() use the default settings of RES_DEFNAMES
452 (qualify_single) but not RES_DNSRCH (search_parents).
454 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
455 spool_read_header() do less string testing, by means of a preliminary
456 switch on the second character of optional "-foo" lines. (This is
457 overdue, caused by the large number of possibilities that now exist.
458 Originally there were few.) While I was there, I also converted the
459 str(n)cmp tests so they don't re-test the leading "-" and the first
460 character, in the hope this might squeeze out yet more improvement.
462 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
463 flag allowing group syntax was set by the header_syntax check but not
464 turned off, possible causing trouble later; (2) The flag was not being
465 set at all for the header_verify test, causing "group"-style headers to
466 be rejected. I have now set it in this case, and also caused header_
467 verify to ignore an empty address taken from a group. While doing this, I
468 came across some other cases where the code for allowing group syntax
469 while scanning a header line wasn't quite right (mostly, not resetting
470 the flag correctly in the right place). These bugs could have caused
471 trouble for malformed header lines. I hope it is now all correct.
473 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
474 with the "reply" argument non-NULL. The code, however (which originally
475 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
476 but it didn't always do it. This confused somebody who was copying the
477 code for some other use. I have removed all the tests.
479 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
480 feature that was used to support insecure browsers during the U.S. crypto
481 embargo. It requires special client support, and Exim is probably the
482 only MTA that supported it -- and would never use it because real RSA is
483 always available. This code has been removed, because it had the bad
484 effect of slowing Exim down by computing (never used) parameters for the
485 RSA_EXPORT functionality.
487 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
488 authenticator to fail if there's a tab character in the incoming data
489 (there should never be unless someone is messing about, as it's supposed
490 to be base64-encoded). Also added, on Timo's advice, the "secured" option
491 if the connection is using TLS or if the remote IP is the same as the
492 local IP, and the "valid-client-cert option" if a client certificate has
495 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
496 authenticators. This can be used for authorization after authentication
497 succeeds. (In the case of plaintext, it servers for both authentication
500 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
501 if any retry times were supplied.
503 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
504 connection, where there is no client IP address to check. In this
505 situation, the verify now always succeeds.
507 PH/25 Applied John Jetmore's -Mset patch.
509 PH/26 Added -bem to be like -Mset, but loading a message from a file.
511 PH/27 In a string expansion for a processed (not raw) header when multiple
512 headers of the same name were present, leading whitespace was being
513 removed from all of them, but trailing whitespace was being removed only
514 from the last one. Now trailing whitespace is removed from each header
515 before concatenation. Completely empty headers in a concatenation (as
518 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
519 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
521 PH/29 [Removed. This was a change that I later backed out, and forgot to
522 correct the ChangeLog entry (that I had efficiently created) before
523 committing the later change.]
525 PH/30 Exim was sometimes attempting to deliver messages that had suffered
526 address errors (4xx response to RCPT) over the same connection as other
527 messages routed to the same hosts. Such deliveries are always "forced",
528 so retry times are not inspected. This resulted in far too many retries
529 for the affected addresses. The effect occurred only when there were more
530 hosts than the hosts_max_try setting in the smtp transport when it had
531 the 4xx errors. Those hosts that it had tried were not added to the list
532 of hosts for which the message was waiting, so if all were tried, there
533 was no problem. Two fixes have been applied:
535 (i) If there are any address or message errors in an SMTP delivery, none
536 of the hosts (tried or untried) are now added to the list of hosts
537 for which the message is waiting, so the message should not be a
538 candidate for sending over the same connection that was used for a
539 successful delivery of some other message. This seems entirely
540 reasonable: after all the message is NOT "waiting for some host".
541 This is so "obvious" that I'm not sure why it wasn't done
542 previously. Hope I haven't missed anything, but it can't do any
543 harm, as the worst effect is to miss an optimization.
545 (ii) If, despite (i), such a delivery is accidentally attempted, the
546 routing retry time is respected, so at least it doesn't keep
547 hammering the server.
549 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
550 in ${readsocket because some servers need this prod.
552 PH/32 Added some extra debug output when updating a wait-xxx database.
554 PH/33 The hint "could be header name not terminated by colon", which has been
555 given for certain expansion errors for a long time, was not being given
556 for the ${if def:h_colon_omitted{... case.
558 PH/34 The spec says: "With one important exception, whenever a domain list is
559 being scanned, $domain contains the subject domain." There was at least
560 one case where this was not true.
562 PH/35 The error "getsockname() failed: connection reset by peer" was being
563 written to the panic log as well as the main log, but it isn't really
564 panic-worthy as it just means the connection died rather early on. I have
565 removed the panic log writing for the ECONNRESET error when getsockname()
568 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
569 runs only) independently of the message's sender address. This meant
570 that, if the 4xx error was in fact related to the sender, a different
571 message to the same recipient with a different sender could confuse
572 things. In particualar, this can happen when sending to a greylisting
573 server, but other circumstances could also provoke similar problems.
574 I have changed the default so that the retry time for these errors is now
575 based a combination of the sender and recipient addresses. This change
576 can be overridden by setting address_retry_include_sender=false in the
579 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
580 remote server are returned as part of bounce messages. This was not
581 happening for LMTP over a pipe (the lmtp transport), but now it is the
582 same for both kinds of LMTP.
584 PH/38 Despite being documented as not happening, Exim was rewriting addresses
585 in header lines that were in fact CNAMEs. This is no longer the case.
587 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
588 and queue runs started by the daemon processed all messages. This has
589 been fixed so that -R and -S can now usefully be given with -q<time>.
591 PH/40 Import PCRE release 6.7 (fixes some bugs).
593 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
595 PH/42 Give an error if -q is specified more than once.
597 PH/43 Renamed the variables $interface_address and $interface_port as
598 $received_ip_address and $received_port, to make it clear that these
599 values apply to message reception, and not to the outgoing interface when
600 a message is delivered. (The old names remain recognized, of course.)
602 PH/44 There was no timeout on the connect() call when using a Unix domain
603 socket in the ${readsocket expansion. There now is.
605 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
606 be meaningful with "accept".
608 SC/01 Eximstats V1.43
609 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
611 SC/02 Eximstats V1.44
612 Use a glob alias rather than an array ref in the generated
613 parser. This improves both readability and performance.
615 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
616 Collect SpamAssassin and rejection statistics.
617 Don't display local sender or destination tables unless
618 there is data to show.
619 Added average volumes into the top table text output.
621 SC/04 Eximstats V1.46
622 Collect data on the number of addresses (recipients)
623 as well as the number of messages.
625 SC/05 Eximstats V1.47
626 Added 'Message too big' to the list of mail rejection
627 reasons (thanks to Marco Gaiarin).
629 SC/06 Eximstats V1.48
630 Mainlog lines which have GMT offsets and are too short to
631 have a flag are now skipped.
633 SC/07 Eximstats V1.49 (Alain Williams)
634 Added the -emptyok flag.
636 SC/08 Eximstats V1.50
637 Fixes for obtaining the IP address from reject messages.
639 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
640 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
641 whitesspace changes from 4.64-PH/27
643 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
646 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
647 are found, allow negative numbers in numeric criteria)
649 JJ/06 exipick.20061117.2, added new $message_body_missing variable
651 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
652 to match changes made in 4.64-PH/43
654 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
656 PH/47 Put in an explicit test for a DNS lookup of an address record where the
657 "domain" is actually an IP address, and force a failure. This locks out
658 those revolvers/nameservers that support "A-for-A" lookups, in
659 contravention of the specifications.
661 PH/48 When a host name was looked up from an IP address, and the subsequent
662 forward lookup of the name timed out, the host name was left in
663 $sender_host_name, contrary to the specification.
665 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
666 restricted to single-key lookups, Exim was not diagnosing an error if
667 * or *@ was used with a query-style lookup.
669 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
671 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
672 long ago, but noone interested enough thought of it. Let's just say that
673 the "1.1" means that there are some new functions that weren't there at
674 some point in the past.
676 PH/51 Error processing for expansion failure of helo_data from an smtp
677 transport during callout processing was broken.
679 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
680 tested/used via the -bh/-bhc/-bs options.
682 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
683 bug, fixed in subsequent PCRE releases).
685 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
686 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
688 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
694 SC/01 Use a glob alias rather than an array ref in eximstats generated
695 parser. This improves both readability and performance.
697 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
698 Don't display local sender or destination tables in eximstats unless
699 there is data to show.
700 Added average volumes into the eximstats top table text output.
702 SC/03 Collect data on the number of addresses (recipients) as well
703 as the number of messages in eximstats.
705 TF/01 Correct an error in the documentation for the redirect router. Exim
706 does (usually) call initgroups() when daemonizing.
708 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
709 with consistent privilege compared to when running as a daemon.
711 TF/03 Note in the spec that $authenticated_id is not set for local
712 submissions from trusted users.
714 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
715 Thanks to Dean Brooks <dean@iglou.com> for the patch.
717 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
718 by adding some example configuration directives to the default
719 configuration file. A little bit of work is required to uncomment the
720 directives and define how usernames and passwords are checked, but
721 there is now a framework to start from.
723 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
724 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
725 without this. I don't know how relevant this is to other LDAP libraries.
727 PH/02 Add the verb name to the "unknown ACL verb" error.
729 PH/03 Magnus Holmgren's patch for filter_prepend_home.
731 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
733 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
734 directory not expanded when it should be if an expanded home directory
735 was set for the address (which is overridden by the transport).
737 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
740 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
741 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
742 because it is too late at that time, and has no effect.
744 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
745 security issue with \' (bugzilla #107). I could not use the
746 PQescapeStringConn() function, because it needs a PGconn value as one of
749 PH/08 When testing addresses using -bt, indicate those final addresses that
750 are duplicates that would not cause an additional delivery. At least one
751 person was confused, thinking that -bt output corresponded to deliveries.
752 (Suppressing duplicates isn't a good idea as you lose the information
753 about possibly different redirections that led to the duplicates.)
755 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
756 systems where poll() doesn't work, in particular OS X.
758 PH/10 Added more information to debugging output for retry time not reached.
760 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
761 operations in malware.c.
763 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
766 PH/13 If write_rejectlog was set false when logging was sent to syslog with
767 syslog_duplication set false, log lines that would normally be written
768 both the the main log and to the reject log were not written to syslog at
771 PH/14 In the default configuration, change the use of "message" in ACL warn
772 statements to "add_header".
774 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
775 not followed by a command (e.g. "seen endif").
777 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
778 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
781 PH/17 Added extra conditions to the default value of delay_warning_condition
785 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
786 { match{$h_precedence:}{(?i)bulk|list|junk} } \
787 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
790 The Auto-Submitted: and various List- headers are standardised, whereas I
791 don't think Precedence: ever was.
793 PH/18 Refactored debugging code in route_finduser() to show more information,
794 in particular, the error code if getpwnam() issues one.
796 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
797 This is apparently needed in addition to the PH/07 change above to avoid
798 any possible encoding problems.
800 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
801 but not after initializing Perl.
803 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
804 output them only if debugging. By default they are written stderr,
805 apparently, which is not desirable.
807 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
810 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
813 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
815 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
816 authenticated or an ident call has been made. Suppress the default
817 values for $authenticated_id and $authenticated_sender (but permit -oMai
818 and -oMas) when testing with -bh.
820 PH/24 Re-jigged the order of the tests in the default configuration so that the
821 tests for valid domains and recipients precede the DNS black list and CSA
822 tests, on the grounds that those ones are more expensive.
824 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
825 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
826 This bug exists in every version of Exim that I still have, right back to
829 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
830 However, an attempt to turn on case-sensitivity in a regex key by
831 including (?-i) didn't work because the subject string was already
832 lowercased, and the effects were non-intuitive. It turns out that a
833 one-line patch can be used to allow (?-i) to work as expected.
839 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
840 other effects) broke the use of negated acl sub-conditions.
842 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
845 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
846 "Deny" causes Exim to reject the incoming connection with a 554 error.
847 Unfortunately, if there is a major crisis, such as a disk failure,
848 tcp-wrappers gives "deny", whereas what one would like would be some
849 kind of temporary error. A kludge has been added to help with this.
850 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
851 554 error is used if errno is still zero or contains ENOENT (which occurs
852 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
855 PH/03 Add -lutil to the default FreeBSD LIBS setting.
857 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
858 errors. Otherwise a message that provokes a temporary error (when other
859 messages do not) can cause a whole host to time out.
861 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
862 addresses were routed directly to files or pipes from a redirect router.
863 File deliveries just didn't batch; pipe deliveries might have suffered
866 PH/06 A failure to get a lock for a hints database would erroneously always say
867 "Failed to get write lock", even when it was really a read lock.
869 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
870 of 0600. This has been changed to use the value of the lockfile_mode
871 option (which defaults to 0600).
873 PH/08 Applied small patch from the Sieve maintainer.
875 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
876 folder from quota calculations, a direct delivery into this folder messed
877 up the contents of the maildirsize file. This was because the regex was
878 used only to exclude .Trash (or whatever) when the size of the mailbox
879 was calculated. There was no check that a delivery was happening into an
880 excluded directory. This bug has been fixed by ignoring all quota
881 processing for deliveries into excluded directories.
883 PH/10 Added the maildirfolder_create_regex option to appendfile.
889 PH/01 The code for finding all the local interface addresses on a FreeBSD
890 system running IPv6 was broken. This may well have applied to all BSD
891 systems, as well as to others that have similar system calls. The broken
892 code found IPv4 interfaces correctly, but gave incorrect values for the
893 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
894 that it would not match correctly against @[] and not recognize the IPv6
897 PH/02 The ipliteral router was not recognizing addresses of the form user@
898 [ipv6:....] because it didn't know about the "ipv6:" prefix.
900 PH/03 Added disable_ipv6.
902 PH/04 Changed $reply_address to use the raw form of the headers instead of the
903 decoded form, because it is most often used to construct To: headers
904 lines in autoreplies, and the decoded form may well be syntactically
905 invalid. However, $reply_address has leading white space removed, and all
906 newlines turned into spaces so that the autoreply transport does not
909 PH/05 If group was specified without a user on a router, and no group or user
910 was specified on a transport, the group from the router was ignored.
912 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
913 for visible compile-time settings that can be used to change these
914 numbers, for those that want even more. Backwards compatibility with old
915 spool files has been maintained. However, going back to a previous Exim
916 release will lost any variables that are in spool files.
918 PH/07 Two small changes when running in the test harness: increase delay when
919 passing a TCP/IP connection to a new process, in case the original
920 process has to generate a bounce, and remove special handling of
921 127.0.0.2 (sic), which is no longer necessary.
923 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
924 be the same on different OS.
926 PH/09 Moved a debug statement in filter processing to avoid a race problem when
929 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
930 whether --show-vars was specified or not
932 JJ/02 exipick: Added support for new ACL variable spool format introduced
935 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
936 syntactically invalid From: or Reply-to: line, and a filter used this to
937 generate an autoreply, and therefore failed to obtain an address for the
938 autoreply, Exim could try to deliver to a non-existent relative file
939 name, causing unrelated and misleading errors. What now happens is that
940 it logs this as a hard delivery error, but does not attempt to create a
943 PH/11 The exinext utility has a -C option for testing purposes, but although
944 the given file was scanned by exinext itself; it wasn't being passed on
945 when Exim was called.
947 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
948 an end-of-file indication when reading a command response.
950 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
951 compiled. In many other places in Exim, IPv6 addresses are always
952 recognized, so I have changed this. It also means that IPv4 domain
953 literals of the form [IPV4:n.n.n.n] are now always recognized.
955 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
956 used if the router is not running as root, for example, when verifying at
957 ACL time, or when using -bh. The debugging output from this situation was
958 non-existent - all you got was a failure to exec. I have made two
961 (a) Failures to set uid/gid, the current directory, or a process leader
962 in a subprocess such as that created by queryprogram now generate
963 suitable debugging ouput when -d is set.
965 (b) The queryprogram router detects when it is not running as root,
966 outputs suitable debugging information if -d is set, and then runs
967 the subprocess without attempting to change uid/gid.
969 PH/15 Minor change to Makefile for building test_host (undocumented testing
972 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
973 additional section of a DNS packet that returns MX or SRV records.
974 Instead, it always explicitly searches for A/AAAA records. This avoids
975 major problems that occur when a DNS server includes only records of one
976 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
977 fixed another bug: if SRV records were looked up and the corresponding
978 address records were *not* found in the additional section, the port
979 values from the SRV records were lost.
981 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
982 using the correct key (the original address) when searching the retry
983 rules in order to find which one to use for generating the retry hint.
985 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
986 adding the default one. Similarly, if it contains a Reply-To: header, the
987 errors_reply_to option, if set, is not used.
989 PH/19 When calculating a retry time, Exim used to measure the "time since
990 failure" by looking at the "first failed" field in the retry record. Now
991 it does not use this if it is later than than the arrival time of the
992 message. Instead it uses the arrival time. This makes for better
993 behaviour in cases where some deliveries succeed, thus re-setting the
994 "first failed" field. An example is a quota failure for a huge message
995 when small messages continue to be delivered. Without this change, the
996 "time since failure" will always be short, possible causing more frequent
997 delivery attempts for the huge message than are intended.
998 [Note: This change was subsequently modified - see PH/04 for 4.62.]
1000 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
1001 $1, $2, $3) because the numerical variables can be reset during some
1002 expansion items (e.g. "match"), thereby losing the authentication data.
1004 PH/21 Make -bV show the size of off_t variables so that the test suite can
1005 decide whether to run tests for quotas > 2G.
1007 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
1008 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1009 filecount value is greater than 2G or if a quota value is greater than 2G
1010 on a system where the size of off_t is not greater than 4, a panic error
1013 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1014 never match. The debug and -bh output now contains an explicit error
1015 message indicating a malformed IPv4 address or mask.
1017 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1018 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1019 PH/23 above applies.
1021 PH/25 Do not write to syslog when running in the test harness. The only
1022 occasion when this arises is a failure to open the main or panic logs
1023 (for which there is an explicit test).
