1 $Cambridge: exim/doc/doc-txt/ChangeLog,v 1.516 2007/06/20 14:13:39 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.
74 MH/01 Fix for bug #448, segfault in Dovecot authenticator when interface_address
75 is unset (happens when testing with -bh and -oMi isn't used). Thanks to
78 PH/01 Added a new log selector smtp_no_mail, to log SMTP sessions that do not
81 PH/02 In an ACL statement such as
83 deny dnslists = X!=127.0.0.2 : X=127.0.0.2
85 if a client was not listed at all, or was listed with a value other than
86 127.0.0.2, in the X list, but was listed with 127.0.0.2 in the Y list,
87 the condition was not true (as it should be), so access was not denied.
88 The bug was that the ! inversion was incorrectly passed on to the second
89 item. This has been fixed.
91 PH/03 Added additional dnslists conditions == and =& which are different from
92 = and & when the dns lookup returns more than one IP address.
94 PH/04 Added gnutls_require_{kx,mac,protocols} to give more control over the
95 cipher suites used by GnuTLS. These options are ignored by OpenSSL.
97 PH/05 After discussion on the list, added a compile time option ENABLE_DISABLE_
98 FSYNC, which compiles an option called disable_fsync that allows for
99 bypassing fsync(). The documentation is heavily laced with warnings.
101 SC/01 Updated eximstats to collate all SpamAssassin rejects into one bucket.
103 PH/06 Some tidies to the infrastructure of the Test Suite that is concerned
104 with the auxiliary C programs that it uses: (1) Arrange for BIND_8_COMPAT
105 to be defined when compiling on OSX (Darwin); (2) Tidies to the Makefile,
106 including adding "make clean"; (3) Added -fPIC when compiling the test
107 dynamically loaded module, to get rid of a warning.
109 MH/02 Fix for bug #451, causing paniclog entries to be written if a bounce
110 message fails, move_frozen_messages = true and ignore_bounce_errors_after
111 = 0s. The bug is otherwise harmless.
113 PH/07 There was a bug in the dovecot authenticator such that the value of
114 $auth1 could be overwritten, and so not correctly preserved, after a
115 successful authentication. This usually meant that the value preserved by
116 the server_setid option was incorrect.
118 PH/08 Added $smtp_count_at_connection_start, deliberately with a long name.
120 PH/09 Installed PCRE release 7.0.
122 PH/10 The acl_not_smtp_start ACL was, contrary to the documentation, not being
123 run for batched SMTP input. It is now run at the start of every message
124 in the batch. While fixing this I discovered that the process information
125 (output by running exiwhat) was not always getting set for -bs and -bS
126 input. This is fixed, and it now also says "batched" for BSMTP.
128 PH/11 Added control=no_pipelining.
130 PH/12 Added $sending_ip_address and $sending_port (mostly Magnus Holmgren's
131 patch, slightly modified), and move the expansion of helo_data till after
132 the connection is made in the smtp transport (so it can use these
135 PH/13 Added ${rfc2047d: to decoded RFC 2047 strings.
137 PH/14 Added log_selector = +pid.
139 PH/15 Flush SMTP output before delaying, unless control=no_delay_flush is set.
141 PH/16 Add ${if forany and ${if forall.
143 PH/17 Added dsn_from option to vary the From: line in DSNs.
145 PH/18 Flush SMTP output before performing a callout, unless control =
146 no_callout_flush is set.
148 PH/19 Change 4.64/PH/36 introduced a bug: when address_retry_include_sender
149 was true (the default) a successful delivery failed to delete the retry
150 item, thus causing premature timeout of the address. The bug is now
153 PH/20 Added hosts_avoid_pipelining to the smtp transport.
155 PH/21 Long custom messages for fakedefer and fakereject are now split up
156 into multiline reponses in the same way that messages for "deny" and
157 other ACL rejections are.
159 PH/22 Applied Jori Hamalainen's speed-up changes and typo fixes to exigrep,
160 with slight modification.
162 PH/23 Applied sieve patches from the maintainer "tracking the latest notify
163 draft, changing the syntax and factoring some duplicate code".
165 PH/24 When the log selector "outgoing_port" was set, the port was shown as -1
166 for deliveries of the second and subsequent messages over the same SMTP
169 PH/25 Applied Magnus Holmgren's patch for ${addresses, ${map, ${filter, and
170 ${reduce, with only minor "tidies".
172 SC/02 Applied Daniel Tiefnig's patch to improve the '($parent) =' pattern match.
174 PH/26 Added a "continue" ACL modifier that does nothing, for the benefit of its
175 expansion side effects.
177 PH/27 When a message times out after an over-quota error from an Exim-imposed
178 quota, the bounce message says "mailbox is full". This message was not
179 being given when it was a system quota that was exceeded. It now should
182 MH/03 Made $recipients available in local_scan(). local_scan() already has
183 better access to the recipient list through recipients_list[], but
184 $recipients can be useful in postmaster-provided expansion strings.
186 PH/28 The $smtp_command and $smtp_command_argument variables were not correct
187 in the case of a MAIL command with additional options following the
188 address, for example: MAIL FROM:<foo@bar> SIZE=1234. The option settings
189 were accidentally chopped off.
191 PH/29 SMTP synchronization checks are implemented when a command is read -
192 there is a check that no more input is waiting when there shouldn't be
193 any. However, for some commands, a delay in an ACL can mean that it is
194 some time before the response is written. In this time, more input might
195 arrive, invalidly. So now there are extra checks after an ACL has run for
196 HELO/EHLO and after the predata ACL, and likewise for MAIL and RCPT when
197 pipelining has not been advertised.
199 PH/30 MH's patch to allow iscntrl() characters to be list separators.
201 PH/31 Unlike :fail:, a custom message specified with :defer: was not being
202 returned in the SMTP response when smtp_return_error_details was false.
205 PH/32 Change the Dovecot authenticator to use read() and write() on the socket
206 instead of the C I/O that was originally supplied, because problems were
209 PH/33 Compile failed with OpenSSL 0.9.8e. This was due to a coding error in
210 Exim which did not show up earlier: it was assuming that a call to
211 SSL_CTX_set_info_callback() might give an error value. In fact, there is
212 no error. In previous releases of OpenSSL, SSL_CTX_set_info_callback()
213 was a macro that became an assignment, so it seemed to work. This has
214 changed to a proper function call with a void return, hence the compile
215 error. Exim's code has been fixed.
217 PH/34 Change HDA_SIZE in oracle.c from 256 to 512. This is needed for 64-bit
220 PH/35 Applied a patch from the Sieve maintainer which fixes a bug in "notify".
222 PH/36 Applied John Jetmore's patch to add -v functionality to exigrep.
224 PH/37 If a message is not accepted after it has had an id assigned (e.g.
225 because it turns out to be too big or there is a timeout) there is no
226 "Completed" line in the log. When some messages of this type were
227 selected by exigrep, they were listed as "not completed". Others were
228 picked up by some special patterns. I have improved the selection
229 criteria to be more general.
231 PH/38 The host_find_failed option in the manualroute router can now be set
232 to "ignore", to completely ignore a host whose IP address cannot be
233 found. If all hosts are ignored, the behaviour is controlled by the new
234 host_all_ignored option.
236 PH/39 In a list of hosts for manualroute, if one item (either because of multi-
237 homing or because of multiple MX records with /mx) generated more than
238 one IP address, and the following item turned out to be the local host,
239 all the secondary addresses of the first item were incorrectly removed
240 from the list, along with the local host and any following hosts (which
241 is what is supposed to happen).
243 PH/40 When Exim receives a message, it writes the login name, uid, and gid of
244 whoever called Exim into the -H file. In the case of the daemon it was
245 behaving confusingly. When first started, it used values for whoever
246 started the daemon, but after a SIGHUP it used the Exim user (because it
247 calls itself on a restart). I have changed the code so that it now always
250 PH/41 (Following a suggestion from Tony Finch) If all the RCPT commands in a
251 message are rejected with the same error (e.g. no authentication or bad
252 sender address), and a DATA command is nevertheless sent (as can happen
253 with PIPELINING or a stupid MUA), the error message that was given to the
254 RCPT commands is included in the rejection of the DATA command. This is
255 intended to be helpful for MUAs that show only the final error to their
258 PH/42 Another patch from the Sieve maintainer.
260 SC/02 Eximstats - Differentiate between permanent and temporary rejects.
261 Eximstats - Fixed some broken HTML links and added missing column headers
263 Eximstats - Fixed Grand Total Summary Domains, Edomains, and Email
264 columns for Rejects, Temp Rejects, Ham, and Spam rows.
266 SC/03 Eximstats - V1.58 Fix to get <> and blackhole to show in edomain tables.
268 PH/43 Yet another patch from the Sieve maintainer.
270 PH/44 I found a way to check for a TCP/IP connection going away before sending
271 the response to the final '.' that terminates a message, but only in the
272 case where the client has not sent further data following the '.'
273 (unfortunately, this is allowed). However, in many cases there won't be
274 any further data because there won't be any more messages to send. A call
275 to select() can be used: if it shows that the input is "ready", there is
276 either input waiting, or the socket has been closed. An attempt to read
277 the next input character can distinguish the two cases. Previously, Exim
278 would have sent an OK response which the client would never have see.
279 This could lead to message repetition. This fix should cure that, at
280 least in a lot of common cases.
282 PH/45 Do not advertise STARTTLS in response to HELP unless it would be
283 advertised in response to EHLO.
289 PH/01 Two more bugs that were introduced by 4.64/PH/07, in addition to the one
290 fixed by 4.65/MH/01 (is this a record?) are fixed:
292 (i) An empty string was always treated as zero by the numeric comparison
293 operators. This behaviour has been restored.
295 (ii) It is documented that the numeric comparison operators always treat
296 their arguments as decimal numbers. This was broken in that numbers
297 starting with 0 were being interpreted as octal.
299 While fixing these problems I realized that there was another issue that
300 hadn't been noticed. Values of message_size_limit (both the global option
301 and the transport option) were treated as octal if they started with 0.
302 The documentation was vague. These values are now always treated as
303 decimal, and I will make that clear in the documentation.
309 TK/01 Disable default definition of HAVE_LINUX_SENDFILE. Clashes with
310 Linux large file support (_FILE_OFFSET_BITS=64) on older glibc
313 MH/01 Don't check that the operands of numeric comparison operators are
314 integers when their expansion is in "skipping" mode (fixes bug
315 introduced by 4.64-PH/07).
317 PH/01 If a system filter or a router generates more than SHRT_MAX (32767)
318 child addresses, Exim now panics and dies. Previously, because the count
319 is held in a short int, deliveries were likely to be lost. As such a
320 large number of recipients for a single message is ridiculous
321 (performance will be very, very poor), I have chosen to impose a limit
322 rather than extend the field.
328 TK/01 Bugzilla #401. Fix DK spooling code so that it can overwrite a
329 leftover -K file (the existence of which was triggered by #402).
330 While we were at it, introduced process PID as part of the -K
331 filename. This should rule out race conditions when creating
334 TK/02 Bugzilla #402. Apply patch from Simon Arlott, speeding up DK signing
335 processing considerably. Previous code took too long for large mails,
336 triggering a timeout which in turn triggers #401.
338 TK/03 Introduced HAVE_LINUX_SENDFILE to os.h-Linux. Currently only used
339 in the DK code in transports.c. sendfile() is not really portable,
340 hence the _LINUX specificness.
342 TF/01 In the add_headers option to the mail command in an Exim filter,
343 there was a bug that Exim would claim a syntax error in any
344 header after the first one which had an odd number of characters
347 PH/01 If a server that rejects MAIL FROM:<> was the target of a sender
348 callout verification, Exim cached a "reject" for the entire domain. This
349 is correct for most verifications, but it is not correct for a recipient
350 verification with use_sender or use_postmaster set, because in that case
351 the callout does not use MAIL FROM:<>. Exim now distinguishes the special
352 case of MAIL FROM:<> rejection from other early rejections (e.g.
353 rejection of HELO). When verifying a recipient using a non-null MAIL
354 address, the cache is ignored if it shows MAIL FROM:<> rejection.
355 Whatever the result of the callout, the value of the domain cache is
356 left unchanged (for any other kind of callout, getting as far as trying
357 RCPT means that the domain itself is ok).
359 PH/02 Tidied a number of unused variable and signed/unsigned warnings that
362 PH/03 On Solaris, an unexpectedly close socket (dropped connection) can
363 manifest itself as EPIPE rather than ECONNECT. When tidying away a
364 session, the daemon ignores ECONNECT errors and logs others; it now
365 ignores EPIPE as well.
367 PH/04 Applied Nico Erfurth's refactoring patch to tidy up mime.c
368 (quoted-printable decoding).
370 PH/05 Applied Nico Erfurth's refactoring patch to tidy up spool_mbox.c, and
371 later the small subsequent patch to fix an introduced bug.
373 PH/06 Installed the latest Cygwin Makefile from the Cygwin maintainer.
375 PH/07 There was no check for overflow in expansions such as ${if >{1}{4096M}}.
377 PH/08 An error is now given if message_size_limit is specified negative.
379 PH/09 Applied and tidied up Jakob Hirsch's patch for allowing ACL variables
380 to be given (somewhat) arbitrary names.
382 JJ/01 exipick 20060919.0, allow for arbitrary acl_ variables introduced
385 JJ/02 exipick 20060919.0, --show-vars args can now be regular expressions,
386 miscellaneous code fixes
388 PH/10 Added the log_reject_target ACL modifier to specify where to log
391 PH/11 Callouts were setting the name used for EHLO/HELO from $smtp_active_
392 hostname. This is wrong, because it relates to the incoming message (and
393 probably the interface on which it is arriving) and not to the outgoing
394 callout (which could be using a different interface). This has been
395 changed to use the value of the helo_data option from the smtp transport
396 instead - this is what is used when a message is actually being sent. If
397 there is no remote transport (possible with a router that sets up host
398 addresses), $smtp_active_hostname is used.
400 PH/12 Installed Andrey Panin's patch to add a dovecot authenticator. Various
401 tweaks were necessary in order to get it to work (see also 21 below):
402 (a) The code assumed that strncpy() returns a negative number on buffer
403 overflow, which isn't the case. Replaced with Exim's string_format()
405 (b) There were several signed/unsigned issues. I just did the minimum
406 hacking in of casts. There is scope for a larger refactoring.
407 (c) The code used strcasecmp() which is not a standard C function.
408 Replaced with Exim's strcmpic() function.
409 (d) The code set only $1; it now sets $auth1 as well.
410 (e) A simple test gave the error "authentication client didn't specify
411 service in request". It would seem that Dovecot has changed its
412 interface. Fortunately there's a specification; I followed it and
413 changed what the client sends and it appears to be working now.
415 PH/13 Added $message_headers_raw to provide the headers without RFC 2047
418 PH/14 Corrected misleading output from -bv when -v was also used. Suppose the
419 address A is aliased to B and C, where B exists and C does not. Without
420 -v the output is "A verified" because verification stops after a
421 successful redirection if more than one address is generated. However,
422 with -v the child addresses are also verified. Exim was outputting "A
423 failed to verify" and then showing the successful verification for C,
424 with its parentage. It now outputs "B failed to verify", showing B's
425 parentage before showing the successful verification of C.
427 PH/15 Applied Michael Deutschmann's patch to allow DNS black list processing to
428 look up a TXT record in a specific list after matching in a combined
431 PH/16 It seems that the options setting for the resolver (RES_DEFNAMES and
432 RES_DNSRCH) can affect the behaviour of gethostbyname() and friends when
433 they consult the DNS. I had assumed they would set it the way they
434 wanted; and indeed my experiments on Linux seem to show that in some
435 cases they do (I could influence IPv6 lookups but not IPv4 lookups).
436 To be on the safe side, however, I have now made the interface to
437 host_find_byname() similar to host_find_bydns(), with an argument
438 containing the DNS resolver options. The host_find_byname() function now
439 sets these options at its start, just as host_find_bydns() does. The smtp
440 transport options dns_qualify_single and dns_search_parents are passed to
441 host_find_byname() when gethostbyname=TRUE in this transport. Other uses
442 of host_find_byname() use the default settings of RES_DEFNAMES
443 (qualify_single) but not RES_DNSRCH (search_parents).
445 PH/17 Applied (a modified version of) Nico Erfurth's patch to make
446 spool_read_header() do less string testing, by means of a preliminary
447 switch on the second character of optional "-foo" lines. (This is
448 overdue, caused by the large number of possibilities that now exist.
449 Originally there were few.) While I was there, I also converted the
450 str(n)cmp tests so they don't re-test the leading "-" and the first
451 character, in the hope this might squeeze out yet more improvement.
453 PH/18 Two problems with "group" syntax in header lines when verifying: (1) The
454 flag allowing group syntax was set by the header_syntax check but not
455 turned off, possible causing trouble later; (2) The flag was not being
456 set at all for the header_verify test, causing "group"-style headers to
457 be rejected. I have now set it in this case, and also caused header_
458 verify to ignore an empty address taken from a group. While doing this, I
459 came across some other cases where the code for allowing group syntax
460 while scanning a header line wasn't quite right (mostly, not resetting
461 the flag correctly in the right place). These bugs could have caused
462 trouble for malformed header lines. I hope it is now all correct.
464 PH/19 The functions {pwcheck,saslauthd}_verify_password() are always called
465 with the "reply" argument non-NULL. The code, however (which originally
466 came from elsewhere) had *some* tests for NULL when it wrote to *reply,
467 but it didn't always do it. This confused somebody who was copying the
468 code for some other use. I have removed all the tests.
470 PH/20 It was discovered that the GnuTLS code had support for RSA_EXPORT, a
471 feature that was used to support insecure browsers during the U.S. crypto
472 embargo. It requires special client support, and Exim is probably the
473 only MTA that supported it -- and would never use it because real RSA is
474 always available. This code has been removed, because it had the bad
475 effect of slowing Exim down by computing (never used) parameters for the
476 RSA_EXPORT functionality.
478 PH/21 On the advice of Timo Sirainen, added a check to the dovecot
479 authenticator to fail if there's a tab character in the incoming data
480 (there should never be unless someone is messing about, as it's supposed
481 to be base64-encoded). Also added, on Timo's advice, the "secured" option
482 if the connection is using TLS or if the remote IP is the same as the
483 local IP, and the "valid-client-cert option" if a client certificate has
486 PH/22 As suggested by Dennis Davis, added a server_condition option to *all*
487 authenticators. This can be used for authorization after authentication
488 succeeds. (In the case of plaintext, it servers for both authentication
491 PH/23 Testing for tls_required and lost_connection in a retry rule didn't work
492 if any retry times were supplied.
494 PH/24 Exim crashed if verify=helo was activated during an incoming -bs
495 connection, where there is no client IP address to check. In this
496 situation, the verify now always succeeds.
498 PH/25 Applied John Jetmore's -Mset patch.
500 PH/26 Added -bem to be like -Mset, but loading a message from a file.
502 PH/27 In a string expansion for a processed (not raw) header when multiple
503 headers of the same name were present, leading whitespace was being
504 removed from all of them, but trailing whitespace was being removed only
505 from the last one. Now trailing whitespace is removed from each header
506 before concatenation. Completely empty headers in a concatenation (as
509 PH/28 Fixed bug in backwards-compatibility feature of PH/09 (thanks to John
510 Jetmore). It would have mis-read ACL variables from pre-4.61 spool files.
512 PH/29 [Removed. This was a change that I later backed out, and forgot to
513 correct the ChangeLog entry (that I had efficiently created) before
514 committing the later change.]
516 PH/30 Exim was sometimes attempting to deliver messages that had suffered
517 address errors (4xx response to RCPT) over the same connection as other
518 messages routed to the same hosts. Such deliveries are always "forced",
519 so retry times are not inspected. This resulted in far too many retries
520 for the affected addresses. The effect occurred only when there were more
521 hosts than the hosts_max_try setting in the smtp transport when it had
522 the 4xx errors. Those hosts that it had tried were not added to the list
523 of hosts for which the message was waiting, so if all were tried, there
524 was no problem. Two fixes have been applied:
526 (i) If there are any address or message errors in an SMTP delivery, none
527 of the hosts (tried or untried) are now added to the list of hosts
528 for which the message is waiting, so the message should not be a
529 candidate for sending over the same connection that was used for a
530 successful delivery of some other message. This seems entirely
531 reasonable: after all the message is NOT "waiting for some host".
532 This is so "obvious" that I'm not sure why it wasn't done
533 previously. Hope I haven't missed anything, but it can't do any
534 harm, as the worst effect is to miss an optimization.
536 (ii) If, despite (i), such a delivery is accidentally attempted, the
537 routing retry time is respected, so at least it doesn't keep
538 hammering the server.
540 PH/31 Installed Andrew Findlay's patch to close the writing end of the socket
541 in ${readsocket because some servers need this prod.
543 PH/32 Added some extra debug output when updating a wait-xxx database.
545 PH/33 The hint "could be header name not terminated by colon", which has been
546 given for certain expansion errors for a long time, was not being given
547 for the ${if def:h_colon_omitted{... case.
549 PH/34 The spec says: "With one important exception, whenever a domain list is
550 being scanned, $domain contains the subject domain." There was at least
551 one case where this was not true.
553 PH/35 The error "getsockname() failed: connection reset by peer" was being
554 written to the panic log as well as the main log, but it isn't really
555 panic-worthy as it just means the connection died rather early on. I have
556 removed the panic log writing for the ECONNRESET error when getsockname()
559 PH/36 After a 4xx response to a RCPT error, that address was delayed (in queue
560 runs only) independently of the message's sender address. This meant
561 that, if the 4xx error was in fact related to the sender, a different
562 message to the same recipient with a different sender could confuse
563 things. In particualar, this can happen when sending to a greylisting
564 server, but other circumstances could also provoke similar problems.
565 I have changed the default so that the retry time for these errors is now
566 based a combination of the sender and recipient addresses. This change
567 can be overridden by setting address_retry_include_sender=false in the
570 PH/37 For LMTP over TCP/IP (the smtp transport), error responses from the
571 remote server are returned as part of bounce messages. This was not
572 happening for LMTP over a pipe (the lmtp transport), but now it is the
573 same for both kinds of LMTP.
575 PH/38 Despite being documented as not happening, Exim was rewriting addresses
576 in header lines that were in fact CNAMEs. This is no longer the case.
578 PH/39 If -R or -S was given with -q<time>, the effect of -R or -S was ignored,
579 and queue runs started by the daemon processed all messages. This has
580 been fixed so that -R and -S can now usefully be given with -q<time>.
582 PH/40 Import PCRE release 6.7 (fixes some bugs).
584 PH/41 Add bitwise logical operations to eval (courtesy Brad Jorsch).
586 PH/42 Give an error if -q is specified more than once.
588 PH/43 Renamed the variables $interface_address and $interface_port as
589 $received_ip_address and $received_port, to make it clear that these
590 values apply to message reception, and not to the outgoing interface when
591 a message is delivered. (The old names remain recognized, of course.)
593 PH/44 There was no timeout on the connect() call when using a Unix domain
594 socket in the ${readsocket expansion. There now is.
596 PH/45 Applied a modified version of Brad Jorsch's patch to allow "message" to
597 be meaningful with "accept".
599 SC/01 Eximstats V1.43
600 Bug fix for V1.42 with -h0 specified. Spotted by Chris Lear.
602 SC/02 Eximstats V1.44
603 Use a glob alias rather than an array ref in the generated
604 parser. This improves both readability and performance.
606 SC/03 Eximstats V1.45 (Marco Gaiarin / Steve Campbell)
607 Collect SpamAssassin and rejection statistics.
608 Don't display local sender or destination tables unless
609 there is data to show.
610 Added average volumes into the top table text output.
612 SC/04 Eximstats V1.46
613 Collect data on the number of addresses (recipients)
614 as well as the number of messages.
616 SC/05 Eximstats V1.47
617 Added 'Message too big' to the list of mail rejection
618 reasons (thanks to Marco Gaiarin).
620 SC/06 Eximstats V1.48
621 Mainlog lines which have GMT offsets and are too short to
622 have a flag are now skipped.
624 SC/07 Eximstats V1.49 (Alain Williams)
625 Added the -emptyok flag.
627 SC/08 Eximstats V1.50
628 Fixes for obtaining the IP address from reject messages.
630 JJ/03 exipick.20061117.2, made header handling as similar to exim as possible
631 (added [br]h_ prefixes, implemented RFC2047 decoding. Fixed
632 whitesspace changes from 4.64-PH/27
634 JJ/04 exipick.20061117.2, fixed format and added $message_headers_raw to
637 JJ/05 exipick.20061117.2, bug fixes (error out sooner when invalid criteria
638 are found, allow negative numbers in numeric criteria)
640 JJ/06 exipick.20061117.2, added new $message_body_missing variable
642 JJ/07 exipick.20061117.2, added $received_ip_address and $received_port
643 to match changes made in 4.64-PH/43
645 PH/46 Applied Jori Hamalainen's patch to add features to exiqsumm.
647 PH/47 Put in an explicit test for a DNS lookup of an address record where the
648 "domain" is actually an IP address, and force a failure. This locks out
649 those revolvers/nameservers that support "A-for-A" lookups, in
650 contravention of the specifications.
652 PH/48 When a host name was looked up from an IP address, and the subsequent
653 forward lookup of the name timed out, the host name was left in
654 $sender_host_name, contrary to the specification.