1025 PH/26 Added the /no_tell option to "control=freeze".
1027 PH/27 If a host name lookup failed very early in a connection, for example, if
1028 the IP address matched host_lookup and the reverse lookup yielded a name
1029 that did not have a forward lookup, an error message of the form "no IP
1030 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1031 could be logged. Now it outputs the IP address instead of "NULL".
1033 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1034 allows the sender and the authenticated sender to be set when
1035 submitting a message from within Exim. Since child_open_exim() is
1036 documented for local_scan(), the new function should be too.
1038 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1039 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1040 results in an empty string is now treated as unset.
1042 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1044 PH/31 Added sender_verify_fail logging option.
1046 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1047 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1048 tidied the source and removed it altogether.
1050 PH/33 When a queue run was abandoned because the load average was too high, a
1051 log line was always written; now it is written only if the queue_run log
1052 selector is set. In addition, the log line for abandonment now contains
1053 information about the queue run such as the pid. This is always present
1054 in "start" and "stop" lines but was omitted from the "abandon" line.
1056 PH/34 Omit spaces between a header name and the colon in the error message that
1057 is given when verify = headers_syntax fails (if there are lots of them,
1058 the message gets confusing).
1060 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1061 names, as there are now some PTR records that contain slashes. This check
1062 is only to protect against broken name servers that fall over on strange
1063 characters, so the fact that it applies to all lookups doesn't matter.
1065 PH/36 Now that the new test suite is complete, we can remove some of the
1066 special code in Exim that was needed for the old test suite. For example,
1067 sorting DNS records because real resolvers return them in an arbitrary
1068 order. The new test suite's fake resolver always returns records in the
1071 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1072 bounces) except when queue_only is set, to avoid logging races between
1073 the different processes.
1075 PH/38 Panic-die if .include specifies a non-absolute path.
1077 PH/39 A tweak to the "H" retry rule from its user.
1079 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1080 a label. They prevented compilation on older perls.
1082 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1083 a warning to be raised on newish perls.
1085 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1086 on queue. Changes to match documented behaviour of showing count of
1087 messages matching specified criteria.
1089 PH/40 Changed the default ident timeout from 30s to 5s.
1091 PH/41 Added support for the use of login_cap features, on those BSD systems
1092 that have them, for controlling the resources used by pipe deliveries.
1094 PH/42 The content-scanning code uses fopen() to create files in which to put
1095 message data. Previously it was not paying any attention to the mode of
1096 the files. Exim runs with umask(0) because the rest of the code creates
1097 files with open(), and sets the required mode explicitly. Thus, these
1098 files were ending up world-writeable. This was not a big issue, because,
1099 being within the spool directory, they were not world-accessible. I have
1100 created a function called modefopen, which takes an additional mode
1101 argument. It sets umask(777), creates the file, chmods it to the required
1102 mode, then resets the umask. All the relevant calls to fopen() in the
1103 content scanning code have been changed to use this function.
1105 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1106 to 24 hours. This avoids potential overflow problems when processing G
1107 and H retry rules. I suspect nobody ever tinkers with this value.
1109 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1111 PH/45 When the plaintext authenticator is running as a client, the server's
1112 challenges are checked to ensure they are valid base64 strings. By
1113 default, the authentication attempt is cancelled if an invalid string is
1114 received. Setting client_ignore_invalid_base64 true ignores these errors.
1115 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1116 they are received. Thus, the responses can be made to depend on the
1117 challenges. If an invalid string is ignored, an empty string is placed in
1120 PH/46 Messages that are created by the autoreply transport now contains a
1121 References: header, in accordance with RFCs 2822 and 3834.
1123 PH/47 Added authenticated_sender_force to the smtp transport.
1125 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1127 PH/49 Installed latest patch from the Sieve maintainer.
1129 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1130 was also set, the appendfile transport was unnecessarily scanning a
1131 directory of message files (e.g. for maildir delivery) to find the count
1132 of files (along with the size), even though it did not need this
1133 information. It now does the scan only if it needs to find either the
1134 size of the count of files.
1136 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1138 PH/52 Two bugs concerned with error handling when the smtp transport is
1141 (i) Exim was not creating retry information for temporary errors given
1142 for individual recipients after the DATA command when the smtp transport
1143 was used in LMTP mode. This meant that they could be retried too
1144 frequently, and not timed out correctly.
1146 (ii) Exim was setting the flag that allows error details to be returned
1147 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1148 recipients that were returned after the DATA command.
1150 PH/53 This is related to PH/52, but is more general: for any failing address,
1151 when detailed error information was permitted to be returned to the
1152 sender, but the error was temporary, then after the final timeout, only
1153 "retry timeout exceeded" was returned. Now it returns the full error as
1154 well as "retry timeout exceeded".
1156 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1157 do this, and (what is worse) MTAs that accept it.
1159 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1160 will now be deprecated.
1162 PH/56 New os.c-cygwin from the Cygwin maintainer.
1164 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1165 formats (previously only available in exim formats via -bpr, -bpru,
1166 and -bpra. Now also available in native and exiqgrep formats)
1168 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1169 with very large, slow to parse queues
1171 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1173 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1175 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1176 responses to authentication challenges, though it was showing the
1177 challenges; (ii) I've removed the CR characters from the debug output for
1180 PH/58 Allow for the insertion of a newline as well as a space when a string
1181 is turned into more than one encoded-word during RFC 2047 encoding. The
1182 Sieve code now uses this.
1184 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1185 data_4xx, lost_connection, tls_required.
1187 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1188 message was being sent as an SMTP response.
1190 PH/61 Add -l and -k options to exicyclog.
1192 PH/62 When verifying, if an address was redirected to one new address, so that
1193 verification continued, and the new address failed or deferred after
1194 having set something in $address_data, the value of $address_data was not
1195 passed back to the ACL. This was different to the case when no
1196 redirection occurred. The value is now passed back in both cases.
1198 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1199 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1200 use login_cap.h, so on its own it isn't the distinguishing feature. The
1201 new name refers directly to the setclassresources() function.
1203 PH/65 Added configuration files for NetBSD3.
1205 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1207 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1208 is preferred over IPv4.
1210 PH/68 The bounce_return_message and bounce_return_body options were not being
1211 honoured for bounces generated during the reception of non-SMTP messages.
1212 In particular, this applied to messages rejected by the ACL. This bug has
1213 been fixed. However, if bounce_return_message is true and bounce_return_
1214 body is false, the headers that are returned for a non-SMTP message
1215 include only those that have been read before the error was detected.
1216 (In the case of an ACL rejection, they have all been read.)
1218 PH/69 The HTML version of the specification is now built in a directory called
1219 spec_html instead of spec.html, because the latter looks like a path with
1220 a MIME-type, and this confuses some software.
1222 PH/70 Catch two compiler warnings in sieve.c.
1224 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1225 function verify_get_ident() calls ip_connect() to connect a socket, but
1226 if the "connect()" function timed out, ip_connect() used to close the
1227 socket. However, verify_get_ident() also closes the socket later, and in
1228 between Exim writes to the log, which may get opened at this point. When
1229 the socket was closed in ip_connect(), the log could get the same file
1230 descriptor number as the socket. This naturally causes chaos. The fix is
1231 not to close the socket in ip_connect(); the socket should be closed by
1232 the function that creates it. There was only one place in the code where
1233 this was missing, in the iplookup router, which I don't think anybody now
1234 uses, but I've fixed it anyway.
1236 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1237 well as to direct DNS lookups. Otherwise the handling of names in host
1238 lists is inconsistent and therefore confusing.
1244 PH/01 Two changes to the default runtime configuration:
1246 (1) Move the checks for relay_from_hosts and authenticated clients from
1247 after to before the (commented out) DNS black list checks.
1249 (2) Add control=submission to the relay_from_hosts and authenticated
1250 clients checks, on the grounds that messages accepted by these
1251 statements are most likely to be submissions.
1253 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1255 (1) Generate an error if the third argument for the ${prvs expansion is
1258 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1261 (3) Reset the variables that are obtained from the first argument of
1262 ${prvscheck and used in the second argument before leaving the code,
1263 because their memory is reclaimed, so using them afterwards may do
1266 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1267 one (it's much easier than Tom thought :-).
1269 (5) Because of (4), we can now allow for the use of $prvscheck_result
1270 inside the third argument.
1272 PH/03 For some reason, the default setting of PATH when running a command from
1273 a pipe transport was just "/usr/bin". I have changed it to
1276 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1277 anything to be listed in the output from -bV.
1279 PH/05 When a filter generated an autoreply, the entire To: header line was
1280 quoted in the delivery log line, like this:
1282 => >A.N.Other <ano@some.domain> <original@ddress> ...
1284 This has been changed so that it extracts the operative address. There
1285 may be more than one such address. If so, they are comma-separated, like
1288 => >ano@some.domain,ona@other.domain <original@ddress> ...
1290 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1291 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1292 not being looked up in the rDNS to get a host name, Exim was showing the
1293 IP address twice in Received: lines, even though the IP addresses were
1294 identical. For example:
1296 Received: from [1.2.3.4] (helo=[1.2.3.4])
1298 However, if the real host name was known, it was omitting the HELO data
1299 if it matched the actual IP address. This has been tidied up so that it
1300 doesn't show the same IP address twice.
1302 PH/07 When both +timestamp and +memory debugging was on, the value given by
1303 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1304 called by the debug printing, thereby overwriting the timestamp buffer.
1305 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1307 PH/08 When the original message was included in an autoreply transport, it
1308 always said "this is a copy of the message, including all the headers",
1309 even if body_only or headers_only was set. It now gives an appropriate
1312 PH/09 Applied a patch from the Sieve maintainer which:
1314 o fixes some comments
1315 o adds the (disabled) notify extension core
1316 o adds some debug output for the result of if/elsif tests
1317 o points to the current vacation draft in the documentation
1318 and documents the missing references header update
1322 o fixes a bug in processing the envelope test (when testing
1323 multiple envelope elements, the last element determinted the
1326 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1327 Electronic Mail") by including:
1329 Auto-submitted: auto-generated
1331 in the messages that it generates (bounce messages and others, such as
1332 warnings). In the case of bounce messages for non-SMTP mesages, there was
1333 also a typo: it was using "Auto_submitted" (underscore instead of
1334 hyphen). Since every message generated by Exim is necessarily in response
1335 to another message, thes have all been changed to:
1337 Auto-Submitted: auto-replied
1339 in accordance with these statements in the RFC:
1341 The auto-replied keyword:
1343 - SHOULD be used on messages sent in direct response to another
1344 message by an automatic process,
1346 - MUST NOT be used on manually-generated messages,
1348 - MAY be used on Delivery Status Notifications (DSNs) and Message
1349 Disposition Notifications (MDNs),
1351 - MUST NOT be used on messages generated by automatic or periodic
1352 processes, except for messages which are automatic responses to
1355 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1356 to the default Received: header definition.
1358 PH/12 Added log selector acl_warn_skipped (default on).
1360 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1361 variables because (a) they are in the wrong storage pool and (b) even if
1362 they were copied, it wouldn't work properly because of the caching.
1364 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1365 checking when decoding. Apparently there are clients that generate
1366 overlong encoded strings. Why am I not surprised?
1368 PH/15 If the first argument of "${if match_address" was not empty, but did not
1369 contain an "@" character, Exim crashed. Now it writes a panic log message
1370 and treats the condition as false.
1372 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1374 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1375 extension "envelope-auth". The code is finished and in agreement with
1376 other implementations, but there is no documentation so far and in fact,
1377 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1378 not changing the active code.
1380 Print executed "if" and "elsif" statements when debugging is used. This
1381 helps a great deal to understand what a filter does.
1383 Document more things not specified clearly in RFC3028. I had all this
1384 sorted out, when out of a sudden new issues came to my mind. Oops."
1386 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1389 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1390 canonical form (as documented). However, after a host name lookup from
1391 the IP address, check_host() was doing a simple string comparison with
1392 addresses acquired from the DNS when checking that the found name did
1393 have the original IP as one of its addresses. Since any found IPv6
1394 addresses are likely to be in abbreviated form, the comparison could
1395 fail. Luckily, there already exists a function for doing the comparison
1396 by converting both addresses to binary, so now that is used instead of
1397 the text comparison.
1399 PH/20 There was another similar case to PH/19, when a complete host name was
1400 given in a host list; looking up its IP address could give an abbreviated
1401 form, whereas the current host's name might or might not be abbreviated.
1402 The same fix has been applied.
1408 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1409 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1412 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1413 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1415 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1417 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1418 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1419 Also, for "Auto-Submitted", treat anything other than "no" as
1420 non-personal, in accordance with RFC 3834. (Previously it treated
1421 anything starting "auto-" as non-personal.)
1423 TF/01 The control=submission/name=... option had a problem with syntax
1424 errors if the name included a slash character. The /name= option
1425 now slurps the rest of the string, so it can include any characters
1426 but it must come last in the list of options (after /sender_retain
1429 PH/05 Some modifications to the interface to the fake nameserver for the new
1437 TK/01 Added the "success_on_redirect" address verification option. See
1438 NewStuff for rationale and an example.
1440 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1442 PH/02 Patch to exigrep to allow it to work on syslog lines.
1444 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1445 fread() to skip over the body file's header line, because in Cygwin the
1446 header line is locked and is inaccessible.
1448 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1449 co-exist for some time) to make it clear that it is the Exim ID that is
1450 referenced, not the Message-ID: header line.
1452 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1453 string_format() function, because snprintf() does not exist on all
1456 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1457 ${stat: expansion item.
1459 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1460 protocol synchronization error", to keep the pedants happy.
1462 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1463 well as for IRIX systems, when gcc is being used. See the host.c source
1466 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1468 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1471 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1472 transport and to the smtp transport in LMTP mode.
1474 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1476 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1477 run a filter in a subprocess. This could lead to confusion in subsequent
1478 lookups in the parent process. There should also be a search_tidyup() at
1479 the end of the subprocess.
1481 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1482 only if the host matched helo_try_verify_hosts, which caused the
1483 verification to occur when the EHLO/HELO command was issued. The ACL just
1484 tested the remembered result. Now, if a previous verification attempt has
1485 not happened, "verify = helo" does it there and then.
1487 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1489 TK/03 Fix log output including CR from clamd.
1491 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1492 exist provoked a memory error which could cause a segfault.
1494 PH/15 Installed PCRE 6.2
1496 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1498 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1499 of the problem. Specifically, suggested +O2 rather than +O1 for the
1502 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1504 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1505 record, the port was not correctly logged when the outgoing_port log
1506 selector was set (it logged the transort's default port).
1508 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1509 fallback_hosts, and "hosts" in the smtp transport.
1511 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1512 host errors such as "Connection refused".
1514 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1515 authentication with radiusclient 0.4.9:
1517 - Error returned from rc_read_config was caught wrongly
1518 - Username/password not passed on to radius server due to wrong length.
1520 The presumption is that some radiusclient API changes for 4.51/PH/17
1521 were not taken care of correctly. The code is still untested by me (my
1522 Linux distribution still has 0.3.2 of radiusclient), but it was
1523 contributed by a Radius user.
1525 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1526 expanding the "port" option of the smtp transport.
1528 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1529 while reading a MIME header. Thanks to Tom Hughes for a patch.
1531 PH/24 Include config.h inside local_scan.h so that configuration settings are
1534 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1535 that in an ACL for RCPT (for example), you can examine exactly what was
1538 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1539 commands, but it was not correctly comparing the address with the actual
1540 client host address. Thus, it would show the EHLO address in Received:
1541 header lines when this was not necessary.
1543 PH/27 Added the % operator to ${eval:}.
1545 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1546 it should be ignoring failures (because with -C, for example, it has lost
1547 privilege). It wasn't ignoring creation failures other than "already
1550 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1553 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1554 because an input line was too long, either on its own, or by virtue of
1555 too many continuations, the temporary file was not being removed, and the
1556 return code was incorrect.
1558 PH/31 Missing "BOOL" in function definition in filtertest.c.
1560 PH/32 Applied Sieve patches from the maintainer.
1562 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1564 PH/33 Added "verify = not_blind".
1566 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1567 Local/Makefile (with some defaults set). These are used in built scripts
1568 such as exicyclog, but they have never been used in the exim_install
1569 script (though there are many overriding facilities there). I have
1570 arranged that the exim_install script now takes note of these two
1573 PH/35 Installed configuration files for Dragonfly.
1575 PH/36 When a locally submitted message by a trusted user did not contain a
1576 From: header, and the sender address was obtained from -f or from an SMTP
1577 MAIL command, and the trusted user did not use -F to supply a sender
1578 name, $originator_name was incorrectly used when constructing a From:
1579 header. Furthermore, $originator_name was used for submission mode
1580 messages from external hosts without From: headers in a similar way,
1581 which is clearly wrong.
1583 PH/37 Added control=suppress_local_fixups.
1585 PH/38 When log_selector = +received_sender was set, and the addition of the
1586 sender made the log line's construction buffer exactly full, or one byte
1587 less than full, an overflow happened when the terminating "\n" was
1590 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1591 when the result of a list match is failure because a DNS lookup failed.
1593 PH/40 RM_COMMAND is now used in the building process.
1595 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1596 the "build-* directories that it finds.
1598 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1599 address in a domain literal was a prefix of an interface address.
1601 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1602 when verifying a sender address, unless rewrite_headers is false.
1604 PH/44 Wrote a long comment about why errors_to addresses are verified as
1605 recipients, not senders.
1607 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1608 the ratelimit ACL was added.
1610 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1612 PH/46 Added extra information about PostgreSQL errors to the error string.
1614 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1615 avoiding the need to install special zones in a real server. This is
1616 backwards compatible; if it can't find the fake resolver, it drops back.
1617 Thus, both old and new test suites can be run.
1619 TF/02 Added util/ratelimit.pl
1621 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1622 clock is set back in time.
1624 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1625 Candler <B.Candler@pobox.com>.
1627 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1628 OK for addresses that are the result of redirections.
1630 PH/48 A number of further additions for the benefit of the new test suite,
1631 including a fake gethostbyname() that interfaces to the fake DNS resolver
1634 TF/06 The fix for widen_domains has also been applied to qualify_single and
1635 search_parents which are the other dnslookup options that can cause
1638 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1641 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1643 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1644 service" if it encountered an MX record with an empty target hostname.
1645 The message is now "an MX or SRV record indicated no SMTP service".
1647 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1648 if the DNS of the sending site is misconfigured. This is quite a
1649 common situation. This change restores the behaviour of treating a
1650 helo verification defer as a failure.
1652 PH/51 If self=fail was set on a router, the bounce message did not include the
1653 actual error message.
1659 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1661 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1662 command itself ended in error, the underlying message about the transport
1663 filter timeout was being overwritten with the pipe command error. Now the
1664 underlying error message should be appended to the second error message.
1666 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1668 PH/02 Reduce the amount of output that "make" produces by default. Full output
1669 can still be requested.
1671 PH/03 The warning log line about a condition test deferring for a "warn" verb
1672 was being output only once per connection, rather than after each
1673 occurrence (because it was using the same function as for successful
1674 "warn" verbs). This seems wrong, so I have changed it.
1676 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1677 it should not have, which might have caused a crash in the right
1678 circumstances, but probably never did.
1680 PH/04 Installed a modified version of Tony Finch's patch to make submission
1681 mode fix the return path as well as the Sender: header line, and to
1682 add a /name= option so that you can make the user's friendly name appear
1685 TF/03 Added the control = fakedefer ACL modifier.
1687 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1688 Mark Lowes for thorough testing.
1690 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1692 TK/03 Merged latest SRS patch from Miles Wilton.
1694 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1695 with the definition in sysexits.h (which is #included earlier).
1696 Fortunately, Exim does not actually use EX_OK. The code used to try to
1697 preserve the sysexits.h value, by assumimg that macro definitions were
1698 scanned for macro replacements. I have been disabused of this notion,
1699 so now the code just undefines EX_OK before #including unistd.h.
1701 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1702 in the smtp transport. When a block could not be written in a single
1703 write() function, the timeout was being re-applied to each part-write.
1704 This seems wrong - if the receiver was accepting one byte at a time it
1705 would take for ever. The timeout is now adjusted when this happens. It
1706 doesn't have to be particularly precise.
1708 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1709 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1711 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1712 without a domain if the check to <postmaster@domain> fails.
1714 SC/01 Eximstats: added -xls and the ability to specify output files
1715 (patch written by Frank Heydlauf).
1717 SC/02 Eximstats: use FileHandles for outputing results.
1719 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1721 SC/04 Eximstats: fixed display of large numbers with -nvr option
1723 SC/05 Eximstats: fixed merging of reports with empty tables.
1725 SC/06 Eximstats: added the -include_original_destination flag
1727 SC/07 Eximstats: removed tabs and trailing whitespace.
1729 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1731 TK/06 MBOX spool code: Add real "From " MBOX separator line
1732 so the .eml file is really in mbox format (even though
1733 most programs do not really care). Patch from Alex Miller.
1735 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1736 The latter is generated from $received_to and is only set if the
1737 message has one envelope recipient. SA can use these headers,
1738 obviously out-of-the-box. Patch from Alex Miller.
1740 PH/08 The ${def test on a variable was returning false if the variable's
1741 value was "0", contrary to what the specification has always said!
1742 The result should be true unless the variable is empty.
1744 PH/09 The syntax error of a character other than { following "${if
1745 def:variable_name" (after optional whitespace) was not being diagnosed.
1746 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1747 accidental colon was present, for example, could give incorrect results.
1749 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1750 result is used (not including appendfile, where other changes are about
1753 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1754 This involved changing a lot of size variables from int to off_t. It
1755 should work with maildirs and everything.
1757 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1758 spamd dying while we are connected to it.
1760 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1763 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1764 function for BDB 4.3.
1766 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1768 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1769 the sources for PCRE have been re-arranged and more files are now
1772 PH/15 The code I had for printing potentially long long variables in PH/11
1773 above was not the best (it lost precision). The length of off_t variables
1774 is now inspected at build time, and an appropriate printing format (%ld
1775 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1776 to be "long long int" or "long int". This is needed for the internal
1777 formatting function string_vformat().
1779 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1780 the configuration file to be ":syslog", then the script "guesses" where
1781 the logs files are, rather than using the compiled in default. In our
1782 case the guess is not the same as the compiled default, so the script
1783 suddenly stopped working when I started to use syslog. The patch checks
1784 to see if log_file_path is "". If so, it attempts to read it from exim
1785 with no configuration file to get the compiled in version, before it
1786 falls back to the previous guessing code."
1788 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1789 implementing BATV in an Exim configuration. See NewStuff for the gory
1792 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1793 Makefile that are specific to HP-UX.
1795 PH/18 If the "use_postmaster" option was set for a recipient callout together
1796 with the "random" option, the postmaster address was used as the MAIL
1797 FROM address for the random test, but not for the subsequent recipient
1798 test. It is now used for both.
1800 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1801 patch removes a few documentation additions to RFC 3028, because the
1802 latest draft now contains them. It adds the new en;ascii-case comparator
1803 and a new error check for 8bit text in MIME parts. Comparator and
1804 require names are now matched exactly. I enabled the subaddress
1805 extension, but it is not well tested yet (read: it works for me)."
1807 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1808 rework some of the code of TK/09 above to avoid the hardwired use of
1809 "%lld" and "long long". Replaced the call to snprintf() with a call to
1812 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1813 records point to non-existent hosts", "retry timeout exceeded", and
1814 "retry time not reached for any host after a long failure period".
1816 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1817 experimental DomainKeys support:
1819 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1820 (2) On an error such as an illegally used "control", the wrong name for
1821 the control was given.
1823 These problems did NOT occur unless DomainKeys support was compiled.
1825 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1827 PH/24 Added ${if match_ip condition.
1829 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1830 databases so that it will be absolutely obvious if a crash occurs in the
1831 DB library. This is a regular occurrence (often caused by mis-matched
1834 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1835 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1836 picked up on a user's system that detects such things. There doesn't seem
1837 to be a gcc warning option for this - only an attribute that has to be
1838 put on the function's prototype. It seems that in Fedora Core 4 they have
1839 set this on a number of new functions. No doubt there will be more in due
1842 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1843 specify a transport. However, if an address that was verified by such a
1844 router was the subject of a callout, Exim crashed because it tried to
1845 read the rcpt_include_affixes from the non-existent transport. Now it
1846 just assumes that the setting of that option is false. This bug was
1847 introduced by 4.51/PH/31.
1849 PH/28 Changed -d+all to exclude +memory, because that information is very
1850 rarely of interest, but it makes the output a lot bigger. People tend to
1851 do -d+all out of habit.
1853 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1854 code in os-type was giving problems when libc.so lives in lib64, like on
1857 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1858 aren't the modern standard, and it seems that some systems' include files
1859 don't always have them. Exim was already checking for some of the newer
1860 ones like T_AAAA, and defining it itself. I've added checks for all the
1861 record types that Exim uses.
1863 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1864 not automatically generating a new one, as it is supposed to. This
1865 prevented TLS from working. If the file did exist, but contained invalid
1866 data, a new version was generated, as expected. It was only the case of a
1867 non-existent file that was broken.
1869 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1870 with a change in libdomainkeys > 0.64.
1872 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1873 from DNS. If the selector record carries the flag, it now has
1874 precedence over the domain-wide flag.
1876 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1878 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1879 the use of a port name that isn't defined in /etc/services) occurred, the
1880 message was deferred as in a normal delivery, and thus remained on the
1881 spool, instead of being failed because of the mua_wrapper setting. This
1882 is now fixed, and I tidied up some of the mua_wrapper messages at the
1885 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1886 the messages in a hash of arrays rather than using individual hashes.
1887 This is a bit cleaner and results in dramatic memory savings, albeit
1888 at a slight CPU cost.
1890 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1891 as requested by Marc Sherman.
1893 SC/10 Eximstats: added histograms for user specified patterns as requested
1896 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1898 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1899 fopen() in the content-scanning modules that did not already have it.
1905 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1906 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1908 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1910 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1912 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1913 to test Sieve filters that use "vacation".
1915 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1916 that changes the way the GnuTLS parameters are stored in the cache file.
1917 The new format can be generated externally. For backward compatibility,
1918 if the data in the cache doesn't make sense, Exim assumes it has read an
1919 old-format file, and it generates new data and writes a new file. This
1920 means that you can't go back to an older release without removing the
1923 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1924 work if there are any delivery delays because "one_time" forces the
1925 parent to be marked "delivered", so its unseen clone is never tried
1926 again. For this reason, Exim now forbids the simultaneous setting of
1929 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1930 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1931 entry if you want to know the details. The fix, however, neglected to
1932 consider the case when local delivery batching is involved. The test for
1933 "previously delivered" was not happening when checking to see if an
1934 address could be batched with a previous (undelivered) one; under
1935 certain circumstances this could lead to multiple deliveries to the same
1938 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1939 in its include files, and this causes problems building Exim.
1941 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1942 header_syntax) but Exim was just ignoring anything given after a slash.
1943 In particular, this caused confusion with an attempt to use "verify =
1944 reverse_host_lookup/defer_ok". An error is now given when options are
1945 supplied for verify items that do not have them. (Maybe reverse_host_
1946 lookup should have a defer_ok option, but that's a different point.)
1948 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1949 defined by RFC 821) to 2048, because there were problems with some AUTH
1950 commands, and RFC 1869 says the size should be increased for extended
1951 SMTP commands that take arguments.
1953 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1956 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1957 "unknown" error; now it says that the functionality isn't in the binary.
1959 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1960 an address' error message when a string expansion fails (syntax or
1961 whatever). Otherwise the password may appear in the log. Following change
1962 PH/42 below, there is no longer a chance of it appearing in a bounce
1965 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1967 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1968 of its addresses was ever tried. (Bugzilla bug #2.)
1970 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1971 the result incorrectly in the debug output. (It correctly added a newline
1972 to what was transported.)
1974 TF/01 Added $received_time.
1976 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1977 commented out examples of how to interface to a virus scanner and to
1978 SpamAssassin. Also added commented examples of av_scanner and
1979 spamd_address settings.
1981 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1982 and controls are allowed in which ACLs. There were a couple of minor
1983 errors. Some of the entries in the conditions table (which is a table of
1984 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1985 negation of where the condition IS allowed.
1987 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1989 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1990 header file does not have a version number, so I've had to invent a new
1991 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1992 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1993 radiusclient), but it was contributed by a Radius user.
1995 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1996 files or directories, for OpenSSL.
1998 PH/19 When an Exim process that is running as root has to create an Exim log
1999 file, it does so in a subprocess that runs as exim:exim so as to get the
2000 ownership right at creation (otherwise, other Exim processes might see
2001 the file with the wrong ownership). There was no test for failure of this
2002 fork() call, which would lead to the process getting stuck as it waited
2003 for a non-existent subprocess. Forks do occasionally fail when resources
2004 run out. I reviewed all the other calls to fork(); they all seem to check
2007 PH/20 When checking for unexpected SMTP input at connect time (before writing
2008 the banner), Exim was not dealing correctly with a non-positive return
2009 from the read() function. If the client had disconnected by this time,
2010 the result was a log entry for a synchronization error with an empty
2011 string after "input=" when read() returned zero. If read() returned -1
2012 (an event I could not check), uninitialized data bytes were printed.
2013 There were reports of junk text (parts of files, etc) appearing after
2016 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2018 PH/22 Added support for macro redefinition, and (re)definition in between
2019 driver and ACL definitions.
2021 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2022 forgetting to use the resulting value; it was using the unexpanded value.
2024 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2025 hadn't been configured. The fix is from Juergen Kreileder, who
2026 understands it better than I do:
2028 "Here's what I see happening with three configured cyrus_sasl
2029 authenticators configured (plain, login, cram-md5):
2031 On startup auth_cyrus_sasl_init() gets called for each of these.
2032 This means three calls to sasl_listmech() without a specified mech_list.
2033 => SASL tests which mechs of all available mechs actually work
2034 => three warnings about OTP not working
2035 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2037 With the patch, sasl_listmech() also gets called three times. But now
2038 SASL's mech_list option is set to the server_mech specified in the the
2039 authenticator. Or in other words, the answer from sasl_listmech()
2040 gets limited to just the mech you're testing for (which is different
2042 => the return list contains just 'plain' or 'login', 'cram-md5' or
2043 nothing depending on the value of ob->server_mech.
2045 I've just tested the patch: Authentication still works fine,
2046 unavailable mechs specified in the exim configuration are still
2047 caught, and the auth.log warnings about OTP are gone."
2049 PH/25 When debugging is enabled, the contents of the command line are added
2050 to the debugging output, even when log_selector=+arguments is not
2053 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2054 answer is "GNU", and only if the return is "GNU/something" is the answer
2057 PH/27 $acl_verify_message is now set immediately after the failure of a
2058 verification in an ACL, and so is available in subsequent modifiers. In
2059 particular, the message can be preserved by coding like this:
2061 warn !verify = sender
2062 set acl_m0 = $acl_verify_message
2064 Previously, $acl_verify_message was set only while expanding "message"
2065 and "log_message" when a very denied access.
2067 PH/28 Modified OS/os.c-Linux with
2069 -#ifndef OS_LOAD_AVERAGE
2070 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2072 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2073 nomenclature these days.)
2075 PH/29 Installed patch from the Sieve maintainer that adds the options
2076 sieve_useraddress and sieve_subaddress to the redirect router.
2078 PH/30 In these circumstances:
2079 . Two addresses routed to the same list of hosts;
2080 . First host does not offer TLS;
2081 . First host accepts first address;
2082 . First host gives temporary error to second address;
2083 . Second host offers TLS and a TLS session is established;
2084 . Second host accepts second address.
2085 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2086 and peerdn, if requested) that were in fact used only for the second
2089 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2090 attention to any local part prefix or suffix that was matched by the
2091 router that accepted the address. It now behaves in the same way as it
2092 does for delivery: the affixes are removed from the local part unless
2093 rcpt_include_affixes is set on the transport.
2095 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2096 timeout during the DATA phase of an incoming message.
2098 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2099 applied a patch sanctioned by the Sieve maintainer.
2101 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2102 the uid or gid is negative. A case of a negative gid caused this to be
2103 noticed. The fix allows for either to be negative.
2105 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2106 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2109 PH/36 The change PH/12 above was broken. Fixed it.
2111 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2112 the grounds that routing the same address twice would always produce the
2113 same answer. This might have been true once, but it is certainly no
2114 longer true now. Routing a child address may depend on the previous
2115 routing that produced that child. Some complicated redirection strategies
2116 went wrong when messages had multiple recipients, and made Exim's
2117 behaviour dependent on the order in which the addresses were given.
2119 I have moved the duplicate checking until after the routing is complete.
2120 Exim scans the addresses that are assigned to local and remote
2121 transports, and removes any duplicates. This means that more work will be
2122 done, as duplicates will always all be routed, but duplicates are
2123 presumably rare, so I don't expect this is of any significance.
2125 For deliveries to pipes, files, and autoreplies, the duplicate checking
2126 still happens during the routing process, since they are not going to be
2129 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2130 It corrects a timeout issue with spamd. This is Ian's comment: "The
2131 background is that sometimes spamd either never reads data from a
2132 connection it has accepted, or it never writes response data. The exiscan
2133 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2134 blindly assumes that writes won't block so it may never time out."
2136 PH/39 Allow G after quota size as well as K and M.
2138 PH/40 The value set for $authenticated_id in an authenticator may not contain
2139 binary zeroes or newlines because the value is written to log lines and
2140 to spool files. There was no check on this. Now the value is run through
2141 the string_printing() function so that such characters are converted to
2142 printable escape sequences.
2144 PH/41 $message_linecount is a new variable that contains the total number of
2145 lines in the message. Compare $body_linecount, which is the count for the
2148 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2149 bounce and delay warning messages, except in certain special cases, which
2152 (a) An SMTP error message from a remote host;
2153 (b) A message specified in a :fail: redirection;
2154 (c) A message specified in a "fail" command in a system filter;
2155 (d) A message specified in a FAIL return from the queryprogram router;
2156 (e) A message specified by the cannot_route_message router option.
2158 In these cases only, Exim does include the error details in bounce and
2159 warning messages. There are also a few cases where bland messages such
2160 as "unrouteable address" or "local delivery error" are given.
2162 PH/43 $value is now also set for the "else" part of a ${run expansion.
2164 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2165 being worked on, but at least Exim now implements the latest version to
2168 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2169 process to complete was treated as a delivery failure, a timeout while
2170 writing the message to the pipe was logged, but erroneously treated as a
2171 successful delivery. Such timeouts include transport filter timeouts. For
2172 consistency with the overall process timeout, these timeouts are now
2173 treated as errors, giving rise to delivery failures by default. However,
2174 there is now a new Boolean option for the pipe transport called
2175 timeout_defer, which, if set TRUE, converts the failures into defers for
2176 both kinds of timeout. A transport filter timeout is now identified in
2179 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2180 systems where "make" and "gmake" are different, calling "gmake" at top
2181 level broke things. I've arranged for the value of $(MAKE) to be passed
2182 from the Makefile to this script so that it can call the same version of
2186 A note about Exim versions 4.44 and 4.50
2187 ----------------------------------------
2189 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2190 changes of various kinds. As a consequence, a big documentation update was
2191 needed. This delayed the release for rather longer than seemed good, especially
2192 in the light of a couple of (minor) security issues. Therefore, the changes
2193 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2194 release. So 4.44 and 4.50 are in effect two different branches that both start
2197 I have left the 4.50 change log unchanged; it contains all the changes since
2198 4.43. The change log for 4.44 is below; many of its items are identical to
2199 those for 4.50. This seems to be the most sensible way to preserve the
2200 historical information.
2206 1. Minor wording change to the doc/README.SIEVE file.
2208 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2209 computation of the current number of files was incorrect.