656 PH/49 Although default lookup types such as lsearch* or cdb*@ have always been
657 restricted to single-key lookups, Exim was not diagnosing an error if
658 * or *@ was used with a query-style lookup.
660 PH/50 Increased the value of DH_BITS in tls-gnu.c from 768 to 1024.
662 MH/01 local_scan ABI version incremented to 1.1. It should have been updated
663 long ago, but noone interested enough thought of it. Let's just say that
664 the "1.1" means that there are some new functions that weren't there at
665 some point in the past.
667 PH/51 Error processing for expansion failure of helo_data from an smtp
668 transport during callout processing was broken.
670 PH/52 Applied John Jetmore's patch to allow tls-on-connect and STARTTLS to be
671 tested/used via the -bh/-bhc/-bs options.
673 PH/53 Added missing "#include <time.h>" to pcre/pcretest.c (this was a PCRE
674 bug, fixed in subsequent PCRE releases).
676 PH/54 Applied Robert Bannocks' patch to avoid a problem with references that
677 arises when using the Solaris LDAP libraries (but not with OpenLDAP).
679 PH/55 Check for a ridiculously long file name in exim_dbmbuild.
685 SC/01 Use a glob alias rather than an array ref in eximstats generated
686 parser. This improves both readability and performance.
688 SC/02 Collect SpamAssassin and rejection statistics in eximstats.
689 Don't display local sender or destination tables in eximstats unless
690 there is data to show.
691 Added average volumes into the eximstats top table text output.
693 SC/03 Collect data on the number of addresses (recipients) as well
694 as the number of messages in eximstats.
696 TF/01 Correct an error in the documentation for the redirect router. Exim
697 does (usually) call initgroups() when daemonizing.
699 TF/02 Call initgroups() when dropping privilege in exim.c, so that Exim runs
700 with consistent privilege compared to when running as a daemon.
702 TF/03 Note in the spec that $authenticated_id is not set for local
703 submissions from trusted users.
705 TF/04 The ratelimit per_rcpt option now works correctly in acl_not_smtp.
706 Thanks to Dean Brooks <dean@iglou.com> for the patch.
708 TF/05 Make it easier to get SMTP authentication and TLS/SSL support working
709 by adding some example configuration directives to the default
710 configuration file. A little bit of work is required to uncomment the
711 directives and define how usernames and passwords are checked, but
712 there is now a framework to start from.
714 PH/01 Added #define LDAP_DEPRECATED 1 to ldap.c because some of the "old"
715 functions that Exim currently uses aren't defined in ldap.h for OpenLDAP
716 without this. I don't know how relevant this is to other LDAP libraries.
718 PH/02 Add the verb name to the "unknown ACL verb" error.
720 PH/03 Magnus Holmgren's patch for filter_prepend_home.
722 PH/03 Fixed Bugzilla #101: macro definition between ACLs doesn't work.
724 PH/04 Applied Magnus Holmgren's patch to fix Bugzilla #98: transport's home
725 directory not expanded when it should be if an expanded home directory
726 was set for the address (which is overridden by the transport).
728 PH/05 Applied Alex Kiernan's patch to fix Bugzilla #99: a problem with
731 PH/06 Added acl_not_smtp_start, based on Johannes Berg's patch, and set the
732 bit to forbid control=suppress_local_fixups in the acl_not_smtp ACL,
733 because it is too late at that time, and has no effect.
735 PH/07 Changed ${quote_pgsql to quote ' as '' instead of \' because of a
736 security issue with \' (bugzilla #107). I could not use the
737 PQescapeStringConn() function, because it needs a PGconn value as one of
740 PH/08 When testing addresses using -bt, indicate those final addresses that
741 are duplicates that would not cause an additional delivery. At least one
742 person was confused, thinking that -bt output corresponded to deliveries.
743 (Suppressing duplicates isn't a good idea as you lose the information
744 about possibly different redirections that led to the duplicates.)
746 PH/09 Applied patch from Erik to use select() instead of poll() in spam.c on
747 systems where poll() doesn't work, in particular OS X.
749 PH/10 Added more information to debugging output for retry time not reached.
751 PH/11 Applied patch from Arkadiusz Miskiewicz to apply a timeout to read
752 operations in malware.c.
754 PH/12 Applied patch from Magnus Holmgren to include the "h" tag in Domain Keys
757 PH/13 If write_rejectlog was set false when logging was sent to syslog with
758 syslog_duplication set false, log lines that would normally be written
759 both the the main log and to the reject log were not written to syslog at
762 PH/14 In the default configuration, change the use of "message" in ACL warn
763 statements to "add_header".
765 PH/15 Diagnose a filter syntax error for "seen", "unseen", or "noerror" if not
766 not followed by a command (e.g. "seen endif").
768 PH/16 Recognize SMTP codes at the start of "message" in ACLs and after :fail:
769 and :defer: in a redirect router. Add forbid_smtp_code to suppress the
772 PH/17 Added extra conditions to the default value of delay_warning_condition
776 { !eq{$h_list-id:$h_list-post:$h_list-subscribe:}{} } \
777 { match{$h_precedence:}{(?i)bulk|list|junk} } \
778 { match{$h_auto-submitted:}{(?i)auto-generated|auto-replied} } \
781 The Auto-Submitted: and various List- headers are standardised, whereas I
782 don't think Precedence: ever was.
784 PH/18 Refactored debugging code in route_finduser() to show more information,
785 in particular, the error code if getpwnam() issues one.
787 PH/19 Added PQsetClientEncoding(conn, "SQL_ASCII") to the pgsql code module.
788 This is apparently needed in addition to the PH/07 change above to avoid
789 any possible encoding problems.
791 PH/20 Perl can change the locale. Exim was resetting it after a ${perl call,
792 but not after initializing Perl.
794 PH/21 Added a call to PQsetNoticeProcessor() to catch pgsql "notices" and
795 output them only if debugging. By default they are written stderr,
796 apparently, which is not desirable.
798 PH/22 Added Alain Williams' LDAP patch to support setting REFERRALS=off on
801 JJ/01 exipick: added --reverse (and -R synonym), --random, --size, --sort and
804 JJ/02 exipick: rewrote --help documentation to hopefully make more clear.
806 PH/23 Made -oMaa and -oMt work with -bh and -bs to pretend the connection is
807 authenticated or an ident call has been made. Suppress the default
808 values for $authenticated_id and $authenticated_sender (but permit -oMai
809 and -oMas) when testing with -bh.
811 PH/24 Re-jigged the order of the tests in the default configuration so that the
812 tests for valid domains and recipients precede the DNS black list and CSA
813 tests, on the grounds that those ones are more expensive.
815 PH/25 Exim was not testing for a space following SMTP commands such as EHLO
816 that require one. Thus, EHLORHUBARB was interpreted as a valid command.
817 This bug exists in every version of Exim that I still have, right back to
820 PH/26 (n)wildlsearch lookups are documented as being done case-insensitively.
821 However, an attempt to turn on case-sensitivity in a regex key by
822 including (?-i) didn't work because the subject string was already
823 lowercased, and the effects were non-intuitive. It turns out that a
824 one-line patch can be used to allow (?-i) to work as expected.
830 TF/01 Fix the add_header change below (4.61 PH/55) which had a bug that (amongst
831 other effects) broke the use of negated acl sub-conditions.
833 PH/01 ${readsocket now supports Internet domain sockets (modified John Jetmore
836 PH/02 When tcp-wrappers is called from Exim, it returns only "deny" or "allow".
837 "Deny" causes Exim to reject the incoming connection with a 554 error.
838 Unfortunately, if there is a major crisis, such as a disk failure,
839 tcp-wrappers gives "deny", whereas what one would like would be some
840 kind of temporary error. A kludge has been added to help with this.
841 Before calling hosts_ctl(), errno is set zero. If the result is "deny", a
842 554 error is used if errno is still zero or contains ENOENT (which occurs
843 if either of the /etc/hosts.{allow,deny} files is missing). Otherwise, a
846 PH/03 Add -lutil to the default FreeBSD LIBS setting.
848 PH/04 Change PH/19 for 4.61 was too wide. It should not be applied to host
849 errors. Otherwise a message that provokes a temporary error (when other
850 messages do not) can cause a whole host to time out.
852 PH/05 Batch deliveries by appendfile and pipe transports did not work when the
853 addresses were routed directly to files or pipes from a redirect router.
854 File deliveries just didn't batch; pipe deliveries might have suffered
857 PH/06 A failure to get a lock for a hints database would erroneously always say
858 "Failed to get write lock", even when it was really a read lock.
860 PH/07 The appendfile transport was creating MBX lock files with a fixed mode
861 of 0600. This has been changed to use the value of the lockfile_mode
862 option (which defaults to 0600).
864 PH/08 Applied small patch from the Sieve maintainer.
866 PH/09 If maildir_quota_directory_regex was set to exclude (say) the .Trash
867 folder from quota calculations, a direct delivery into this folder messed
868 up the contents of the maildirsize file. This was because the regex was
869 used only to exclude .Trash (or whatever) when the size of the mailbox
870 was calculated. There was no check that a delivery was happening into an
871 excluded directory. This bug has been fixed by ignoring all quota
872 processing for deliveries into excluded directories.
874 PH/10 Added the maildirfolder_create_regex option to appendfile.
880 PH/01 The code for finding all the local interface addresses on a FreeBSD
881 system running IPv6 was broken. This may well have applied to all BSD
882 systems, as well as to others that have similar system calls. The broken
883 code found IPv4 interfaces correctly, but gave incorrect values for the
884 IPv6 interfaces. In particular, ::1 was not found. The effect in Exim was
885 that it would not match correctly against @[] and not recognize the IPv6
888 PH/02 The ipliteral router was not recognizing addresses of the form user@
889 [ipv6:....] because it didn't know about the "ipv6:" prefix.
891 PH/03 Added disable_ipv6.
893 PH/04 Changed $reply_address to use the raw form of the headers instead of the
894 decoded form, because it is most often used to construct To: headers
895 lines in autoreplies, and the decoded form may well be syntactically
896 invalid. However, $reply_address has leading white space removed, and all
897 newlines turned into spaces so that the autoreply transport does not
900 PH/05 If group was specified without a user on a router, and no group or user
901 was specified on a transport, the group from the router was ignored.
903 PH/06 Increased the number of ACL variables to 20 of each type, and arranged
904 for visible compile-time settings that can be used to change these
905 numbers, for those that want even more. Backwards compatibility with old
906 spool files has been maintained. However, going back to a previous Exim
907 release will lost any variables that are in spool files.
909 PH/07 Two small changes when running in the test harness: increase delay when
910 passing a TCP/IP connection to a new process, in case the original
911 process has to generate a bounce, and remove special handling of
912 127.0.0.2 (sic), which is no longer necessary.
914 PH/08 Changed debug output of dbfn_open() flags from numbers to names, so as to
915 be the same on different OS.
917 PH/09 Moved a debug statement in filter processing to avoid a race problem when
920 JJ/01 exipick: fixed bug where -b (brief) output option showed "Vars:"
921 whether --show-vars was specified or not
923 JJ/02 exipick: Added support for new ACL variable spool format introduced
926 PH/10 Fixed another bug related to PH/04 above: if an incoming message had a
927 syntactically invalid From: or Reply-to: line, and a filter used this to
928 generate an autoreply, and therefore failed to obtain an address for the
929 autoreply, Exim could try to deliver to a non-existent relative file
930 name, causing unrelated and misleading errors. What now happens is that
931 it logs this as a hard delivery error, but does not attempt to create a
934 PH/11 The exinext utility has a -C option for testing purposes, but although
935 the given file was scanned by exinext itself; it wasn't being passed on
936 when Exim was called.
938 PH/12 In the smtp transport, treat an explicit ECONNRESET error the same as
939 an end-of-file indication when reading a command response.
941 PH/13 Domain literals for IPv6 were not recognized unless IPv6 support was
942 compiled. In many other places in Exim, IPv6 addresses are always
943 recognized, so I have changed this. It also means that IPv4 domain
944 literals of the form [IPV4:n.n.n.n] are now always recognized.
946 PH/14 When a uid/gid is specified for the queryprogram router, it cannot be
947 used if the router is not running as root, for example, when verifying at
948 ACL time, or when using -bh. The debugging output from this situation was
949 non-existent - all you got was a failure to exec. I have made two
952 (a) Failures to set uid/gid, the current directory, or a process leader
953 in a subprocess such as that created by queryprogram now generate
954 suitable debugging ouput when -d is set.
956 (b) The queryprogram router detects when it is not running as root,
957 outputs suitable debugging information if -d is set, and then runs
958 the subprocess without attempting to change uid/gid.
960 PH/15 Minor change to Makefile for building test_host (undocumented testing
963 PH/16 As discussed on the list in Nov/Dec: Exim no longer looks at the
964 additional section of a DNS packet that returns MX or SRV records.
965 Instead, it always explicitly searches for A/AAAA records. This avoids
966 major problems that occur when a DNS server includes only records of one
967 type (A or AAAA) in an MX/SRV packet. A byproduct of this change has
968 fixed another bug: if SRV records were looked up and the corresponding
969 address records were *not* found in the additional section, the port
970 values from the SRV records were lost.
972 PH/17 If a delivery to a pipe, file, or autoreply was deferred, Exim was not
973 using the correct key (the original address) when searching the retry
974 rules in order to find which one to use for generating the retry hint.
976 PH/18 If quota_warn_message contains a From: header, Exim now refrains from
977 adding the default one. Similarly, if it contains a Reply-To: header, the
978 errors_reply_to option, if set, is not used.
980 PH/19 When calculating a retry time, Exim used to measure the "time since
981 failure" by looking at the "first failed" field in the retry record. Now
982 it does not use this if it is later than than the arrival time of the
983 message. Instead it uses the arrival time. This makes for better
984 behaviour in cases where some deliveries succeed, thus re-setting the
985 "first failed" field. An example is a quota failure for a huge message
986 when small messages continue to be delivered. Without this change, the
987 "time since failure" will always be short, possible causing more frequent
988 delivery attempts for the huge message than are intended.
989 [Note: This change was subsequently modified - see PH/04 for 4.62.]
991 PH/20 Added $auth1, $auth2, $auth3 to contain authentication data (as well as
992 $1, $2, $3) because the numerical variables can be reset during some
993 expansion items (e.g. "match"), thereby losing the authentication data.
995 PH/21 Make -bV show the size of off_t variables so that the test suite can
996 decide whether to run tests for quotas > 2G.
998 PH/22 Test the values given for quota, quota_filecount, quota_warn_threshold,
999 mailbox_size, and mailbox_filecount in the appendfile transport. If a
1000 filecount value is greater than 2G or if a quota value is greater than 2G
1001 on a system where the size of off_t is not greater than 4, a panic error
1004 PH/23 When a malformed item such as 1.2.3/24 appears in a host list, it can
1005 never match. The debug and -bh output now contains an explicit error
1006 message indicating a malformed IPv4 address or mask.
1008 PH/24 An host item such as 1.2.3.4/abc was being treated as the IP address
1009 1.2.3.4 without a mask. Now it is not recognized as an IP address, and
1010 PH/23 above applies.
1012 PH/25 Do not write to syslog when running in the test harness. The only
1013 occasion when this arises is a failure to open the main or panic logs
1014 (for which there is an explicit test).
1016 PH/26 Added the /no_tell option to "control=freeze".
1018 PH/27 If a host name lookup failed very early in a connection, for example, if
1019 the IP address matched host_lookup and the reverse lookup yielded a name
1020 that did not have a forward lookup, an error message of the form "no IP
1021 address found for host xxx.xxx.xxx (during SMTP connection from NULL)"
1022 could be logged. Now it outputs the IP address instead of "NULL".
1024 PH/28 An enabling patch from MH: add new function child_open_exim2() which
1025 allows the sender and the authenticated sender to be set when
1026 submitting a message from within Exim. Since child_open_exim() is
1027 documented for local_scan(), the new function should be too.
1029 PH/29 In GnuTLS, a forced expansion failure for tls_privatekey was not being
1030 ignored. In both GnuTLS and OpenSSL, an expansion of tls_privatekey that
1031 results in an empty string is now treated as unset.
1033 PH/30 Fix eximon buffer overflow bug (Bugzilla #73).
1035 PH/31 Added sender_verify_fail logging option.
1037 PH/32 In November 2003, the code in Exim that added an empty Bcc: header when
1038 needed by RFC 822 but not by RFC 2822 was commented out. I have now
1039 tidied the source and removed it altogether.
1041 PH/33 When a queue run was abandoned because the load average was too high, a
1042 log line was always written; now it is written only if the queue_run log
1043 selector is set. In addition, the log line for abandonment now contains
1044 information about the queue run such as the pid. This is always present
1045 in "start" and "stop" lines but was omitted from the "abandon" line.
1047 PH/34 Omit spaces between a header name and the colon in the error message that
1048 is given when verify = headers_syntax fails (if there are lots of them,
1049 the message gets confusing).
1051 PH/35 Change the default for dns_check_names_pattern to allow slashes within
1052 names, as there are now some PTR records that contain slashes. This check
1053 is only to protect against broken name servers that fall over on strange
1054 characters, so the fact that it applies to all lookups doesn't matter.
1056 PH/36 Now that the new test suite is complete, we can remove some of the
1057 special code in Exim that was needed for the old test suite. For example,
1058 sorting DNS records because real resolvers return them in an arbitrary
1059 order. The new test suite's fake resolver always returns records in the
1062 PH/37 When running in the test harness, use -odi for submitted messages (e.g.
1063 bounces) except when queue_only is set, to avoid logging races between
1064 the different processes.
1066 PH/38 Panic-die if .include specifies a non-absolute path.
1068 PH/39 A tweak to the "H" retry rule from its user.
1070 JJ/03 exipick: Removed parentheses from 'next' and 'last' calls that specified
1071 a label. They prevented compilation on older perls.
1073 JJ/04 exipick: Refactored code to prevent implicit split to @_ which caused
1074 a warning to be raised on newish perls.
1076 JJ/05 exipick: Fixed bug where -bpc always showed a count of all messages
1077 on queue. Changes to match documented behaviour of showing count of
1078 messages matching specified criteria.
1080 PH/40 Changed the default ident timeout from 30s to 5s.
1082 PH/41 Added support for the use of login_cap features, on those BSD systems
1083 that have them, for controlling the resources used by pipe deliveries.
1085 PH/42 The content-scanning code uses fopen() to create files in which to put
1086 message data. Previously it was not paying any attention to the mode of
1087 the files. Exim runs with umask(0) because the rest of the code creates
1088 files with open(), and sets the required mode explicitly. Thus, these
1089 files were ending up world-writeable. This was not a big issue, because,
1090 being within the spool directory, they were not world-accessible. I have
1091 created a function called modefopen, which takes an additional mode
1092 argument. It sets umask(777), creates the file, chmods it to the required
1093 mode, then resets the umask. All the relevant calls to fopen() in the
1094 content scanning code have been changed to use this function.
1096 PH/43 If retry_interval_max is set greater than 24 hours, it is quietly reset
1097 to 24 hours. This avoids potential overflow problems when processing G
1098 and H retry rules. I suspect nobody ever tinkers with this value.
1100 PH/44 Added STRIP_COMMAND=/usr/bin/strip to the FreeBSD Makefile.
1102 PH/45 When the plaintext authenticator is running as a client, the server's
1103 challenges are checked to ensure they are valid base64 strings. By
1104 default, the authentication attempt is cancelled if an invalid string is
1105 received. Setting client_ignore_invalid_base64 true ignores these errors.
1106 The decoded challenge strings are now placed in $auth1, $auth2, etc. as
1107 they are received. Thus, the responses can be made to depend on the
1108 challenges. If an invalid string is ignored, an empty string is placed in
1111 PH/46 Messages that are created by the autoreply transport now contains a
1112 References: header, in accordance with RFCs 2822 and 3834.
1114 PH/47 Added authenticated_sender_force to the smtp transport.
1116 PH/48 The ${prvs expansion was broken on systems where time_t was long long.
1118 PH/49 Installed latest patch from the Sieve maintainer.
1120 PH/50 When an Exim quota was set without a file count quota, and mailbox_size
1121 was also set, the appendfile transport was unnecessarily scanning a
1122 directory of message files (e.g. for maildir delivery) to find the count
1123 of files (along with the size), even though it did not need this
1124 information. It now does the scan only if it needs to find either the
1125 size of the count of files.
1127 PH/51 Added ${time_eval: to convert Exim time strings into seconds.
1129 PH/52 Two bugs concerned with error handling when the smtp transport is
1132 (i) Exim was not creating retry information for temporary errors given
1133 for individual recipients after the DATA command when the smtp transport
1134 was used in LMTP mode. This meant that they could be retried too
1135 frequently, and not timed out correctly.
1137 (ii) Exim was setting the flag that allows error details to be returned
1138 for LMTP errors on RCPT commands, but not for LMTP errors for individual
1139 recipients that were returned after the DATA command.
1141 PH/53 This is related to PH/52, but is more general: for any failing address,
1142 when detailed error information was permitted to be returned to the
1143 sender, but the error was temporary, then after the final timeout, only
1144 "retry timeout exceeded" was returned. Now it returns the full error as
1145 well as "retry timeout exceeded".
1147 PH/54 Added control=allow_auth_unadvertised, as it seems there are clients that
1148 do this, and (what is worse) MTAs that accept it.
1150 PH/55 Added the add_header modified to ACLs. The use of "message" with "warn"
1151 will now be deprecated.
1153 PH/56 New os.c-cygwin from the Cygwin maintainer.
1155 JJ/06 exipick: added --unsorted option to allow unsorted output in all output
1156 formats (previously only available in exim formats via -bpr, -bpru,
1157 and -bpra. Now also available in native and exiqgrep formats)
1159 JJ/07 exipick: added --freeze and --thaw options to allow faster interaction
1160 with very large, slow to parse queues
1162 JJ/08 exipick: added ! as generic prefix to negate any criteria format
1164 JJ/09 exipick: miscellaneous performance enhancements (~24% improvements)
1166 PH/57 Tidies in SMTP dialogue display in debug output: (i) It was not showing
1167 responses to authentication challenges, though it was showing the
1168 challenges; (ii) I've removed the CR characters from the debug output for
1171 PH/58 Allow for the insertion of a newline as well as a space when a string
1172 is turned into more than one encoded-word during RFC 2047 encoding. The
1173 Sieve code now uses this.
1175 PH/59 Added the following errors that can be detected in retry rules: mail_4xx,
1176 data_4xx, lost_connection, tls_required.
1178 PH/60 When a VRFY deferred or FAILED, the log message rather than the user
1179 message was being sent as an SMTP response.
1181 PH/61 Add -l and -k options to exicyclog.
1183 PH/62 When verifying, if an address was redirected to one new address, so that
1184 verification continued, and the new address failed or deferred after
1185 having set something in $address_data, the value of $address_data was not
1186 passed back to the ACL. This was different to the case when no
1187 redirection occurred. The value is now passed back in both cases.
1189 PH/63 Changed the macro HAVE_LOGIN_CAP (see PH/41 for this release above) to
1190 HAVE_SETCLASSRESOURCES because there are different APIs in use that all
1191 use login_cap.h, so on its own it isn't the distinguishing feature. The
1192 new name refers directly to the setclassresources() function.
1194 PH/65 Added configuration files for NetBSD3.
1196 PH/66 Updated OS/Makefile-HP-UX for gcc 4.1.0 with HP-UX 11.
1198 PH/67 Fixed minor infelicity in the sorting of addresses to ensure that IPv6
1199 is preferred over IPv4.
1201 PH/68 The bounce_return_message and bounce_return_body options were not being
1202 honoured for bounces generated during the reception of non-SMTP messages.
1203 In particular, this applied to messages rejected by the ACL. This bug has
1204 been fixed. However, if bounce_return_message is true and bounce_return_
1205 body is false, the headers that are returned for a non-SMTP message
1206 include only those that have been read before the error was detected.
1207 (In the case of an ACL rejection, they have all been read.)
1209 PH/69 The HTML version of the specification is now built in a directory called
1210 spec_html instead of spec.html, because the latter looks like a path with
1211 a MIME-type, and this confuses some software.
1213 PH/70 Catch two compiler warnings in sieve.c.
1215 PH/71 Fixed an obscure and subtle bug (thanks Alexander & Matthias). The
1216 function verify_get_ident() calls ip_connect() to connect a socket, but
1217 if the "connect()" function timed out, ip_connect() used to close the
1218 socket. However, verify_get_ident() also closes the socket later, and in
1219 between Exim writes to the log, which may get opened at this point. When
1220 the socket was closed in ip_connect(), the log could get the same file
1221 descriptor number as the socket. This naturally causes chaos. The fix is
1222 not to close the socket in ip_connect(); the socket should be closed by
1223 the function that creates it. There was only one place in the code where
1224 this was missing, in the iplookup router, which I don't think anybody now
1225 uses, but I've fixed it anyway.
1227 PH/72 Make dns_again_means_nonexist apply to lookups using gethostbyname() as
1228 well as to direct DNS lookups. Otherwise the handling of names in host
1229 lists is inconsistent and therefore confusing.
1235 PH/01 Two changes to the default runtime configuration:
1237 (1) Move the checks for relay_from_hosts and authenticated clients from
1238 after to before the (commented out) DNS black list checks.