2211 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2212 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2215 4. Give more explanation in the error message when the command for a transport
2216 filter fails to execute.
2218 5. There are several places where Exim runs a non-Exim command in a
2219 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2220 was being done only for the command run by the queryprogram router. It is
2221 now done for all such subprocesses. The other cases are: ${run, transport
2222 filters, and the commands run by the lmtp and pipe transports.
2224 6. Added CONFIGURE_GROUP build-time option.
2226 7. Some older OS have a limit of 256 on the maximum number of file
2227 descriptors. Exim was using setrlimit() to set 1000 as a large value
2228 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2229 systems. I've change it so that if it can't get 1000, it tries for 256.
2231 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2232 was an oversight, and furthermore, ever since the addition of extra
2233 controls (e.g. 4.43/32), the checks on when to allow different forms of
2234 "control" were broken. There should now be diagnostics for all cases when a
2235 control that does not make sense is encountered.
2237 9. Added the /retain_sender option to "control=submission".
2239 10. $recipients is now available in the predata ACL (oversight).
2241 11. Tidy the search cache before the fork to do a delivery from a message
2242 received from the command line. Otherwise the child will trigger a lookup
2243 failure and thereby defer the delivery if it tries to use (for example) a
2244 cached ldap connection that the parent has called unbind on.
2246 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2247 of $address_data from the recipient verification was clobbered by the
2248 sender verification.
2250 13. The value of address_data from a sender verification is now available in
2251 $sender_address_data in subsequent conditions in the ACL statement.
2253 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2255 15. Added a new option "connect=<time>" to callout options, to set a different
2258 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2259 was its contents. (It was OK if the option was not defined at all.)
2261 17. A "Completed" log line is now written for messages that are removed from
2262 the spool by the -Mrm option.
2264 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2265 information about exactly what failed.
2267 19. Added -dd to debug only the daemon process.
2269 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2270 handles timeouts, both on the server side and network timeouts. Renamed the
2271 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2273 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2274 It is now set to "smtps".
2276 22. $host_address is now set to the target address during the checking of
2277 ignore_target_hosts.
2279 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2280 being passed; this would have caused $sender_host_name to have been used if
2281 matching the list had actually called for a host name (not very likely,
2282 since this list is usually IP addresses). A host name is now passed as
2285 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2286 code that specifies a non-restarting handler (typically sigaction() in
2287 modern systems) in an attempt to fix a rare and obscure crash bug.
2289 25. Narrowed the window for a race in the daemon that could cause it to ignore
2290 SIGCHLD signals. This is not a major problem, because they are used only to
2291 wake it up if nothing else does.
2293 26. A malformed maildirsize file could cause Exim to calculate negative values
2294 for the mailbox size or file count. Odd effects could occur as a result.
2295 The maildirsize information is now recalculated if the size or filecount
2298 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2299 support for a long time. Removed HAVE_SYS_VFS_H.
2301 28. Installed the latest version of exipick from John Jetmore.
2303 29. In an address list, if the pattern was not a regular expression, an empty
2304 subject address (from a bounce message) matched only if the pattern was an
2305 empty string. Non-empty patterns were not even tested. This was the wrong
2306 because it is perfectly reasonable to use an empty address as part of a
2307 database query. An empty address is now tested by patterns that are
2308 lookups. However, all the other forms of pattern expect the subject to
2309 contain a local part and a domain, and therefore, for them, an empty
2310 address still always fails if the pattern is not itself empty.
2312 30. Exim went into a mad DNS loop when attempting to do a callout where the
2313 host was specified on an smtp transport, and looking it up yielded more
2314 than one IP address.
2316 31. Re-factored the code for checking spool and log partition space into a
2317 function that finds that data and another that does the check. The former
2318 is then used to implement four new variables: $spool_space, $log_space,
2319 $spool_inodes, and $log_inodes.
2321 32. The RFC2047 encoding function was originally intended for short strings
2322 such as real names; it was not keeping to the 75-character limit for
2323 encoded words that the RFC imposes. It now respects the limit, and
2324 generates multiple encoded words if necessary. To be on the safe side, I
2325 have increased the buffer size for the ${rfc2047: expansion operator from
2328 33. It is now permitted to omit both strings after an "if" condition; if the
2329 condition is true, the result is "true". As before, when the second string
2330 is omitted, a false condition yields an empty string. This makes it less
2331 cumbersome to write custom ACL and router conditions.
2333 34. Failure to deliver a bounce message always caused it to be frozen, even if
2334 there was an errors_to setting on the router. The errors_to setting is now
2337 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2338 canonical form (fully expanded) before being placed in
2339 $sender_host_address.
2341 36. The table in the code that translates DNS record types into text (T_A to
2342 "A" for instance) was missing entries for NS and CNAME. It is just possible
2343 that this could have caused confusion if both these types were looked up
2344 for the same domain, because the text type is used as part of Exim's
2345 per-process caching. But the chance of anyone hitting this buglet seems
2348 37. The dnsdb lookup has been extended in a number of ways.
2350 (1) There is a new type, "zns", which walks up the domain tree until it
2351 finds some nameserver records. It should be used with care.
2353 (2) There is a new type, "mxh", which is like "mx" except that it returns
2354 just the host names, not the priorities.
2356 (3) It is now possible to give a list of domains (or IP addresses) to be
2357 looked up. The behaviour when one of the lookups defers can be
2358 controlled by a keyword.
2360 (4) It is now possible to specify the separator character for use when
2361 multiple records are returned.
2363 38. The dnslists ACL condition has been extended: it is now possible to supply
2364 a list of IP addresses and/or domains to be looked up in a particular DNS
2367 39. Added log_selector=+queue_time_overall.
2369 40. When running the queue in the test harness, wait just a tad after forking a
2370 delivery process, to get repeatability of debugging output.
2372 41. Include certificate and key file names in error message when GnuTLS fails
2373 to set them up, because the GnuTLS error message doesn't include the name
2374 of the failing file when there is a problem reading it.
2376 42. Allow both -bf and -bF in the same test run.
2378 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2380 44. The "Exiscan patch" is now merged into the mainline Exim source.
2382 45. Sometimes the final signoff response after QUIT could fail to get
2383 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2384 < 0 before doing a fflush(). This bug looks as though it goes back to the
2385 introduction of TLS in release 3.20, but "sometimes" must have been rare
2386 because the tests only now provoked it.
2388 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2389 (this can affect the format of dates).
2391 47. exim_tidydb, when checking for the continued existence of a message for
2392 which it has found a message-specific retry record, was not finding
2393 messages that were in split spool directories. Consequently, it was
2394 deleting retry records that should have stayed in existence.
2396 48. Steve fixed some bugs in eximstats.
2398 49. The SPA authentication driver was not abandoning authentication and moving
2399 on to the next authenticator when an expansion was forced to fail,
2400 contradicting the general specification for all authenticators. Instead it
2401 was generating a temporary error. It now behaves as specified.
2403 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2404 (the order specifies the preference for clients). The order is now AES256,
2405 AES128, 3DES, ARCFOUR128.
2407 51. Small patch to Sieve code - explicitly set From: when generating an
2410 52. Exim crashed if a remote delivery caused a very long error message to be
2411 recorded - for instance if somebody sent an entire SpamAssassin report back
2412 as a large number of 550 error lines. This bug was coincidentally fixed by
2413 increasing the size of one of Exim's internal buffers (big_buffer) that
2414 happened as part of the Exiscan merge. However, to be on the safe side, I
2415 have made the code more robust (and fixed the comments that describe what
2418 53. Now that there can be additional text after "Completed" in log lines (if
2419 the queue_time_overall log selector is set), a one-byte patch to exigrep
2420 was needed to allow it to recognize "Completed" as not the last thing in
2423 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2424 patch that reportedly fixes this has been added. I am not expert enough to
2425 create a test for it. This is what the patch creator wrote:
2427 "I found a little strange behaviour of ldap code when working with
2428 Windows 2003 AD Domain, where users was placed in more than one
2429 Organization Units. When I tried to give exim partial DN, the exit code
2430 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2431 But simultaneously result of request was absolutely normal ldap result,
2432 so I produce this patch..."
2434 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2435 REFERENCE, so I have modified the code to exclude the patch when that macro
2438 55. Some experimental protocols are using DNS PTR records for new purposes. The
2439 keys for these records are domain names, not reversed IP addresses. The
2440 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2441 leaves it alone. Component reversal etc. now happens only for IP addresses.
2444 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2446 57. Double the size of the debug message buffer (to 2048) so that more of very
2447 long debug lines gets shown.
2449 58. The exicyclog utility now does better if the number of log files to keep
2450 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2452 59. Two changes related to the smtp_active_hostname option:
2454 (1) $smtp_active_hostname is now available as a variable.
2455 (2) The default for smtp_banner uses $smtp_active_hostname instead
2456 of $primary_hostname.
2458 60. The host_aton() function is supposed to be passed a string that is known
2459 to be a valid IP address. However, in the case of IPv6 addresses, it was
2460 not checking this. This is a hostage to fortune. Exim now panics and dies
2461 if the condition is not met. A case was found where this could be provoked
2462 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2463 components; fortuitously, this particular loophole had already been fixed
2464 by change 4.50/55 above.
2466 If there are any other similar loopholes, the new check in host_aton()
2467 itself should stop them being exploited. The report I received stated that
2468 data on the command line could provoke the exploit when Exim was running as
2469 exim, but did not say which command line option was involved. All I could
2470 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2471 running as the user.
2474 61. There was a buffer overflow vulnerability in the SPA authentication code
2475 (which came originally from the Samba project). I have added a test to the
2476 spa_base64_to_bits() function which I hope fixes it.
2479 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2480 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2482 63. The daemon start-up calls getloadavg() while still root for those OS that
2483 need the first call to be done as root, but it missed one case: when
2484 deliver_queue_load_max is set with deliver_drop_privilege. This is
2485 necessary for the benefit of the queue runner, because there is no re-exec
2486 when deliver_drop_privilege is set.
2488 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2489 This has been fixed.
2491 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2492 was in use, was not putting the data itself into the right store pool;
2493 consequently, it could be overwritten for a subsequent message in the same
2494 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2497 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2499 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2500 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2501 the function were treating the return as a boolean value, which happened to
2502 work because 0=false and not-0=true, but is not correct code.
2504 68. The host_aton() function was not handling scoped IPv6 addresses (those
2505 with, for example, "%eth0" on the end) correctly.
2507 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2508 negated items (that is, ~something) in unsigned ints. Some compilers
2509 apparently mutter when there is no cast.
2511 70. If an address verification called from an ACL failed, and did not produce a
2512 user-specific message (i.e. there was only a "system" message), nothing was
2513 put in $acl_verify_message. In this situation, it now puts the system
2516 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2517 session; change 4.31/43 added the unwanted input to the log line - except
2518 that it did not do this in the start of session case. It now does.
2520 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2521 This is wrong and can cause the other end to generate a synchronization
2522 error if it is another Exim or anything else that does the synchronization
2523 check. A QUIT command is no longer sent after a timeout.
2525 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2526 during host lookups.
2528 74. The defer_ok option of callout verification was not working if it was used
2529 when verifying addresses in header lines, that is, for this case:
2531 verify = header_sender/callout=defer_ok
2533 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2534 those file descriptors could be used for SMTP connections. If anything
2535 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2536 could be sent to the SMTP client, causing chaos. The daemon now opens
2537 stdin, stdout, and stderr to /dev/null when it puts itself into the
2540 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2541 log by default. The user can override this with suitable Perl magic.
2543 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2544 the log message when discard triggers, was not working for the DATA ACL or
2545 for the non-SMTP ACL.
2547 78. Error message wording change in sieve.c.
2549 79. If smtp_accept_max_per_host was set, the number of connections could be
2550 restricted to fewer than expected, because the daemon was trying to set up
2551 a new connection before checking whether the processes handling previous
2552 connections had finished. The check for completed processes is now done
2553 earlier. On busy systems, this bug wouldn't be noticed because something
2554 else would have woken the daemon, and it would have reaped the completed
2557 80. If a message was submitted locally by a user whose login name contained one
2558 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2559 It caused a spool format error. I have fixed the spool reading code. A
2560 related problem was that the "from" clause in the Received: line became
2561 illegal because of the space(s). It is now covered by ${quote_local_part.
2563 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2566 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2568 83. Patch from Sieve maintainer to fix unterminated string problem in
2569 "vacation" handling.
2571 84. Some minor changes to the Linux configuration files to help with other
2572 OS variants using glibc.
2574 85. One more patch for Sieve to update vacation handling to latest spec.
2577 ----------------------------------------------------
2578 See the note above about the 4.44 and 4.50 releases.
2579 ----------------------------------------------------
2585 1. Change 4.43/35 introduced a bug that caused file counts to be
2586 incorrectly computed when quota_filecount was set in an appendfile
2589 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2590 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2593 3. Give more explanation in the error message when the command for a transport
2594 filter fails to execute.
2596 4. There are several places where Exim runs a non-Exim command in a
2597 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2598 was being done only for the command run by the queryprogram router. It is
2599 now done for all such subprocesses. The other cases are: ${run, transport
2600 filters, and the commands run by the lmtp and pipe transports.
2602 5. Some older OS have a limit of 256 on the maximum number of file
2603 descriptors. Exim was using setrlimit() to set 1000 as a large value
2604 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2605 systems. I've change it so that if it can't get 1000, it tries for 256.
2607 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2608 was an oversight, and furthermore, ever since the addition of extra
2609 controls (e.g. 4.43/32), the checks on when to allow different forms of
2610 "control" were broken. There should now be diagnostics for all cases when a
2611 control that does not make sense is encountered.
2613 7. $recipients is now available in the predata ACL (oversight).
2615 8. Tidy the search cache before the fork to do a delivery from a message
2616 received from the command line. Otherwise the child will trigger a lookup
2617 failure and thereby defer the delivery if it tries to use (for example) a
2618 cached ldap connection that the parent has called unbind on.
2620 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2621 of $address_data from the recipient verification was clobbered by the
2622 sender verification.
2624 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2625 was its contents. (It was OK if the option was not defined at all.)
2627 11. A "Completed" log line is now written for messages that are removed from
2628 the spool by the -Mrm option.
2630 12. $host_address is now set to the target address during the checking of
2631 ignore_target_hosts.
2633 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2634 being passed; this would have caused $sender_host_name to have been used if
2635 matching the list had actually called for a host name (not very likely,
2636 since this list is usually IP addresses). A host name is now passed as
2639 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2640 code that specifies a non-restarting handler (typically sigaction() in
2641 modern systems) in an attempt to fix a rare and obscure crash bug.
2643 15. Narrowed the window for a race in the daemon that could cause it to ignore
2644 SIGCHLD signals. This is not a major problem, because they are used only to
2645 wake it up if nothing else does.
2647 16. A malformed maildirsize file could cause Exim to calculate negative values
2648 for the mailbox size or file count. Odd effects could occur as a result.
2649 The maildirsize information is now recalculated if the size or filecount
2652 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2653 support for a long time. Removed HAVE_SYS_VFS_H.
2655 18. Updated exipick to current release from John Jetmore.
2657 19. Allow an empty sender to be matched against a lookup in an address list.
2658 Previously the only cases considered were a regular expression, or an
2661 20. Exim went into a mad DNS lookup loop when doing a callout where the
2662 host was specified on the transport, if the DNS lookup yielded more than
2665 21. The RFC2047 encoding function was originally intended for short strings
2666 such as real names; it was not keeping to the 75-character limit for
2667 encoded words that the RFC imposes. It now respects the limit, and
2668 generates multiple encoded words if necessary. To be on the safe side, I
2669 have increased the buffer size for the ${rfc2047: expansion operator from
2672 22. Failure to deliver a bounce message always caused it to be frozen, even if
2673 there was an errors_to setting on the router. The errors_to setting is now
2676 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2677 canonical form (fully expanded) before being placed in
2678 $sender_host_address.
2680 24. Updated eximstats to version 1.33
2682 25. Include certificate and key file names in error message when GnuTLS fails
2683 to set them up, because the GnuTLS error message doesn't include the name
2684 of the failing file when there is a problem reading it.
2686 26. Expand error message when OpenSSL has problems setting up cert/key files.
2689 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2690 (this can affect the format of dates).
2692 28. exim_tidydb, when checking for the continued existence of a message for
2693 which it has found a message-specific retry record, was not finding
2694 messages that were in split spool directories. Consequently, it was
2695 deleting retry records that should have stayed in existence.
2697 29. eximstats updated to version 1.35
2698 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2699 1.35 - bugfix such that pie charts by volume are generated correctly
2701 30. The SPA authentication driver was not abandoning authentication and moving
2702 on to the next authenticator when an expansion was forced to fail,
2703 contradicting the general specification for all authenticators. Instead it
2704 was generating a temporary error. It now behaves as specified.
2706 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2707 (the order specifies the preference for clients). The order is now AES256,
2708 AES128, 3DES, ARCFOUR128.
2710 31. Small patch to Sieve code - explicitly set From: when generating an
2713 32. Exim crashed if a remote delivery caused a very long error message to be
2714 recorded - for instance if somebody sent an entire SpamAssassin report back
2715 as a large number of 550 error lines. This bug was coincidentally fixed by
2716 increasing the size of one of Exim's internal buffers (big_buffer) that
2717 happened as part of the Exiscan merge. However, to be on the safe side, I
2718 have made the code more robust (and fixed the comments that describe what
2721 33. Some experimental protocols are using DNS PTR records for new purposes. The
2722 keys for these records are domain names, not reversed IP addresses. The
2723 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2724 leaves it alone. Component reversal etc. now happens only for IP addresses.
2727 34. The host_aton() function is supposed to be passed a string that is known
2728 to be a valid IP address. However, in the case of IPv6 addresses, it was
2729 not checking this. This is a hostage to fortune. Exim now panics and dies
2730 if the condition is not met. A case was found where this could be provoked
2731 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2732 components; fortuitously, this particular loophole had already been fixed
2733 by change 4.50/55 or 4.44/33 above.
2735 If there are any other similar loopholes, the new check in host_aton()
2736 itself should stop them being exploited. The report I received stated that
2737 data on the command line could provoke the exploit when Exim was running as
2738 exim, but did not say which command line option was involved. All I could
2739 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2740 running as the user.
2743 35. There was a buffer overflow vulnerability in the SPA authentication code
2744 (which came originally from the Samba project). I have added a test to the
2745 spa_base64_to_bits() function which I hope fixes it.
2748 36. The daemon start-up calls getloadavg() while still root for those OS that
2749 need the first call to be done as root, but it missed one case: when
2750 deliver_queue_load_max is set with deliver_drop_privilege. This is
2751 necessary for the benefit of the queue runner, because there is no re-exec
2752 when deliver_drop_privilege is set.
2754 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2755 was in use, was not putting the data itself into the right store pool;
2756 consequently, it could be overwritten for a subsequent message in the same
2757 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2760 38. Sometimes the final signoff response after QUIT could fail to get
2761 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2762 < 0 before doing a fflush(). This bug looks as though it goes back to the
2763 introduction of TLS in release 3.20, but "sometimes" must have been rare
2764 because the tests only now provoked it.
2770 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2771 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2772 Now it returns an integer. A number of calls were still expecting a T/F
2773 return. Fortuitously, in all cases, the tests worked in OK situations,
2774 which is the norm. However, things would have gone wrong on any write
2775 failures on the smtp file descriptor. This function is used when sending
2776 messages over SMTP and also when doing verify callouts.
2778 2. When Exim is called to do synchronous delivery of a locally submitted
2779 message (the -odf or -odi options), it no longer closes stderr before doing
2782 3. Implemented the mua_wrapper option.
2784 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2786 5. Implemented the functions header_remove(), header_testname(),
2787 header_add_at_position(), and receive_remove_recipient(), and exported them
2790 6. If an ACL "warn" statement specified the addition of headers, Exim already
2791 inserted X-ACL-Warn: at the start if there was no header name. However, it
2792 was not making this test for the second and subsequent header lines if
2793 there were newlines in the string. This meant that an invalid header could
2794 be inserted if Exim was badly configured.
2796 7. Allow an ACL "warn" statement to add header lines at the start or after all
2797 the Received: headers, as well as at the end.
2799 8. Added the rcpt_4xx retry error code.
2801 9. Added postmaster_mailfrom=xxx to callout verification option.
2803 10. Added mailfrom=xxxx to the callout verification option, for verify=
2806 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2807 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2809 12. Inserted some casts to stop certain compilers warning when using pointer
2810 differences as field lengths or precisions in printf-type calls (mostly
2811 affecting debugging statements).
2813 13. Added optional readline() support for -be (dynamically loaded).
2815 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2816 same clock tick as a message's arrival, so that its received time was the
2817 same as the "first fail" time on the retry record, and that message
2818 remained on the queue past the ultimate address timeout, every queue runner
2819 would try a delivery (because it was past the ultimate address timeout) but
2820 after another failure, the ultimate address timeout, which should have then
2821 bounced the address, did not kick in. This was a "< instead of <=" error;
2822 in most cases the first failure would have been in the next clock tick
2823 after the received time, and all would be well.
2825 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2826 being recognized when the domain list was tested by the match_domain
2827 condition in an expansion string.
2829 16. Added the ${str2b64: operator.
2831 17. Exim was always calling setrlimit() to set a large limit for the number of
2832 processes, without checking whether the existing limit was already
2833 adequate. (It did check for the limit on file descriptors.) Furthermore,
2834 errors from getrlimit() and setrlimit() were being ignored. Now they are
2835 logged to the main and panic logs, but Exim does carry on, to try to do its
2836 job under whatever limits there are.
2838 18. Imported PCRE 5.0.
2840 19. Trivial typo in log message " temporarily refused connection" (the leading
2843 20. If the log selector return_path_on_delivery was set and an address was
2844 redirected to /dev/null, the delivery process crashed because it assumed
2845 that a return path would always be set for a "successful" delivery. In this
2846 case, the whole delivery is bypassed as an optimization, and therefore no
2849 21. Internal re-arrangement: the function for sending a challenge and reading
2850 a response while authentication was assuming a zero-terminated challenge
2851 string. It's now changed to take a pointer and a length, to allow for
2852 binary data in such strings.
2854 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2856 23. Exim was not respecting finduser_retries when seeking the login of the
2857 uid under which it was called; it was always trying 10 times. (The default
2858 setting of finduser_retries is zero.) Also, it was sleeping after the final
2859 failure, which is pointless.
2861 24. Implemented tls_on_connect_ports.
2863 25. Implemented acl_smtp_predata.
2865 26. If the domain in control=submission is set empty, Exim assumes that the
2866 authenticated id is a complete email address when it generates From: or
2867 Sender: header lines.
2869 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2870 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2871 chown and chgrp in /bin and hostname in /usr/bin.
2873 28. Exim was keeping the "process log" file open after each use, just as it
2874 does for the main log. This opens the possibility of it remaining open for
2875 long periods when the USR1 signal hits a daemon. Occasional processlog
2876 errors were reported, that could have been caused by this. Anyway, it seems
2877 much more sensible not to leave this file open at all, so that is what now
2880 29. The long-running daemon process does not normally write to the log once it
2881 has entered its main loop, and it closes the log before doing so. This is
2882 so that log files can straightforwardly be renamed and moved. However,
2883 there are a couple of unusual error situations where the daemon does write
2884 log entries, and I had neglected to close the log afterwards.
2886 30. The text of an SMTP error response that was received during a remote
2887 delivery was being truncated at 512 bytes. This is too short for some of
2888 the long messages that one sometimes sees. I've increased the limit to
2891 31. It is now possible to make retry rules that apply only when a message has a
2892 specific sender, in particular, an empty sender.
2894 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2895 it possible to be selective about when SMTP synchronization is enforced.
2897 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2899 32. Implemented hosts_connection_nolog.
2901 33. Added an ACL for QUIT.
2903 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2906 35. Added mailbox_size and mailbox_filecount to appendfile.
2908 36. Added control = no_multiline_responses to ACLs.
2910 37. There was a bug in the logic of the code that waits for the clock to tick
2911 in the case where the clock went backwards by a substantial amount such
2912 that the microsecond fraction of "now" was more than the microsecond
2913 fraction of "then" (but the whole seconds number was less).
2915 38. Added support for the libradius Radius client library this is found on
2916 FreeBSD (previously only the radiusclient library was supported).
2922 1. When certain lookups returned multiple values in the form name=value, the
2923 quoting of the values was not always being done properly. Specifically:
2924 (a) If the value started with a double quote, but contained no whitespace,
2926 (b) If the value contained whitespace other than a space character (i.e.
2927 tabs or newlines or carriage returns) it was not quoted.
2928 This fix has been applied to the mysql and pgsql lookups by writing a
2929 separate quoting function and calling it from the lookup code. The fix
2930 should probably also be applied to nisplus, ibase and oracle lookups, but
2931 since I cannot test any of those, I have not disturbed their existing code.
2933 2. A hit in the callout cache for a specific address caused a log line with no
2934 reason for rejecting RCPT. Now it says "Previous (cached) callout
2935 verification failure".
2937 3. There was an off-by-one bug in the queryprogram router. An over-long
2938 return line was truncated at 256 instead of 255 characters, thereby
2939 overflowing its buffer with the terminating zero. As well as fixing this, I
2940 have increased the buffer size to 1024 (and made a note to document this).
2942 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2943 when Exim is waiting for an SMTP response from a remote server, Exim
2944 restarts its select() call on the socket, thereby resetting its timeout.
2945 This is not a problem when such interrupts are rare. Somebody set up a cron
2946 job to run exiwhat every 2 minutes, which is less than the normal select()
2947 timeout (5 or 10 minutes). This meant that the select() timeout never
2948 kicked in because it was always reset. I have fixed this by comparing the
2949 time when an interrupt arrives with the time at the start of the first call
2950 to select(). If more time than the timeout has elapsed, the interrupt is
2951 treated as a timeout.
2953 5. Some internal re-factoring in preparation for the addition of Sieve
2954 extensions (by MH). In particular, the "personal" test is moved to a
2955 separate function, and given an option for scanning Cc: and Bcc: (which is
2956 not set for Exim filters).
2958 6. When Exim created an email address using the login of the caller as the
2959 local part (e.g. when creating a From: or Sender: header line), it was not
2960 quoting the local part when it contained special characters such as @.
2962 7. Installed new OpenBSD configuration files.
2964 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2965 try to make them clearer.
2967 9. Callout options, other than the timeout value, were being ignored when
2968 verifying sender addresses in header lines. For example, when using
2970 verify = header_sender/callout=no_cache
2972 the cache was (incorrectly) being used.
2974 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2975 only the Cygwin environment.
2977 11. When return_path_on_delivery was set as a log selector, if different remote
2978 addresses in the same message used different return paths and parallel
2979 remote delivery occurred, the wrong values would sometimes be logged.
2980 (Whenever a remote delivery process finished, the return path value from
2981 the most recently started remote delivery process was logged.)
2983 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2984 lines when AUTH and/or TLS are in use. This is the "received protocol"
2985 field. Exim used to use "asmtp" for authenticated SMTP, without any
2986 indication (in the protocol name) for TLS use. Now it follows the RFC and
2987 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2988 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2989 names appear in log lines as well as in Received: header lines.
2991 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2992 extensions, and comparison tests, and to fix some bugs.
2994 14. Changes to the "personal" filter test:
2996 (1) The test was buggy in that it was just doing the equivalent of
2997 "contains" tests on header lines. For example, if a user's address was
2998 anne@some.where, the "personal" test would incorrectly be true for
3000 To: susanne@some.where
3002 This test is now done by extracting each address from the header in turn,
3003 and checking the entire address. Other tests that are part of "personal"
3004 are now done using regular expressions (for example, to check local parts
3005 of addresses in From: header lines).
3007 (2) The list of non-personal local parts in From: addresses has been
3008 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3009 taken from the Sieve specification recommendations.
3011 (3) If the message contains any header line starting with "List-" it is
3012 treated as non-personal.
3014 (4) The test for "circular" in the Subject: header line has been removed
3015 because it now seems ill-conceived.
3017 15. Minor typos in src/EDITME comments corrected.
3019 16. Installed latest exipick from John Jetmore.
3021 17. If headers_add on a router specified a text string that was too long for
3022 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3023 of string_sprintf() is now avoided.
3025 18. $message_body_size was not set (it was always zero) when running the DATA
3026 ACL and the local_scan() function.
3028 19. For the "mail" command in an Exim filter, no default was being set for
3029 the once_repeat time, causing a random time value to be used if "once" was
3030 specified. (If the value happened to be <= 0, no repeat happened.) The
3031 default is now 0s, meaning "never repeat". The "vacation" command was OK
3032 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3033 (I found it when inspecting the code).
3035 20. There is now an overall timeout for performing a callout verification. It
3036 defaults to 4 times the callout timeout, which applies to individual SMTP
3037 commands during the callout. The overall timeout applies when there is more
3038 than one host that can be tried. The timeout is checked before trying the
3039 next host. This prevents very long delays if there are a large number of
3040 hosts and all are timing out (e.g. when the network connections are timing
3041 out). The value of the overall timeout can be changed by specifying an
3042 additional sub-option for "callout", called "maxwait". For example:
3044 verify = sender/callout=5s,maxwait=20s
3046 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3047 to the end before writing, but this should make it even safer).
3049 22. Exim was forgetting that it had advertised PIPELINING for the second and
3050 subsequent messages on an SMTP connection. It was also not resetting its
3051 memory on STARTTLS and an internal HELO.
3053 23. When Exim logs an SMTP synchronization error within a session, it now
3054 records whether PIPELINING has been advertised or not.
3056 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3057 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3058 rather than long int.
3060 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3062 26. Added the never_mail option to autoreply.
3068 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3069 crash if the getsockname() call failed; this can happen if a connection is
3070 closed very soon after it is established. The problem was simply in the
3071 order in which certain operations were done, causing Exim to try to write
3072 to the SMTP stream before it had set up the file descriptor. The bug has
3073 been fixed by making things happen in the correct order.
3079 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3080 before the connection was closed, thus losing the rejection response.
3082 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3083 some early Solaris releases, but causes trouble in current releases where
3084 socklen_t is defined.
3086 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3089 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3092 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3093 that releasing the top of it at the end releases what was used for sub-
3094 expansions (unless the block got too big). However, discard this block if
3095 the first thing is a variable or header, so that we can use its block when
3096 it is dynamic (useful for very large $message_headers, for example).
3098 6. Lookups now cache *every* query, not just the most recent. A new, separate
3099 store pool is used for this. It can be recovered when all lookup caches are
3100 flushed. Lookups now release memory at the end of their result strings.
3101 This has involved some general refactoring of the lookup sources.
3103 7. Some code has been added to the store_xxx() functions to reduce the amount
3104 of flapping under certain conditions.
3106 8. log_incoming_interface used to affect only the <= reception log lines. Now
3107 it causes the local interface and port to be added to several more SMTP log
3108 lines, for example "SMTP connection from", and rejection lines.
3110 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3112 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3114 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3115 could be overwritten at the end of the current message (or the start of a
3116 new message if it was set in a HELO ACL). The value is now preserved for
3117 the duration of the SMTP connection.
3119 12. If a transport had a headers_rewrite setting, and a matching header line
3120 contained an unqualified address, that address was qualified, even if it
3121 did not match any rewriting rules. The underlying bug was that the values
3122 of the flags that permit the existence of unqualified sender and recipient
3123 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3124 non-local messages, and by -bnq for local messages) were not being
3125 preserved with the message after it was received.
3127 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3128 "next input=" as part of the text comprising the host identity instead of
3129 the correct text. The code was using the same buffer for two different
3130 strings. However, depending on which order the printing function evaluated
3131 its arguments, the bug did not always show up. Under Linux, for example, my
3132 test suite worked just fine.
3134 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3135 doesn't work with some older versions of Perl. It has been changed to "my",
3136 which in any case is probably the better facility to use.
3138 15. A really picky compiler found some instances of statements for creating
3139 error messages that either had too many or two few arguments for the format
3142 16. The size of the buffer for calls to the DNS resolver has been increased
3143 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3144 for addresses that have a lot of PTR records. This alleviates a problem; it
3145 does not fully solve it.
3147 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3148 buffer now truncates the list and logs the incident, which is the same
3149 action as happens when Exim is looking up a host name and its aliases.
3150 Previously in this situation something unpredictable would happen;
3151 sometimes it was "internal error: store_reset failed".
3153 18. If a server dropped the connection unexpectedly when an Exim client was
3154 using GnuTLS and trying to read a response, the client delivery process
3155 crashed while trying to generate an error log message.
3157 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3158 string, for example:
3160 warn message = H1: something\nH2: something
3162 the text was added as a single header line from Exim's point of view
3163 though it ended up OK in the delivered message. However, searching for the
3164 second and subsequent header lines using $h_h2: did not work. This has been
3165 fixed. Similarly, if a system filter added multiple headers in this way,
3166 the routers could not see them.
3168 20. Expanded the error message when iplsearch is called with an invalid key to
3169 suggest using net-iplsearch in a host list.
3171 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3172 ACLs are no longer actually imposed (and a message to that effect is
3175 22. If a "gecos" field in a passwd entry contained escaped characters, in
3176 particular, if it contained a \" sequence, Exim got it wrong when building
3177 a From: or a Sender: header from that name. A second bug also caused
3178 incorrect handling when an unquoted " was present following a character
3179 that needed quoting.
3181 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3182 was not being matched caselessly.
3184 24. Arranged for all hyphens in the exim.8 source to be escaped with
3187 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3188 properly. Recipient callout cache records were still being keyed to include
3189 the sender, even when use_sender was set false. This led to far more
3190 callouts that were necessary. The sender is no longer included in the key
3191 when use_sender is false.
3193 26. Added "control = submission" modifier to ACLs.
3195 27. Added the ${base62d: operator to decode base 62 numbers.
3197 28. dnsdb lookups can now access SRV records.
3199 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3200 the configuration file.
3202 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3203 (-v) mode. This makes the output for a verbose queue run more intelligible.
3205 31. Added a use_postmaster feature to recipient callouts.
3207 32. Added the $body_zerocount variable, containing the number of binary zero
3208 bytes in the message body.
3210 33. The time of last modification of the "new" subdirectory is now used as the
3211 "mailbox time last read" when there is a quota error for a maildir
3214 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3216 35. Added +ignore_unknown as a special item in host lists.
3218 36. Code for decoding IPv6 addresses in host lists is now included, even if
3219 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3220 address was recognized as an IP address, but was then not correctly decoded
3221 into binary, causing unexpected and incorrect effects when compared with
3228 1. Very minor rewording of debugging text in manualroute to say "list of
3229 hosts" instead of "hostlist".
3231 2. If verify=header_syntax was set, and a header line with an unqualified
3232 address (no domain) and a large number of spaces between the end of the
3233 name and the colon was received, the reception process suffered a buffer
3234 overflow, and (when I tested it) crashed. This was caused by some obsolete
3235 code that should have been removed. The fix is to remove it!
3237 3. When running in the test harness, delay a bit after writing a bounce
3238 message to get a bit more predictability in the log output.
3240 4. Added a call to search_tidyup() just before forking a reception process. In
3241 theory, someone could use a lookup in the expansion of smtp_accept_max_
3242 per_host which, without the tidyup, could leave open a database connection.
3244 5. Added the variables $recipient_data and $sender_data which get set from a
3245 lookup success in an ACL "recipients" or "senders" condition, or a router
3246 "senders" option, similar to $domain_data and $local_part_data.
3248 6. Moved the writing of debug_print from before to after the "senders" test
3251 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3252 problems for message scanning, either using a data ACL, or using
3253 local_scan() because the Received: header was not generated till after they
3254 were called (in order to set the time as the time of reception completion).