1240 (2) Add control=submission to the relay_from_hosts and authenticated
1241 clients checks, on the grounds that messages accepted by these
1242 statements are most likely to be submissions.
1244 PH/02 Several tidies to the handling of ${prvs and ${prvscheck:
1246 (1) Generate an error if the third argument for the ${prvs expansion is
1249 (2) Treat a missing third argument of ${prvscheck as if it were an empty
1252 (3) Reset the variables that are obtained from the first argument of
1253 ${prvscheck and used in the second argument before leaving the code,
1254 because their memory is reclaimed, so using them afterwards may do
1257 (4) Tidy up the code for expanding the arguments of ${prvscheck one by
1258 one (it's much easier than Tom thought :-).
1260 (5) Because of (4), we can now allow for the use of $prvscheck_result
1261 inside the third argument.
1263 PH/03 For some reason, the default setting of PATH when running a command from
1264 a pipe transport was just "/usr/bin". I have changed it to
1267 PH/04 SUPPORT_TRANSLATE_IP_ADDRESS and MOVE_FROZEN_MESSAGES did not cause
1268 anything to be listed in the output from -bV.
1270 PH/05 When a filter generated an autoreply, the entire To: header line was
1271 quoted in the delivery log line, like this:
1273 => >A.N.Other <ano@some.domain> <original@ddress> ...
1275 This has been changed so that it extracts the operative address. There
1276 may be more than one such address. If so, they are comma-separated, like
1279 => >ano@some.domain,ona@other.domain <original@ddress> ...
1281 PH/06 When a client host used a correct literal IP address in a HELO or EHLO
1282 command, (for example, EHLO [1.2.3.4]) and the client's IP address was
1283 not being looked up in the rDNS to get a host name, Exim was showing the
1284 IP address twice in Received: lines, even though the IP addresses were
1285 identical. For example:
1287 Received: from [1.2.3.4] (helo=[1.2.3.4])
1289 However, if the real host name was known, it was omitting the HELO data
1290 if it matched the actual IP address. This has been tidied up so that it
1291 doesn't show the same IP address twice.
1293 PH/07 When both +timestamp and +memory debugging was on, the value given by
1294 $tod_xxx expansions could be wrong, because the tod_stamp() function was
1295 called by the debug printing, thereby overwriting the timestamp buffer.
1296 Debugging no longer uses the tod_stamp() function when +timestamp is set.
1298 PH/08 When the original message was included in an autoreply transport, it
1299 always said "this is a copy of the message, including all the headers",
1300 even if body_only or headers_only was set. It now gives an appropriate
1303 PH/09 Applied a patch from the Sieve maintainer which:
1305 o fixes some comments
1306 o adds the (disabled) notify extension core
1307 o adds some debug output for the result of if/elsif tests
1308 o points to the current vacation draft in the documentation
1309 and documents the missing references header update
1313 o fixes a bug in processing the envelope test (when testing
1314 multiple envelope elements, the last element determinted the
1317 PH/10 Exim was violating RFC 3834 ("Recommendations for Automatic Responses to
1318 Electronic Mail") by including:
1320 Auto-submitted: auto-generated
1322 in the messages that it generates (bounce messages and others, such as
1323 warnings). In the case of bounce messages for non-SMTP mesages, there was
1324 also a typo: it was using "Auto_submitted" (underscore instead of
1325 hyphen). Since every message generated by Exim is necessarily in response
1326 to another message, thes have all been changed to:
1328 Auto-Submitted: auto-replied
1330 in accordance with these statements in the RFC:
1332 The auto-replied keyword:
1334 - SHOULD be used on messages sent in direct response to another
1335 message by an automatic process,
1337 - MUST NOT be used on manually-generated messages,
1339 - MAY be used on Delivery Status Notifications (DSNs) and Message
1340 Disposition Notifications (MDNs),
1342 - MUST NOT be used on messages generated by automatic or periodic
1343 processes, except for messages which are automatic responses to
1346 PH/11 Added "${if def:sender_address {(envelope-from <$sender_address>)\n\t}}"
1347 to the default Received: header definition.
1349 PH/12 Added log selector acl_warn_skipped (default on).
1351 PH/13 After a successful wildlsearch lookup, discard the values of numeric
1352 variables because (a) they are in the wrong storage pool and (b) even if
1353 they were copied, it wouldn't work properly because of the caching.
1355 PH/14 Add check_rfc2047_length to disable enforcement of RFC 2047 length
1356 checking when decoding. Apparently there are clients that generate
1357 overlong encoded strings. Why am I not surprised?
1359 PH/15 If the first argument of "${if match_address" was not empty, but did not
1360 contain an "@" character, Exim crashed. Now it writes a panic log message
1361 and treats the condition as false.
1363 PH/16 In autoreply, treat an empty string for "once" the same as unset.
1365 PH/17 A further patch from the Sieve maintainer: "Introduce the new Sieve
1366 extension "envelope-auth". The code is finished and in agreement with
1367 other implementations, but there is no documentation so far and in fact,
1368 nobody wrote the draft yet. This extension is currently #undef'ed, thus
1369 not changing the active code.
1371 Print executed "if" and "elsif" statements when debugging is used. This
1372 helps a great deal to understand what a filter does.
1374 Document more things not specified clearly in RFC3028. I had all this
1375 sorted out, when out of a sudden new issues came to my mind. Oops."
1377 PH/18 Exim was not recognizing the "net-" search type prefix in match_ip lists
1380 PH/19 Exim expands the IPv6 address given to -bh to its full non-abbreviated
1381 canonical form (as documented). However, after a host name lookup from
1382 the IP address, check_host() was doing a simple string comparison with
1383 addresses acquired from the DNS when checking that the found name did
1384 have the original IP as one of its addresses. Since any found IPv6
1385 addresses are likely to be in abbreviated form, the comparison could
1386 fail. Luckily, there already exists a function for doing the comparison
1387 by converting both addresses to binary, so now that is used instead of
1388 the text comparison.
1390 PH/20 There was another similar case to PH/19, when a complete host name was
1391 given in a host list; looking up its IP address could give an abbreviated
1392 form, whereas the current host's name might or might not be abbreviated.
1393 The same fix has been applied.
1399 PH/01 The ${base62: operator adjusted itself to base 36 when BASE_62 was
1400 set to 36 (for Darwin and Cygwin), but the ${base62d: operator did not.
1403 PH/02 Two minor problems detected in Cygwin: the os.{c,h} files had lost */ on
1404 the CVS lines, and there was a missing #if HAVE_IPV6 in host.c.
1406 PH/03 Typo: missing ".o" in src/pcre/Makefile.
1408 PH/04 Tighten up "personal" tests: Instead of testing for any "List-"
1409 header line, restrict the check to what is listed in RFCs 2369 and 2929.
1410 Also, for "Auto-Submitted", treat anything other than "no" as
1411 non-personal, in accordance with RFC 3834. (Previously it treated
1412 anything starting "auto-" as non-personal.)
1414 TF/01 The control=submission/name=... option had a problem with syntax
1415 errors if the name included a slash character. The /name= option
1416 now slurps the rest of the string, so it can include any characters
1417 but it must come last in the list of options (after /sender_retain
1420 PH/05 Some modifications to the interface to the fake nameserver for the new
1428 TK/01 Added the "success_on_redirect" address verification option. See
1429 NewStuff for rationale and an example.
1431 PH/01 Added support for SQLite, basic code supplied by David Woodhouse.
1433 PH/02 Patch to exigrep to allow it to work on syslog lines.
1435 PH/03 When creating an mbox file for a virus/spam scan, use fseek() instead of
1436 fread() to skip over the body file's header line, because in Cygwin the
1437 header line is locked and is inaccessible.
1439 PH/04 Added $message_exim_id, ultimately to replace $message_id (they will both
1440 co-exist for some time) to make it clear that it is the Exim ID that is
1441 referenced, not the Message-ID: header line.
1443 PH/05 Replaced all Tom's calls to snprintf() with calls to the internal
1444 string_format() function, because snprintf() does not exist on all
1447 PH/06 The use of forbid_filter_existstest now also locks out the use of the
1448 ${stat: expansion item.
1450 PH/07 Changed "SMTP protocol violation: synchronization error" into "SMTP
1451 protocol synchronization error", to keep the pedants happy.
1453 PH/08 Arrange for USE_INET_NTOA_FIX to be set in config.h for AIX systems as
1454 well as for IRIX systems, when gcc is being used. See the host.c source
1457 PH/09 Installed latest Cygwin configuration files from the Cygwin maintainer.
1459 PH/10 Named domain lists were not working if used in a queue_smtp_domains
1462 PH/11 Added support for the IGNOREQUOTA extension to LMTP, both to the lmtp
1463 transport and to the smtp transport in LMTP mode.
1465 TK/02 Remove one case of BASE64 error detection FTTB (undocumented anyway).
1467 PH/12 There was a missing call to search_tidyup() before the fork() in rda.c to
1468 run a filter in a subprocess. This could lead to confusion in subsequent
1469 lookups in the parent process. There should also be a search_tidyup() at
1470 the end of the subprocess.
1472 PH/13 Previously, if "verify = helo" was set in an ACL, the condition was true
1473 only if the host matched helo_try_verify_hosts, which caused the
1474 verification to occur when the EHLO/HELO command was issued. The ACL just
1475 tested the remembered result. Now, if a previous verification attempt has
1476 not happened, "verify = helo" does it there and then.
1478 JJ/01 exipick: added $message_exim_id variable (see 4.53-PH/04)
1480 TK/03 Fix log output including CR from clamd.
1482 PH/14 A reference to $reply_address when Reply-to: was empty and From: did not
1483 exist provoked a memory error which could cause a segfault.
1485 PH/15 Installed PCRE 6.2
1487 PH/17 Defined BIND_8_COMPAT in the Darwin os.h file.
1489 PH/18 Reversed 4.52/PH/17 because the HP-UX user found it wasn't the cause
1490 of the problem. Specifically, suggested +O2 rather than +O1 for the
1493 PH/19 Added sqlite_lock_timeout option (David Woodhouse's patch).
1495 PH/20 If a delivery was routed to a non-standard port by means of an SRV
1496 record, the port was not correctly logged when the outgoing_port log
1497 selector was set (it logged the transort's default port).
1499 PH/21 Added support for host-specific ports to manualroute, queryprogram,
1500 fallback_hosts, and "hosts" in the smtp transport.
1502 PH/22 If the log selector "outgoing_port" is set, the port is now also given on
1503 host errors such as "Connection refused".
1505 PH/23 Applied a patch to fix problems with exim-4.52 while doing radius
1506 authentication with radiusclient 0.4.9:
1508 - Error returned from rc_read_config was caught wrongly
1509 - Username/password not passed on to radius server due to wrong length.
1511 The presumption is that some radiusclient API changes for 4.51/PH/17
1512 were not taken care of correctly. The code is still untested by me (my
1513 Linux distribution still has 0.3.2 of radiusclient), but it was
1514 contributed by a Radius user.
1516 PH/24 When doing a callout, the value of $domain wasn't set correctly when
1517 expanding the "port" option of the smtp transport.
1519 TK/04 MIME ACL: Fix buffer underrun that occurs when EOF condition is met
1520 while reading a MIME header. Thanks to Tom Hughes for a patch.
1522 PH/24 Include config.h inside local_scan.h so that configuration settings are
1525 PH/25 Make $smtp_command_argument available after all SMTP commands. This means
1526 that in an ACL for RCPT (for example), you can examine exactly what was
1529 PH/26 Exim was recognizing IPv6 addresses of the form [IPv6:....] in EHLO
1530 commands, but it was not correctly comparing the address with the actual
1531 client host address. Thus, it would show the EHLO address in Received:
1532 header lines when this was not necessary.
1534 PH/27 Added the % operator to ${eval:}.
1536 PH/28 Exim tries to create and chdir to its spool directory when it starts;
1537 it should be ignoring failures (because with -C, for example, it has lost
1538 privilege). It wasn't ignoring creation failures other than "already
1541 PH/29 Added "crypteq" to the list of supported features that Exim outputs when
1544 PH/30 Fixed (presumably very longstanding) bug in exim_dbmbuild: if it failed
1545 because an input line was too long, either on its own, or by virtue of
1546 too many continuations, the temporary file was not being removed, and the
1547 return code was incorrect.
1549 PH/31 Missing "BOOL" in function definition in filtertest.c.
1551 PH/32 Applied Sieve patches from the maintainer.
1553 TK/05 Domainkeys: Accomodate for a minor API change in libdomainkeys 0.67.
1555 PH/33 Added "verify = not_blind".
1557 PH/34 There are settings for CHOWN_COMMAND and MV_COMMAND that can be used in
1558 Local/Makefile (with some defaults set). These are used in built scripts
1559 such as exicyclog, but they have never been used in the exim_install
1560 script (though there are many overriding facilities there). I have
1561 arranged that the exim_install script now takes note of these two
1564 PH/35 Installed configuration files for Dragonfly.
1566 PH/36 When a locally submitted message by a trusted user did not contain a
1567 From: header, and the sender address was obtained from -f or from an SMTP
1568 MAIL command, and the trusted user did not use -F to supply a sender
1569 name, $originator_name was incorrectly used when constructing a From:
1570 header. Furthermore, $originator_name was used for submission mode
1571 messages from external hosts without From: headers in a similar way,
1572 which is clearly wrong.
1574 PH/37 Added control=suppress_local_fixups.
1576 PH/38 When log_selector = +received_sender was set, and the addition of the
1577 sender made the log line's construction buffer exactly full, or one byte
1578 less than full, an overflow happened when the terminating "\n" was
1581 PH/39 Added a new log selector, "unknown_in_list", which provokes a log entry
1582 when the result of a list match is failure because a DNS lookup failed.
1584 PH/40 RM_COMMAND is now used in the building process.
1586 PH/41 Added a "distclean" target to the top-level Makefile; it deletes all
1587 the "build-* directories that it finds.
1589 PH/42 (But a TF fix): In a domain list, Exim incorrectly matched @[] if the IP
1590 address in a domain literal was a prefix of an interface address.
1592 PH/43 (Again a TF fix): In the dnslookup router, do not apply widen_domains
1593 when verifying a sender address, unless rewrite_headers is false.
1595 PH/44 Wrote a long comment about why errors_to addresses are verified as
1596 recipients, not senders.
1598 TF/01 Add missing LIBS=-lm to OS/Makefile-OpenBSD which was overlooked when
1599 the ratelimit ACL was added.
1601 PH/45 Added $smtp_command for the full command (cf $smtp_command_argument).
1603 PH/46 Added extra information about PostgreSQL errors to the error string.
1605 PH/47 Added an interface to a fake DNS resolver for use by the new test suite,
1606 avoiding the need to install special zones in a real server. This is
1607 backwards compatible; if it can't find the fake resolver, it drops back.
1608 Thus, both old and new test suites can be run.
1610 TF/02 Added util/ratelimit.pl
1612 TF/03 Minor fix to the ratelimit code to improve its behaviour in case the
1613 clock is set back in time.
1615 TF/04 Fix the ratelimit support in exim_fixdb. Patch provided by Brian
1616 Candler <B.Candler@pobox.com>.
1618 TF/05 The fix for PH/43 was not completely correct; widen_domains is always
1619 OK for addresses that are the result of redirections.
1621 PH/48 A number of further additions for the benefit of the new test suite,
1622 including a fake gethostbyname() that interfaces to the fake DNS resolver
1625 TF/06 The fix for widen_domains has also been applied to qualify_single and
1626 search_parents which are the other dnslookup options that can cause
1629 PH/49 Michael Haardt's randomized retrying, but as a separate retry parameter
1632 PH/50 Make never_users, trusted_users, admin_groups, trusted_groups expandable.
1634 TF/07 Exim produced the error message "an SRV record indicated no SMTP
1635 service" if it encountered an MX record with an empty target hostname.
1636 The message is now "an MX or SRV record indicated no SMTP service".
1638 TF/08 Change PH/13 introduced the possibility that verify=helo may defer,
1639 if the DNS of the sending site is misconfigured. This is quite a
1640 common situation. This change restores the behaviour of treating a
1641 helo verification defer as a failure.
1643 PH/51 If self=fail was set on a router, the bounce message did not include the
1644 actual error message.
1650 TF/01 Added support for Client SMTP Authorization. See NewStuff for details.
1652 PH/01 When a transport filter timed out in a pipe delivery, and the pipe
1653 command itself ended in error, the underlying message about the transport
1654 filter timeout was being overwritten with the pipe command error. Now the
1655 underlying error message should be appended to the second error message.
1657 TK/01 Fix poll() being unavailable on Mac OSX 10.2.
1659 PH/02 Reduce the amount of output that "make" produces by default. Full output
1660 can still be requested.
1662 PH/03 The warning log line about a condition test deferring for a "warn" verb
1663 was being output only once per connection, rather than after each
1664 occurrence (because it was using the same function as for successful
1665 "warn" verbs). This seems wrong, so I have changed it.
1667 TF/02 Two buglets in acl.c which caused Exim to read a few bytes of memory that
1668 it should not have, which might have caused a crash in the right
1669 circumstances, but probably never did.
1671 PH/04 Installed a modified version of Tony Finch's patch to make submission
1672 mode fix the return path as well as the Sender: header line, and to
1673 add a /name= option so that you can make the user's friendly name appear
1676 TF/03 Added the control = fakedefer ACL modifier.
1678 TF/04 Added the ratelimit ACL condition. See NewStuff for details. Thanks to
1679 Mark Lowes for thorough testing.
1681 TK/02 Rewrote SPF support to work with libspf2 versions >1.2.0.
1683 TK/03 Merged latest SRS patch from Miles Wilton.
1685 PH/05 There's a shambles in IRIX6 - it defines EX_OK in unistd.h which conflicts
1686 with the definition in sysexits.h (which is #included earlier).
1687 Fortunately, Exim does not actually use EX_OK. The code used to try to
1688 preserve the sysexits.h value, by assumimg that macro definitions were
1689 scanned for macro replacements. I have been disabused of this notion,
1690 so now the code just undefines EX_OK before #including unistd.h.
1692 PH/06 There is a timeout for writing blocks of data, set by, e.g. data_timeout
1693 in the smtp transport. When a block could not be written in a single
1694 write() function, the timeout was being re-applied to each part-write.
1695 This seems wrong - if the receiver was accepting one byte at a time it
1696 would take for ever. The timeout is now adjusted when this happens. It
1697 doesn't have to be particularly precise.
1699 TK/04 Added simple SPF lookup method in EXPERIMENTAL_SPF. See NewStuff for
1700 details. Thanks to Chris Webb <chris@arachsys.com> for the patch!
1702 PH/07 Added "fullpostmaster" verify option, which does a check to <postmaster>
1703 without a domain if the check to <postmaster@domain> fails.
1705 SC/01 Eximstats: added -xls and the ability to specify output files
1706 (patch written by Frank Heydlauf).
1708 SC/02 Eximstats: use FileHandles for outputing results.
1710 SC/03 Eximstats: allow any combination of xls, txt, and html output.
1712 SC/04 Eximstats: fixed display of large numbers with -nvr option
1714 SC/05 Eximstats: fixed merging of reports with empty tables.
1716 SC/06 Eximstats: added the -include_original_destination flag
1718 SC/07 Eximstats: removed tabs and trailing whitespace.
1720 TK/05 Malware: Improve on aveserver error handling. Patch from Alex Miller.
1722 TK/06 MBOX spool code: Add real "From " MBOX separator line
1723 so the .eml file is really in mbox format (even though
1724 most programs do not really care). Patch from Alex Miller.
1726 TK/07 MBOX spool code: Add X-Envelope-From: and X-Envelope-To: headers.
1727 The latter is generated from $received_to and is only set if the
1728 message has one envelope recipient. SA can use these headers,
1729 obviously out-of-the-box. Patch from Alex Miller.
1731 PH/08 The ${def test on a variable was returning false if the variable's
1732 value was "0", contrary to what the specification has always said!
1733 The result should be true unless the variable is empty.
1735 PH/09 The syntax error of a character other than { following "${if
1736 def:variable_name" (after optional whitespace) was not being diagnosed.
1737 An expansion such as ${if def:sender_ident:{xxx}{yyy}} in which an
1738 accidental colon was present, for example, could give incorrect results.
1740 PH/10 Tidied the code in a number of places where the st_size field of a stat()
1741 result is used (not including appendfile, where other changes are about
1744 PH/11 Upgraded appendfile so that quotas larger than 2G are now supported.
1745 This involved changing a lot of size variables from int to off_t. It
1746 should work with maildirs and everything.
1748 TK/08 Apply fix provided by Michael Haardt to prevent deadlock in case of
1749 spamd dying while we are connected to it.
1751 TF/05 Fixed a ${extract error message typo reported by Jeremy Harris
1754 PH/12 Applied Alex Kiernan's patch for the API change for the error callback
1755 function for BDB 4.3.
1757 PH/13 Changed auto_thaw such that it does not apply to bounce messages.
1759 PH/14 Imported PCRE 6.0; this was more than just a trivial operation because
1760 the sources for PCRE have been re-arranged and more files are now
1763 PH/15 The code I had for printing potentially long long variables in PH/11
1764 above was not the best (it lost precision). The length of off_t variables
1765 is now inspected at build time, and an appropriate printing format (%ld
1766 or %lld) is chosen and #defined by OFF_T_FMT. We also define LONGLONG_T
1767 to be "long long int" or "long int". This is needed for the internal
1768 formatting function string_vformat().
1770 PH/16 Applied Matthew Newton's patch to exicyclog: "If log_file_path is set in
1771 the configuration file to be ":syslog", then the script "guesses" where
1772 the logs files are, rather than using the compiled in default. In our
1773 case the guess is not the same as the compiled default, so the script
1774 suddenly stopped working when I started to use syslog. The patch checks
1775 to see if log_file_path is "". If so, it attempts to read it from exim
1776 with no configuration file to get the compiled in version, before it
1777 falls back to the previous guessing code."
1779 TK/09 Added "prvs" and "prvscheck" expansion items. These help a lot with
1780 implementing BATV in an Exim configuration. See NewStuff for the gory
1783 PH/17 Applied Michael Haardt's patch for HP-UX, affecting only the os.h and
1784 Makefile that are specific to HP-UX.
1786 PH/18 If the "use_postmaster" option was set for a recipient callout together
1787 with the "random" option, the postmaster address was used as the MAIL
1788 FROM address for the random test, but not for the subsequent recipient
1789 test. It is now used for both.
1791 PH/19 Applied Michael Haardt's patch to update Sieve to RFC3028bis. "The
1792 patch removes a few documentation additions to RFC 3028, because the
1793 latest draft now contains them. It adds the new en;ascii-case comparator
1794 and a new error check for 8bit text in MIME parts. Comparator and
1795 require names are now matched exactly. I enabled the subaddress
1796 extension, but it is not well tested yet (read: it works for me)."
1798 PH/20 Added macros for time_t as for off_t (see PH/15 above) and used them to
1799 rework some of the code of TK/09 above to avoid the hardwired use of
1800 "%lld" and "long long". Replaced the call to snprintf() with a call to
1803 PH/21 Added some other messages to those in 4.51/PH/42, namely "All relevant MX
1804 records point to non-existent hosts", "retry timeout exceeded", and
1805 "retry time not reached for any host after a long failure period".
1807 PH/22 Fixed some oversights/typos causing bugs when Exim is compiled with
1808 experimental DomainKeys support:
1810 (1) The filter variables $n0-$n9 and $sn0-$sn9 were broken.
1811 (2) On an error such as an illegally used "control", the wrong name for
1812 the control was given.
1814 These problems did NOT occur unless DomainKeys support was compiled.
1816 PH/23 Added daemon_startup_retries and daemon_startup_sleep.
1818 PH/24 Added ${if match_ip condition.
1820 PH/25 Put debug statements on either side of calls to EXIM_DBOPEN() for hints
1821 databases so that it will be absolutely obvious if a crash occurs in the
1822 DB library. This is a regular occurrence (often caused by mis-matched
1825 PH/26 Insert a lot of missing (void) casts for functions such as chown(),
1826 chmod(), fcntl(), sscanf(), and other functions from stdio.h. These were
1827 picked up on a user's system that detects such things. There doesn't seem
1828 to be a gcc warning option for this - only an attribute that has to be
1829 put on the function's prototype. It seems that in Fedora Core 4 they have
1830 set this on a number of new functions. No doubt there will be more in due
1833 PH/27 If a dnslookup or manualroute router is set with verify=only, it need not
1834 specify a transport. However, if an address that was verified by such a
1835 router was the subject of a callout, Exim crashed because it tried to
1836 read the rcpt_include_affixes from the non-existent transport. Now it
1837 just assumes that the setting of that option is false. This bug was
1838 introduced by 4.51/PH/31.
1840 PH/28 Changed -d+all to exclude +memory, because that information is very
1841 rarely of interest, but it makes the output a lot bigger. People tend to
1842 do -d+all out of habit.
1844 PH/29 Removed support for the Linux-libc5 build, as it is obsolete and the
1845 code in os-type was giving problems when libc.so lives in lib64, like on
1848 PH/30 Exim's DNS code uses the original T_xxx names for DNS record times. These
1849 aren't the modern standard, and it seems that some systems' include files
1850 don't always have them. Exim was already checking for some of the newer
1851 ones like T_AAAA, and defining it itself. I've added checks for all the
1852 record types that Exim uses.