3255 I have revised the way this works. The header is now generated after the
3256 body is received, but before the ACL or local_scan() are called. After they
3257 are run, the timestamp in the header is updated.
3263 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3264 before starting a queue runner without re-exec. This happened only when
3265 deliver_drop_privilege was set or when the Exim user was set to root. The
3266 effect of the bug was that timeouts during subsequent deliveries caused
3267 crashes instead of being properly handled. The handler is now left at its
3268 default (and expected) setting.
3270 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3271 message, again when deliver_drop_privilege is set or Exim is run as root.
3272 The bug described in (1) was not present in this case, but the tidying up
3273 of the other signals was missing. I have made the two cases consistent.
3275 3. The ignore_target_hosts setting on a manualroute router was being ignored
3276 for hosts that were looked up using the /MX notation.
3278 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3281 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3282 operated on the sender address. After changing the $sender_address to <>
3283 for the sender address verify, Exim was re-instated it as the original
3284 (before rewriting) address, but remembering that it had rewritten it, so it
3285 wasn't rewriting it again. This bug also had the effect of breaking the
3286 sender address verification caching when the sender address was rewritten.
3288 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3289 This has been changed so that if the ip literal address matches
3290 ignore_target_hosts, the router declines.
3292 7. Added expansion conditions match_domain, match_address, and match_local_
3293 part (NOT match_host).
3295 8. The placeholder for the Received: header didn't have a length field set.
3297 9. Added code to Exim itself and to exim_lock to test for a specific race
3298 condition that could lead to file corruption when using MBX delivery. The
3299 issue is with the lockfile that is created in /tmp. If this file is removed
3300 after a process has opened it but before that process has acquired a lock,
3301 there is the potential for a second process to recreate the file and also
3302 acquire a lock. This could lead to two Exim processes writing to the file
3303 at the same time. The added code performs the same test as UW imapd; it
3304 checks after acquiring the lock that its file descriptor still refers to
3305 the same named file.
3307 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3308 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3311 11. Added the smtp_active_hostname option. If used, this will typically be made
3312 to depend on the incoming interface address. Because $interface_address is
3313 not set up until the daemon has forked a reception process, error responses
3314 that can happen earlier (such as "too many connections") no longer contain
3317 12. If an expansion in a condition on a "warn" statement fails because a lookup
3318 defers, the "warn" statement is abandoned, and the next ACL statement is
3319 processed. Previously this caused the whole ACL to be aborted.
3321 13. Added the iplsearch lookup type.
3323 14. Added ident_timeout as a log selector.
3325 15. Added tls_certificate_verified as a log selector.
3327 16. Added a global option tls_require_ciphers (compare the smtp transport
3328 option of the same name). This controls incoming TLS connections.
3330 17. I finally figured out how to make tls_require_ciphers do a similar thing
3331 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3332 before starting the TLS session.
3334 18. Tabs are now shown as \t in -bP output.
3336 19. If the log selector return_path_on_delivery was set, Exim crashed when
3337 bouncing a message because it had too many Received: header lines.
3339 20. If two routers both had headers_remove settings, and the first one included
3340 a superfluous trailing colon, the final name in the first list and the
3341 first name in the second list were incorrectly joined into one item (with a
3342 colon in the middle).
3348 1. Added -C and -D options to the exinext utility, mainly to make it easier
3349 to include in the automated testing, but these could be helpful when
3350 multiple configurations are in use.
3352 2. The exinext utility was not formatting the output nicely when there was
3353 an alternate port involved in the retry record key, nor when there was a
3354 message id as well (for retries that were specific to a specific message
3355 and a specific host). It was also confused by IPv6 addresses, because of
3356 the additional colons they contain. I have fixed the IPv4 problem, and
3357 patched it up to do a reasonable job for IPv6.
3359 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3360 delivery, the log line now contains "pipelined" if PIPELINING was used.
3362 4. An SMTP transport process used to panic and die if the bind() call to set
3363 an explicit outgoing interface failed. This has been changed; it is now
3364 treated in the same way as a connect() failure.
3366 5. A reference to $sender_host_name in the part of a conditional expansion
3367 that was being skipped was still causing a DNS lookup. This no longer
3370 6. The def: expansion condition was not recognizing references to header lines
3371 that used bh_ and bheader_.
3373 7. Added the _cache feature to named lists.
3375 8. The code for checking quota_filecount in the appendfile transport was
3376 allowing one more file than it should have been.
3378 9. For compatibility with Sendmail, the command line option
3386 and sets the incoming protocol and host name (for trusted callers). The
3387 host name and its colon can be omitted when only the protocol is to be set.
3388 Note the Exim already has two private options, -pd and -ps, that refer to
3389 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3390 "s", but I don't think that's a major issue.
3392 10. A number of refactoring changes to the code, none of which should affect
3395 (a) The number of logging options was getting close to filling up the
3396 32-bit word that was used as a bit map. I have split them into two classes:
3397 those that are passed in the argument to log_write(), and those that are
3398 only ever tested independently outside of that function. These are now in
3399 separate 32-bit words, so there is plenty of room for expansion again.
3400 There is no change in the user interface or the logging behaviour.
3402 (b) When building, for example, log lines, the code previously used a
3403 macro that called string_cat() twice, in order to add two strings. This is
3404 not really sufficiently general. Furthermore, there was one instance where
3405 it was actually wrong because one of the argument was used twice, and in
3406 one call a function was used. (As it happened, calling the function twice
3407 did not affect the overall behaviour.) The macro has been replaced by a
3408 function that can join an arbitrary number of extra strings onto a growing
3411 (c) The code for expansion conditions now uses a table and a binary chop
3412 instead of a serial search (which was left over from when there were very
3413 few conditions). Also, it now recognizes conditions like "pam" even when
3414 the relevant support is not compiled in: a suitably worded error message is
3415 given if an attempt is made to use such a condition.
3417 11. Added ${time_interval:xxxxx}.
3419 12. A bug was causing one of the ddress fields not to be passed back correctly
3420 from remote delivery subprocesses. The field in question was not being
3421 subsequently used, so this caused to problems in practice.
3423 13. Added new log selectors queue_time and deliver_time.
3425 14. Might have fixed a bug in maildirsizefile handling that threw up
3426 "unexpected character" debug warnings, and recalculated the data
3427 unnecessarily. In any case, I expanded the warning message to give more
3430 15. Added the message "Restricted characters in address" to the statements in
3431 the default ACL that block characters like @ and % in local parts.
3433 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3434 Three changes have been made:
3436 (a) There was a serious bug; a negative response to MAIL caused the whole
3437 recipient domain to be cached as invalid, thereby blocking all messages
3438 to all local parts at the same domain, from all senders. This bug has
3439 been fixed. The domain is no longer cached after a negative response to
3440 MAIL if the sender used is not empty.
3442 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3445 (c) A new callout option, "use_sender" has been added for people who want
3446 the modified behaviour.
3452 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3455 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3456 indeed breaks things for older releases.
3458 3. Added additional logging to the case where there is a problem reading data
3459 from a filter that is running in a subprocess using a pipe, in order to
3460 try to track down a specific problem.
3462 4. Testing facility fudge: when running in the test harness and attempting
3463 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3464 getting "No route to host". Convert this to a timeout.
3466 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3469 6. Some OS don't have socklen_t but use size_t instead. This affects the
3470 fifth argument of getsockopt() amongst other things. This is now
3471 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3472 can be set for individual OS. I have set it for SunOS5, OSF1, and
3473 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3474 some earlier ones do not.
3476 7. Change 4.30/15 was not doing the test caselessly.
3478 8. The standard form for an IPv6 address literal was being rejected by address
3479 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3480 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3481 this, as well as the form without the "IPv6" on the front (but only when
3482 address literals are enabled, of course).
3484 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3486 10. Exim crashed if a message with an empty sender address specified by -f
3487 encountered a router with an errors_to setting. This could be provoked only
3488 by a command such as
3492 where an empty string was supplied; "<>" did not hit this bug.
3494 11. Installed PCRE release 4.5.
3496 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3497 remained set. It is now erased.
3499 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3500 times from message ids (which are base 36 rather than the normal 62).
3502 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3503 were being counted as actual protocol errors, and logged if the log
3504 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3505 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3506 and DATA following a set of rejected RCPTs do not count as protocol errors.
3507 In other words, Exim assumes they were pipelined, though this may not
3508 actually be the case. Of course, in all cases the client gets an
3509 appropriate error code.
3511 15. If a lookup fails in an ACL condition, a message about the failure may
3512 be available; it is used if testing the ACL cannot continue, because most
3513 such messages specify what the cause of the deferral is. However, some
3514 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3515 that caused an old message to be retained and used if a later statement
3516 caused a defer, replacing the real cause of the deferral.
3518 16. If an IP address had so many PTR records that the DNS lookup buffer
3519 was not large enough to hold them, Exim could crash while trying to process
3520 the truncated data. It now detects and logs this case.
3522 17. Further to 4.21/58, another change has been made: if (and only if) the
3523 first line of a message (the first header line) ends with CRLF, a bare LF
3524 in a subsequent header line has a space inserted after it, so as not to
3525 terminate the header.
3527 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3528 unnecessarily, used atoi() instead of strtol(), and didn't check the
3529 termination when getting file sizes from file names by regex.
3531 19. Completely re-implemented the support for maildirsize files, in the light
3532 of a number of problems with the previous contributed implementation
3533 (4.30/29). In particular:
3535 . If the quota is zero, the maildirsize file is maintained, but no quota is
3538 . If the maildir directory does not exist, it is created before any attempt
3539 to write a maildirsize file.
3541 . The quota value in the file is just a cache; if the quota is changed in
3542 the transport, the new value overrides.
3544 . A regular expression is available for excluding directories from the
3547 20. The autoreply transport checks the characters in options that define the
3548 message's headers; it allows continued headers, but it was checking with
3549 isspace() after an embedded newline instead of explicitly looking for a
3552 21. If all the "regular" hosts to which an address was routed had passed their
3553 expiry times, and had not reached their retry times, the address was
3554 bounced, even if fallback hosts were defined. Now Exim should go on to try
3557 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3558 equivalent code in the SMTP transport. Some hosts send humungous responses
3559 to HELO/EHLO, more than 1024 it seems.
3561 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3562 gives compiler warnings in some environments. I've now done it "properly",
3565 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3566 (because of problems with the built-in one) was declared to return uschar *
3567 instead of char *, causing compiler failure.
3569 25. Fixed a file descriptor leak when processing alias/forward files.
3571 26. Fixed a minor format string issue in dbfn.c.
3573 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3575 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3576 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3579 29. When a sender address is verified, it is cached, to save repeating the test
3580 when there is more than one recipient in a message. However, when the
3581 verification involves a callout, it is possible for different callout
3582 options to be set for different recipients. It is too complicated to keep
3583 track of this in the cache, so now Exim always runs a verification when a
3584 callout is required, relying on the callout cache for the optimization.
3585 The overhead is duplication of the address routing, but this should not be
3588 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3589 to be verified with callout=postmaster, and the main callout worked but the
3590 postmaster check failed, the verification correctly failed. However, if a
3591 subsequent RCPT command asked for sender verification *without* the
3592 postmaster check, incorrect caching caused this verification also to fail,
3595 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3596 it was not caching the DNS options (qualify_single, search_parents) that
3597 were used when the lookup failed. A subsequent lookup with different
3598 options therefore always gave the same answer, though there were cases
3599 where it should not have. (Example: a "domains = !$mx_any" option on a
3600 dnslookup router: the "domains" option is always processed without any
3601 widening, but the router might have qualify_single set.) Now Exim uses the
3602 cached value only when the same options are set.
3604 32. Added John Jetmore's "exipick" utility to the distribution.
3606 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3607 than a timeout (e.g. a certificate is required, and is not provided), an
3608 Exim server now closes the connection immediately. Previously it waited for
3609 the client to close - but if the client is SSL, it seems that they each
3610 wait for each other, leading to a delay before one of them times out.
3612 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3613 maintained 0.8.x compatibility because I don't think many are using it, and
3614 it is clearly obsolete.
3616 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3619 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3620 client certificate was expired. A simple patch fixes this, though I don't
3621 understand the full logic of why the verify callback is called multiple
3624 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3625 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3626 which causes problems with some clients (such as the Certicom SSL Plus
3627 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3628 disables the coutermeasure allowing Eudora to connect."
3630 38. Exim was not checking that a write() to a log file succeeded. This could
3631 lead to Bad Things if a log got too big, in particular if it hit a file
3632 size limit. Exim now panics and dies if it cannot write to a log file, just
3633 as it does if it cannot open a log file.
3635 39. Modified OS/Makefile-Linux so that it now contains
3637 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3639 The two -D definitions ensure that Exim is compiled with large file
3640 support, which makes it possible to handle log files that are bigger than
3643 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3644 instance) a domain was checked against a named list that involved a lookup,
3645 causing $domain_data to be set, then another domain was checked against the
3646 same list, then the first domain was re-checked, the value of $domain_data
3647 after the final check could be wrong. In particular, if the second check
3648 failed, it could be set empty. This bug probably also applied to
3651 41. The strip_trailing_dot option was not being applied to the address given
3652 with the -f command-line option.
3654 42. The code for reading a message's header from the spool was incrementing
3655 $received_count, but never initializing it. This meant that the value was
3656 incorrect (doubled) while delivering a message in the same process in which
3657 it was received. In the most common configuration of Exim, this never
3658 happens - a fresh exec is done - but it can happen when
3659 deliver_drop_privilege is set.
3661 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3662 it now includes up to 150 characters of the unexpected data in the log
3665 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3666 and building data strings. The size of both of these buffers was 10 000
3667 bytes - far larger than anybody would *ever* want, thought I. Needless to
3668 say, somebody hit the limit. I have increased the maximum line length to
3669 20 000 and the maximum data length of concatenated lines to 100 000. I have
3670 also fixed two bugs, because there was no checking on these buffers. Tsk,
3671 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3672 buffer is too small.
3674 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3675 lsearch lookups. Now it does.
3677 46. When parsing a route_list item in a manualroute router, a fixed-length
3678 buffer was used for the list of hosts. I made this 1024 bytes long,
3679 thinking that nobody would ever have a list of hosts that long. Wrong.
3680 Somebody had a whole pile of complicated expansion conditions, and the
3681 string was silently truncated, leading to an expansion error. It turns out
3682 that it is easier to change to an unlimited length (owing to other changes
3683 that have happened since this code was originally written) than to build
3684 structure for giving a limitation error. The length of the item that
3685 expands into the list of hosts is now unlimited.
3687 47. The lsearch lookup could not handle data where the length of text line was
3688 more than 4095 characters. Such lines were truncated, leading to shortened
3689 data being returned. It should now handle lines of any length.
3691 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3692 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3695 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3696 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3697 no longer happens in comment lines. A list of replacements is now placed
3698 at the head of all of the source files, except those whose only change is
3699 to replace PERL_COMMAND in the very first #! line.
3701 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3702 messages on the queue, with a bottom-up merge sort, using code contributed
3703 by Michael Haardt. This should make operations like -bp somewhat faster on
3704 large queues. It won't affect queue runners, except when queue_run_in_order
3707 51. Installed eximstats 1.31 in the distribution.
3709 52. Added support for SRV lookups to the dnslookup router.
3711 53. If an ACL referred to $message_body or $message_body_end, the value was not
3712 reset for any messages that followed in the same SMTP session.
3714 54. The store-handling optimization for building very long strings was not
3715 differentiating between the different store pools. I don't think this
3716 actually made any difference in practice, but I've tidied it.
3718 55. While running the routers to verify a sender address, $sender_address
3719 was still set to the sender address. This is wrong, because when routing to
3720 send a bounce to the sender, it would be empty. Therefore, I have changed
3721 it so that, while verifying a sender address, $sender_address is set to <>.
3722 (There is no change to what happens when verifying a recipient address.)
3724 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3725 target A or AAAA records (if not already returned) without resetting the
3726 qualify_single or search_parents options of the DNS resolver. These are
3727 inappropriate in this case because the targets of MX and SRV records must
3728 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3729 target that, when qualified, matched something in the local domain. These
3730 two options are now turned off when doing these lookups.
3732 57. It seems that at least some releases of Reiserfs (which does not have the
3733 concept of a fixed number of inodes) returns zero and not -1 for the
3734 number of available inodes. This interacted badly with check_spool_inodes,
3735 which assumed that -1 was the "no such thing" setting. What I have done is
3736 to check that the total number of inodes is greater than zero before doing
3737 the test of how many are available.
3739 58. When a "warn" ACL statement has a log_message modifier, the message is
3740 remembered, and not repeated. This is to avoid a lot of repetition when a
3741 message has many recipients that cause the same warning to be written.
3742 Howewer, Exim was preserving the list of already written lines for an
3743 entire SMTP session, which doesn't seem right. The memory is now reset if a
3744 new message is started.
3746 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3747 debugging output unless log_rewrite was also set.
3749 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3750 of (int)(handle) when we know that handle contains (void *)(-1).
3752 61. The Exim daemon panic-logs an error return when it closes the incoming
3753 connection. However "connection reset by peer" seems to be common, and
3754 isn't really an error worthy of noting specially, so that particular error
3757 62. When Exim is trying to find all the local interfaces, it used to panic and
3758 die if the ioctl to get the interface flags failed. However, it seems that
3759 on at least one OS (Solaris 9) it is possible to have an interface that is
3760 included in the list of interfaces, but for which you get a failure error
3761 for this call. This happens when the interface is not "plumbed" into a
3762 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3763 failure of the "get flags" call assumes that the interface is down.
3765 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3766 makes life easier for people who are doing arithmetic on fields extracted
3767 from dates, where you often get leading zeros that should not be
3768 interpreted as octal.
3770 64. Added qualify_domain to the redirect router, to override the global
3773 65. If a pathologically long header line contained very many addresses (the
3774 report of this problem mentioned 10 000) and each of them was rewritten,
3775 Exim could use up a very large amount of memory. (It kept on making new
3776 copies of the header line as it rewrote, and never released the old ones.)