1854 PH/31 When using GnuTLS, if the parameters cache file did not exist, Exim was
1855 not automatically generating a new one, as it is supposed to. This
1856 prevented TLS from working. If the file did exist, but contained invalid
1857 data, a new version was generated, as expected. It was only the case of a
1858 non-existent file that was broken.
1860 TK/10 Domainkeys: Fix a bug in verification that caused a crash in conjunction
1861 with a change in libdomainkeys > 0.64.
1863 TK/11 Domainkeys: Change the logic how the "testing" policy flag is retrieved
1864 from DNS. If the selector record carries the flag, it now has
1865 precedence over the domain-wide flag.
1867 TK/12 Cleared some compiler warnings related to SPF, SRS and DK code.
1869 PH/32 In mua_wrapper mode, if an smtp transport configuration error (such as
1870 the use of a port name that isn't defined in /etc/services) occurred, the
1871 message was deferred as in a normal delivery, and thus remained on the
1872 spool, instead of being failed because of the mua_wrapper setting. This
1873 is now fixed, and I tidied up some of the mua_wrapper messages at the
1876 SC/08 Eximstats: whilst parsing the mainlog(s), store information about
1877 the messages in a hash of arrays rather than using individual hashes.
1878 This is a bit cleaner and results in dramatic memory savings, albeit
1879 at a slight CPU cost.
1881 SC/09 Eximstats: added the -show_rt<list> and the -show_dt<list> flags
1882 as requested by Marc Sherman.
1884 SC/10 Eximstats: added histograms for user specified patterns as requested
1887 SC/11 Eximstats: v1.43 - bugfix for pattern histograms with -h0 specified.
1889 PH/33 Patch from the Cygwin maintainer to add "b" to all occurences of
1890 fopen() in the content-scanning modules that did not already have it.
1896 TK/01 Added Yahoo DomainKeys support via libdomainkeys. See
1897 doc/experimental-spec.txt for details. (http://domainkeys.sf.net)
1899 TK/02 Fix ACL "control" statement not being available in MIME ACL.
1901 TK/03 Fix ACL "regex" condition not being available in MIME ACL.
1903 PH/01 Installed a patch from the Sieve maintainer that allows -bf to be used
1904 to test Sieve filters that use "vacation".
1906 PH/02 Installed a slightly modified version of Nikos Mavrogiannopoulos' patch
1907 that changes the way the GnuTLS parameters are stored in the cache file.
1908 The new format can be generated externally. For backward compatibility,
1909 if the data in the cache doesn't make sense, Exim assumes it has read an
1910 old-format file, and it generates new data and writes a new file. This
1911 means that you can't go back to an older release without removing the
1914 PH/03 A redirect router that has both "unseen" and "one_time" set does not
1915 work if there are any delivery delays because "one_time" forces the
1916 parent to be marked "delivered", so its unseen clone is never tried
1917 again. For this reason, Exim now forbids the simultaneous setting of
1920 PH/04 Change 4.11/85 fixed an obscure bug concerned with addresses that are
1921 redirected to themselves ("homonym" addresses). Read the long ChangeLog
1922 entry if you want to know the details. The fix, however, neglected to
1923 consider the case when local delivery batching is involved. The test for
1924 "previously delivered" was not happening when checking to see if an
1925 address could be batched with a previous (undelivered) one; under
1926 certain circumstances this could lead to multiple deliveries to the same
1929 PH/05 Renamed the macro SOCKLEN_T as EXIM_SOCKLEN_T because AIX uses SOCKLEN_T
1930 in its include files, and this causes problems building Exim.
1932 PH/06 A number of "verify =" ACL conditions have no options (e.g. verify =
1933 header_syntax) but Exim was just ignoring anything given after a slash.
1934 In particular, this caused confusion with an attempt to use "verify =
1935 reverse_host_lookup/defer_ok". An error is now given when options are
1936 supplied for verify items that do not have them. (Maybe reverse_host_
1937 lookup should have a defer_ok option, but that's a different point.)
1939 PH/07 Increase the size of the buffer for incoming SMTP commands from 512 (as
1940 defined by RFC 821) to 2048, because there were problems with some AUTH
1941 commands, and RFC 1869 says the size should be increased for extended
1942 SMTP commands that take arguments.
1944 PH/08 Added ${dlfunc dynamically loaded function for expansion (code from Tony
1947 PH/09 Previously, an attempt to use ${perl when it wasn't compiled gave an
1948 "unknown" error; now it says that the functionality isn't in the binary.
1950 PH/10 Added a nasty fudge to try to recognize and flatten LDAP passwords in
1951 an address' error message when a string expansion fails (syntax or
1952 whatever). Otherwise the password may appear in the log. Following change
1953 PH/42 below, there is no longer a chance of it appearing in a bounce
1956 PH/11 Installed exipick version 20050225.0 from John Jetmore.
1958 PH/12 If the last host in a fallback_hosts list was multihomed, only the first
1959 of its addresses was ever tried. (Bugzilla bug #2.)
1961 PH/13 If "headers_add" in a transport didn't end in a newline, Exim printed
1962 the result incorrectly in the debug output. (It correctly added a newline
1963 to what was transported.)
1965 TF/01 Added $received_time.
1967 PH/14 Modified the default configuration to add an acl_smtp_data ACL, with
1968 commented out examples of how to interface to a virus scanner and to
1969 SpamAssassin. Also added commented examples of av_scanner and
1970 spamd_address settings.
1972 PH/15 Further to TK/02 and TK/03 above, tidied up the tables of what conditions
1973 and controls are allowed in which ACLs. There were a couple of minor
1974 errors. Some of the entries in the conditions table (which is a table of
1975 where they are NOT allowed) were getting very unwieldy; rewrote them as a
1976 negation of where the condition IS allowed.
1978 PH/16 Installed updated OS/os.c-cygwin from the Cygwin maintainer.
1980 PH/17 The API for radiusclient changed at release 0.4.0. Unfortunately, the
1981 header file does not have a version number, so I've had to invent a new
1982 value for RADIUS_LIB_TYPE, namely "RADIUSCLIENTNEW" to request the new
1983 API. The code is untested by me (my Linux distribution still has 0.3.2 of
1984 radiusclient), but it was contributed by a Radius user.
1986 PH/18 Installed Lars Mainka's patch for the support of CRL collections in
1987 files or directories, for OpenSSL.
1989 PH/19 When an Exim process that is running as root has to create an Exim log
1990 file, it does so in a subprocess that runs as exim:exim so as to get the
1991 ownership right at creation (otherwise, other Exim processes might see
1992 the file with the wrong ownership). There was no test for failure of this
1993 fork() call, which would lead to the process getting stuck as it waited
1994 for a non-existent subprocess. Forks do occasionally fail when resources
1995 run out. I reviewed all the other calls to fork(); they all seem to check
1998 PH/20 When checking for unexpected SMTP input at connect time (before writing
1999 the banner), Exim was not dealing correctly with a non-positive return
2000 from the read() function. If the client had disconnected by this time,
2001 the result was a log entry for a synchronization error with an empty
2002 string after "input=" when read() returned zero. If read() returned -1
2003 (an event I could not check), uninitialized data bytes were printed.
2004 There were reports of junk text (parts of files, etc) appearing after
2007 PH/21 Added acl_not_smtp_mime to allow for MIME scanning for non-SMTP messages.
2009 PH/22 Added support for macro redefinition, and (re)definition in between
2010 driver and ACL definitions.
2012 PH/23 The cyrus_sasl authenticator was expanding server_hostname, but then
2013 forgetting to use the resulting value; it was using the unexpanded value.
2015 PH/24 The cyrus_sasl authenticator was advertising mechanisms for which it
2016 hadn't been configured. The fix is from Juergen Kreileder, who
2017 understands it better than I do:
2019 "Here's what I see happening with three configured cyrus_sasl
2020 authenticators configured (plain, login, cram-md5):
2022 On startup auth_cyrus_sasl_init() gets called for each of these.
2023 This means three calls to sasl_listmech() without a specified mech_list.
2024 => SASL tests which mechs of all available mechs actually work
2025 => three warnings about OTP not working
2026 => the returned list contains: plain, login, cram-md5, digest-md5, ...
2028 With the patch, sasl_listmech() also gets called three times. But now
2029 SASL's mech_list option is set to the server_mech specified in the the
2030 authenticator. Or in other words, the answer from sasl_listmech()
2031 gets limited to just the mech you're testing for (which is different
2033 => the return list contains just 'plain' or 'login', 'cram-md5' or
2034 nothing depending on the value of ob->server_mech.
2036 I've just tested the patch: Authentication still works fine,
2037 unavailable mechs specified in the exim configuration are still
2038 caught, and the auth.log warnings about OTP are gone."
2040 PH/25 When debugging is enabled, the contents of the command line are added
2041 to the debugging output, even when log_selector=+arguments is not
2044 PH/26 Change scripts/os-type so that when "uname -s" returns just "GNU", the
2045 answer is "GNU", and only if the return is "GNU/something" is the answer
2048 PH/27 $acl_verify_message is now set immediately after the failure of a
2049 verification in an ACL, and so is available in subsequent modifiers. In
2050 particular, the message can be preserved by coding like this:
2052 warn !verify = sender
2053 set acl_m0 = $acl_verify_message
2055 Previously, $acl_verify_message was set only while expanding "message"
2056 and "log_message" when a very denied access.
2058 PH/28 Modified OS/os.c-Linux with
2060 -#ifndef OS_LOAD_AVERAGE
2061 +#if !defined(OS_LOAD_AVERAGE) && defined(__linux__)
2063 to make Exim compile on kfreebsd-gnu. (I'm totally confused about the
2064 nomenclature these days.)
2066 PH/29 Installed patch from the Sieve maintainer that adds the options
2067 sieve_useraddress and sieve_subaddress to the redirect router.
2069 PH/30 In these circumstances:
2070 . Two addresses routed to the same list of hosts;
2071 . First host does not offer TLS;
2072 . First host accepts first address;
2073 . First host gives temporary error to second address;
2074 . Second host offers TLS and a TLS session is established;
2075 . Second host accepts second address.
2076 Exim incorrectly logged both deliveries with the TLS parameters (cipher
2077 and peerdn, if requested) that were in fact used only for the second
2080 PH/31 When doing a callout as part of verifying an address, Exim was not paying
2081 attention to any local part prefix or suffix that was matched by the
2082 router that accepted the address. It now behaves in the same way as it
2083 does for delivery: the affixes are removed from the local part unless
2084 rcpt_include_affixes is set on the transport.
2086 PH/32 Add the sender address, as F=<...>, to the log line when logging a
2087 timeout during the DATA phase of an incoming message.
2089 PH/33 Sieve envelope tests were broken for match types other than :is. I have
2090 applied a patch sanctioned by the Sieve maintainer.
2092 PH/34 Change 4.50/80 broke Exim in that it could no longer handle cases where
2093 the uid or gid is negative. A case of a negative gid caused this to be
2094 noticed. The fix allows for either to be negative.
2096 PH/35 ACL_WHERE_MIME is now declared unconditionally, to avoid too much code
2097 clutter, but the tables that are indexed by ACL_WHERE_xxx values had been
2100 PH/36 The change PH/12 above was broken. Fixed it.
2102 PH/37 Exim used to check for duplicate addresses in the middle of routing, on
2103 the grounds that routing the same address twice would always produce the
2104 same answer. This might have been true once, but it is certainly no
2105 longer true now. Routing a child address may depend on the previous
2106 routing that produced that child. Some complicated redirection strategies
2107 went wrong when messages had multiple recipients, and made Exim's
2108 behaviour dependent on the order in which the addresses were given.
2110 I have moved the duplicate checking until after the routing is complete.
2111 Exim scans the addresses that are assigned to local and remote
2112 transports, and removes any duplicates. This means that more work will be
2113 done, as duplicates will always all be routed, but duplicates are
2114 presumably rare, so I don't expect this is of any significance.
2116 For deliveries to pipes, files, and autoreplies, the duplicate checking
2117 still happens during the routing process, since they are not going to be
2120 PH/38 Installed a patch from Ian Freislich, with the agreement of Tom Kistner.
2121 It corrects a timeout issue with spamd. This is Ian's comment: "The
2122 background is that sometimes spamd either never reads data from a
2123 connection it has accepted, or it never writes response data. The exiscan
2124 spam.[ch] uses a 3600 second timeout on spamd socket reads, further, it
2125 blindly assumes that writes won't block so it may never time out."
2127 PH/39 Allow G after quota size as well as K and M.
2129 PH/40 The value set for $authenticated_id in an authenticator may not contain
2130 binary zeroes or newlines because the value is written to log lines and
2131 to spool files. There was no check on this. Now the value is run through
2132 the string_printing() function so that such characters are converted to
2133 printable escape sequences.
2135 PH/41 $message_linecount is a new variable that contains the total number of
2136 lines in the message. Compare $body_linecount, which is the count for the
2139 PH/42 Exim no longer gives details of delivery errors for specific addresses in
2140 bounce and delay warning messages, except in certain special cases, which
2143 (a) An SMTP error message from a remote host;
2144 (b) A message specified in a :fail: redirection;
2145 (c) A message specified in a "fail" command in a system filter;
2146 (d) A message specified in a FAIL return from the queryprogram router;
2147 (e) A message specified by the cannot_route_message router option.
2149 In these cases only, Exim does include the error details in bounce and
2150 warning messages. There are also a few cases where bland messages such
2151 as "unrouteable address" or "local delivery error" are given.
2153 PH/43 $value is now also set for the "else" part of a ${run expansion.
2155 PH/44 Applied patch from the Sieve maintainer: "The vacation draft is still
2156 being worked on, but at least Exim now implements the latest version to
2159 PH/45 In a pipe transport, although a timeout while waiting for the pipe
2160 process to complete was treated as a delivery failure, a timeout while
2161 writing the message to the pipe was logged, but erroneously treated as a
2162 successful delivery. Such timeouts include transport filter timeouts. For
2163 consistency with the overall process timeout, these timeouts are now
2164 treated as errors, giving rise to delivery failures by default. However,
2165 there is now a new Boolean option for the pipe transport called
2166 timeout_defer, which, if set TRUE, converts the failures into defers for
2167 both kinds of timeout. A transport filter timeout is now identified in
2170 PH/46 The "scripts/Configure-config.h" script calls "make" at one point. On
2171 systems where "make" and "gmake" are different, calling "gmake" at top
2172 level broke things. I've arranged for the value of $(MAKE) to be passed
2173 from the Makefile to this script so that it can call the same version of
2177 A note about Exim versions 4.44 and 4.50
2178 ----------------------------------------
2180 Exim 4.50 was meant to be the next release after 4.43. It contains a lot of
2181 changes of various kinds. As a consequence, a big documentation update was
2182 needed. This delayed the release for rather longer than seemed good, especially
2183 in the light of a couple of (minor) security issues. Therefore, the changes
2184 that fixed bugs were backported into 4.43, to create a 4.44 maintenance
2185 release. So 4.44 and 4.50 are in effect two different branches that both start
2188 I have left the 4.50 change log unchanged; it contains all the changes since
2189 4.43. The change log for 4.44 is below; many of its items are identical to
2190 those for 4.50. This seems to be the most sensible way to preserve the
2191 historical information.
2197 1. Minor wording change to the doc/README.SIEVE file.
2199 2. Change 4.43/35 introduced a bug: if quota_filecount was set, the
2200 computation of the current number of files was incorrect.
2202 3. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2203 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2206 4. Give more explanation in the error message when the command for a transport
2207 filter fails to execute.
2209 5. There are several places where Exim runs a non-Exim command in a
2210 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2211 was being done only for the command run by the queryprogram router. It is
2212 now done for all such subprocesses. The other cases are: ${run, transport
2213 filters, and the commands run by the lmtp and pipe transports.
2215 6. Added CONFIGURE_GROUP build-time option.
2217 7. Some older OS have a limit of 256 on the maximum number of file
2218 descriptors. Exim was using setrlimit() to set 1000 as a large value
2219 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2220 systems. I've change it so that if it can't get 1000, it tries for 256.
2222 8. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2223 was an oversight, and furthermore, ever since the addition of extra
2224 controls (e.g. 4.43/32), the checks on when to allow different forms of
2225 "control" were broken. There should now be diagnostics for all cases when a
2226 control that does not make sense is encountered.
2228 9. Added the /retain_sender option to "control=submission".
2230 10. $recipients is now available in the predata ACL (oversight).
2232 11. Tidy the search cache before the fork to do a delivery from a message
2233 received from the command line. Otherwise the child will trigger a lookup
2234 failure and thereby defer the delivery if it tries to use (for example) a
2235 cached ldap connection that the parent has called unbind on.
2237 12. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2238 of $address_data from the recipient verification was clobbered by the
2239 sender verification.
2241 13. The value of address_data from a sender verification is now available in
2242 $sender_address_data in subsequent conditions in the ACL statement.
2244 14. Added forbid_sieve_filter and forbid_exim_filter to the redirect router.
2246 15. Added a new option "connect=<time>" to callout options, to set a different
2249 16. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2250 was its contents. (It was OK if the option was not defined at all.)
2252 17. A "Completed" log line is now written for messages that are removed from
2253 the spool by the -Mrm option.
2255 18. New variables $sender_verify_failure and $recipient_verify_failure contain
2256 information about exactly what failed.
2258 19. Added -dd to debug only the daemon process.
2260 20. Incorporated Michael Haardt's patch to ldap.c for improving the way it
2261 handles timeouts, both on the server side and network timeouts. Renamed the
2262 CONNECT parameter as NETTIMEOUT (but kept the old name for compatibility).
2264 21. The rare case of EHLO->STARTTLS->HELO was setting the protocol to "smtp".
2265 It is now set to "smtps".
2267 22. $host_address is now set to the target address during the checking of
2268 ignore_target_hosts.
2270 23. When checking ignore_target_hosts for an ipliteral router, no host name was
2271 being passed; this would have caused $sender_host_name to have been used if
2272 matching the list had actually called for a host name (not very likely,
2273 since this list is usually IP addresses). A host name is now passed as
2276 24. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2277 code that specifies a non-restarting handler (typically sigaction() in
2278 modern systems) in an attempt to fix a rare and obscure crash bug.
2280 25. Narrowed the window for a race in the daemon that could cause it to ignore
2281 SIGCHLD signals. This is not a major problem, because they are used only to
2282 wake it up if nothing else does.
2284 26. A malformed maildirsize file could cause Exim to calculate negative values
2285 for the mailbox size or file count. Odd effects could occur as a result.
2286 The maildirsize information is now recalculated if the size or filecount
2289 27. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2290 support for a long time. Removed HAVE_SYS_VFS_H.
2292 28. Installed the latest version of exipick from John Jetmore.
2294 29. In an address list, if the pattern was not a regular expression, an empty
2295 subject address (from a bounce message) matched only if the pattern was an
2296 empty string. Non-empty patterns were not even tested. This was the wrong
2297 because it is perfectly reasonable to use an empty address as part of a
2298 database query. An empty address is now tested by patterns that are
2299 lookups. However, all the other forms of pattern expect the subject to
2300 contain a local part and a domain, and therefore, for them, an empty
2301 address still always fails if the pattern is not itself empty.
2303 30. Exim went into a mad DNS loop when attempting to do a callout where the
2304 host was specified on an smtp transport, and looking it up yielded more
2305 than one IP address.
2307 31. Re-factored the code for checking spool and log partition space into a
2308 function that finds that data and another that does the check. The former
2309 is then used to implement four new variables: $spool_space, $log_space,
2310 $spool_inodes, and $log_inodes.
2312 32. The RFC2047 encoding function was originally intended for short strings
2313 such as real names; it was not keeping to the 75-character limit for
2314 encoded words that the RFC imposes. It now respects the limit, and
2315 generates multiple encoded words if necessary. To be on the safe side, I
2316 have increased the buffer size for the ${rfc2047: expansion operator from
2319 33. It is now permitted to omit both strings after an "if" condition; if the
2320 condition is true, the result is "true". As before, when the second string
2321 is omitted, a false condition yields an empty string. This makes it less
2322 cumbersome to write custom ACL and router conditions.
2324 34. Failure to deliver a bounce message always caused it to be frozen, even if
2325 there was an errors_to setting on the router. The errors_to setting is now
2328 35. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2329 canonical form (fully expanded) before being placed in
2330 $sender_host_address.
2332 36. The table in the code that translates DNS record types into text (T_A to
2333 "A" for instance) was missing entries for NS and CNAME. It is just possible
2334 that this could have caused confusion if both these types were looked up
2335 for the same domain, because the text type is used as part of Exim's
2336 per-process caching. But the chance of anyone hitting this buglet seems
2339 37. The dnsdb lookup has been extended in a number of ways.
2341 (1) There is a new type, "zns", which walks up the domain tree until it
2342 finds some nameserver records. It should be used with care.
2344 (2) There is a new type, "mxh", which is like "mx" except that it returns
2345 just the host names, not the priorities.
2347 (3) It is now possible to give a list of domains (or IP addresses) to be
2348 looked up. The behaviour when one of the lookups defers can be
2349 controlled by a keyword.
2351 (4) It is now possible to specify the separator character for use when
2352 multiple records are returned.
2354 38. The dnslists ACL condition has been extended: it is now possible to supply
2355 a list of IP addresses and/or domains to be looked up in a particular DNS
2358 39. Added log_selector=+queue_time_overall.
2360 40. When running the queue in the test harness, wait just a tad after forking a
2361 delivery process, to get repeatability of debugging output.
2363 41. Include certificate and key file names in error message when GnuTLS fails
2364 to set them up, because the GnuTLS error message doesn't include the name
2365 of the failing file when there is a problem reading it.
2367 42. Allow both -bf and -bF in the same test run.
2369 43. Did the same fix as 41 above for OpenSSL, which had the same infelicity.
2371 44. The "Exiscan patch" is now merged into the mainline Exim source.
2373 45. Sometimes the final signoff response after QUIT could fail to get
2374 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2375 < 0 before doing a fflush(). This bug looks as though it goes back to the
2376 introduction of TLS in release 3.20, but "sometimes" must have been rare
2377 because the tests only now provoked it.
2379 46. Reset the locale to "C" after calling embedded Perl, in case it was changed
2380 (this can affect the format of dates).
2382 47. exim_tidydb, when checking for the continued existence of a message for
2383 which it has found a message-specific retry record, was not finding
2384 messages that were in split spool directories. Consequently, it was
2385 deleting retry records that should have stayed in existence.
2387 48. Steve fixed some bugs in eximstats.
2389 49. The SPA authentication driver was not abandoning authentication and moving
2390 on to the next authenticator when an expansion was forced to fail,
2391 contradicting the general specification for all authenticators. Instead it
2392 was generating a temporary error. It now behaves as specified.
2394 50. The default ordering of permitted cipher suites for GnuTLS was pessimal
2395 (the order specifies the preference for clients). The order is now AES256,
2396 AES128, 3DES, ARCFOUR128.
2398 51. Small patch to Sieve code - explicitly set From: when generating an
2401 52. Exim crashed if a remote delivery caused a very long error message to be
2402 recorded - for instance if somebody sent an entire SpamAssassin report back
2403 as a large number of 550 error lines. This bug was coincidentally fixed by
2404 increasing the size of one of Exim's internal buffers (big_buffer) that
2405 happened as part of the Exiscan merge. However, to be on the safe side, I
2406 have made the code more robust (and fixed the comments that describe what
2409 53. Now that there can be additional text after "Completed" in log lines (if
2410 the queue_time_overall log selector is set), a one-byte patch to exigrep
2411 was needed to allow it to recognize "Completed" as not the last thing in
2414 54. The LDAP lookup was not handling a return of LDAP_RES_SEARCH_REFERENCE. A
2415 patch that reportedly fixes this has been added. I am not expert enough to
2416 create a test for it. This is what the patch creator wrote:
2418 "I found a little strange behaviour of ldap code when working with
2419 Windows 2003 AD Domain, where users was placed in more than one
2420 Organization Units. When I tried to give exim partial DN, the exit code
2421 of ldap_search was unknown to exim because of LDAP_RES_SEARCH_REFERENCE.
2422 But simultaneously result of request was absolutely normal ldap result,
2423 so I produce this patch..."
2425 Later: it seems that not all versions of LDAP support LDAP_RES_SEARCH_
2426 REFERENCE, so I have modified the code to exclude the patch when that macro
2429 55. Some experimental protocols are using DNS PTR records for new purposes. The
2430 keys for these records are domain names, not reversed IP addresses. The
2431 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2432 leaves it alone. Component reversal etc. now happens only for IP addresses.
2435 56. Improve error message when ldap_search() fails in OpenLDAP or Solaris LDAP.
2437 57. Double the size of the debug message buffer (to 2048) so that more of very
2438 long debug lines gets shown.
2440 58. The exicyclog utility now does better if the number of log files to keep
2441 exceeds 99. In this case, it numbers them 001, 002 ... instead of 01, 02...
2443 59. Two changes related to the smtp_active_hostname option:
2445 (1) $smtp_active_hostname is now available as a variable.
2446 (2) The default for smtp_banner uses $smtp_active_hostname instead
2447 of $primary_hostname.
2449 60. The host_aton() function is supposed to be passed a string that is known
2450 to be a valid IP address. However, in the case of IPv6 addresses, it was
2451 not checking this. This is a hostage to fortune. Exim now panics and dies
2452 if the condition is not met. A case was found where this could be provoked
2453 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2454 components; fortuitously, this particular loophole had already been fixed
2455 by change 4.50/55 above.