3777 At the expense of a bit more processing, the header rewriting function has
3778 been changed so that it no longer eats memory in this way.
3780 66. The generation of the Received: header has been moved from the time that a
3781 message starts to be received, to the time that it finishes. The timestamp
3782 in the Received: header should now be very close to that of the <= log
3783 line. There are two side-effects of this change:
3785 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3786 logged header lines no longer include the local Received: line, because
3787 it has not yet been created. The same applies to a copy of the message
3788 that is returned to a non-SMTP sender when a message is rejected.
3790 (b) When a filter file is tested using -bf, no additional Received: header
3791 is added to the test message. After some thought, I decided that this
3794 This change does not affect the value of $received_for. It is still set
3795 after address rewriting, but before local_scan() is called.
3797 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3799 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3800 gave an unhelpful panic error message, and a defer error. I have managed to
3801 change this behaviour so that it now rejects any supplied certificate,
3802 which seems right, as the list of acceptable certificates is empty.
3804 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3805 gave an unhelpful defer error. I have not managed to make this reject any
3806 supplied certificates, but the error message it gives is "no certificate
3807 supplied", which is not helpful.
3809 70. exigrep's output now also includes lines that are not associated with any
3810 message, but which match the given pattern. Implemented by a patch from
3811 Martin Sluka, which also tidied up the Perl a bit.
3813 71. Recipient callout verification, like sender verification, was using <> in
3814 the MAIL FROM command. This isn't really the right thing, since the actual
3815 sender may affect whether the remote host accepts the recipient or not. I
3816 have changed it to use the actual sender in the callout; this means that
3817 the cache record is now keyed on a recipient/sender pair, not just the
3818 recipient address. There doesn't seem to be a real danger of callout loops,
3819 since a callout by the remote host to check the sender would use <>.
3820 [SEE ABOVE: changed after hitting problems.]
3822 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3823 temporary errors. However, in the case of such a code being given after
3824 the end of a data transmission (i.e. after ".") Exim was failing to write
3825 a retry record for the message. (Yes, there was some broken host that was
3826 actually sending 8xx at this point.)
3828 73. An unknown lookup type in a host list could cause Exim to panic-die when
3829 the list was checked. (An example that provoked this was putting <; in the
3830 middle of a list instead of at the start.) If this happened during a DATA
3831 ACL check, a -D file could be left lying around. This kind of configuration
3832 error no longer causes Exim to die; instead it causes a defer errror. The
3833 incident is still logged to the main and panic logs.
3835 74. Buglet left over from Exim 3 conversion. The message "too many messages
3836 in one connection" was written to the rejectlog but not the mainlog, except
3837 when address rewriting (yes!) was being logged.
3839 75. Added write_rejectlog option.
3841 76. When a system filter was run not as root (that is, when system_filter_user
3842 was set), the values of the $n variables were not being returned to the
3843 main process; thus, they were not subsequently available in the $sn
3846 77. Added +return_path_on_delivery log selector.
3848 78. A connection timeout was being treated differently from recipients deferred
3849 when testing hosts_max_try with a message that was older than the host's
3850 retry timeout. (The host should not be counted, thus allowing all hosts to
3851 be tried at least once before bouncing.) This may have been the cause of an
3852 occasionally reported bug whereby a message would remain on the queue
3853 longer than the retry timeout, but would be bounced if a delivery was
3854 forced. I say "may" because I never totally pinned down the problem;
3855 setting up timeout/retry tests is difficult. See also the next item.
3857 79. The ultimate address timeout was not being applied to errors that involved
3858 a combination of host plus message (for example, a timeout on a MAIL
3859 command). When an address resolved to a number of possible hosts, and they
3860 were not all tried for each delivery (e.g. because of hosts_max_try), a
3861 message could remain on the queue longer than the retry timeout.
3863 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3866 81. Fixed an obscure SMTP outgoing bug which required at least the following
3867 conditions: (a) there was another message waiting for the same server;
3868 (b) the server returned 5xx to all RCPT commands in the first message so
3869 that the message was not completed; (c) the server dropped the connection
3870 or gave a negative response to the RSET that Exim sends to abort the
3871 transaction. The observed case was a dropped connection after DATA that had
3872 been sent in pipelining mode. That is, the server had advertised PIPELINING
3873 but was not implementing it correctly. The effect of the bug was incorrect
3874 behaviour, such as trying another host, and this could lead to a crash.
3880 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3881 and daemon.c were passed as pointers to ints; they should have been
3882 pointers to socklen_t variables (which are typically unsigned ints).
3884 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3887 3. Fixed a really odd bug that affected only the testing scheme; patching a
3888 certain fixed string in the binary changed the value of another string that
3889 happened to be identical to the end of the original first string.
3891 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3892 name", it returns that address as the IP address. On some operating
3893 systems (e.g. Solaris), it also passes back the IP address string as the
3894 "host name". However, on others (e.g. Linux), it passes back an empty
3895 string. Exim wasn't checking for this, and was changing the host name to an
3896 empty string, assuming it had been canonicized.
3898 5. Although rare, it is permitted to have more than one PTR record for a given
3899 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3900 all the names associated with an address, because they do in Solaris.
3901 However, it seems that they do not in Linux for data that comes from the
3902 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3903 I found this out when I moved to a new Linux workstation and tried to run
3904 the Exim test suite.
3906 To get round this problem I have changed the code so that it now does its
3907 own call to the DNS to look up PTR records when searching for a host name.
3908 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3909 addresses that are only in /etc/hosts are still found.
3911 This behaviour is, however, controlled by an option called host_lookup_
3912 order, which defaults to "bydns:byaddr". If people want to use the other
3913 order, or indeed, just use one or the other means of lookup, they can
3914 specify it in this variable.
3916 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3917 some operating systems, this comes back from gethostbyaddr() as an empty
3918 string, and this is what Exim used to test for. However, it seems that in
3919 other systems, "." is yielded. Exim now tests for this case too.
3921 7. The values of check_spool_space and check_log_space are now held internally
3922 as a number of kilobytes instead of an absolute number of bytes. If a
3923 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3924 kilobyte. This means that much larger values can be stored.
3926 8. Exim monitor: an attempt to get the action menu when not actually pointing
3927 at a message produces an empty menu entitled "No message selected". This
3928 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3929 no entries in it ("Shell widget menu has zero width and/or height"). So I
3930 have added a single, blank menu entry in this case.
3932 9. Added ${quote_local_part.
3934 10. MIME decoding is now applied to the contents of Subject: header lines when
3937 11. Now that a reference to $sender_host_address automatically causes a reverse
3938 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3939 host lookup before query-style lookups in lists that might use this
3940 variable. This has therefore been abolished, and the "net-" prefix is no
3941 longer necessary for query-style lookups.
3943 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3944 have been a typo for LFLAGS, so it has been changed.
3946 13. The install script calls Exim with "-C /dev/null" in order to find the
3947 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3948 to be output. Howeve, since Exim outputs its version number before the
3949 error, it didn't break the script. It just looked ugly. I fixed this by
3950 always allowing "-C /dev/null" if the caller is root.
3952 14. Ignore overlarge ACL variable number when reading spool file - insurance
3953 against a later release with more variables having written the file.
3955 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3956 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3957 this, as well as the form without the "IPv6" on the front.
3959 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3960 OS/Makefile-Darwin file.
3962 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3963 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3965 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3966 variable. However, some OS use ints for inodes. I've added cast to long int
3967 to get rid of the compiler warning.
3969 19. I had forgotten to lock out "/../" in configuration file names when
3970 ALT_CONFIG_PREFIX was set.
3972 20. Routers used for verification do not need to specify transports. However,
3973 if such a router generated a host list, and callout was configured, Exim
3974 crashed, because it could not find a port number from the (non-existent)
3975 transport. It now assumes port 25 in this circumstance.
3977 21. Added the -t option to exigrep.
3979 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3980 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3981 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3982 them is set, LOOKUP_LSEARCH is forced.
3984 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3986 24. Added sender and host information to the "rejected by local_scan()" log
3987 line; previously there was no indication of these.
3989 25. Added .include_if_exists.
3991 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3992 Linux. It turns out that not all file systems support this. Apparently some
3993 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3994 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3995 not supported on the file descriptor, is now ignored when Exim is trying to
3996 sync a directory. This applies only to Linux.
3998 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
4000 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
4001 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
4004 29. Added support for maildirsize files from supplied patch (modified a bit).
4006 30. The use of :fail: followed by an empty string could lead Exim to respond to
4007 sender verification failures with (e.g.):
4009 550 Verification failed for <xxx>
4010 550 Sender verify failed
4012 where the first response line was missing the '-' that indicates it is not
4013 the final line of the response.
4015 31. The loop for finding the name of the user that called Exim had a hardwired
4016 limit of 10; it now uses the value of finduser_retries, which is used for
4017 all other user lookups.
4019 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4022 33. Exim was neglecting to zero errno before one call of strtol() when
4023 expanding a string and expecting an integer value. On some systems this
4024 resulted in spurious "integer overflow" errors. Also, it was casting the
4025 result into an int without checking.
4027 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4028 did not work. The code looks as if it has *never* worked, though it appears
4029 to have been documented since at least releast 1.62. I have made it work.
4031 35. The "timeout_DNS" error in retry rules, also documented since at least
4032 1.62, also never worked. As it isn't clear exactly what this means, and
4033 clearly it isn't a major issue, I have abolished the feature by treating it
4034 as "timeout", and writing a warning to the main and panic logs.
4036 36. The display of retry rules for -brt wasn't always showing the error code
4039 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4040 timeout_connect_A, timeout_connect_MX.
4042 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4043 to the empty sender.
4045 39. The daemon was not analysing the content of -oX till after it had closed
4046 stderr and disconnected from the controlling terminal. This meant that any
4047 syntax errors were only noted on the panic log, and the return code from
4048 the command was 0. By re-arranging the code a little, I've made the
4049 decoding happen first, so such errors now appear on stderr, and the return
4050 code is 1. However, the actual setting up of the sockets still happens in
4051 the disconnected process, so errors there are still only recorded on the
4054 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4055 connections (as happens on some IP stacks) was logged at start up time as
4056 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4057 it from "IPv6 and IPv4", which means that two separate sockets are being
4060 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4061 says whether AF_INET or AF_INET6 was passed as an argument.
4063 42. Exiwhat output was messed up when time zones were included in log
4066 43. Exiwhat now gives more information about the daemon's listening ports,
4067 and whether -tls-on-connect was used.
4069 44. The "port" option of the smtp transport is now expanded.
4071 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4072 silently ignored. Now an error message is written to the main and panic
4075 46. There's a new ACL modifier called "logwrite" which writes to a log file
4076 as soon as it is encountered.
4078 47. Added $local_user_uid and $local_user_gid at routing time.
4080 48. Exim crashed when trying to verify a sender address that was being
4083 49. Exim was recognizing only a space character after ".include". It now also
4084 recognizes a tab character.
4086 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4087 extracting the relevant information from the specification. The man page no
4088 longer contains scrambled data for the -d option, and I've added a section
4089 at the front about calling Exim under different names.
4091 51. Added "extra_headers" argument to the "mail" command in filter files.
4093 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4096 53. Installed eximstats 1.29.
4098 54. Added transport_filter_timeout as a generic transport option.
4100 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4101 Cc: header lines. This was required by RFC 822, but it not required by RFC
4104 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4105 incoming messages that did not have them. Now it does so only if the
4106 message originates locally, that is, if there is no associated remote host
4107 address. When Resent- header lines are present, this applies to the Resent-
4108 lines rather than the non-Resent- lines.
4110 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4111 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4113 58. Messages for configuration errors now include the name of the main
4114 configuration file - useful now that there may be more than one file in a
4115 list (.included file names were always shown).
4117 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4118 for those rare installations that do not start the daemon as root or run it
4119 setuid root. I've cut out the call to initgroups() if the daemon is not
4122 60. The Exim user and group can now be bound into the binary as text strings
4123 that are looked up at the start of Exim's processing.
4125 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4127 62. Added $mailstore_basename variable.
4129 63. Installed patch to sieve.c from Michael Haardt.
4131 64. When Exim failed to open the panic log after failing to open the main log,
4132 the original message it was trying to log was written to stderr and debug
4133 output, but if they were not available (the usual case in production), it
4134 was lost. Now it is written to syslog before the two lines that record the
4135 failures to open the logs.
4137 65. Users' Exim filters run in subprocesses under the user's uid. It is
4138 possible for a "deliver" command or an alias in a "personal" command to
4139 provoke an address rewrite. If logging of address rewriting is configured,
4140 this fails because the process is not running as root or exim. There may be
4141 a better way of dealing with this, but for the moment (because 4.30 needs
4142 to be released), I have disabled address rewrite logging when running a
4143 filter in a non-root, non-exim process.
4149 1. The buildconfig auxiliary program wasn't quoting the value set for
4150 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4151 not defined. This bug was masked in 4.22 by the effect that was fixed in
4154 2. Some messages that were rejected after a message id was allocated were
4155 shown as "incomplete" by exigrep. It no longer does this for messages that
4156 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4158 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4159 have allow_domain_literals set, the ID did not get logged in the <= line.
4160 Domain literals are now always recognized in Message-ID: header lines.
4162 4. The first argument for a ${extract expansion item is the key name or field
4163 number. Leading and trailing spaces in this item were not being ignored,
4164 causing some misleading effects.
4166 5. When deliver_drop_privilege was set, single queue runner processes started
4167 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4168 same command in the process it spins off) were not dropping privilege.
4170 6. When the daemon running as "exim" started a queue runner, it always
4171 re-executed Exim in the spun-off process. This is a waste of effort when
4172 deliver_drop_privilege is set. The new process now just calls the
4173 queue-runner function directly.
4179 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4182 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4183 ignored. Though the use of -oP was forcing the writing of a pid file, it
4184 was always written to the default place.
4186 3. If the message "no IP address found for host xxxx" is generated during
4187 incoming verification, it is now followed by identification of the incoming
4188 connection (so you can more easily find what provoked it).
4190 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4192 5. Added some features to "harden" Exim a bit more against certain attacks:
4194 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4195 be put in Local/Makefile. This is like the never_users runtime option,
4196 but it cannot be overridden. The default setting is "root".
4198 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4199 prefix string with which any file named in a -C command line option
4202 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4203 is retained for -C and -D only if the caller of Exim is root. Without
4204 it, the exim user may also use -C and -D and retain privilege.
4206 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4207 command line option is disabled.
4209 6. Macro names set by the -D option must start with an upper case letter, just
4210 like macro names defined in the configuration file.
4212 7. Added "dereference=" facility to LDAP.
4214 8. Two instances of the typo "uknown" in the source files are fixed.
4216 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4217 the Configure-Makefile script screwed up while processing it.
4219 10. Incorporated PCRE 4.4.
4221 11. The SMTP synchronization check was not operating right at the start of an
4222 SMTP session. For example, it could not catch a HELO sent before the client
4223 waited for the greeting. There is now a check for outstanding input at the
4224 point when the greeting is written. Because of the duplex, asynchronous
4225 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4226 way, but not yet received, when the check is performed.
4228 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4229 on TCP/IP sockets, because this apparently causes some broken clients to
4232 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4233 unchanged) from the Cygwin maintainer.
4235 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4236 was supported instead of testing for mbx. Effectively a typo.
4238 15. The spa authenticator server code was not checking that the input it
4239 received was valid base64.
4241 16. The debug output line for the "set" modifier in ACLs was not showing the
4242 name of the variable that was being set.
4244 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4246 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4247 lookup on its own. Something else was needed to trigger the lookup. For
4248 example, a match in host_lookup or the need for a host name in a host list.
4249 Now, if $sender_host_name is referenced and the host name has not yet been
4250 looked up, a lookup is performed. If the lookup fails, the variable remains
4251 empty, and $host_lookup_failed is set to "1".
4253 19. Added "eqi" as a case-independent comparison operator.
4255 20. The saslauthd authentication condition could segfault if neither service
4256 nor realm was specified.
4258 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4259 error message that was logged was misleading, and incoming SMTP
4260 connections were dropped. The message is now more accurate, and temporary
4261 errors are given to SMTP connections.
4263 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4264 (except RSET) with a 421 error, until QUIT is received. However, it was
4265 failing to send a response to QUIT.
4267 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4268 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4269 ACL is now run after the helo_try_verify_hosts code.
4271 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4272 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4273 case-independent, so other case variants are also recognized). Apparently
4274 some systems use these upper case variants.
4276 25. If more than two messages were waiting for the same host, and a transport
4277 filter was specified for the transport, Exim sent two messages over the
4278 same TCP/IP connection, and then failed with "socket operation on non-
4279 socket" when it tried to send the third.
4281 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4283 27. The extern definition of crypt16() in expand.c was not being excluded when
4284 the OS had its own crypt16() function.
4286 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4287 as a preferred synonym for return_size_limit, both as an option and as an
4290 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4292 30. Changed the default configuration ACL to relax the local part checking rule
4293 for addresses that are not in any local domains. For these addresses,
4294 slashes and pipe symbols are allowed within local parts, but the sequence
4295 /../ is explicitly forbidden.
4297 31. SPA server authentication was not clearing the challenge buffer before
4300 32. log_message in a "warn" ACL statement was writing to the reject log as
4301 well as to the main log, which contradicts the documentation and doesn't
4302 seem right (because no rejection is happening). So I have stopped it.
4304 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4305 However, I am unable to do any testing of this.
4307 34. Fixed an infelicity in the appendfile transport. When checking directories
4308 for a mailbox, to see if any needed to be created, it was accidentally
4309 using path names with one or more superfluous leading slashes; tracing
4310 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4312 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4313 contents are added to the log line that is written for every discarded
4314 recipient. (Previously a log_message setting was ignored.)
4316 36. The ${quote: operator now quotes the string if it is empty.
4318 37. The install script runs exim in order to find its version number. If for
4319 some reason other than non-existence or emptiness, which it checks, it
4320 could not run './exim', it was installing it with an empty version number,
4321 i.e. as "exim-". This error state is now caught, and the installation is
4324 38. An argument was missing from the function that creates an error message
4325 when Exim fails to connect to the socket for saslauthd authentication.