2457 If there are any other similar loopholes, the new check in host_aton()
2458 itself should stop them being exploited. The report I received stated that
2459 data on the command line could provoke the exploit when Exim was running as
2460 exim, but did not say which command line option was involved. All I could
2461 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2462 running as the user.
2465 61. There was a buffer overflow vulnerability in the SPA authentication code
2466 (which came originally from the Samba project). I have added a test to the
2467 spa_base64_to_bits() function which I hope fixes it.
2470 62. Configuration update for GNU/Hurd and variations. Updated Makefile-GNU and
2471 os.h-GNU, and added configuration files for GNUkFreeBSD and GNUkNetBSD.
2473 63. The daemon start-up calls getloadavg() while still root for those OS that
2474 need the first call to be done as root, but it missed one case: when
2475 deliver_queue_load_max is set with deliver_drop_privilege. This is
2476 necessary for the benefit of the queue runner, because there is no re-exec
2477 when deliver_drop_privilege is set.
2479 64. A call to exiwhat cut short delays set up by "delay" modifiers in ACLs.
2480 This has been fixed.
2482 65. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2483 was in use, was not putting the data itself into the right store pool;
2484 consequently, it could be overwritten for a subsequent message in the same
2485 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2488 66. Added hosts_max_try_hardlimit to the smtp transport, default 50.
2490 67. The string_is_ip_address() function returns 0, 4, or 6, for "no an IP
2491 address", "IPv4 address", and "IPv6 address", respectively. Some calls of
2492 the function were treating the return as a boolean value, which happened to
2493 work because 0=false and not-0=true, but is not correct code.
2495 68. The host_aton() function was not handling scoped IPv6 addresses (those
2496 with, for example, "%eth0" on the end) correctly.
2498 69. Fixed some compiler warnings in acl.c for the bitmaps specified with
2499 negated items (that is, ~something) in unsigned ints. Some compilers
2500 apparently mutter when there is no cast.
2502 70. If an address verification called from an ACL failed, and did not produce a
2503 user-specific message (i.e. there was only a "system" message), nothing was
2504 put in $acl_verify_message. In this situation, it now puts the system
2507 71. Change 4.23/11 added synchronization checking at the start of an SMTP
2508 session; change 4.31/43 added the unwanted input to the log line - except
2509 that it did not do this in the start of session case. It now does.
2511 72. After a timeout in a callout SMTP session, Exim still sent a QUIT command.
2512 This is wrong and can cause the other end to generate a synchronization
2513 error if it is another Exim or anything else that does the synchronization
2514 check. A QUIT command is no longer sent after a timeout.
2516 73. $host_lookup_deferred has been added, to make it easier to detect DEFERs
2517 during host lookups.
2519 74. The defer_ok option of callout verification was not working if it was used
2520 when verifying addresses in header lines, that is, for this case:
2522 verify = header_sender/callout=defer_ok
2524 75. A backgrounded daemon closed stdin/stdout/stderr on entry; this meant that
2525 those file descriptors could be used for SMTP connections. If anything
2526 wrote to stderr (the example that came up was "warn" in embedded Perl), it
2527 could be sent to the SMTP client, causing chaos. The daemon now opens
2528 stdin, stdout, and stderr to /dev/null when it puts itself into the
2531 76. Arrange for output from Perl's "warn" command to be written to Exim's main
2532 log by default. The user can override this with suitable Perl magic.
2534 77. The use of log_message on a "discard" ACL verb, which is supposed to add to
2535 the log message when discard triggers, was not working for the DATA ACL or
2536 for the non-SMTP ACL.
2538 78. Error message wording change in sieve.c.
2540 79. If smtp_accept_max_per_host was set, the number of connections could be
2541 restricted to fewer than expected, because the daemon was trying to set up
2542 a new connection before checking whether the processes handling previous
2543 connections had finished. The check for completed processes is now done
2544 earlier. On busy systems, this bug wouldn't be noticed because something
2545 else would have woken the daemon, and it would have reaped the completed
2548 80. If a message was submitted locally by a user whose login name contained one
2549 or more spaces (ugh!), the spool file that Exim wrote was not re-readable.
2550 It caused a spool format error. I have fixed the spool reading code. A
2551 related problem was that the "from" clause in the Received: line became
2552 illegal because of the space(s). It is now covered by ${quote_local_part.
2554 81. Included the latest eximstats from Steve (adds average sizes to HTML Top
2557 82. Updated OS/Makefile-AIX as per message from Mike Meredith.
2559 83. Patch from Sieve maintainer to fix unterminated string problem in
2560 "vacation" handling.
2562 84. Some minor changes to the Linux configuration files to help with other
2563 OS variants using glibc.
2565 85. One more patch for Sieve to update vacation handling to latest spec.
2568 ----------------------------------------------------
2569 See the note above about the 4.44 and 4.50 releases.
2570 ----------------------------------------------------
2576 1. Change 4.43/35 introduced a bug that caused file counts to be
2577 incorrectly computed when quota_filecount was set in an appendfile
2580 2. Closing a stable door: arrange to panic-die if setitimer() ever fails. The
2581 bug fixed in 4.43/37 would have been diagnosed quickly if this had been in
2584 3. Give more explanation in the error message when the command for a transport
2585 filter fails to execute.
2587 4. There are several places where Exim runs a non-Exim command in a
2588 subprocess. The SIGUSR1 signal should be disabled for these processes. This
2589 was being done only for the command run by the queryprogram router. It is
2590 now done for all such subprocesses. The other cases are: ${run, transport
2591 filters, and the commands run by the lmtp and pipe transports.
2593 5. Some older OS have a limit of 256 on the maximum number of file
2594 descriptors. Exim was using setrlimit() to set 1000 as a large value
2595 unlikely to be exceeded. Change 4.43/17 caused a lot of logging on these
2596 systems. I've change it so that if it can't get 1000, it tries for 256.
2598 6. "control=submission" was allowed, but had no effect, in a DATA ACL. This
2599 was an oversight, and furthermore, ever since the addition of extra
2600 controls (e.g. 4.43/32), the checks on when to allow different forms of
2601 "control" were broken. There should now be diagnostics for all cases when a
2602 control that does not make sense is encountered.
2604 7. $recipients is now available in the predata ACL (oversight).
2606 8. Tidy the search cache before the fork to do a delivery from a message
2607 received from the command line. Otherwise the child will trigger a lookup
2608 failure and thereby defer the delivery if it tries to use (for example) a
2609 cached ldap connection that the parent has called unbind on.
2611 9. If verify=recipient was followed by verify=sender in a RCPT ACL, the value
2612 of $address_data from the recipient verification was clobbered by the
2613 sender verification.
2615 10. If FIXED_NEVER_USERS was defined, but empty, Exim was assuming the uid 0
2616 was its contents. (It was OK if the option was not defined at all.)
2618 11. A "Completed" log line is now written for messages that are removed from
2619 the spool by the -Mrm option.
2621 12. $host_address is now set to the target address during the checking of
2622 ignore_target_hosts.
2624 13. When checking ignore_target_hosts for an ipliteral router, no host name was
2625 being passed; this would have caused $sender_host_name to have been used if
2626 matching the list had actually called for a host name (not very likely,
2627 since this list is usually IP addresses). A host name is now passed as
2630 14. Changed the calls that set up the SIGCHLD handler in the daemon to use the
2631 code that specifies a non-restarting handler (typically sigaction() in
2632 modern systems) in an attempt to fix a rare and obscure crash bug.
2634 15. Narrowed the window for a race in the daemon that could cause it to ignore
2635 SIGCHLD signals. This is not a major problem, because they are used only to
2636 wake it up if nothing else does.
2638 16. A malformed maildirsize file could cause Exim to calculate negative values
2639 for the mailbox size or file count. Odd effects could occur as a result.
2640 The maildirsize information is now recalculated if the size or filecount
2643 17. Added HAVE_SYS_STATVFS_H to the os.h file for Linux, as it has had this
2644 support for a long time. Removed HAVE_SYS_VFS_H.
2646 18. Updated exipick to current release from John Jetmore.
2648 19. Allow an empty sender to be matched against a lookup in an address list.
2649 Previously the only cases considered were a regular expression, or an
2652 20. Exim went into a mad DNS lookup loop when doing a callout where the
2653 host was specified on the transport, if the DNS lookup yielded more than
2656 21. The RFC2047 encoding function was originally intended for short strings
2657 such as real names; it was not keeping to the 75-character limit for
2658 encoded words that the RFC imposes. It now respects the limit, and
2659 generates multiple encoded words if necessary. To be on the safe side, I
2660 have increased the buffer size for the ${rfc2047: expansion operator from
2663 22. Failure to deliver a bounce message always caused it to be frozen, even if
2664 there was an errors_to setting on the router. The errors_to setting is now
2667 23. If an IPv6 address is given for -bh or -bhc, it is now converted to the
2668 canonical form (fully expanded) before being placed in
2669 $sender_host_address.
2671 24. Updated eximstats to version 1.33
2673 25. Include certificate and key file names in error message when GnuTLS fails
2674 to set them up, because the GnuTLS error message doesn't include the name
2675 of the failing file when there is a problem reading it.
2677 26. Expand error message when OpenSSL has problems setting up cert/key files.
2680 27. Reset the locale to "C" after calling embedded Perl, in case it was changed
2681 (this can affect the format of dates).
2683 28. exim_tidydb, when checking for the continued existence of a message for
2684 which it has found a message-specific retry record, was not finding
2685 messages that were in split spool directories. Consequently, it was
2686 deleting retry records that should have stayed in existence.
2688 29. eximstats updated to version 1.35
2689 1.34 - allow eximstats to parse syslog lines as well as mainlog lines
2690 1.35 - bugfix such that pie charts by volume are generated correctly
2692 30. The SPA authentication driver was not abandoning authentication and moving
2693 on to the next authenticator when an expansion was forced to fail,
2694 contradicting the general specification for all authenticators. Instead it
2695 was generating a temporary error. It now behaves as specified.
2697 31. The default ordering of permitted cipher suites for GnuTLS was pessimal
2698 (the order specifies the preference for clients). The order is now AES256,
2699 AES128, 3DES, ARCFOUR128.
2701 31. Small patch to Sieve code - explicitly set From: when generating an
2704 32. Exim crashed if a remote delivery caused a very long error message to be
2705 recorded - for instance if somebody sent an entire SpamAssassin report back
2706 as a large number of 550 error lines. This bug was coincidentally fixed by
2707 increasing the size of one of Exim's internal buffers (big_buffer) that
2708 happened as part of the Exiscan merge. However, to be on the safe side, I
2709 have made the code more robust (and fixed the comments that describe what
2712 33. Some experimental protocols are using DNS PTR records for new purposes. The
2713 keys for these records are domain names, not reversed IP addresses. The
2714 dnsdb PTR lookup now tests whether its key is an IP address. If not, it
2715 leaves it alone. Component reversal etc. now happens only for IP addresses.
2718 34. The host_aton() function is supposed to be passed a string that is known
2719 to be a valid IP address. However, in the case of IPv6 addresses, it was
2720 not checking this. This is a hostage to fortune. Exim now panics and dies
2721 if the condition is not met. A case was found where this could be provoked
2722 from a dnsdb PTR lookup with an IPv6 address that had more than 8
2723 components; fortuitously, this particular loophole had already been fixed
2724 by change 4.50/55 or 4.44/33 above.
2726 If there are any other similar loopholes, the new check in host_aton()
2727 itself should stop them being exploited. The report I received stated that
2728 data on the command line could provoke the exploit when Exim was running as
2729 exim, but did not say which command line option was involved. All I could
2730 find was the use of -be with a bad dnsdb PTR lookup, and in that case it is
2731 running as the user.
2734 35. There was a buffer overflow vulnerability in the SPA authentication code
2735 (which came originally from the Samba project). I have added a test to the
2736 spa_base64_to_bits() function which I hope fixes it.
2739 36. The daemon start-up calls getloadavg() while still root for those OS that
2740 need the first call to be done as root, but it missed one case: when
2741 deliver_queue_load_max is set with deliver_drop_privilege. This is
2742 necessary for the benefit of the queue runner, because there is no re-exec
2743 when deliver_drop_privilege is set.
2745 37. Caching of lookup data for "hosts =" ACL conditions, when a named host list
2746 was in use, was not putting the data itself into the right store pool;
2747 consequently, it could be overwritten for a subsequent message in the same
2748 SMTP connection. (Fix 4.40/11 dealt with the non-cache case, but overlooked
2751 38. Sometimes the final signoff response after QUIT could fail to get
2752 transmitted in the non-TLS case. Testing !tls_active instead of tls_active
2753 < 0 before doing a fflush(). This bug looks as though it goes back to the
2754 introduction of TLS in release 3.20, but "sometimes" must have been rare
2755 because the tests only now provoked it.
2761 1. Fixed a longstanding but relatively impotent bug: a long time ago, before
2762 PIPELINING, the function smtp_write_command() used to return TRUE or FALSE.
2763 Now it returns an integer. A number of calls were still expecting a T/F
2764 return. Fortuitously, in all cases, the tests worked in OK situations,
2765 which is the norm. However, things would have gone wrong on any write
2766 failures on the smtp file descriptor. This function is used when sending
2767 messages over SMTP and also when doing verify callouts.
2769 2. When Exim is called to do synchronous delivery of a locally submitted
2770 message (the -odf or -odi options), it no longer closes stderr before doing
2773 3. Implemented the mua_wrapper option.
2775 4. Implemented mx_fail_domains and srv_fail_domains for the dnslookup router.
2777 5. Implemented the functions header_remove(), header_testname(),
2778 header_add_at_position(), and receive_remove_recipient(), and exported them
2781 6. If an ACL "warn" statement specified the addition of headers, Exim already
2782 inserted X-ACL-Warn: at the start if there was no header name. However, it
2783 was not making this test for the second and subsequent header lines if
2784 there were newlines in the string. This meant that an invalid header could
2785 be inserted if Exim was badly configured.
2787 7. Allow an ACL "warn" statement to add header lines at the start or after all
2788 the Received: headers, as well as at the end.
2790 8. Added the rcpt_4xx retry error code.
2792 9. Added postmaster_mailfrom=xxx to callout verification option.
2794 10. Added mailfrom=xxxx to the callout verification option, for verify=
2797 11. ${substr_1_:xxxx} and ${substr__3:xxxx} are now diagnosed as syntax errors
2798 (they previously behaved as ${substr_1_0:xxxx} and ${substr:_0_3:xxxx}).
2800 12. Inserted some casts to stop certain compilers warning when using pointer
2801 differences as field lengths or precisions in printf-type calls (mostly
2802 affecting debugging statements).
2804 13. Added optional readline() support for -be (dynamically loaded).
2806 14. Obscure bug fix: if a message error (e.g. 4xx to MAIL) happened within the
2807 same clock tick as a message's arrival, so that its received time was the
2808 same as the "first fail" time on the retry record, and that message
2809 remained on the queue past the ultimate address timeout, every queue runner
2810 would try a delivery (because it was past the ultimate address timeout) but
2811 after another failure, the ultimate address timeout, which should have then
2812 bounced the address, did not kick in. This was a "< instead of <=" error;
2813 in most cases the first failure would have been in the next clock tick
2814 after the received time, and all would be well.
2816 15. The special items beginning with @ in domain lists (e.g. @mx_any) were not
2817 being recognized when the domain list was tested by the match_domain
2818 condition in an expansion string.
2820 16. Added the ${str2b64: operator.
2822 17. Exim was always calling setrlimit() to set a large limit for the number of
2823 processes, without checking whether the existing limit was already
2824 adequate. (It did check for the limit on file descriptors.) Furthermore,
2825 errors from getrlimit() and setrlimit() were being ignored. Now they are
2826 logged to the main and panic logs, but Exim does carry on, to try to do its
2827 job under whatever limits there are.
2829 18. Imported PCRE 5.0.
2831 19. Trivial typo in log message " temporarily refused connection" (the leading
2834 20. If the log selector return_path_on_delivery was set and an address was
2835 redirected to /dev/null, the delivery process crashed because it assumed
2836 that a return path would always be set for a "successful" delivery. In this
2837 case, the whole delivery is bypassed as an optimization, and therefore no
2840 21. Internal re-arrangement: the function for sending a challenge and reading
2841 a response while authentication was assuming a zero-terminated challenge
2842 string. It's now changed to take a pointer and a length, to allow for
2843 binary data in such strings.
2845 22. Added the cyrus_sasl authenticator (code supplied by MBM).
2847 23. Exim was not respecting finduser_retries when seeking the login of the
2848 uid under which it was called; it was always trying 10 times. (The default
2849 setting of finduser_retries is zero.) Also, it was sleeping after the final
2850 failure, which is pointless.
2852 24. Implemented tls_on_connect_ports.
2854 25. Implemented acl_smtp_predata.
2856 26. If the domain in control=submission is set empty, Exim assumes that the
2857 authenticated id is a complete email address when it generates From: or
2858 Sender: header lines.
2860 27. Added "#define SOCKLEN_T int" to OS/os.h-SCO and OS/os.h-SCO_SV. Also added
2861 definitions to OS/Makefile-SCO and OS/Makefile-SCO_SV that put basename,
2862 chown and chgrp in /bin and hostname in /usr/bin.
2864 28. Exim was keeping the "process log" file open after each use, just as it
2865 does for the main log. This opens the possibility of it remaining open for
2866 long periods when the USR1 signal hits a daemon. Occasional processlog
2867 errors were reported, that could have been caused by this. Anyway, it seems
2868 much more sensible not to leave this file open at all, so that is what now
2871 29. The long-running daemon process does not normally write to the log once it
2872 has entered its main loop, and it closes the log before doing so. This is
2873 so that log files can straightforwardly be renamed and moved. However,
2874 there are a couple of unusual error situations where the daemon does write
2875 log entries, and I had neglected to close the log afterwards.
2877 30. The text of an SMTP error response that was received during a remote
2878 delivery was being truncated at 512 bytes. This is too short for some of
2879 the long messages that one sometimes sees. I've increased the limit to
2882 31. It is now possible to make retry rules that apply only when a message has a
2883 specific sender, in particular, an empty sender.
2885 32. Added "control = enforce_sync" and "control = no_enforce_sync". This makes
2886 it possible to be selective about when SMTP synchronization is enforced.
2888 33. Added "control = caseful_local_part" and "control = "caselower_local_part".
2890 32. Implemented hosts_connection_nolog.
2892 33. Added an ACL for QUIT.
2894 34. Setting "delay_warning=" to disable warnings was not working; it gave a
2897 35. Added mailbox_size and mailbox_filecount to appendfile.
2899 36. Added control = no_multiline_responses to ACLs.
2901 37. There was a bug in the logic of the code that waits for the clock to tick
2902 in the case where the clock went backwards by a substantial amount such
2903 that the microsecond fraction of "now" was more than the microsecond
2904 fraction of "then" (but the whole seconds number was less).
2906 38. Added support for the libradius Radius client library this is found on
2907 FreeBSD (previously only the radiusclient library was supported).
2913 1. When certain lookups returned multiple values in the form name=value, the
2914 quoting of the values was not always being done properly. Specifically:
2915 (a) If the value started with a double quote, but contained no whitespace,
2917 (b) If the value contained whitespace other than a space character (i.e.
2918 tabs or newlines or carriage returns) it was not quoted.
2919 This fix has been applied to the mysql and pgsql lookups by writing a
2920 separate quoting function and calling it from the lookup code. The fix
2921 should probably also be applied to nisplus, ibase and oracle lookups, but
2922 since I cannot test any of those, I have not disturbed their existing code.
2924 2. A hit in the callout cache for a specific address caused a log line with no
2925 reason for rejecting RCPT. Now it says "Previous (cached) callout
2926 verification failure".
2928 3. There was an off-by-one bug in the queryprogram router. An over-long
2929 return line was truncated at 256 instead of 255 characters, thereby
2930 overflowing its buffer with the terminating zero. As well as fixing this, I
2931 have increased the buffer size to 1024 (and made a note to document this).
2933 4. If an interrupt, such as the USR1 signal that is send by exiwhat, arrives
2934 when Exim is waiting for an SMTP response from a remote server, Exim
2935 restarts its select() call on the socket, thereby resetting its timeout.
2936 This is not a problem when such interrupts are rare. Somebody set up a cron
2937 job to run exiwhat every 2 minutes, which is less than the normal select()
2938 timeout (5 or 10 minutes). This meant that the select() timeout never
2939 kicked in because it was always reset. I have fixed this by comparing the
2940 time when an interrupt arrives with the time at the start of the first call
2941 to select(). If more time than the timeout has elapsed, the interrupt is
2942 treated as a timeout.
2944 5. Some internal re-factoring in preparation for the addition of Sieve
2945 extensions (by MH). In particular, the "personal" test is moved to a
2946 separate function, and given an option for scanning Cc: and Bcc: (which is
2947 not set for Exim filters).
2949 6. When Exim created an email address using the login of the caller as the
2950 local part (e.g. when creating a From: or Sender: header line), it was not
2951 quoting the local part when it contained special characters such as @.
2953 7. Installed new OpenBSD configuration files.
2955 8. Reworded some messages for syntax errors in "and" and "or" conditions to
2956 try to make them clearer.
2958 9. Callout options, other than the timeout value, were being ignored when
2959 verifying sender addresses in header lines. For example, when using
2961 verify = header_sender/callout=no_cache
2963 the cache was (incorrectly) being used.
2965 10. Added a missing instance of ${EXE} to the exim_install script; this affects
2966 only the Cygwin environment.
2968 11. When return_path_on_delivery was set as a log selector, if different remote
2969 addresses in the same message used different return paths and parallel
2970 remote delivery occurred, the wrong values would sometimes be logged.
2971 (Whenever a remote delivery process finished, the return path value from
2972 the most recently started remote delivery process was logged.)
2974 12. RFC 3848 specifies standard names for the "with" phrase in Received: header
2975 lines when AUTH and/or TLS are in use. This is the "received protocol"
2976 field. Exim used to use "asmtp" for authenticated SMTP, without any
2977 indication (in the protocol name) for TLS use. Now it follows the RFC and
2978 uses "esmtpa" if the connection is authenticated, "esmtps" if it is
2979 encrypted, and "esmtpsa" if it is both encrypted and authenticated. These
2980 names appear in log lines as well as in Received: header lines.
2982 13. Installed MH's patches for Sieve to add the "copy" and "vacation"
2983 extensions, and comparison tests, and to fix some bugs.
2985 14. Changes to the "personal" filter test:
2987 (1) The test was buggy in that it was just doing the equivalent of
2988 "contains" tests on header lines. For example, if a user's address was
2989 anne@some.where, the "personal" test would incorrectly be true for
2991 To: susanne@some.where
2993 This test is now done by extracting each address from the header in turn,
2994 and checking the entire address. Other tests that are part of "personal"
2995 are now done using regular expressions (for example, to check local parts
2996 of addresses in From: header lines).
2998 (2) The list of non-personal local parts in From: addresses has been
2999 extended to include "listserv", "majordomo", "*-request", and "owner-*",
3000 taken from the Sieve specification recommendations.
3002 (3) If the message contains any header line starting with "List-" it is
3003 treated as non-personal.
3005 (4) The test for "circular" in the Subject: header line has been removed
3006 because it now seems ill-conceived.
3008 15. Minor typos in src/EDITME comments corrected.
3010 16. Installed latest exipick from John Jetmore.
3012 17. If headers_add on a router specified a text string that was too long for
3013 string_sprintf() - that is, longer than 8192 bytes - Exim panicked. The use
3014 of string_sprintf() is now avoided.
3016 18. $message_body_size was not set (it was always zero) when running the DATA
3017 ACL and the local_scan() function.
3019 19. For the "mail" command in an Exim filter, no default was being set for
3020 the once_repeat time, causing a random time value to be used if "once" was
3021 specified. (If the value happened to be <= 0, no repeat happened.) The
3022 default is now 0s, meaning "never repeat". The "vacation" command was OK
3023 (its default is 7d). It's somewhat surprising nobody ever noticed this bug
3024 (I found it when inspecting the code).
3026 20. There is now an overall timeout for performing a callout verification. It
3027 defaults to 4 times the callout timeout, which applies to individual SMTP
3028 commands during the callout. The overall timeout applies when there is more
3029 than one host that can be tried. The timeout is checked before trying the
3030 next host. This prevents very long delays if there are a large number of
3031 hosts and all are timing out (e.g. when the network connections are timing
3032 out). The value of the overall timeout can be changed by specifying an
3033 additional sub-option for "callout", called "maxwait". For example:
3035 verify = sender/callout=5s,maxwait=20s
3037 21. Add O_APPEND to the open() call for maildirsize files (Exim already seeks
3038 to the end before writing, but this should make it even safer).
3040 22. Exim was forgetting that it had advertised PIPELINING for the second and
3041 subsequent messages on an SMTP connection. It was also not resetting its
3042 memory on STARTTLS and an internal HELO.
3044 23. When Exim logs an SMTP synchronization error within a session, it now
3045 records whether PIPELINING has been advertised or not.
3047 24. Added 3 instances of "(long int)" casts to time_t variables that were being
3048 formatted using %ld, because on OpenBSD (and perhaps others), time_t is int
3049 rather than long int.