4326 This could cause Exim to crash, or give a corrupted message.
4328 39. Added isip, isip4, and isip6 to ${if conditions.
4330 40. The ACL variables $acl_xx are now saved with the message, and can be
4331 accessed later in routers, transports, and filters.
4333 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4334 strings in the file are not string-expanded.
4336 42. If a MAIL command specified a SIZE value that was too large to fit into an
4337 int variable, the check against message_size_limit failed. Such values are
4338 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4339 day this will have to be increased, but I don't think I want to be around
4340 when emails are that large.
4347 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4348 iconv() is not standard in FreeBSD.
4350 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4351 IPv6 enabled. The observed symptom was a segmentation fault on return from
4352 the function os_common_find_running_interfaces() in src/os.c.
4354 3. In the check_special_case() function in daemon.c I had used "errno" as an
4355 argument name, which causes warnings on some systems. This was basically a
4356 typo, since it was named "eno" in the comments!
4358 4. The code that waits for the clock to tick (at a resolution of some fraction
4359 of a second) so as to ensure message-id uniqueness was always waiting for
4360 at least one whole tick, when it could have waited for less. [This is
4361 almost certainly not relevant at current processor speeds, where it is
4362 unlikely to ever wait at all. But we try to future-proof.]
4364 5. The function that sleeps for a time interval that includes fractions of a
4365 second contained a race. It did not block SIGALRM between setting the
4366 timer, and suspending (a couple of lines later). If the interval was short
4367 and the sigsuspend() was delayed until after it had expired, the suspension
4368 never ended. On busy systems this could lead to processes getting stuck for
4371 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4372 process, before it forks any delivery processes. The open lookup caching
4373 mechanism meant that the open file or database connection was passed into
4374 the delivery process. The problem was that delivery processes always tidy
4375 up cached lookup data. This could cause a problem for the next delivery
4376 process started by the queue runner, because the external queue runner
4377 process does not know about the closure. So the next delivery process
4378 still has data in the lookup cache. In the case of a file lookup, there was
4379 no problem because closing a file descriptor in a subprocess doesn't affect
4380 the parent. However, if the lookup was caching a connection to a database,
4381 the connection was closed, and the second delivery process was likely to
4382 see errors such as "PGSQL: query failed: server closed the connection
4383 unexpectedly". The problem has been fixed by closing all cached lookups
4384 in a queue runner before running a delivery process.
4386 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4387 seem to have the "const" qualifier which it has on other OS. I've
4390 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4391 *of the same type* (client or server) with the same public name that an
4392 error should be diagnosed.
4394 9. When Exim looked up a host name for an IP address, but failed to find the
4395 original IP address when looking up the host name (a safety check), it
4396 output the message "<ip address> does not match any IP for NULL", which was
4397 confusing, to say the least. The bug was that the host name should have
4398 appeared instead of "NULL".
4400 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4401 user that is built into the binary, and the -C or -D options is used, root
4402 privilege is dropped before the configuration file is read. In addition,
4403 logging is switched to stderr instead of the normal log files. If the
4404 configuration then re-defines the Exim user, the unprivileged environment
4405 is probably not what is expected, so Exim logs a panic warning message (but
4408 However, if deliver_drop_privilege is set, the unprivileged state may well
4409 be exactly what is intended, so the warning has been cut out in that case,
4410 and Exim is allowed to try to write to its normal log files.
4416 1. smtp_return_error_details was not giving details for temporary sender
4417 or receiver verification errors.
4419 2. Diagnose a configuration error if two authenticators have the same public
4422 3. Exim used not to create the message log file for a message until the first
4423 delivery attempt. This could be confusing when incoming messages were held
4424 for policy or load reasons. The message log file is now created at the time
4425 the message is received, and an initial "Received" line is written to it.
4427 4. The automatically generated man page for command line options had a minor
4428 bug that caused no ill effects; however, a more serious problem was that
4429 the procedure for building the man page automatically didn't always
4430 operate. Consequently, release 4.20 contains an out-of-date version. This
4431 shouldn't happen again.
4433 5. When building Exim with embedded Perl support, the script that builds the
4434 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4435 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4437 6. The freeze_tell option was not being used for messages that were frozen on
4438 arrival, either by an ACL or by local_scan().
4440 7. Added the smtp_incomplete_transaction log selector.
4442 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4443 was accepting AUTH without a new EHLO.
4445 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4446 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4449 10. Logging of TCP/IP connections (when configured) now happens in the main
4450 daemon process instead of the child process, so that the TCP/IP connection
4451 count is more accurate (but it can never be perfect).
4453 11. The use of "drop" in a nested ACL was not being handled correctly in the
4454 outer ACL. Now, if condition failure induced by the nested "drop" causes
4455 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4456 or "require"), the connection is dropped.
4458 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4459 that yield "discard" can now be used with an "accept" or a "discard" verb,
4460 but an error is generated for any others (because I can't see a useful way
4461 to define what should happen).
4463 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4464 are now processed in the same way as lines in the Exim configuration file.
4465 In particular, continuation lines are supported.
4467 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4469 15. Added -ti meaning -t -i.
4471 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4472 domains, and warn by logging if others are found.
4474 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4475 returned from GIFCONF when Exim is trying to find the list of interfaces on
4476 a host. The code in os.c has been modified to copy each ifreq to an aligned
4477 structure in all cases.
4479 Also, in some cases, the returned ifreq's were being copied to a 'struct
4480 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4481 means the last couple of bytes of an IPv6 address could be chopped if the
4482 ifreq contained only a normal sockaddr (14 bytes storage).
4484 18. Named domain lists were not supported in the hosts_treat_as_local option.
4485 An entry such as +xxxx was not recognized, and was treated as a literal
4488 19. Ensure that header lines added by a DATA ACL are included in the reject log
4489 if the ACL subsequently rejects the message.
4491 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4492 MD5 (which is deprecated).
4494 21. When testing a filter file using -bf, Exim was writing a message when it
4495 took the sender from a "From " line in the message, but it was not doing so
4496 when it took $return_path from a Return-Path: header line. It now does.
4498 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4499 with a valid header line field name (a series of printing characters
4500 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4502 23. Changed "disc" in the source to "disk" to conform to the documentation and
4503 the book and for uniformity.
4505 24. Ignore Sendmail's -Ooption=value command line item.
4507 25. When execve() failed while trying to run a command in a pipe transport,
4508 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4509 could be confused with a return value of 69 from the command itself. This
4510 has been changed to 127, the value the shell returns if it is asked to run
4511 a non-existent command. The wording for the related log line suggests a
4512 non-existent command as the problem.
4514 26. If received_header_text expands to an empty string, do not add a Received:
4515 header line to the message. (Well, it adds a token one on the spool, but
4516 marks it "old" so that it doesn't get used or transmitted.)
4518 27. Installed eximstats 1.28 (addition of -nt option).
4520 28. There was no check for failure on the call to getsockname() in the daemon
4521 code. This can fail if there is a shortage of resources on the system, with
4522 ENOMEM, for example. A temporary error is now given on failure.
4524 29. Contrary to the C standard, it seems that in some environments, the
4525 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4526 program. Exim now does this explicitly; it affects the formatting of
4527 timestamps using strftime().
4529 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4530 complaints. I've now initialized all the variables, to avoid this.
4532 32. Header lines added by a system filter were not being "seen" during
4533 transport-time rewrites.
4535 33. The info_callback() function passed to OpenSSL is set up with type void
4536 (*)(SSL *, int, int), as described somewhere. However, when calling the
4537 function (actually a macro) that sets it up, the type void(*)() is
4538 expected. I've put in a cast to prevent warnings from picky compilers.
4540 34. If a DNS black list lookup found a CNAME record, but there were no A
4541 records associated with the domain it pointed at, Exim crashed.
4543 35. If a DNS black list lookup returned more than one A record, Exim ignored
4544 all but the first. It now scans all returned addresses if a particular IP
4545 value is being sought. In this situation, the contents of the
4546 $dnslist_value variable are a list of all the addresses, separated by a
4549 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4550 to accept a host name and all its aliases if the forward lookup for any of
4551 them yielded the IP address of the incoming connection. Now it accepts only
4552 those names whose forward lookup yields the correct IP address. Any other
4553 names are discarded. This closes a loophole whereby a rogue DNS
4554 administrator could create reverse DNS records to break through a
4555 wildcarded host restriction in an ACL.
4557 37. If a user filter or a system filter that ran in a subprocess used any of
4558 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4559 the wrong values were passed to the pipe command ($thisaddress had the
4560 value of $0, $0 had the value of $1, etc). This bug was introduced by
4561 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4563 38. Improved the line breaking for long SMTP error messages from ACLs.
4564 Previously, if there was no break point between 40 and 75 characters, Exim
4565 left the rest of the message alone. Two changes have been made: (a) I've
4566 reduced the minimum length to 35 characters; (b) if it can't find a break
4567 point between 35 and 75 characters, it looks ahead and uses the first one
4568 that it finds. This may give the occasional overlong line, but at least the
4569 remaining text gets split now.
4571 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4572 file descriptors might be low, and that setting 1000 would always raise it.
4573 It turns out that in some environments, the limit is already over 1000 and
4574 that lowering it causes trouble. So now Exim takes care not to decrease it.
4576 40. When delivering a message, the value of $return_path is set to $sender_
4577 address at the start of routing (routers may change the value). By an
4578 oversight, this default was not being set up when an address was tested by
4579 -bt or -bv, which affected the outcome if any router or filter referred to
4582 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4583 log only when "message" or "log_message" are set. However, if one of the
4584 conditions was an address verification, or a call to a nested ACL, the
4585 messages generated by the underlying test were being passed through. This
4586 no longer happens. The underlying message is available in $acl_verify_
4587 message for both "message" and "log_message" expansions, so it can be
4588 passed through if needed.
4590 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4591 new expansion $bh_ to give the encoded byte string without charset
4592 translation. Translation happens only if iconv() is available; HAVE_ICONV
4593 indicates this at build time. HEADERS_CHARSET gives the charset to
4594 translate to; headers_charset can change it in the configuration, and
4595 "headers charset" can change it in an individual filter file.
4597 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4598 that creates RFC 2047 encoded "words" labels them as that charset instead
4599 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4600 expansion operator; (ii) when Exim creates a From: line for a local
4601 message; (iii) when a header line is rewritten to include a "phrase" part.
4603 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4604 buggy, causing it to skip the first lines of messages whose message ID
4605 ended in 'D'. This would not have bitten before Exim release 4.14, because
4606 message IDs were unlikely to end in 'D' before then. The effect was to have
4607 incorrect size information for certain domains.
4609 45. #include "config.h" was missing at the start of the crypt16.c module. This
4610 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4613 46. If there was a timeout during a "random" callout check, Exim treated it as
4614 a failure of the random address, and carried on sending RSET and the real
4615 address. If the delay was just some slowness somewhere, the response to the
4616 original RCPT would be taken as a response to RSET and so on, causing
4617 mayhem of various kinds.
4619 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4620 when I implemented it. It didn't allow for the fact that some option values
4621 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4622 the right test for positive values.
4624 48. Domain names in DNS records are case-independent. Exim always looks them up
4625 in lower case. Some resolvers return domain names in exactly the case they
4626 appear in the zone file, that is, they may contain uppercase letters. Not
4627 all resolvers do this - some return always lower case. Exim was treating a
4628 change of case by a resolver as a change of domain, similar to a widening
4629 of a domain abbreviation. This triggered its re-routing code and so it was
4630 trying to route what was effectively the same domain again. This normally
4631 caused routing to fail (because the router wouldn't handle the domain
4632 twice). Now Exim checks for this case specially, and just changes the
4633 casing of the domain that it ultimately uses when it transmits the message
4636 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4639 50. If a filter generated a file delivery with a non-absolute name (possible if
4640 no home directory exists for the router), the forbid_file option was not
4643 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4644 the existing equality matching.
4646 52. Exim was using ints instead of ino_t variables in some places where it was
4647 dealing with inode numbers.
4649 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4650 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4651 called TMPDIR, and if it finds it is different, it changes its value.
4653 54. The smtp_printf() function is now made available to local_scan() so
4654 additional output lines can be written before returning. There is also an
4655 smtp_fflush() function to enable the detection of a dropped connection.
4656 The variables smtp_input and smtp_batched_input are exported to
4659 55. Changed the default runtime configuration: the message "Unknown user"
4660 has been removed from the ACL, and instead placed on the localuser router,
4661 using the cannot_route_message feature. This means that any verification
4662 failures that generate their own messages won't get overridden. Similarly,
4663 the "Unrouteable address" message that was in the ACL for unverifiable
4664 relay addresses has also been removed.
4666 56. Added hosts_avoid_esmtp to the smtp transport.
4668 57. The exicyclog script was not checking for the esoteric option
4669 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4670 will work only if exicyclog is run under the appropriate euid.
4672 58. Following a discussion on the list, the rules by which Exim recognises line
4673 endings on incoming messages have been changed. The -dropcr and drop_cr
4674 options are now no-ops, retained only for backwards compatibility. The
4675 following line terminators are recognized: LF CRLF CR. However, special
4676 processing applies to CR:
4678 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4679 nor a local message in the state where . is a terminator.
4681 (ii) If a bare CR is encountered in a header line, an extra space is added
4682 after the line terminator so as not to end the header. The reasoning
4683 behind this is that bare CRs in header lines are most likely either
4684 to be mistakes, or people trying to play silly games.
4686 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4687 was being incorrectly given as 18 bytes larger than it should have been.
4688 This is a VOB (very old bug).
4690 60. This may never have affected anything current, but just in case it has:
4691 When the local host is found other than at the start of a list of hosts,
4692 the local host, those with the same MX, and any that follow, are discarded.
4693 When the list in question was part of a longer list of hosts, the following
4694 hosts (not currently being processed) were also being discarded. This no
4695 longer happens. I'm not sure if this situation could ever has previously
4698 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4701 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4702 header. This is something that is recommended in a new Internet Draft, and
4703 is something that is documented as being done by Sendmail. There are two
4704 possible values. For messages generated by the autoreply transport, Exim
4707 Auto-Submitted: auto-replied
4709 whereas for all other generated messages (e.g. bounces) it adds
4711 Auto-Submitted: auto-generated
4713 63. The "personal" condition in filters now includes a test for the
4714 Auto-Submitted: header. If it contains the string "auto-" the message it
4715 not considered personal.
4717 64. Added rcpt_include_affixes as a generic transport option.
4719 65. Added queue_only_override (default true).
4721 66. Added the syslog_duplication option.
4723 67. If what should have been the first header line of a message consisted of
4724 a space followed by a colon, Exim was mis-interpreting it as a header line.
4725 It isn't of course - it is syntactically invalid and should therefore be
4726 treated as the start of the message body. The misbehaviour could have
4727 caused a number of strange effects, including loss of data in subsequent
4728 header lines, and spool format errors.
4730 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4731 client host had authenticated. This control can now be exercised by an ACL
4732 for more flexibility.
4734 69. By default, callouts do not happen when testing with -bh. There is now a
4735 variant, -bhc, which does actually run the callout code, including
4736 consulting and updating the callout cache.
4738 70. Added support for saslauthd authentication, courtesy of Alexander
4741 71. If statvfs() failed on the spool or log directories while checking their
4742 size for availability, Exim confusingly gave the error "space shortage".
4743 Furthermore, in debugging mode it crashed with a floating point exception.
4744 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4745 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4746 serious problem, Exim now writes to the main and panic logs when this
4747 happens, with details of the failure. It then refuses to accept the
4748 incoming message, giving the message "spool directory problem" or "log
4749 directory problem" with a 421 code for SMTP messages.
4751 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4752 0, 1, and 2 exist, because some OS complain for execs without them (see
4753 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4754 descriptors. However, the code omitted to check that the open succeeded,
4755 causing mysterious errors if for some reason the permissions on /dev/null
4756 got screwed. Now Exim writes a message to the main and panic logs, and
4757 bombs out if it can't open /dev/null.
4759 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4760 interact so that it is all more flexible. It is supposed to remain
4761 backwards compatible. Also added extra_local_interfaces.
4763 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4764 to bomb out with an assertion failure - to the client this appears as a
4765 connection drop. This problem occurs in the part of the code that was taken
4766 from the Samba project. Fortunately, the assertion is in a very simple
4767 function, so I have fixed this by reproducing the function inline in the
4768 one place where it is called, and arranging for authentication to fail
4769 instead of killing the process with assert().
4771 75. The SPA client code was not working when the server requested OEM rather
4772 than Unicode encoding.
4774 76. Added code to make require_files with a specific uid setting more usable in
4775 the case where statting the file as root fails - usually a non-root-mounted
4776 NFS file system. When this happens and the failure is EACCES, Exim now
4777 forks a subprocess and does the per-uid checking as the relevant uid.
4779 77. Added process_log_path.
4781 78. If log_file_path was not explicitly set, a setting of check_log_space or
4782 check_log_inodes was ignored.
4784 79. If a space check for the spool or log partitions fails, the incident is now
4785 logged. Of course, in the latter case the data may get lost...
4787 80. Added the %p formatting code to string_format() so that it can be used to
4788 print addresses in debug_print(). Adjusted all the address printing in the
4789 debugging in store.c to use %p rather than %d.
4791 81. There was a concern that a line of code in smtp_in.c could overflow a
4792 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4793 initially expressed, the concern was not well-founded, because trailing
4794 spaces are removed early. However, if the trailing spaces were followed by
4795 a NULL, they did not get removed, so the overflow was possible. Two fixes
4798 (a) I re-wrote the offending code in a cleaner fashion.
4799 (b) If an incoming SMTP command contains a NULL character, it is rejected
4802 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4803 runs initgroups(), so that if the Exim user is in any additional groups,
4804 they will be used during message reception.
4810 The change log for 4.20 and earlier releases has been archived.
projects / exim.git / blob