3051 25. Installed the latest Cygwin configuration files from the Cygwin maintainer.
3053 26. Added the never_mail option to autoreply.
3059 1. A reorganization of the code in order to implement 4.40/8 caused a daemon
3060 crash if the getsockname() call failed; this can happen if a connection is
3061 closed very soon after it is established. The problem was simply in the
3062 order in which certain operations were done, causing Exim to try to write
3063 to the SMTP stream before it had set up the file descriptor. The bug has
3064 been fixed by making things happen in the correct order.
3070 1. If "drop" was used in a DATA ACL, the SMTP output buffer was not flushed
3071 before the connection was closed, thus losing the rejection response.
3073 2. Commented out the definition of SOCKLEN_T in os.h-SunOS5. It is needed for
3074 some early Solaris releases, but causes trouble in current releases where
3075 socklen_t is defined.
3077 3. When std{in,out,err} are closed, re-open them to /dev/null so that they
3080 4. Minor refactoring of os.c-Linux to avoid compiler warning when IPv6 is not
3083 5. Refactoring in expand.c to improve memory usage. Pre-allocate a block so
3084 that releasing the top of it at the end releases what was used for sub-
3085 expansions (unless the block got too big). However, discard this block if
3086 the first thing is a variable or header, so that we can use its block when
3087 it is dynamic (useful for very large $message_headers, for example).
3089 6. Lookups now cache *every* query, not just the most recent. A new, separate
3090 store pool is used for this. It can be recovered when all lookup caches are
3091 flushed. Lookups now release memory at the end of their result strings.
3092 This has involved some general refactoring of the lookup sources.
3094 7. Some code has been added to the store_xxx() functions to reduce the amount
3095 of flapping under certain conditions.
3097 8. log_incoming_interface used to affect only the <= reception log lines. Now
3098 it causes the local interface and port to be added to several more SMTP log
3099 lines, for example "SMTP connection from", and rejection lines.
3101 9. The Sieve author supplied some patches for the doc/README.SIEVE file.
3103 10. Added a conditional definition of _BSD_SOCKLEN_T to os.h-Darwin.
3105 11. If $host_data was set by virtue of a hosts lookup in an ACL, its value
3106 could be overwritten at the end of the current message (or the start of a
3107 new message if it was set in a HELO ACL). The value is now preserved for
3108 the duration of the SMTP connection.
3110 12. If a transport had a headers_rewrite setting, and a matching header line
3111 contained an unqualified address, that address was qualified, even if it
3112 did not match any rewriting rules. The underlying bug was that the values
3113 of the flags that permit the existence of unqualified sender and recipient
3114 addresses in header lines (set by {sender,recipient}_unqualified_hosts for
3115 non-local messages, and by -bnq for local messages) were not being
3116 preserved with the message after it was received.
3118 13. When Exim was logging an SMTP synchronization error, it could sometimes log
3119 "next input=" as part of the text comprising the host identity instead of
3120 the correct text. The code was using the same buffer for two different
3121 strings. However, depending on which order the printing function evaluated
3122 its arguments, the bug did not always show up. Under Linux, for example, my
3123 test suite worked just fine.
3125 14. Exigrep contained a use of Perl's "our" scoping after change 4.31/70. This
3126 doesn't work with some older versions of Perl. It has been changed to "my",
3127 which in any case is probably the better facility to use.
3129 15. A really picky compiler found some instances of statements for creating
3130 error messages that either had too many or two few arguments for the format
3133 16. The size of the buffer for calls to the DNS resolver has been increased
3134 from 1024 to 2048. A larger buffer is needed when performing PTR lookups
3135 for addresses that have a lot of PTR records. This alleviates a problem; it
3136 does not fully solve it.
3138 17. A dnsdb lookup for PTR records that receives more data than will fit in the
3139 buffer now truncates the list and logs the incident, which is the same
3140 action as happens when Exim is looking up a host name and its aliases.
3141 Previously in this situation something unpredictable would happen;
3142 sometimes it was "internal error: store_reset failed".
3144 18. If a server dropped the connection unexpectedly when an Exim client was
3145 using GnuTLS and trying to read a response, the client delivery process
3146 crashed while trying to generate an error log message.
3148 19. If a "warn" verb in an ACL added multiple headers to a message in a single
3149 string, for example:
3151 warn message = H1: something\nH2: something
3153 the text was added as a single header line from Exim's point of view
3154 though it ended up OK in the delivered message. However, searching for the
3155 second and subsequent header lines using $h_h2: did not work. This has been
3156 fixed. Similarly, if a system filter added multiple headers in this way,
3157 the routers could not see them.
3159 20. Expanded the error message when iplsearch is called with an invalid key to
3160 suggest using net-iplsearch in a host list.
3162 21. When running tests using -bh, any delays imposed by "delay" modifiers in
3163 ACLs are no longer actually imposed (and a message to that effect is
3166 22. If a "gecos" field in a passwd entry contained escaped characters, in
3167 particular, if it contained a \" sequence, Exim got it wrong when building
3168 a From: or a Sender: header from that name. A second bug also caused
3169 incorrect handling when an unquoted " was present following a character
3170 that needed quoting.
3172 23. "{crypt}" as a password encryption mechanism for a "crypteq" expansion item
3173 was not being matched caselessly.
3175 24. Arranged for all hyphens in the exim.8 source to be escaped with
3178 25. Change 16 of 4.32, which reversed 71 or 4.31 didn't quite do the job
3179 properly. Recipient callout cache records were still being keyed to include
3180 the sender, even when use_sender was set false. This led to far more
3181 callouts that were necessary. The sender is no longer included in the key
3182 when use_sender is false.
3184 26. Added "control = submission" modifier to ACLs.
3186 27. Added the ${base62d: operator to decode base 62 numbers.
3188 28. dnsdb lookups can now access SRV records.
3190 29. CONFIGURE_OWNER can be set at build time to define an alternative owner for
3191 the configuration file.
3193 30. The debug message "delivering xxxxxx-xxxxxx-xx" is now output in verbose
3194 (-v) mode. This makes the output for a verbose queue run more intelligible.
3196 31. Added a use_postmaster feature to recipient callouts.
3198 32. Added the $body_zerocount variable, containing the number of binary zero
3199 bytes in the message body.
3201 33. The time of last modification of the "new" subdirectory is now used as the
3202 "mailbox time last read" when there is a quota error for a maildir
3205 34. Added string comparison operators lt, lti, le, lei, gt, gti, ge, gei.
3207 35. Added +ignore_unknown as a special item in host lists.
3209 36. Code for decoding IPv6 addresses in host lists is now included, even if
3210 IPv6 support is not being compiled. This fixes a bug in which an IPv6
3211 address was recognized as an IP address, but was then not correctly decoded
3212 into binary, causing unexpected and incorrect effects when compared with
3219 1. Very minor rewording of debugging text in manualroute to say "list of
3220 hosts" instead of "hostlist".
3222 2. If verify=header_syntax was set, and a header line with an unqualified
3223 address (no domain) and a large number of spaces between the end of the
3224 name and the colon was received, the reception process suffered a buffer
3225 overflow, and (when I tested it) crashed. This was caused by some obsolete
3226 code that should have been removed. The fix is to remove it!
3228 3. When running in the test harness, delay a bit after writing a bounce
3229 message to get a bit more predictability in the log output.
3231 4. Added a call to search_tidyup() just before forking a reception process. In
3232 theory, someone could use a lookup in the expansion of smtp_accept_max_
3233 per_host which, without the tidyup, could leave open a database connection.
3235 5. Added the variables $recipient_data and $sender_data which get set from a
3236 lookup success in an ACL "recipients" or "senders" condition, or a router
3237 "senders" option, similar to $domain_data and $local_part_data.
3239 6. Moved the writing of debug_print from before to after the "senders" test
3242 7. Change 4.31/66 (moving the time when the Received: is generated) caused
3243 problems for message scanning, either using a data ACL, or using
3244 local_scan() because the Received: header was not generated till after they
3245 were called (in order to set the time as the time of reception completion).
3246 I have revised the way this works. The header is now generated after the
3247 body is received, but before the ACL or local_scan() are called. After they
3248 are run, the timestamp in the header is updated.
3254 1. Change 4.24/6 introduced a bug because the SIGALRM handler was disabled
3255 before starting a queue runner without re-exec. This happened only when
3256 deliver_drop_privilege was set or when the Exim user was set to root. The
3257 effect of the bug was that timeouts during subsequent deliveries caused
3258 crashes instead of being properly handled. The handler is now left at its
3259 default (and expected) setting.
3261 2. The other case in which a daemon avoids a re-exec is to deliver an incoming
3262 message, again when deliver_drop_privilege is set or Exim is run as root.
3263 The bug described in (1) was not present in this case, but the tidying up
3264 of the other signals was missing. I have made the two cases consistent.
3266 3. The ignore_target_hosts setting on a manualroute router was being ignored
3267 for hosts that were looked up using the /MX notation.
3269 4. Added /ignore=<ip list> feature to @mx_any, @mx_primary, and @mx_secondary
3272 5. Change 4.31/55 was buggy, and broke when there was a rewriting rule that
3273 operated on the sender address. After changing the $sender_address to <>
3274 for the sender address verify, Exim was re-instated it as the original
3275 (before rewriting) address, but remembering that it had rewritten it, so it
3276 wasn't rewriting it again. This bug also had the effect of breaking the
3277 sender address verification caching when the sender address was rewritten.
3279 6. The ignore_target_hosts option was being ignored by the ipliteral router.
3280 This has been changed so that if the ip literal address matches
3281 ignore_target_hosts, the router declines.
3283 7. Added expansion conditions match_domain, match_address, and match_local_
3284 part (NOT match_host).
3286 8. The placeholder for the Received: header didn't have a length field set.
3288 9. Added code to Exim itself and to exim_lock to test for a specific race
3289 condition that could lead to file corruption when using MBX delivery. The
3290 issue is with the lockfile that is created in /tmp. If this file is removed
3291 after a process has opened it but before that process has acquired a lock,
3292 there is the potential for a second process to recreate the file and also
3293 acquire a lock. This could lead to two Exim processes writing to the file
3294 at the same time. The added code performs the same test as UW imapd; it
3295 checks after acquiring the lock that its file descriptor still refers to
3296 the same named file.
3298 10. The buffer for building added header lines was of fixed size, 8192 bytes.
3299 It is now parameterized by HEADER_ADD_BUFFER_SIZE and this can be adjusted
3302 11. Added the smtp_active_hostname option. If used, this will typically be made
3303 to depend on the incoming interface address. Because $interface_address is
3304 not set up until the daemon has forked a reception process, error responses
3305 that can happen earlier (such as "too many connections") no longer contain
3308 12. If an expansion in a condition on a "warn" statement fails because a lookup
3309 defers, the "warn" statement is abandoned, and the next ACL statement is
3310 processed. Previously this caused the whole ACL to be aborted.
3312 13. Added the iplsearch lookup type.
3314 14. Added ident_timeout as a log selector.
3316 15. Added tls_certificate_verified as a log selector.
3318 16. Added a global option tls_require_ciphers (compare the smtp transport
3319 option of the same name). This controls incoming TLS connections.
3321 17. I finally figured out how to make tls_require_ciphers do a similar thing
3322 in GNUtls to what it does in OpenSSL, that is, set up an appropriate list
3323 before starting the TLS session.
3325 18. Tabs are now shown as \t in -bP output.
3327 19. If the log selector return_path_on_delivery was set, Exim crashed when
3328 bouncing a message because it had too many Received: header lines.
3330 20. If two routers both had headers_remove settings, and the first one included
3331 a superfluous trailing colon, the final name in the first list and the
3332 first name in the second list were incorrectly joined into one item (with a
3333 colon in the middle).
3339 1. Added -C and -D options to the exinext utility, mainly to make it easier
3340 to include in the automated testing, but these could be helpful when
3341 multiple configurations are in use.
3343 2. The exinext utility was not formatting the output nicely when there was
3344 an alternate port involved in the retry record key, nor when there was a
3345 message id as well (for retries that were specific to a specific message
3346 and a specific host). It was also confused by IPv6 addresses, because of
3347 the additional colons they contain. I have fixed the IPv4 problem, and
3348 patched it up to do a reasonable job for IPv6.
3350 3. When there is an error after a MAIL, RCPT, or DATA SMTP command during
3351 delivery, the log line now contains "pipelined" if PIPELINING was used.
3353 4. An SMTP transport process used to panic and die if the bind() call to set
3354 an explicit outgoing interface failed. This has been changed; it is now
3355 treated in the same way as a connect() failure.
3357 5. A reference to $sender_host_name in the part of a conditional expansion
3358 that was being skipped was still causing a DNS lookup. This no longer
3361 6. The def: expansion condition was not recognizing references to header lines
3362 that used bh_ and bheader_.
3364 7. Added the _cache feature to named lists.
3366 8. The code for checking quota_filecount in the appendfile transport was
3367 allowing one more file than it should have been.
3369 9. For compatibility with Sendmail, the command line option
3377 and sets the incoming protocol and host name (for trusted callers). The
3378 host name and its colon can be omitted when only the protocol is to be set.
3379 Note the Exim already has two private options, -pd and -ps, that refer to
3380 embedded Perl. It is therefore impossible to set a protocol value of "d" or
3381 "s", but I don't think that's a major issue.
3383 10. A number of refactoring changes to the code, none of which should affect
3386 (a) The number of logging options was getting close to filling up the
3387 32-bit word that was used as a bit map. I have split them into two classes:
3388 those that are passed in the argument to log_write(), and those that are
3389 only ever tested independently outside of that function. These are now in
3390 separate 32-bit words, so there is plenty of room for expansion again.
3391 There is no change in the user interface or the logging behaviour.
3393 (b) When building, for example, log lines, the code previously used a
3394 macro that called string_cat() twice, in order to add two strings. This is
3395 not really sufficiently general. Furthermore, there was one instance where
3396 it was actually wrong because one of the argument was used twice, and in
3397 one call a function was used. (As it happened, calling the function twice
3398 did not affect the overall behaviour.) The macro has been replaced by a
3399 function that can join an arbitrary number of extra strings onto a growing
3402 (c) The code for expansion conditions now uses a table and a binary chop
3403 instead of a serial search (which was left over from when there were very
3404 few conditions). Also, it now recognizes conditions like "pam" even when
3405 the relevant support is not compiled in: a suitably worded error message is
3406 given if an attempt is made to use such a condition.
3408 11. Added ${time_interval:xxxxx}.
3410 12. A bug was causing one of the ddress fields not to be passed back correctly
3411 from remote delivery subprocesses. The field in question was not being
3412 subsequently used, so this caused to problems in practice.
3414 13. Added new log selectors queue_time and deliver_time.
3416 14. Might have fixed a bug in maildirsizefile handling that threw up
3417 "unexpected character" debug warnings, and recalculated the data
3418 unnecessarily. In any case, I expanded the warning message to give more
3421 15. Added the message "Restricted characters in address" to the statements in
3422 the default ACL that block characters like @ and % in local parts.
3424 16. Change 71 for release 4.31 proved to be much less benign that I imagined.
3425 Three changes have been made:
3427 (a) There was a serious bug; a negative response to MAIL caused the whole
3428 recipient domain to be cached as invalid, thereby blocking all messages
3429 to all local parts at the same domain, from all senders. This bug has
3430 been fixed. The domain is no longer cached after a negative response to
3431 MAIL if the sender used is not empty.
3433 (b) The default behaviour of using MAIL FROM:<> for recipient callouts has
3436 (c) A new callout option, "use_sender" has been added for people who want
3437 the modified behaviour.
3443 1. Removed "EXTRALIBS=-lwrap" from OS/Makefile-Unixware7 on the advice of
3446 2. Removed "LIBS = -lresolv" from OS/Makefile-Darwin as it is not needed, and
3447 indeed breaks things for older releases.
3449 3. Added additional logging to the case where there is a problem reading data
3450 from a filter that is running in a subprocess using a pipe, in order to
3451 try to track down a specific problem.
3453 4. Testing facility fudge: when running in the test harness and attempting
3454 to connect to 10.x.x.x (expecting a connection timeout) I'm now sometimes
3455 getting "No route to host". Convert this to a timeout.
3457 5. Define ICONV_ARG2_TYPE as "char **" for Unixware7 to avoid compiler
3460 6. Some OS don't have socklen_t but use size_t instead. This affects the
3461 fifth argument of getsockopt() amongst other things. This is now
3462 configurable by a macro called SOCKLEN_T which defaults to socklen_t, but
3463 can be set for individual OS. I have set it for SunOS5, OSF1, and
3464 Unixware7. Current versions of SunOS5 (aka Solaris) do have socklen_t, but
3465 some earlier ones do not.
3467 7. Change 4.30/15 was not doing the test caselessly.
3469 8. The standard form for an IPv6 address literal was being rejected by address
3470 parsing in, for example, MAIL and RCPT commands. An example of this kind of
3471 address is [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3472 this, as well as the form without the "IPv6" on the front (but only when
3473 address literals are enabled, of course).
3475 9. Added some casts to avoid compiler warnings in OS/os.c-Linux.
3477 10. Exim crashed if a message with an empty sender address specified by -f
3478 encountered a router with an errors_to setting. This could be provoked only
3479 by a command such as
3483 where an empty string was supplied; "<>" did not hit this bug.
3485 11. Installed PCRE release 4.5.
3487 12. If EHLO/HELO was rejected by an ACL, the value of $sender_helo_name
3488 remained set. It is now erased.
3490 13. exiqgrep wasn't working on MacOS X because it didn't correctly compute
3491 times from message ids (which are base 36 rather than the normal 62).
3493 14. "Expected" SMTP protocol errors that can arise when PIPELINING is in use
3494 were being counted as actual protocol errors, and logged if the log
3495 selector +smtp_protocol_error was set. One cannot be perfect in this test,
3496 but now, if PIPELINING has been advertised, RCPT following a rejected MAIL,
3497 and DATA following a set of rejected RCPTs do not count as protocol errors.
3498 In other words, Exim assumes they were pipelined, though this may not
3499 actually be the case. Of course, in all cases the client gets an
3500 appropriate error code.
3502 15. If a lookup fails in an ACL condition, a message about the failure may
3503 be available; it is used if testing the ACL cannot continue, because most
3504 such messages specify what the cause of the deferral is. However, some
3505 messages (e.g. "MYSQL: no data found") do not cause a defer. There was bug
3506 that caused an old message to be retained and used if a later statement
3507 caused a defer, replacing the real cause of the deferral.
3509 16. If an IP address had so many PTR records that the DNS lookup buffer
3510 was not large enough to hold them, Exim could crash while trying to process
3511 the truncated data. It now detects and logs this case.
3513 17. Further to 4.21/58, another change has been made: if (and only if) the
3514 first line of a message (the first header line) ends with CRLF, a bare LF
3515 in a subsequent header line has a space inserted after it, so as not to
3516 terminate the header.
3518 18. Refactoring: tidied an ugly bit of code in appendfile that copied data
3519 unnecessarily, used atoi() instead of strtol(), and didn't check the
3520 termination when getting file sizes from file names by regex.
3522 19. Completely re-implemented the support for maildirsize files, in the light
3523 of a number of problems with the previous contributed implementation
3524 (4.30/29). In particular:
3526 . If the quota is zero, the maildirsize file is maintained, but no quota is
3529 . If the maildir directory does not exist, it is created before any attempt
3530 to write a maildirsize file.
3532 . The quota value in the file is just a cache; if the quota is changed in
3533 the transport, the new value overrides.
3535 . A regular expression is available for excluding directories from the
3538 20. The autoreply transport checks the characters in options that define the
3539 message's headers; it allows continued headers, but it was checking with
3540 isspace() after an embedded newline instead of explicitly looking for a
3543 21. If all the "regular" hosts to which an address was routed had passed their
3544 expiry times, and had not reached their retry times, the address was
3545 bounced, even if fallback hosts were defined. Now Exim should go on to try
3548 22. Increased buffer sizes in the callout code from 1024 to 4096 to match the
3549 equivalent code in the SMTP transport. Some hosts send humungous responses
3550 to HELO/EHLO, more than 1024 it seems.
3552 23. Refactoring: code in filter.c used (void *) for "any old type" but this
3553 gives compiler warnings in some environments. I've now done it "properly",
3556 24. The replacement for inet_ntoa() that is used with gcc on IRIX systems
3557 (because of problems with the built-in one) was declared to return uschar *
3558 instead of char *, causing compiler failure.
3560 25. Fixed a file descriptor leak when processing alias/forward files.
3562 26. Fixed a minor format string issue in dbfn.c.
3564 27. Typo in exim.c: ("dmbnz" for "dbmnz").
3566 28. If a filter file refered to $h_xxx or $message_headers, and the headers
3567 contained RFC 2047 "words", Exim's memory could, under certain conditions,
3570 29. When a sender address is verified, it is cached, to save repeating the test
3571 when there is more than one recipient in a message. However, when the
3572 verification involves a callout, it is possible for different callout
3573 options to be set for different recipients. It is too complicated to keep
3574 track of this in the cache, so now Exim always runs a verification when a
3575 callout is required, relying on the callout cache for the optimization.
3576 The overhead is duplication of the address routing, but this should not be
3579 30. Fixed a bug in callout caching. If a RCPT command caused the sender address
3580 to be verified with callout=postmaster, and the main callout worked but the
3581 postmaster check failed, the verification correctly failed. However, if a
3582 subsequent RCPT command asked for sender verification *without* the
3583 postmaster check, incorrect caching caused this verification also to fail,
3586 31. Exim caches DNS lookup failures so as to avoid multiple timeouts; however,
3587 it was not caching the DNS options (qualify_single, search_parents) that
3588 were used when the lookup failed. A subsequent lookup with different
3589 options therefore always gave the same answer, though there were cases
3590 where it should not have. (Example: a "domains = !$mx_any" option on a
3591 dnslookup router: the "domains" option is always processed without any
3592 widening, but the router might have qualify_single set.) Now Exim uses the
3593 cached value only when the same options are set.
3595 32. Added John Jetmore's "exipick" utility to the distribution.
3597 33. GnuTLS: When an attempt to start a TLS session fails for any reason other
3598 than a timeout (e.g. a certificate is required, and is not provided), an
3599 Exim server now closes the connection immediately. Previously it waited for
3600 the client to close - but if the client is SSL, it seems that they each
3601 wait for each other, leading to a delay before one of them times out.
3603 34: GnuTLS: Updated the code to use the new GnuTLS 1.0.0 API. I have not
3604 maintained 0.8.x compatibility because I don't think many are using it, and
3605 it is clearly obsolete.
3607 35. Added TLS support for CRLs: a tls_crl global option and one for the smtp
3610 36. OpenSSL: $tls_certificate_verified was being set to 1 even if the
3611 client certificate was expired. A simple patch fixes this, though I don't
3612 understand the full logic of why the verify callback is called multiple
3615 37. OpenSSL: a patch from Robert Roselius: "Enable client-bug workaround.
3616 Versions of OpenSSL as of 0.9.6d include a 'CBC countermeasure' feature,
3617 which causes problems with some clients (such as the Certicom SSL Plus
3618 library used by Eudora). This option, SSL_OP_DONT_INSERT_EMPTY_FRAGMENTS,
3619 disables the coutermeasure allowing Eudora to connect."
3621 38. Exim was not checking that a write() to a log file succeeded. This could
3622 lead to Bad Things if a log got too big, in particular if it hit a file
3623 size limit. Exim now panics and dies if it cannot write to a log file, just
3624 as it does if it cannot open a log file.
3626 39. Modified OS/Makefile-Linux so that it now contains
3628 CFLAGS=-O -D_FILE_OFFSET_BITS=64 -D_LARGEFILE_SOURCE
3630 The two -D definitions ensure that Exim is compiled with large file
3631 support, which makes it possible to handle log files that are bigger than
3634 40. Fixed a subtle caching bug: if (in an ACL or a set of routers, for
3635 instance) a domain was checked against a named list that involved a lookup,
3636 causing $domain_data to be set, then another domain was checked against the
3637 same list, then the first domain was re-checked, the value of $domain_data
3638 after the final check could be wrong. In particular, if the second check
3639 failed, it could be set empty. This bug probably also applied to
3642 41. The strip_trailing_dot option was not being applied to the address given
3643 with the -f command-line option.
3645 42. The code for reading a message's header from the spool was incrementing
3646 $received_count, but never initializing it. This meant that the value was
3647 incorrect (doubled) while delivering a message in the same process in which
3648 it was received. In the most common configuration of Exim, this never
3649 happens - a fresh exec is done - but it can happen when
3650 deliver_drop_privilege is set.
3652 43. When Exim logs an SMTP synchronization error - client data sent too soon -
3653 it now includes up to 150 characters of the unexpected data in the log
3656 44. The exim_dbmbuild utility uses fixed size buffers for reading input lines
3657 and building data strings. The size of both of these buffers was 10 000
3658 bytes - far larger than anybody would *ever* want, thought I. Needless to
3659 say, somebody hit the limit. I have increased the maximum line length to
3660 20 000 and the maximum data length of concatenated lines to 100 000. I have
3661 also fixed two bugs, because there was no checking on these buffers. Tsk,
3662 tsk. Now exim_dbmbuild gives a message and exits with an error code if a
3663 buffer is too small.
3665 45. The exim_dbmbuild utility did not support quoted keys, as Exim does in
3666 lsearch lookups. Now it does.
3668 46. When parsing a route_list item in a manualroute router, a fixed-length
3669 buffer was used for the list of hosts. I made this 1024 bytes long,
3670 thinking that nobody would ever have a list of hosts that long. Wrong.
3671 Somebody had a whole pile of complicated expansion conditions, and the
3672 string was silently truncated, leading to an expansion error. It turns out
3673 that it is easier to change to an unlimited length (owing to other changes
3674 that have happened since this code was originally written) than to build
3675 structure for giving a limitation error. The length of the item that
3676 expands into the list of hosts is now unlimited.
3678 47. The lsearch lookup could not handle data where the length of text line was
3679 more than 4095 characters. Such lines were truncated, leading to shortened
3680 data being returned. It should now handle lines of any length.
3682 48. Minor wording revision: "cannot test xxx in yyy ACL" becomes "cannot test
3683 xxx condition in yyy ACL" (e.g. "cannot test domains condition in DATA
3686 49. Cosmetic tidy to scripts like exicyclog that are generated by globally
3687 replacing strings such as BIN_DIRECTORY in a source file: the replacement
3688 no longer happens in comment lines. A list of replacements is now placed
3689 at the head of all of the source files, except those whose only change is
3690 to replace PERL_COMMAND in the very first #! line.
3692 50. Replaced the slow insertion sort in queue.c, for sorting the list of
3693 messages on the queue, with a bottom-up merge sort, using code contributed
3694 by Michael Haardt. This should make operations like -bp somewhat faster on
3695 large queues. It won't affect queue runners, except when queue_run_in_order
3698 51. Installed eximstats 1.31 in the distribution.
3700 52. Added support for SRV lookups to the dnslookup router.
3702 53. If an ACL referred to $message_body or $message_body_end, the value was not
3703 reset for any messages that followed in the same SMTP session.
3705 54. The store-handling optimization for building very long strings was not
3706 differentiating between the different store pools. I don't think this
3707 actually made any difference in practice, but I've tidied it.
3709 55. While running the routers to verify a sender address, $sender_address
3710 was still set to the sender address. This is wrong, because when routing to
3711 send a bounce to the sender, it would be empty. Therefore, I have changed
3712 it so that, while verifying a sender address, $sender_address is set to <>.
3713 (There is no change to what happens when verifying a recipient address.)
3715 56. After finding MX (or SRV) records, Exim was doing a DNS lookup for the
3716 target A or AAAA records (if not already returned) without resetting the
3717 qualify_single or search_parents options of the DNS resolver. These are
3718 inappropriate in this case because the targets of MX and SRV records must
3719 be FQDNs. A broken DNS record could cause trouble if it happened to have a
3720 target that, when qualified, matched something in the local domain. These
3721 two options are now turned off when doing these lookups.
3723 57. It seems that at least some releases of Reiserfs (which does not have the
3724 concept of a fixed number of inodes) returns zero and not -1 for the
3725 number of available inodes. This interacted badly with check_spool_inodes,
3726 which assumed that -1 was the "no such thing" setting. What I have done is
3727 to check that the total number of inodes is greater than zero before doing
3728 the test of how many are available.
3730 58. When a "warn" ACL statement has a log_message modifier, the message is
3731 remembered, and not repeated. This is to avoid a lot of repetition when a
3732 message has many recipients that cause the same warning to be written.
3733 Howewer, Exim was preserving the list of already written lines for an
3734 entire SMTP session, which doesn't seem right. The memory is now reset if a
3735 new message is started.
3737 59. The "rewrite" debugging flag was not showing the result of rewriting in the
3738 debugging output unless log_rewrite was also set.
3740 60. Avoid a compiler warning on 64-bit systems in dsearch.c by avoiding the use
3741 of (int)(handle) when we know that handle contains (void *)(-1).
3743 61. The Exim daemon panic-logs an error return when it closes the incoming
3744 connection. However "connection reset by peer" seems to be common, and
3745 isn't really an error worthy of noting specially, so that particular error
3748 62. When Exim is trying to find all the local interfaces, it used to panic and
3749 die if the ioctl to get the interface flags failed. However, it seems that
3750 on at least one OS (Solaris 9) it is possible to have an interface that is
3751 included in the list of interfaces, but for which you get a failure error
3752 for this call. This happens when the interface is not "plumbed" into a
3753 protocol (i.e. neither IPv4 nor IPv6). I've changed the code so that a
3754 failure of the "get flags" call assumes that the interface is down.
3756 63. Added a ${eval10: operator, which assumes all numbers are decimal. This
3757 makes life easier for people who are doing arithmetic on fields extracted
3758 from dates, where you often get leading zeros that should not be
3759 interpreted as octal.
3761 64. Added qualify_domain to the redirect router, to override the global
3764 65. If a pathologically long header line contained very many addresses (the
3765 report of this problem mentioned 10 000) and each of them was rewritten,
3766 Exim could use up a very large amount of memory. (It kept on making new
3767 copies of the header line as it rewrote, and never released the old ones.)
3768 At the expense of a bit more processing, the header rewriting function has
3769 been changed so that it no longer eats memory in this way.
3771 66. The generation of the Received: header has been moved from the time that a
3772 message starts to be received, to the time that it finishes. The timestamp
3773 in the Received: header should now be very close to that of the <= log
3774 line. There are two side-effects of this change:
3776 (a) If a message is rejected by a DATA or non-SMTP ACL or local_scan(), the
3777 logged header lines no longer include the local Received: line, because
3778 it has not yet been created. The same applies to a copy of the message
3779 that is returned to a non-SMTP sender when a message is rejected.
3781 (b) When a filter file is tested using -bf, no additional Received: header
3782 is added to the test message. After some thought, I decided that this
3785 This change does not affect the value of $received_for. It is still set
3786 after address rewriting, but before local_scan() is called.
3788 67. Installed the latest Cygwin-specific files from the Cygwin maintainer.
3790 68. GnuTLS: If an empty file is specified for tls_verify_certificates, GnuTLS
3791 gave an unhelpful panic error message, and a defer error. I have managed to
3792 change this behaviour so that it now rejects any supplied certificate,
3793 which seems right, as the list of acceptable certificates is empty.
3795 69. OpenSSL: If an empty file is specified for tls_verify_certificates, OpenSSL
3796 gave an unhelpful defer error. I have not managed to make this reject any
3797 supplied certificates, but the error message it gives is "no certificate
3798 supplied", which is not helpful.
3800 70. exigrep's output now also includes lines that are not associated with any
3801 message, but which match the given pattern. Implemented by a patch from
3802 Martin Sluka, which also tidied up the Perl a bit.
3804 71. Recipient callout verification, like sender verification, was using <> in
3805 the MAIL FROM command. This isn't really the right thing, since the actual
3806 sender may affect whether the remote host accepts the recipient or not. I
3807 have changed it to use the actual sender in the callout; this means that
3808 the cache record is now keyed on a recipient/sender pair, not just the
3809 recipient address. There doesn't seem to be a real danger of callout loops,
3810 since a callout by the remote host to check the sender would use <>.
3811 [SEE ABOVE: changed after hitting problems.]
3813 72. Exim treats illegal SMTP error codes that do not begin with 4 or 5 as
3814 temporary errors. However, in the case of such a code being given after
3815 the end of a data transmission (i.e. after ".") Exim was failing to write
3816 a retry record for the message. (Yes, there was some broken host that was
3817 actually sending 8xx at this point.)
3819 73. An unknown lookup type in a host list could cause Exim to panic-die when
3820 the list was checked. (An example that provoked this was putting <; in the
3821 middle of a list instead of at the start.) If this happened during a DATA
3822 ACL check, a -D file could be left lying around. This kind of configuration
3823 error no longer causes Exim to die; instead it causes a defer errror. The
3824 incident is still logged to the main and panic logs.
3826 74. Buglet left over from Exim 3 conversion. The message "too many messages
3827 in one connection" was written to the rejectlog but not the mainlog, except
3828 when address rewriting (yes!) was being logged.
3830 75. Added write_rejectlog option.
3832 76. When a system filter was run not as root (that is, when system_filter_user
3833 was set), the values of the $n variables were not being returned to the
3834 main process; thus, they were not subsequently available in the $sn
3837 77. Added +return_path_on_delivery log selector.
3839 78. A connection timeout was being treated differently from recipients deferred
3840 when testing hosts_max_try with a message that was older than the host's
3841 retry timeout. (The host should not be counted, thus allowing all hosts to
3842 be tried at least once before bouncing.) This may have been the cause of an
3843 occasionally reported bug whereby a message would remain on the queue
3844 longer than the retry timeout, but would be bounced if a delivery was
3845 forced. I say "may" because I never totally pinned down the problem;
3846 setting up timeout/retry tests is difficult. See also the next item.
3848 79. The ultimate address timeout was not being applied to errors that involved
3849 a combination of host plus message (for example, a timeout on a MAIL
3850 command). When an address resolved to a number of possible hosts, and they
3851 were not all tried for each delivery (e.g. because of hosts_max_try), a
3852 message could remain on the queue longer than the retry timeout.
3854 80. Sieve bug: "stop" inside "elsif" was broken. Applied a patch from Michael
3857 81. Fixed an obscure SMTP outgoing bug which required at least the following
3858 conditions: (a) there was another message waiting for the same server;
3859 (b) the server returned 5xx to all RCPT commands in the first message so
3860 that the message was not completed; (c) the server dropped the connection
3861 or gave a negative response to the RSET that Exim sends to abort the
3862 transaction. The observed case was a dropped connection after DATA that had
3863 been sent in pipelining mode. That is, the server had advertised PIPELINING
3864 but was not implementing it correctly. The effect of the bug was incorrect
3865 behaviour, such as trying another host, and this could lead to a crash.
3871 1. The 3rd arguments to getsockname(), getpeername(), and accept() in exim.c
3872 and daemon.c were passed as pointers to ints; they should have been
3873 pointers to socklen_t variables (which are typically unsigned ints).
3875 2. Some signed/unsigned type warnings in the os.c file for Linux have been
3878 3. Fixed a really odd bug that affected only the testing scheme; patching a
3879 certain fixed string in the binary changed the value of another string that
3880 happened to be identical to the end of the original first string.
3882 4. When gethostbyname() (or equivalent) is passed an IP address as a "host
3883 name", it returns that address as the IP address. On some operating
3884 systems (e.g. Solaris), it also passes back the IP address string as the
3885 "host name". However, on others (e.g. Linux), it passes back an empty
3886 string. Exim wasn't checking for this, and was changing the host name to an
3887 empty string, assuming it had been canonicized.
3889 5. Although rare, it is permitted to have more than one PTR record for a given
3890 IP address. I thought that gethostbyaddr() or getipnodebyaddr() always gave
3891 all the names associated with an address, because they do in Solaris.
3892 However, it seems that they do not in Linux for data that comes from the
3893 DNS. If an address in /etc/hosts has multiple names, they _are_ all given.
3894 I found this out when I moved to a new Linux workstation and tried to run
3895 the Exim test suite.
3897 To get round this problem I have changed the code so that it now does its
3898 own call to the DNS to look up PTR records when searching for a host name.
3899 If nothing can be found in the DNS, it tries gethostbyaddr(), so that
3900 addresses that are only in /etc/hosts are still found.
3902 This behaviour is, however, controlled by an option called host_lookup_
3903 order, which defaults to "bydns:byaddr". If people want to use the other
3904 order, or indeed, just use one or the other means of lookup, they can
3905 specify it in this variable.
3907 6. If a PTR record yields an empty name, Exim treats it as non-existent. In
3908 some operating systems, this comes back from gethostbyaddr() as an empty
3909 string, and this is what Exim used to test for. However, it seems that in
3910 other systems, "." is yielded. Exim now tests for this case too.
3912 7. The values of check_spool_space and check_log_space are now held internally
3913 as a number of kilobytes instead of an absolute number of bytes. If a
3914 numbers is specified without 'K' or 'M', it is rounded up to the nearest
3915 kilobyte. This means that much larger values can be stored.
3917 8. Exim monitor: an attempt to get the action menu when not actually pointing
3918 at a message produces an empty menu entitled "No message selected". This
3919 works on Solaris (OpenWindows). However, XFree86 does not like a menu with
3920 no entries in it ("Shell widget menu has zero width and/or height"). So I
3921 have added a single, blank menu entry in this case.
3923 9. Added ${quote_local_part.
3925 10. MIME decoding is now applied to the contents of Subject: header lines when
3928 11. Now that a reference to $sender_host_address automatically causes a reverse
3929 lookup to occur if necessary (4.13/18), there is no need to arrange for a
3930 host lookup before query-style lookups in lists that might use this
3931 variable. This has therefore been abolished, and the "net-" prefix is no
3932 longer necessary for query-style lookups.
3934 12. The Makefile for SCO_SV contained a setting of LDFLAGS. This appears to
3935 have been a typo for LFLAGS, so it has been changed.
3937 13. The install script calls Exim with "-C /dev/null" in order to find the
3938 version number. If ALT_CONFIG_PREFIX was set, this caused an error message
3939 to be output. Howeve, since Exim outputs its version number before the
3940 error, it didn't break the script. It just looked ugly. I fixed this by
3941 always allowing "-C /dev/null" if the caller is root.
3943 14. Ignore overlarge ACL variable number when reading spool file - insurance
3944 against a later release with more variables having written the file.
3946 15. The standard form for an IPv6 address literal was being rejected by EHLO.
3947 Example: [IPv6:2002:c1ed:8229:10:202:2dff:fe07:a42a]. Exim now accepts
3948 this, as well as the form without the "IPv6" on the front.
3950 16. Added CHOWN_COMMAND=/usr/sbin/chown and LIBS=-lresolv to the
3951 OS/Makefile-Darwin file.
3953 17. Fixed typo in lookups/ldap.c: D_LOOKUP should be D_lookup. This applied
3954 only to LDAP libraries that do not have LDAP_OPT_DEREF.
3956 18. After change 4.21/52, "%ld" was used to format the contents of the $inode
3957 variable. However, some OS use ints for inodes. I've added cast to long int
3958 to get rid of the compiler warning.
3960 19. I had forgotten to lock out "/../" in configuration file names when
3961 ALT_CONFIG_PREFIX was set.
3963 20. Routers used for verification do not need to specify transports. However,
3964 if such a router generated a host list, and callout was configured, Exim
3965 crashed, because it could not find a port number from the (non-existent)
3966 transport. It now assumes port 25 in this circumstance.
3968 21. Added the -t option to exigrep.
3970 22. If LOOKUP_LSEARCH is defined, all three linear search methods (lsearch,
3971 wildlsearch, nwildlsearch) are compiled. LOOKUP_WILDLSEARCH and LOOKUP_
3972 NWILDLSEARCH are now obsolete, but retained for compatibility. If either of
3973 them is set, LOOKUP_LSEARCH is forced.
3975 23. "exim -bV" now outputs a list of lookups that are included in the binary.
3977 24. Added sender and host information to the "rejected by local_scan()" log
3978 line; previously there was no indication of these.
3980 25. Added .include_if_exists.
3982 26. Change 3.952/11 added an explicit directory sync on top of a file sync for
3983 Linux. It turns out that not all file systems support this. Apparently some
3984 versions of NFS do not. (It's rare to put Exim's spool on NFS, but people
3985 do it.) To cope with this, the error EINVAL, which means that sync-ing is
3986 not supported on the file descriptor, is now ignored when Exim is trying to
3987 sync a directory. This applies only to Linux.
3989 27. Added -DBIND_8_COMPAT to the CLFAGS setting for Darwin.
3991 28. In Darwin (MacOS X), the PAM headers are in /usr/include/pam and not in
3992 /usr/include/security. There's now a flag in OS/os.h-Darwin to cope with
3995 29. Added support for maildirsize files from supplied patch (modified a bit).
3997 30. The use of :fail: followed by an empty string could lead Exim to respond to
3998 sender verification failures with (e.g.):
4000 550 Verification failed for <xxx>
4001 550 Sender verify failed
4003 where the first response line was missing the '-' that indicates it is not
4004 the final line of the response.
4006 31. The loop for finding the name of the user that called Exim had a hardwired
4007 limit of 10; it now uses the value of finduser_retries, which is used for
4008 all other user lookups.
4010 32. Added $received_count variable, available in data and not_smtp ACLs, and at
4013 33. Exim was neglecting to zero errno before one call of strtol() when
4014 expanding a string and expecting an integer value. On some systems this
4015 resulted in spurious "integer overflow" errors. Also, it was casting the
4016 result into an int without checking.
4018 34. Testing for a connection timeout using "timeout_connect" in the retry rules
4019 did not work. The code looks as if it has *never* worked, though it appears
4020 to have been documented since at least releast 1.62. I have made it work.
4022 35. The "timeout_DNS" error in retry rules, also documented since at least
4023 1.62, also never worked. As it isn't clear exactly what this means, and
4024 clearly it isn't a major issue, I have abolished the feature by treating it
4025 as "timeout", and writing a warning to the main and panic logs.
4027 36. The display of retry rules for -brt wasn't always showing the error code
4030 37. Added new error conditions to retry rules: timeout_A, timeout_MX,
4031 timeout_connect_A, timeout_connect_MX.
4033 38. Rewriting the envelope sender at SMTP time did not allow it to be rewritten
4034 to the empty sender.
4036 39. The daemon was not analysing the content of -oX till after it had closed
4037 stderr and disconnected from the controlling terminal. This meant that any
4038 syntax errors were only noted on the panic log, and the return code from
4039 the command was 0. By re-arranging the code a little, I've made the
4040 decoding happen first, so such errors now appear on stderr, and the return
4041 code is 1. However, the actual setting up of the sockets still happens in
4042 the disconnected process, so errors there are still only recorded on the
4045 40. A daemon listener on a wildcard IPv6 socket that also accepts IPv4
4046 connections (as happens on some IP stacks) was logged at start up time as
4047 just listening for IPv6. It now logs "IPv6 with IPv4". This differentiates
4048 it from "IPv6 and IPv4", which means that two separate sockets are being
4051 41. The debug output for gethostbyname2() or getipnodebyname() failures now
4052 says whether AF_INET or AF_INET6 was passed as an argument.
4054 42. Exiwhat output was messed up when time zones were included in log
4057 43. Exiwhat now gives more information about the daemon's listening ports,
4058 and whether -tls-on-connect was used.
4060 44. The "port" option of the smtp transport is now expanded.
4062 45. A "message" modifier in a "warn" statement in a non-message ACL was being
4063 silently ignored. Now an error message is written to the main and panic
4066 46. There's a new ACL modifier called "logwrite" which writes to a log file
4067 as soon as it is encountered.
4069 47. Added $local_user_uid and $local_user_gid at routing time.
4071 48. Exim crashed when trying to verify a sender address that was being
4074 49. Exim was recognizing only a space character after ".include". It now also
4075 recognizes a tab character.
4077 50. Fixed several bugs in the Perl script that creates the exim.8 man page by
4078 extracting the relevant information from the specification. The man page no
4079 longer contains scrambled data for the -d option, and I've added a section
4080 at the front about calling Exim under different names.
4082 51. Added "extra_headers" argument to the "mail" command in filter files.
4084 52. Redirecting mail to an unqualified address in a Sieve filter caused Exim to
4087 53. Installed eximstats 1.29.
4089 54. Added transport_filter_timeout as a generic transport option.
4091 55. Exim no longer adds an empty Bcc: header to messages that have no To: or
4092 Cc: header lines. This was required by RFC 822, but it not required by RFC
4095 56. Exim used to add From:, Date:, and Message-Id: header lines to any
4096 incoming messages that did not have them. Now it does so only if the
4097 message originates locally, that is, if there is no associated remote host
4098 address. When Resent- header lines are present, this applies to the Resent-
4099 lines rather than the non-Resent- lines.
4101 57. Drop incoming SMTP connection after too many syntax or protocol errors. The
4102 limit is controlled by smtp_max_synprot_errors, defaulting to 3.
4104 58. Messages for configuration errors now include the name of the main
4105 configuration file - useful now that there may be more than one file in a
4106 list (.included file names were always shown).
4108 59. Change 4.21/82 (run initgroups() when starting the daemon) causes problems
4109 for those rare installations that do not start the daemon as root or run it
4110 setuid root. I've cut out the call to initgroups() if the daemon is not
4113 60. The Exim user and group can now be bound into the binary as text strings
4114 that are looked up at the start of Exim's processing.
4116 61. Applied a small patch for the Interbase code, supplied by Ard Biesheuvel.
4118 62. Added $mailstore_basename variable.
4120 63. Installed patch to sieve.c from Michael Haardt.
4122 64. When Exim failed to open the panic log after failing to open the main log,
4123 the original message it was trying to log was written to stderr and debug
4124 output, but if they were not available (the usual case in production), it
4125 was lost. Now it is written to syslog before the two lines that record the
4126 failures to open the logs.
4128 65. Users' Exim filters run in subprocesses under the user's uid. It is
4129 possible for a "deliver" command or an alias in a "personal" command to
4130 provoke an address rewrite. If logging of address rewriting is configured,
4131 this fails because the process is not running as root or exim. There may be
4132 a better way of dealing with this, but for the moment (because 4.30 needs
4133 to be released), I have disabled address rewrite logging when running a
4134 filter in a non-root, non-exim process.
4140 1. The buildconfig auxiliary program wasn't quoting the value set for
4141 HEADERS_CHARSET. This caused a compilation error complaining that 'ISO' was
4142 not defined. This bug was masked in 4.22 by the effect that was fixed in
4145 2. Some messages that were rejected after a message id was allocated were
4146 shown as "incomplete" by exigrep. It no longer does this for messages that
4147 are rejected by local_scan() or the DATA or non-SMTP ACLs.
4149 3. If a Message-ID: header used a domain literal in the ID, and Exim did not
4150 have allow_domain_literals set, the ID did not get logged in the <= line.
4151 Domain literals are now always recognized in Message-ID: header lines.
4153 4. The first argument for a ${extract expansion item is the key name or field
4154 number. Leading and trailing spaces in this item were not being ignored,
4155 causing some misleading effects.
4157 5. When deliver_drop_privilege was set, single queue runner processes started
4158 manually (i.e. by the command "exim -q") or by the daemon (which uses the
4159 same command in the process it spins off) were not dropping privilege.
4161 6. When the daemon running as "exim" started a queue runner, it always
4162 re-executed Exim in the spun-off process. This is a waste of effort when
4163 deliver_drop_privilege is set. The new process now just calls the
4164 queue-runner function directly.
4170 1. Typo in the src/EDITME file: it referred to HEADERS_DECODE_TO instead of
4173 2. Change 4.21/73 introduced a bug. The pid file path set by -oP was being
4174 ignored. Though the use of -oP was forcing the writing of a pid file, it
4175 was always written to the default place.
4177 3. If the message "no IP address found for host xxxx" is generated during
4178 incoming verification, it is now followed by identification of the incoming
4179 connection (so you can more easily find what provoked it).
4181 4. Bug fix for Sieve filters: "stop" inside a block was not working properly.
4183 5. Added some features to "harden" Exim a bit more against certain attacks:
4185 (a) There is now a build-time option called FIXED_NEVER_USERS that can
4186 be put in Local/Makefile. This is like the never_users runtime option,
4187 but it cannot be overridden. The default setting is "root".
4189 (b) If ALT_CONFIG_PREFIX is defined in Local/Makefile, it specifies a
4190 prefix string with which any file named in a -C command line option
4193 (c) If ALT_CONFIG_ROOT_ONLY is defined in Local/Makefile, root privilege
4194 is retained for -C and -D only if the caller of Exim is root. Without
4195 it, the exim user may also use -C and -D and retain privilege.
4197 (d) If DISABLE_D_OPTION is defined in Local/Makefile, the use of the -D
4198 command line option is disabled.
4200 6. Macro names set by the -D option must start with an upper case letter, just
4201 like macro names defined in the configuration file.
4203 7. Added "dereference=" facility to LDAP.
4205 8. Two instances of the typo "uknown" in the source files are fixed.
4207 9. If a PERL_COMMAND setting in Local/Makefile was not at the start of a line,
4208 the Configure-Makefile script screwed up while processing it.
4210 10. Incorporated PCRE 4.4.
4212 11. The SMTP synchronization check was not operating right at the start of an
4213 SMTP session. For example, it could not catch a HELO sent before the client
4214 waited for the greeting. There is now a check for outstanding input at the
4215 point when the greeting is written. Because of the duplex, asynchronous
4216 nature of TCP/IP, it cannot be perfect - the incorrect input may be on its
4217 way, but not yet received, when the check is performed.
4219 12. Added tcp_nodelay to make it possible to turn of the setting of TCP_NODELAY
4220 on TCP/IP sockets, because this apparently causes some broken clients to
4223 13. Installed revised OS/Makefile-CYGWIN and OS/os.c-cygwin (the .h file was
4224 unchanged) from the Cygwin maintainer.
4226 14. The code for -bV that shows what is in the binary showed "mbx" when maildir
4227 was supported instead of testing for mbx. Effectively a typo.
4229 15. The spa authenticator server code was not checking that the input it
4230 received was valid base64.
4232 16. The debug output line for the "set" modifier in ACLs was not showing the
4233 name of the variable that was being set.
4235 17. Code tidy: the variable type "vtype_string" was never used. Removed it.
4237 18. Previously, a reference to $sender_host_name did not cause a DNS reverse
4238 lookup on its own. Something else was needed to trigger the lookup. For
4239 example, a match in host_lookup or the need for a host name in a host list.
4240 Now, if $sender_host_name is referenced and the host name has not yet been
4241 looked up, a lookup is performed. If the lookup fails, the variable remains
4242 empty, and $host_lookup_failed is set to "1".
4244 19. Added "eqi" as a case-independent comparison operator.
4246 20. The saslauthd authentication condition could segfault if neither service
4247 nor realm was specified.
4249 21. If an overflowing value such as "2048M" was set for message_size_limit, the
4250 error message that was logged was misleading, and incoming SMTP
4251 connections were dropped. The message is now more accurate, and temporary
4252 errors are given to SMTP connections.
4254 22. In some error situations (such as 21 above) Exim rejects all SMTP commands
4255 (except RSET) with a 421 error, until QUIT is received. However, it was
4256 failing to send a response to QUIT.
4258 23. The HELO ACL was being run before the code for helo_try_verify_hosts,
4259 which made it impossible to use "verify = helo" in the HELO ACL. The HELO
4260 ACL is now run after the helo_try_verify_hosts code.
4262 24. "{MD5}" and "{SHA1}" are now recognized as equivalent to "{md5"} and
4263 "{sha1}" in the "crypteq" expansion condition (in fact the comparison is
4264 case-independent, so other case variants are also recognized). Apparently
4265 some systems use these upper case variants.
4267 25. If more than two messages were waiting for the same host, and a transport
4268 filter was specified for the transport, Exim sent two messages over the
4269 same TCP/IP connection, and then failed with "socket operation on non-
4270 socket" when it tried to send the third.
4272 26. Added Exim::debug_write and Exim::log_write for embedded Perl use.
4274 27. The extern definition of crypt16() in expand.c was not being excluded when
4275 the OS had its own crypt16() function.
4277 28. Added bounce_return_body as a new option, and bounce_return_size_limit
4278 as a preferred synonym for return_size_limit, both as an option and as an
4281 29. Added LIBS=-liconv to OS/Makefile-OSF1.
4283 30. Changed the default configuration ACL to relax the local part checking rule
4284 for addresses that are not in any local domains. For these addresses,
4285 slashes and pipe symbols are allowed within local parts, but the sequence
4286 /../ is explicitly forbidden.
4288 31. SPA server authentication was not clearing the challenge buffer before
4291 32. log_message in a "warn" ACL statement was writing to the reject log as
4292 well as to the main log, which contradicts the documentation and doesn't
4293 seem right (because no rejection is happening). So I have stopped it.
4295 33. Added Ard Biesheuvel's lookup code for accessing an Interbase database.
4296 However, I am unable to do any testing of this.
4298 34. Fixed an infelicity in the appendfile transport. When checking directories
4299 for a mailbox, to see if any needed to be created, it was accidentally
4300 using path names with one or more superfluous leading slashes; tracing
4301 would show up entries such as stat("///home/ph10", 0xFFBEEA48).
4303 35. If log_message is set on a "discard" verb in a MAIL or RCPT ACL, its
4304 contents are added to the log line that is written for every discarded
4305 recipient. (Previously a log_message setting was ignored.)
4307 36. The ${quote: operator now quotes the string if it is empty.
4309 37. The install script runs exim in order to find its version number. If for
4310 some reason other than non-existence or emptiness, which it checks, it
4311 could not run './exim', it was installing it with an empty version number,
4312 i.e. as "exim-". This error state is now caught, and the installation is
4315 38. An argument was missing from the function that creates an error message
4316 when Exim fails to connect to the socket for saslauthd authentication.
4317 This could cause Exim to crash, or give a corrupted message.
4319 39. Added isip, isip4, and isip6 to ${if conditions.
4321 40. The ACL variables $acl_xx are now saved with the message, and can be
4322 accessed later in routers, transports, and filters.
4324 41. The new lookup type nwildlsearch is like wildlsearch, except that the key
4325 strings in the file are not string-expanded.
4327 42. If a MAIL command specified a SIZE value that was too large to fit into an
4328 int variable, the check against message_size_limit failed. Such values are
4329 now forced to INT_MAX, which is around 2Gb for a 32-bit variable. Maybe one
4330 day this will have to be increased, but I don't think I want to be around
4331 when emails are that large.
4338 1. Removed HAVE_ICONV=yes from OS/Makefile-FreeBSD, since it seems that
4339 iconv() is not standard in FreeBSD.
4341 2. Change 4.21/17 was buggy and could cause stack overwriting on a system with
4342 IPv6 enabled. The observed symptom was a segmentation fault on return from
4343 the function os_common_find_running_interfaces() in src/os.c.
4345 3. In the check_special_case() function in daemon.c I had used "errno" as an
4346 argument name, which causes warnings on some systems. This was basically a
4347 typo, since it was named "eno" in the comments!
4349 4. The code that waits for the clock to tick (at a resolution of some fraction
4350 of a second) so as to ensure message-id uniqueness was always waiting for
4351 at least one whole tick, when it could have waited for less. [This is
4352 almost certainly not relevant at current processor speeds, where it is
4353 unlikely to ever wait at all. But we try to future-proof.]
4355 5. The function that sleeps for a time interval that includes fractions of a
4356 second contained a race. It did not block SIGALRM between setting the
4357 timer, and suspending (a couple of lines later). If the interval was short
4358 and the sigsuspend() was delayed until after it had expired, the suspension
4359 never ended. On busy systems this could lead to processes getting stuck for
4362 6. Some uncommon configurations may cause a lookup to happen in a queue runner
4363 process, before it forks any delivery processes. The open lookup caching
4364 mechanism meant that the open file or database connection was passed into
4365 the delivery process. The problem was that delivery processes always tidy
4366 up cached lookup data. This could cause a problem for the next delivery
4367 process started by the queue runner, because the external queue runner
4368 process does not know about the closure. So the next delivery process
4369 still has data in the lookup cache. In the case of a file lookup, there was
4370 no problem because closing a file descriptor in a subprocess doesn't affect
4371 the parent. However, if the lookup was caching a connection to a database,
4372 the connection was closed, and the second delivery process was likely to
4373 see errors such as "PGSQL: query failed: server closed the connection
4374 unexpectedly". The problem has been fixed by closing all cached lookups
4375 in a queue runner before running a delivery process.
4377 7. Compiler warning on Linux for the second argument of iconv(), which doesn't
4378 seem to have the "const" qualifier which it has on other OS. I've
4381 8. Change 4.21/2 was too strict. It is only if there are two authenticators
4382 *of the same type* (client or server) with the same public name that an
4383 error should be diagnosed.
4385 9. When Exim looked up a host name for an IP address, but failed to find the
4386 original IP address when looking up the host name (a safety check), it
4387 output the message "<ip address> does not match any IP for NULL", which was
4388 confusing, to say the least. The bug was that the host name should have
4389 appeared instead of "NULL".
4391 10. Since release 3.03, if Exim is called by a uid other than root or the Exim
4392 user that is built into the binary, and the -C or -D options is used, root
4393 privilege is dropped before the configuration file is read. In addition,
4394 logging is switched to stderr instead of the normal log files. If the
4395 configuration then re-defines the Exim user, the unprivileged environment
4396 is probably not what is expected, so Exim logs a panic warning message (but
4399 However, if deliver_drop_privilege is set, the unprivileged state may well
4400 be exactly what is intended, so the warning has been cut out in that case,
4401 and Exim is allowed to try to write to its normal log files.
4407 1. smtp_return_error_details was not giving details for temporary sender
4408 or receiver verification errors.
4410 2. Diagnose a configuration error if two authenticators have the same public
4413 3. Exim used not to create the message log file for a message until the first
4414 delivery attempt. This could be confusing when incoming messages were held
4415 for policy or load reasons. The message log file is now created at the time
4416 the message is received, and an initial "Received" line is written to it.
4418 4. The automatically generated man page for command line options had a minor
4419 bug that caused no ill effects; however, a more serious problem was that
4420 the procedure for building the man page automatically didn't always
4421 operate. Consequently, release 4.20 contains an out-of-date version. This
4422 shouldn't happen again.
4424 5. When building Exim with embedded Perl support, the script that builds the
4425 Makefile was calling 'perl' to find its compile-time parameters, ignoring
4426 any setting of PERL_COMMAND in Local/Makefile. This is now fixed.
4428 6. The freeze_tell option was not being used for messages that were frozen on
4429 arrival, either by an ACL or by local_scan().
4431 7. Added the smtp_incomplete_transaction log selector.
4433 8. After STARTTLS, Exim was not forgetting that it had advertised AUTH, so it
4434 was accepting AUTH without a new EHLO.
4436 9. Added tls_remember_esmtp to cope with YAEB. This allows AUTH and other
4437 ESMTP extensions after STARTTLS without a new EHLO, in contravention of the
4440 10. Logging of TCP/IP connections (when configured) now happens in the main
4441 daemon process instead of the child process, so that the TCP/IP connection
4442 count is more accurate (but it can never be perfect).
4444 11. The use of "drop" in a nested ACL was not being handled correctly in the
4445 outer ACL. Now, if condition failure induced by the nested "drop" causes
4446 the outer ACL verb to deny access ("accept" or "discard" after "endpass",
4447 or "require"), the connection is dropped.
4449 12. Similarly, "discard" in a nested ACL wasn't being handled. A nested ACL
4450 that yield "discard" can now be used with an "accept" or a "discard" verb,
4451 but an error is generated for any others (because I can't see a useful way
4452 to define what should happen).
4454 13. When an ACL is read dynamically from a file (or anywhere else), the lines
4455 are now processed in the same way as lines in the Exim configuration file.
4456 In particular, continuation lines are supported.
4458 14. Added the "dnslists = a.b.c!=n.n.n.n" feature.
4460 15. Added -ti meaning -t -i.
4462 16. Check for letters, digits, hyphens, and dots in the names of dnslist
4463 domains, and warn by logging if others are found.
4465 17. At least on BSD, alignment is not guarenteed for the array of ifreq's
4466 returned from GIFCONF when Exim is trying to find the list of interfaces on
4467 a host. The code in os.c has been modified to copy each ifreq to an aligned
4468 structure in all cases.
4470 Also, in some cases, the returned ifreq's were being copied to a 'struct
4471 ifreq' on the stack, which was subsequently passed to host_ntoa(). That
4472 means the last couple of bytes of an IPv6 address could be chopped if the
4473 ifreq contained only a normal sockaddr (14 bytes storage).
4475 18. Named domain lists were not supported in the hosts_treat_as_local option.
4476 An entry such as +xxxx was not recognized, and was treated as a literal
4479 19. Ensure that header lines added by a DATA ACL are included in the reject log
4480 if the ACL subsequently rejects the message.
4482 20. Upgrade the cramtest.pl utility script to use Digest::MD5 instead of just
4483 MD5 (which is deprecated).
4485 21. When testing a filter file using -bf, Exim was writing a message when it
4486 took the sender from a "From " line in the message, but it was not doing so
4487 when it took $return_path from a Return-Path: header line. It now does.
4489 22. If the contents of a "message" modifier for a "warn" ACL verb do not begin
4490 with a valid header line field name (a series of printing characters
4491 terminated by a colon, Exim now inserts X-ACL-Warn: at the beginning.
4493 23. Changed "disc" in the source to "disk" to conform to the documentation and
4494 the book and for uniformity.
4496 24. Ignore Sendmail's -Ooption=value command line item.
4498 25. When execve() failed while trying to run a command in a pipe transport,
4499 Exim was returning EX_UNAVAILBLE (69) from the subprocess. However, this
4500 could be confused with a return value of 69 from the command itself. This
4501 has been changed to 127, the value the shell returns if it is asked to run
4502 a non-existent command. The wording for the related log line suggests a
4503 non-existent command as the problem.
4505 26. If received_header_text expands to an empty string, do not add a Received:
4506 header line to the message. (Well, it adds a token one on the spool, but
4507 marks it "old" so that it doesn't get used or transmitted.)
4509 27. Installed eximstats 1.28 (addition of -nt option).
4511 28. There was no check for failure on the call to getsockname() in the daemon
4512 code. This can fail if there is a shortage of resources on the system, with
4513 ENOMEM, for example. A temporary error is now given on failure.
4515 29. Contrary to the C standard, it seems that in some environments, the
4516 equivalent of setlocale(LC_ALL, "C") is not obeyed at the start of a C
4517 program. Exim now does this explicitly; it affects the formatting of
4518 timestamps using strftime().
4520 30. If exiqsumm was given junk data, it threw up some uninitialized variable
4521 complaints. I've now initialized all the variables, to avoid this.
4523 32. Header lines added by a system filter were not being "seen" during
4524 transport-time rewrites.
4526 33. The info_callback() function passed to OpenSSL is set up with type void
4527 (*)(SSL *, int, int), as described somewhere. However, when calling the
4528 function (actually a macro) that sets it up, the type void(*)() is
4529 expected. I've put in a cast to prevent warnings from picky compilers.
4531 34. If a DNS black list lookup found a CNAME record, but there were no A
4532 records associated with the domain it pointed at, Exim crashed.
4534 35. If a DNS black list lookup returned more than one A record, Exim ignored
4535 all but the first. It now scans all returned addresses if a particular IP
4536 value is being sought. In this situation, the contents of the
4537 $dnslist_value variable are a list of all the addresses, separated by a
4540 36. Tightened up the rules for host name lookups using reverse DNS. Exim used
4541 to accept a host name and all its aliases if the forward lookup for any of
4542 them yielded the IP address of the incoming connection. Now it accepts only
4543 those names whose forward lookup yields the correct IP address. Any other
4544 names are discarded. This closes a loophole whereby a rogue DNS
4545 administrator could create reverse DNS records to break through a
4546 wildcarded host restriction in an ACL.
4548 37. If a user filter or a system filter that ran in a subprocess used any of
4549 the numerical variables ($1, $2 etc), or $thisaddress, in a pipe command,
4550 the wrong values were passed to the pipe command ($thisaddress had the
4551 value of $0, $0 had the value of $1, etc). This bug was introduced by
4552 change 4.11/101, and not discovered because I wrote an inadequate test. :-(
4554 38. Improved the line breaking for long SMTP error messages from ACLs.
4555 Previously, if there was no break point between 40 and 75 characters, Exim
4556 left the rest of the message alone. Two changes have been made: (a) I've
4557 reduced the minimum length to 35 characters; (b) if it can't find a break
4558 point between 35 and 75 characters, it looks ahead and uses the first one
4559 that it finds. This may give the occasional overlong line, but at least the
4560 remaining text gets split now.
4562 39. Change 82 of 4.11 was unimaginative. It assumed the limit on the number of
4563 file descriptors might be low, and that setting 1000 would always raise it.
4564 It turns out that in some environments, the limit is already over 1000 and
4565 that lowering it causes trouble. So now Exim takes care not to decrease it.
4567 40. When delivering a message, the value of $return_path is set to $sender_
4568 address at the start of routing (routers may change the value). By an
4569 oversight, this default was not being set up when an address was tested by
4570 -bt or -bv, which affected the outcome if any router or filter referred to
4573 41. The idea of the "warn" ACL verb is that it adds a header or writes to the
4574 log only when "message" or "log_message" are set. However, if one of the
4575 conditions was an address verification, or a call to a nested ACL, the
4576 messages generated by the underlying test were being passed through. This
4577 no longer happens. The underlying message is available in $acl_verify_
4578 message for both "message" and "log_message" expansions, so it can be
4579 passed through if needed.
4581 42. Added RFC 2047 interpretation of header lines for $h_ expansions, with a
4582 new expansion $bh_ to give the encoded byte string without charset
4583 translation. Translation happens only if iconv() is available; HAVE_ICONV
4584 indicates this at build time. HEADERS_CHARSET gives the charset to
4585 translate to; headers_charset can change it in the configuration, and
4586 "headers charset" can change it in an individual filter file.
4588 43. Now that we have a default RFC 2047 charset (see above), the code in Exim
4589 that creates RFC 2047 encoded "words" labels them as that charset instead
4590 of always using iso-8859-1. The cases are (i) the explicit ${rfc2047:
4591 expansion operator; (ii) when Exim creates a From: line for a local
4592 message; (iii) when a header line is rewritten to include a "phrase" part.
4594 44. Nasty bug in exiqsumm: the regex to skip already-delivered addresses was
4595 buggy, causing it to skip the first lines of messages whose message ID
4596 ended in 'D'. This would not have bitten before Exim release 4.14, because
4597 message IDs were unlikely to end in 'D' before then. The effect was to have
4598 incorrect size information for certain domains.
4600 45. #include "config.h" was missing at the start of the crypt16.c module. This
4601 caused trouble on Tru64 (aka OSF1) systems, because HAVE_CRYPT16 was not
4604 46. If there was a timeout during a "random" callout check, Exim treated it as
4605 a failure of the random address, and carried on sending RSET and the real
4606 address. If the delay was just some slowness somewhere, the response to the
4607 original RCPT would be taken as a response to RSET and so on, causing
4608 mayhem of various kinds.
4610 47. Change 50 for 4.20 was a heap of junk. I don't know what I was thinking
4611 when I implemented it. It didn't allow for the fact that some option values
4612 may legitimatetly be negative (e.g. size_addition), and it didn't even do
4613 the right test for positive values.
4615 48. Domain names in DNS records are case-independent. Exim always looks them up
4616 in lower case. Some resolvers return domain names in exactly the case they
4617 appear in the zone file, that is, they may contain uppercase letters. Not
4618 all resolvers do this - some return always lower case. Exim was treating a
4619 change of case by a resolver as a change of domain, similar to a widening
4620 of a domain abbreviation. This triggered its re-routing code and so it was
4621 trying to route what was effectively the same domain again. This normally
4622 caused routing to fail (because the router wouldn't handle the domain
4623 twice). Now Exim checks for this case specially, and just changes the
4624 casing of the domain that it ultimately uses when it transmits the message
4627 49. Added Sieve (RFC 3028) support, courtesy of Michael Haardt's contributed
4630 50. If a filter generated a file delivery with a non-absolute name (possible if
4631 no home directory exists for the router), the forbid_file option was not
4634 51. Added '&' feature to dnslists, to provide bit mask matching in addition to
4635 the existing equality matching.
4637 52. Exim was using ints instead of ino_t variables in some places where it was
4638 dealing with inode numbers.
4640 53. If TMPDIR is defined in Local/Makefile (default in src/EDITME is
4641 TMPDIR="/tmp"), Exim checks for the presence of an environment variable
4642 called TMPDIR, and if it finds it is different, it changes its value.
4644 54. The smtp_printf() function is now made available to local_scan() so
4645 additional output lines can be written before returning. There is also an
4646 smtp_fflush() function to enable the detection of a dropped connection.
4647 The variables smtp_input and smtp_batched_input are exported to
4650 55. Changed the default runtime configuration: the message "Unknown user"
4651 has been removed from the ACL, and instead placed on the localuser router,
4652 using the cannot_route_message feature. This means that any verification
4653 failures that generate their own messages won't get overridden. Similarly,
4654 the "Unrouteable address" message that was in the ACL for unverifiable
4655 relay addresses has also been removed.
4657 56. Added hosts_avoid_esmtp to the smtp transport.
4659 57. The exicyclog script was not checking for the esoteric option
4660 CONFIGURE_FILE_USE_EUID in the Local/Makefile. It now does this, but it
4661 will work only if exicyclog is run under the appropriate euid.
4663 58. Following a discussion on the list, the rules by which Exim recognises line
4664 endings on incoming messages have been changed. The -dropcr and drop_cr
4665 options are now no-ops, retained only for backwards compatibility. The
4666 following line terminators are recognized: LF CRLF CR. However, special
4667 processing applies to CR:
4669 (i) The sequence CR . CR does *not* terminate an incoming SMTP message,
4670 nor a local message in the state where . is a terminator.
4672 (ii) If a bare CR is encountered in a header line, an extra space is added
4673 after the line terminator so as not to end the header. The reasoning
4674 behind this is that bare CRs in header lines are most likely either
4675 to be mistakes, or people trying to play silly games.
4677 59. The size of a message, as listed by "-bp" or in the Exim monitor window,
4678 was being incorrectly given as 18 bytes larger than it should have been.
4679 This is a VOB (very old bug).
4681 60. This may never have affected anything current, but just in case it has:
4682 When the local host is found other than at the start of a list of hosts,
4683 the local host, those with the same MX, and any that follow, are discarded.
4684 When the list in question was part of a longer list of hosts, the following
4685 hosts (not currently being processed) were also being discarded. This no
4686 longer happens. I'm not sure if this situation could ever has previously
4689 61. Added the "/MX" feature to lists of hosts in the manualroute and query
4692 62. Whenever Exim generates a new message, it now adds an Auto-Submitted:
4693 header. This is something that is recommended in a new Internet Draft, and
4694 is something that is documented as being done by Sendmail. There are two
4695 possible values. For messages generated by the autoreply transport, Exim
4698 Auto-Submitted: auto-replied
4700 whereas for all other generated messages (e.g. bounces) it adds
4702 Auto-Submitted: auto-generated
4704 63. The "personal" condition in filters now includes a test for the
4705 Auto-Submitted: header. If it contains the string "auto-" the message it
4706 not considered personal.
4708 64. Added rcpt_include_affixes as a generic transport option.
4710 65. Added queue_only_override (default true).
4712 66. Added the syslog_duplication option.
4714 67. If what should have been the first header line of a message consisted of
4715 a space followed by a colon, Exim was mis-interpreting it as a header line.
4716 It isn't of course - it is syntactically invalid and should therefore be
4717 treated as the start of the message body. The misbehaviour could have
4718 caused a number of strange effects, including loss of data in subsequent
4719 header lines, and spool format errors.
4721 68. Formerly, the AUTH parameter on a MAIL command was trusted only if the
4722 client host had authenticated. This control can now be exercised by an ACL
4723 for more flexibility.
4725 69. By default, callouts do not happen when testing with -bh. There is now a
4726 variant, -bhc, which does actually run the callout code, including
4727 consulting and updating the callout cache.
4729 70. Added support for saslauthd authentication, courtesy of Alexander
4732 71. If statvfs() failed on the spool or log directories while checking their
4733 size for availability, Exim confusingly gave the error "space shortage".
4734 Furthermore, in debugging mode it crashed with a floating point exception.
4735 These checks are done if check_{spool,log}_{space,inodes} are set, and when
4736 an SMTP message arrives with SIZE= on the MAIL command. As this is a really
4737 serious problem, Exim now writes to the main and panic logs when this
4738 happens, with details of the failure. It then refuses to accept the
4739 incoming message, giving the message "spool directory problem" or "log
4740 directory problem" with a 421 code for SMTP messages.
4742 72. When Exim is about to re-exec itself, it ensures that the file descriptors
4743 0, 1, and 2 exist, because some OS complain for execs without them (see
4744 ChangeLog 4.05/30). If necessary, Exim opens /dev/null to use for these
4745 descriptors. However, the code omitted to check that the open succeeded,
4746 causing mysterious errors if for some reason the permissions on /dev/null
4747 got screwed. Now Exim writes a message to the main and panic logs, and
4748 bombs out if it can't open /dev/null.
4750 73. Re-vamped the way daemon_smtp_port, local_interfaces, and -oX work and
4751 interact so that it is all more flexible. It is supposed to remain
4752 backwards compatible. Also added extra_local_interfaces.
4754 74. Invalid data sent to a SPA (NTLM) server authenticator could cause the code
4755 to bomb out with an assertion failure - to the client this appears as a
4756 connection drop. This problem occurs in the part of the code that was taken
4757 from the Samba project. Fortunately, the assertion is in a very simple
4758 function, so I have fixed this by reproducing the function inline in the
4759 one place where it is called, and arranging for authentication to fail
4760 instead of killing the process with assert().
4762 75. The SPA client code was not working when the server requested OEM rather
4763 than Unicode encoding.
4765 76. Added code to make require_files with a specific uid setting more usable in
4766 the case where statting the file as root fails - usually a non-root-mounted
4767 NFS file system. When this happens and the failure is EACCES, Exim now
4768 forks a subprocess and does the per-uid checking as the relevant uid.
4770 77. Added process_log_path.
4772 78. If log_file_path was not explicitly set, a setting of check_log_space or
4773 check_log_inodes was ignored.
4775 79. If a space check for the spool or log partitions fails, the incident is now
4776 logged. Of course, in the latter case the data may get lost...
4778 80. Added the %p formatting code to string_format() so that it can be used to
4779 print addresses in debug_print(). Adjusted all the address printing in the
4780 debugging in store.c to use %p rather than %d.
4782 81. There was a concern that a line of code in smtp_in.c could overflow a
4783 buffer if a HELO/EHLO command was given followed by 500 or so spaces. As
4784 initially expressed, the concern was not well-founded, because trailing
4785 spaces are removed early. However, if the trailing spaces were followed by
4786 a NULL, they did not get removed, so the overflow was possible. Two fixes
4789 (a) I re-wrote the offending code in a cleaner fashion.
4790 (b) If an incoming SMTP command contains a NULL character, it is rejected
4793 82. When Exim changes uid/gid to the Exim user at daemon start time, it now
4794 runs initgroups(), so that if the Exim user is in any additional groups,
4795 they will be used during message reception.
4801 The change log for 4.20 and earlier releases has been archived.
git://git.exim.org / exim.git / blob